The Frontiers of the Ottoman World 0197264425, 9780197264423

This is the first major comparative study of the frontiers of the Ottoman Empire, one of the crucial forces that shaped

415 68 9MB

English Pages 618 [544] Year 2009

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
1
Title Pages
A.C.S. Peacock
Title Pages
(p.i) The Frontiers of the Ottoman World (p.ii) (p.iii) The Frontiers of the Ottoman World
Title Pages
2
(p.viii) Figures and Tables
A.C.S. Peacock
(p.viii) Figures and Tables
(p.viii) Figures and Tables
(p.viii) Figures and Tables
(p.viii) Figures and Tables
3
(p.xii) (p.xiii) Notes on Contributors
A.C.S. Peacock
(p.xii) (p.xiii) Notes on Contributors
(p.xii) (p.xiii) Notes on Contributors
(p.xii) (p.xiii) Notes on Contributors
(p.xii) (p.xiii) Notes on Contributors
(p.xii) (p.xiii) Notes on Contributors
(p.xii) (p.xiii) Notes on Contributors
4
(p.xxi) Acknowledgements
A.C.S. Peacock
(p.xxi) Acknowledgements
5
(p.xxii) A Note on Transliteration, Place Names and the Organisation of this Volume
A.C.S. Peacock
(p.xxii) A Note on Transliteration, Place Names and the Organisation of this Volume
6
(p.xxiii) Abbreviations
A.C.S. Peacock
(p.xxiii) Abbreviations
7
Introduction: The Ottoman Empire and its Frontiers
A.C.S. Peacock
Introduction: The Ottoman Empire and its Frontiers
A. C. S. PEACOCK
Abstract and Keywords
Introduction: The Ottoman Empire and its Frontiers
Introduction: The Ottoman Empire and its Frontiers
Ottoman Frontiers: A Historical Overview
Introduction: The Ottoman Empire and its Frontiers
Introduction: The Ottoman Empire and its Frontiers
Introduction: The Ottoman Empire and its Frontiers
Introduction: The Ottoman Empire and its Frontiers
Introduction: The Ottoman Empire and its Frontiers
The Ideology of Ottoman Expansion from Beylik to Empire
Introduction: The Ottoman Empire and its Frontiers
Introduction: The Ottoman Empire and its Frontiers
Introduction: The Ottoman Empire and its Frontiers
Frontiers, Frontiersmen and the Centre
Introduction: The Ottoman Empire and its Frontiers
Introduction: The Ottoman Empire and its Frontiers
(p.19) Frontier Fortifications
Introduction: The Ottoman Empire and its Frontiers
The Administration of the Frontier
Introduction: The Ottoman Empire and its Frontiers
Frontier Society: Rulers, Ruled and Revolt
Introduction: The Ottoman Empire and its Frontiers
Introduction: The Ottoman Empire and its Frontiers
The Economy of the Frontier
Introduction: The Ottoman Empire and its Frontiers
Introduction: The Ottoman Empire and its Frontiers
Notes:
Introduction: The Ottoman Empire and its Frontiers
Introduction: The Ottoman Empire and its Frontiers
Introduction: The Ottoman Empire and its Frontiers
Introduction: The Ottoman Empire and its Frontiers
8
The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries
A.C.S. Peacock
The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries
PALMIRA BRUMMETT
Abstract and Keywords
The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries
Imagining Empire and its Limits
Sites of Confrontation
The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries
The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries
The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries
The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries
The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries
The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries
The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries
Ottoman Mappings
The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries
The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries
Notes:
The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries
The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries
The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries
The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries
The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries
9
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
A.C.S. Peacock
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
GÁBOR ÁGOSTON
Abstract and Keywords
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
River Systems, Terrain and Fortifications
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
(p.72) Landscape, Climate and Frontier33
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
Conclusion
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
Notes:
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary
10
The Ottoman Conquest of Arabia and the Syrian Hajj Route
A.C.S. Peacock
The Ottoman Conquest of Arabia and the Syrian Hajj Route
ANDREW PETERSEN
Abstract and Keywords
Ottoman Rule in Arabia
Period One: Conquest, Confrontation and Defeat
The Ottoman Conquest of Arabia and the Syrian Hajj Route
The Ottoman Conquest of Arabia and the Syrian Hajj Route
Period Two: The Rise of the Wahhabis
The Ottoman Conquest of Arabia and the Syrian Hajj Route
The Syrian Hajj Route
The Ottoman Conquest of Arabia and the Syrian Hajj Route
Phase One: Sixteenth Century
The Ottoman Conquest of Arabia and the Syrian Hajj Route
The Ottoman Conquest of Arabia and the Syrian Hajj Route
(p.90) Phase Two: Eighteenth Century
The Ottoman Conquest of Arabia and the Syrian Hajj Route
The Ottoman Conquest of Arabia and the Syrian Hajj Route
Conclusion
The Ottoman Conquest of Arabia and the Syrian Hajj Route
Notes:
The Ottoman Conquest of Arabia and the Syrian Hajj Route
The Ottoman Conquest of Arabia and the Syrian Hajj Route
11
‘Aqaba Castle in the Ottoman Period, 1517–1917
A.C.S. Peacock
‘Aqaba Castle in the Ottoman Period, 1517–1917
DENYS PRINGLE
Abstract and Keywords
‘Aqaba Castle in the Ottoman Period, 1517–1917
‘Aqaba Castle in the Ottoman Period, 1517–1917
‘Aqaba Castle in the Ottoman Period, 1517–1917
‘Aqaba Castle in the Ottoman Period, 1517–1917
‘Aqaba Castle in the Ottoman Period, 1517–1917
‘Aqaba Castle in the Ottoman Period, 1517–1917
‘Aqaba Castle in the Ottoman Period, 1517–1917
‘Aqaba Castle in the Ottoman Period, 1517–1917
‘Aqaba Castle in the Ottoman Period, 1517–1917
Notes:
‘Aqaba Castle in the Ottoman Period, 1517–1917
‘Aqaba Castle in the Ottoman Period, 1517–1917
‘Aqaba Castle in the Ottoman Period, 1517–1917
12
Evliya Çelebi, the Mani and the Fortress of Kelefa
A.C.S. Peacock
Evliya Çelebi, the Mani and the Fortress of Kelefa
MALCOLM WAGSTAFF
Abstract and Keywords
Evliya Çelebi, the Mani and the Fortress of Kelefa
(p.114) The Mani
Evliya Çelebi, the Mani and the Fortress of Kelefa
Evliya Çelebi, the Mani and the Fortress of Kelefa
The Chronicle of the Morea
Evliya Çelebi, the Mani and the Fortress of Kelefa
Evliya Çelebi, the Mani and the Fortress of Kelefa
Western Literary and Documentary Evidence
Evliya Çelebi, the Mani and the Fortress of Kelefa
The Seyahatname of Evliya Çelebi
Evliya Çelebi, the Mani and the Fortress of Kelefa
Evliya Çelebi, the Mani and the Fortress of Kelefa
Evliya Çelebi, the Mani and the Fortress of Kelefa
Other Evidence
Evliya Çelebi, the Mani and the Fortress of Kelefa
(p.132) Conclusion
Evliya Çelebi, the Mani and the Fortress of Kelefa
Evliya Çelebi, the Mani and the Fortress of Kelefa
Notes:
Evliya Çelebi, the Mani and the Fortress of Kelefa
Evliya Çelebi, the Mani and the Fortress of Kelefa
Evliya Çelebi, the Mani and the Fortress of Kelefa
Evliya Çelebi, the Mani and the Fortress of Kelefa
13
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
A.C.S. Peacock
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
VICTOR OSTAPCHUK
SVITLANA BILYAYEVA
Abstract and Keywords
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
(p.141) Akkerman in its Frontier Setting
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
The Akkerman Fortress Project
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
Akkerman in the Fifteenth to Seventeenth Centuries
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
(p.148) The Port Yard and Associated Structures
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
Documentary Evidence for Decision Making in Fortress Works
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
Akkerman in the Eighteenth and Nineteenth Centuries
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
(p.156) The Ottoman Response to Advances in Fortification Technology
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
(p.160) The Infrastructure of Maintenance and Construction
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
Archaeological Evidence for Frontier Life at Akkerman
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
Conclusion
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
Notes:
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project
14
Palanka Forts and Construction Activity in the Late Ottoman Balkans
A.C.S. Peacock
Palanka Forts and Construction Activity in the Late Ottoman Balkans
BURCU ÖZGÜVEN
Abstract and Keywords
Palanka Forts and Construction Activity in the Late Ottoman Balkans
(p.173) Geography
Palanka Forts and Construction Activity in the Late Ottoman Balkans
(p.175) Late Eighteenth-Century Historical Background
Palanka Forts and Construction Activity in the Late Ottoman Balkans
(p.178) Late Military Buildings in the Balkans
The Hapsburg Frontier
Palanka Forts and Construction Activity in the Late Ottoman Balkans
Palanka Forts and Construction Activity in the Late Ottoman Balkans
The Danube Frontier
The Black Sea Frontier
Palanka Forts and Construction Activity in the Late Ottoman Balkans
Late Civil Buildings in the Balkans
Palanka Forts and Construction Activity in the Late Ottoman Balkans
Conclusion
Palanka Forts and Construction Activity in the Late Ottoman Balkans
Palanka Forts and Construction Activity in the Late Ottoman Balkans
Notes:
Palanka Forts and Construction Activity in the Late Ottoman Balkans
Palanka Forts and Construction Activity in the Late Ottoman Balkans
Palanka Forts and Construction Activity in the Late Ottoman Balkans
Palanka Forts and Construction Activity in the Late Ottoman Balkans
Palanka Forts and Construction Activity in the Late Ottoman Balkans
15
Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey
A.C.S. Peacock
Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey
LUCIENNE THYS-ŞENOCAK
RAHMI NURHAN ÇELİK
ARZU ÖZSAVAŞÇI
GÜLSÜN TANYELİ
Abstract and Keywords
Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey
Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey
Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey
(p.194) The Survey and Restoration Projects at the Fortress of Seddülbahir
Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey
Representational Records
Engravings
Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey
Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey
Historical Photographs
Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey
Archival Documents
Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey
Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey
Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey
(p.207) Conclusion
Notes:
Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey
Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey
Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey
Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey
16
Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century
A.C.S. Peacock
Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century
TOM SINCLAIR
Abstract and Keywords
Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century
Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century
Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century
Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century
Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century
Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century
Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century
Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century
Notes:
Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century
Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century
Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century
Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century
17
Ottoman Frontier Policies in North-East Africa, 1517–1914
A.C.S. Peacock
Ottoman Frontier Policies in North-East Africa, 1517–1914
JOHN ALEXANDER
Abstract and Keywords
Ottoman Frontier Policies in North-East Africa, 1517–1914
(p.226) Period One: 1517–55
Ottoman Frontier Policies in North-East Africa, 1517–1914
Period Two: 1555–89
Period Three: 1589–1820
Ottoman Frontier Policies in North-East Africa, 1517–1914
Ottoman Frontier Policies in North-East Africa, 1517–1914
Period Four: 1820–1914
Ottoman Frontier Policies in North-East Africa, 1517–1914
(p.234) Conclusions
Ottoman Frontier Policies in North-East Africa, 1517–1914
Notes:
Ottoman Frontier Policies in North-East Africa, 1517–1914
Ottoman Frontier Policies in North-East Africa, 1517–1914
18
Continuities in Ottoman Centre–Periphery Relations, 1787–1915
A.C.S. Peacock
Continuities in Ottoman Centre–Periphery Relations, 1787–1915
FREDERICK ANSCOMBE
Abstract and Keywords
Continuities in Ottoman Centre–Periphery Relations, 1787–1915
Albania, 1787–1820
Continuities in Ottoman Centre–Periphery Relations, 1787–1915
Continuities in Ottoman Centre–Periphery Relations, 1787–1915
Continuities in Ottoman Centre–Periphery Relations, 1787–1915
Continuities in Ottoman Centre–Periphery Relations, 1787–1915
Continuities in Ottoman Centre–Periphery Relations, 1787–1915
Continuities in Ottoman Centre–Periphery Relations, 1787–1915
Eastern Arabia, 1870–1915
Continuities in Ottoman Centre–Periphery Relations, 1787–1915
Continuities in Ottoman Centre–Periphery Relations, 1787–1915
Continuities in Ottoman Centre–Periphery Relations, 1787–1915
Conclusion
Notes:
Continuities in Ottoman Centre–Periphery Relations, 1787–1915
Continuities in Ottoman Centre–Periphery Relations, 1787–1915
Continuities in Ottoman Centre–Periphery Relations, 1787–1915
Continuities in Ottoman Centre–Periphery Relations, 1787–1915
19
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
A.C.S. Peacock
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
KAHRAMAN ŞAKUL
Abstract and Keywords
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
How to Run the Adriatic Frontier: Diplomatic Problems
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
(p.262) How to Run a Frontier: Local Interests versus Imperial Priorities
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
Conclusion
Notes:
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars
20
Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century
A.C.S. Peacock
Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century
GÖKHAN ÇETİNSAYA
Abstract and Keywords
Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century
Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century
(p.275) Dilemmas of Reform in the Iraqi Provinces
Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century
Chronic Conflicts in Provincial Administration
Continuous Tribal Unrest in the Iraqi Provinces
Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century
The Iranian (Shi‘i) Challenge
Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century
Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century
Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century
The British Penetration of Iraq and the Gulf
Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century
Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century
Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century
Notes:
Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century
Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century
Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century
Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century
Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century
21
The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914
A.C.S. Peacock
The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914
ISA BLUMI
Abstract and Keywords
The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914
A Context for Modern Imperial Statecraft
The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914
The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914
The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914
The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914
A Contingent British Response
The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914
Harnessing Imperial Patronage
The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914
The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914
Territorialising Yemen
The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914
Conclusion
The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914
Notes:
The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914
The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914
The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914
The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914
22
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
A.C.S. Peacock
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
SARA NUR YILDIZ
Abstract and Keywords
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
(p.311) Geography of the Karamanid Frontier
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
Ottoman-Karamanid Relations in the Fifteenth Century and the Karamanid Succession Struggle, 1464–8
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
Ottoman Interference in Karamanid Succession Politics
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
Gevele and Konya
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
The Importance of the Hinterland
The Displacement of the Turgutoğlu from the Hinterland
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
Conclusion
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
Notes:
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468
23
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
A.C.S. Peacock
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
ROSSITSA GRADEVA
Abstract and Keywords
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
(p.333) Material Structures
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
Human Resources
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
(p.347) Administrative System and Agrarian Regime
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
Conclusion
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
Notes:
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries
24
The Garrison and its Hinterland in the Ottoman East, 1578–1605
A.C.S. Peacock
The Garrison and its Hinterland in the Ottoman East, 1578–1605
RHOADS MURPHEY
Abstract and Keywords
The Garrison and its Hinterland in the Ottoman East, 1578–1605
Front Formation and Maintaining Fortified Frontiers in the Balkans and North Africa
The Garrison and its Hinterland in the Ottoman East, 1578–1605
The Garrison and its Hinterland in the Ottoman East, 1578–1605
The Garrison and its Hinterland in the Ottoman East, 1578–1605
(p.358) Case One: Shirvan Province
The Garrison and its Hinterland in the Ottoman East, 1578–1605
Case Two: Kars
The Garrison and its Hinterland in the Ottoman East, 1578–1605
The Garrison and its Hinterland in the Ottoman East, 1578–1605
Case Three: Military-Civilian Conflict in Peacetime Erzurum as Judged by the Events of 1592–4
The Garrison and its Hinterland in the Ottoman East, 1578–1605
Case Four: An Account of the Change in Community Attitudes and Evolution of Military–Civilian Relations in Tabriz During Times of War (1585–90) and Times of Peace (1590–1603)
Phase One: Winning the War on the Azeri Front, 1585–90
The Garrison and its Hinterland in the Ottoman East, 1578–1605
The Garrison and its Hinterland in the Ottoman East, 1578–1605
Phase Two: Establishing Peace in Ottoman Azerbaijan, 1590–1603
The Garrison and its Hinterland in the Ottoman East, 1578–1605
Conclusion and General Findings
The Garrison and its Hinterland in the Ottoman East, 1578–1605
Notes:
The Garrison and its Hinterland in the Ottoman East, 1578–1605
The Garrison and its Hinterland in the Ottoman East, 1578–1605
25
Ottoman Archaeology of the Middle Nile Valley in the Sudan
A.C.S. Peacock
Ottoman Archaeology of the Middle Nile Valley in the Sudan
INTISAR ELZEIN
Abstract and Keywords
Ottoman Archaeology of the Middle Nile Valley in the Sudan
The Ottomans in the Middle Nile Valley
Ottoman Archaeology of the Middle Nile Valley in the Sudan
Berberistan: The Funj-Ottoman Frontier Region according to Archaeological Evidence and Evliya Çelebi
Ottoman Archaeology of the Middle Nile Valley in the Sudan
Qal‘at Sai
Ottoman Archaeology of the Middle Nile Valley in the Sudan
(p.378) Tinare
Jabal Sese
Ottoman Archaeology of the Middle Nile Valley in the Sudan
Kada Musa
Arduan Island
Simit Island
Ottoman Archaeology of the Middle Nile Valley in the Sudan
Hannek
Unidentified sites in Berberistan
Ottoman Archaeology of the Middle Nile Valley in the Sudan
Conclusion
Ottoman Archaeology of the Middle Nile Valley in the Sudan
Notes:
Ottoman Archaeology of the Middle Nile Valley in the Sudan
Ottoman Archaeology of the Middle Nile Valley in the Sudan
26
Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds
A.C.S. Peacock
Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds
IBOLYA GERELYES
Abstract and Keywords
Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds
New Ceramic Types in Ottoman Hungary
Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds
Classification of the New Ceramic Types
First Group
Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds
Second Group
Third Group
Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds
Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds
(p.395) Ottoman Castles and Ceramic Finds
Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds
Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds
Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds
Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds
(p.398) Conclusions
The Composition of the Garrison in Question
Ethnicity of the ‘Supply Hinterlands’
Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds
Ethnic Changes as Reflected in the Ceramic Corpus
Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds
Notes:
Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds
Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds
Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds
Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds
27
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
A.C.S. Peacock
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
RICHARD CARLTON
ALAN RUSHWORTH
Abstract and Keywords
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
Castles and Fortifications
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
(p.413) Ethnography
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
Rural Houses
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
(p.418) Funerary Stelae
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
Watermills
Pottery Making
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
Local Production Centres
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
Non-local Sources of Pottery
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
Conclusions
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
Notes:
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
The Krajina Project: Exploring the Ottoman–Hapsburg Borderland
28
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
A.C.S. Peacock
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
NEIL FAULKNER
NICHOLAS J. SAUNDERS
Abstract and Keywords
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
The Archaeology of Modern Conflict
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
The Great Arab Revolt Project
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
The Ottoman Empire Goes to War
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
Lawrence’s Conception of Guerrilla Warfare
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
The Archaeology of Imperial Collapse 1: A Trench Fortress at Ma‘an
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
The Archaeology of Imperial Collapse 2: A Fortified Railway Station at Wadi Rutm
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
The Archaeology of Imperial Collapse 3: The First Battle of Ma‘an
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
Conclusions
War without frontiers
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
War of Detachment
War of the Flea
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
Notes:
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
War without Frontiers: The Archaeology of the Arab Revolt, 1916–18
29
Military Service and Material Gain on the Ottoman–Hapsburg Frontier
A.C.S. Peacock
Military Service and Material Gain on the Ottoman–Hapsburg Frontier
MARK STEIN
Abstract and Keywords
Military Service and Material Gain on the Ottoman–Hapsburg Frontier
Military Service and Material Gain on the Ottoman–Hapsburg Frontier
Military Service and Material Gain on the Ottoman–Hapsburg Frontier
Military Service and Material Gain on the Ottoman–Hapsburg Frontier
Military Service and Material Gain on the Ottoman–Hapsburg Frontier
Military Service and Material Gain on the Ottoman–Hapsburg Frontier
Military Service and Material Gain on the Ottoman–Hapsburg Frontier
Military Service and Material Gain on the Ottoman–Hapsburg Frontier
Notes:
Military Service and Material Gain on the Ottoman–Hapsburg Frontier
Military Service and Material Gain on the Ottoman–Hapsburg Frontier
Military Service and Material Gain on the Ottoman–Hapsburg Frontier
Military Service and Material Gain on the Ottoman–Hapsburg Frontier
Military Service and Material Gain on the Ottoman–Hapsburg Frontier
30
Ottoman Suakin 1541–1865: Lost and Found
A.C.S. Peacock
Ottoman Suakin 1541–1865: Lost and Found
MICHAEL MALLINSON
LAURENCE SMITH
COLIN BREEN
WES FORSYTHE
JACKE PHILLIPS
Abstract and Keywords
Ottoman Suakin 1541–1865: Lost and Found
(p.471) Site History and Archaeology
Ottoman Suakin 1541–1865: Lost and Found
Ottoman Suakin 1541–1865: Lost and Found
The 2002–7 Project
Ottoman Suakin 1541–1865: Lost and Found
(p.476) Excavations of Main Ottoman Buildings
Bayt al-Mufti
Ottoman Suakin 1541–1865: Lost and Found
Bayt Khurshid Efendi
Ottoman Suakin 1541–1865: Lost and Found
(p.480) Bayt al-Basha
Ottoman Suakin 1541–1865: Lost and Found
Suakin Port and its Trade
The Suitability of the Harbour
Ottoman Suakin 1541–1865: Lost and Found
The Economy of the Port
Ottoman Suakin 1541–1865: Lost and Found
Ottoman Suakin 1541–1865: Lost and Found
Ottoman Suakin 1541–1865: Lost and Found
(p.486) The Extent of Trading Links of the Port
Ottoman Suakin 1541–1865: Lost and Found
Ottoman Suakin 1541–1865: Lost and Found
Conclusion
Ottoman Suakin 1541–1865: Lost and Found
Ottoman Suakin 1541–1865: Lost and Found
Notes:
Ottoman Suakin 1541–1865: Lost and Found
Ottoman Suakin 1541–1865: Lost and Found
Ottoman Suakin 1541–1865: Lost and Found
31
A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760
A.C.S. Peacock
A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760
COLIN HEYWOOD
Abstract and Keywords
I
A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760
A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760
II
A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760
III
A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760
A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760
IV
A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760
A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760
V
A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760
Notes:
A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760
A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760
A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760
A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760
A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760
A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760
32
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
A.C.S. Peacock
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
PAUL LANE
DOUGLAS JOHNSON
Abstract and Keywords
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
Ottoman Egypt
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
Slavery in Ottoman Egypt in the Nineteenth Century
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
Slave Trading in the Sudan
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
The Zariba System
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
Later History of Rumbek
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
Recent Archaeological Survey
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
Pendit
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
Meen Atol
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
Lol Nhom
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
Matching the Historical Sources with Archaeological Evidence
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
Conclusion
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
Notes:
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
The Archaeology and History of Slavery in South Sudan in the Nineteenth Century
33
Afterword
A.C.S. Peacock
Afterword
CAROLINE FINKEL
Abstract and Keywords
The Ottoman Frontier
Afterword
Afterword
Afterword
History and Archaeology—and Historical Archaeology
Afterword
Ottoman Archaeology
Afterword
Afterword
A Way Forward
Afterword
Afterword
Notes:
Afterword
Afterword
34
(p.551) Glossary
A.C.S. Peacock
(p.551) Glossary
(p.551) Glossary
(p.551) Glossary
(p.551) Glossary
35
(p.554) (p.555) Ottoman Sultans
A.C.S. Peacock
(p.554) (p.555) Ottoman Sultans
(p.554) (p.555) Ottoman Sultans
36
(p.556) (p.557) Select Bibliography
A.C.S. Peacock
(p.556) (p.557) Select Bibliography
General
(p.556) (p.557) Select Bibliography
(p.558) Anatolian Frontiers
(p.556) (p.557) Select Bibliography
Arabia, Iraq, the Gulf and Yemen
(p.556) (p.557) Select Bibliography
(p.556) (p.557) Select Bibliography
Central and Eastern Europe
(p.556) (p.557) Select Bibliography
(p.556) (p.557) Select Bibliography
(p.556) (p.557) Select Bibliography
(p.563) The Eastern Mediterranean Frontier and Greece
(p.556) (p.557) Select Bibliography
(p.556) (p.557) Select Bibliography
The Western Mediterranean Frontier and North Africa
(p.556) (p.557) Select Bibliography
(p.565) Sub-Saharan Africa and the Sudan
(p.556) (p.557) Select Bibliography
(p.556) (p.557) Select Bibliography
37
(p.567) Index
A.C.S. Peacock
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
(p.567) Index
Recommend Papers

The Frontiers of the Ottoman World
 0197264425, 9780197264423

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Title Pages

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Title Pages (p.i) The Frontiers of the Ottoman World (p.ii) (p.iii) The Frontiers of the Ottoman World PROCEEDINGS OF THE BRITISH ACADEMY · 156

(p.iv) Oxford University Press, Great Clarendon Street, Oxford OX2 6DP Oxford NewYork Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With offices in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam Published in the United States by Oxford University Press Inc., New York © The British Academy 2009 Database right The British Academy (maker) First published 2009 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of the British Academy, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction Page 1 of 2 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Title Pages outside the scope of the above should be sent to the Publications Department, The British Academy, 10 Carlton House Terrace, London SW1Y 5AH You must not circulate this book in any other binding or cover and you must impose this same condition on any acquirer British Library Cataloguing in Publication Data Data available Library of Congress Cataloging in Publication Data Data available Typeset by New Leaf Design, Scarborough, North Yorkshire Printed in Great Britain on acid-free paper by CPI Antony Rowe Chippenham, Wiltshire ISBN 978-0-19-726442-3 ISSN 0068-1202

Access brought to you by:

Page 2 of 2 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Figures and Tables

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

(p.viii) Figures and Tables Figures 1.1 Map of the expansion of the Ottoman Empire. 6 2.1 The defeat of the Ottomans at Hatvan. J. S., ‘Abris der Vöstung Hadtwan’, [1596]. 34 2.2 The conquest of Győr by the Ottomans. Alexander Mair, ‘Iavarinum sive Raab’, [1598]. 36 2.3 Map of the Fortress of Nayhaysel. Giovanni Battista Chiarello, ‘Pianta della Fortezza di Nayhaysel’. 38 2.4 Map of the fortress of Nicosia by Simon Pinargenti. 41 2.5 Map of the fortress of Nicosia by Giuseppe Rosaccio. 43 2.6 Map of Isola di Santa Maura by Vincenzo Coronelli. 45 2.7 Fortress detail from map of Isola di Santa Maura by Vincenzo Coronelli. 46 2.8 Title cartouche from map of Isola di Santa Maura by Vincenzo Coronelli. 47 2.9 Clissa, from ‘Molina Family Atlas’, mid-seventeenth century. 49 2.10 The submission of Prince John Sigismund Zapolya of Transylvania to Sultan Süleyman I. 52 2.11 The Ottoman–Safavid border. Matrakçı Nasuh, Menazil-i Sefer-i ‘Irakeyn. 54 3.1 Map of major Ottoman campaigns in Hungary in the sixteenth century. 59 3.2 Map showing fortifications and geography of sixteenth-century Hungary. 69 4.1 Map of Arabia in the sixteenth century showing the extent of the Ottoman Empire and the area of Portuguese influence. 83 4.2 South face of the sixteenth-century fort at Qatrana. 88 4.3 Ground plan of the sixteenth-century fort at Ma‘an. 89 4.4 South face of the recently restored sixteenth-century fort at Ma‘an with original entrance. 89 4.5 Map of the hajj route showing locations of forts in the eighteenth century. 91 4.6 First-floor plan of the eighteenth-century fort of Fassu‘a with projecting corner towers. 92 4.7 Qal‘at Fassu‘a: north-west corner tower. 92 4.8 Qal‘at Fassu‘a: north face showing projecting corner towers. 93 5.1 Map of ‘Aqaba and surrounding region. 96 5.2 Plan of ‘Aqaba Castle. 99 5.3 ‘Aqaba Castle: the gatehouse. 100 Page 1 of 4 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Figures and Tables 5.4 ‘Aqaba Castle: the gate within the gatehouse. 100 5.5 ‘Aqaba Castle: the gate and north range seen from inside the castle. 101 5.6 ‘Aqaba Castle: view from the north-east, by Léon de Laborde, 1828. 105 5.7 ‘Aqaba Castle: plan by Léon de Laborde, 1828. 106 5.8 ‘Aqaba: air photograph taken from the north-west during a bombing raid on British and Arab forces on 26 August 1918. 109 5.9 ‘Aqaba Castle: interior looking north, August 1918. 111 5.10 ‘Aqaba Castle: interior looking south-west, 1917–18. 111 (p.ix) 6.1 Map showing the political situation in the Mediterranean region, c.1600. 115 6.2 Map of the Ottoman maritime frontier in south-western Greece, c.1670. 116 6.3 Map of Evliya Çelebi’s itinerary in the Mani Peninsula, 1670. 125 6.4 Venetian plans of the fortress of Kelefa, c.1685. 128 6.5 Pictogram of Vittolo, c.1685. 131 6.6 An aerial view of the fortress of Kelefa. 134 7.1 Map of the Black Sea region. 138 7.2a Akkerman fortress: panorama from the liman. 144 7.2b Akkerman fortress: citadel and garrison yard walls. 144 7.2c Akkerman fortress: Guardian Tower and part of Wheel Tower. 145 7.2d Akkerman fortress: the ditch at Storeyed Tower. 145 7.3a Akkerman fortress plan showing locations identified from Ottoman documents. 146 7.3b Akkerman fortress: satellite image. 146 7.4 Barbican of Akkerman fortress viewed from the top of the civil yard wall, next to Fireplace Tower. 147 7.5 Port yard hamam of Akkerman fortress (between Fireplace Tower and Water Gate). 147 7.6 Sample pages from a building register including works carried out at Akkerman in 1709. 155 7.7 Akkerman fortress: Kauffer plan (1793). 159 7.8 Akkerman fortress: Förster plan (1807). 159 7.9 Akkerman fortress: Kauffer plan detail (1793). 159 7.10 Sgrafitto shards excavated in the port yard of Akkerman fortress. 166 7.11 Iznik shards excavated in the port yard of Akkerman fortress. 166 7.12 Kütahya coffee cup excavated in the port yard of Akkerman fortress. 166 7.13 China shards excavated in the port yard of Akkerman fortress. 166 7.14 Pipes excavated in the port yard of Akkerman fortress. 168 8.1 Palanka in L. F., Stato militare dell’Impero Ottomanno, 1732. 172 8.2 Map of the Ottoman Balkans c.1800. 174 8.3 Austrian map depicting the Ottoman–Hapsburg frontier in Croatia. 177 9.1 Late seventeenth-century engraving showing Turhan Sultan’s fortresses of Seddülbahir (Chateau Neuf d’Europe) and Kumkale (Chateau Neuf d’Asie). 190 9.2 Drawing of Seddülbahir in Museo Correr, Venice. 197 9.3 Engraving of Seddülbahir, 1717. 198 9.4 French plan showing proposed changes to be made at the fortress of Seddülbahir. 199 9.5 View of south section of Seddülbahir as seen from the entrance to the Dardanelles. 200 9.6 View of the bunkers in the north section of Kumkale. 200 9.7 The newly completed Ottoman officer barracks in the upper interior section of Seddülbahir prior to its destruction in the First World War. 201 Page 2 of 4 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Figures and Tables 9.8 Pre-First World War view of the long west wall of Seddülbahir. 201 9.9 The results of the bombardment of the west tower of Seddülbahir (the Başkule, or Principal Tower) during the Allied attacks of April 1915. 202 9.10 Restitution drawing of Seddülbahir for the late eighteenth-century phase of the fortress. 204 9.11 The ruins of the military hospital just outside the north-west wall of the Seddülbahir fortress. 206 10.1 Map showing the topography of the Lake Van region, in relation to modern political boundaries. 212 11.1 Map of north-east Africa in the late sixteenth century. 229 (p.x) 12.1 Map of Greece and Albania, showing the main places mentioned in the text in relation to modern political borders. 237 12.2 Map of the Gulf, showing the main places mentioned in the text in relation to modern political borders. 246 13.1 Map showing the borders of the Ottoman Empire, Venice and Dubrovnik. 254 13.2 The Flag of the Republic of the Seven United Islands. 260 14.1 Map of Ottoman Iraq in the late nineteenth century. 272 15.1 Map of Yemen in the late nineteenth century, showing the ‘tribes’ under British protection. 291 16.1 Relief map of Anatolia, showing principal places mentioned in the text. 310 16.2 Map of Konya and the surrounding region. 313 17.1 Map of the Ottomans’ Danube frontier, showing principal places mentioned in the text. 334 18.1 Map of the Ottoman–Safavid border in Eastern Anatolia and the Caucasus c.1612. 357 19.1 Tombs of the nineteenth-century Circassian governors of the Sudan, Khartoum. 372 19.2 Map of the Mahas region. 373 19.3 Qal‘at Sai, showing the reinforced south-west bastion. 377 20.1 Glazed Turkish ware, sixteenth to seventeenth century. 388 20.2 Line drawing of glazed Turkish ware, sixteenth to seventeenth century, Buda Castle. 389 20.3 Reduction-fired black jug, from seventeenth-century Hungary. 391 20.4 Hungarian pottery, seventeenth century. 393 20.5 Pots made on a hand-turned wheel, from late sixteenth-century Hungary. 393 20.6 (1–2) Pots made on a hand-turned wheel, late sixteenth century. (3–7) Hungarian pottery, late sixteenth century, all from Ozora Castle. 394 21.1 Map showing the location of the Bihaćka Krajina within Bosnia and Herzegovina, with successive Ottoman–Hapsburg frontier lines marked. 404 21.2 Map showing main places in the Krajina mentioned in the text, with modern political borders. 405 21.3 Plan of Pećigrad castle. 407 21.4 View of Pećigrad from the south-east showing castle and mosque within. 408 21.5 The gateway of Stijena castle surmounted by an inscription of AH 1185 / AD 1771. 411 21.6 A house in the Catholic/Orthodox tradition at Šiljkovača. 416 21.7 A typical house in the Muslim tradition of the Bihaćka Krajina at Grabovac, with construction detail. 417 21.8 Carved motifs on nišani in the cemetery at Lučka near Pećigrad. 419

Page 3 of 4 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Figures and Tables 21.9 Cutaway reconstruction drawing of a typical watermill, based on an example at Gornji Šiljkovača. 420 21.10 The potter Fehim Javurdzija photographed at Demiševci, north-west Bosnia in 1989. 422 21.11 Cooking pots from Demiševci, Brezova Kosa and Kaluđerovac. 422 21.12 Potter from the Jontić family photographed at Vrginmost (now Gvozd), Croatia in 1990. 428 22.1 Map of southern Jordan showing the location of Great Arab Revolt Project sites. 434 22.2 Plan of the area west of Ma‘an station showing recorded Ottoman defence works. 442 22.3 Plan of Wadi Rutm showing recorded late Ottoman sites. 445 22.4 Plot of expended munitions recovered in and in front of the north-west-facing firing trench on the Hill of the Birds at Ma‘an. 448 24.1 Map showing location of Suakin and other sites mentioned in the text. 470 (p.xi) 24.2 Plan showing location of main sites investigated and other features on Suakin Island. 475 24.3 View of the south wall, Khurshid Efendi’s diwan, c.1929. 478 24.4 Plan of Bayt al-Basha (House 184), showing diwan, dihliz, yard and ancillary buildings. 480 24.5 Sherds of green and white glazed ware, PS40. 487 24.6 Sherds PS20, PS27, PS01, PS06 and PS09. 487 24.7 Sherd PS03, part of open bowl. 488 24.8 Sherds PS24, PS25 and PS23. 488 25.1 Map of North Africa, showing the main places mentioned in the text. 496 26.1 Map showing location of Rumbek, South Sudan and other places mentioned in the text. 510 26.2 General location and plans of Pendit, Meen Atol and Lol Nhom zaribas. 526 26.3 Main enclosure ditch on the north side of Pendit zariba, Rumbek, looking east. 528 26.4 Enclosure bank on the east side of Meen Atol zariba, from the enclosure interior. 530 26.5 Southern corner of main enclosure ditch, Lol Nhom zariba, looking north-west. 531 26.6 Digitised version of Schweinfurth’s 1872 general map, showing the location of different zaribas, with those ‘owned’ by Ghattas highlighted. 533 26.7 Enlarged, digitised section of Schweinfurth’s 1872 map showing location of de Malzac’s, Ronga and Adael zaribas compared with location of Pendit, Meen Atol and Lol Nhom zaribas as logged in 2007. 534 Tables 7.1 Percentages of finds from the port yard at Akkerman fortress. 164 7.2 Percentages of ceramic wares from the port yard at Akkerman fortress. 164 20.1 Ceramic finds in Ottoman castles of sixteenth- to seventeenth-century Hungary. 396 24.1 Summary of trade contacts and trade items mentioned in European accounts of Suakin between the sixteenth and mid-nineteenth century. 484 24.2 Tonnage of trade from Suakin in the second half of the nineteenth century. 485

Access brought to you by: Page 4 of 4 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Notes on Contributors

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

(p.xii) (p.xiii) Notes on Contributors Gábor Ágoston is Associate Professor at the Department of History of Georgetown University. His research focuses on Ottoman economic and military history from the fifteenth to the eighteenth century, early modern Hungarian history, and the comparative study of the Ottoman and Hapsburg Empires. In addition to his five Hungarian-language books and numerous articles, he is the author of Guns for the Sultan: Military Power and the Weapons Industry in the Ottoman Empire (2005) and co-editor and co-author with Bruce Masters of Encyclopedia of the Ottoman Empire (2009). John Alexander is a fellow of St John’s College, Cambridge, and was formerly University Lecturer in Archaeology at the University of Cambridge. His research interests include the prehistory of Africa and Europe, frontier theory, and the beginnings of iron-using and urbanism, and he has worked extensively at the Ottoman forts of Qasr Ibrim and Sai in Nubia. Publications include The Directing of Archaeological Excavations (1970). Frederick Anscombe is Senior Lecturer in Contemporary History at Birkbeck College, University of London. His main research interest is the history of the Ottoman Empire from the late seventeenth to the early twentieth century as viewed from the provinces. He is the author of The Ottoman Gulf: The Creation of Kuwait, Saudi Arabia, and Qatar (1997) and editor of The Ottoman Balkans, 1750–1830 (2006). Svitlana Bilyayeva is a Senior Research Associate at the Institute of Archaeology of the National Academy of Sciences of Ukraine. She is a specialist in Old Rus’ as well as Golden Horde and Ottoman archaeology. Aside from Ottoman BilhorodDnistrovskyi (Akkerman), she has led several excavations at Ochakiv (Özi). Isa Blumi is Assistant Professor of Middle Eastern and Balkan history at Georgia State University in Atlanta. His work extends from the early nineteenth-century Ottoman administration in the Balkans and Red Sea region to post-Ottoman Arabia and the Balkans. In addition to over two dozen articles, he has authored Rethinking the Late Ottoman Empire: A Comparative Social and Political History of Albania and Yemen, 1878–1918 (2003). (p.xiv) Colin Breen is Senior Lecturer in Archaeology at the University of Ulster, Northern Ireland, having previously worked at the Queen’s University of Belfast and as an Irish government marine archaeologist. His research interests lie in historical Page 1 of 6 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Notes on Contributors and maritime archaeology with a particular focus on Britain, Ireland and Africa. His most recent book is An Archaeology of Southwest Ireland, 1570–1670 (2007). Palmira Brummett is Professor of History and Distinguished Professor of Humanities at the University of Tennessee. She is editor of and contributor to The Book of Travels: Genre, Itinerary, Ethnology and Pilgrimage, 1250–1700, forthcoming from Brill. Her work concerns the rhetorics (narrative and visual) of interaction between the Ottoman Empire and the Christian kingdoms of Europe in the early modern era, and her current project is a monograph on early modern mapping of Ottoman space and sovereignty. Richard Carlton is a visiting fellow at the University of Newcastle upon Tyne, School of Historical Studies and a Director of The Archaeological Practice, a cultural heritage consultancy based in Newcastle upon Tyne. His principal area of individual research is ceramic technology, focusing particularly on the Western Balkans, where he has undertaken ethnographic and archaeological fieldwork since the 1980s. He is the author of various publications on this theme, including, most recently, The Role and Status of Women in the Pottery-Making Traditions of the Western Balkans (2008). Rahmi Nurhan Çelik is an Associate Professor in the Geodesy division of the Department of Geodesy and Photogrammetry at Istanbul Technical University. His research interests include the geodetic infrastructure of geographical information systems, geodetic network design and advanced electrometry. With Lucienne ThysŞenocak he has co-directed a survey and architectural documentation project at the Ottoman fortresses of Kumkale and Seddülbahir, and since 2005 the conservation project at the latter fortress. Gökhan Çetinsaya is a Professor of History at Istanbul Technical University, Department of Humanities and Social Sciences. He works on Ottoman and Middle Eastern history. He is the author of Ottoman Administration of Iraq, 1890–1908 (2006). Intisar Soghayroun Elzein is an Associate Professor in the Department of Archaeology, University of Khartoum, Sudan. Her main field of research is Islamic archaeology, and her current research work is a cultural and archaeological survey of Northern Dongola. She is the author of Islamic Archaeology in the Sudan (2004). Neil Faulkner is an Honorary Research Fellow at the University of Bristol and Codirector of the Sedgeford Historical and Archaeological Research Project (in Britain) (p.xv) and of the Great Arab Revolt Project (in Jordan). Adopting a Marxist approach, he is interested in the dynamics of imperial systems, in the rise and fall of empires, and in the way in which empire is experienced and contested by subject classes. His most recent book is Rome: Empire of the Eagles (2008), but his main research focus is now on the period 1914–23, with particular reference to the Middle East, where he undertakes modern conflict archaeology. Caroline Finkel is a historian of the Ottoman Empire. She is an Honorary Fellow at the universities of Edinburgh and Exeter. She is co-director with Svitlana Bilyayeva and Victor Ostapchuk of the Akkerman Fortress Project in Ukraine, and an organising participant in the Evliya Çelebi Ride, an international project of historical re-enactment tracing the route of this seventeenth-century Ottoman traveller in north-western Anatolia. Her most noted work is Osman’s Dream: The Story of the Ottoman Empire, 1300–1923 (2005), a narrative history now published in several languages.

Page 2 of 6 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Notes on Contributors Wes Forsythe is a lecturer at the Centre for Maritime Archaeology, University of Ulster. His research interests lie in the maritime activities, economies and infrastructure of urban centres along the East African littoral, encompassing both underwater and coastal expressions of material culture. His work has focused on the Swahili of Tanzania and Kenya, as well as Islamic and post-medieval Sudan. Ibolya Gerelyes is Deputy Director of the Archaeological Department of the Hungarian National Museum in Budapest. Her doctoral dissertation investigated probate inventories compiled for Turkish inhabitants of Buda during the sixteenth century. She has taken part in excavations at a number of Ottoman castles and has made a special study of ceramics discovered there. She has edited the volumes Archaeology of the Ottoman Period in Hungary (2003) (with Gyöngyi Kovács) and Turkish Flowers: Studies on Ottoman Art in Hungary (2005). With support from the Max van Berchem Foundation in Geneva, she is currently writing an Englishlanguage catalogue of Ottoman metalwork held by the Hungarian National Museum. Rossitsa Gradeva is Associate Professor of Ottoman and Balkan History at the American University in Bulgaria and Research Fellow at the Institute of Balkan Studies, Bulgarian Academy of Sciences. Her research interests are mainly in the field of Ottoman institutions of provincial administration, application of Islamic law in the Ottoman Empire, various aspects of the status of non-Muslim communities and everyday life in the Ottoman Balkan provinces, the Ottoman–Danube frontier and the development of the decentralisation processes in the region in the pre-Tanzimat period. Selections from her studies in these fields were published in Rumeli under the Ottomans, 15th–18th Centuries: Institutions and Communities (2004) and War and Peace in Rumeli: 15th to Beginning of 19th Century (2008). (p.xvi) Colin Heywood is an Honorary Research Fellow of the Maritime Historical Studies Centre, University of Hull. He is currently working on aspects of the history of English shipping in the Mediterranean in the seventeenth and eighteenth centuries and is coeditor of a collective volume on the theme of After Braudel: The ‘Northern Invasion’ and the Mediterranean Maritime World, 1580–1798, which will appear shortly. Douglas H. Johnson is a historian of the Sudan. He is the author of Nuer Prophets (1994), and The Root Causes of Sudan’s Civil Wars (2003). He has conducted research on military slavery in nineteenth- and twentieth-century North East Africa. Paul Lane is a Senior Lecturer in Archaeology at the University of York and Director of the Historical Ecologies of East African Landscapes (HEEAL) project. He is an archaeologist who specialises in the study of the later Holocene archaeology of Africa. His doctoral dissertation (1986) was an ethnoarchaeological study of space and time among the Dogon in Mali, West Africa. Before joining the University of York in 2007, he taught at the universities of Dar es Salaam, Tanzania (1989–91) and Botswana (1992–7), and was Director of the British Institute in Eastern Africa (1998– 2006). His current research interests are in African landscape and historical archaeology, the materialisation of memory, the transition to farming in Africa, and indigenous archaeologies. With Andrew Reid, he is editor of African Historical Archaeologies (2004), and joint editor with Andrew Reid and Alinah Segobye of Ditswa Mmung: The Archaeology of Botswana (1998). Michael Mallinson is a co-director of the Suakin Project and a director of Mallinson Architects and Engineers Ltd, who has been responsible for designing and building museums in Egypt, Sudan and Ethiopia with their governments’ antiquities services and as a consultant for UNESCO and the World Bank. He has worked on Page 3 of 6 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Notes on Contributors archaeological projects since 1981 with the Egypt Exploration Society, the British Institute in Eastern Africa, the Royal Ontario Museum, the Sudan Archaeological Research Society, and the McDonald Institute for Archaeological Research in Cambridge. Rhoads Murphey is Reader in Ottoman Studies at the Institute for Archaeology and Antiquity in the College of Arts & Law, University of Birmingham. His research interests range broadly over topics relating to the social, economic, cultural and institutional history of the Ottoman lands in the ‘middle’ centuries (c. 1500–c. 1750) of Ottoman rule. His latest book is Exploring Ottoman Sovereignty: Tradition, Image and Practice in the Ottoman Imperial Household, 1400–1800 (2008). Victor Ostapchuk is an Associate Professor at the Department of Near and Middle Civilizations of the University of Toronto. His interests include the Ottoman Black Sea and the empire’s relations with Muscovy, Poland and Ukraine as well as Ottoman (p.xvii) historical archaeology. He is the author of Warfare and Diplomacy across Sea and Steppe: The Ottoman Black Sea Frontier in the Early Seventeenth Century (forthcoming). Burcu Özgüven is Associate Professor of Architecture at the Faculty of Engineering and Architecture, İzzet Baysal University in Bolu, Turkey. She is an architect and architectural historian specialising in military architecture and urbanism within the Ottoman territories from the sixteenth to the nineteenth century, and buildings in Istanbul during the period of westernisation. Her publications include a book on Ottoman-Hungarian fortifications and towns, Osmanlı Macaristan’ında Kentler, Kaleler (2001). She is currently conducting research on architectural developments in Istanbul during the inter-war period. Arzu Özsavaşçı is a professional architect who received her MA degree in 1999 from the History of Architecture Program at the Architecture Department of Istanbul Technical University. She is the project architect and the architectural team leader of the Seddülbahir fortress survey and conservation project. Her research interests include the application of new surveying technologies in architectural conservation and restoration projects, and the implications of these new methods for architectural practices. Currently she is working in Istanbul at the architectural firm ETUD Mimarlik. Andrew Peacock is Assistant Director of the British Institute at Ankara, and is a specialist in Islamic history. Publications include Mediaeval Islamic Historiography and Political Legitimacy (2007) and Early Seljuq History: A New Interpretation (in press). He is co-director of a British Academy-sponsored project investigating links between the Ottoman Empire and South-East Asia, Islam, Trade and Politics across the Indian Ocean. Andrew Petersen is lecturer in Islamic Archaeology and Director of Research on Islamic Culture at the University of Wales, Lampeter. He has worked on archaeological projects throughout the Muslim world and is director of the Ottoman Hajj Forts project. Recent publications include Gazetteer of Buildings in Muslim Palestine (2001) and The Towns of Muslim Palestine (2005). His current research interests include a project which focuses on the archaeological evidence for Ottoman rule in Arabia. Jacke Phillips is a Teaching Fellow at the School of Oriental and African Studies, University of London. Her research focuses on intercultural contact, trade and interaction, particularly the material and ideological transmission of artefacts and images through peripheral regions within and beyond the East Mediterranean and Page 4 of 6 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Notes on Contributors North-East Africa, including the Bronze Age Aegean, ancient Egypt, Aksumite Ethiopia, Christian Nubia and Islamic Sudan. She has excavated and published in all these fields, her most recent book being Aegyptiaca on the Island of Crete in their Chronological Context: A (p.xviii) Critical Review (2008). Her current research considers East-West land connections between the southern Red Sea and the Nile Valley. Denys Pringle is a Professor in the Cardiff School of History and Archaeology, Cardiff University. His research interests embrace the material culture of the Mediterranean world from late antiquity onwards, with a particular focus on the medieval Crusader states. He has recently completed the fourth and final volume of The Churches of the Crusader Kingdom of Jerusalem: A Corpus (1993–). His other recent publications include An Expatriate Community in Tunis 1648–1885: St George’s Protestant Cemetery and its Inscriptions (2008). Alan Rushworth is a director of the Archaeological Practice, Newcastle upon Tyne, and has been involved in the investigation of a wide range of sites and periods, most recently completing the writing up of a long programme of fieldwork by Newcastle University and others at Housesteads on Hadrian’s Wall—The Grandest Station: Excavation and Survey at Housesteads Roman Fort 1954–95—which is currently in press. Following his doctoral dissertation (1992) on the interaction of the imperial army and tribal society along the frontiers of late Roman Africa, his principal research interest has focused on frontier society, settlement and fortifications from antiquity to the early modern era. Kahraman Şakul is a Ph.D. candidate in the Department of History, Georgetown University. The topic of his doctoral dissertation is Ottoman–Russian relations in the Age of Napoleon. His research interests include military and technological aspects of Ottoman reforms as well as political culture of the Ottoman court in the transitional period of 1774–1826. He is currently a research fellow at the Research Center for Anatolian Civilizations, Koç University, Istanbul, where he is working on images of Russia on the eve of Ottoman modernisation (1774–1826). Nicholas J. Saunders is Senior Lecturer in the Department of Archaeology and Anthropology at the University of Bristol, and is co-director of the Great Arab Revolt Project in Jordan. Between 1998 and 2001, he was British Academy Institutional Fellow at University College London, making the first anthropological study of the material culture of the First World War. Since then, he has played a leading role in the development of a multidisciplinary archaeology of twentieth-century conflict, publishing many books and articles on this topic. He is particularly interested in the relationship between material culture and conflict memory. Tom Sinclair is an Associate Professor in Turkish History at the University of Cyprus. He works on eastern Asia Minor in the late Middle Ages and the early Ottoman period. At present the biggest project on which he is engaged is a study of the Van region. This (p.xix) involves micro-studies of individual cities and dynasties, and will compare the region’s administration in the late Middle Ages with that in the first century of Ottoman rule. He has also prepared a book on the Ayas-Tabriz trade route of the late thirteenth and early fourteenth centuries. He is the author of Eastern Turkey: An Architectural and Archaeological Survey (4 vols., 1987–90). Laurence Smith is a Research Fellow at the McDonald Institute, Cambridge University and a Senior Member of Wolfson College. Initially, he undertook studies of trade in Nubia during the Meroitic and Christian periods through provenancing pottery, and worked as pottery assistant on three excavations and a survey in Page 5 of 6 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Notes on Contributors northern Sudan. Subsequently, he has co-directed (with Michael Mallinson) an archaeological survey along the road north of Khartoum and the current project at Suakin, beginning in 2002. Mark L. Stein is Associate Professor of History at Muhlenberg College. His research focuses on the social and economic conditions along Ottoman frontiers as well as how state elites viewed the frontiers of the empire. His recent publications include Guarding the Frontier: Ottoman Border Forts and Garrisons in Europe (2007) and Osmanlı Kaleleri Avrupa’da Hudut Boyları (2007). Gülsün Tanyeli is an Assistant Professor of Architectural Conservation in the Department of Architecture at Istanbul Technical University. Her research focuses on the history of Ottoman building technology and industrial archaeology. She is a member of the archeological and historical research team investigating DemirköySamakocuk, an Ottoman iron production centre in Thrace. Lucienne Thys-Şenocak is an Associate Professor in the Department of Archaeology and the History of Art at Koç University, Istanbul. She is also the coordinator of the Anatolian Civilizations and Cultural Heritage Management Graduate Program at the same institution. Her research interests include the architectural patronage of imperial Ottoman women in the early modern era, Ottoman fortifications, and the history of the Gallipoli region of Turkey. With Rahmi N. Çelik she has co-directed the survey and architectural documentation project at the Ottoman fortresses of Kumkale and Seddülbahir, and since 2005, the conservation project at the latter fortress. Among her recent publications is Ottoman Women Builders: The Architectural Patronage of Hadice Turhan Sultan (2006). Malcolm Wagstaff is Professor Emeritus in the University of Southampton and former visiting fellow at the British Schools in Athens and Jerusalem, as well as a long-term member of the British Institute in Ankara. He has research interests in the development of settlements and land use in the former Ottoman territories, especially Greece (p.xx) and Turkey. His book Greece: Ethnicity and Sovereignty, 1820–1994 was published by Archive Editions in 2002. Sara Nur Yıldız is currently a post-doctoral fellow at the German Orient-Institut Istanbul, while on research leave from the History Department, Istanbul Bilgi University. She is a historian of medieval and early Ottoman Anatolia with interests in empire-building and frontier politics, political culture and historical writing. She is completing a monograph based on her doctoral dissertation, Mongol Rule in Seljuk Anatolia: The Politics of Conquest and History Writing, 1243–1282 (forthcoming), and is writing a general study of Seljuk Anatolia, tentatively entitled The Seljuks of Anatolia: A Muslim Empire on the Frontier.

Access brought to you by:

Page 6 of 6 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Acknowledgements

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

(p.xxi) Acknowledgements The title of this volume, and many of the contributions in it, derive from a workshop entitled ‘The Frontiers of the Ottoman World: Fortifications, Trade, Pilgrimage and Slavery’, held at the British Academy on 15 and 16 February 2007. This was supported by a grant from the Board for Academy-Sponsored Institutes and Societies, and organised by the British Institute at Ankara in collaboration with the British Institute in Eastern Africa, the British Institute for the Study of Iraq, the British School at Athens, the Council for British Research in the Levant, the Egypt Exploration Society and the Society for Libyan Studies. I would like to take this opportunity to express my especial thanks to Gina Coulthard, who undertook much of the burden of organising the workshop. I am very grateful to the numerous individuals who have given of their expertise and advice in preparing this volume. Particular thanks are due to Ben Fortna and Alison Gascoigne for their counsel both in arranging the original workshop and subsequently. I am indebted to Machiel Kiel and Caroline Finkel for their advice. I am also very grateful to the numerous anonymous referees who gave their time to comment on individual chapters. I would also like to thank the copyright-holders who gave permission for illustrations to be reproduced here; a full acknowledgment will be found by each illustration. The maps were prepared for publication by Pieter Collet, whose expertise and patience with a complicated task is much appreciated. I am also very grateful to the volume’s copy-editor, Susan Milligan, for her meticulous work. Finally, I would like to thank the British Academy and the Board for Academy-Sponsored Institutes and Societies for providing the funding that made both the original workshop and this volume possible. A. C. S. Peacock Ankara, May 2008

Access brought to you by:

Page 1 of 1 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

A Note on Transliteration, Place Names and the Organisation of this Volume

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

(p.xxii) A Note on Transliteration, Place Names and the Organisation of this Volume In general, words have been transliterated according to the rules of Turkish orthography rather than Arabic, so vakıf is preferred to waqf. However, it was felt on occasion that anachronism was more desirable than an unsightly consistency, and this rule has not been applied rigorously for discussions of the Arab provinces of the empire. The English form Pasha has generally been preferred to paşa except in quotations from Ottoman. Complete consistency was also elusive with regard to place names. Especially in the Balkans, names often have several different guises in Ottoman, Hungarian and Slavic languages. Moreover, to refer to well-known places in Greece by their obscure Ottoman names did not seem helpful to the reader. On the other hand, to use only the modern names gave rise to even more anachronisms. Ottoman and modern versions of the name are supplied wherever possible, although it should be noted that in some instances it is not possible to link the name given in the Ottoman texts with a known settlement and vice versa. Except in Greece, the Ottoman version of the name is generally given first. In the Arab world, places are referred to by the names they are known by today in the west (e.g. Damascus) or by a transliteration of their Arabic names. This volume is organised around four key themes: fortifications, administration, frontier society, and the economy of the frontier, which aim to allow the reader to compare historical developments at different times in far-flung provinces. Within each section a broadly chronological arrangement has been followed. The Select Bibliography, however, is arranged according to region, so readers wishing to explore a specific frontier in more detail should consult this.

Access brought to you by:

Page 1 of 1 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Abbreviations

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

(p.xxiii) Abbreviations Ar. Arabic BOA Başbakanlık Osmanlı Arşivi (Prime Ministry Ottoman Archive), Istanbul EI1 Encyclopedia of Islam (1st edn., Leiden: Brill, 1913–36) EI2 Encyclopaedia of Islam (2nd edn., Leiden: Brill, 1960–2005) EIr Encyclopaedia Iranica (Columbia University, 1985–) Evliya Çelebi, Seyahatname Evliya Çelebi, Seyahatnamesi, ed. and transcribed by S. A. Kahraman, Y. Dağlı and R. Dankoff et al. (Istanbul, Yapı Kredi Yayınları, 10 vols., 1999–2007) Gr. Greek Hu. Hungarian İA İslam Ansiklopedisi (Istanbul, Ankara: Maarif Matbaası, Milli Eğitim Bakanlığı, 1940–1986) Ott. Ottoman Pol. Polish Rom. Romanian Serb. Serbian Sl. Slovak TNA The National Archives (formerly Public Records Office), Kew, London TDVİA Türkiye Diyanet Vakfı İslam Ansiklopedisi (Ankara: Türkiye Diyanet Vakfı, 1988–) TSA Topkapı Sarayı Arşivi (Topkapı Palace Archive), Istanbul TTK Türk Tarih Kurumu

Access brought to you by:

Page 1 of 1 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Introduction: The Ottoman Empire and its Frontiers A. C. S. PEACOCK

DOI:10.5871/bacad/9780197264423.003.0001

Abstract and Keywords Stretching across Europe, Asia and Africa for half a millennium bridging the end of the Middle Ages and the early twentieth century, the Ottoman Empire was one of the major forces that forged the modern world. The chapters in this book focus on four key themes: frontier fortifications, the administration of the frontier, frontier society and relations between rulers and ruled, and the economy of the frontier. Through snapshots of aspects of Ottoman frontier policies in such diverse times and places as fifteenth-century Anatolia, seventeenth-century Hungary, nineteenth-century Iraq or twentieth-century Jordan, the book provides a richer picture than hitherto available of how this complex empire coped with the challenge of administering and defending disparate territories in an age of comparatively primitive communications. By way of introduction, this chapter seeks to provide an overview of these four themes in the history of Ottoman frontiers. Keywords:   Ottoman Empire, frontier fortifications, frontier society, Middle Ages, Ottoman frontiers, Hungary, Anatolia

STRETCHING ACROSS EUROPE, ASIA AND AFRICA for half a millennium bridging the end of the Middle Ages and the early twentieth century, the Ottoman Empire was one of the forces that forged the modern world. The conversion of Balkan communities to Islam under Ottoman rule has left a legacy evident in the region’s conflicts in the late twentieth century, while the states of the modern Arab world were largely built out of the ruins of the empire after the First World War. From Kosovo to Cairo, a rich—if often underappreciated—architectural heritage testifies to the Ottoman presence. Yet Ottoman history is still seriously neglected and under-researched compared to that of its contemporaries and rivals in Europe such as the Hapsburg or Russian empires. Although we now know much about certain regions of the Ottoman Empire, and certain periods, others remain in almost total obscurity. Above all, in contrast to other empires of the past such as Rome or China, the Ottomans’ frontiers have rarely been examined. They are, however, particularly important to understanding the development of the empire. They were places both of interaction with the outside world through trade and of conflict in war, places

Page 1 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers where the empire’s prestige and authority were both displayed and challenged. The essays gathered in this book attempt to fill a historiographical lacuna by investigating the Ottomans and their frontiers over the fifteenth to twentieth centuries. However, the volume aims not just to survey our current state of knowledge of Ottoman frontiers, but also to signal future possible avenues of research. For every modern state, its frontiers constitute a vital part of its identity, marking—in theory at least—the extent of its government’s authority and serving as a barrier or bridge to the world beyond. Frontiers today are usually precisely delineated, with the exception of a few areas of the world where geography is too inhospitable to allow this, such as the Empty Quarter between Saudi Arabia, Oman and Yemen, or where they are under dispute, such as the Himalayan frontiers of Pakistan, India and China. Thus the modern understanding of frontiers and borders relies largely on relatively recent or specific technology and concepts: precise maps, only really sufficiently developed for (p.2) exact border delineation in the nineteenth century (and not always then);1 and governments on both sides of the frontier that accept geographical limitations to their sovereignty. Some historians have wondered whether the concept of frontiers is valid at all for the pre-modern world, at least until the development of reasonably sophisticated cartography in the Renaissance.2 However, every political entity has a geographical point at which it comes into contact with the world beyond. The continuing stream of historical works dealing with frontiers in antiquity and the Middle Ages from Europe to China indicates there is still widespread interest in and acceptance of the existence of pre-modern frontiers, even if the nature of these may be far removed from our modern conceptions.3 Ottoman frontiers were constantly shifting as the empire expanded and contracted over its history. They also varied radically in their nature, according to geography, local circumstances, and the nature of the powers contesting Istanbul’s authority. In Europe, on the front line with the Hapsburgs, something akin to a modern frontier had developed by the sixteenth century, if not before, with the two powers confronting each other across a delineated border from behind a network of fortifications, even if this frontier line was not always taken particularly seriously in practice.4 In sixteenth- to seventeenth-century Iraq, the border with the Safavids of Iran was more ambiguous, with numerous tribes of fluctuating loyalties inhabiting the no-man’s land between the two empires, but for merchants and pilgrims clearly defined stations nonetheless marked the transition from one political jurisdiction to another.5 In Africa, on the other hand, it remains unclear how far the empire’s reach stretched up the Nile, and it seems more likely that rather than there having been a direct frontier with the Ottomans’ rivals, the Funj Sultanate of Sinnar, the authority of both sides simply petered out on the Middle Nile, leaving the intervening space to be ruled by local rulers who might pay nominal allegiance to one power or another (see Elzein in Chapter 19 of this volume). In the Hijaz, the Ottomans struggled to control the roads connecting the pilgrimage centres of Mecca and Medina with Syria and ultimately Istanbul. The sultan’s prestige could (p.3) be challenged by local forces such as the Bedouin of Arabia as well as rival states and empires. In the Mediterranean, Ottoman authority was contested by the navies of western powers and by corsairs. Thus the varying approaches of the essays in this volume reflect the complexity and diversity of the topic itself. The essays also reflect the different methodologies that may be applied to the study of the empire. Traditionally, the study of Ottoman history has been based on textual sources: chronicles and archival materials. The astonishing wealth of the Ottoman archives (mainly in Istanbul, but also with lesser-known collections elsewhere such as Sofia and Cairo) means that this field of research is not likely to be exhausted for the foreseeable future. The surveys (tahrir) carried

Page 2 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers after the conquest of a new province offer an insight into its economy and population. The mühimme defteris, registers of important events, record correspondence between Istanbul and officials in provincial centres. Financial records inform us of the expense of maintaining soldiers on the frontiers and the costs of constructing or repairing fortresses. Repair records show us where the materials for such fortifications were brought from and how far the central administration in Istanbul intervened in the business of maintaining its frontier defences. Taxation registers and court records offer an insight into daily life on the frontiers. These are but a few of examples of the types of information and records that are extant, and naturally, a good number of the essays presented here exploit these sources. Ottoman archaeology, on the other hand, is a much more recent field of enquiry, the potential of which is only slowly coming to be realised.6 It will never be possible to attempt to write the history of the empire on the basis of archaeology alone, but it can substantially enrich our knowledge, especially in areas like the Sudan, where written sources for the Ottoman presence are few and far between. Even elsewhere archaeology offers a unique insight into frontier life that the archives alone cannot provide. The location and structure of fortifications can tell us much about how the Ottomans perceived threats and how they responded to them (see Chapters 3, 4, 5 and 8, by Ágoston, Petersen, Pringle and Özgüven respectively). Even simple small finds, such as fragments of pottery or smoking-pipes, can give us an impression of the extent to which Ottoman culture penetrated distant provinces (see Chapter 7 by Ostapchuk and Bilyayeva for examples from Akkerman fortress in the Ukraine). Some of the essays in this volume rely largely on one type of evidence—textual or archaeological —and some attempt to combine both. The latter is rarely as simple a matter as it may seem, for textual and archaeological evidence does not always neatly fit together, as Lane and Johnson note in their chapter on slavery in the Sudan (Chapter 26), and as some evidence from Akkerman also suggests. This in itself raises interesting (p.4) methodological problems: why does this seeming conflict arise, even when the textual evidence is archival, and so less likely to be subject to the whims of a more literary author? How can one reconstruct the history of a site in the face of these contradictions?7 We are far from being able to resolve such issues, but it is hoped that drawing attention to them in the context of Ottoman and Islamic archaeology will in itself be a valuable service to both historians and archaeologists of the Middle East and Ottoman Europe, who often work in ignorance of each other’s research. The essays focus on four key themes: frontier fortifications, the administration of the frontier, frontier society and relations between rulers and ruled, and the economy of the frontier. It is far beyond the capabilities of a single volume to give a comprehensive overview of the empire’s complex frontiers over 500 years. However, through snapshots of aspects of Ottoman frontier policies in such diverse times and places as fifteenth-century Anatolia, seventeenth-century Hungary, nineteenth-century Iraq or twentieth-century Jordan we can provide a richer picture than hitherto available of how this complex empire coped with the challenge of administering and defending disparate territories in an age of comparatively primitive communications. By way of introduction, this chapter seeks to provide an overview of these four themes in the history of Ottoman frontiers. First, however, we shall examine the historical context of the evolution of the Ottoman frontiers.

Ottoman Frontiers: A Historical Overview The study of Ottoman frontiers is of course intimately linked to that of Ottoman expansion and contraction.8 A place like Vidin (on the Danube in modern Bulgaria; see Gradeva’s study in Chapter 17) was a frontier town in the immediate aftermath of its conquest at the end of the Page 3 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers fourteenth century. With Ottoman conquests in Europe in the fifteenth and sixteenth centuries, it became part of the core Ottoman lands, but Ottoman defeats at the hands of its European rivals meant that by the end of the seventeenth century Vidin was on the frontier again, where it remained until the Ottomans definitively lost control of it in the nineteenth century. Although the traditional paradigm of the empire’s ‘decline’ lasting from the seventeenth century to the First World War is widely rejected by scholars today,9 it is evident that there were periods (p.5) when the empire was broadly gaining territory at the expense of its neighbours, and those when it was losing it. Although any attempt at periodisation of the empire is bound to vary according to the criteria used, we may discern four broad phases of the physical expanse of the empire’s boundaries, with the caveat that these are extremely rough and ready outlines, with plenty of exceptions (see Figure 1.1). The first phase was the early Ottoman principality or beylik, which originated c.1299 in northwestern Anatolia. As the principality gradually expanded into Europe and Anatolia, its rulers, the descendants of an obscure Türkmen chief named Osman b. Ertuğrul, arrogated to themselves the Islamic title of sultan, indicating their aspiration to be accepted as legitimate in the Muslim world. For much of the first hundred years, until the reign of Sultan Bayezid I (1389–1402), the Ottomans concentrated on conquests less in Anatolia than in Europe, occupying Thrace in 1352 and seizing Balkan cities such as Serbian Ni8 (1386) and Byzantine Thessalonike (1387). With Bayezid came a more concerted effort to advance to the east, and this led to the first major setback to Ottoman expansion. This came from the Muslim not the Christian world, with the catastrophic defeat of Bayezid by Timur (Tamerlane) at the Battle of Ankara in 1402. The Ottomans lost most of their territories in Anatolia and a fierce civil war raged between the descendants of Bayezid over the succession. However, by the mid-fifteenth century the Ottomans had fully recovered their former territories, and the completion of the conquest of Byzantium with the fall Constantinople in 1453 to Sultan Mehmed II (‘the Conqueror’) marks the end of the first phase of the Ottoman state. At this point the frontiers of the empire stretched across eastern Europe, including much of northern Greece, Bulgaria and Serbia. In Asia, the Ottomans had re-incorporated most of western Anatolia into their state, but central principalities like the Karamanids retained a precarious autonomy as Ottoman vassals, and the east remained apart from Ottoman rule.10 This seemingly invincible expansion continued for most of the second phase of the empire, the remainder of the fifteenth and sixteenth centuries. Mehmed himself, before his death in 1481, continued his conquests in both directions, in Europe incorporating the Peloponnese, Bosnia and the remainder of Serbia into his empire, and largely completing the subjugation of Albania initiated in the 1380s. In Asia he put an end to the last outpost of Byzantium with the conquest of Trebizond in 1461, and expanded further into Anatolia with the defeat of the powerful Akkoyunlu tribal confederation that dominated eastern Anatolia, northern Iran and Iraq from their capital at Diyarbakır at the Battle of BaĿkent in 1473. The Black Sea too had been turned into an Ottoman lake with the gradual expulsion of the Genoese colonies and the acceptance of Ottoman sovereignty by the Crimean Tatars in 1478. Thus by the late fifteenth century the Ottoman Empire stretched from the southern borders of Georgia to modern Croatia. (p.6)

Page 4 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers (p.7) Expansion continued under Bayezid II (1481–1512). This resulted in wars with the Mamluks of Egypt and Syria for control of the approaches to Syria, with Iran over eastern Anatolia, where the new Safavid dynasty based its support on the Shi’i Türkmen tribesmen (the ‘KızılbaĿ’) who remained a persistent nuisance to the Ottomans, and with Venice for control of the Mediterranean. Fewer than a hundred years after the disastrous defeat at Ankara the Ottomans had risen from virtual extinction to being a world power of the first order. Figure 1.1. Map of the expansion of the Ottoman Empire. The fate of prince Cem, Mehmed II’s younger son and Bayezid’s rival for the throne, illustrates how even the empire’s internal affairs could resonate throughout Europe and Middle East. After Cem’s failure to seize the throne, he fled abroad, initially to Egypt, and then was held in captivity variously in Rhodes, France and Italy, as the powers of the day—the Knights of St John (his original captors), the Mamluks and the Pope—all haggled with each other for custody of this valuable prisoner, while extorting subsidies from his sibling, Bayezid, in return for keeping him.11 Succession disputes apart, Ottoman victories continued apace with Selim I’s (1512–20) defeat of the Mamluks and incorporation of their Arab-speaking lands in Syria, Egypt and the Hijaz into the empire. This has been seen by some scholars as the true turning point in Ottoman history12 Previously, most of the empire’s subjects had been Christian, but now for the first time the majority of the population was Muslim. The earlier Ottoman policy of rule over Christians by accommodation (istimalet) was replaced by a more classically Islamic, Arab-influenced system.13 Further conquests followed under Süleyman I (known in the west as ‘the Magnificent’ and in Turkish as Kanuni, ‘the Lawgiver’), in Hungary, Yemen, Iraq and Abyssinia. Yet in retrospect Süleyman’s reign also marked the beginning of the end of expansion, although it doubtless did not seem so to his Ottoman contemporaries, for whom defeats like the failure of the first Ottoman siege of Vienna (1529) probably appeared to be temporary setbacks that could later be overcome.14 The Hapsburgs of Austria blocked further Ottoman advances into Europe, and in the far west, although much of North Africa fell under Ottoman control during the sixteenth century, Ottoman attempts to annex Morocco were ultimately unsuccessful, despite briefly installing a dependent ruler in Fez.15 The Sublime Porte’s efforts to thwart Portuguese ambitions in the Arabian Gulf (p.8) and the Indian Ocean also ended in failure,16 while the accession of Ivan IV ‘The Terrible’ as Russian Tsar in 1552 and Muscovy’s subsequent annexation of Muslim khanates in Kazan and Astrakhan marked the rise of a powerful new enemy to the Ottomans. Nonetheless, the empire’s borders reached their largest extent under the reigns of Süleyman and his successors Selim II (1566–74) and Murad III (1574–95). Despite the Treaty of Amasya in 1555 which had brought peace in the east (and substantial territorial gains for the Ottomans at Iran’s expense), war with the Safavids resumed in 1578, resulting in the temporary annexation of much of Caucasia and the capture of the Safavid capital, Tabriz, in 1585. Further afield, the Ottomans were in diplomatic contact with polities as remote as Bornu by Lake Chad,17 and Aceh in Indonesia, to which they gave military support against the Portuguese.18

Page 5 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers Despite the successes against Iran, the late sixteenth century is traditionally seen as having marked the start of the Ottomans’ decline, beginning with the defeat of Selim II’s navy by the Holy League at the Battle of Lepanto in 1571.19 More recently, scholars have preferred to describe the period as one of adaptation to new challenges, and it seems that the Ottomans maintained military parity with western and central European powers until c.1700.20 However, even if the importance of defeats like Lepanto from an Ottoman perspective has been overstated, and some territories were still added to the empire, it is true that problems had become evident by around 1590, in the wake of the costly war with Iran. This period, stretching to the early nineteenth century, may be described as marking the third phase of the empire, in which its borders gradually contracted under pressure from neighbouring powers. Admittedly, the Ottoman borders in Europe remained relatively stable until the latter half of the seventeenth century, but the empire suffered from numerous internal rebellions in the Ottoman heartlands of the Balkans and Anatolia. Further afield, the picture was even more gloomy. The Ottomans were obliged to retreat from Yemen in the early seventeenth century and control of Iraq was constantly contested by the Safavids. It is not clear exactly how or when effective Ottoman control over provinces like HabeĿ in north-east Africa or al-Hasa (Ott. Lahsa in eastern Arabia, annexed c.1555) came to an end, but the fact that documentation about their affairs in the mühimme defteris tails off after the late sixteenth century indicates that the central government had lost (p.9) the ability or will to intervene there.21 Meanwhile, even the empire’s hold over those provinces which did remain within it was increasingly tenuous. Egypt was controlled by mamluk dynasties with little reference to Istanbul, and in Lebanon and much of Syria the Druze emir Fakhr al-Din Ma‘an (executed 1635) set himself up as a de facto independent ruler. Yet despite the growing political autonomy of these peripheral regions, by culture their elite remained linked to Istanbul and the Ottoman Empire, on which they relied for legitimacy. In 1683 the second Ottoman attempt to capture Vienna failed, followed by the loss of most of Hungary to the Hapsburgs. Yet it was not a universal pattern of retreat. Crete was conquered in 1669, and Russian attempts to intervene in the Crimea were repelled. However, the gradual encroachment of the Austrians and Russians could not be prevented in the long term.22 By the early eighteenth century the border with the Hapsburgs had come as far south as Niš in Serbia and losses to the Russians were confirmed by the Treaty of Küçük Kaynarca (1774), which definitively marked the end of Ottoman dominance in the Black Sea. A long and debilitating war with Iran (1725–45) ended with a peace treaty but no major border alterations. The tendency of rulers in peripheral areas to assert a de facto independence within the framework of the Ottoman Empire was strengthened, with the foundation of local Ottomanised dynasties (supported by ‘localised’ Ottomans) in North Africa.23 Even in the Balkans, relatively close to Istanbul, nominally Ottoman officials like Tepedelenli Ali Pasha of Ioannina could exercise power with little reference to the Sublime Porte. The final phase of Ottoman rule stretches from the early nineteenth century until the First World War. This too has in the past been characterised as a period of decline as the empire continued to lose territory, but is distinguished from the third phase by two main factors, the attempts to modernise and expand the empire. The Ottoman government took persistent, if often faltering, steps to modernise its armed forces, as it increasingly realised that the traditional Janissary system, which had originated in the fourteenth century (even if by now it resembled very little the original elite corps raised by the devĿirme), was ineffective against modern military technology. In some respects, this process had started earlier in the eighteenth century, with Ottoman attempts to modernise their fortifications to meet the challenge of modern artillery. Later, between 1793 and 1807, Selim III introduced a new reformed army, the Nizam-ı Cedid, Page 6 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers which failed ultimately due to Janissary opposition. It was not until 1826 that the Janissaries were finally abolished, followed in 1839 by the beginning of the Tanzimat period of (p.10) reforms inaugurated by the Gülhane Decree that promised equality for all the empire’s citizens, administrative reforms and the modernisation of the army. The effort to modernise continued— with varying degrees of enthusiasm—until the First World War. Accompanying this were ever increasing levels of foreign investment, as the Ottomans tried to use advances in technology to weld together their far-flung empire, with projects such as the Baghdad and Hijaz railways intended to bring its most distant parts more closely under Istanbul’s control and to enhance the sultans’ prestige. Modernity brought technology, but was accompanied by the growth of the nationalism that would destroy the empire and which lingers in many of its successor states as an obstacle to a balanced understanding of the Ottoman past. Inspired by the tumult of the French Revolution and encouraged by hellenophile Romantics like Byron, the struggle for independence broke out in Greece in 1821. As we have noted above, separatist tendencies were nothing new in the Ottoman Empire. In the late eighteenth to nineteenth century alone, Egypt’s autonomy grew to effective independence under the mamluk dynasties and, more famously, the Albanian soldier of fortune Mehmed Ali (Muhammad ‘Ali), while Tepedelenli Ali Pasha held northern Greece and Albania, and Osman Pasvanoğlu controlled much of Bulgaria and Serbia. Yet none of the above sought acceptance as independent rulers. Rather, their legitimacy depended on their positions as governors being recognised by the Sublime Porte. All, whatever their ethnic origins, were Muslims and Turkish speakers who amied to maximise their power within an Islamic, Ottoman context. The Greeks were different. They sought not autonomy under Ottoman sovereignty but outright independence, and derived legitimacy for their movement from the ideals of classical Greece, of Byzantium and of Orthodox Christianity. Comparable sentiments spread among the empire’s other Christian peoples, in Bulgaria, in Serbia and among the Armenians. However, the Arab Muslim population of the empire generally remained loyal till nearly the end, despite the spread of nationalist ideology to a limited extent and attempts at centralisation and Turkification on the part of some governments in Istanbul, notably the Committee of Union and Progress (1908–18)24. Yet the final century of the empire is not a simple story of decline as its frontiers receded, with most of the Balkans being lost by the early twentieth century and a Russian presence in eastern Anatolia from 1885. On the contrary, the vigour with which the Ottomans sought to protect themselves from European encroachment meant that in the late nineteenth to early twentieth centuries the Ottomans’ frontiers started expanding again, as they resurrected long-neglected claims to remote areas that had not witnessed an Ottoman presence since the sixteenth or seventeenth centuries. In eastern Arabia, the Ottomans revived their claims over al-Hasa and sought successfully to bring Kuwait and Qatar under their sovereignty. European competition over Yemen, (p.11) combined with Mehmed Ali’s intervention there, reminded the Ottomans of their own claim, and in 1872 they reoccupied San‘a’. In North Africa, the Ottomans continued to recognise Mehmed Ali’s descendants as hereditary governors of Egypt on the sultan’s behalf, but in Libya, again prompted by the threat of European intervention, they re-established their rule in 1835, which lasted until 1911. This imperial revival was accompanied by a growth of Ottoman influence abroad, not in Europe but in Africa and south-east Asia. The sultans of Zanzibar, who maintained close relations with the Sublime Porte, even visiting Istanbul in person, acknowledged the Ottoman sultan’s suzerainty25 In south-east Asia, Muslim rulers sought Ottoman protection, while some of the

Page 7 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers ulema of Java came to study in Istanbul.26 This late imperial expansion was not universally successful. Despite their nominal sovereignty over most of North Africa (excluding Morocco), by the end of the nineteenth century effective Ottoman control was limited to the main towns of Libya, contested as it was by British and French imperial ambitions and local forces. However, this burst of dynamism in itself presents a corrective to traditional assumptions of Ottoman decline. Indeed, the mere fact that the empire survived four years of war (1914–18) with most of its Asian Arab lands under its control until nearly the end, and the Ottoman garrison at remote Medina holding out until 1919, suggests that it was more resilient and commanded greater loyalties than many of its detractors believed. The Ideology of Ottoman Expansion from Beylik to Empire The Ottomans’ origins in the late thirteenth to early fourteenth centuries were as what is often described as a ‘frontier principality’ (uc beyliği) in north-western Anatolia on the borders of the Byzantine and Islamic worlds. Yet as observed above, what is meant by a frontier with regard to the Middle Ages is complex. Although the concept of finite sovereignty implied by the idea of a frontier may seem obvious today, it was by no means always accepted in the past. Most famously, the Mongol Empire aspired to universal rule under the influence of the Chinese concept of the Mandate of Heaven. This tradition remained current even after its rulers’ conversion to Islam and influenced subsequent Turkish polities, including the Ottomans, as is discussed further below. Even before the Mongols, the concept of a frontier was rather alien to the classical Islamic world. Although the word hadd (pl. hudud), which in modern Arabic means ‘frontier’, is widely attested in the writings of Islamic geographers (though not used at all by the famous Idrisi in the twelfth century), this usually meant a region or even a town rather than a line. Sovereignty was perceived as gradually diminishing the further one travelled from the capital, until it ultimately fizzled out altogether. Even the border (p.12) between the Christian and Islamic worlds was conceptualised in these terms, as a region of negligible sovereignty and constant fighting.27 For the Ottomans, this marchland zone was known as the uc.28 Thus the Ottoman beylik was born in the uc of medieval Bithynia beyond the reach of Byzantine and Mongol authority. The Byzantines held the major cities of the region, such as Bursa (which fell to the Ottomans in 1326), and the Ottoman beylik was initially centred on the rather obscure but strategically located town of Söğüt, to the north of EskiĿehir.29 To the south lay another, much more powerful beylik, the Germiyanids of Kütahya, which would eventually fall to the Ottomans. Medieval Anatolia was covered by a patchwork of these beyliks which had arisen in the vacuum left by the collapse of the Seljuk state in Anatolia in the late thirteenth century and the Mongols’ lack of interest in expansion to the west as their energies were consumed by war with their great rivals, the Golden Horde of the southern Eurasian steppe and the Mamluks of Egypt and Syria. All of these beyliks would eventually be incorporated into the Ottoman Empire, some by force, some by more peaceful means such as marriage alliances, but this process was not completed until the early sixteenth century. In the fourteenth century, however, more established beyliks such as the Karamanids controlled Anatolia’s economic and cultural centres and commanded prestige on the basis of their links with the Seljuk house (or in other cases, such as the Eretnids of Kayseri, their connections with the Mongols), and had diplomatic links as far afield as Venice and Egypt. Claims of the unknown Ottomans to legitimacy in medieval Anatolia would have seemed tenuous at best in comparison, which perhaps helps explain their early concentration on Europe.

Page 8 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers The nature of the early Ottoman beylik has been hotly debated in recent years, largely because there are few early sources for it and almost no archival material. Traditionally, it was argued that Ottoman expansion was motivated by a desire for gaza (holy war) against Christianity and that the early Ottoman sultans portrayed themselves in their titulature as ‘sultan of the gazis (religious warriors)’. This ‘gazi mentality’ has been widely seen as having fundamentally influenced the subsequent course of the empire’s expansion.30 A gazi, motivated by religious ideals, naturally does not accept a clearly delineated frontier with non-Muslims. Rather, he seeks constantly to expand the uc into enemy territory (the Dar al-Harb), while areas that were formerly marchlands become absorbed into the Muslim world, the Dar al-Islam. This probably still represents at least the popular, if not the scholarly, consensus on Ottoman expansion into (p.13) Europe. Earlier scholars, notably Paul Wittek, saw the ‘gazi mentality’ as having survived until the First World War, when the Ottomans’ alliance with the Christian Hapsburgs finally rendered it unsustainable.31 In contrast, Rifa‘at Abou-el-Hajj and Colin Heywood have argued that the idea of the empire as a ‘gazi state’ came to an end with the Treaty of Karlowitz in 1699, when the Ottomans were finally forced to accept joint border commissions and a delineated frontier with the Hapsburgs.32 In any event, the old ideas continued to have a certain resonance. Sultan Mahmud II (1808–39) attempted to use the language of holy war to justify his military reforms and his campaigns against Iran,33 and despite the Ottomans’ Austro-Hungarian and German alliance, the First World War was still proclaimed to be jihad.34 However, the view of the early Ottoman conquests as propelled by Islamic ideals has been challenged by scholars who point to the presence of numerous Christians in the early Ottoman war-bands, and who debate whether gazi really does imply a religiously motivated warrior, at least in the context of medieval Anatolia. In addition, while the main Ottoman thrust was directed against Europe, they were also happy to fight their Muslim neighbours, annexing the minor western Anatolia beylik of Karasi as early as the 1340s, and probably clashing with the Germiyanids even earlier.35 However, for medieval Anatolian rulers there was not necessarily a contradiction between using the ideals of holy war to legitimise themselves as ‘religious warriors’ and in practice having a pragmatic approach towards their non-Muslim subjects and neighbours. For example, among the Anatolian Seljuk sultans, from whom later chroniclers depicted the Ottomans as deriving their authority,36 the twelfth-century sultan Kılıç Arslan II was taunted by his Muslim rival Zangi of Aleppo for being soft on Crusaders and Christians.37 Despite, or perhaps because of this, Kılıç Arslan emphasised his credentials as a religious warrior on inscriptions in his capital city, Konya, proclaiming himself to be ‘the killer of the polytheists’ (i.e. Christians) (qatil al-mushrikin).38 Very similar titles (qahir al-kuffar wa-lmushrikin) were used 250 years later by Mehmed II, who presumably had less reason to be embarrassed about his successes.39 (p.14) Thus early Ottoman proclamations of their gazi credentials, as found in a much debated inscription of 1337 in Bursa,40 need not be taken at face value. Throughout their history, the Ottomans were ready to form alliances with Christian powers, starting with the marriage of the second Ottoman ruler, Orhan son of Osman, to the Byzantine princess Theodora in 1346, and the frequent employment of Ottomans by the Byzantine emperors as mercenaries. Certainly, when Mehmed II finally captured Constantinople, he merely commanded some prominent churches like the Haghia Sophia to be converted into mosques. The Greek patriarchs were granted jurisdiction over their flocks by the sultan, and Christians and Jews as well as Muslims were encouraged to settle in the new capital, despite Ottoman chroniclers renaming it with the pun Islambol, ‘abounding in Islam’.41 Even in the great age of imperial expansion, in the fifteenth and sixteenth centuries, the Ottomans sought alliances with Christian states. France, for Page 9 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers instance, had friendly relations with Süleyman the Magnificent, bound by a common hostility to Austria. Of course, northern Europe was far from the focus of Ottoman expansion in Hungary, but even in Ottoman territories there was only occasionally evidence of a gazi mentality aiming at the conversion of subject peoples. When the latter did convert to Islam, they perhaps usually did so in search of tax privileges, as was often the case with conversion in the pre-modern Islamic world. Nonetheless, the picture was not uniformly rosy. We find instances of eighteenthand nineteenth-century Balkan Muslims asserting their rights over Christians according to a mysterious kanun-i serhad, ‘law of the frontier’, the details of which are as yet unclear.42 If Ottoman expansion was usually in practice pragmatic towards newly subject peoples, this does not mean, however, that the theoretical underpinning for conquest was so. Expansion into both the Muslim and the non-Muslim world was justified in Islamic terms. Chronicles (usually written for court consumption) regularly describe campaigns as gaza, whether dealing with Mehmed II’s fifteenth-century campaigns in Christian eastern Europe or Murad IV’s seventeenth-century operations against the Muslim Safavid Empire of Persia.43 Campaigns against the Safavids were relatively easy to justify, as they were Shi‘is, who could conveniently be regarded as heretical by the Sunni Ottomans. Somewhat more problematic, from a legalistic point of view, were Ottoman hostilities against their fellow Sunnis, the Mamluks who ruled much of the Arab Middle East. Chronicles sought to excuse the destruction in 1516–17 of the Mamluk regime, with whom the Ottomans had been quarrelling for control of southeastern (p.15) Anatolia for much of the later fifteenth century,44 by suggesting that the Mamluks were in league with the Shi‘i Safavids, and anyway had brought about their own destruction by their unfitness for rule45. While religious legitimacy provided a cloak for the Selim I’s expansionist ambitions, it could also be useful in times of failure. In the early seventeenth century, in response to the Ottoman dynasty’s most traumatic crisis since the Battle of Ankara, with numerous revolts in (predominantly Muslim) Anatolia, and a gloomy international situation with Ottoman power challenged in the Balkans by the Hapsburgs and in the Mediterranean by the Venetians, several sultans resorted to emphasising their Islamic credentials both in propaganda and in deeds. Murad IV’s anti-Safavid campaigns were preceded by an emphasis on moral regeneration and the enforcement of sumptuary laws on non-Muslims, while Mehmed IV embarked on an ambitious campaign of forced Islamisation in the Balkans, alongside a programme of enforcing a stricter Islamic morality and sharia law on the existing Muslim population of the empire.46 In a similar way, in the late nineteenth century Abdülhamid II sought to assert his Islamic credentials, backed up by his use of the title of caliph, to rally the Muslim population of the empire around him. The Ottoman claim to the caliphate only dated from the eighteenth century (although the title had been used intermittently in earlier times), when it had been asserted in the face of Russian attempts to claim the allegiance of the Ottomans’ Orthodox population. Although the Ottoman claim came late, it met with some success in attracting the loyalties of non-Ottoman Muslims, as was shown as late as 1924, when Indian Muslims protested against Atatürk’s abolition of the office.47 The Mongol tradition of universal rulership, the other prop of Ottoman expansionist ideology, has been much less studied and remains poorly understood.48 The Ottomans had perhaps only a tenuous claim to being inheritors of the Mongol tradition—their vassals, the Girays, khans of Crimea, had a much stronger one as they were ultimately descendants of Genghis Khan. Nonetheless, until the end the Ottomans used the title khan on their official documents, although sahib-kiran, which also designated claims to universal rulership, stopped being used rather earlier. Ottoman universal claims were also reflected in the forms of address used with

Page 10 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers other rulers: the Ottoman sultans were referred to in terms far more exalted than the addressee, until a degree of equality was imposed on the Ottomans after Karlowitz at (p.16) the Russians’ insistence. Thus while Ottoman attempts to legitimise themselves as universal rulers seem to have faltered after the mid-sixteenth century, elements of the tradition lived on. Furthermore, the Ottoman sultans were also the inheritors of Byzantine imperial traditions: the late Byzantine author George of Trebizond addressed Mehmed II in apocalyptic terms, seeing him as an emperor who would unite the entire world under his rule.49 Behind the rhetoric, the Ottomans had a realistic view of what they called the serhad-i mansura ‘the victorious frontier’ (the term which seems to have usually replaced uc after the fifteenth century). The nature and direction of the serhad-i mansura’s expansion and contraction was dictated less by ideology than by the nature of the threats to Ottoman interests. As Gábor Ágoston discusses in Chapter 3, sixteenth-century Ottoman plans to build a series of canals in Egypt, Russia and Anatolia to facilitate military movements, even if ultimately unsuccessful, reveal the importance of strategy to the Sublime Porte. Likewise, Ottoman intervention in Yemen and the Gulf in the sixteenth and nineteenth centuries can be seen as a direct strategic response to European interest in these areas—first on the part of the Portuguese, later of the British, and to some extent the Italians and the French. The absence of Ottoman rule in these areas for most of the seventeenth and eighteenth centuries coincides with a period in which there was little direct European threat, making it not worth the Sublime Porte’s while to invest in keeping a military presence there when there were greater challenges elsewhere, in central Europe and on the Safavid frontier. However, John Alexander’s contribution (Chapter 11) suggests that the importance of ideology should not be underestimated when considering the Ottomans’ expansion into Africa and their alliances with local Muslims such as the rulers of Harar against Ethiopia. Frontiers, Frontiersmen and the Centre The rhetoric of Ottoman chronicles and the nature of the archival sources—largely documents exchanged between the frontier administration and the centre—perhaps result in a certain tendency to see Ottoman frontiers from the point of view of the central government: territories to be administered, troops to be paid, fortresses to be built, neighbours to be negotiated with or fought. This Istanbul-centred view obscures some of the realities of frontier life. Firstly, while this book is largely concerned with the empire’s frontiers with the outside world, it should be recalled that there were also internal frontiers. The division between northern Albania and Montenegro was one such,50 while in the former Karamanid beylik there was a mountainous frontier between the (p.17) coastal and interior areas. As Burcu Özgüven, Sara Nur Yıldız and Intisar Elzein all point out (Chapters 8, 16 and 19 respectively), the dynamics of some frontier areas reveal continuities stretching back to antiquity. External frontiers too often reflected those of earlier empires: thus those of the Ottoman Empire in western Arabia and Egypt followed closely the frontiers of the Mamluk sultanate at its height. Furthermore, there were often various connections between different frontiers themselves as well as with the centre. Individuals in the sultan’s service might spend their careers moving between disparate frontier provinces. Özdemiroğlu Osman Pasha, for instance, the son of the Circassian mamluk (slave soldier) Özdemir Pasha who conquered Yemen and HabeĿ for the Ottomans, started his career as governor of HabeĿ, was entrusted with command of the campaigns in Caucasia against the Safavids in 1578–90, and ended up as Grand Vizier.51 However, links between disparate frontiers existed on a much more profound level. For instance, Egypt and Iraq in the eighteenth century were governed by mamluks, generally of Georgian origin. Georgia itself was a sensitive region on the Russian–Safavid–Ottoman border, but

Page 11 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers nonetheless these mamluks maintained close links with their families, who would sometimes—at least in the case of Egypt—come out to visit them, and who could expect favours to be done for them in their home in the Caucasus by their powerful relatives.52 Connections between distant frontier provinces might be remembered for centuries, even when physical links had presumably long since ceased. The garrison of Qasr Ibrim, on the Nile south of Aswan, was comprised largely of Janissaries of European origin. Although in theory the corps should have been made up exclusively of youths gathered in the devĿirme, in practice Janissaries were often able to pass on their positions to their descendants, especially from the seventeenth century onwards. This is what seems to have happened at Qasr Ibrim: in the nineteenth century Burckhardt found the Janissaries of Qasr Ibrim had inherited their positions, but still had not assimilated with the local population, remembering their Bosnian and Ottoman roots.53 The displacement or relocation of individuals is a characteristic of every empire. In the Ottoman Empire, whole groups were moved, for various reasons, but often connected with the expansion or contraction of the empire’s frontiers. On the conquest of Egypt, Selim applied the policy of sürgün, forced migration, to take craftsmen of every kind to beautify his capital. In the seventeenth-century Balkans, the southern advance of the Hapsburg frontier and consequent Ottoman reprisals against the Christian (p.18) population prompted many Serbs to flee Kosovo for a new life in Hapsburg lands. Their place was taken by Muslim Albanians, sowing the seeds of conflict over the region in the late twentieth century54 In the nineteenth century the Muslim inhabitants of northwestern Caucasia were expelled by the advancing Russians and forced into Anatolia to serve as agricultural slaves.55 Indeed, the slave trade (discussed further in Part IV, ‘The Economy of the Frontier’) was perhaps one of the most important factors in the displacement of peoples. Ottoman influences could also play a surprisingly important part in the lives of the sultan’s subjects, even in distant areas that were only nominally part of the empire. A good example of this is the nineteenth-century Sudan. Although there had been a brief sixteenth-century Ottoman presence on the Middle Nile, it only re-entered the empire in the nineteenth century when it was conquered by Mehmed Ali, the Ottoman vali (governor) of Egypt. Ottoman sovereignty was thus indirect, for although Mehmed Ali was in theory the sultan’s vassal, Istanbul had no real power to intervene in Egypt. Nonetheless, the Egyptian elite maintained their Ottoman identity, and Turkish was the official administrative language until 1858, and remained in use until the end of the nineteenth century56 As the officers that Mehmed Ali used to conquer the Sudan were part of this Turkish-speaking elite, this period of Egyptian domination (and nominal Ottoman sovereignty) in the Sudan became known as the Turkiyya. However, as the case of the Arabic poet al-Barudi (1839–1904) shows, even for those who started life in this remote province, the Ottoman Empire was still a very real presence. Al-Barudi was born in Dongola, the son of a Circassian army officer. After completing training at the military academy in Cairo, he went to Istanbul, where he served in the Foreign Ministry. After his return to Egypt, he was later sent back to Europe to assist the Ottomans in crushing a rebellion in Crete (1865), and was awarded an Ottoman medal. Alongside his famous Arabic diwan, he also wrote poetry in Turkish. AlBarudi is generally associated with the rise of Egyptian nationalism—he was later exiled to Ceylon by the British—but, evidently, certain affinities tied him to the Ottoman world, as is attested by the Arabic poem he wrote condemning the Cretans’ rebellion against the sultan and celebrating its suppression.57

Page 12 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers (p.19) Frontier Fortifications Pragmatism, as we shall see, may have been the Ottomans’ preferred means of administration, but it could not secure the conquest of new territories. Furthermore, new subjects were by no means necessarily quiescent in their incorporation into the empire. Force was needed to add to the empire, to defend its frontiers, and to suppress revolt. The extent to which it could be applied naturally varied according to the strength of the central administration and the challenges the Ottomans were facing elsewhere. Conquest was also limited by physical constraints. Campaigns were of course constrained by the climate, and by natural barriers— mountains, marshes, deserts. These could be surmounted, but only at great effort and expense. Inevitably, Ottoman conquests concentrated on more accessible areas, which explains the successful Ottoman domination of eastern and central Europe compared to their tenuous hold over the Caucasus. Such considerations applied equally at sea.58 At the same time, however, as Gábor Ágoston shows in Chapter 3, the environment of the frontier was to some extent created by human hands: rivers might be diverted to create protective swamps around fortresses, forests might be cut down to satisfy garrisons’ need for fuel, or as part of a deliberate scorched-earth policy. However, as Palmira Brummett argues in Chapter 2, it was the fortress that was at the centrepiece of the way frontiers were imagined and described. Naturally, there was no single standard of Ottoman fortress design. Often earlier structures were reused (as Malcolm Wagstaff in Chapter 6 suggests may have happened with the fortress of Kelefa in the Mani, and as certainly was the case at Akkerman). In eastern Europe, perhaps the most common type of fortress was known as the palanka (discussed by Burcu Özgüven in Chapter 8, and see also the contributions of Gradeva and Stein in Chapters 17 and 23 respectively). Typically, a palanka was a wooden structure, but the term also seems to have been applied to stone fortresses occasionally. Apart from the comparatively small palankas were fortress cities like Revan (Yerevan in modern Armenia), which were located near a sensitive frontier and the importance of which largely derived from servicing the local garrison. The garrison of a fortress city might be several thousand strong in the sixteenth century59 In Arabia, however, as Petersen and Pringle show in Chapters 4 and 5, the forts designed to guard the hajj route were fairly small, with even the larger ones only accommodating a garrison of a few dozen. Their purpose was not to control the surrounding area but to protect the pilgrims—although the evidence presented by Pringle suggests that such forts were not necessarily particularly effective at doing that, even if it was beyond the Bedouins’ capabilities to capture them. Paying garrisons and constructing and maintaining fortresses represented a massive expense and, where possible, Istanbul attempted to encourage local patrons to take (p.20) responsibility for the construction and maintenance of frontier fortifications—and naturally to bear the costs. It is not surprising, then, that although the Ottomans were well aware of the latest technology in fortress-building, like the bastioned fort that was introduced from the sixteenth century onwards, such forts were only sporadically built in certain key locations, mainly on the border with the Hapsburgs. Budgetary constraints were a constant limit on what was possible, both for the Ottomans and for rival powers.60 A simpler and cheaper option was the restoration and adaptation of earlier structures, such as the modifications at Akkerman to take cannon-platforms in the early eighteenth century. In the eighteenth century there was a programme of fortress restoration in response to the numerous threats the empire faced. Yet, evidently, financial pressures hindered this: it is telling that, as Lucienne Thys-S¸enocak notes (Chapter 9) in her study of Seddülbahir fortress on the Dardanelles guarding the way to the capital, despite the appearance of the Russian navy in this strategically vital region in 1770, the fortifications were not strengthened for another fifteen years. Indeed, in this case, restoration Page 13 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers seems to have been undertaken not because of the evident military threat but because of destruction caused by earthquakes.

The Administration of the Frontier Istanbul did not seek to impose a single type of administration on its provinces (eyalet).61 Traditionally, scholars have seen the provinces in ‘the classical age’ (fifteenth to sixteenth centuries) as falling into two main categories based on the revenue arrangements: timarlı and salyaneli. In the former, which predominated in Europe and Anatolia, lands were assigned to sipahis (cavalrymen) and their commanders (sancakbeyis) in return for military service. The sancaks made up a province (beylerbeyilik) governed by a beylerbeyi. In the salyaneli provinces —most common in the Arab world—revenues were collected by tax farmers (mültezim), and after the costs of the province’s administration and military had been met, the surplus was sent to Istanbul. However, as Gábor Ágoston has pointed out, this masks a rather more complex reality,62 for the Ottomans often preferred to maintain or adapt existing administrative arrangements. In some instances, especially in eastern Anatolia (an area incorporated, as we will recall, comparatively late into the empire), hereditary beys were often kept in place, often the Kurdish or Türkmen chiefs who had ruled there before the Ottoman conquest. At various times the central government attempted to absorb them more (p.21) fully into the empire, not necessarily successfully, and the chapters by Sinclair (Chapter 10, covering the Lake Van region in the sixteenth century), Çetinsaya (Chapter 14, discussing nineteenth-century Iraq) and Blumi (Chapter 15, on late Ottoman Yemen) emphasise how dealing with the tribes, whether Kurdish or Arab, with their fickle loyalties, was an especially sensitive matter for Istanbul. In the hükümet type of province—admittedly usually ones distant from sensitive borders—it was even specified that no Ottoman troops should be stationed there. Istanbul received no revenue from either of the first two types of province. A third type of province was tributepaying (haracgüzar), such as the Principalities of Wallachia and Moldavia, where the local Christian rulers remained in place with substantial autonomy, provided they participated in Ottoman military campaigns, gathered intelligence on the activities of neighbouring states, and protected the frontier, in return for paying tribute (harac or cizye).63 The khans of the Crimea, on the other hand, although also vassals, were the recipients of subsidies from the Ottoman treasury. A fourth type of province was the semi-autonomous kind that predominated in the Arab world. Although as salyaneli provinces they did (in theory) remit income to Istanbul, in practice they often enjoyed wide autonomy. However, such provinces were often an immense drain on Ottoman resources (although not Egypt, which was an extremely profitable venture, at least initially). Places like the Hijaz and Yemen were distant, cost a lot of money in military operations, and did not realise a return for the government. They were kept either for reasons of prestige—control over the holy places of the Hijaz—or to safeguard the empire from external threats, as in the case of Yemen. However, even where typical Ottoman administrative arrangements appear to have been in place—the timar for instance—the situation on the ground was often more complex. Thus although Hungary was a timarlı province, in reality villages might pay tax not just to the Ottomans but also to their Hapsburg overlord.64 This sort of dual administration casts doubt on how meaningful even the delineated frontiers of Hungary (which might anyway be several miles wide) were for either the Hapsburgs or the Ottomans. Beyond the Ottoman heartlands of central and western Anatolia and the Balkans even the use of Ottoman coinage was limited, and depended on the administrative arrangements in place for that province.65 A single administrative policy for the Ottoman frontiers was not feasible, and the empire had to rely on

Page 14 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers local elites, often of different religions, to assist in the administration of these remote areas. Although there were at various times efforts to assert more central control, especially in the nineteenth (p.22) century, compromise and pragmatism were more effective at securing Ottoman sovereignty than force, although this remained an option in extremis. As a result of this diversity, it is difficult to determine even any general principles for frontier administrative arrangements. Thus—to give an extreme example—as Kahraman Şakul discusses in Chapter 13, the Sublime Porte was happy for the Ionian Islands in the early nineteenth century to be run as an autonomous republic under joint Russo-Ottoman protection. Nonetheless, frontier administration was not a complete free-for-all. Even in HabeĿ, the most distant province of the empire, at least in the sixteenth century, governors were appointed from Istanbul, instructions were exchanged between the governor and Istanbul relating to matters ranging from the security of the water supply of the provincial capital, Suakin, the structure of the province’s administration, the provision of munitions and soldiers, and the infiltration of Portuguese spies.66 In practice, of course, with communications between HabeĿ and Istanbul taking a good five to six months,67 the governor must have had considerable discretion, but it was clearly not the case that the further one got from Istanbul, the more minimal the capital’s intervention was. The Danubian Principalities were much closer but seem to have had more autonomy. Likewise, it is perhaps tempting to overestimate the continuities between the preOttoman and Ottoman administration. While these existed in most areas, they did not necessarily endure. Gradeva shows that in Vidin, in the early post-conquest period the administration was modified by the introduction of timars and kadıs, and many Christians served the Ottomans. By the eighteenth century, in the wake of the numerous wars with AustroHungary and its allies, Christians were widely distrusted, and were no longer allowed to live in the fortified heart of the city. It is interesting to note that this hostility to the Christian population seems to have originated among the local Muslim population, who forced their views on a reluctant administration. The impact of the ambitious programme of modernisation embarked upon in the nineteenth century varied extensively from province to province. Again, as Fred Anscombe argues (Chapter 12) in his comparison of pre-Tanzimat Albania with the late nineteenth-century Gulf, pragmatism rather than ideals seems to have held sway. On sensitive frontiers with a religiously mixed population, the reforms aimed at reducing intercommunal tensions and increasing administrative efficiency would be introduced, but elsewhere, in strongly Muslim areas, the Ottomans were prepared to allow local elites to remain in control until the twentieth century. However, as Gökhan Çetinsaya (Chapter 14) shows, even a Muslim area like Iraq could be highly sensitive, due to its substantial Shi‘i population, its proximity to Iran, large tribal population and British competition. Thus, over the nineteenth century, it was gradually but (p.23) slowly transformed with the introduction of the Tanzimat reforms and a moderately successful attempt to settle the tribes.

Frontier Society: Rulers, Ruled and Revolt This flexible approach to administration raises the question of how Ottoman actually were frontier areas of the empire. Here the archives, which are so crucial to our understanding of administration, are less help. They record very little of daily life, so archaeological evidence is thus of paramount importance in illustrating the extent to which Ottoman culture penetrated the borderlands. Ibolya Gerelyes in Chapter 20 shows how ceramics can offer an insight into the links between garrisons and local communities in Hungary, while Carlton and Rushworth in Chapter 21 combine archaeological and ethnographic research to show how a distinctive frontier

Page 15 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers culture developed on the Ottoman–Hapsburg border in Hungary, as is attested by the warrior motifs found on Muslim funeral stelae in the region. Evidence from Akkerman fortress (Chapter 7) on the Black Sea coast of modern Ukraine, in the form of Iznik pottery and Ottoman smoking pipes, indicates a perhaps surprisingly high degree of Ottoman cultural penetration on the edges of the Eurasian steppe. Nonetheless, it remains questionable how representative these influences are, which in all these instances seem to be linked to the nearby presence of garrisons. In contrast, literary evidence from eighteenth-century Çıldır Eyaleti (modern south Georgia and north-eastern Turkey) suggests that the ethnically Georgian hereditary governors—descendants of the Georgian Jaqeli family that had been influential in the area since the twelfth century—spoke Turkish, adopted Ottoman customs, and had long since embraced Islam, but the bulk of the population remained Christian, and spoke Georgian, with little real alteration in their way of life having been brought about by 300 years of Ottoman rule.68 Despite the welcome given to Ottoman forces by much of the local Muslim population on their return to Yemen in the nineteenth century, and the fact that Turkish was introduced and widely used in the bureaucracy and legal system, there is not much evidence of ‘Ottomanisation’ of local elites.69 Of course, societies and individuals could have a dual identity: the eighteenth-century Moldavian prince Dimitri Cantemir is an example of this. Educated in Istanbul, he knew Ottoman, Persian and Arabic, but wrote most of his scholarly works in Latin for a European audience. (p.24) Thus the extent of Ottomanisation of frontier provinces remains a problematic question, on which much further research is needed.70 What cannot be doubted is that the exchange was not merely one way, and frontiers exerted a considerable influence on the Ottoman centre. This came in multiple forms, from the physical presence of slaves in Istanbul to the role played by men from the frontiers in politics. Erik Zürcher has noted, for instance, that most of the leading members of the Committee of Union and Progress that governed Turkey between 1908 and 1919 were from frontier provinces, as were many of the early leading Turkish nationalists. He has suggested that the sensitive nature of the provinces means that debates on identity and the future of the empire started earlier and were more acute there than they were in the capital itself71 The other element of frontier society that concerns us is the nature of relations between rulers and ruled in frontier provinces. Rhoads Murphey in Chapter 18 discusses how even in Anatolia, relations between the soldiery and the inhabitants of a town like Erzurum could swiftly turn sour over the burden of provisioning the troops. As he also shows, the Ottoman authorities did not necessarily seek to rule through force alone, seeking reconciliation with the people of Tabriz and their goodwill. However, threats did not come from the local populace alone. Perhaps one of the most damaging rebellions in the history of the empire was that of Abaza Mehmed Pasha, the Abkhazian commander of Erzurum, the main military base for the Ottomans’ Caucasian and Iranian fronts—a frontiersman both by origin and position. Although such rebel pashas were a persistent problem in Ottoman history, they were perhaps not the greatest challenge to the sultan’s authority. If force against them failed, they could generally be bought off by promotion and gifts, for their aim was usually to enhance their position and prestige, not to overthrow the dynasty. In some ways the Bedouin tribes of Arabia presented a more serious threat to the dynasty’s legitimacy. Inhabiting the desert, their aim was not so much to overthrow Ottoman authority (until the Arab Revolt, at least), but rather to profit both by plundering pilgrims on their way to the holy cities of the Hijaz and by extorting cash

Page 16 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers payments from the Ottoman government. This made controlling Arabia an expensive proposition, for it required substantial investment in fortifications, as Petersen and Pringle show in Chapters 4 and 5. When the Bedouin finally did resort to outright revolt against the empire during the First World War, they seriously damaged the Ottoman war effort in the Middle East by tying down large numbers of Ottoman troops, perhaps more than the regular British army in Palestine did, as Faulkner and Saunders argue in Chapter 22. (p.25) Revolts and tense relationships between rulers and ruled were not, of course, limited to the periphery. The seventeenth-century Celali rebellions were based in Anatolia, and the Patrona Rebellion of 1703 came from the capital itself. Indeed, even when there was no open revolt, one should not necessarily assume that Ottoman authority was universally accepted, even in the Ottoman core of Anatolia. The early Ottoman poet Şeyhi lamented the passing of the Germiyanid beylik in which he had been born,72 but such local patriotism could survive for centuries. In Karaman, for instance, a certain loyalty to, or at the very least interest in, the defunct preOttoman dynasty persisted as late as the eighteenth century, as is attested by the copies of Şikari’s history of the Karamanids made in the period.73 However, it is probably true that on the frontiers of the empire revolts were more dangerous than in the heartland, if only for the simple reason that there were rival powers to which the rebels could resort, such as the Safavids, with whom the Ottomans remained in competition for the loyalty of the tribes of eastern Anatolia and Iraq.

The Economy of the Frontier Frontiers were not merely a drain on the central treasury. They also presented immense opportunities for enrichment both of the state and of individuals. In the European and Anatolian timarlı provinces, after their conquest land was packaged out to the cavalrymen, in return for military service, and a survey of the province’s wealth would be undertaken by the Ottoman state to assess its taxation potential (although, as Sara Nur Yıldız shows with her discussion in Chapter 16 of the annexation of the Karamanid beylik in the fifteenth century, sometimes the reallocation and survey might take place when a vassal prince still theoretically ruled the area, to prepare the way for its conquest). At any rate, both soldiers and the state benefited—in theory, although the costs of administering and fortifying a crucial and hotly contested frontier like Hungary in fact needed to be heavily subsidised, to the tune of half the annual revenue of Egypt, the empire’s most productive province in the sixteenth century.74 Individual soldiers could profit from being stationed to a frontier by engaging in raiding, while the construction of fortresses could often help the local economy. As Mark Stein discusses in Chapter 23, there was a widespread trade in captives for ransoms on the Hungarian–Ottoman frontier. With the opportunities for self-enrichment that military service offered, it is scarcely surprising that Christian troops served the Ottomans too, and civilians claimed membership of the Janissary corps. Frontiers like (p.26) those in North Africa discussed by Colin Heywood in Chaper 25, which simultaneously looked south to the desert and north to the Mediterranean and Europe, illustrate the variety of money-making opportunities available. Caravans from the desert brought slaves for sale in the ports of the North African littoral, while the activities of corsairs caught further captives, the white slaves who were prized in the slave markets of Algiers and Tripoli and might be sold as far away as distant Bornu. Opportunities to speculate in the shipping market might bring a seventeenth-century Englishman to Algiers, and as Heywood notes, there was even a vigorous trade importing wine to Tripoli. Pilgrimage provided another— rather more reputable—way for ports to make money, and the lasting importance of Suakin on the Sudanese Red Sea coast (whose Ottoman history and archaeology are discussed by

Page 17 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers Mallinson et al. in Chapter 24) doubtless derived partly from its position as a transport hub for African Muslims crossing to Jeddah, the port for Mecca and Medina. However, like most ports, it served a variety of functions, and the archaeological evidence confirms that it was part of a trading system reaching as far away as south-east Asia. Indeed, Suakin was a major trading entrepôt not just for the Ottomans but also for the Funj Sultanate of Sinnar, the other main Muslim power in the early-modern Sudan. Nonetheless, slavery in its various forms perhaps represents the quintessential form of frontier economic activity, ranging from the sale of captive soldiers—for a ransom if they were lucky, or at a slave market if they were not—to the seizure of communities en masse described by Lane and Johnson in their study (Chapter 26) of the slave trade in the southernmost edge of the Ottoman Empire, the nineteenth-century Turco-Egyptian Sudan. Tens of thousands were seized each year by slavers like the notorious Zubayr Pasha, to feed the slave markets of Cairo and Istanbul. Despite European pressure for the abolition of the practice, the palace continued to purchase black slaves until the twentieth century, the Ottoman family owning 194 eunuchs as late as 1903.75 Alongside the better-known trade in African slaves and European ones, the trade in East European white slaves was also substantial, and was not restricted to Circassians for the harem. It is hard to assess the consequences of this massive displacement of peoples: perhaps one of the more surprising ones is that as a proportion—gradually decreasing, but still significant in the nineteenth century—of the elite were of slave origins or descent, they were rather more reluctant to abolish the trade, perceiving slavery in a rather different way from their western European counterparts.76 *** The study of the Ottomans’ frontiers raises profound questions about the nature of this state. It obliges us to consider to what extent the Sublime Porte was capable of exerting its power and making its will reality in these often distant and inaccessible areas, giving (p.27) an insight into the effectiveness of Ottoman governance; it raises questions about how the Ottomans coopted local elites and what the latter’s role was during the disintegration of the empire; it forces us to re-examine how the Ottomans interacted with their neighbours, and their motives for and their methods of expansion and conquest. On a more mundane level, we are also obliged to consider what the impact of Ottoman rule was on the peoples of the frontier, to what extent the Ottoman presence influenced their daily lives, and the nature of their relations with the Ottoman garrisons and administration. These are difficult problems, the answers to which will vary according to the period and region studied. The essays in this book do not, therefore, pretend to offer conclusive answers. The study of Ottoman frontiers is in its infancy and the quantity and quality of work (and evidence) varies massively according to region and period. For instance, the Ottomans’ frontiers with the Hapsburgs are quite well known, and have been the subject of numerous monographs and edited collections; the Ottoman frontier with the Safavids and the Russians has been much less studied, especially from an archaeological point of view; while the Ottoman frontier with the Funj is almost entirely unknown. Inevitably, this unevenness is reflected in this volume, although it does present a more comprehensive treatment of the empire’s frontiers than has yet appeared in print. It does not, however, claim to do more than to shed some patches of light on some of the problems outlined above and to point the way to further avenues of research. This may be by the exploitation of as yet neglected archives (as argued for by Wagstaff and Heywood in Chapters 6 and 25), further research in the better-known collections of the BaĿbakanlık Osmanlı ArĿivi in Istanbul, or further archaeological work, or an

Page 18 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers interdisciplinary approach, combining archaeological and textual evidence. It is hoped that this volume will stimulate further interest in and research on this rich but neglected field. Notes: Proceedings of the British Academy 156, 1–27. © The British Academy 2009. (1) See Chapter 15 by Isa Blumi in this volume for the problems caused by inadequate maps in delineating the British–Ottoman frontier in late nineteenth- to early twentieth-century Yemen. (2) See the discussion in Ronnie Ellenblum, Crusader Castles and Modern Histories (Cambridge: Cambridge University Press, 2007), 118–45. (3) For a sampling of recent scholarship, see Daniel Power and Naomi Standen (eds.), Frontiers in Question: Eurasian Borderlands, 700–1700 (Basingstoke: Macmillan, 1999); on Rome and China see, for example, C. R. Whittaker, The Frontiers of the Roman Empire: A Social and Economic Study (Baltimore: Johns Hopkins University Press, 1994); Hugh Elton, The Frontiers of the Roman Empire (Bloomington: Indiana University Press, 1996); Thomas J. Barfield, The Perilous Frontier: Nomadic Empires and China, 221 BC to AD 1757 (Oxford: Blackwell, 1989). (4) Suraiya Faroqhi, The Ottoman Empire and the World Around It (London: I. B. Tauris, 2004), 21. (5) Rudi Mathee, ‘The Safavid-Ottoman frontier: ‘Iraq-i Arab as seen by the Safavids’, in Kemal H. Karpat with Robert W. Zens (eds.), Ottoman Borderlands: Issues, Personalities and Political Changes (Madison: University of Wisconsin Press, 2003), 168 (also published as International Journal of Turkish Studies, 9 (2003)). (6) For examples of recent work on Ottoman archaeology, see (apart from the essays in this volume), Uzi Baram and Lynda Carroll (eds.), A Historical Archaeology of the Ottoman Empire: Breaking New Ground (New York: Kluwer Academic/Plenum Publishers, 2000); Ibolya Gerelyes and Gyöngyi Kovács (eds.), Archaeology of the Ottoman Period in Hungary (Budapest: Hungarian National Museum, 2003). (7) Similar issues are of course well known from archaeology in the west; for an introduction see Dan Hicks and Mary C. Beaudry (eds.), The Cambridge Companion to Historical Archaeology (Cambridge: Cambridge University Press, 2006), esp. 13–33. (8) For a convenient survey of Ottoman history, see Caroline Finkel, Osman’s Dream: The Story of the Ottoman Empire, 1300–1923 (London: John Murray, 2005). (9) See, for example, Rifa‘at ‘Ali Abou-El-Haj, Formation of the Modern State: The Ottoman Empire, Sixteenth to Eighteenth Centuries (Syracuse: Syracuse University Press, 2005); Suraiya Faroqhi, ‘Crisis and change, 1590–1699’, in Halil İnalcık with Donald Quataert (eds.), An Economic and Social History of the Ottoman Empire, 1300–1914 (Cambridge: Cambridge University Press, 1994), 413 ff. (10) For a detailed study of the early Ottomans see Colin Imber, The Ottoman Empire, 1300–1481 (Istanbul: Isis, 1990). (11) On Cem see Nicolas Vatin, Sultan Djem: un prince ottoman dans l’Europe du XV siècle d’après deux oeuvres contemporaines (Ankara: TTK, 1997). Page 19 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers (12) Irène Beldiceanu-Steinherr, ‘Le règne de Selim Ier: tournant dans la vie politique et réligieuse de l’empire ottoman’, Turcica, 6 (1975), 35–48. (13) Heath W. Lowry, The Nature of the Early Ottoman State (Albany, NY: State University of New York Press, 2003), 106–14. (14) Faroqhi, The Ottoman Empire and the World Around It, 38. (15) See Abderrahman el-Moudden, ‘The Sharif and the Padishah: some remarks on MoroccanOttoman relations in the sixteenth century’, in Selim Deringil and Selim Kuneralp (eds.), Studies on Ottoman Diplomatic History, V: The Ottomans and Africa (Istanbul: Isis, 1990), 27–34 and references therein. (16) Salih Özbaran, ‘Ottoman naval policy in the south’, in Metin Kunt and Christine Woodhead (eds.), Süleyman the Magnificent and his Age (London: Longman, 1995), 58–64. (17) B. G. Martin, ‘Maî Idrîs of Bornu and the Ottoman Turks, 1576–78’, International Journal of Middle East Studies, 3 (1972), 470–90. (18) İsmail Hakkı Göksoy, Güneydoğu Asya’da Osmanlı-Türk Tesirleri (Isparta: Fakülte Kitabevi, 2004), 18–56; Anthony Reid, ‘Sixteenth century Turkish influence in western Indonesia’, Journal of Southeast Asian History, 10/3 (1969) (Special Issue, International Trade and Politics in Southeast Asia, 1500–1800), 395–414. (19) See Andrew Hess, ‘The Battle of Lepanto and its place in Mediterranean history’, Past & Present, 57 (1972), 52–73. (20) Rhoads Murphey, Ottoman Warfare, 1500–1700 (New Brunswick: Rutgers University Press, 1999), 15. (21) On Habeş see Cengiz Orhonlu, Osmanlı İmparatorlugu’nun Güney Siyaseti: Habeş Eyaleti (Ankara: TTK, 1996); on al-Hasa see Jon E. Mandaville, ‘The Ottoman province of al-Hasa in the sixteenth and seventeenth centuries’, Journal of the American Oriental Society, 90 (1970), 486– 531. (22) On this period see Virgina H. Aksan, Ottoman Wars 1700–1870: An Empire Besieged (Harlow: Longman, 2007). (23) Most notably in Tripoli and Tunis; Algiers also sought autonomy, although it relied on recruitment to its elite from Anatolia, which gave the sultan greater leverage over it, and the processes at work were rather different. See Tal Shuval, ‘The Ottoman Algerian elite and its ideology’, International Journal of Middle East Studies, 32 (2000), 323–44. (24) Hasan Kayalı, Arabs and Young Turks: Ottomanism, Arabism and Islamism in the Ottoman Empire, 1908–1918 (Berkeley: University of California Press, 1997). (25) Hatice Uğur, Osmanlı Afrikası’nda bir Sultanlik: Zengibar (Istanbul: Küre Yayınları, 2005). (26) Göksoy, Güneydoğu Asya’da, 63 ff. (27) Ralph W. Brauer, ‘Boundaries and frontiers in medieval Muslim geography’, Transactions of the American Philosophical Society, n.s. 85/6 (1995), 1–30. Page 20 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers (28) Colin Heywood, ‘The frontier in Ottoman history: old ideas and new myths’, in Power and Standen (eds.), Frontiers in Question, 233; reprinted in C. Heywood, Writing Ottoman History: Documents and Interpretations (Aldershot: Variorum, 2002). (29) Rudi Paul Lindner, Explorations in Ottoman Prehistory (Ann Arbor: University of Michigan Press, 2007), 35–53. (30) See the discussion in Heywood, ‘The frontier in Ottoman history’, 232–3. (31) Ibid., 241. (32) Ibid., 242; Rifaat A. Abou-el-Haj, ‘The formal closure of the Ottoman frontier in Europe: 1699–1703’, Journal of the American Oriental Society, 89 (1969), 467–75. (33) Finkel, Osman’s Dream, 434. (34) Ibid., 529. (35) Lindner, Explorations, 67–80. (36) e.g. Mehmed NeĿri, Kitab-ı Cihan-nüma (Neşri Tarihi), ed. Faik Reşit Unat and Mehmed Köymen (Ankara: TTK, 1949), 62–5. (37) Michel Balivet, Romanie Byzantine et Pays de Rûm Turc: histoire d’un espace d’imbrication gréco-turque (Istanbul: Isis, 1994), 84. (38) Remzi Duran, Selçuklu Devri Konya Yapı Kitabeleri (İnşa ve Tamir) (Ankara: TTK, 2001), 69; interestingly, Kılıç Arslan II is the only Seljuk ruler who used such a title, according to the (admittedly limited) collection of inscriptions published by Duran (cf. ibid., 15–18). (39) For example, see Akdes Nimat Kurat, Topkapı Sarayı Müzesi Arşivindeki Altın Ordu, Kırım ve Türkistan Hanlarına Ait Yarlık ve Bitkiler (Istanbul: Burhaneddin Matbbası, 1940), 108–9, l. 10. (40) Lowry, The Nature of the Early Ottoman State, 33–44. (41) Finkel, Osman’s Dream, 56–7. (42) See Chapter 17 below, Rossitsa Gradeva, ‘Between hinterland and frontier: Ottoman Vidin, fifteenth to eighteenth centuries’, pp. 344–5, 349. (43) For example, in Kıvami’s Fetihname, ed. Ceyhun Vedat Uygur (Yapı Kredi Yayınları, 2007), virtually all of Mehmed’s campaigns are described as gaza, with the exception of those against Muslim Karaman; for Murad IV see Sıdki PaĿa, Gazavat-ı Sultân Murâd-ı Rabi’ (IV. Murad’ın Revan Seferi), ed. M. Arslan (Istanbul: Kitabevi, 2006), passim. (44) Shai Har-El, Struggle for Domination in the Middle East: The Ottoman-Mamluk War 1485– 1491 (Leiden: Brill, 1995). (45) See, for example, İdris-i Bidlisi, Selim Şah-name, trans. Hicabi Kılangıç (Ankara: Kültür Bakanlıgı, 2001), 303.

Page 21 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers (46) Marc D. Baer, Honored by the Glory of Islam: Conversion and Conquest in Ottoman Europe (New York: Oxford University Press, 2008). (47) Azmi Özcan, Pan-Islamism: Indian Muslims, the Ottomans and Britain (1877–1914) (Leiden: Brill, 1997), 184–204. (48) See the discussion in Cornell H. Fleischer, Bureaucrat and Intellectual in the Ottoman Empire: The Historian Mustafa Âli (1541–1600) (Princeton: Princeton University Press, 1986), 272–92. (49) Nancy Bisaha, Creating East and West: Renaissance Humanists and the Ottoman Turks (Philadelphia: University of Pennsylvania Press, 2006), 153–6. (50) Maurus Reinkowski, ‘Double struggle no income: Ottoman borderlands in northern Albania’, in Karpat and Zens (eds.), Ottoman Borderlands, 240. (51) J. R. Blackburn, ‘Othman Pasha’, EI2, 8: 183–5. (52) Daniel Crecelius and Gotcha Djaparidze, ‘Relations of the Georgian mamluks of Egypt with their homeland in the last decades of the eighteenth century’, Journal of the Economic and Social History of the Orient, 45 (2002), 320–41. (53) J.-L. Burckhardt, Travels in Nubia (London: John Murray, 1819), 34; on Ottoman Qasr Ibrim and its garrison see Martin Hinds and Hamdi Sakkout, Arabic Documents from the Ottoman Period from Qasr Ibrim: Texts from Excavations (London: Egypt Exploration Society, 1986); Martin Hinds and Victor Ménage, Qasr Ibrim in the Ottoman Period: Turkish and Further Arabic Documents (London: Egypt Exploration Society, 1991). (54) See the discussion in Antonina Zhelyazkova, ‘Islamization in the Balkans as an historiographic problem: the southeast-European perspective’, in Fikret Adanır and Suraiya Faroqhi (eds.), The Ottomans and the Balkans: A Discussion of Historiography (Leiden: Brill, 2002), 237–44. (55) Ehud R. Toledano, Slavery and Abolition in the Ottoman Middle East (Seattle: University of Washington Press, 1998), 81–5. (56) Ehud R. Toledano, State and Society in Mid-Nineteenth-Century Egypt (Cambridge: Cambridge University Press, 1990), 157–60. (57) For the Cretan qasida (a nuniyya starting akhadha l-kara bi-ma‘aqid al-ajfani) see Diwan alBarudi, ed. ‘Ali al- Jarim and Muhammad Shafiq Ma‘ruf (Beirut: Dar al-‘Awda, 1998), 643–8, and see the introduction to the same for more details on the poet’s life. (58) Murphey, Ottoman Warfare, 20–5. (59) Ibid., 54–5. (60) Murphey, Ottoman Warfare, 18–19, 53–9, 111–15. (61) Eyalet was the term generally used from the late sixteenth to late nineteenth century; in earlier and later times (post-1865), the word vilayet was used, while beylerbeyilik could also be found in the fifteenth and sixteenth centuries. See Halil İnalcık, ‘Eyalet’, EI2, 3: 721–4. Page 22 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Introduction: The Ottoman Empire and its Frontiers (62) Gábor Ágoston, ‘A flexible empire: authority and its limits on the Ottoman frontiers’, in Karpat and Zens (eds.), Ottoman Borderlands, 17; see also Karpat’s introduction in ibid., 2–3. (63) Viorel Panaite, ‘The Voivodes of the Danubian Principalities as Haracgüzarlar of the Ottoman sultans’, in Karpat and Zens (eds.), Ottoman Borderlands, 59–78. (64) Ágoston, ‘A flexible empire’, 24–5. (65) S¸evket Pamuk, ‘The Ottoman monetary system and frontier territories in Europe: 1600– 1700’, in Karpat and Zens (eds.), Ottoman Borderlands, 175–82. (66) See, for example, the documents published in Orhonlu, HabeĿ Eyaleti, 186, 196, 199, 230. (67) Ibid., 204. (68) M.-F. Brosset, Histoire de la Géorgie (St Petersburg: Académie imperial, 1849), 2: 79. (69) On the Ottomans’ (far from universal) popularity in Yemen and Turkish-speaking local administrators, see Paul Dresch, A History of Modern Yemen (Cambridge: Cambridge University Press, 2000), 4–5; on the lack of an Ottomanised provincial elite, see Thomas Kühn, ‘An imperial borderland as colony: knowledge production and the elaboration of difference in Ottoman Yemen, 1872–1914’, The MIT Electronic Journal of Middle East Studies, 3 (Spring 2003), 14. (70) For the late Ottoman period, see Ehud Toledano, ‘The emergence of Ottoman-local elites 1700–1900: a framework for research’, in Moshe Maoz and Ilan Pappe (eds.), Middle Eastern Politics and Ideas: A History from Within (London: I. B. Tauris, 1998), 145–62. (71) Erik-Jan Zürcher, ‘The Young Turks: children of the borderlands?’ in Karpat and Zens (eds.), Ottoman Borderlands, 275–85. (72) Şeyhi Divanı, ed. Mustafa İsen and Cemâl Kurnaz (Ankara: Akçağ, 1990), 91–3. (73) Pers. comm. Sara Nur Yıldız. See Şikari, Tarih-i al-i Karaman, ed. M. Mesud Koman and M. Ferid Uğur as Şikari’nin Karamanoğulları Tarihi (Konya: Yeni Kitap Basımevi, 1946), 2–3, 8, 16. (74) Murphey, Ottoman Warfare, 58. (75) Toledano, Slavery and Abolition, 12, 47. (76) Ibid., 21–31, 126–34.

Access brought to you by:

Page 23 of 23 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries PALMIRA BRUMMETT

DOI:10.5871/bacad/9780197264423.003.0002

Abstract and Keywords The frontiers of the Ottoman Empire, like frontiers elsewhere in the early modern world, were not defined and represented in terms of linear boundaries. The fortress was possessed space, occupied by the soldiers or subordinates (long-term or temporary) of a sovereign entity. Control of territory and of trade routes was counted, in terms of the submission of fortresses. It is that counting and mapping which this chapter proposes to consider. Early modern fortress images vary from the architecturally correct, complete with keys to various features, to the highly impressionistic, to the simply iconic. Maps also show the rhetorical fortress — an emblem of possession. To illustrate that characteristic, this chapter presents a set of maps of fortresses on the Ottoman-Hapsburg-Venetian frontier. These images suggest the ways in which the fortress served to define Ottoman frontiers in the early modern imagination and to stamp sovereignty onto contested regional space. Keywords:   fortress maps, geography, Ottoman Empire, confrontation sites, Ottoman frontiers, OttomanHapsburg-Venetian frontier

THE FRONTIERS OF THE OTTOMAN EMPIRE, like frontiers elsewhere in the early modern world, were not defined and represented in terms of linear boundaries. They were defined, instead, in terms of physical features, sovereign claims, units of taxation, the reach of armies, and the memories of elderly residents. Within that matrix of defining elements, in the narratives and maps of the sixteenth- and seventeenth-century Eurasian world, the fortress occupied a special position. It was the quintessential marker of frontier space; it marked not so much the edge of empire as points of control in both land and seascapes. The fortress was possessed space, occupied by the soldiers or subordinates (long-term or temporary) of a sovereign entity; it could be designated ‘ours’ or ‘theirs’ and marked accordingly; it could be won or lost. Control of territory and of trade routes was ‘counted’, in terms of the submission of fortresses; and the occupation of fortresses was recorded in the histories and commemorated in the maps as an

Page 1 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries indicator of the success and failure of empires and their commanders. It is that ‘counting’ and mapping which this chapter proposes to consider. Maps were not simply registers of geography in the early modern world; they were (as they are today) a special form of knowledge and communication. Maps of fortresses might function as news (conveying to an audience what purported to be true and accurate images of sieges as they happened), as strategic plans designed for navigational or military purposes, or as artistic representations, with pleasing landscapes or historical vignettes. Early modern fortress images vary from the architecturally ‘correct’, complete with keys to various features, to the highly impressionistic, to the simply iconic.1 But maps also show us the rhetorical fortress—an emblem of possession. (p.32) In almost every case, they suggest owned space, which is somehow labelled or marked with the sign of its ‘owners’ as space which can be enjoyed, defended, exploited and conquered. To illustrate that characteristic, this chapter will present a set of maps of fortresses on the Ottoman–Hapsburg–Venetian frontier. These images suggest the ways in which the fortress served to define Ottoman frontiers in the early modern imagination and to stamp sovereignty onto contested regional space.

Imagining Empire and its Limits The Ottoman Empire was imagined in a variety of ways in the literatures of early modern Europe: as a dynasty inserted into the capital of Constantine; as a Muslim power established on three continents and controlling an empire that spanned thousands of miles and thousands of souls; as a potential trading partner and ally in European wars for political and religious hegemony; and as a monolithic Muslim threat, among other options. Its boundaries were conceptualised as contested or ambiguous—that is, depictions of where Europe ‘ended’ and where the Ottoman Empire ‘began’ were highly variable—but the Ottomans were certainly construed as occupying or pressing into the territories of Christendom. If one assesses those various representations and the ways in which the Ottomans were crafted in space by their contemporaries, one can speak about the Ottoman Empire as an entity measured in terms of a set of land- and sea-based points of encounter, aggression, exchange and defence. Those visible points of encounter were fortresses, variously defined—fixed spaces that changed hands, and possessed multiple and often ambiguous identities. Such fortresses—located at sea ports, near mountain passes, in commercial centres or in expanses of agricultural land—were limit-points. They were mapped as possessed space that ‘belonged’, however tangentially, to sovereigns who were labelled ‘Christian’ and ‘Turk’—primarily the Hapsburg emperor and the Ottoman sultan, but also the Republic of Venice, and a series of petty kings and warlords. ‘Fortress’, of course, is a term which is broadly construed; but I think ‘stronghold’ is a useful definition: a place built of wood, earth and stone (ranging from an outpost to a city) that serves to garrison troops and to defend and celebrate the limits and resources of sovereign space.

Sites of Confrontation Sometimes fortresses were depicted as tranquil sites, devoid of action, embedded in bucolic landscapes. But the mapped fortress was often contested space, the focal point for a series of confrontations and claims in the Ottoman–Hapsburg-Venetian struggle for hegemony along a combined land and sea front reaching from the Adriatic to the Black Sea. News maps depicting battle scenes, especially sieges, became quite popular (p.33) in Italy, for example, in the second half of the sixteenth century and particularly in the aftermath of the land-battle of Szigetvár in 1566 and the sea-battle of Lepanto in 1571.2 German map-makers were active in this trade as well; and both Italian and German maps were disseminated to Holland, France and England. News (in maps and broadsheets) of battle with the ‘Turks’ thus travelled relatively fast, Page 2 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries providing ammunition for political tracts and sermon literature, particularly in the context of the Ottoman–Hapsburg Long War of 1593–1606.3 For European publics and officials, such maps served to notify, to inflame, to educate, to commemorate, and possibly even to enhance Christian solidarity, although that was a difficult task in the context of the Reformation era’s complex, inter-European, political–religious struggles. A German siege map, complete with key, for example, commemorates the defeat of the Ottomans at Hatvan, north-east of Budapest (Figure 2.1).4 It shows the fortified town set into the countryside surrounded by the tents, military units, and bulwarks and entrenchments of defenders and attackers. Those attackers are identified in both key and legend as ‘Christian’. The crosses on their flags also declare that identity5 The stronghold under siege is marked as Ottoman, indicated by a crescent flag hanging from the castle tower. The legend banner across the top of the map suggests the immediacy of events, providing the date of the conflict as 3 September 1596. The key and the animated figures spread across the siege map, firing cannon and riding horses, draw the distant viewer into the conflict, both suggesting an eyewitness presentation of events and allowing the reader to savour the victory. One more point of encounter, one more segment of territory has been reallocated to the ‘Christian’ side. (p.34) (p.35) Another German map, by Alexander Mair, published in Augsburg, also shows the moving of the frontiers (Figure 2.2).6 On this map, yet another ‘Christian’ force captures Iavarinum (Győr or Yanık) on the Danube

Figure 2.1. Defeat of the Ottomans at Hatvan. J. S., ‘Abris der Vöstung Hadtwan’, and Raab rivers in 1598.7 Here the [1596]. British Library, Maps C.7.e.2(.31). besieging army is rather more generic and Copyright British Library Board. All Rights less animated than that in the previous Reserved. siege map. But unlike that previous map, this fortress is presented to the viewer in a somewhat broader geographical context and embedded in a more elaborate set of celebratory Latin and explanatory German texts. An effort is made to depict the contours of fortress architecture with its angled bastions, thereby lending the image a sense of ‘accuracy’.8 A fire, presumably from gunpowder stores blowing, lights up one of the bastions. This is a fairly common figural device, lending further immediacy to the news map’s portrayal. And the fortress is shown situated at the confluence of the Danube and Raab channels, thus signalling its strategic importance. Though the modes of artistic representation are quite different, these siege maps of Hatvan and Győr taken together suggest some standard options for the envisioning of contested space on the Ottoman–Hapsburg frontier at the end of the sixteenth century. Their focus on the individual fortress highlights the notion that the frontiers are measured in the points where armies meet, not in blocks of territory. One does not see a system of defensive sites or a line of military advance; rather one sees sovereign power condensed into a set of walls, bastions, towers and flags. On Mair’s map, the celebratory mood is made more explicit with text as well as iconography. A flag with a large cross has been raised over the wall, denoting Christian identity. Other large banners show the Hapsburg double-headed eagle, and the coat of arms of Hungary. But an even more prominent symbol of possession is located in the foreground, where one finds a cartouche in the form of a monument. It is dedicated to Adolph Schwarzenberg (1547–1600), one of the

Page 3 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries ‘liberators’ of the citadel, and invokes the name of the Hapsburg Rudolph II (Holy Roman Emperor, 1576–1612).9 On this monument, an angel raises two victory wreaths over two ‘classically’ inspired obelisks. Tied beneath her feet are captive Turks, one wearing a turban and another wearing the peaked headgear of the Janissary. Győr had been captured by the Ottoman general Sinan Pasha in September 1594, and the forces of royal Hungary had tried to retake it in 1597 without success. The town was thus emblematic of the advancing Ottoman frontier. Its reconquest in 1598 was a symbol of salvation. That salvation, in turn, became a vision of the frontier—stamped onto a map and circulated to an audience preoccupied with the question of how far the Ottoman armies might go and how far they might be pushed back.10 To seal the message of salvation (and punishment) a Latin passage from the Vulgate serves as a legend across the bottom of the map, and as a reminder of those to whom the ultimate victory must belong. It reads: ‘Just like the fire which burns down the forest, and just like a flame burning up the mountains, thus will you destroy them in your fury, and scatter them utterly in your rage’ (Sicut ignis qui comburit silvam, et sicut flamma comburens montes; ita persequêris illos in tempestate tua, et in ira tua turbabis eos).11 (p.36) (p.37) Almost 100 years later, conventions for mapping the Ottoman–Hapsburg frontier had not radically altered.12 A Venetian map by Giovanni Battista Chiarello (Figure 2.3) shows the 1685 siege of Nayhaysel (Neuhäusel or Uyvar), on the River Neutra (Nitra), in Royal Hungary, two years after the failed battle for Vienna. The map is one of several in Chiarello’s 1687 history of the Hapsburg wars against what he calls ‘rebels

Figure 2.2. The conquest of Győr by the Ottomans. Alexander Mair, ‘Iavarinum sive Raab’, [1598]. British Library, Maps 28225– 5. Copyright British Library Board. All Rights Reserved.

and Ottomans’.13 In the dedication to his book, Chiarello touts the cause of Christendom, and fervently hopes for the return of the Cross to the Orient, which has been subject to ‘the tyranny of Muslim impiety’ under the ‘Turks’.14 In a note to his readers (Cortese Lettore), Chiarello characterises the combatants in these wars: on one side are the Christian potentates and on the other is the ‘universal enemy’, the Ottoman sultan.15 (p.38) The borders between the two, he suggests, are necessarily contested and variable, because ‘the sceptre does not rest securely in the right hand of the monarch unless, together with the sceptre, he also grips the sword.’16 (p.39) In this particular map, one of several in Chiarello’s text, that gripping of the sword is embodied in the struggle for Nayhaysel. Here we see a schematic design of the bastioned fortress surrounded by Figure 2.3. Giovanni Battista Chiarello, Pianta della Fortezza di Nayhaysel, in iconographic units of soldiers. Like the map Historia degl’avvenimenti dell’armi Imperiali of Hatvan, this map includes a key, contro a’ Ribelli, et Ottomani…1687, Folger explaining the action of the siege. It Shakespeare Library, 246080. By Permission indicates points of attack, locations of of the Folger Shakespeare Library. artillery, routes of approach, some military units, waterways, and natural features like hills and swamps. The fortress itself is empty of buildings or people, stripped of its human identity. But that human dimension is embodied in a cartouche celebrating the defeat of the

Page 4 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries ‘Turks’, represented by characters whose facial expressions are visible. Two lie, fallen in battle, at the base of the cartouche (which takes the form of a drape or banner). Another ‘Turk’ figure rubs his head in dismay and points to his comrades. At the top right of the cartouche, not so readily visible, are the heads of two ‘Turks’ mounted on pikes, one head wearing a turban, the other shaved and top-knotted. Just beneath those heads is the smiling face of an unknown figure, pleased apparently at this display of trophies. Chiarello’s fortress, like many other such images, floats in undesignated space. Without the cartouche and key this could be any one of many early modern European fortress images—places where battles were fought against Muslim or Christian foes. But labelled and dated it becomes a specific point of possession, part of the broader frontier and of one sovereign entity or another, now Hapsburg, now Ottoman. It is the text and its decoration of heads, a specific kind of head, that is, which place the fortress in context, giving it a history, and providing it with an identity. Another interesting map-pairing may be found in two images (not shown here), published a century apart, of the fortress of Agria (Erlau, Eger, Ott. Egri) in northern Hungary. One, dated 1568, shows a defensive wall with seven rounded bastions and a central fortress flying the double-headed eagle flag of the Hapsburgs.17 A river flows in front of the fortress and the cartographer notes on a mountain behind the fortress that this elevation is higher than that of the fortress (a strategic consideration). A second map of the same fortress from a century later provides a different view, and more detail.18 This map has a combination of rounded and pointed bastions. The river flows around and also through the city, which is set on two levels with the ‘castle’ on a higher plane. A key at the bottom of the map locates the castle, the river, the ‘city’, the cathedral and the Palace of the Magistrates. Interestingly, in terms of indicators of possession, the tall buildings (including the cathedral and the castle) all have crescents mounted at their tops, marking the city as Ottoman. Turbaned figures ride in the countryside as if to reaffirm that identity and date the map to the period of Ottoman rule (p.40) (1596–1687). From one map to the other, Eger, a significant point in the frontier zone, thus shifts from the Hapsburg to the Ottoman ‘side’. The problematic nature of location and sovereignty in images that may be detached from their texts and from surrounding regional contexts is apparent in a map of the fortress of Nicosia on Cyprus (Figure 2.4) which is contained in a Venetian atlas, the isolario (book of islands) of Simon Pinargenti.19 The isolario, if intact, provides a geographic frame, a progression of sites (ports and islands) in the sea frontier zone between Venice and Istanbul. Looking at the map of Nicosia alone, however, its cartouche unfinished, one could easily assume that it showed a city embedded in the Balkan peninsula, rather than the central point of an island kingdom. Although the fortress remains a focal point for imagining sovereignty, and, just like the fortresses depicted earlier in this essay, it is subject to attack and the transfer of power, the besieging armies must arrive and depart in ships. And so they had arrived, in 1570, landing on the south coast of the island of Cyprus and marching to Nicosia, which fell to the Ottomans after a forty-five-day siege.20 Pinargenti’s 1573 isolario commemorates a series of such battles along the long Ottoman Venetian sea frontier—one island, one fortress, one port at a time—invoking the struggle between ‘Christians’ and ‘Turks’, and the movement between one imperial capital and another, for a readership which saw those distant, contested, sea spaces as part of its own history and destiny. In this version of Pinargenti’s isolario, which like other such atlases, contains a variety of maps by different map-makers, the fortress is generic. Its central space, however, is filled with defenders, and buildings, including the major church, Santa Sofia, which is labelled as if to let

Page 5 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries the reader know that this is indeed Nicosia, as the legend proclaims.21 Outside the fortress wall, yet another battle rages. The fortress is surrounded by Ottoman attackers, identified by the crescents on their tents and banners. The viewer sees the puffs of smoke issuing from cannon which are aimed at the apparently formidable walls. Individual military units are labelled according to type or function (Janissaries, cavalry, archers). In the foreground two groups of figures, sketched out only very roughly, confront each other. Unwilling to leave identity to the imagination of the viewer or to the iconography of crescent flags, the map-maker has labelled one force ‘Turchi’ and one ‘Cristiani.’ Thus the map returns its readers to the notion of religious confrontation. This is not simply a siege taking place in some far-off place, it is a struggle between Christians and Turks. If the Turks succeed, one more point of Christendom will be lost.22 (p.41) (p.42) The church of Santa Sofia and its surrounding buildings appear again, in Giuseppe Rosaccio’s 1598 illustrated narrative of the journey from Venice to Constantinople.23 But in Rosaccio’s map of Nicosia there is no siege and no indication of possession. The fortress presents a tranquil cityscape in a bucolic countryside, and the Ottomans are nowhere in evidence, despite their having been in command of the

Figure 2.4. Simon Pinargenti, Nicosia, in Isole che son da Venetia nella Dalmatia…, 1573. British Library, Maps C.24.g.10(.42). Copyright British Library Board. All Rights Reserved.

city since 1570 and of the island since 1571 (Figure 2.5).24 No crescents deck the walls or buildings. Thus, who in fact possesses this fortified site is unclear or even hidden. Nicosia might be one timeless space in the chain of cities the traveller visited or witnessed on the long journey from Venice to Istanbul. As in Pinargenti’s map of Nicosia, there is no indication that this fortress is located on an island, one point of encounter in a chain of such points defining the Mediterranean sea frontier contested by Porte and Signoria.25 In fact, in only one of his series of maps does Rosaccio suggest, by means of a dotted line, the broader sea-based frontiers (confini) dividing his opposing sides, the same ‘sides’ advanced by Chiarello: ‘Turks’ and ‘Christian princes’. The reader must look beyond the individual maps to the author’s narrative descriptions of place to find a more elaborate delineation of which spaces belong to whom. In the full title of his book, Viaggio da Venetia, a Costantinopoli per Mare, e per Terra, & insieme quello di Terra Santa. Da Gioseppe Rosaccio con Brevita Descritto nel quale, oltre à Settantadui Disegni, di Geografia, e Corografia si discorre, quanto in esso Viaggio, si ritrova. Cioè Città, Castelli, Porti, Golfi, Isole, Monti, Fiumi, é Mari, Opera utile, à Mercanti Marinari, & à Studiosi di Geografia, Rosaccio notes the units by which the early modern space is measured: cities, castles, ports, gulfs, islands, mountains, rivers and seas. The ‘castles’ and ‘cities’ are the equivalent of our fortresses, the sites of occupation which (along with routes of travel and trade) determine sovereign possession, especially in the broad frontier zones, as opposed to sovereignty, which was claimed in terms of blocks of territory (or in units of taxable land).26 (p.43)

Page 6 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries (p.44) For a fortress placed in broader context, demonstrating a sea frontier, we have the Venetian cosmographer, Vincenzo Coronelli’s late seventeenth-century map of another space in the contested Ottoman– Figure 2.5. Giuseppe Rosaccio, Nicosia, in Viaggio da Venetia, a Costantinopoli per Venetian frontier zone: Santa Maura (on the Mare…1598. British Library, Maps C.27.b. Ionian island of Levkas). In this map, which 26(.46). Copyright British Library Board. All zooms out from the fortress and the island Rights Reserved. itself to show the surrounding seas, the viewer gets a sense of geographic context without losing the sense of past or impending military confrontation (Figure 2.6). The island fills most of the map space, its connection to the Ionian mainland (along with the location of the fortress) indicated in the upper right-hand corner of the map. Each of the other three corners is occupied by a cartouche. One elaborate cartouche, in the upper left-hand corner, presents the fortress of Santa Maura (Figure 2.7).27 It includes a fine little schematic of the walls and fortifications and a decorative owl bearing a scale of passi Veneti to indicate the size of the fortress in paces or steps. This inset emphasises the location of the fortress (at the northern end of the island) with the lagoon on one side and the mainland on the other. The map is undated but presumably was produced to celebrate Venice’s taking of the island from the Ottomans in 1684. That conquest is indicated, in rather grizzly fashion, by the legend cartouche in the lower right-hand corner of the map (Figure 2.8). The legend is emblazoned upon the skinned body of a ‘Turk’ held in the teeth of the winged lion of St Mark, mascot of Venice. That ‘Turk’ identity is signified by the figure’s shaved head and topknot and by the crescent upon which the lion rests his foot. Coronelli dedicated the map to Matteo Sanuto, procuratore of Venice. But no doubt he also had in mind Marcantonio Bragadino, commander of Famagusta on Cyprus when it was forced to surrender to the Ottomans in 1571. At that time, the victorious Mustafa Pasha had Bragadino flayed alive and his skin sent to the sultan.28 That humiliation rankled—even 100 years later—thus, through this cartouche, Coronelli gained a bit of vengeance for his Venetian audience. Such cartouches on early modern European maps, whether the relatively mild depiction of dominance included in Mair’s Győr, the understated but grisly image of defeat in Chiarello’s map of Nayhaysel, or the sly and vicious ‘parchment’ in Coronelli’s Santa Maura, framed the fortress in the context of a ‘universal’ struggle between the representatives of Christian and Muslim kings. Just as these fortress images simplified architectural, geographical, and military realities, so their rhetorics of possession might reduce the complex cultural, political and ethno-religious realities of the broad land and sea frontiers to a contest between ‘Christians’ and ‘Turks’.29 (p.45) (p.46)

Figure 2.6. Vincenzo Coronelli, Isola di Santa Maura. British Library, Maps C.27.g. 16(.61). Copyright British Library Board. All Rights Reserved.

Page 7 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries (p.47) The notion that the Ottoman–Hapsburg or Ottoman–Venetian frontiers were neatly divided between Christians and Turks is, of course, just as simplistic as the notion that (p.48) they were neatly divided into land and sea frontiers. One might, rather, divide frontiers into those that were subject to attacks mounted overland and those that were subject to attack via sea-based (or seatransported) forces. In the island–coast zones of the Adriatic and Aegean, forces mobilised at sea often penetrated inland, assisted by local forces. Such was the case in the years 1645 to 1648 when the fleet of the Venetian commander, Leonardo Foscolo, conducted a series of raids along the

Figure 2.7. Vincenzo Coronelli, Isola di Santa Maura, inset, fortress detail. British Library, Maps C.27.g.16(.61). Copyright British Library Board. All Rights Reserved.

Figure 2.8. Vincenzo Coronelli, Isola di Santa Maura, inset, title cartouche. British Library, Maps C.27.g.16(.61). Copyright British Library Board. All Rights Reserved.

Adriatic coast.30 That campaign culminated in the Venetian conquest of Clissa (Klis) a few miles south-east of Spalato (Split). The conquest is commemorated in a series of coloured maps (assembled in an ‘atlas’ dedicated to the Molina family) which portray siege warfare in a rather more direct and personal manner than that envisioned in the maps treated so far. One such image, showing more realistic people (that is, those drawn as two-dimensional individuals, rather than as iconographic military units) attacking Clissa, combines a variety of artistic and cartographic techniques (Figure 2.9).31 The fortress, parts of which are labelled in the key, is shown in profile mounted on a hill while the Venetian forces fire upon it with their batteries. Crescents mark the banners and the mosque of the town, clearly indicating whose territory is under siege. The key notes the location of the mosque. Scattered on the ground in front of the advancing forces are what look like books and two lanterns (or receptacles) with crescents on top. It is unclear what these objects represent; they are marked for the key with the letter ‘G’, but I am unable to make out the identification. Even without identification, however, their crescents mark them as iconic representations of Muslim faith and of the Ottoman polity. Thus the map designates this assault not simply as a victory, but as a victory over Islam and its representatives. Although the Ottomans controlled the bulk of the Balkan peninsula for well over a hundred years, location near the sea made fortresses vulnerable to the fleets of attackers on either side. It was far easier to offload cannons from ships than it was to haul them overland from Istanbul, Venice or Vienna. Once taken, Clissa would be garrisoned by its conquerors who would then withdraw back to their ships.32 For the moment, however, the siege of Clissa is frozen in time in this elaborately decorated image, right down to the sight lines showing the trajectory of balls hurled from the attackers’ cannon. While the crescents of the defenders still stand defiant atop the major edifices of the fortress, the message here is that they will not (or did not) endure for long.33 (p.49)

Page 8 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries (p.50) The images treated so far are a small sample of the options for mapping Ottoman frontiers and the fortresses that stood as markers of sovereign space in the sixteenth and seventeenth centuries. Such Figure 2.9. Clissa, ‘Molina Family Atlas’, [mid-seventeenth century]. British Library, maps, I would argue, not only recorded K.Top. 78.31.b (Table 6, no. 5). Copyright events but embodied a sense of possession, British Library Board. All Rights Reserved. making it real, personal and palpable. This was especially the case when fortress maps were accompanied by surrounding texts and narratives which provided historical vignettes and biblical allusions, highlighted points of contact and conflict, and elaborated upon the divisions between Christian and ‘Turk’.34 Possession was visualised in terms of these contested, conquerable points of urban space. Sovereign territory was, of course, also measured in routes, and stages of a journey through the frontier. Thus the isolario of Pinargenti and the Viaggio of Rosaccio charted movement from port to island to port in the journey from the lands of the ‘Christians’ to the lands of the ‘Turk’, in the process incorporating their fortresses into a larger matrix of travel, trade and imperial reach.

Ottoman Mappings Ottoman maps also envisioned space in terms of movement from port to port, or fortress to fortress, particularly in the narration of campaigns. Ottoman campaign ‘maps’, showing the stages of the campaign journey, took different forms in such narrations, which were designed to glorify the Ottoman sultans and to demonstrate their entitlement to conquered territories. For example, Lokman ibn Seyyid Hüseyin (d. 1601/2), the Ottoman court panegyrist, in his Book of Accomplishments (Hünername), shows the Ottoman sultan Süleyman the Magnificent moving through the Balkans from one fortress to another, at each one demanding and receiving acts of submission from his vassals (old or new).35 In one such image the young prince of (p.51) Transylvania, John Sigismund Zapolya, is brought before the sultan in 1541, after Süleyman had redeemed Buda from an assault by the Hapsburg King Ferdinand I (Figure 2.10). The child’s father, John Zapolya, had controlled part of Hungary as an Ottoman vassal, and the sultan here accepts the son’s submission and grants his family’s claim to tributary status in the principality of Transylvania.36 What is interesting for our purposes, however, is not the complex family dynamics of the struggle for Balkan lands but the miniature’s image of the fortified town, with the sultan’s tents and cannons sprawled before it, acting as backdrop for the certification of sovereignty37 Süleyman’s cannon are still ‘aimed’ at the city walls, and Buda itself bristles with cannon, a threatened and contested space. But, at least temporarily, the struggle has ceased.38 Muezzins give the call to prayer from a minaret within the city, confirming the sense of Ottoman possession and identity. The soldiers of the sultan, and the inhabitants of the town (at left centre) look on—witnesses to Süleyman’s hegemony. It is at such stopping places, the fortresses within the frontier zone, that the main action of the story of Ottoman expansion takes place. Sovereignty is thus mapped, through miniatures, across the Balkans, from Istanbul all the way, or almost, to Vienna. And it is that declaration of sovereignty, rather than Buda’s shape, location, or the nature of its defences, which is important in these representations. Like those in many European maps, the fortress images in the Hünername (despite their different artistic styles) tend to emphasise strength or weakness and the exchange of possession rather than the specifics of fortification or defence.39

Page 9 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries (p.52) (p.53) Another type of campaign mapping was produced in the elaborately illustrated works of Matrakçı Nasuh (d. 1564), an Ottoman pasha who accompanied Sultan Süleyman on his Baghdad campaign in the 1530s, and participated in later campaigns on the European front.40 Matrakçı was particularly concerned to show the stages of the campaign journey; in fact his work is often referred to by the simple shortened

Figure 2.10. The submission of Prince John Sigismund Zapolya of Transylvania to Sultan Süleyman I. Lokman ibn Seyyid Hüseyin, Hünername, Topkapı Sarayı Müzesi, MS Hazine 1524, fol. 266a. Courtesy of T. C. Topkapı Sarayı Müzesi.

title, Menazil (stages).41 Unlike the miniatures in Lokman’s text, however, Matrakçı’s maps are devoid of people. 42 They show the phases of the journey in terms of cities, fortresses, shrines, wells, palaces, roads, rivers and mountain passes. It is left to the surrounding text to provide the cast of characters, the celebration of Ottoman power, and the assertion of territorial entitlement and possession. Matrakçı’s fortresses are set in the surrounding countryside and directly linked to the routes of passage from one stopping place to another. For example, one of his illustrations—and the only one in the Menazil which bears a legend denoting a frontier (serhad)—shows the border zone between Arab and Persian Iraq, and between the Ottoman and Safavid empires (Figure 2.11).43 This ‘border’ is not imagined as a line but as a claimed territorial space delimited by physical features and by edifices: the fortress (kale) of Yeni Imam at the top, Kasr-ı Şirin—a ‘sweet’ summer palace or castle—in the middle section, and the town of Hanekiyye, in the bottom section of the map, which is marked with the legend box denoting its name as well as the serhad designation. These are the spaces past which Süleyman’s armies marched on the way to Baghdad and back, over the roads and rivers and through the mountain passes. Matrakçı shows each way station in the journey as part of the larger whole of empire, linked to the imperial capital by the systems of conquest and of artistic patronage in which he himself was an active participant.44 (p.54) (p.55) As we can see from these few examples, the options for mapping possessed and contested space on the frontiers of the Ottoman Empire in the early modern era varied according to cultural and Figure 2.11. The Ottoman–Safavid border. Matrakçı Nasuh, Menazil-i Sefer-i ‘Irakeyn, artistic norms, individual artists, the texts in İstanbul Universite Kütüphanesi, MSS which the maps were embedded (if they TY5964, fol. 42b. Courtesy of T. C. İstanbul were embedded in texts at all and not Üniversite Kütüphanesi. distributed as broadsheets), and the objectives of the maps’ makers and their audiences. But the fortress was the centrepiece, in the early modern imagination, for the marking of frontiers and possession. It is the point of battle, the meeting place of Christendom and Islam, the way station for travellers, the scene of ceremonies of transfer, submission, defeat or acquiescence. Mapped, it may be the site of the action or a tranquil edifice in designated or

Page 10 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries undesignated space, its history (and predictions of its future) supplied by the imagination of the viewer and suggested by the iconic symbols and legends of possession. Notes: Proceedings of the British Academy 156, 31–55. © The British Academy 2009. (1) Many maps of the era advertise themselves in their legends as ‘true and accurate’ representations. Fortresses were depicted with greater or lesser degrees of architectural accuracy (e.g. towers, numbers of bastions, major edifices inside), so ‘correct’ here is a relative term. But it is important to note the devices (direct and indirect) used to suggest ‘accuracy’ to a distant audience of map consumers. (2) Brendan Dooley, ‘The wages of war: battles, prints and entrepreneurs in late seventeenthcentury Venice’, Word and Image, 17 (2001), 7–24; and Mario Infelice, ‘The war, the news, and the curious: military gazettes in Italy’, in Brendan Dooley and Sabrina Baron (eds.), The Politics of Information in Early Modern Europe (London: Routledge, 2001), 216–35. Also on the function and dissemination of maps, see George Tolias, ‘Nikolaos Sophiano’s Totius Graeciae Descriptio: the resources, diffusion, and function of a sixteenth century antiquarian map of Greece’, Imago Mundi, 58 (2006), 150–82. On Lepanto, see Andrew Hess, ‘The Battle of Lepanto and its place in Mediterranean history’, Past & Present, 57 (1972), 53–73. See also Halil Inalcık, ‘Lepanto in the Ottoman documents’, in Gino Benzoni (ed.), Il Mediterraneo nella seconda metà del ’500 alla luce di Lepanto (Florence: Leo S. Olschki, 1974), 185–92; and John Guilmartin, Gunpowder and Galleys: Changing Technology and Mediterranean Warfare at Sea in the Sixteenth Century (Cambridge: Cambridge University Press, 1980), 221–52. (3) On the Ottoman–Hapsburg wars, see Caroline Finkel, The Administration of Warfare: The Ottoman Military Campaigns in Hungary, 1593–1606 (Vienna: VWGÖ, 1988). See also Ferenc Szakály, ‘The early Ottoman period, including Royal Hungary, 1526–1606’, in Peter Sugar et al. (eds.), A History of Hungary (Bloomington: Indiana University Press, 1994), 83–99; Géza Dávid and Pál Fodor (eds.), Hungarian-Ottoman Military and Diplomatic Relations in the Age of Süleyman the Magnificent (Budapest: ELTE, 1994); Géza Dávid and Pál Fodor (eds.), Ottomans, Hungarians, and Habsburgs in Central Europe: The Military Confines in the Era of Ottoman Conquest (Leiden: Brill, 2000). Virginia Aksan, Ottomans and Europeans: Contacts and Conflicts (Istanbul: Isis, 2004), 154–5, characterises the ‘effective fighting force of palace troops in this period’ (1593–1606) as about 50,000 infantry and cavalry. It is often difficult to tell exactly how fast such news and images travelled. Many maps and broadsheets are undated. But Venice, for example, received news of Balkan sieges via couriers from cities like Ragusa/Dubrovnik and agents posted on the battlefields. (4) J. S., ‘Abris der Vöstung Hadtwan, Von den Christen Belegert und Eröbert. Den. 2 Septemb: A: 1596’, 1596. British Library, Maps. C.7.e.2(.31). The key marks the encampments of the general and commanders; rivers, roads, moat and entrenchments, and the attackers’ gun emplacements with cannon and mortars. (5) On mapping the emblems of ‘Christian’ and ‘Turk’, see Palmira Brummett, ‘“Turks” and “Christians”: the iconography of possession in the depiction of the Ottoman–Venetian-Hapsburg frontiers, 1550–1689’, in Matthew Dimmock and Andrew Hadfield (eds.), The Religions of the Book: Co-existence and Conflict, 1400–1660 (Basingstoke: Palgrave, 2008), 110–39.

Page 11 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries (6) Alexander Mair, ‘Iavarinum sive Raab a Christianis captum 29 die Martij Anno Christi 1598’, 1598. British Library, Maps 28225–5. (7) See Szakály, ‘The early Ottoman period’, 96–7. (8) Of course, many such representations are formulaic or iconic rather than expressive of the individual fortress’s actual lines. Generic models were often used for fortresses and any given map image might or might not reflect a sense of that fortress’s actual site, geographic context and fortifications. Szakály, ‘The early Ottoman period’, 97, gives the date of the conquest as 19 March, whereas the map legend says 29 March. For a series of images of fortresses and a treatment of siege warfare in this area, see Christopher Duffy, Siege Warfare: The Fortress in the Early Modern World (New York: Barnes and Noble, 1979), 191–219. (9) In the upper left, a smaller legend offers the map to ‘Mariae Fuggerae’, possibly a patron of the map-maker. (10) That, of course, was a question for the Ottomans as well, who had to keep an eye on the Safavid front in the east, where warfare was soon to resume. (11) Vulgate, 82:15–16. Another legend ribbon at the top of the map, the relevance of which is more obscure, cites a passage from 1 Samuel 3:11. (12) Indeed, the plates for many such fortress plans were recycled or copied over the years. This is not to say that late seventeenth-century maps did not benefit from technical improvements in measurement and production. Rather it is to say that many of the conventions of representation remained in place or were employed together with newer conventions (such as more formal and exacting demarcation of state borders). Designs of fortifications, whether in the sixteenth or seventeenth century, might be exacting or fanciful. See, for two examples, Cartographic Treasures of the Newberry Library (Chicago: Newberry Library, 2001), 44–5, plates 35 and 36, showing an early news map of the Battle of Pavia in 1525, and one of Alexander de Groote’s fortification designs with imagined battle from 1617. (13) Giovanni Battista Chiarello, Historia degl’avvenimenti dell’armi Imperiali contro a’ Ribelli, et Ottomani, Confederationi, e Trattati sequiti frà le Potenze di Cesare, Polonia, Venetia, e Moscovia, Negotiati, & Aleanze del Conte Tekely con la Porta Ottomana. Accampamenti, Guerre, Assedij, Piazze, e Conquiste di Città, e Provincie. Battaglie, Rotte, e Vittorie variamente successe nelle quatro Campagne degl’Anni 1683, 1684, 1685, 1686 (Venice: Presso Steffani Curti. Con Licenza de’ Superiori, e Privilegio, 1687). Map: ‘Spiegatione dell’Assedio di Nayhay sel Seguito l’Anno MDCLXXXV’, located after page 346. Folger Shakespeare Library, 246080. The first siege of Vienna was in 1529. (14) Chiarello, Historia, unnumbered page, first dedication. (15) Ibid., second dedication. (16) Chiarello, Historia, unnumbered section entitled ‘Notitie Historiche e Geografiche del Regno dell’Hungaria, Schiavonia, e Croatia’, which is located after the first map, ‘Hungaria Millitaria’, in the front matter. (17) L. P., ‘Agria, fortezza nel paese di Ongheria nel modo che a presente si trova, 1568’, British Library K.Top. 110.112. Defenders had fought off an Ottoman siege in 1552. Page 12 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries (18) Agria, [1683?], British Library K.Top. 110.113. The Ottomans took the fortress in 1596 and surrendered it in 1687. (19) Simon Pinargenti, Isole che son da Venetia nella Dalmatia, et per tutto l’Archipelago, fino a Costantinopoli, con le loro Fortezze e con le terre piu notabili di Dalmatia, nuovamente poste in disegno a beneficio de gli Studiosi di Geografia (Vinegia: Appresso Simon Pinargenti et compagni, 1573). British Library, Maps C.24.g.10(.42). (20) John Julius Norwich, Venice: The Greatness and the Fall (London: Allen Lane, 1981), 215–17. (21) The Leventis Foundation on Nicosia has published various treatments of the fortress (unseen by this author), for example, G. M. Perbellini, The Fortress of Nicosia: Prototype of European Military Architecture (Nicosia: Anastasios G. Leventis Foundation, 1994). (22) Of course, the real ethno-religious identities of the participants in such struggles along the long Ottoman frontiers were quite complex. Conversions and intermarriage were common; and those living in frontier areas were often cited as people of questionable loyalties. See, for example, the Ottoman raconteur Evliya Çelebi’s often snide comments as he travels through the southern Balkan peninsula: Evliya, Seyahatname, 8: 302, 304, 324, and Evliya Çelebi in Albania and Adjacent Regions (Kosovo, Montenegro, Ohrid: The Relevant Sections of the Seyahatname), ed. and trans. Robert Dankoff and Robert Elsie (Leiden: Brill, 2000), 85, 93, 191. Identities were also malleable over time, space and situation, depending on social, economic and political variables. Even the designations ‘Christian’ and ‘Turk’ suggest the lumping of many peoples into simple (if not homogenising) categories. (23) See Gioseppe Rosaccio, Viaggio da Venetia, a Costantinopoli per Mare, e per Terra, & insieme quello di Terra Santa. Da Gioseppe Rosaccio con Brevita Descritto nel quale, oltre à Settantadui Disegni, di Geografia, e Corografia si discorre, quanto in esso Viaggio, si ritrova. Cioè Città, Castelli, Porti, Golfi, Isole, Monti, Fiumi, é Mari, Opera utile, à Mercanti Marinari, & à Studiosi di Geografia ([Seal:] Con Privileggio. Venice: Appresso Giacomo Franco, 1598). Such narratives of the stages of travel, often illustrated with maps, were a particular genre of the day, especially detailing the land and sea journeys from Vienna or Venice to Istanbul or Jerusalem. (24) Rosaccio, Viaggio, ‘Nicosia’, British Library, Maps C.27.b.26(.46). Note that there are different editions of Rosaccio’s Viaggio; as with other similar narratives and isolarii, the number and order of the maps may be different from version to version, a function both of multiple editions and of volumes compiled to the specifications of customers. (25) On the Ottoman–Venetian sea frontiers, see Maria Pia Pedani, Dalla frontiera al confine, Quaderni di Studi Arabi, Studi e Testi, 5 (Venice: Università Ca’ Foscari di Venezia and Herder Editrice, 2002), esp. 39–51, on fortresses. See also, for details on coastal fortifications in North Africa, Neji Djelloul, Les Fortifications côtières ottomanes de la Régence de Tunis (XVIe–XIXe Siècles) (Zaghouan: FTERSI, 1995). (26) As Gábor Ágoston, ‘A flexible empire: authority and its limits on the Ottoman frontiers’, in Kemal H. Karpat with Robert W. Zens (eds.), Ottoman Borderlands: Issues, Personalities, and Political Changes (Madison: University of Wisconsin Press, 2003), esp. 23–7, has demonstrated, the Ottoman and Hapsburg sovereigns or their underlings could and did tax the same units of space in the frontier zone, causing significant complaints among the populace. Rural populations

Page 13 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries in such frontier zones also suffered plundering by the garrisons from border fortresses on both ‘sides’. (27) Vincenzo Coronelli, ‘Fortezza di S. Maura’, [1695], British Library, Maps, C.27.g.16(.61). (28) Norwich, Venice, 220–1. (29) On some of the complexities of frontier zones, see: Colin Heywood, ‘The frontier in Ottoman history: old ideas and new myths’, in Daniel Power and Naomi Standen (eds.), Frontiers in Question: Eurasian Borderlands 700–1700 (Basingstoke: Macmillan, 1999), 228–50, reprinted in C. Heywood, Writing Ottoman History: Documents and Interpretations (Aldershot: Variorum, 2002); Colin Imber, Keiko Kiyotake and Rhoads Murphey (eds.), Frontiers of Ottoman Studies, II: State, Province, and the West (London: I. B. Tauris, 2005); Géza Pálffy, ‘The origins and development of the border defense system against the Ottoman Empire in Hungary (up to the early eighteenth century)’, in Géza Dávid and Pál Fodor (eds.), Ottomans, Hungarians, and Habsburgs in Central Europe: The Military Confines in the Era of Ottoman Conquest (Leiden: Brill, 2000), 3–70. (30) Norwich, Venice, 300. (31) ‘Clissa’, The Molina Family Atlas [mid-seventeenth century]. British Library, K.Top. 78.31.b (Table 6, no. 5). The coat of arms on this map, repeated elsewhere in the atlas, would seem to be the coat of arms of the Molina family, whose participation or at least interest in the siege is reflected in the atlas itself. (32) One could say that one of the ways that Ottomans ‘mapped’ space was through a counting of garrison troops, the personnel whereby fortresses were defended and maintained. For one such example, of a ‘roll-call’ register, see Ottoman Garrisons on the Middle Danube: Based on Austrian National Library MS MXT 562 of 956/1549–1550, ed. and trans. Asparuch Velkov and Evgeniy Radushev (Budapest: Akadémiai Kiadó, 1996), 12–27. Such registers were a counterpart of the tahrir surveys for tax purposes, both avenues for managing conquered space. (33) The map and others in this atlas give the impression of sketches from an eyewitness artist. That impression, however, may be an illusion. (34) This is true even when the maps have been detached from their texts. (35) See on Lokman, H. Sohrweide, ‘Lukman b. Sayyid Husayn’, EI2, 5: 813–14. On the Hünername, see Metin And, Turkish Miniature Painting: The Ottoman Period (Istanbul: Dost, 1987), 32, 105–10, 114. Most of its images are attributed to Üstad Osman and his workshop. See also Selmin Kangal (ed.), The Sultan’s Portrait: Picturing the House of Osman (Istanbul: İşbank, 1999); and Esin Atıl, ‘The image of Süleymân in Ottoman art’, in Halil İnalcık and Cemal Kafadar (eds.), Süleymân the Second and His Time (Istanbul: Isis Press, 1993), 333–41. On the rhetorics of such acts of submission, see Palmira Brummett, ‘A kiss is just a kiss: rituals of submission along the east–west divide’, in Matthew Birchwood and Matthew Dimmock (eds.), Cultural Encounters Between East and West: 1453–1699 (Newcastle-upon-Tyne: Cambridge Scholars Press, 2005), 107–31; and Rhoads Murphey, ‘The cultural and political meaning of Ottoman rituals of welcome: a text-linked analysis based on accounts by three keyOttoman historians’, in Markus Köhbach, Gisela Procházka-Eisl and Claudia Römer (eds.), Acta Viennesia Ottomanica, Akten des 13 CIEPO-Symposiums (Comité International des Études Pré-Ottomanes et

Page 14 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries Ottomanes), vom 21 bis 25 September, 1998 (Vienna: Selbstverlag des Instituts für Orientalistik, 1999), 247–55. (36) Lokman, Hünername, Topkapı Sarayı Müzesi, MS Hazine 1524, fol. 266a. Buda, in Hungary, on the right bank of the Danube, was held by the Ottomans from 1541 to 1686. This rite of submission took place on 29 August 1541, and the miniature takes some liberties with the scene as John Sigismund was in fact only an infant at the time. On John Zapolya, see Gábor Barta, ‘IV. The first period of the principality of Transylvania (1526–1606)’, in Béla Köpecki and Bennett Kovrig (eds.), History of Transylvania, trans. Péter Szaffkó (New York: Columbia University Press, 2001), 1: 606–19. (37) For two different types of Ottoman siege views, see the 1565 map, or siege (kuşatma) plan of Mustafa Pasha for Malta and the illustration of the siege of Inebahtı from Katib Çelebi’s Tuhfet ül-kibar, Topkapı Sarayı Müzesi, MS R. 1192, fol. 17a, in Idris Bostan, Kürekli ve Yelkenli Osmanlı Gemileri (Istanbul: Bilge, 2005), 78–9, 88–9. Although we have few Ottoman maps from the sixteenth century, it is clear that the Ottomans drew strategic maps for treaty and military purposes and did systematic ‘mapping’ for tax purposes in tahrir surveys. Gábor Ágoston has noted the surprisingly accurate map of the region around Kanije/Kanizsa prepared by Üveys Pasha, Ottoman governor of Buda from 1578 to 1580; see Chapter 3 below, p. 66, and idem, ‘Information, ideology, and limits of imperial policy: Ottoman grand strategy in the context of Ottoman-Habsburg rivalry’, in Virginia H. Aksan and Daniel Goffman (eds.), The Early Modern Ottomans: Remapping the Empire (Cambridge: Cambridge University Press, 2007), 89. (38) Süleyman’s army came to the relief of Buda, which had been attacked by King Ferdinand. Gábor Ágoston, Guns for the Sultan: Military Power and the Weapons Industry in the Ottoman Empire (Cambridge: Cambridge University Press, 2005), 194, has noted: ‘Until well into the seventeenth century, but especially in the fifteenth and sixteenth centuries, Ottoman artillery proved to be superior against European fortifications…Between 1521 and 1566 only thirteen Hungarian forts were able to resist Ottoman firepower for more than ten days, merely nine castles for more than twenty days, and altogether four fortresses were able fully to withstand Ottoman assaults.’ (39) Evliya Çelebi provides some interesting word-pictures of fortresses in his various narratives of travels. See Seyahatname, 5:295; 8: 315, 327, and Evliya Çelebi in Albania, 25, 147, 205. On the fortress of Kaçanik, in Kosova, for example, he writes: ‘Sinan Pasha, the conqueror of Yemen, constructed a beautiful stonework fortress at the mouth of the gorge and on the banks of the Lepenca river. It is square in shape and 800 paces in circumference. One drawback is that it is situated in the valley and so has many higher points surrounding it. The castle has a warden, 50 garrison soldiers, two cannons, and one gate. Inside are 40 or 50 houses to accommodate the soldiers, but no public buildings. Outside the wall there are another 100 houses, all with tiled roofs and gardens’ (Seyahatname, 5: 295; trans. from Evliya Çelebi in Albania, 25). (40) See on Matrakçı Nasuh, Halil Sahillioğlu, ‘Dördüncü Muradın Bağdat Seferi Menzilnamesi’, Belgeler 2/3–4 (1965), 1–36; and Hedda Reindl, ‘Zu einigen Miniaturen und Karten aus Handschriften Matraqcı Nasuh’s’, Islamkundliche Abhandlungen, Beiträge zur Kenntnis Südosteuropas und des Nahen Orients, 17 (1974), 146–71. (41) See for the Turkish edition, Nasuhü’s-Silahi (Matrakçı), Beyan-ı menazil-i sefer-i ‘Irakeyn-i Süleyman Han, ed. H. G. Yurdaydın (Ankara: TTK, 1976).

Page 15 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Fortress: Defining and Mapping the Ottoman Frontier in the Sixteenth and Seventeenth Centuries (42) For a set of variations on Ottoman mapping and other fortress views in an Ottoman context see J. M. Rogers, ‘Itineraries and town views in Ottoman histories’, in J. B. Harley and David Woodward (eds.), The History of Cartography, II/1: Cartography in the Traditional Islamic and South Asian Societies (Chicago: University of Chicago Press, 1992), 228–55, esp. 234–45; and Kathryn Ebel, ‘City Views, Imperial Visions: Cartography and the Visual Culture of Urban Space in the Ottoman Empire, 1453–1603’ (Unpublished Ph.D. thesis, University of Texas, 2002). (43) Matrakçı Nasuh, Menazil-i Sefer-i ‘Irakeyn, İstanbul Universite Kütüphanesi, MS TY5964, fol. 42b. On the Ottoman–Safavid frontier, see Rudi Matthee, ‘The Safavid-Ottoman frontier: Iraq-i ‘Arab as seen by the Safavids’, in Karpat and Zens (eds.), Ottoman Borderlands, 157–73. Matrakçı also included illustrations of fortresses and routes in his accounts of the Balkan campaigns, but I do not have a suitable image from one of those accounts. (44) Matrakçı’s campaign volume was not meant as a strategic plan or roadmap (in the modern sense) for the journey to Baghdad; like Lokman’s work it was both a history and a celebration of sovereign power. But it does constitute a commemoration of a journey portrayed in such a way as to show the important markers of space.

Access brought to you by:

Page 16 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary GÁBOR ÁGOSTON

DOI:10.5871/bacad/9780197264423.003.0003

Abstract and Keywords It has been fashionable in the generalist literature to argue that the Ottomans lacked knowledge in European geography and politics. This chapter first offers some comments regarding Istanbul's understanding of geography and environment in the context of Ottoman strategy and frontier warfare. The, it presents a short overview of the importance of rivers, marshlands and mountains with regard to the formation of the opposing Hapsburg and Ottoman defence systems in Hungary. The last part of the chapter deals with the relationship between landscape, climate and fortifications and offers some preliminary results and tentative observations regarding deforestation and marshlands. The discussion also argues, although in somewhat different ways, that the Ottomans and their Hapsburg rivals both had a keen interest in geography and mapped their empires and resources, and possessed adequate information as to the terrain and river systems of their lands and frontiers. Keywords:   European geography, Hungary, Ottoman, deforestation, frontier warfare

IT HAS BEEN FASHIONABLE IN THE GENERALIST LITERATURE to argue that the Ottomans lacked knowledge in European geography and politics. In several of my recent studies, I have demonstrated that Istanbul possessed a multi-layered information-gathering system that provided the Ottoman government both in the centre and in the provinces with sufficient information about their adversaries.1 In this chapter, I first offer some comments regarding Istanbul’s understanding of geography and environment in the context of Ottoman strategy and frontier warfare. Then I proceed with a short overview of the importance of rivers, marshlands and mountains with regard to the formation of the opposing Hapsburg and Ottoman defence systems in Hungary. The last part of the chapter deals with the relationship between landscape, climate and fortifications and offers some preliminary results and tentative observations regarding deforestation and marshlands. In the absence of specialised studies on the subject, one cannot do more. Nevertheless, these are important, as they underline the interconnected nature of landscape and frontiers, and the significance of such studies for Ottomanist historians,

Page 1 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary for whom both modern military and environmental history are new disciplines. The chapter also shows that we do in fact possess a wide array of both data and sources that, used in novel ways, could shed light on hitherto neglected aspects of Ottoman military, economic and social history. (p.58) With regard to the Ottomans’ understanding of geography, the available evidence suggests that Ottoman policy-makers not only understood geography but clearly were capable of thinking in larger strategic terms. As examples one can point to the gradual and systematic conquest of the Black Sea coast and the Danube Delta up to the 1480s, and the capture and construction of strategically important forts along major river routes, such as the Danube, the Tigris and the Euphrates. The Ottomans recognised the importance of the Danube as early as the late fourteenth century and occupied all strategically vital fortresses along the river during the next 150 years. These included the forts at Kilia (Ott. Kili, 1484), Silistra (Ott. Silistre, 1388), Ruse (Ott. Rusçuk, 1388), Nikopol (Ott. Niğbolu, 1395), Vidin (1396), Severin (1524), Orşova (1522), Golubac (Ott. Güvercinlik, 1427, 1458), Haram (1483), Smederevo (Ott. Semendire, 1439, 1459), Zemun (Ott. Zemin, 1521), Belgrade (1521), Petrovaradin (1526), Buda (Ott. Budin/ Budun, 1541), and its twin city Pest (Ott. Peşte, 1541), Vác (Ott. Vaç, 1543), Visegrád (Ott. ViĿegrad, 1544), and Esztergom (Ott. Estergon, 1543) (see Figure 3.1). The possession of these forts proved instrumental for the Ottomans to control the Danube, the major watercourse in their European theatre of war. Belgrade and Buda became Ottoman administrative centres following their conquests, also acting as major logistical bases during Ottoman campaigns against Hapsburg Hungary in the sixteenth and seventeenth centuries. The Ottomans established naval arsenals at Rusçuk, Niğbolu and Vidin, and smaller shipbuilding sites at Güvercinlik, Semendire, Belgrade, İzvornik (Serb. Zvornik), Alacahisar (Serb. Kruševac), Pojega, Mohaç (Hu. Mohács), Buda and Estergon. Semendire and İzvornik were each capable of constructing some 200 to 250 river boats for the Hungarian campaigns of Süleyman I in the 1540s and 1560s. These forts and their hinterland were also responsible for providing bridgebuilding material and pontoon bridges.2 The Danube played a crucial role in Ottoman campaign logistics.3 Ordnance and ammunition, especially heavy stone cannon balls, were transported on special ships, called gun ship (top gemisi) and stone ship (taş gemisi). For the transport of gunpowder the Ottomans used special ‘covered’ (örtülü) boats. However, the Danube was not fully navigable. During campaigns against Hungary and the Hapsburgs, the Ottomans used the waterway only from Rusçuk (or Belgrade) up to Buda (or Estergon); thus cannon, ammunition and victuals shipped from Istanbul via the Black Sea to Varna were transported on carts from Varna to Rusçuk (or to Belgrade), where they were again loaded on ships.4 Securing crossing points on the Danube and its tributaries, and building bridges and pontoon bridges were essential for campaign logistics and for the arduous and dangerous business of river crossing. The failure to provide bridges and pontoons and other infrastructure of river crossings often resulted in disasters and/or mutinies.5 (p.59)

Page 2 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary (p.60) Apart from the conquest and securing of major rivers, other obvious examples for the Ottomans’ strategic thinking include the sixteenth-century Don– Volga and Suez Canal projects.6 The idea of connecting the Don and Volga rivers at their closest point was first conceived in 1563 under Süleyman the Magnificent. With the help of the planned canal the Ottomans hoped to check Muscovy’s expansion in the north Caucasus, by dislodging them from the lower Volga, and especially from Astrakhan, captured in 1555 by Tsar Ivan Figure 3.1. Major Ottoman campaigns in IV’s (r. 1547–84) forces. However, the plan Hungary in the sixteenth century. did not get support from the sultan, whose strategic interest lay in the Mediterranean (Malta, Cyprus) and Hungary. In 1568–9, following Süleyman’s death and the accession of Selim II, Grand Vizier Sokollu Mehmed Pasha revived the plan. With the canal, the grand vizier planned to transport Ottoman warships carrying siege cannons, ammunition and provisions from the River Don to the Volga. It would have enabled the Ottomans to block Russian relief forces coming downstream on the Volga, and to capture Astrakhan with the flotilla and the accompanying Ottoman and Crimean Tatar land forces. In July 1569, the Ottoman flotilla ascended the Don from Azak (Azov) and stopped south of the River Ilovlya, tributary of the Don, somewhere opposite presentday Volgograd (formerly Tsaritsyn and Stalingrad). They attempted to dig a canal at what seemed to be the two rivers’ closest point. However, the rivers were still some 50 kilometres apart and the ground was hilly. The assertion by some historians that the Ottomans did dig about one-third of the estimated 50-kilometre-long canal is unsubstantiated. Because of insurmountable physical obstacles, the Ottomans abandoned the plan shortly after their arrival and sent the flotilla back to Azak with most of the siege artillery. Although Ottoman and Tatar land forces reached Astrakhan by mid-September, having only light field pieces and perhaps as few as two siege cannons, they failed to capture the fort. Ending the siege after just ten days, the army reached Azak by 23 October, after a devastating one-month-long march during which hundreds died of hunger and thirst. Although Kasım Bey, the governor of Azak and the commander of the expedition, wanted to renew the operation next spring, Istanbul was already preparing for the Cyprus campaign, which demanded all available troops and resources. Thus ended the Don–Volga canal project, with which Sokollu Mehmed Pasha hoped to dislodge the Russians from Astrakhan and the Lower Volga, and, by (p.61) transporting the Ottoman Black Sea fleet on to the Caspian Sea, attack Safavid Persia from the north, conquering the province of Shirvan.7 Moreover, Ottoman control of the steppes and forests of the lower Volga ‘would have strengthened the empire’s ecological portfolio, providing grain, horses and timber in abundance’,8 and thus would have further enhanced the Ottoman strategic position on this frontier and in general. With the Suez Canal, Istanbul wanted to transport its main Mediterranean navy on to the Red Sea and the Indian Ocean to check Portuguese expansion there. Ottoman attempts to dislodge the Portuguese from the Indian Ocean had failed spectacularly under Süleyman I. Although the expedition launched from Suez by the governor-general of Egypt, Süleyman Pasha, in 1538 secured Aden and Yemen, it never met the Portuguese navy. Another expedition in 1552, led by Piri Reis (c.1470–1554), the renowned Ottoman corsair, marine and cartographer and author of Page 3 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary the famous naval manual (Kitab-i Bahriye, or Book of Seafaring/Maritime Matters) and maker of the first Ottoman map of the New World, ended in failure, and cost Piri Reis his life in 1554. Another attempt launched from Basra by Seydi (Sidi) Ali Reis in 1554 had the same unfortunate end.9 The idea to deploy part of their Mediterranean fleet in the Indian Ocean was revived by Sokollu Mehmed Pasha in 1568. In a decree (ferman) written in January 1568, Sultan Selim II ordered the governor-general of Egypt to investigate the feasibility of such a canal. The beylerbeyi had to dispatch knowledgeable local architects and engineers who would, in their turn, study the possibility of the construction of a canal between the Mediterranean and the Red Seas. The governor was also to report on the length and width of the prospective canal. As in earlier attempts, the underlying concern behind the decree was the threat that Portuguese expansion meant for the Holy Cities of Mecca and Medina and for the Muslim pilgrims coming from India.10 However, nothing materialised from the plan. A third ambitious project was the Sakarya–İzmit canal plan. This would have connected the Black Sea with the Sea of Marmara through the Sakarya River (the third largest river of presentday Turkey after the Kızılırmak and Euphrates) that empties (p.62) into the Black Sea and Lake Sapanca, an elongated lake some 17 kilometres long and 6 kilometres wide that lies some 5 kilometres to the west of the river. From the latter another canal would have led to the Bay of İzmit. The project would have enabled the Ottomans to bring timber from the forests of the sancak of Kocaeli to the Istanbul Naval Arsenal as well as firewood to the capital via canals, rivers and lakes, rather than using more expensive and cumbersome overland transportation. The idea was first conceived under Süleyman the Magnificent, who commissioned Mimar Sinan and a Greek architect to undertake the work. Although the project was launched and allegedly some 15 kilometres of canal was built, Süleyman’s land campaigns diverted attention and resources from the project. The plan got new support in 1591, when Istanbul pondered largescale fleet construction in order to strengthen the Ottoman navy, which had deteriorated in the relatively peaceful years after the Battle of Lepanto in 1571. However, the scheme was abandoned in 1593 when a renewed war with the Hapsburgs in Hungary demanded all available resources. Istanbul revived the plan in the seventeenth, eighteenth and nineteenth centuries, but every time without success, due largely to infighting and intrigues within the Ottoman elite.11 While none of the above-mentioned canal projects materialised, they point to the strategic thinking of the Ottoman leadership. On the other hand, they also reveal the lack of detailed knowledge with regard to the topography of remote lands, especially in the case of the Don– Volga canal project. We should not forget, however, that such ambitious projects also failed elsewhere in Europe. In 1697, Peter the Great (r. 1682–1725) also attempted to connect the Volga and Don rivers by a canal between their tributaries, the Kamyshin and Ilovlya, in order to gain access to the Black Sea, but the obstacles proved insurmountable for his richer and strategically better-suited empire too. The project did not materialise until Soviet times. The construction of the Volga–Don Ship Canal south of Volgograd started in 1948 and was completed in four years. Another example is the Rhine–Maas canal. First attempted by the Spanish Hapsburgs in 1626–9 and again by Napoleon in 1804, this ambitious project also ended as a costly failure.12 (p.63) Like their arch-rivals, the Spanish and Austrian Hapsburgs, the Ottomans also attempted mapping their empire and frontiers, although the techniques and forms by which the Ottomans collated, stored and displayed the relevant information were somewhat different from those used in Europe. In addition to a ‘General Visitation’, ordered by Philip II in 1559 to gather

Page 4 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary information about his dominions in Italy, the king also commissioned cityscapes and maps of the Iberian Peninsula. On Philip’s order, in the 1560s, the Flemish court painter Anton van den Wyngaerde travelled extensively in Spain and North Africa, preparing a series of topographical views of towns, of which sixty-two finished ones survive. It seems that Philip planned to have them engraved and published, but the cityscapes became scattered in Vienna, London and Oxford and were rediscovered only in the nineteenth century. A more ambitious project involved the mapping of the entire Iberian Peninsula. In the 1570s and 1580s a team of cartographers surveyed the whole peninsula, some 500,000 square kilometres, presenting their results in an atlas of twenty-one detailed and remarkably accurate maps on a scale of 1:430,000. While the Escorial ‘Atlas of Spain’ remained, along with the planned description of Spanish America to be based on the comprehensive questionnaires of the Relaciones topográficas, incomplete and unpublished, the maps were unique in their time, for no other European state in the sixteenth century had anything similar to them.13 In the Ottoman Empire, the most valuable information acquired from ‘mapping’ was collated and stored in the land or revenue surveys, known as tahrir defterleri, which were undertaken regularly and systematically in the fifteenth and sixteenth centuries. These surveys afforded the Istanbul government and its provincial administrators a detailed database regarding the size, composition and economic conditions of the population of various sancaks and provinces (beylerbeyiliks, vilayets, or eyalets). In frontier provinces, these tahrirs were often used to delineate borders and solve frontier disputes. The most famous of such surveys along the Hapsburg–Ottoman frontier were the very first Ottoman tahrirs of the sancaks of the province of Buda, compiled by Halil Bey in 1546 and known in contemporary Hungarian parlance as Halil Bey’s defters. All concerned parties regarded villages that had been included into Halil Bey’s survey as settlements under Ottoman rule and taxation. Naturally, in later years Ottoman officials tried to include further villages in these defters, against which the Hungarian and Hapsburg authorities protested vehemently, considering such attempts as an unlawful broadening of the Ottoman frontiers.14 Controversies that surrounded Halil Bey’s defters remind us that we should not limit Ottoman ‘mapping’ to cartographical undertakings, although these were not unknown to the Ottomans either. (p.64) Sultan Mehmed II’s keen interest in maps as tools of military reconnaissance and intelligence is well documented.15 The existence in the archives of the Topkapı Palace, the empire’s administrative centre, of military maps and siege plans, such as those of Kiev (from the time of Sultan Bayezid II, 1481–1512), Belgrade (1521?), Malta (1565), Szigetvár (1566), Vienna (1683), Van in Asia Minor, Adakale (1738), and diagrams of the battles of Haçova/Mezőkeresztes (1596) and Prut (1711), to name only the bestknown ones, illustrate the employment of mapmakers during Ottoman campaigns and the use of cartographical representations in military communications. The mid-seventeenth-century regional map, drawn on eight double-folio sheets of paper, that charts the Euphrates and Tigris Rivers and indicates important sites, roads and mountains along the rivers or the map of the region to the north of the Black Sea (1768–9), shows Ottoman interest in mapping waterways and roads for both commercial and military purposes.16 Ottoman corsairs and marines also gathered valuable information on ports and port cities. While many of the topographical illustrations in Piri Reis’s Kitab-i Bahriye are schematic and inaccurate, his personal observations are evident with regard to ports, harbours and harbour fortifications which—like those at Ragusa/Dubrovnik, Ancona and Venice—are carefully portrayed. As a former corsair, Piri Reis showed special interest in harbours, and it has been

Page 5 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary suggested that the depiction of the lagoon fortifications of Venice along with that of the Venetian arsenal in the Kitab-i Bahriye might have been based on the work of a spy or renegade.17 A new interpretation of the town views in Matrakçı Nasuh’s (d. 1564) works argues that the Ottomans were also keen to map their new territorial acquisitions and the geographical extent of their empire. Matrakçı Nasuh’s Beyan-i Menazil-i Sefer-i Irakeyn-i Sultan Süleyman Han (The Description of the Stages of His Imperial Majesty Sultan Süleyman’s Campaign in the Two Iraqs [modern Iraq and western Iran], usually known by its shorter title Mecmua-i Menazil), and his Tarih-i Feth-i Şikloş, Üstürgon ve İstolni Belgrad (History of the Conquest of Siklós, Esztergom and Székesfehérvár), which narrate and illustrate Sultan Süleyman’s conquests during his eastern and Hungarian campaigns in 1534–5 and 1543, respectively, have been studied extensively, not least because of their splendid town views which, while resembling contemporary European bird’s eye views, represent a specifically Ottoman genre. However, it has only recently been noticed that all the prominently displayed towns in double-folio city plans in the Mecmua-i Menazil—those of Tabriz and Sultaniye in western Iran, Baghdad in central Iraq, Najaf, Karbala, and Hilla in lower Mesopotamia—were newly conquered frontier (p.65) towns which effectively defined the Ottoman–Safavid frontier in 1535. The same pattern is observable in Matrakçı Nasuh’s other works, where the most prominently described towns are those on the Empire’s Hungarian and eastern frontiers.18 Awareness of geography worked differently in global and local contexts. Close familiarity with topography during imperial campaigns and smaller military undertakings was indispensable for identifying possible river crossings and drinking-water sources for soldiers and animals alike. Ottoman campaign journals, chronicles and geographical descriptions not surprisingly make frequent references to river crossings, as well as to local road guides (kılavuz). The employment of kılavuzes was instrumental especially in regions where climate and vegetation were hostile. Such a frontier was Hungary, which, contrary to general belief, was far from being the western extension of the large waterless Eurasian steppe. Without excellent knowledge of the region’s geography, vegetation and climate, communication and military undertakings were unthinkable in this country, where entire regions were dominated by swamps and marshes and where forests had extended over areas much wider than they do today. The fact that the Ottomans hired road guides as late as 1697 when marching from Niš to Belgrade, a familiar terrain and a usual campaign route along the Morava River, underlines the importance of up-to-date knowledge of topography and illustrates the difficulties and hazards associated with moving large forces even in familiar territory19 As in other parts of the Muslim–Christian divide, the Ottoman–Hapsburg border in Hungary was also a cultural and linguistic frontier that made information gathering more difficult. However, local societies provided the personnel who possessed the necessary language skills and acted as mediators between the two sides of the frontier. Since few Ottomans knew Latin or German, and perhaps even fewer Hapsburg officials knew Ottoman Turkish, Hungarian became the language of diplomacy and information gathering along the frontier, replacing Slavic, which had played the same role up until the early sixteenth century. The Hungarian interpreters of Ottoman beylerbeyis and sancakbeyis played a crucial role in this regard, and in the Austrian and Hungarian archives there are hundreds of letters written in Hungarian that provide precious information concerning daily life along the frontier.20 Governors and commanders in Hungary employed spies and informants to gather information about the other side of the frontier. They certainly felt that they knew a great deal about their enemy. In 1561, when the sancakbeyi of İstolni Belgrad (Hu. Székesfehérvár), Hamza Bey, was threatened that a large Hapsburg army

Page 6 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary would be sent against him, he reminded the Hungarian king, Ferdinand I of Hapsburg (r. 1526– 64), that he could not possibly have (p.66) enough soldiers at his disposal for the undertaking. Had Ferdinand had enough soldiers, Hamza would have been informed, because I have had a spy living in Vienna for six years, whose wife and child are there, a man who can say mass if he wants, or be a scribe, or a German, a Hungarian, or a good improviser, a soldier, a man with a limp, or someone who walks as steady as you do with a good knowledge of every language.21 With the help of their informants living in castles and villages in Hapsburg Hungary, Ottoman officials closely monitored the Hapsburg frontier. The detailed and surprisingly accurate Ottoman map of the Hapsburg border fortresses around Kanije (Hu. Kanizsa) is an example of such activity. The map was commissioned by Üveys Pasha, beylerbeyi of Buda (1578–80), and sent to Istanbul. It identified all the forts and castles, as well as major river crossings of this recently reorganised section of the Hapsburg frontier around Kanije in Transdanubian Hungary.22 Information coming from the provinces was assessed in Istanbul with the help of the Sublime Porte’s renegade interpreters, who being Germans, Italians, Greeks or Hungarians, were familiar with Europe’s geography and had access to European maps and atlases. While research regarding the relationship between European and Ottoman cartography is in its infancy, sporadic evidence suggests that Istanbul was more aware of European cartographical knowledge than previously thought. In 1573, one of the Porte’s renegade interpreters, Tercuman Mahmud (alias Sebold von Pribach of Vienna), ordered from Vienna two copies of Abraham Ortelius’ (1527–98) Theatrum Orbis Terrarum (Theatre of the World), the first modern atlas that contained uniform maps and supporting text. Mahmud was surprisingly up-to-date, for Ortelius first published his map-book in 1570.23

River Systems, Terrain and Fortifications As in other parts of the world, border defence systems along the Ottoman-Hungarian/ Hapsburg frontier often followed major rivers and used river systems, marshlands, (p.67) mountains and other natural defensive features that geography offered. The Danube played a crucial role as a natural border of empires from Roman times. The medieval Hungarian kingdom (1000–1526) built, from the late fourteenth century on, its anti-Ottoman defence system along the Danube and Sava Rivers. This defence line successfully halted Ottoman advance for 150 years, until it collapsed in the 1520s (see Figure 3.1).24 Ottoman authorities also studied the defence line of the countries they wanted to conquer. During their 1541 Hungarian campaign, which ended with the capture of Buda, the capital of the medieval Kingdom of Hungary, the Ottomans prepared a ‘plan of conquest’ that listed some of the strategically important Hungarian castles whose capture, according to the unknown author, was especially warranted. The document listed the forts according to their owners, the most prominent aristocrats and politicians of Hungary. It also gave the location of the castles, accompanied by short comments regarding their immediate past.25 By the 1550s, the Ottomans controlled the Danube as far as Estergon. Most of the former Hungarian castles between the Danube and Lake Balaton in Transdanubia, the forts in the Novigrad (Hu. Nográd) Mountains north of Ottoman Budun (Hu. Buda), and all the major castles along the River Tisza and its tributaries in the eastern parts of the country had also been in Ottoman hands. The Ottoman authorities repaired, strengthened and rebuilt these recently Page 7 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary captured castles and forts when they thought them suitable for defence, or demolished them when they considered them useless. The Hapsburgs for their part responded to these Ottoman conquests by organising their own border defence system. This crescent-shaped new defence line, established from the midsixteenth century onward, stretched some 1,000 kilometres in length from the Adriatic Sea to northern and north-eastern Hungary and comprised 120 to 130 large and small forts and watchtowers. Not surprisingly, it followed the mountains of Transdanubia and northern Hungary, that is, the only possible natural defence line offered by the geography of the region after the collapse of the former defence system that had relied on the Danube–Sava Rivers. Since Vienna lay just 180 to 200 kilometres from the major Ottoman garrisons in Hungary, the Hapsburg administration rebuilt and modernised the key fortresses of the Croatian and Hungarian Military Border according to the state-of-the-art standards of fortress building (trace italienne). In the 1570s and 1580s, some 22,000 soldiers guarded this border, of whom 15 per cent were (p.68) German, Italian and Spanish mercenaries, stationed in the key fortresses, whereas the rest of the guards were comprised of Hungarians, Serbs and Croats.26 The number of forts and the size of the garrison forces defending Ottoman Hungary in the second half of the sixteenth century were comparable to that of the Hungarian/Hapsburg defence line. In their two Hungarian provinces, Budun and Temeşvar, the Ottomans had about 120 to 130 castles and smaller forts which were guarded by some 18,000 garrison soldiers and 7,000 cavalrymen remunerated by military fiefs (timarlı sipahi). Ottoman pay registers and other archival sources indicate that the Ottomans closely monitored the Hungarian/Hapsburg frontier and adjusted the size and composition of their garrisons accordingly. The majority of Ottoman forts in Hungary were captured from the Hungarians. The Ottomans also built new forts and altered the inherited ones when their defence strategy so required. Moreover, their conquests and the organisation of new provinces and sub-provinces show that they took into consideration environmental and strategic considerations (see Figure 3.2). The establishment of a defensive ring of forts around Buda/Budun, the administrative and logistical centre of the first Ottoman province in Hungary by the same name (vilayet-i Budun), illustrates this strategic thinking. Estergon was the most important fortress that guarded Budun from Hungarian and Hapsburg attacks from the west. Since it also controlled both river and overland communication routes, Estergon received special attention from the Ottoman authorities. They not only reinforced Estergon proper by building several new bastions both in the fortress and in the ‘Lower Castle’ and a gunpowder work (baruthane), but also erected two new forts, mainly palisades, known as parkan in Ottoman Turkish and parkány in Hungarian: Tepedelen (Hu. Szent Tamáshegy), right next to Estergon, and Ciğirdelen (Hu. Párkány), on the other side of the Danube.27 In the Danube bend, Višegrad and Vaç guarded Budun from the north/north-west, whereas Hamzabey Sarayı and Korkmaz/Cankurtalan at the southern tip of Csepel Island near Adony, secured the city from the south. From Vaç, the protecting ring defending Budun from the north followed the line of the Novigrad Mountains, where most of the Hungarian castles, such as Drégely, Szécsény, Hollókő and Buják (Ott. Diregel, Seçen, Holloka and Buyak) were captured in 1552. South of Buyak, the Ottomans built a new fort at Hatvan (1544), and made it the centre of a newly established sancak, for it controlled both the Zagyva River, a tributary of the Tisza, and the royal route (via regia) coming from Budun. The capture of Siklós (Ott. Şikloş), Pécs (Ott. Peç) and Fehérvár (Ott. İstolni Belgrad) in Transdanubia (west of the River Danube) in 1543 helped to connect Budun with Ottoman-held Sirem (Serb. Srem/Cr. Srjem; between the Sava and Danube Rivers in present-day Croatia and Serbia). All these forts were reinforced and became sancak

Page 8 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary centres. The conquest of Ozora, Tamási (Ott. Tomaşin) and Simontornya (Ott. Şimontorna) in 1545 secured Ottoman communication and transport along the right bank of the Danube.28 (p.69) (p.70) The 1551–2 campaigns are significant, for several of the forts occupied during these campaigns secured important river routes and crossings. In September 1551, Sokollu Mehmed Pasha, governorgeneral of Rumeli, the European parts of the empire, captured Becse (Ott. Beçe) south of Szeged (Ott. Segedin or Seged) on the River Tisza, which after Szeged (conquered in 1543), became the Ottomans’ second major port and river crossing point on the lower Tisza, especially during their campaigns and raids against the castles of the region of Temesvár (Ott. Temeşvar / Rom. Timişoara). From Becse, Sokollu Figure 3.2. Fortifications and geography of proceeded south-east, taking Becskerek sixteenth-century Hungary. (Ott. Beçkerek) on the River Béga, a tributary of the Tisza. His other conquests proved short-lived: Csanád (Ott. Çanad), whose rectangular fortress with corner bastions guarded a crossing point on the River Maros (Rom. Mureş), a tributary of the Tisza, and Lippa (Ott./Rom. Lipova), east of Csanád on the left bank of the Maros, with an elaborate defence system, were both retaken by Hapsburg and Hungarian forces within a month. Sokollu also failed to conquer Temesvár, the most important fortress in south-eastern Hungary on the River Temes (Rom. TimiĿ), a tributary of the Danube, whose defences were strengthened just before the 1551 siege. Not surprisingly, the campaigns of 1552 also targeted some of the most important castles that commanded the river routes of the Trans-Tisza region in south-eastern Hungary. At the end of July, after a siege of more than a month, Temesvár fell to second vizier Kara (‘Black’) Ahmed Pasha, by which the Ottomans secured the River Temes. Lippa soon followed, as did Arad, both on the Maros River, and several other castles along the Maros and Temes Rivers. Then Ahmed Pasha turned to the north, and in early September he conquered Szolnok (Ott. Solnok), which controlled the confluence of the Tisza and Zagyva Rivers, and the only permanent bridge over the Tisza.29 (p.71) That the conquests of 1551–2 were the result of careful planning and knowledge of geography is apparent from the fact that the aforementioned ‘conquest plan’ of 1541 also identified these forts as important targets. In 1541, the concern was the defence of the Ottoman northern frontier in Sirem. ‘Until the fortresses called Lippa, Temesvár, Becskerek and Becse have been taken Sirem province will not be free from evildoers’, commented the unknown author of our source.30 The strategic significance of these forts is illustrated by the fact that in 1552 Temeşvar became the centre of a new province, the second beylerbeyilik established in Hungary. Beçe-Beçkerek, Çanad, Arad and Lipova all became sancak centres, belonging to the new province of Temeşvar, Page 9 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary and their first tahrir surveys were promptly completed in 1554.31 The Ottomans also created a new sancak around Solnok, and attached it to the province of Budun. The Ottomans used their forts in Hungary for both defensive and offensive purposes. During the long Hapsburg–Ottoman war at the turn of the century (1593–1606), the Ottomans endeavoured to capture Vienna again; by occupying Veszprém (Ott. Bespirim), Palota, Tata, Győr, Szentmárton (Pannonhalma), Pápa and Tihany (Ott. Tihon) they came very close to attaining their goal. However, in the second phase of the war, because of the counter-attacks of the Hapsburg and Hungarian forces, the Ottomans lost all their recent conquests, except for Palota. Still, by occupying Eger (Ott. Eğri, 1596) and Kanizsa (Ott. Kanije, 1600), they further expanded their realm in Hungary. Further serious border changes took place only in the middle of the seventeenth century. The occupation of Várad (Ott. Varad, Rom. Oradea) in 1660 ceded control of the Sebes-Körös River (Rom. Crişul Repede), and thus of Transylvania. The occupation of Érsekújvár (Ott. Uyvar, Sl. Nové Zámky) in 1664 drove a wedge into the Hungarian defensive ring around Vienna and, as the unsuccessful Hapsburg siege of Buda in 1684 proved, Uyvar significantly increased the protection of Ottoman Budun. Terrain also influenced the composition of garrisons. Whereas cavalrymen usually dominated garrisons on both sides of the border in Hungary, in and around Kanije, which was surrounded by swamps and marshlands, infantry comprised the majority of the forces, for they could move around more easily in swamps than cavalry. In Ottoman Kanije in the 1650s, the infantry–cavalry ratio was 60:37. It is hardly surprising that in the fourteen smaller Hungarian/Hapsburg border forts charged with the task of opposing Ottoman Kanije—known as Gegen Canischa werts Liegenten Granitzen— both the size and the distribution of the garrisons were comparable with that of Kanije: 59–64 per cent of the soldiers being infantry, and only 36–41 per cent cavalry32

(p.72) Landscape, Climate and Frontier33 The most characteristic landscape-forming transformations in Hungary in the sixteenth and seventeenth centuries were deforestation and the spreading of marshes and swamps. While the causes, the extent and the consequence of these changes require further research, the simplistic and biased views regarding the negative effects of the ‘Turkish conquests’ on landscape and population put forward by Gyula Szekfű, the most influential historian in inter-war Hungary and the father of the Hungarian Geistesgeschichte school, are hardly tenable. Known for his proHapsburg and anti-Turkish views, indeed for his Turkophobia, Szekfű blamed the Turks not only for the general dislocation of the country’s historical evolution, economic and social decay, depopulation and changing ratio of Magyar to non-Magyar populations, but also for the destruction of environment and changes in climate. In his view, ‘the puszta [semidesert] character of the Great Hungarian Plain, the puszta-vegetation, the arid puszta- climate in which extreme heat waves alternate with severe cold weather, the lack of trees and water are all the results of the Turkish era, that is, the consequences of the Turkish conquest.’34 While his multivolume Magyar History (1928–34), co-authored with his noted medievalist colleague Bálint Hóman, has enjoyed startling popularity in the past two decades and his pro-Hapsburg views got a new spin with Hungary’s joining the European Union, his theses regarding landscape were challenged as early as 1940. Pál Teleki and his colleagues refuted Szekfu’s views, by pointing out that the puszta- character of the Great Hungarian Plain was not created by the Turks, but was instead the result of the major drainage projects of the nineteenth century35 It is also clear that deforestation was not unique to Hungary and that its causes were similar to those observable elsewhere in Europe. Around 1500, Europe had substantially more forests and wooded areas than today. However, the growth in Europe’s population from about 1500 and the Page 10 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary resulting expansion of agriculture led to forest (p.73) clearings, whereas mining and smelting, war-related industries (cannon casting, saltpetre and gunpowder production), fortress building, and the construction of evergrowing navies, all brought about by the so-called early-modern ‘gunpowder/military revolution’, required substantially more fuelwood, charcoal and timber, and resulted in deforestation. Geographers and environmental historians have identified several causes of deforestation in the early modern period, of which clearing for agricultural expansion (because of population increases), fuelwood consumption, shipbuilding, charcoal and iron making are usually ranked as the most important activities that contributed to deforestation.36 Since the population of Hungary, unlike that of Western Europe, remained stagnant or increased only modestly thanks mainly to warfare and its impact,37 domestic fuel consumption probably did not rise substantially. Agricultural clearing, on the other hand, must have intensified. The sixteenth century was the golden era of the Hungarian cattle trade, associated with the upsurge of increased meat consumption of a booming European population and with the nature of a frontier economy. In war zones people preferred investment in movable assets, such as cattle, which could be saved more easily than grain or wine, whose transport depended on road conditions and the availability of the means of transportation. By the 1570s, total annual cattle exports from Hungary could have topped 140,000–150,000 head, of which two-thirds were sent to Austria and Upper Germany and the rest to Venice. By this time a new breed, known as the Hungarian ‘white-and-grey’ cattle, weighing up to 500 kg as compared to 200 kg of the smaller peasant cattle, dominated the market. All this required new pasture lands and the clearing of forests. In addition, tens of thousands of cattle and sheep grazed in woodlands, causing irreparable damage. Indeed, literature on woodlands and forests has named cattle and sheep breeding as the number one cause for deforestation in the Great Hungarian Plain, which, contrary to general belief, was covered amply by forests in the sixteenth and seventeenth centuries.38 It seems that in the mid-sixteenth century there were still plenty of wooded areas around castles and fortifications to meet their need for timber and fuelwood. However, after construction works in the major forts gathered pace from the mid-1550s onwards, less and less wood was available in the vicinity of these fortresses. The account books of the castle of Eger (northern Hungary) show that most of the timber in the 1550s came from the neighbouring forests situated at a distance of 10 to 15 kilometres around (p.74) Eger, whereas lumber, wood lath and shingles were purchased from some twenty-nine villages in the neighbouring Gömör county.39 Similarly, according to the decrees of the Hungarian Diet of 1563, the forests of the Csallóköz were designated to meet the timber need of the fortresses of Komárom and Győr in north-western Hungary. The decision seemed logical, for the forests of the Csallóköz—a large island in northwestern Hungary (present-day south-western Slovakia) between the main branch of the River Danube and the Little Danube, known formerly as the Csalló River—lay in the immediate vicinity of the two forts. Meeting the timber and fuelwood needs of the two fortresses, however, must have taken severe toll on the Csallóköz forests, especially after major rebuilding started in earnest in Komárom and Győr in the mid-1550s. Thus, the Diet of 1600 allowed the cutting of trees in the remote forests of Hungary’s northernmost counties (Turóc, Liptó, and Árva), along the borders of Poland, some 120–250 kilometres north of Komárom. Increased defence- and war-related activities on both sides of the frontier all drew heavily on forests. The account books of the Upper Hungarian Chamber show that the Hungarian garrisons under the command of the captain-general of Upper Hungary, that is the eastern parts of

Page 11 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary northern Hungary,40 used 24,542 cartloads of firewood during the winter of 1674. This would have equalled 41,741 cubic metres of wood, or some 14,600 metric tons of pine and 25,000 metric tons of oak,41 which seems a lot given that the number of soldiers in the forts of Upper Hungary could not have been more than 2,800 and was perhaps much less.42 While estimates regarding per capita per annum fuelwood consumption in pre-industrial Europe vary significantly, our figure is still two to ten times higher than available estimates for Northern Europe and (p.75) Germany.43 This can only be explained if we consider the following possible reasons: (1) The total number of people living in these fortresses usually exceeded those on the regular payroll. (2) Fortresses consumed substantially more firewood than average villagers or city dwellers, for they employed an array of wood-burning trades (gunpowder-makers, smiths, founders, potters, bakers, and so on). (3) For many of these trades they needed charcoal rather than firewood, which considerably increased their firewood needs, for contemporaries used five to ten kilograms of wood to make one kilogram of charcoal.44 (4) Winters in the early 1670s were colder than usual.45 Since soldiers on the royal payroll in the Upper Hungarian forts constituted only about 25 per cent of the total garrison forces deployed by the Hapsburgs along the Hungarian Border Defence Line in the seventeenth century, the fuelwood need of all the garrisons under Hapsburg command could have reached 166,964 cubic metres per winter. However, if we add to this the fuelwood consumption of the Ottoman garrisons (whose size in the 1660s was about twice that of the Hungarian/Hapsburg forces)46 we arrive at a staggering amount of 500,892 cubic metres of fuelwood burned by the opposing military garrisons along the Ottoman–Hapsburg frontier in Hungary per winter. Depending on the type of wood, this would equal the annual yield of 7,012 square kilometres of pine and 12,000 square kilometres of oak woodland.47 Fuel-hungry iron forges, saltpetre and gunpowder works put an additional strain on the forests of the frontier. The manufacturing of an estimated yearly amount of 108 metric tons (roughly 2,000 kantars) of gunpowder in Ottoman Hungary’s two most important Ottoman powder works (Budun and Temeşvar), required some 81 metric tons of saltpetre, 13.5 metric tons of charcoal and the like amount of sulphur (not counting the loss during production).48 To produce this amount of saltpetre, the powder works had to burn some 1,300 metric tons of firewood.49 It took an additional 67.5 to 135 metric tons of wood to make 13.5 metric tons of charcoal used in these (p.76) gunpowder works. The two together would consume the annual yield of almost 58 square kilometres of woodland. Another activity that took a serious toll on woodlands was deliberate destruction of forests and villages. The ‘scorched earth policy’, that is, the devastation of the hinterland of the Ottoman forts, had been suggested time and again by the Hapsburg military commanders (the most notable being the 1577 Viennese military conference), but the plan was eventually rejected for military, economic and political considerations. However, at least on two occasions (in 1598 and 1664), Hungarian commanders brought havoc to Ottoman Hungary. In July 1598, Miklós Pálffy (1552–1600), district captain-general of Érsekújvár (1589–1600), burned down a large territory as far as 130 kilometres to the south of Buda. In January 1664, Count Miklós Zrínyi (Nikola Zrinski, 1620–64), defender of the Muraköz Region (the area between the Rivers Drava and Mura in southern Hungary, now Međimurje in northern Croatia) scorched the whole area along the Drava, and burned down Süleyman I’s famous timber bridge at Eszék (Osijek), the main crossing point over the Drava and its surrounding swamps.50 Burning of villages, grain fields and woodlands was also frequent during smaller raids by both parties.

Page 12 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary Deforestation had an effect on the region’s hydraulic system too. Streams poured into rivers from deforested hills, bringing a lot of silt with them, banking up valleys, and blocking riverbeds. All these were facilitated by war conditions, which hindered water regulation works that had helped maintain riverbeds in earlier centuries. War conditions demanded military work on waterways, such as creating protective marshes. On plain lands, protection was secured by building castles in river estuaries and river bends or by routing water via canals from nearby rivers, streams or marshes into ditches dug around the castles. Such waters protected the riverside and ‘swamp forts’ of Győr, Tokaj, Szolnok, Gyula, Temesvár, Szigetvár, Kaposvár, Ecsed and Tata. As contemporaries, both Europeans and Ottomans, noted, many of these forts appeared to the observer as islands within lakes and marshes (see Figure 3.2). In his Tercüme-i Coğrafya-i Kebir (Translation of the Great Atlas) Behram Dimişki (d. 1690/1) commented on these features of several forts in Hungary. Thanks to its deep moat, Temeşvar looked like an island, whereas its surroundings were inaccessible swamp. Behind its inner castle there was a lake created from the waters of the Temes River. He had similar comments on Beçkerek, which stood in the Temes River on an island, surrounded by bogs.51 Kanije was enclosed by the Zakany River. It was inaccessible on every side for at least a mil (or about three kilometres),52 because of its swamps. (p.77) Sigetvar was also built in the centre of a large lake, encircled by waters and marshes. Its separate two fortresses (in fact, there were three) were connected by a bridge.53 Contemporary Ottoman and European topographical illustrations of Sigetvar and other ‘swamp forts’ were also in accord with these descriptions, and the maps prepared by Luigi Ferdinando Marsigli at the end of the seventeenth and the beginning of the eighteenth centuries provide vivid illustrations of the marshlands of Hungary and the Hapsburg–Ottoman frontier.54 In marshy and swampy areas, soil humidity and evaporation were extremely high. Foreign mercenaries fighting in Hungary in the sixteenth and seventeenth centuries complained about the harmful evaporation and dampness of the soil, which was so high that the shirts of soldiers who slept in tents were soaked in water by morning, despite the fact that the tents had multiple layers for better protection against the weather. This was coupled with the extreme weather: summers were hot and winters were very cold. May was already very warm, with unbearable heat in July and August. Not only were the seasons extreme, but temperature discrepancies during the day meant that summer days were too warm and nights were too cold and damp. On the other hand, evenings became very cool by the end of September, the first snow fell in October, and everything was covered by thick snow from November through February. Especially cold and harsh winters were recorded in the middle and at the end of the sixteenth century and in the middle and at the end of the seventeenth century. All these were the effects of the so-called ‘Little Ice Age’. The observations of contemporaries have been confirmed by climate historians, although changes in seasonal climate were considerable. While summers in 1534–6 and 1556–9 were exceptionally warm and dry, those in 1541–4 were cooler. Winters in 1565–96 and 1687–98 were especially severe. In the last third of the sixteenth century in every season (except autumns) temperature fell more than 1° C, and rainfall usually concentrated in midsummer.55 Persistently snowy winters and extraordinarily rainy summers were often followed by strikingly bad crops and famine. This occurred in the frosty 1690s, when the crops all over Europe were at their worst. The quantity of rainfall increased, and reached its peak at the turn of the eighteenth century. Periods with strikingly high rainfall coincided with the Long Hungarian War (1593– Page 13 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary 1606), the Ottoman campaigns of 1663–4, and the Holy League’s War of 1683–99, which greatly reinforced the negative effects of wars on the economy and society. (p.78) Wars claimed their victims by causing famines and epidemics and by destroying the normal functioning of the economy and society. Underfed, weak soldiers were easily susceptible to epidemics, which the armies spread among frontier populations. In the period between the sixteenth and the eighteenth century, the morbus hungaricus was the illness that made Hungary notorious. Medical history has identified the morbus hungaricus with epidemic or ‘louse-borne’ typhus, a disease of war and famine, which, in 1542, killed some 30,000 German and Italian soldiers during Joachim of Brandenburg’s Hungarian campaign. Although by then it had already been known in Italy, Spain and France, infected German mercenaries returning from Hungary spread it to other countries in Europe, causing typhus outbreaks in Bohemia, Austria and Germany. It seems that typhus was known as a localised endemic disease in Hungary and the Balkans, and it is probable that the Hungarians and the Ottomans had developed some kind of immunity against it. Besides their more careful hygienic measures, this relative immunity must have been responsible for the fact that the disease did not do as much harm among the Hungarians and Turks as it did among the German mercenaries fighting in Hungary. The German mercenaries were infected via lice and, since they did not have immunity, what had earlier been a localised endemic disease soon assumed epidemic proportions and claimed its victims by the thousands.56 The Hapsburg–Ottoman Long War of 1593–1606 saw a succession of different epidemics and calamities. Plague broke out in Pozsony County in 1599, famine and plague decimated the population in several areas of Hungary and Transylvania in 1600, plague and cattle-plague broke out in Upper Hungary in 1601, and in 1602–3 famine and plague devastated Transylvania thanks to exceptionally bad crops and the war. The plague claimed numerous victims in the seventeenth century, especially among the urban population. In 1621, within four months, out of Debrecen’s 12,000 citizens 2,000 people fell victim to the plague, while in June 1645, 2,200 people died of the plague in Lőcse. The plague of 1655 killed almost half of the inhabitants of Sopron, and the number of people living in the city did not reach the pre-plague population of 4,000 until the end of the century57

Conclusion This chapter has shown some possible themes where environmental and military historians can mutually benefit from each other’s research. It argued that, although in somewhat different ways, the Ottomans and their Hapsburg rivals both had keen interest in geography and mapped their empires and resources, and possessed adequate (p.79) information as to the terrain and river systems of their lands and frontiers. The Ottomans occupied and reinforced dozens of forts along the River Danube and its tributaries, demolishing those deemed useless or impossible to defend. The chapter has also demonstrated that the formation of the Ottoman–Hapsburg military frontier was significantly influenced by Hungary’s terrain and river systems. On the other hand, war and defence-related activities—such as construction of forts, shipbuilding, cannon casting, saltpetre and gunpowder production—significantly affected the environment of the frontier by clearing its woodlands and causing considerable deforestation along and beyond the frontier. This deforestation, in turn, influenced the region’s hydraulic system and accelerated the spread of swamps, as did the establishment of dozens of ‘swamp-forts’. The spread of marshlands, in turn, negatively affected the micro-climate and created situations in which epidemics, such as morbus hungaricus, spread more rapidly.

Page 14 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary Nevertheless, wetlands and swamps also offered refuge for tens of thousands of people. Whole villages ‘disappeared’ and were relocated during campaigns, and during Ottoman, Hungarian and Hapsburg raids, as well as before the taxation season (24 April and 26 October, in the case of Ottoman Hungary). While both the Ottomans and the Hapsburgs tried to familiarise themselves with the country’s topography, the locals obviously enjoyed advantages in this regard. To follow villagers into the marshes meant great risks for soldiers and officials. The authorities had to rely on local guides, whose loyalties were often questionable. We know of several stories, some undoubtedly embellished during the centuries, telling of how these guides led Ottoman and Hapsburg soldiers and officials into swamps only to abandon them there. Whatever the historical accuracy of such stories regarding the details, they show that marshes, and geography in general, played a more important role in warfare and the daily life in the frontier than is usually acknowledged. Note. I would like to express my thanks to Professor John McNeill and the anonymous readers for their comments on an earlier version of this paper. I am also indebted to Andrew Peacock for his patience and exceptional editorial assistance and for cartographer Pieter Collet for his excellent work on the maps. Notes: Proceedings of the British Academy 156, 57–79. © The British Academy 2009. (1) See, for instance, Gábor Ágoston, ‘Information, ideology, and limits of imperial policy: Ottoman grand strategy in the context of Ottoman-Habsburg rivalry’, in Virginia H. Aksan and Daniel Goffman (eds.), The Early Modern Ottomans: Remapping the Empire (Cambridge: Cambridge University Press, 2007), 75–103; idem, ‘Birodalom és információ: Konstantinápoly, mint a koraújkori Európa információs központja’, in Gábor Hausner (ed.), Az értelem bátorsága. Tanulmányok Perjés Géza Emlékére (Budapest: Argumentum, 2005), 31–60. (2) Jusuf Gülderen, ‘Turska brodogradilišta na Dunavu i njegovim pritokama u drugoj polovini XVI veka’, in Vasa Čubrilović (ed.), Plovidba na Dunavu i njegovim pritokama kroz vekove (Belgrade, 1983), 179–91. (3) On Ottoman campaign logistics in Europe see Caroline Finkel, The Administration of Warfare: The Ottoman Military Campaigns in Hungary, 1593–1606 (Vienna: VWGÖ, 1988). (4) Gábor Ágoston, Guns for the Sultan: Military Power and the Weapons Industry in the Ottoman Empire (Cambridge: Cambridge University Press, 2005), 48–56. (5) See Palmira Brummett, ‘The river crossing: breaking points (metaphorical and “real”) in Ottoman mutiny’, in Jane Hathaway (ed.), Rebellion, Repression, Reinvention: Mutiny in Comparative Perspective (Westport, Conn., London: Praeger, 2001), 215–31. (6) On these see İsmail Hakkı Uzunçarşılı, Osmanlı Tarihi (Ankara: TTK, 1983), 3/1: 31–9. (7) Akdes Nimet Kurat, ‘The Turkish expedition to Astrakhan in 1569 and the problem of the Don–Volga Canal’, Slavonic and East European Review, 40 (1961), 7–23, which uses Ottoman, European and Russian sources and corrects some of the mistakes of earlier studies. See also Halil İnalcık, ‘The origins of the Ottoman–Russian rivalry and the Don–Volga Canal’, Annales de

Page 15 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary l’Université d’Ankara, 1 (1947), 47–110; Michael Khodarkovsky, Russia’s Steppe Frontier: The Making of a Colonial Empire, 1500–1800 (Bloomington: Indiana University Press, 2002), 115–17. (8) John McNeill, ‘Ecology and strategy in the Mediterranean: points of intersection’, in John B. Hattendorf (ed.), Naval Strategy and Policy in the Mediterranean: Past, Present, and Future (London: Frank Cass, 2000), 378. (9) Salih Özbaran, ‘Expansion in the southern seas’, in idem, The Ottoman Response to European Expansion: Studies on Ottoman-Portuguese Relations in the Indian Ocean and Ottoman Administration in the Arab Lands during the Sixteenth Century (Istanbul: Isis, 1994), 77–87. (10) BOA, Mühimme Defteri (MD) 7, no. 721, p. 258. See also Murat Şener, Nurullah İsler, and Hacı Osman Yıldırım (eds.), 7 Numaralı Mühimme Defteri: Tıpkıbasım (Ankara: Başbakanlık Devlet Arşivleri Genel Müdürlügü, 1997), 1: 258, where the original decree is published in facsimile. This decree has been known for some time. See, for example, Uzunçarşılı, Osmanlı Tarihi, 3/1, 33–4. (11) İsmail Hakkı Uzunçarşılı, ‘Sakarya Nehrinin İzmit Körfezine Akıtılmasıyla Marmara ve Karadenizin Birleştirilmesi Hakkında’, Belleten, 4/14–15 (1940), 149–74, esp. 150–7; Caroline Finkel and Aykut Barka, ‘The Sakarya River–Lake Sapanca–İzmit Bay canal project: a reappraisal of the historical record in the light of new morphological evidence’, Istanbuller Mitteilungen, 47 (1997), 429–42; Pál Fodor, ‘Between two continental wars: the Ottoman naval preparations in 1590–1592’, in idem, In Quest of the Golden Apple: Imperial Ideology, Politics, and Military Administration in the Ottoman Empire (Istanbul: Isis, 2000), 189. Recently, the project has been revitalised by young Turkish environmentalists in the form of an ambitious second ‘strait’ plan, submitted to the International Environmental Project Olympics (INEPO, Fatih College, Istanbul, Turkey), aimed at students aged 13 to 19. This suggested second ‘strait’ would follow the old Sakarya–Sapanca–İzmit canal, in order to divert the traffic of oil tankers from the Bosporus, and thus diminish the environmental hazards and pollution there. (12) On the Rhine–Maas canal project see Jonathan I. Israel, ‘A Spanish project to defeat the Dutch without fighting: the Rhine–Maas Canal, 1624–9’, in idem, Conflicts of Empires: Spain, the Low Countries and the Struggle for World Supremacy, 1585–1713 (London: Hambledon Press, 1997), 45–62. (13) Geoffrey Parker, The Grand Strategy of Philip II (New Haven: Yale University Press, 1998), 59–65. (14) Gyula Káldy-Nagy, ‘The administration of the sanjaq registrations in Hungary’, Acta Orientalia Academiae Scientiarum Hungaricae, 21 (1968), 183–4. (15) Franz Babinger, ‘An Italian map of the Balkans, presumably owned by Mehmed II, the Conqueror (1452–53)’, Imago Mundi, 8 (1951), 8–15. (16) Ahmet T. Karamustafa, ‘Military, administrative, and scholarly maps and plans’, in J. B. Harley and David Woodward (eds.), The History of Cartography, II/1: Cartography in the Traditional Islamic and South Asian Societies (Chicago: University of Chicago Press, 1992), 209– 27. (17) J. M. Rogers, ‘Itineraries and town views in Ottoman histories’, in Harley and Woodward (eds.), The History of Cartography, II/1, 231. Page 16 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary (18) Kathryn A. Ebel, ‘Representations of the frontier in Ottoman town views of the sixteenth century’, Imago Mundi 60 (2008), 1–22; cf. Palmira Brummett in Chapter 2 in this volume. (19) For the 1697 campaign, twenty road guides (kılavuzan) were hired. See BOA, Maliyeden Müdevver Defterleri (MAD) 2731, 92. (20) See, for instance, Gustav Bayerle, The Hungarian Letters of Ali Pasha of Buda, 1604–1616 (Budapest: Akadémiai Kiadó, 1991). (21) Sándor Takáts, ‘Kalauzok és kémek a török világban’, in idem, Rajzok a török világból (Budapest: Magyar Tudományos Akadémia, 1915–32), 2: 170. (22) In Istanbul, Üveys Pasha’s map was copied and its inscriptions translated into Italian for Joachim von Sinzendorf, Hapsburg ambassador to Istanbul (1578–81), who sent it back to Vienna, indicating that the Hapsburg ambassadors, too, were efficient in counter-intelligence. The Italian copy of the original Turkish map is to be found in the Viennese Archives (Österreichische Staatsarchiv, Haus-, Hof- und Staatsarchiv Turcica Karton 43. Konv. 2. fol. 50) and has been published in facsimile by Géza Pálffy, Európa védelmében (Pápa: Jókai Mór Városi Könyvtár, 2000), facsimile III. (23) Österreichische Staatsarchiv, Haus-, Hof- und Staatsarchiv, Turcica Karton 30. Konv. I. fol. 29 (22 January 1574). I am indebted to my colleague and friend, archivist István Fazekas, for this reference. See also Sándor Takáts, ‘Magyar és török íródeákok’, in idem, Művelődéstörténeti tanulmányok a 16–17. századból, ed. Kálmán Benda (Budapest: Gondolat Kiadó, 1961), 179. (24) Ferenc Szakály, ‘The Hungarian-Croatian border defense system and its collapse’, in János M. Bak and Béla K. Király (eds.), From Hunyadi to Rákóczi: War and Society in Late Medieval and Early Modern Hungary (New York: Brooklyn College, 1982), 141–58. (25) The document is published in Pál Fodor, ‘Ottoman policy towards Hungary, 1520–1541’, in idem, In Quest of the Golden Apple, 146–51. (26) Gábor Ágoston, ‘Habsburgs and Ottomans: defense, military change and shifts in power’, Turkish Studies Association Bulletin, 22 (1998), 126–41; Géza Pálffy, ‘The origins and development of the border defence system against the Ottoman Empire in Hungary (up to the early eighteenth century)’, in Géza Dávid and Pál Fodor (eds.), Ottomans, Hungarians and Habsburgs in Central Europe: The Military Confines in the Era of Ottoman Conquest (Leiden: Brill, 2000), 3–69; Géza Pálffy, ‘Die Türkenabwehr in Ungarn im 16. und 17. Jahrhundert: ein Forschungsdesiderat’, Anzeiger der philosophisch-historischen Klasse, 137 (2002), 99–131. (27) István Horváth, ‘Ottoman military construction in Esztergom’, in Ibolya Gerelyes and Gyöngyi Kovács (eds.), Archaeology of the Ottoman Period in Hungary (Budapest: Hungarian National Museum, 2003), 75–87. (28) Regarding the Ottoman fortress system (its development, military force etc.) in Hungary see Klára Hegyi’s threevolume definitive work: A török hódoltság várai és várkatonái (Budapest: História, 2007), which, however, was unavailable at the time of the writing of this chapter. See also her earlier studies in English: Klára Hegyi, ‘Balkan garrison troops and soldier-peasants in the Vilayet of Buda’, in Gerelyes and Kovács (eds.), Archaeology of the Ottoman Period in Hungary, 23–40; Hegyi, ‘The Ottoman network of fortresses in Hungary’, in Dávid and Fodor (eds.), Ottomans, Hungarians, and Habsburgs, 163–93; Hegyi, ‘The Ottoman military force in Page 17 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary Hungary’, in Géza Dávid and Pál Fodor (eds.), Hungarian-Ottoman Military and Diplomatic Relations in the Age of Süleyman the Magnificent (Budapest: ELTE, 1994), 131–48. (29) The Ottoman campaigns of 1551–2 have been the subjects of numerous detailed studies, and the history of many of the aforementioned castles is discussed in separate monographs, which cannot be listed here. On Ottoman building activities in the province of Temeşvar see Ferenc Csortán, ‘Ottoman architecture in the Vilayet of Temesvár’, in Gerelyes and Kovács (eds.), Archaeology of the Ottoman Period in Hungary, 187–96. (30) Fodor, ‘Ottoman policy towards Hungary’, 150. (31) Gyula Káldy-Nagy, A csanádi szandzsák 1567. és 1579. évi összeírása (Szeged: Csongrád Megyei Levéltár, 2000), 7. (32) Gábor Ágoston, ‘The costs of the Ottoman fortress-system in Hungary in the sixteenth and seventeenth centuries’, in Dávid and Fodor (eds.), Ottomans, Hungarians and Habsburgs, 202–3; József Kelenik, ‘A Kanizsa elleni végvidék katonai erejének változásai 1633–1638’, Zalai Gyűjtemény, 36 (1995), 5–51. The fourteen Hungarian forts were Körmend, Egerszeg, Pölöske, Kapornak, Egervár, Kemend, Lövő, Magyarosd, Tótfalu, Kiskomárom, Zalavár, Szentgrót, Szentgyörgy and Bér. On Kanije and Uyvar in English see also Mark Stein, Guarding the Frontier: Ottoman Border Forts and Garrisons in Europe (London: I. B. Tauris, 2007). See my review in the Journal of the Economic and Social History of the Orient, 52 (2009), 159–63. (33) This section of the paper is based on my earlier studies. See, for example, Gábor Ágoston, A hódolt Magyarország (Budapest: Adams, 1992), 92–125; Gábor Ágoston and Teréz Oborni, A tizenhetedik század története (Budapest: Pannonica, 2000), 86–104. A short summary in English is Gábor Ágoston, ‘Ottoman conquest and the Ottoman military frontier in Hungary’, in Béla Király and László Veszprémy (eds.), A Millennium of Hungarian Military History (Boulder, Co.: Atlantic Research and Publications, 2002), 103–7. (34) Bálint Hóman and Gyula Szekfű, Magyar történet, V: Gyula Szekfű, A tizenhetedik század (Budapest: Királyi Magyar Egyetemi Nyomda, n.d), 7. See also Steven Béla Várdy, ‘The changing image of the Turks in twentieth- century Hungarian historiography’, in idem, Clio’s Art in Hungary and in Hungarian-America (Boulder, Co.: East European Monographs, 1985), 147–70, who also quotes Szekfű. (35) Ágnes R. Várkonyi, ‘Környezet és végvár’, in Tivadar Petercsák and Jolán Szabó (eds.), Végvárak és régiók a XVI–XVII. Században (Eger: Heves Megyei Múzeum, 1993), 18. (36) See, for example, Michael Williams, Deforesting the Earth: From Prehistory to Global Crisis (Chicago and London: University of Chicago Press, 2003), 168–209. See also John McNeill, ‘Woods and warfare in world history’, Environmental History, 9 (2004), 388–410. (37) Recent studies estimate the population of Hungary at 3.1–3.5 million around 1500, at 3.5–4 million in the 1570s–1590s, and at 4 million around 1720. See the essays by András Kubinyi, Géza Dávid, Vera Zimányi and Teréz Oborni, in József Kovacsics (ed.), Magyarország történeti demográfiája, 896–1995 (Budapest, 1997); see also Géza Dávid, Studies in the Demographic and Administrative History of Ottoman Hungary (Istanbul: Isis, 1997). (38) Károly Kaán, Alföldi kérdések. Erdők és vizek az Alföld kérdéseiben (Budapest: Stádium, 1939). Page 18 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary (39) István Sugár, ‘Az egri vár építőanyagainak beszerzési helyei 1548–1564’, in Tivadar Petercsák and Ernő Pető (eds.), Végvár és környezet (Eger: Heves Megyei Múzeum, 1995), 178. (40) From administrative and military-administrative points of view the Viennese and Hungarian authorities divided northern Hungary into two parts along the line stretching from the western borders of Gömör-Kishont (or Szepes) county through the Tisza River. Territories to the west of this line (and thus closer to Vienna, the point of reference) were called Lower Hungary, while those to the east were Upper Hungary. Counties under the command of the captain-general of Upper Hungary included (from west to east) Szepes, Gömör-Kishont, Torna, Borsod, Heves, Sáros, Abaúj, Zemplén, Szabolcs, Ung, Bereg, Ugocsa and Szatmár. See Pálffy, ‘The origins and development of the border defence system’, 68. (41) The closest data I could find for cartloads used for transporting wood is from Veszprem and comes from 1696: 1 cartload = 1.7 cubic metres. See István Bogdán, Magyarországi űr-, térfogat-, súly- és darabmértékek 1874–ig (Budapest: Akadémiai Kiadó, 1991), 421. This seems reasonable. Calculating with the conservative 350 kg per m3 for seasoned and dry pine and 600 kg per m3 for oak, one cart would be between 595 kg and 1,020 kg, which compared with other data is more than plausible. Of course, these figures could have been much higher depending on the types of pine and oak wood, and their moisture content could also influence the calculation. (42) Unfortunately, we do not know the number of soldiers stationed in these forts. In 1670, their number on paper was 2,817. However, according to the ‘decree of reduction’ of 1671 this number had to be reduced to 1,000 soldiers. According to an official report dated 10 October 1675, the king still paid 1,562 soldiers (1,168 hussars and 394 hajdús) in the Upper Hungarian castles. See István Czigány, Reform vagy kudarc? Kísérletek a magyarországi katonaság beillesztésére a Habsburg Birodalom haderejébe (Budapest: Balassi Kiadó, 2004), 125–7. (43) According to Williams, in pre-industrial northern Europe annual per capita consumption could have averaged 2 tons. He also states that rural peasant families in seventeenth-century southern Germany could have reckoned to use 25–50 m3 of wood per annum (Williams, Deforesting the Earth, 181–3). See also Paul Warde, Ecology, Economy and State Formation in Early Modern Germany (Cambridge: Cambridge University Press, 2006), 267, whose highest per capita fuelwood consumption estimate is only 3.9 m3, and who suggest that 1.2 m3 of solid timber consumption for fuel is a reasonable estimate. (44) See J. R. McNeill, The Mountains of the Mediterranean World: An Environmental History (Cambridge: Cambridge University Press, 1992), 137. (45) See Lajos Rácz, ‘Variations of climate in Hungary (1540–1779)’, in Burkhard Frenzel, Christian Pfister and Birgit Gläser (eds.), European Climate Reconstructed from Documentary Data: Methods and Results (Stuttgart: Gustav Fischer Verlag, 1992), 132. (46) See Ágoston, ‘Ottoman conquest’, 95. (47) Assuming a production of 25 tons per km2 of managed woodland. See Williams, Deforesting the Earth, 183. (48) Estimates and calculations are based on Ágoston, Guns for the Sultan, 135–8, 151–8. (49) The saltpetre works in the province of Karaman burned on average some 15–16 kg firewood to produce 1 kg of saltpetre in the 1640s. BOA, MAD 5685, pp. 2–6. Page 19 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Where Environmental and Frontier Studies Meet: Rivers, Forests, Marshes and Forts along the Ottoman–Hapsburg Frontier in Hungary (50) Géza Pálffy, ‘Elképzelések a török hódoltság elpusztításáról a XVI–XVII században’, in Gyöngyi Kovács (ed.), Quasi Liber et Pictura: Tanulmányok Kubinyi András hetvenedik születésnapjára (Budapest: ELTE, 2004), 387–401. (51) Lajos Fekete, ‘A hódoltságkori törökség Magyarországra vonatkozó földrajzi ismeretei’, Hadtörténelmi Közlemények, 31 (1930), 15–16. (52) Fekete claims that 1 mil = 1/2 farsah, that is the distance that a horseman could travel in half an hour. 1 farsah was about 6,230 metres in Hungary, and about 5,760 metres in the Arab lands (Fekete, A hódoltságkori törökség’, 12). According to Walter Hinz (Islamische Masse und Gewichte umgerechnet ins metrische System (Leiden: Brill, 1955), 63) 1 mil = 1/3 farsah, that is about two kilometres. (53) Fekete, ‘A hódoltságkori törökség’, 141–2. (54) In English see John Stoye, Marsigli’s Europe, 1680–1730: The Life and Times of Luigi Ferdinando Marsigli, Soldier and Virtuoso (New Haven: Yale University Press, 1994). (55) Lajos Rácz, ‘The climate history of central Europe in the modern age’, in József Laszlovszky and Péter Szabó (eds.), People and Nature in Historical Perspective (Budapest: Central European University Press, 2003), 229–45. (56) Tibor Győry, Morbus hungaricus (Budapest, 1910); Hans Zinsser, Rats, Lice and History (New York, 1960), 162–71, 202–5. (57) Ágoston and Oborni, A tizenhetedik század, 99–104.

Access brought to you by:

Page 20 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Conquest of Arabia and the Syrian Hajj Route

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

The Ottoman Conquest of Arabia and the Syrian Hajj Route ANDREW PETERSEN

DOI:10.5871/bacad/9780197264423.003.0004

Abstract and Keywords This chapter discusses the Ottoman presence in Arabia, from the initial conquests in the sixteenth century up to the end of the eighteenth century, before the French invasion of Egypt followed by the rise of Mehmed Ali Pasha fundamentally changed the balance of power in the region. In particular, it focuses on the evolution of the Syrian hajj route and compares it to wider developments elsewhere in the Arabian Peninsula. The first part of the chapter focuses on the expansion and contraction of Ottoman rule in Arabia. The second part focuses on the Hijaz and the fortification of the hajj route linking it to Mecca. Keywords:   Ottoman Empire, Arabia, Mehmed Ali Pasha, Syrian hajj, Arabian Peninsula, Hijaz

THIS CHAPTER WILL DISCUSS THE OTTOMAN PRESENCE IN ARABIA, from the initial conquests in the sixteenth century up to the end of the eighteenth century, before the French invasion of Egypt followed by the rise of Mehmed Ali Pasha fundamentally changed the balance of power in the region. In particular, it will focus on the evolution of the Syrian hajj route and will compare it to wider developments elsewhere in the Arabian Peninsula. The first part of the chapter will focus on the expansion and contraction of Ottoman rule in Arabia and the second part will focus on the Hijaz and the fortification of the hajj route linking it to Mecca.

Ottoman Rule in Arabia The Ottoman experience in each part of Arabia was different, although there are certain trends which are noticeable throughout the region and can conveniently be divided into two main periods. These can be summarised as Period One, 1516–1700 and Period Two, 1700–1800. Period One: Conquest, Confrontation and Defeat Before 1516, the Ottomans had no direct connection with Arabia and their knowledge of the region derived either from the accounts of individual pilgrims such as Şehzade Korkud1 or from contacts with the Mamluk regime that ruled Syria, Egypt and the Hijaz, such as Sultan Bayezid’s naval alliance of 1510. However, this all changed when the Ottomans defeated the Mamluks at the Battle of Marj Dabiq near Aleppo in 1516 and subsequently took possession of both Page 1 of 12 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Conquest of Arabia and the Syrian Hajj Route Damascus and Cairo. For the first time, the (p.82) Ottomans came into direct contact with northern and western Arabia and inherited the ambiguous Mamluk relationship between Turkic rulers and local Arab notables. The first part of Arabia to succumb to Ottoman rule was the Hijaz, which accepted Ottoman suzerainty in 1517 when the Sharif of Mecca, Barakat II, agreed to send his thirteen-year-old son, Abu Numayy, to Istanbul.2 Following the occupation of the Hijaz, the Ottomans extended their rule into Yemen, occupying Aden in 1538.3 Twelve years later, the Ottomans were also able to add north-eastern Arabia to their list of territories when they occupied al-Hasa (Lahsa) oasis, with its capital at Hufuf. It is likely that the Ottomans intended to extend their rule into south-eastern Arabia, as can be seen from their abortive attempt to occupy the straits of Hurmuz.4 The Ottomans claimed that their chief interest in Arabia was the protection of the Holy Cities of Mecca and Medina, although in reality they may have been more concerned with the potential financial benefits. Their main object of interest in the Peninsula was control of the Indian spice trade, which had been seized by the Portuguese in the late fifteenth century after they had found a sea route to India via the Cape. As early as 1538 a truce had been achieved with the Portuguese, establishing respective zones of activity. However, the relationship with the Portuguese remained tense and there were often attempts to alter the status quo, such as the Ottoman attempts to capture Bahrain or Hurmuz. The Portuguese had superior naval power with larger and better equipped ships, while the Ottomans had the advantage of land-based forces in the region. Thus by the middle of the sixteenth century, the Ottoman presence in Arabia was established in three distinct areas: the Hijaz, Yemen and al-Hasa (Figure 4.1). The pattern of Ottoman control in each of these areas was complex and often incomplete; it appears that the most successful Ottoman province was the Hijaz, even though economically it had considerably less to offer the Ottoman centre in terms of agricultural revenues or trade. In this chapter I will suggest that the principal reason for the successful retention of the Hijaz was the hajj from Damascus, and that this was maintained by effective roadside security, for which there is considerable archaeological evidence. Before discussing the Syrian hajj route in detail and its implications for Ottoman control in the region, it is important to set it within the wider context of Arabia from the sixteenth to the end of the eighteenth century. During the seventeenth century the Ottomans suffered a number of defeats, the most notable of which was their defeat at Vienna in 1683, which was followed by an aggressive counter-offensive by the Hapsburg Empire.5 On the eastern front, the (p.83) Ottomans led an intensified campaign against the Safavids, which drew manpower from other parts of the empire and in particular from Syria and Arabia. It may well be that the reduced Portuguese threat in the Indian Ocean and the Gulf meant that the Ottomans attached less emphasis to their Arabian territories. In any case, by the middle of the seventeenth century they had lost control of two of their three Arabian provinces.

Page 2 of 12 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Conquest of Arabia and the Syrian Hajj Route Ottoman involvement in Yemen started in 1516 soon after their conquest of Egypt, when an Ottoman force defeated the ruling Tahirids near Zabid. However, full-scale (p. 84) conquest did not take place until 1538, when a large Ottoman expedition on its way to India arrived at Karaman island in the Red Sea, capturing Aden in 1538, Ta‘izz in 1539, and finally San‘a’ in 1547.6 Ottoman rule was based on several fortresses, and the degree of Ottoman control beyond these strongholds was always debatable. The Ottomans’ control of much of Yemen was lost as early as 1566 when the Zaydi Imam Mutahhar expelled the Turkish garrison in San‘a’. Turkish rule was later re-established after an expensive campaign led by Sinan Pasha, the Ottoman governor of Egypt.7 A series of rebellions in the early 1600s led to a final Ottoman defeat and withdrawal in

Figure 4.1. Map of Arabia in the sixteenth century showing the extent of the Ottoman Empire and the area of Portuguese influence.

1636.8 The Ottoman hold on Yemen was always precarious, partly because of its geographical remoteness from Istanbul and partly because many of its inhabitants were Shi‘a. Communication with the rest of the Ottoman Empire was maintained through Egypt via the Red Sea and through Syria via the Hijaz and the overland pilgrimage routes to Mecca. The Ottoman occupation of al-Hasa was in theory less problematic than that of Yemen. Although there was a Shi‘a presence within the province, there were also large numbers of Sunnis who apparently coexisted peacefully9 Unlike Yemen, there was no developed form of political control to rival the Ottomans, and the area had previously been ruled by local tribal chiefs. Like Yemen, al-Hasa was strategically important, as it not only protected the southern approach to Iraq but also served as a check on the expansion of Safavid power to the west side of the Arabian Gulf. In addition, communications within the province were relatively good, with caravan routes running to the coast at al-Qatif to Basra and Iraq in the north and westwards to Najd and the Hijaz. The province itself was agriculturally rich by Arabian standards and was capable of producing surpluses for the Ottoman Empire.10 However, the Ottomans were not able to retain control of this area beyond the first half of the seventeenth century. The reasons for the loss of this province are unclear, although one recent writer has suggested that the Ottomans did not engage with the local population, extorted punitive taxes, and relied on large fortresses rather than establishing a presence in the landscape.11 Period Two: The Rise of the Wahhabis By the early eighteenth century, direct Ottoman control in Arabia was limited to the Hijaz, where it was frequently being threatened both by Bedouin and Wahhabi forces. (p.85) At this time, the Ottoman provinces peripheral to Arabia were increasingly under the control of strongmen such as Ali Bey in Egypt, Zahir al-‘Umar in Palestine and Büyük Süleyman Pasha in Iraq. These were effectively independent rulers who sought legitimacy and titles from the Ottoman sultan in return for nominal allegiance. Sometimes the actions of these rulers were in accordance with Ottoman policy and sometimes in direct conflict, although there was rarely outright opposition Page 3 of 12 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Conquest of Arabia and the Syrian Hajj Route to the concept of the Ottoman Empire, which would have entailed a rejection of the caliphate, one of the foundations of Sunni Muslim unity. In these circumstances, it is not surprising that the strongest opposition to the Ottoman presence in Arabia came from an area which had never been conquered by the Turks and was expressed in terms of religious revivalism. Najd comprised a series of settlements based on a complex of oases located on a plateau in the centre of Arabia. Although Najd was outside the area formally conquered by the Ottomans, it was surrounded by territories that had been occupied by the Turks in the sixteenth century and had considerable trading links with the empire. To the north-east was the Ottoman province of Basra, to the north-west was the province of Damascus, to the west the Hijaz and Yemen. The only area with no Ottoman control was the south and south-east, comprising the sparsely inhabited Empty Quarter. As a result of its geographical position, most of the inhabitants of Najd would have had some contact with the Ottoman Empire either through trade or warfare, but the region clearly remained independent. Prior to the eighteenth century, this was an area with little security, where different tribes or family groups raided each other’s settlements and livestock in order to gain some ascendancy. In the eighteenth century, the Muslim cleric ‘Abd al-Wahhab called for a return to fundamental Islamic values, specifically the abandonment of all local ancestral shrines and any religious practices which were perceived to be innovations since the time of the Prophet Muhammad.12 In 1745 Muhammad b. Sa‘ud, the ruler of al-Diriya, gave refuge to ‘Abd al-Wahhab and began an alliance which was to transform the Arabian Peninsula and lay the foundations of the future Saudi state. The arrangement between Ibn Sa‘ud and ‘Abd al-Wahhab enabled the former to continue raids and conquests against neighbouring settlements and tribes with the added bonus of religious sanction, which elevated his raids to the status of a jihad. This provided ‘Abd al-Wahhab with a sure supply of new followers who could see the advantages of the new doctrine, which ensured a steady supply of food and money as well as increased personal security.13 (p.86) Two groups were particularly at risk from the rising Wahhabi state, the Shi‘a of al-Hasa and Iraq and the people of the Hijaz, whose Sunni religious beliefs were closer to those of the Ottoman Turks. In the first wave of their expansion the Wahhabis raided and eventually conquered most of al-Hasa oasis, destroying ancestral shrines and cutting down sacred trees wherever they found them.14 From al-Hasa the Wahhabis raided deep into Iraq and in 1802 sacked the Shi‘i shrine at Karbala.15 The Wahhabis had for some time tried to expand their influence to the Hijaz, and as early as 1745 ‘Abd al-Wahhab had sent missionaries to Mecca. During the next half-century relations between the Wahhabis and the Hijaz gradually deteriorated, and in 1790 the Sharif of Mecca sent an expedition against Najd. Although the expedition was ultimately unsuccessful, the Wahhabis were not able to take advantage of their superior numbers to occupy the Hijaz until 1805, when Ottoman resources were engaged elsewhere (for example, with the French occupation of Egypt).

The Syrian Hajj Route If we compare the development of the Syrian hajj route with Ottoman rule elsewhere in Arabia as outlined above, we can see that a very different pattern emerges. Before describing the developments on the route in detail it is necessary to set them within their historical and cultural context. Historically the Syrian hajj route was the most important of the routes connecting Mecca with the Muslim world. During the early Islamic period, it was used by the Arab armies invading Syria, and under the Umayyads it remained the principal route connecting the caliphs to Mecca and their Arabian heritage.16 Although the Abbasid revolution was coordinated from one of the stations on the Syrian hajj route (Hummayma in southern Jordan), the

Page 4 of 12 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Conquest of Arabia and the Syrian Hajj Route new regime established its capital in Iraq and the Syrian hajj route ceased to be an imperial axis linking the political capital with the religious centre. Subsequent events, such as the rise of the Fatimids in Egypt in the tenth century and the Crusader invasions, meant that the Syrian route was of little importance. Under the Mamluks (1260–1516) some stability returned to the region and the Syrian hajj route regained some prominence.17 However, it was the Ottoman conquest of Syria and Egypt in 1516 that ensured that it became once again a central axis of empire, forming part of an overland route linking Mecca (p.87) directly with Istanbul.18 As a result of the new importance of the route a series of facilities was constructed by the Ottoman Sultans. Between Istanbul and Damascus the route was provided with hans, mosques and other travellers’ facilities,19 while to the south the security of the route was improved with a series of small forts. In Damascus itself, a magnificent pilgrimage complex was built, known as the Tekiyye, which included hans, mosques and a medrese built according to a design of the Ottoman imperial architect Sinan.20 The fortification of the Syrian hajj route may be divided into two phases, which parallel those of the conquest of Arabia as a whole. During the first phase a series of forts were established at important water sources along the route, and during the second phase the entire distance between Damascus and Mecca was provided with fortified stopping places. Phase One: Sixteenth Century The first three forts were built within five years of the Ottoman conquest during the reign of Sultan Selim I and were all located less than 100 kilometres from Damascus.21 The largest of these forts was that at Muzayrib; it comprised a large square enclosure (approximately 80 metres each side) with square corner towers and a range of rooms built along the south side of the entrance.22 In 1563 it had a garrison of fifty-six men and was designed to act as an assembly point and administrative centre for the hajj.23 The other two forts, Sanamayn and Tall Far‘un (Mafraq), were on a smaller scale, comprising small square courtyard buildings measuring approximately 20 metres each side. This became the standard design for forts on the Syrian hajj route, which continued to be built with minor variations up to the end of the eighteenth century. During the reign of Sultan Süleyman the Magnificent, forts were built further south, into the area of present-day Jordan and the Hijaz. In 1534 the first fort to be built by Sultan Süleyman was at Ukhaydhir, approximately halfway between Damascus and Medina. According to Mehmed Edib, this fort was built as a specific response to (p.88) attacks by the Bani Lam Bedouin,24 but it may also have been built to mark the halfway point of the hajj route. Other forts built during the reign of Süleyman include Qatrana (Figure 4.2), Ma‘an (Figures 4.3–4.4), Dhat al-Hajj and Tabuk. Each of these forts was built at a place of long-term strategic importance where there is evidence of earlier settlement and plentiful water resources. For example, the fort at Qatrana is built next to a large Roman reservoir, Ma‘an is an ancient town with Roman and early Islamic (Umayyad) water reservoirs,25 Dhat al-Hajj lies in a hollow where there is plentiful fresh water, Tabuk is a major oasis with plentiful water and a long history of settlement, and Ukhaydhir was built next to a large pool.26 Another fort which may also be attributed to Sultan Süleyman is at al-‘Ula, which was also located in an area of ancient settlements. The process of fortress construction continued under Selim II (1566–74), with forts built at ‘Unayza, north of Ma‘an, and at Hadiyya, forty-two hours’ journey north of Medina.

Page 5 of 12 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Conquest of Arabia and the Syrian Hajj Route (p.89)

Figure 4.2. South face of sixteenth-century fort at Qatrana. Note gun loops and machicolation above door.

Figure 4.3. Ground plan of the sixteenthcentury fort at Ma‘an.

Page 6 of 12 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Conquest of Arabia and the Syrian Hajj Route (p.90) Phase Two: Eighteenth Century During the 160 years between 1574 and the 1730s, there is no evidence of fortress construction, although the existing forts appear to have been repaired and remained in use.27 In the eighteenth century, eight new forts were built, at al-Hasa, Dar alHamra’, Fassu‘a (Zahr al-‘Aqaba), Mudawwara, Nakhlatayn, Qalandariyya, Mada’in Salih and Dab‘a (Figure 4.5). Although the design of these forts was similar to earlier ones, their locations marked a departure from earlier practice. Instead of being located at historical strategic sites with plentiful water resources, the new forts were built at sites with scant water resources and little historical significance. It is possible that some of these forts were actually repairs of earlier buildings. Thus an inscription (now lost)

Figure 4.4. South face of the recently restored sixteenth-century fort at Ma‘an with original entrance. Only one original gun loop survives; the others have been converted to rectangular windows.

at Qal‘at Balqa states that it was restored in

28

1767. The fortresses now represented a continuous line of fortification stretching from Damascus to Medina with forts at regular intervals of a day’s march or less (that is, 40 kilometres). In addition to the forts, the Ottomans also invested in the roads themselves, and sections of paved road are still visible at al-Mudawwara and Qal‘at al-Hasa, where there is also a bridge. As well as forming a continuous line of fortification, the hajj route also formed part of a wider network of fortification, which on the one hand included Ottoman fortresses in Palestine29 and on the other fortresses on both sides of the Red Sea.30 The eighteenth-century forts had new features that modified the basic design of the forts, which had remained the same from the 1500s. The main new feature was the projecting corner tower, which was located at first-floor level and rested on projecting corbels (Figures 4.6–4.8). Projecting corner towers had the advantage of enabling the soldiers inside the fortress to cover the entire area beneath it. Forts without projecting corner turrets (or with conventional corner towers which rise from ground level) would not have provided cover of the area immediately beneath the fortress walls, which was particularly important to prevent them being undermined. Another innovation of the (p.91) eighteenth century was the provision of a large number of smaller openings for guns (see Figure 4.7). Whereas the sixteenth-century forts had a few large embrasures designed for small cannon, the eighteenth-century forts had a large number of slits, many of which were nearly invisible on the exterior. These slits were designed for the use of smaller hand-held guns and provided the fortress soldier with more protection.

Page 7 of 12 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Conquest of Arabia and the Syrian Hajj Route (p.92)

Figure 4.5. Map of the hajj route showing locations of forts in the eighteenth century.

Figure 4.6. First-floor plan of the eighteenth-century fort of Fassu‘a with projecting corner towers.

Page 8 of 12 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Conquest of Arabia and the Syrian Hajj Route (p.93)

Conclusion Both the increased frequency of fortresses on the hajj route and the more sophisticated design indicate the seriousness of the Ottoman sultans’ intention to provide a secure route to the holy cities. While the eighteenth century is often regarded as the period when the empire was in decline, the maintenance of the hajj route and construction of additional forts suggest an active policy of retaining and increasing hold on territory. In many respects this reflects the increasing number and potency of Bedouin attacks on the Ottoman hajj route. Whereas in the sixteenth century it had often been sufficient to fire a cannon to disperse the Bedouin, by the eighteenth century the latter had acquired hand-held guns which could be fired from horseback. The most devastating display of Bedouin firepower came in 1757, when the Bani Sakhr massacred 20,000 pilgrims at al-‘Ula in the latest of a series of increasingly

Figure 4.7. Qal‘at Fassu‘a. North-west corner tower. Note the numerous small gun slits.

ferocious attacks.31 In spite of such serious setbacks, the Ottomans continued to maintain their ability to send hajj caravans to Mecca, which now appeared to be one of the few places in Arabia to acknowledge Ottoman rule. If we compare the Hijaz with the other Ottoman territories in Arabia, one of its Figure 4.8. Qal‘at Fassu‘a. North face main advantages was the existence of a showing projecting corner towers. The large secure fortified land route linking it to the opening above the entrance was originally a machicolation which later collapsed. centre of the empire. The existence of a secure road for pilgrims also ensured that there (p.94) was a military route, which enabled the Ottomans to maintain an active presence in Mecca and the Hijaz in general. Often the hajj would be combined with a military expedition, as in 1672 when Evliya Çelebi accompanied Hüseyin Pasha and 8,000 troops to demand redress for the killing of the Ottoman governor of Jeddah.32 Although al-Hasa had several caravan routes linking it to Basra in the north and the Hijaz in the west, these routes were not maintained by the Ottomans and there was no protection for those using the routes. Indeed, the hajj route from al-Hasa to Mecca was closed for most of the period of the Ottoman occupation because of a fear that Shi‘a pilgrims would use the road. Anyone coming by this route was liable to be imprisoned by the governor of al-Hasa.33 Similarly,

Page 9 of 12 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Conquest of Arabia and the Syrian Hajj Route although Yemen had long-established sea routes linking it to Egypt, the land routes needed to enable the movement of armies were less secure. It is notable that Sinan Pasha had first to secure Mecca as a base for his reconquest of Yemen and then marched his men overland from there to San‘a’. Although the route between Damascus and Mecca was well fortified, that from Mecca to Yemen was less well developed and did not have the fortified posts of the Syrian hajj route. It is possible that if the Ottomans had extended their fortified road network to other parts of Arabia they might have retained their Arabian territories and contained the Wahhabi threat. Certainly, when al-Hasa was re-occupied in the 1870s, Midhat Pasha recommended the construction of roads with fortified posts, both as a way of improving security for the inhabitants of the province and as a way of maintaining control.34 Similarly, the construction of the Hijaz railway on the line of the Syrian hajj route was regarded as of major strategic significance by both the Ottomans and their enemies. Whether these fortified routes amount to a frontier marking the boundary between the power of a state and the area outside is a matter of debate, although there is certainly a strong resemblance to modern political frontiers. Notes: Proceedings of the British Academy 156, 81–94. © The British Academy 2009. (1) Nabil al-Tikriti, ‘The hajj as justifiable self-exile: Şehzade Korkud’s Wasīlat al-ahbāb (915– 916/1509–1510)’, Al-Masaq: Islam and the Medieval Mediterranean, 17 (2005), 125. (2) J.-L. Bacqué-Grammont and A. Kroel, Mamlouks, Ottomans et Portugais en Mer Rouge: L’affaire de Djedda en 1517 (Cairo: IFAO, 1988), 17–18 and 24. (3) G. R. Smith, ‘al-Yaman, 3. History’, EI2 11: 271–4. (4) Salih Özbaran, ‘The Ottoman Turks and the Portuguese in the Persian Gulf, 1534–1581’, in R. B. Serjeant and B. L. Bidwell (eds.), Arabian Studies IV (London: Hurst and Company, 1978), 125–32. (5) Caroline Finkel, Osman’s Dream: The Story of the Ottoman Empire, 1300–1923 (London: John Murray, 2005), 288. (6) Smith, ‘al-Yaman’, 271–4. (7) C. K. Smith, ‘Kawkaban, the key to Sinan Pasha’s campaign in the Yemen (March 1569-March 1571)’, Proceedings of the Seminar for Arabian Studies, 32 (2002), 287–94. (8) Smith, ‘al-Yaman’, 273. (9) U. Haarman, ‘Murtada b. Ali b. Alawan’s journey through Arabia’, in A. M. Abdullah, S. alSakar and R. T. Mortel (eds.), Sources for the History of Arabia (Riyad: Riyad University Press, 1979) 1/2: 248. (10) F. S. Vidal, The ARAMCO Reports on al-Hasa and Oman 1950–1955, II (Dharan: Aramco, [1950] 1990). (11) A. Vassiliev, The History of Saudi Arabia (London: Saqi Books, 1998), 59–61. (12) Ibid., 78.

Page 10 of 12 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Conquest of Arabia and the Syrian Hajj Route (13) Ibid., 83–8. (14) Vassiliev, History of Saudi Arabia, 88–91. (15) Ibid., 96–8. (16) Y. Frenkel, ‘Road stations in southern Bilad al-Sham in the 7th–8th centuries’, Aram, 8 (1996), 177–88. (17) A. ‘Ankawi, ‘The pilgrimage to Mecca in Mamluk times’, in R. B. Serjeant and B. L. Bidwell (eds.), Arabian Studies I (London: Hurst and Company, 1974), 116–70; A. D. Petersen, ‘The medieval hajj route through Syria and Jordan’, in K. D’hustler and A. van Tongerloo (eds.), Festschrift Professor Urban Vermeulen: Fatimid, Ayyubid and Mamluk Studies (Leuven: Peeters, 2007). (18) For a full discussion of the Ottoman period see Suraiya Faroqhi, Pilgrims and Sultans: The Hajj under the Ottomans (London: I. B. Tauris, 1994). (19) J. Sauvaget, ‘Les caravanserais Syriens du Hadjdj de Constantinople’, Ars Islamica, 4 (1937), 98–121. (20) G. Goodwin, Ottoman Architecture (London: Thames and Hudson, 1971), 255–7. (21) M. A. Bakhit, The Ottoman Province of Damascus in the Sixteenth Century (Beirut: Librairie du Liban, 1982), 97–8. (22) M. Meinicke, Patterns of Change in Islamic Architecture: Local Traditions Versus Migrating Artists (New York: New York University Press, 1996), 47; M. Kiel, ‘The carvansaray and civic centre of Defterdar Murad Çelebi in Ma‘arat an-Numan and the külliye of Fatih Sinan Pasha in Sa‘sa’’, in Seven Centuries of Ottoman Architecture: A Supra-National Heritage’ (Istanbul: Yem Yayın for Istanbul Chamber of Architects, 2001), 104–5. (23) Bakhit, Ottoman Province, 113. (24) Mehmed Edib b. Mehmed Derviş, Menasik-ı hacc-ı şerif (Istanbul, 1232); trans. M. Bianchi, Itinéraire de Constantinople à la Mecque (Paris: Société de Géographie, 1825), 132. (25) D. Genequand, ‘Ma‘an, an early Islamic settlement in southern Jordan: preliminary report on a survey in 2002’, Annual of the Department of Antiquities of Jordan, 47 (2003), 25–35. (26) Bakhit, Ottoman Province, 16. (27) K. Barbir, Ottoman Rule in Damascus 1708–1758 (Princeton: Princeton University Press, 1980), 137. (28) R. E. Brünnow and A. von Domaszewski, Die Provincia Arabia (Strasbourg: Karl J. Tübner, 1905), 2: 74–6. (29) A. D. Petersen, ‘The fortification of the pilgrimage route during the first three centuries of Ottoman rule (1516–1757)’, in K. ‘Amr, F. Zayadine and M. Zaghoul (eds.), Studies in the

Page 11 of 12 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Conquest of Arabia and the Syrian Hajj Route Archaeology of Jordan V (Amman: Department of Antiquities, 1995), 299–305; Idem, ‘Qal‘at Ras al-‘Ayn: a sixteenth century Ottoman fortress’, Levant, 30 (1998), 97–112. (30) C. Le Quesne, Quseir: An Ottoman and Napoleonic Fortress on the Red Sea Coast of Egypt (Cairo: AUC, 2007). (31) Barbir, Ottoman Rule, 200–1. (32) Evliya Çelebi, Seyahatname, 9: 286 ff.; K. M. Kortepeter, ‘A source for the history of OttomanHijaz relations: the Seyahatname’, in Abdullah, al-Sakar and Mortel (eds.), Sources for the History of Arabia, 1/2: 233. (33) Jon E. Mandaville, ‘The Ottoman province of al-Hasa in the sixteenth and seventeenth centuries’, Journal of the American Oriental Society, 90 (1970), 498. (34) Frederick Anscombe, The Ottoman Gulf: The Creation of Kuwait, Saudi Arabia, and Qatar (New York: Columbia University Press, 1997), 35.

Access brought to you by:

Page 12 of 12 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

‘Aqaba Castle in the Ottoman Period, 1517–1917

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

‘Aqaba Castle in the Ottoman Period, 1517–1917 DENYS PRINGLE

DOI:10.5871/bacad/9780197264423.003.0005

Abstract and Keywords For most of the period during which 'Aqaba belonged to the Ottoman Empire, the precise nature of its frontier status needs to be nuanced, since, in theory at least, all of the provinces adjoining it formed part of the same political unit, and the Red Sea itself was a largely Ottoman lake. In practice, however, Ottoman political and military control in the Syrian and Arabian deserts was often tenuous and reliant on individual deals struck with Bedouin leaders, often within the context of the sultan's fulfilment of his obligation to protect the pilgrimage to the Holy Places of Arabia. From this perspective, forts along the pilgrimage routes such as that at 'Aqaba, were none the less frontier places, mediating relations, whether peaceable or belligerent, between populations who regarded themselves as either belonging to or extraneous to the polity of the Ottoman state. Keywords:   Ottoman Empire, Syria, Red Sea, pilgrimage routes, Arabia

SITUATED ON THE BORDERS OF ARABIA, SYRIA-PALESTINE AND EGYPT and at one end of a maritime trade route which in various periods has stretched to the Persian Gulf, India and the East Indies, the credentials of the Jordanian Red Sea port of ‘Aqaba as a frontier settlement should be beyond dispute. In Crusader and Ayyubid times, the principal fortification in ‘Aqaba (or Ayla) had been a castle located on Jazirat Fara‘un, an offshore island situated on the western (or Egyptian) side of the gulf, some 15 kilometres south-west of the present town of ‘Aqaba (Figure 5.1). Its function had been at first to interrupt—but then, following its capture and reconstruction by Saladin in 1170, to protect—the desert route between Cairo and Damascus that passed around the eastern and southern fringes of the Frankish kingdom of Jerusalem. Under the Mamluk Sultan Baybars (1260–77), however, with the Frankish threat to the region removed, ‘Aqaba came to prominence as a staging post for Egyptian (and some Palestinian) pilgrims travelling to and from Mecca and Medina, the holy sites of Arabia. A particularly vulnerable part of the route was the pass, known as ‘Aqabat Ayla (or Naqb al-‘Aqaba), where the road climbed steeply from the valley floor up on to the Sinai plateau to the west. Here and in the

Page 1 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

‘Aqaba Castle in the Ottoman Period, 1517–1917 oasis of ‘Aqaba itself, where the pilgrims camped on their outward and return journeys, the caravan was particularly at risk of attack from the Bedouin of the surrounding area. It was under Baybars that the pilgrimage came under official protection, and although further work remains to be done on the dating evidence from the recent excavations,1 it seems likely that to this period also belongs the construction of the first of a succession of hans, whose remains partly underlie the present castle. At any rate, there was certainly some form of fortification on the shore by c.1331, in the third reign of the Mamluk Sultan al-Nasir Muhammad (1309–40), when Abu ’l-Fida’ records the Egyptian governor moving his residence to it from the island castle, which was then abandoned.2 The present chapter, however, will focus on the history of the existing castle from the time of its construction in the last years of Mamluk rule until its capture by the Arab army in 1917. It should be stressed that, since analysis of the structure itself is still incomplete, the conclusions presented here should be regarded as provisional at this stage. (p.96) (p.97) The building—or, as we now know, the rebuilding—of the han in ‘Aqaba is recorded by the chronicler Ibn Iyas under his entry for August 1508 (Rabi‘ II, AH 914), in the reign of the penultimate Mamluk sultan, al-Ashraf Qansuh al-Ghawri. He records that in that month the sultan ordered Khayirbak Mi‘mar [otherwise known as Khayir Bey al-‘Ala’i] to go to the pass of ‘Aqaba at the head of a team of masons and engineers. The sultan had undertaken various works at ‘Aqaba: a caravanserai, towers and cisterns at the points where the pilgrims assembled, and a paved road along the coast. He also ordered that the narrow passes of the mountains of ‘Aqaba, which the pilgrims were only able to traverse with

Figure 5.1. Map of ‘Aqaba and surrounding region.

the most extreme difficulties, should be appropriately arranged.3 The works were therefore evidently already under way by August 1508; and in April and May of the following year (Muharram AH 915), the pilgrims returning from Mecca were able to benefit from them. Ibn Iyas records: The journey had been easy and fortunate, having taken place in the most complete security, and the state of health [of the pilgrims] had been excellent. On their return the pilgrims praised the useful works undertaken by the sultan in the vicinity of ‘Aqaba, namely a caravanserai, storerooms that could be used as depositories; and watch-towers, with garrisons of Turks who were relieved once a year. The ‘Aqaba road had been improved and the steepest sections of the passes had been levelled. A tower had been built at Ajrud, and another at Nakhl. A certain number of watering points on the Mecca road had been rearranged. Many other buildings of the same (p.98) type had been built; and finally a tower had been established at Aznam. Measures had been taken to install in all the guard posts a detachment of Mamluks, who were relieved annually4

Page 2 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

‘Aqaba Castle in the Ottoman Period, 1517–1917 It is quite possible, however, that building work on the han was still in progress at this time, for a monumental inscription built into its gatehouse, while attributing the construction to Qansuh alGhawri and his amir, Khayir Bey al-‘Ala’i, appears to give the date as 920/1514–15, the year before the sultan’s death.5 The list of building works undertaken by the late sultan that was subsequently given by Ibn Iyas also includes mention of his remaking of the ‘Aqaba road and protecting it against blown sand, and his construction of ‘a caravanserai with a gate flanked by towers’, in which were built magazines in which the pilgrims could store their goods.6 The caravanserai built in ‘Aqaba by Qansuh al-Ghawri lay close to the shore, some 900 metres south of the abandoned site of early Islamic Ayla, and took the form of a rectangular enclosure, some 56.5 metres north–south by 58 metres east–west, with polygonal towers at the four corners and a gatehouse projecting forward in the centre of the north wall (Figure 5.2). It was apparently built throughout in ashlar masonry. The gate itself consisted of a broad rounded arch, set between rounded turrets and covered by a projecting box-machicolation (Figure 5.3). The archway opened into an outer porch, or liwan, covered by a groin-vault, within which was the gate itself: originally a two-leaf door protected by a slit machicolation. Behind it, doors in the side walls of the passage opened into guard chambers, equipped with latrines and stairs to the upper floors (Figure 5.4). After these came an inner liwan, covered by a low domical vault. Around the inside faces of the fort’s walls were arranged series of barrel-vaulted rooms, each independently accessed from the central courtyard (Figure 5.5). These evidently represented the store-rooms in which the pilgrims and merchants were permitted to store their goods while camping in the surrounding oasis. In the south wall was a mosque, now represented by no more than its qibla wall, containing a stilted-arched mihrab flanked by colonnettes with muqarnas capitals. The principal function of the caravanserai at ‘Aqaba from the time of Qansuh al-Ghawri, as with its predecessors, appears to have been to serve both as a base for a detachment of mamluks to protect the pilgrim route and as a safe depository in which returning pilgrims and merchants could store their goods against raids by the neighbouring Bedouin. It was also at ‘Aqaba that the returning caravan would be met by a supply caravan sent from Cairo. In 1468, for example, it was reported that the Bedouin had stolen the victuals deposited in ‘Aqaba for the use of the returning pilgrims and a new supply caravan guarded by troops had quickly to be sent from Cairo.7 By the beginning of the sixteenth century the practice of sending a supply caravan to meet the pilgrims on their return was well established.8 These arrangements also continued after the Ottoman takeover in 1517.9 In 1521, it also became customary for the pilgrimage itself to be provided with an armed escort. In that year, Sultan Süleyman the Magnificent provided 360 troops and the amir al-hajj apparently some 100 mamluks and regular soldiers.10 But even this did not prevent eighty pilgrims dying from cold and lack of food between ‘Aqaba and Cairo in December that year.11 Later the size of the escort was reduced. In 1535 and 1555, for instance, the sultan’s escort was 240 troops, and in 1553 the amir al-hajj contributed only sixty, including the relief garrisons for the forts at ‘Aqaba and Aznam, which numbered ten each.12 AlJazari records that for the supply of the relief caravan going to meet the pilgrims at ‘Aqaba, the amir al-hajj had the right to five cantars of refined sugar, two and a half cantars of sweets, 125 ardebbs of broken beans (majrush) for the animals and twenty-five ardebbs of barley.13 He also mentions a firework being let off at ‘Aqaba on the return journey.14 ‘Aqaba was also the place where the goods of merchants returning from Mecca were inventoried for the purpose of taxation, though the tax itself was not levied until just before the caravan reached Cairo.15 (p.99)

Page 3 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

‘Aqaba Castle in the Ottoman Period, 1517–1917 (p.100) (p.101) Figure 5.2. ‘Aqaba Castle: plan, surveyed by Antoni A. Ostrasz and drawn by C. Rustum, June 1988. Reproduced by kind permission of the Jordanian Department of Antiquities.

Figure 5.3. ‘Aqaba Castle: the gatehouse, photographed by M. R. Savignac c.1928. Photograph courtesy of the École Biblique et Archéologique Française, Jerusalem.

Figure 5.4. ‘Aqaba Castle: the gate within the gatehouse, seen from the inner liwan, showing the secondary single-leaf door. Photograph D. Pringle.

Page 4 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

‘Aqaba Castle in the Ottoman Period, 1517–1917 Two later inscriptions carved on marble medallions, one either side of the castle gate, record a restoration of the castle (qal‘a) carried out by the Ottoman Sultan (p.102) Murad III in 1588.16 It is uncertain, however, whether this work entailed any significant modifications or additions to the Mamluk structure, or represented simply its completion. Various structural changes may nevertheless be identified as having taken place between the sixteenth century and the early Figure 5.5. ‘Aqaba Castle: the gate and nineteenth, but at present it is difficult to north range seen from inside the castle. date them with any precision. They include Photograph D. Pringle. the reconstruction of the western side of the gatehouse, including its turret, and the insertion of a single-leaf door in the outer gateway arch, which had a smaller arch built within it to accommodate it (Figure 5.3); the replacement of the two-leaf inner door by a similar singleleaf one (Figure 5.4); the rebuilding of the courtyard façade of the north range to the east of the gate, leaving the rubble-built vaulted rooms behind in their original state (Figure 5.5); the rebuilding of the west and south ranges with timber roofs as opposed to vaults; and the addition of upper floors to all the ranges, if they did not already have them. The latter were lightly constructed in rubble set in a mortar consisting largely of clay, with timber including palm logs used for the floors and ceilings and as lacing within the walls. Around 1685, the English convert to Islam, Joseph Pitts, travelled from Algiers to Mecca and Medina and returned with the Egyptian pilgrims through ‘Aqaba. When we came within fifteen days’ journey of Grand Cairo, we were met by many people who came from thence, with their camels laden with presents for the hagges, sent from their friends and relatives, as sweetmeats, etc. But some of them came rather for profit, to sell provisions to the hagges and trade with them. About ten days before we got to Cairo, we came to a very long steep hill, called Ackaba, which the hagges are usually much afraid how they shall be able to get up. Those who can will walk it. The poor camels, having no hoofs, find it very hard work, and many drop here. They were all untied and we dealt gently with them, moving very slowly and often halting. Before we came to this hill I observed no descent, and when we were at the top there was none, but all plain as before.17 The principal threat to the Egyptian pilgrims travelling to and from Mecca and Medina came from the Bedouin through whose lands they had to pass. As the Ottoman sultans had only limited control over the Arabian desert, they adopted the practice—no doubt already used by their Mamluk predecessors—of paying subsidies (sürre) to the Bedouin leaders in order to fulfil their duty of protecting pilgrims to the Holy Places. Such payments were officially justified in terms of the provision of the food and especially water that the pilgrims required along the route. There was also a political dimension, however, as may be seen by the honorific gifts that were sometimes conferred along with them, such as robes or titles like amir al-‘Arab. On the other hand the (p.103) Bedouin were also quite capable of asserting their own independence by

Page 5 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

‘Aqaba Castle in the Ottoman Period, 1517–1917 raiding a caravan if they considered the payment of their subsidy to have been insufficient or tardy18 In the eighteenth century, the district around ‘Aqaba remained as dangerous for pilgrims as it had been in earlier centuries. In 1732, for example, finding their approach to the citadel of ‘Aqaba blocked by Bedouin, the returning pilgrims sent a camelleer to ask for help from the governor of Cairo. He dispatched an expedition under ‘Ali Bey Zayn al-Faqar. The chronicler alDamurdashi describes how the relief expedition captured the four guard posts along the road through ‘Aqaba, one by one, until it was able to link up with the caravan. ‘Ali Bey then pursued the Bedouin and cut off the heads of some of their musketmen to serve as an example for the rest.19 In 1741, the returning pilgrimage under ‘Umar Bey Qatamish also learned that the Bedouin had occupied ‘Aqaba. Yahya Pasha in Cairo therefore sent ‘Ali Bey al-Dumyati to punish the Bedouin and bring the caravan safely back to Cairo. Al-Damurdashi relates: He came upon the Washshasha and people from al-‘Aqaba who had made a [defensive] circle of their loads and beasts of burden outside the city [which was occupied by hostile Bedouins]. The bedouins used a different tactic…[This time] they took positions behind rocks where they remained unseen by ‘Ali Bey al-Dumyati’s retainers as they passed the first guard post (halazun) and advanced towards the fourth post. Not a single bedouin was spotted…The pilgrim caravan lodged overnight in the citadel to avoid the heat [of the day]. [The next day] he descended with his baggage and crates, his troops, along with the Washshasha and people of al-‘Aqaba, past the first guard post. They continued with the loads, but when they reached the fourth guard post the bedouins hiding between the third and fourth guard posts caught them in a crossfire, rushed upon them and looted the baggage and possessions of the Washshasha and people of al-‘Aqaba. They made off with the entire load. Then another group of bedouins returned to resell the items to the pilgrims. They displayed the goods and [the pilgrims] had to buy them back. The bedouins who had seized the third guard post forced ‘Ali Bey al-Dumyati to abandon his tents, provisions and supplies. He withdrew to the fort at Nakhl where, for lack of anything else, they cooked beans without meat.20 It may be suspected that the enterprise shown by the Bedouin on this occasion resulted at least in part from the failure of the organisers of the pilgrimage to offer them sufficient payment for traversing their territory in the first place. By this period, however, the garrison at ‘Aqaba also seems to have become somewhat run down. At the time of Khalil Bey’s pilgrimage in 1746–7, for example, the mamluks are described as young negro slaves, so despised that ‘they stood in the guard posts at al-‘Aqaba begging (p.104) from the troops’.21 Shortly after this, under ‘Ali Bey, the practice of sending a supply caravan to ‘Aqaba was discontinued.22 In May 1816, John Lewis Burckhardt visited ‘Aqaba from Cairo to escape the plague then raging in the city: The castle itself stands at a few hundred paces from the sea, and is surrounded with large groves of date-trees. It is a square building, with strong walls, erected, as it now stands, by Sultan el Ghoury of Egypt, in the sixteenth century. In its interior are many Arab huts; a market is held there, which is frequented by Hedjaz and Syrian Arabs; and small caravans arrive sometimes from Khalyl. The castle has tolerably good water in deep wells. The Pasha of Egypt, keeps here a garrison of about thirty soldiers, to guard the provisions

Page 6 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

‘Aqaba Castle in the Ottoman Period, 1517–1917 deposited for the supply of the Hadj, and for the use of the cavalry on their passage by this route to join the army in the Hedjaz. The isolated position of the garrison, however, often caused the men to become rebellious. Burckhardt relates that, some three years before, the ağa, or governor, had married a daughter of the chief of the Banu ‘Umran and declared his independence, fleeing the castle to the protection of his wife’s kinsfolk only when troops were sent against him from Cairo.23 The ambivalence of the governor’s relationships with his masters in Cairo on the one hand and with the local tribal leaders on the other is a recurring theme in nineteenth-century travellers’ accounts. It was often only through his good offices that they could obtain safe passage through the surrounding tribal areas. Despite present uncertainty over the phasing and dating of some of the postMamluk building works, it appears that the castle survived more or less in its sixteenthcentury form until as late as the early nineteenth century, when it was visited and described by Léon de Laborde. De Laborde spent several days in ‘Aqaba in 1828 as a guest of the governor, Hasan Agha, who was holding it for Mehmed Ali, governor of Egypt. Hasan’s staff included a gunner (topçu), a secretary and a captain of the troops.24 While there de Laborde made drawings of the castle from the south-west and from the north, adding to the latter for dramatic effect an imaginary caravan arriving from Mecca (Figure 5.6). With the help—or, as he describes it, well-meaning hindrance—of the garrison, he also made a plan of the building (Figure 5.7), which he describes as follows: The fortress of ‘Aqaba, built to a regular plan, shows the same disposition and the same system as all those that were built for the protection of the caravan of Mecca. It is maintained externally in a fairly good state, to resist the caprices of the tribes, who, although fighters, have none of the means that would allow them to undertake a successful attack. On the inside, neglect has allowed several rooms to fall down in ruins. The governor has kept for himself the south-west bastion, and has enlarged it with a new layout of rooms; the topschi or military chief lives in the south-east bastion, and old gunner that he is he sleeps near a cannon. The piece, which is 12 [cm] calibre [i.e. roughly equivalent to a 12-pounder] and the one that is to be seen in the north-east tower are the only ones that could be fired in the case of attack, an event which fortunately is not anticipated. The gunner has added to his military occupations the much more peaceful one of merchant, and he has made his shop out of the mosque, which is now in ruins. A newly constructed well and a palm tree are the only objects to draw the attention in a badly levelled court, enclosed by buildings, smoke-blackened and in disorder. Some shacks belonging to Arabs who live off the small profits that they make from the sale of butter and meat to the soldiers of the garrison, some tombs of people from the fortress and of pilgrims stopped in their pious excursions by sickness…occupy the northern approaches to the fortress. To the east, the prolongation of the Jabal al-‘Aqaba comes down as sand hills right up to its walls. To the south and west, plantations of acacias, nabobs and palm trees raise up their rich vegetation.25 (p.105)

Page 7 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

‘Aqaba Castle in the Ottoman Period, 1517–1917 (p.106) (p.107) Although, as de Laborde indicates, parts of the fort’s interior were in a somewhat ramshackle state, by this period its military establishment had been put on a firmer footing. This was no doubt connected with the coming to power of Mehmed Ali as governor of Egypt in 1805, resulting in his subsequent seizure of Palestine and southern Syria from the sultan in 1831. Although forced to surrender those conquests in 1841, he was still permitted by the terms of the Convention of London to keep hold of ‘Aqaba and a number of other Red Sea garrison towns. In April-May 1822, Eduard Rüppell had

Figure 5.6. ‘Aqaba Castle: view from the north-east, drawn by Léon de Laborde in 1828. Courtesy of the Bodleian Library, University of Oxford: shelfmark 20608a.1, pl. opp. p. 45.

Figure 5.7. ‘Aqaba Castle: plan drawn by Léon de Laborde in 1828. Courtesy of the Bodleian Library, University of Oxford: shelfmark 20608a.1, pl. opp. p. 49.

numbered the garrison at ‘Aqaba at forty,26 though Burckhardt and others put it at or around thirty. In 1836, the governor was Süleyman Ağa.27 When Edward Robinson arrived in ‘Aqaba in April 1838, however, he was met by a young governor only recently arrived, his predecessor having been recalled, it was alleged, for showing incivility to travellers. The garrison, who were described as Maghribis but seemed to Robinson more probably Bedouin from Upper Egypt, comprised thirty-three undisciplined soldiers, in addition to whom there was also a captain of the gate, a gunner, and a commissary (or wakil) in charge of stores.28 The complement appears to have been the same at the time of W. H. Bartlett’s visit in October 1845.29 (p.108) De Laborde’s plan and drawings indicate that in 1828 all four corner towers of the fort were polygonal. The lower, ashlar courses of the external walls were painted alternately red and white, while the rubble-built upper parts were plastered white. Soon after his visit, however, some significant rebuilding works seem to have taken place. There is perhaps a hint of them in the drawing made by David Roberts in February–March 1839, on which the north-eastern corner tower appears to be shown rounded.30 Roberts, however, had already visited de Laborde’s companion, M. Linant, in Cairo and had access to de Laborde’s earlier drawings.31 It is possible therefore that he simply misrepresented what he saw on the other’s drawing. Further doubt concerning the accuracy of his drawing is also cast by his depiction of the south-west tower as rounded. In fact it is still polygonal and was shown thus on de Laborde’s drawing from the south-west and on W. H. Bartlett’s, made in 1845.32 Nonetheless, the surviving structural evidence indicates that the north-east and north-west towers were both largely rebuilt in the nineteenth century with rounded as opposed to polygonal exteriors. Decorative panels set in the wall faces of these towers are unfortunately of little help in dating, and may in any case have been salvaged from the pre-existing structure. But the gun ports in the south-east tower were provided with vertically sliding shutters, similar to those found in naval vessels of the early to mid nineteenth century. Another alteration made at this time was the demolition of the northern part of the west range to create a walled enclosure, entered through a broad rounded-arched gate on the east (see Figure 5.2). The purpose of this is uncertain, though it may perhaps have

Page 8 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

‘Aqaba Castle in the Ottoman Period, 1517–1917 been a compound for corralling transport animals, or possibly even field artillery. In a later phase, the gate to the compound was blocked and other buildings were constructed in front of it. When Aloïs Musil came to ‘Aqaba in 1910, while surveying the route of the Hijaz Railway, he found a company of Turkish regular troops stationed there under a governor (kaymakam). They had been there at least since the previous year, but the governor whom he met had only arrived a few days before. The latter informed Musil that his aim was to bring the surrounding region under Turkish rule; but, although the forces at his disposal were supposed to number 150 regulars, relieved every six months, as well as twenty-six camel-mounted ‘Agayl, some 130 of his soldiers were sick, there were only two camels for the ‘Agayl, and he had no money, food or ammunition. These troops, however, were not stationed in the old fort, whose north and east walls were by now half ruined, but in a new timber encampment placed some distance north of the village. The position of this may be seen on an air photograph (Figure 5.8), which was taken by Erich Steiner, commander of German Air Squadron 305 based in Dar‘a, during a bombing raid on ‘Aqaba on 26 August 1918.33 The government building occupied by the governor, ‘a low, dirty hut with two rooms and a single window’, stood between the village and the camp.34 (p.109) The fort saw military action in the ItaloTurkish war of 1911–12, when it was bombarded by the Italian navy. Following a visit made incognito on 6–7 November 1912, W. B. Drury reported to General Wingate: (p.110) I went ashore at Akaba; it is a miserable sort of village hidden in a large date palm grove; the buildings & fort had been knocked about by the Italians; there was one co[mpan]y of Turkish regulars there, commanded by a Mulazim Awal.35 In February 1914, however, T. E. Lawrence visited ‘Aqaba while completing the archaeological survey of the Negev for the Palestine Exploration Fund and found that the fort, though partly ruined, was still

Figure 5.8. ‘Aqaba: air photograph taken from the north-west during a bombing raid on British and Arab forces there on 26 August 1918 by Erich Steiner, commander of German Squadron 305. The castle lies among the palm trees close to the southern jetty (top), while the Turkish military cantonment appears to have been near the northern jetty (bottom), where the photograph shows rows of British military tents. Reproduced by kind permission of the Department of Geography, Aerial Photographs Unit, Hebrew University of Jerusalem.

acting as military headquarters and had troops living in it.36 On 30 April 1917, the Royal Navy put a party ashore to attack the Turks entrenched in ‘Aqaba.37 On 4 June, some thirty ‘Agayl deserted to Wajh on the Red Sea coast and reported that the Turks had abandoned ‘Aqaba;38 but a week later it was reoccupied by about fifty Turkish and Syrian regulars, 300 Arab irregulars, and a small German detachment.39 After further bombardments by British and French warships, ‘Aqaba fell to the Arab army of Sharif Husayn on 6 July 1917.40 ‘Aqaba became the Sharif’s headquarters for the next six months (see Figures 5.9–5.10). As a result of the bombardments of 1911–12 and 1917, the west, or seaward, wall of the castle and its north-western corner tower were effectively destroyed, their collapsed remains being buried under a mound of rubble; but although it would thereafter have been barely serviceable as a military work, the castle’s continued importance as a symbol of authority for the new Hashemite amirate—and later kingdom—of Transjordan is illustrated by the coat of arms that was quickly Page 9 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

‘Aqaba Castle in the Ottoman Period, 1517–1917 set up above its gateway. With this it entered the most recent phase of its history as a national monument of a new nation state. (p.111) (p.112) For most of the period during which ‘Aqaba belonged to the Ottoman Empire, the precise nature of its frontier status needs to be nuanced, since, in theory at least, all of the provinces adjoining it formed part of the same political unit, and the Red Sea itself was a largely Ottoman lake. In practice, however, as we have seen, Ottoman political and military control in the Syrian and Arabian deserts was often tenuous and reliant on individual deals struck with Bedouin leaders, often within the context of the sultan’s fulfilment of his obligation to protect the pilgrimage to the

Figure 5.9. ‘Aqaba Castle: interior looking north, photographed by Capt. Raymond Goslett in August 1918. Photograph courtesy of the Imperial War Museum, London, neg. no. Q 59289.

Figure 5.10. ‘Aqaba Castle: interior looking Holy Places of Arabia.41 From this south-west, photographed by Major perspective forts along the pilgrimage Mackintosh in 1917–18. Photograph courtesy routes such as that at ‘Aqaba, although of the Imperial War Museum, London, neg. technically sited well within the no. Q 59750. geographical boundaries of the Ottoman Empire, were none the less frontier places, mediating relations, whether peaceable or belligerent, between populations who regarded themselves as either belonging to or extraneous to the polity of the Ottoman state. The rise to power of Mehmed Ali as a virtually independent ruler of Egypt in 1805 would have made ‘Aqaba remoter still from Istanbul, but it would not have significantly altered its frontier status as thus defined. A little over a century later, however, in 1910, an active—though apparently mostly ineffectual—attempt was being made to bring ‘Aqaba and its surrounding area under direct Ottoman control; and for a brief period, until its capture by the Hashemites in 1917, it did at last come into its own as a frontier fortress in the more conventional sense of the term. Notes: Proceedings of the British Academy 156, 95–112. © The British Academy 2009. (1) Between 2000 and 2003, three seasons of survey and excavation were carried out on the late Mamluk fortified han or castle of ‘Aqaba by a joint Belgian and British team, directed by John De Meulemeester of the University of Ghent and the writer. See John De Meulemeester and Denys Pringle, ‘The archaeological investigation of ‘Aqaba castle’, Newsletter of the Council for British Research in the Levant (2001), 19–21; eidem, ‘The ‘Aqaba Castle Project 2001’, Newsletter of the Council for British Research in the Levant (2002), 19–20; eidem, ‘The ‘Aqaba Castle Project 2003’, Newsletter of the Council for British Research in the Levant (2004), 19–20; Sawsan alFakhri, ‘The history of the al-‘Aqaba Castle in the light of recent excavation’, Annual of the Department of Antiquities of Jordan, 45 (2001), Arabic section, 59–62. The work was carried out under the auspices of the Jordanian Department of Antiquities and was sponsored by the Wallonne Regional Government and the Council for British Research in the Levant. Since 2003, Professor De Meulemeester has carried out a further three seasons of excavation, which have

Page 10 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

‘Aqaba Castle in the Ottoman Period, 1517–1917 extended considerably the area under investigation and added further to our knowledge of the series of Mamluk hans that preceded the one that we see today. See John De Meulemeester and Denys Pringle, ‘The ‘Aqaba Castle Project 2004–5’, Newsletter of the Council for British Research in the Levant (2005), 42; John De Meulemeester and Reem al-Shqour, ‘The ‘Aqaba Castle Project 2004–5’, Bulletin of the Council for British Research in the Levant, 1 (2006), 27–8. (2) Abu ’l-Fida’, Taqwim al-Buldan, ed. J. T. Reinaud and Macguckin de Slane as Géographie d’Aboulfeda (Paris 1840), 2: 86–7; Denys Pringle, ‘The Castles of Ayla (al-‘Aqaba) in the Crusader, Ayyubid and Mamluk periods’, in U. Vermeulen and J. Van Steenberg (eds.), Egypt and Syria in the Fatimid, Ayyubid and Mamluk Eras, 4 (Leuven: Peeters, 2005), 333–53. (3) Ibn Iyas, Journal d’un bourgeois du Caire: Chronique d’Ibn Iyâs, trans. G. Wiet (Paris: Bibliothèque général de l’École Pratique des Hautes Études, VIe section, 1955–60), 1: 130. (4) Ibn Iyas, Journal, 1: 148. (5) Harold W. Glidden, ‘The Mamluk origins of the fortified khan at al-‘Aqaba, Jordan’, in G. C. Miles (ed.), Archaeologica Orientalia in Memoriam Ernst Herzfeld (New York: J. J. Augustin, 1952), 116–18; cf. Moshe Sharon, Corpus Inscriptionum Arabicarum Palaestinae (CIAP) (Leiden: Brill, 1997), 1: 100–1. (6) Ibn Iyas, Journal, 2: 91–2. (7) J. Jomier, Le Mahmal et la caravane égyptienne des pèlerins de la Mecque (XIIIe–XXe siècles) (Cairo: IFAO, 1953), 90. (8) Ibid., 91; cf. Ibn Iyas, Journal, 1: 403, 410. (9) Ibn Iyas, Journal, 2: 265; cf. 2: 443. (10) Jomier, Le Maḥmal, 97. (11) Ibn Iyas, Journal, 2: 414. (12) Jomier, Le Maḥmal, 97. (13) Ibid., 106. (14) Ibid., 122. (15) Ibid., 216. (16) Sharon, CIAP, 1: 101, figs. 46a, 47; Mehmet Tütüncü, Turkish Jerusalem (1516–1917): Ottoman Inscriptions from Jerusalem and Other Palestinian Cities (Haarlem: Turkestan and Azerbaijan Research Centre, 2006), 203–4, no. 106. (17) ‘An Account by Joseph Pitts of his Journey from Algiers to Mecca and Medina and back, c. 1685’, in W. Forster (ed.), The Red Sea and Adjacent Countries at the Close of the Seventeenth Century, as described by Joseph Pitts, William Daniel and Charles Jacques Poncet (London: Hakluyt Society, 1949), 47–8.

Page 11 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

‘Aqaba Castle in the Ottoman Period, 1517–1917 (18) See the discussion of caravan security by Suraiya Faroqhi, Pilgrims and Sultans: the Hajj under the Ottomans 1517–1683 (London: I. B. Tauris, 1994), 54–73. (19) al-Damurdashi, al-Durra al-Musana fi Akhbar al-Kinana, trans. and annotated by D. Crecelius and ‘Abd al-Wahhab Bakr as al-Damurdashi’s Chronicle of Egypt 1688–1755 (Leiden: Brill, 1991), 303–4. (20) Ibid., 344–5. (21) al-Damurdashi, al-Durra al-Musana fi Akhbar al-Kinana, 54–73. (22) Jomier, Le Maḥmal, 134. (23) J.-L. Burckhardt, Travels in Syria and the Holy Land (London: John Murray, 1822), 509–10. (24) Léon de Laborde, Voyage de l’Arabie Pétrée par Léon de Laborde et Linant (Paris: Giard, 1830), 45–52; idem, Journey through Arabia Petræa, to Mount Sinai and the Excavated City of Petra (London: John Murray, 1836), 95–101. (25) De Laborde, Voyage de l’Arabie Pétrée, 49–50. This passage, omitted from the English edition of de Laborde’s book, is here translated from the French. (26) Carl Ritter, The Comparative Geography of Palestine and the Sinaitic Peninsula, translated and adapted to the use of Biblical students by W. L. Gage (London 1865; repr. New York: Haskell House, 1969), 1: 51, 74. (27) John Lloyd Stephens, Incidents of Travel in Egypt, Arabia Petræa, and the Holy Land, ed. V. W. von Hagen (Norman: University of Oklahoma Press, 1970), 226–32. (28) Edward Robinson, Biblical Researches in Palestine, and in the Adjacent Regions: A Journal of Travels in the Year 1838, by E. Robinson and E. Smith (Boston: Crocker and Brewster, 1860), 1: 164. (29) W. H. Bartlett, Forty Days in the Desert on the Track of the Israelites; or, A Journey from Cairo, by Wady Feiran, to Mount Sinai and Petra, 5th edition (London, n.d. [1845]), 104–5; cf. E. H. Palmer in Picturesque Palestine, Sinai and Egypt, ed. Charles W. Wilson (London, n.d. [1880]), 3: 227–8. (30) The Holy Land, Syria, Idumea, Egypt, & Nubia, from drawings made on the spot by David Roberts R.A., with Historical Descriptions by the Revd. George Crolly L.L.D., lithographed by Louis Haghe (London: F. G. Moon, 1842–9); reprinted as The Holy Land (London: Studio Editions, 1989), 5: 56–7, pl. 122. (31) Ibid., 1: 9. (32) Bartlett, Forty Days in the Desert, pl. 28 (opp. p. 100). (33) B. Z. Kedar, The Changing Land between the Jordan and the Sea: Aerial Photographs from 1917 to the Present ([Jerusalem]: Yad Ben-Zvi Press/Ministry of Defence, 1999), 45. I am grateful to Professor Kedar for providing me with the copy of this photograph reproduced here.

Page 12 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

‘Aqaba Castle in the Ottoman Period, 1517–1917 (34) Aloïs Musil, The Northern Heğâz: A Topographical Itinerary (New York: American Geographical Society, 1926), 81–7, figs. 7–9; cf. Barnabé Meistermann, Guide du Nil au Jourdain (Paris: A. Picard, 1909), 195. (35) Letter from W. B. Drury to General Wingate (Port Sudan, 12 November 1912), in Durham University Library, Sudan Archive: SAD/134/1/16–21. (36) C. Leonard Woolley and T. E. Lawrence, The Wilderness of Zin (London: Cape, 1936), 141–3. (37) Arab Bulletin, 49 (Cairo, 30 April 1917), 193. (38) Ibid., 278. (39) Arab Bulletin, 55 (Cairo, 28 June 1917), 290–1. (40) Arab Bulletin, 57 (Cairo, 24 June 1917); Frederick G. Peake, History of the Tribes of Jordan (Miami: University of Miami Press, 1958), 98; T. E. Lawrence, Seven Pillars of Wisdom (Harmondsworth: Penguin, 1962), 322. (41) See also Chapter 4 above, Andrew Petersen, ‘The Ottoman conquest of Arabia and the Syrian hajj route’.

Access brought to you by:

Page 13 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Evliya Çelebi, the Mani and the Fortress of Kelefa

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Evliya Çelebi, the Mani and the Fortress of Kelefa MALCOLM WAGSTAFF

DOI:10.5871/bacad/9780197264423.003.0006

Abstract and Keywords The fortress of Kelefa on the western side of the Mani peninsula in southern Greece is one of a series of fortifications constructed or refurbished to defend the Ottomans' maritime frontier against Venice in the late seventeenth century. This chapter addresses whether the castle is indeed of Ottoman construction, or whether it is in fact an earlier, Frankish, structure that was merely restored by the Ottomans. It outlines what The Chronicle of the Morea says and how it can be interpreted. It looks at what light later literary and other sources can throw on the problem, discussing Evliya Çelebi in the process. It identifies Grand Magne with a fortress which others have claimed is of Ottoman construction and dates from the seventeenth rather than the thirteenth century. The chapter also provides a sketch of the Mani as part of the Ottoman frontier zone in the Mediterranean region. Keywords:   Kelefa, Mani, Grand Magne, Evliya Çelebi, Ottoman frontier, Chronicle of the Morea

THE FORTRESS OF KELEFA ON THE WESTERN SIDE OF the Mani peninsula in southern Greece is one of a series of fortifications constructed or refurbished to defend the Ottomans’ maritime frontier against Venice in the late seventeenth century. The key problem addressed in this chapter is whether the castle is indeed of Ottoman construction, as has usually been thought, or whether it is in fact an earlier, Frankish, structure that was merely restored by the Ottomans. This may shed light on a further important problem in the history of the Mani, whether Kelefa is—or is not—the castle known as Grand Magne, built on the orders of the Frankish Prince of Achaea, Guillaume de Villehardouin, c.1248. The identification and precise location of Grand Magne has troubled scholars more or less from the time when the Greek text of the principal source for its construction, The Chronicle of the Morea, was first published in 1904.1 Versions of the Chronicle exist in Aragonese, French, Greek (the only one in verse) and Italian, but it took its final form around 1340.2 Grand Magne is the name used in the French version. Mégali Máïni and Palaiá Máïni are employed in the Greek text.

Page 1 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Evliya Çelebi, the Mani and the Fortress of Kelefa In this chapter I will outline what The Chronicle of the Morea says and how it can be interpreted. Then I will look at what light later literary and other sources can throw on the problem. This is where Evliya Çelebi comes in, and I will deal with his material at some length. Finally, I will draw some conclusions: I identify Grand Magne with a fortress which others have claimed is of Ottoman construction and dates from the seventeenth rather than the thirteenth century. But first I will provide a sketch of the Mani as part of the Ottoman frontier zone in the Mediterranean region.

(p.114) The Mani The Mani is the southernmost peninsula of mainland Europe. It is the continuation of the Taygetos range of mountains—an arid and barren land of rocks and thin soils, especially in the south, and with few other resources. It ends in Cape Matapan, ancient Tainaron. The region’s strategic location in the Ottoman Empire is clear from the map of political control in the Mediterranean region c.1600, roughly halfway through the period of Ottoman rule over what is now southern Greece (Figure 6.1). Spain dominated the northern section of the western basin, while the Barbary States were emerging along the southern shore; the Ottomans controlled the eastern basin. Venice, though, still ruled the Ionian Islands, Tinos and Crete. The Knights Hospitallers were entrenched in Malta. The Mani lies at the intersection of these circles of power. Defended to some extent on the north by mountains and difficult to penetrate by land because of extremely rugged terrain, it is exposed to the exercise of sea power. For example, Venice occupied it in the 1460s,3 while in June 1570 a squadron of Venetian galleys from Crete under Marco Querini, Captain of the Gulf, captured the recently built fortress above Porto Kaio, named Maini or Maina, and destroyed it.4 In 1585–7 and again in 1601 Spanish forces attacked the Maniat coast, with the fortress of Passava, in the north-west, a particular target.5 In 1770 the Mani was one of those peripheral regions of the Ottoman Empire where Russia stirred up rebellion during the war of 1768–74, and Russian troops were landed in Oitylo Bay.6 The French became interested in the region later in the eighteenth century and into the early nineteenth century, capitalising on its possibilities for clandestine operations.7 Figure 6.2 focuses on the Ottoman maritime frontier in south-western Greece and attempts a tentative reconstruction for 1670, when the Turkish traveller, Evliya Çelebi, visited the Mani.8 At this date the sultan controlled the mainland and, following the fall of Candia (Irakleion) in 1669, the whole of Crete. Harbours and anchorages are scattered all along the coasts. Many of the major ones were defended by fortresses, some of (p.115) which were of considerable age but had been refurbished after the Ottoman conquests of the fifteenth and early sixteenth centuries. Notable examples are Coron (Ott. Koron, Gr. Koroni) and Modon (Gr. Methoni), taken from the Venetians in 1500 and held for the sultan despite attacks by the Genoese Admiral, Andrea Doria (Coron in 1532), the Knights of Malta (Modon 1531) and Don John of Austria (Modon 1572). These were vital ports of call for vessels going east from Italy and were described as the ‘eyes of Venice’ when the Republic had significant commercial and colonial interest in the eastern Mediterranean. Of even greater consequence, though, were the fortresses built by the Ottomans. The most important of these are the fortresses at the mouth of the Gulf of Lepanto/ Corinth and New Navarino. The Bay of Navarino is sheltered by the island of Sphacteria, where a Spartan force was blockaded by the Athenians during the Pelopennesian War (425 BC), and it is the ‘largest harbour in the Pelopónnisos’.9 In 1827 it was the scene of the last fleet action fought under sail; defeat of the Ottoman fleet opened the way to the liberation of southern Greece from Ottoman rule. The fortress of New Navarino (Pylos today) was built in 1573 to

Page 2 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Evliya Çelebi, the Mani and the Fortress of Kelefa command the southern entrance into the bay—the only practical way in after Ottoman forces blocked the northern entrance with rocks following an attack by Don John of Austria. The Castle of Roumeli (today Andirion) and the Castle of the Morea (Rion) were built over three months in 1499. They are low lying, right on the seashore, and were designed to house heavy guns, which would prevent ships from entering the strait leading into the Gulf of Lepanto/Corinth and thus penetrating the heart of Ottoman domains in southern Greece. The celebrated Battle of Lepanto (7 October 1571) was fought outside the straits; the Ottoman fleet was based at Lepanto (Navpaktos) following attacks on the Venetian-held Ionian Islands in the summer. (p.116) (p.117) Off the mainland of western and south-western Greece lie the Ionian Islands, including Cerigo (Cythera) and Cerigotto (Anticythera). All were ruled by the Venetian Republic in 1670, and major fortresses were located at Corfu and Santa Maura (Levkas). Venice still controlled a few towns on the mainland of Epirus as well. The Venetian territories provided operational bases against the Ottoman Empire for the Republic and its allies, notably during the Cyprus War (1570–3), of which the Lepanto campaign was a part, and the long War of Candia (1645–69). They were also targeted from time to time by the Ottoman fleet, as in 1537, 1570, 1661, 1667 and during the Morean War (1684–99), as well as its remise in 1715–18. The significance of the maritime frontier continued into the nineteenth century, when the French occupations of the Ionian Islands (1797–9, 1807–10, 1814) were seen as the precursor to an attack on the Balkan provinces of the Ottoman Empire and a young artillery officer, Captain William Martin Leake, was sent by the English government in 1804 to find out about the vulnerability of the western coast to attack, including the state of the fortresses, and to advise the Ottoman

Figure 6.1. The political situation in the Mediterranean region, c.1600.

authorities on their refurbishment.10 We can conclude, then, that the maritime frontier of Figure 6.2. The Ottoman maritime frontier the Ottoman Empire in south-western in south-western Greece, c.1670, showing places mentioned in the text. Greece from the fifteenth to the nineteenth century was as much a contested space, or a ‘zone of strife’, as the (p.118) frontiers in the Danubian plains or the Caucasus. This was the result of the spatial relationships between the islands, the adjacent mainland and the linking seaways.

Page 3 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Evliya Çelebi, the Mani and the Fortress of Kelefa The Mani was particularly important to the Ottoman strategy, though that varied with the risk of attack from the west. Control of its harbours was essential not only to the safe movement of shipping in the dangerous waters around Cape Matapan but also to the policing of the sea routes, which converged south-eastwards in the passage between the mainland and Crete. Both were strengthened when the same power ruled both sides of the seaway, something which the Ottomans achieved in the War of Candia. Shipping moving eastwards either from the Strait of Messina and the Strait of Otranto or from Malta passed this way, as did that going westwards from the Archipelago and beyond. The space between Crete and Libya was used much less. Sailing ships had difficulty weathering Cape Matapan in northerly winds, while sudden squalls and down draughts from the mountainous spine of the Mani and its steep cliffs could take them aback or put them in stays. Calms left them unmanoeuvrable and vulnerable. Bays immediately around Matapan offered shelter, but it was risky. The Maniats preyed on disabled and sheltering ships as pirates, privateers and wreckers until well into the nineteenth century. In such a situation secure harbours were important to the Ottoman Empire. This is why the fortress was built at Porto Kaio, on the eastern side of Cape Matapan, and why the castle at Kelefa was built or, as I shall argue, was refurbished in 1670. Kelefa would help to control the Maniats and secure the adjacent harbour (Porto Vitylo/Oitylo) at a time of heightened tension following the War of Candia. The Chronicle of the Morea The Chronicle tells how, after the capture of Monemvasia c.1248, the Frankish Prince of Achaea, Guillaume de Villehardouin, founded Mistra, a short way from Lakedaimonia, the successor to ancient Sparta. The context suggests that Mistra was established primarily to control a Slav tribe known as the Melingoi, which lived in the Taygetos mountains.11 In order to contain this disrespectful group still further, Guillaume with his council took the decision to establish another castle ‘somewhere around these mountains’. The Chronicle goes on: Thereupon, the prince himself made a tour on horseback, following the directions of the people of the land, and he passed Passava and journeyed to Máïni, and there he found an awesome crag on a promontory. Because he found it pleasing, he built there a castle and named it Máïni, as it is still called.12 (p.119) A third castle was added some time later to complete the containment of the Melingoi. This is Beaufort or Leftro in northern Mani.13 After the foundation story, the Chronicle usually refers to Villehardouin’s castle as Mégali Máïni or Palaiá Máïni in the Greek version and Grant or Grand Magne in French.14 The use of a qualifying term in The Chronicle of the Morea—Grand—might suggest that the author was drawing a distinction between at least two places with the same name. It complicates our task, as does the use of the name Maina for the fortress above Porto Kaio.15 For this chapter, I wish to sidestep that issue and focus on the main question: Where was Grand Magne? Continuity in the name Mani, as well as reference to Mistra and Passava in the Chronicle, suggest a location in the Mani peninsula, but where? Establishing the precise location of the castle is vital, for it would help in understanding the geography of Villehardouin’s strategy of containing the remnant Slavs in the southern Peloponnese and deciding how much of the Mani he claimed to control. It is also essential to making sense of the spatiality of the Byzantine revival in southern Greece following the Frankish defeat in the plain of Pelagonia (western Macedonia) in 1259 and the ceding three years later to the Emperor Michael Palaeologos by

Page 4 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Evliya Çelebi, the Mani and the Fortress of Kelefa Guillaume de Villehardouin of Geraki, Maini, Mistra and Monemvasia, together with ‘all the theme around Kisterna’, to secure his release.16 Various sites have been suggested for the location of Grand Magne, four in particular. Deciding which one has the best claim involves various difficulties. These concern the date and continuity of place names, their meaning and location; problems of interpretation, not only of the source material but also of the topography of the southern Peloponnese, both of which are conditioned by the interpreter’s basic assumptions; and the typologies of fortification and masonry. I thought in 1991 that I had resolved the issues by applying six criteria to the four candidates, set out in this list of criteria for the identification of Grand Magne:17 A. From The Chronicle of the Morea Candidate sites must: 1. make sense in a strategy of containment for the Melingoi and in relation to the other princely castles (Mistra and Leftro) and the baronial castles of Passava and Gritsena; 2. lie ‘beyond’ Passava when riding from Mistra; 3. be situated in Máïni/Máni; 4. (p.120) be capable of description as ‘an awesome crag on a promontory’; 5. be on a harbour. B. From other sources of information: Candidate sites must also: 6. make strategic sense in the pattern of fortresses ceded with all the ‘theme of Kisterna’ in 1262; 7. have fortifications containing at least some Frankish elements. My conclusion was that Kelefa provided the best match. This has been disputed and claimants in the southern part of the Mani have been promoted,18 though the criteria have been generally accepted. Indeed, Takis Katsafados has extended them to eleven, refined them and assessed quantitatively the degree to which each claimant meets them, calculating a coefficient to evaluate the degree of conformity.19 However, I will stick to my simpler scheme for the moment, but add a new criterion, No. 5. I introduce this to take account of a passage in the French text of the Chronicle which I had overlooked. This tells how a later Prince of Achaea, Floret of Hainhaut, ‘came to the point where the Chastel de la Grand Mange is’ and mentions a harbour.20 Hetherington locates Grand Magne on the rugged, arid Tigani peninsula in southern Mani, where there are the remains of an extensive settlement, and identifies the harbour mentioned by the chronicler with Mezapo Bay21 Shelter for small vessels is certainly found in the eastern angle of this bay, but Vitylo Bay (Porto Vitylo) to the north provides a larger harbour and fits in with my identification of Grand Magne with Kelefa.22 There are two weaknesses to my original argument, which I readily concede: the apparent lack of fortifications typical of the Frankish period, particularly of a donjon or keep; and the evidence in some sources that Kelefa may have been built de novo c. 1670. To summarise my position on the fortifications at Kelefa, I argued, first, that keeps were not a universal feature of Frankish castles in the Peloponnese and pointed out that no evidence for one has been found at nearby Passava, the Frankish origins of which have not been disputed. I went on to argue that the old fortifications could have been remodelled, particularly to cope with guns, and principally by adding bastions to the walls and lowering their height. Kougeas has suggested that the order of Manuel II Palaeologos from 1415 to demolish the Maniat fortifications applied to the fortresses (p.121) of Leftro, Máïni and Passava.23 On the other hand, the period of Venetian occupation, 1463–5,24 might have been a time for refurbishing pre-existing fortifications. However, the real Page 5 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Evliya Çelebi, the Mani and the Fortress of Kelefa problem is that the fortifications at Kelefa have never been adequately examined. The only published studies of which I am aware are those of Traquair (1906), Bon (published 1969, but based upon fieldwork carried out across the whole Peloponnese in 1925–36 and 1938), and Andrews (1953, reprinted 2006).25 Traquair simply says that the fourth of his five types of masonry distinguished in fortresses across the Peloponnese was ‘rubble masonry without tiles, found at Kelefa. This is Turkish work [my italics], and tiles were probably not available; it is quite exceptional.’ Bon’s view was that, although the situation of Kelefa might indicate construction at the same time as Passava, its appearance (rectangular outline, strong round corner towers scarcely protruding beyond the curtain walls) ‘invites the thought’ that it was later. Andrews pointed to the deficiencies in the fortifications at Kelefa—over-extended, thin curtain walls, round bastions, and lack of external works such as ditch, counter-scarp or glacis—and stated that these were due to Turkish ineptitude. Unfortunately, William Martin Leake, the observant nineteenth-century topographer of Greece, did not visit Kelefa in 1805 when he was in the region, but he reported that ‘it is said to be of Venetian construction, but could only have been a repair or reconstruction of the original fortress, as both Kelefá and Zarnáta existed before the Venetian conquest, and were taken by them from the Turks in the year 1685.’26 If the fortress at Kelefa proves to be of seventeenth-century construction, then its creation must be seen, at least in part, as a major investment to strengthen the Ottoman frontier in south-western Greece at time of increased risk of attack.

Western Literary and Documentary Evidence In the absence of better archaeological information from Kelefa, we can turn to other evidence. A few medieval documents mention a place, fortress or settlement called Mani. For example, some Greeks were alleged to have made an attack upon a Venetian (p.122) ship at the ‘castle of Mayna’ some time before 1278,27 and ‘le Meyne’ appeared in a tax list from 1391.28 Some degree of continuity is suggested. Much later a Maniat petition to the captain-general of the Venetian forces, Francesco Morosini, dated 1684 (that is, right at the beginning of the Venetian attack upon the Ottoman Peloponnese), states that the Ottomans ‘…built a castle on the harbour of Vitulo’ to subject them to his will.29 No date is given, but the implication is that construction took place some time after the Ottoman conquest c.1465 and possibly fairly recently. Kelefa stands above the head of this harbour, on the opposite side of a ravine from Vitylo/Oitylo, and could be the castle mentioned. Two—arguably three—literary sources appear to confirm its relatively late date. Dr Jacob Spon (1645–85) from Lyons and George Wheler (1651–1724) published separate accounts of their travels together in Greece in 1675–6.30 Wheler describes how, while their vessel lay becalmed during a night in the Gulf of Messenia, opposite Koroni (see Figure 6.2), ‘some Mariners of this place, that were on board’ regaled them with ‘many diverting Stories’ about the people of the Mani, including their habit of cutting the cables of anchored vessels so that they would run ashore and be wrecked. Their informants told Spon and Wheler that The Magnoti, who are the Inhabitants of that Country are famous Pirates by Sea, and Pestilent Robbers by Land. They have always bravely defended themselves against the Turks, and maintained their Liberty, till lately by this Stratagem the Turks were too hard for them. They got their consent to build two forts upon their Coasts, which they did so advantageously, as soon made them Masters of their City, and them. And now none of them are exempted from paying Tribute, but a few in the Mountains.31 Kelefa is assumed to be one of these forts; the other is to the north, at Zarnáta.

Page 6 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Evliya Çelebi, the Mani and the Fortress of Kelefa The Seyahatname of Evliya Çelebi A different literary source is the Seyahatname or ‘Book of Travels’, written in Turkish by the man known only by his nom de plume, Evliya Çelebi (1611–c.1684). This contains a specific reference to Kelefa. The Seyahatname survives in several manuscripts, but what is considered the archetype is clearly a rough draft, full of gaps and annotated (p.123) with notes for inclusion and improvement, on which the author may have been working when he died.32 Although it appears to be a ‘sprawling and rambling account’33 of travels over a forty-year period (1640–80), there is a degree of organisation and structure to the work, as Dankoff has shown.34 However, the historical value of the Seyahatname has been questioned.35 Evliya Çelebi clearly describes places he did not visit. The influence of professional story-telling is apparent in his delight in the fanciful and mythical. Folk-tales and oral traditions are repeated uncritically. The text is demonstrably full of error, exaggeration and carelessness, particularly over numbers. It should be no surprise, then, that doubts have been expressed over his claim to have visited Mani.36 On the other hand, much can be said in favour of the general accuracy of Evliya’s accounts of places. He placed himself in the tradition of Islamic geographers, following a standard organisational pattern in his descriptions. Throughout, Evliya was at pains to assure his readers of his credibility and he does seem to have tried to ensure the veracity of his information. ‘Lying is forbidden’, he says at one point. He mentions consulting local officials and documents. He took care not only over orientation with respect to the qibla, but also over measurements. Where Evliya Çelebi’s accounts have been carefully checked, notably those for Bitlis and Diyarbakir in what is now eastern Turkey, he has been shown to be precise and correct.37 Moreover, much of the exaggerated statistical information has been shown to be the fault of the original editor of the printed text.38 Not only did Ahmet Cevdet work with an imperfect manuscript, but he also bowdlerised his readings, omitting some passages and changing others, while inserting ridiculous figures in places where the archetype leaves a gap. In other words, Evliya Çelebi cannot simply be dismissed as unreliable.39 Fortunately, the recent edition by Kahraman, Dağli and Dankoff provides a much better basis for understanding the work than was previously available. I have used both this text and a Greek translation by Demetri Loupis.40 (p.124) Some thirty years ago I demonstrated that spatial sequencing in historical documents can be used to uncover genuine itineraries.41 According to Suraiya Faroqhi, the same claim was made by Pierre Mackay ‘many years ago’:42 mapping would weed out invented travels for Evliya Çelebi. The case is essentially this. If a written itinerary, or a non-alphabetical list of toponyms, follows a sequence of places which, when identified, makes spatial sense, then it is likely that the author possessed sensible information about the area concerned, whether from personal observation or from some other traveller, and there is at least some probability that he made the journey himself, as claimed. When large-scale, detailed maps had not been produced and those maps which were available were small-scale and full of error (jumbling names of different eras with bogus places and wrongly locating genuine names), there can be few other explanations for the level of coherence seen in historical itineraries and lists. Plotting the places in the Mani which Evliya Çelebi claimed to have visited produces reasonable results. Figure 6.3 suggests a connected itinerary consistent with the topography of the area north of the Milolangada. There are a few gaps (where was Ayo Varvara, for example?). Information from south of the pass is too sparse and fragmentary to reconstruct an itinerary for that area, though Evliya Çelebi claimed to have been there and the places mentioned can be identified. Evliya Çelebi went to the Mani with an army sent by the Grand Vizier, Fazıl Ahmed Köprülüzade, after the fall of Candia in September 1669. Its objective was to put down the Maniat rebels who

Page 7 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Evliya Çelebi, the Mani and the Fortress of Kelefa had provided assistance to the Venetians by running supplies into Candia and harassing Ottoman shipping. The genuineness of the expedition against the Mani is confirmed by the Ottoman historian, Silahdar Findıklı Mehmed Ağa.43 The Ottoman army, largely local troops from the Morea, concentrated on Mezistre (Mistra) in 1670–1 and was put under the command of Hüseyin Ali Pasha. Evliya Çelebi claimed that there were 10,000 troops, but the number included workmen to clear and prepare the way for the eighty cannon. When the army eventually set off for Londar (Leontari) and Kalamata, Evliya Çelebi stayed behind for a few days to finish off some work. He describes the narrow road and precipitous crags of the pass over Taygetos, which he eventually took, and mentions that his horse slipped on the bare rock. As the expedition proceeded into the vilayet of Manya (Maina/Mani), Evliya Çelebi noted the names of some of the villages passed and reported a number of others in the vicinity, ‘every one’, he said, ‘composed of fifty to sixty houses which resemble castles’.44 He noted the number of armed men that could be supplied, the presence of vineyards and gardens, silk production, land ownership and tax status. ‘The infidels (p.125) here always have their minds on war’, he reported.45 They reached Kalamata, which Evliya Çelebi claimed to have visited in 1077/ 1666–7. A delegation of elders and younger men from the Mani came to visit the serasker, Hüseyin Ali Pasha, pointing out to him that the vilayet of Manya had not been invaded by a Muslim for 200 years, but inviting him to visit the castle of Zarnáta with their blessing. The army (allegedly 20,000 strong now!) moved forward along the coast. We can follow the route to Zarnáta, which was put under siege (see Figure 6.3). Evliya Çelebi, who claimed that he had a beautiful singing voice, tells us that he sang the call to prayer when the fortress surrendered. The fortifications were apparently in a poor state and the serasker ordered that they should be pulled down and completely rebuilt. It took two months, according to Evliya Çelebi, who said he composed the dating inscription. The walls, he said, were 12 ars (9 metres)46 high and the circuit 1,070 paces. Then he gave further details of the fortress and described its varos (suburb, or extra-mural settlement).47 (p.126) The sequence of place names in the next section of Evliya Çelebi’s account suggests an excursion north-east from Zarnáta to Gaçiça (Gaïtsa), up to Piğa (Pigadia), Padalonya (Pano or Pende Alonia) and a narrow pass to the north (the Rindomo Gorge?) resembling the Nakura or Rende in Arabia. Here Evliya Çelebi turned back, in the direction of the qibla. The next place mentioned is the port of Mila (identified as Kardamyli/Skardamoula), followed by the plain of Lesin and the village of Leftini, the latter of which John Chapman has suggested might be Leptini, just west of Tseria. The Island of Pirastoz (modern Meropi) comes next and Evliya Çelebi tells how the gazis of Koron (Koroni), the large fortress across the gulf, attacked and destroyed it, transferring the inhabitants to the mountain inland, where they built the present village of Pirastoz (Prasteio). They were always in contention

Page 8 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Evliya Çelebi, the Mani and the Fortress of Kelefa with other villages, Evliya Çelebi said. ‘Utter infidels’, they kidnapped Muslims, especially ‘their gentle and beautiful girls’.

Figure 6.3. Evliya Çelebi’s itinerary in the Mani Peninsula, 1670.

The itinerary goes on in logical fashion—Ardovişta (Androuvitsa), Kastaniya (Kastania) and Pilaça (Platsa)—until we reach Kutufari (Koutifari) and Lankada, where the spatial sequence appears to break down. Lankada and Nomçin (Nomitsa) appear to reverse the route Evliya Çelebi was following, but then he wrote that he went in the direction of qibla for an hour to reach Ayo Varvara (unidentified but possibly Kivella or Garbelia), before going on for another hour towards Mila (Milia). He seemed to bypass the villages along the western edge of the mountains. The route then ran through Demirhisar (Sidirokastro), near which Evliya Çelebi reported that thousands of Muslim soldiers were slaughtered in 1037/ 1627–8, to Kıryopoli (Karioupolis). From there he must have gone westward along the Milolangada to Kirya Nero (Kryoneri) and round northwards to the ‘large village of Vitiloz’ (Vitylo/Oitylo). It had (p.127) a thousand stone houses, Evliya Çelebi tells us, ‘tall like castles, with small windows and tile roofs’. The inhabitants dressed like Franks, and lived from the sea, particularly hunting for slaves—Muslims whom they sold to the Franks, Christians whom they sold to the Muslims. ‘The harbour of Vitiloz is large’ and was said to accommodate ‘a thousand ships’ but was open to the south. The Venetian fleet often came there, Evliya Çelebi reported, and local people assisted it. At this point in his narrative, Evliya Çelebi tells that the serasker accepted his suggestion to build a fortress to defend the harbour. Hüseyin Ali Pasha told Evliya Çelebi to determine the size and shape of the fortress, the number of the garrison, and the quantity of provisions required. A conference seems to have been convened at which ‘the elders and priests of the village’ (not named) were present, together with gazis from Koron and Modon. They agreed that a fortress like Zarnáta would protect local ships coming from and going to Crete, Tunisia, Tripoli and Algeria, and it was necessary to build it. Evliya Çelebi, however, does not describe the actual building of a fortress. His next stop, though, was at Kelefa. It consisted of a hundred houses, built on a platform on the edge of ‘a precipitous ravine’. While Evliya Çelebi writes that the ‘vineyards and gardens are remarkable’ and notes that there were ‘300 demolished houses’ at Kelefa,48 he did not report a fortress. From Kelefa Evliya Çelebi went deeper into the Mani. The first place named is Kaliya (Karea). This is the name of a village which lies north-east of Kelefa, but since the next place name is Cibova (Tzimona, modern Areopolis), Evliya Çelebi may have confused Kaliya (Karea) with the similar sounding name, Keria, the name of a village south of Areopolis. From here Evliya Çelebi appears to have gone further south, even as far as Vatya (Vatheia) and Kaya Pasha Limani (Porto Kaio), a gunshot away from the fortress of Güllü (Keli, according to Dimitri Loupis49), where ‘the gate and walls still remain strong’, but which was ‘now’ deserted. However, the spatial sequencing does not fit the topography of this part of the peninsula. In fact, Evliya Çelebi’s itinerary was interrupted by the urgent need to secure food and drink, which required a hasty visit to the fortress of Pasova (Passava), and which he claimed to have visited in 1666–7.50 Returned to Zarnáta, taking with him 170 Muslims whom he had freed from slavery, Evliya Çelebi presented Hüseyin Ali Pasha with the results of a tax survey of seventy-seven villages. The compilation of this document was probably the reason for his travels in the Mani. What are we to conclude from Evliya Çelebi? He did not mention a pre-existing fortress at Kelefa. Nor does he report that one was actually built there—only proposed and agreed for the defence of Oitylo Bay. Zarnáta was rebuilt, apparently therefore one of the two forts mentioned

Page 9 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Evliya Çelebi, the Mani and the Fortress of Kelefa by Spon and Wheler a few years later, as we have seen. It is (p.128) logical to conclude that the other must be Kelefa. There was certainly a fortress at Kelefa on 25 September 1685 because it is reported to have surrendered on that day to German troops in Venetian service, nearly three weeks after Zarnáta had capitulated (6 September 1685).51 Military engineers in Venetian service planned the fortress (Figure 6.4). The plan published by the Geographer to the Venetian Republic, Father Vincenzo Coronelli, must have been drawn soon after Kelefa surrendered.52 That by Vasieur (also known as Le Vaseur, Vaseur and Vassor) and published by Andrews was probably produced during the time when Francesco Grimani was Provveditore Generale dell’ Armi in Morea (1699–1701) or when he was Governor General (1706–8).53 The refurbishment of all the Morea fortresses was under consideration, but the Ottoman army returned before much work could be carried out. Kelefa surrendered to an Ottoman naval squadron on 9 or 10 August 1715 during the lightning campaign which restored the Morea to the sultan’s rule: its walls were in poor repair, the garrison was under strength and the water supply had been cut. Zarnáta surrendered a week later.54 Both appear never again to have been garrisoned by Ottoman troops and fell into disrepair. (p.129) So, when was Kelefa built? Was it built in or shortly after 1670? If Zarnáta was actually rebuilt, rather than initiated from scratch by the Ottomans, then why not Kelefa as well? Could it be much older, ie. mid-thirteenth century? There is some evidence to suggest that there was something—even a castle—on the site between c.1248 and 1670.

Other Evidence The evidence consists mainly of brief references in a few travel accounts and portolans (sailing directions), but there are Figure 6.4. Venetian plans of the fortress of Kelefa, c. 1685. also intriguing references in a manuscript account of the Venetian campaign in the Morea in 1684–8, which need to be taken into account. The earliest literary source dates from 1335–52, and tells how the Irish pilgrim, Symon Semeonis, on his way to the Holy Land, sailed from Methoni ‘by Corone, a Venetian possession, past Mayna and Compana, a fort (castrum) of the Greek Emperor, to Porto Quaglio’.55 So a fort was visible on the Maniat coast from the arc traced by a vessel going from Koroni on the western side of the Gulf of Messenia to round Cape (p.130) Matapan at the end of the peninsula. The merchant antiquarian, Ciriaco of Ancona (1391–c.1455), visited the region in 1447. He reported crossing from Coroni to the Tainaron peninsula and landing at ‘Messenian Pylos’, a place where there were ‘still elegant remnants of the city’ and which an inscription allowed him to identify as ancient Bitylos. Ciriaco describes how, ‘no more than five stadia from the shore’,56 the party ‘saw a citadel built by later inhabitants out of [materials taken from] ancient buildings’.57 Was this Kelefa? The name itself first appears in a Venetian document of 1495,58 while a letter of Sultan Mehmed II of 1480 reports the killing of the subaşı of Megalo Mane—Grand Magne.59 But the linking of the name Vitylo/Oitylo with a fortress, as Ciriaco appears to do, and the possible confusion of Vitylo/Oitylo with Kelefa, is also apparent in a late seventeenth-century source.60 This manuscript account of the Venetian conquest of the Morea mentions both Vittolo and Celefà, but refers to both as fortezza. The manuscript also contains a Page 10 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Evliya Çelebi, the Mani and the Fortress of Kelefa pictogram of Vittolo (Figure 6.5). It depicts the fortress as densely built up and surrounded by crenellated walls. Round towers are shown at the corners and in the middle of the western and northern sides, while a rectangular gatehouse is shown at the centre of the eastern wall. Near the centre, surrounded by buildings, is a minaret. The drawing looks very like Kelefa (compare Figures 6.4 and 6.5). Examination of portolans and maps for the region dating from before 1670 might be expected to provide evidence for a fortress on the Bay of Oitylo, if one did indeed exist. The early maps are particularly disappointing, and while these have been the subject of biblio-cartographic research,61 analysis of their contents has been limited.62 No historical maps of the Mani itself are known. Instead the peninsula is shown, usually distorted in shape, as part of a wider region, whether the Morea, the Aegean Sea or the Mediterranean Sea. Maina, or a variant, is often marked as a place, but though there is something of a consensus that it lay to the west of Cape Matapan, the precise location varies from map to map. More might be expected of the portolans. As the Admiralty Pilot says, Porto Vitylo/Oitylo is the ‘best natural harbour’ on the Gulf of Messenia,63 and one might have thought it was worth including. Several of the published portolans do not cover the interior parts of the Gulf of Messenia above the latitude of Koroni at all, but take the navigator coming from the west straight across the mouth of the Gulf of Messenia to Porto Kaio on the eastern side of Cape Matapan (see Figure 6.2) (e.g. Portolan Parma-Magliabecchi of the early to mid-fifteenth century; Portolan des Gratiosus Benicasa of 143564). However, a Greek portolan of 1534 published by Delatte mentions a kastro as a navigational mark on the harbour of Vitylo.65 While kastro can mean a walled town in medieval and modern Greek, its primary meaning is the same as castrum, from which it is derived—fort or castle.66 The famous Kitab-ı Bahriye produced by the Ottoman admiral, Piri Reis (c. 1470–1554) in 1520 is just as ambiguous, if not more so.67 It notes Kavu de Viteroza and a ‘good anchorage’ on its eastern side. Despite the similarity in name—Viteroza/Vetiloz/Vitylo— I think that the headland is probably Kavo Grosso and the anchorage Mezapo Bay (see Figure 6.2), because the next navigational mark noted is Koron, forty miles to the west-northwest. However, Piri Reis goes on to note a castle called Kastah-ı Fireng, ‘which is a castle located inside a bay on the north-western side of Koron castle’. Although the editor of the text identifies this with a ‘small castle near the city of Kalamai at the head of the gulf of Messenia’,68 Kalamata does not bear north-west from Koroni. Nor does Porto Vitylo, for that matter but, as noted above, it is the most significant harbour on the Gulf and likely to have been noted by Piri Reis. In fact, the fortress could be Kelefa. The name ‘Castle of the Franks’ is suggestive. (p.131)

(p.132) Conclusion The currently available evidence, which I have set out, is inconclusive. However, I think it is more compelling than the evidence used to support other claimants to Figure 6.5. Pictogram of Vittolo, c.1685. being the site of Grand Magne. In some Querini-Stampalia C1. IV, Cod. Xciii, n. 1347, 77. Published with the permission of the ways they are claimants only because some Fondazione Querini Stampalia Onlus scholars have been unconvinced by the case Venezia. for Kelefa and have sought alternatives. The question of whether Kelefa is—or is not— Grand Magne can only be resolved if new evidence is brought into the argument. This may be

Page 11 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Evliya Çelebi, the Mani and the Fortress of Kelefa obtainable from the Venetian and Ottoman archives and from a careful archaeological study of Kelefa. The State Archives in Venice contain the reports and correspondence of the commanders of the Maniat fortresses from 1698 to 1718.69 These should be examined in the hope of finding references to the state of the fortifications at Kelefa when the Venetians took over, to any building work undertaken and to the nature of the buildings within the enceinte. The end of term reports (relatione) of the governors-general might also contain useful snippets of information, despite their high rhetorical language. Particular attention should be given to the papers of Francesco Grimani in the Grimani ai Servi Archive, now part of the State Archives, for they may contain relevant correspondence and reports.70 More information might be gained from a firsthand (p.133) study of Vasieur’s coloured plan of Kelefa in the file of maps associated with Grimani in the Gennadeion in Athens and now published in the reprint of Andrews’s book.71 It shows relatively few buildings inside the walls and these appear to be storehouses and cisterns, but include a barrack block and, more or less at the centre, a church (see Figure 6.4). Coronelli’s plan (Figure 6.4) shows an L-shaped block in the bottom righthand corner (presumably C and E on Vasieur’s plan) and three other buildings (the more central one is probably the church). Albrizzi’s pictogram (1687) (see Figure 6.5),72 contemporaneous with Coronelli’s plan, shows the same ground plan as the other two drawings (possibly as a mirror image), but perhaps suggests that the church had once been a mosque—at least, the central building appears to have had a dome and a tower (or minaret?). More plans may survive in the State Archives at Venice. A number of the Venetian maps of the Morea were taken to Vienna in the nineteenth century, and the collection in the War Ministry Archive should be re-examined.73 Comparative study of the Venetian plans could provide a framework for the careful examination of the extant remains of the fortress at Kelefa. The work of Bon and Traquair, mentioned earlier, seems to have been rapid and superficial, while Andrews tended to follow Traquair and dismissed the fortifications, both layout and masonry, as ‘Turkish’. Fifty to a hundred years on, the circuit of the walls and the bastions are very clear (Figure 6.6), but details of their construction need to be examined to establish when they were built. It is perhaps too much to expect that somewhere in the rubble a Venetian or Ottoman inscription has survived from one or other of the gates. The interior of the enceinte has been divided into closes and the former buildings are now heaps of stones, but the Venetian plans provide a starting point for identifying what they were. Any archaeological survey of the interior should probably focus on the substantial remains of the central building, marked as a church by Vasieur. The form and orientation of the building will be crucial to identifying its original purpose. If it consists of a single chamber, perhaps with an open narthex, and it is orientated towards the qibla (roughly 135° or south-east from Kelefa), then the chances are that the structure was built as a mosque. On the other hand, an orientation more towards the east (90°) would suggest that the building was originally a church. Recovery of the ground plan should confirm a tentative identification based on orientation. Any remodelling to convert a church into a mosque, or vice versa, is unlikely to have changed the basic ground plan. An archaeological survey is thus fundamental to the resolution of the problem of whether Kelefa was Grand Magne or not. (p.134)

Page 12 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Evliya Çelebi, the Mani and the Fortress of Kelefa Finally, let us not forget the Ottoman Archives in Sofia and Istanbul. Their potential is considerable, but—as far as I am aware—no attempt has been made to search them for specifically Maniat material. If Figure 6.6. An aerial view of the fortress of Kelefa. Courtesy of P. S. Katsafados. Kelefa existed when the Ottomans took over the Mani in the 1460s, and if they used it after that, there ought to be some references to its governor, garrison and provisioning, as well as to any refurbishment work. Curiously, Evliya Çelebi does not mention a fortress at Kelefa when he visited the village of that name in 1670, unless the ruins, which he described as ‘houses’, were it. If the Ottomans built or rebuilt the fortress, authorisations, building registers and progress reports should exist from 1670 onwards. They ought to reveal whether a new fortress was built or whether a pre-existing one was renovated, as Evliya Çelebi said was the case at Zarnáta. Correspondence from the serasker, Hüseyin Ali Pasha, and the Grand Vizier, Fazil Ahmed Köprülüzade, might also survive. On his return to Zarnáta from the south of the Mani, Evliya Çelebi says he was ordered by the serasker to take the keys of that fortress to Istanbul, together with a report for the archives, not only on the capture and rebuilding of the fortress but also about its ‘form and construction…, the gate and the wall, the buildings which are found in its interior’.74 This document and the specifications for a fortress to defend the harbour of Vitylo, which Evliya (p.135) Çelebi said he was asked to prepare, may be in the Prime Ministry Archive in Istanbul. If they really exist, a byproduct of the research might be to discover Evliya Çelebi’s real name. Note. I am grateful to John Chapman for many helpful suggestions on bibliography and topography, to Takis Katsafados and Jon van Leuven for reviving the debate on the location and identity of Grand Magne, to Victor Ménage for help with the Ottoman text of Evliya Çelebi in the early days of my researches on the Mani, and to the Cartographic Unit of the University of Southampton for providing the illustrations. Notes: Proceedings of the British Academy 156, 113–135. © The British Academy 2009. (1) The Chronicle of the Morea, ed. J. Schmitt (London: Methuen, 1904). (2) David Jacoby, ‘Quelques considérations sur les versions de la Chronique de Morée’, Journal des Savants (1968), 133–89. (3) C. Hopf, Chroniques Gréco-Romaines inédites ou peu connues (Berlin: Weidmann, 1873), 385. (4) V. Coronelli, An Historical and Geographical Account of the Morea, Negropont, and the Maritime Places as far as Thessalonika, Englished by G. W. Gent. (London: Matth. Gillyflower, 1687), 102–3; P. Paruta, The History of Venice, translated by Henry, Earl of Monmouth (London, 1658), 40. (5) P. Dallas (ed.), The Adventures of Captain Alonson de Contreras (New York: Paragon House, 1989), 25; George Finlay, A History of Greece from its Conquest by the Romans to the Present Time, 146 B. C. to A.D. 1864 (Oxford: Clarendon Press, 1877), 5: 108–14; A. Tenenti, Naufrages, corsaires et assurances maritimes à Venise, 1592–1609 (Paris, 1959), 45–7; Joseph von HammerPurgstall, Histoire de L’Empire Ottoman (Paris: Bellizard et al., 1838), 7: 205.

Page 13 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Evliya Çelebi, the Mani and the Fortress of Kelefa (6) R. C. Anderson, Naval Wars in the Levant, 1559–1853 (Liverpool: Liverpool University Press, 1952), 277–308. (7) J. M. Wagstaff, ‘John Philip Morier’s account of the Mani, 1804’, in Y. Saïtis (ed.), Máni: Témoignes sur l’espace et la société (Athens: Institut de Recherches Néohelléniques, 1996), 273– 87. (8) Most of the material in this and the next paragraph is drawn from Anderson, Naval Wars, chs. 2–6, 13–15. (9) The Mediterranean Pilot (London: Hydrographic Department, Admiralty, 1957), 3: 63. (10) Wagstaff, ‘John Philip Morier’s account’. (11) Crusaders as Conquerors: The Chronicle of the Morea, trans. H. E. Lurier (New York: Columbia University Press, 1964), 158–9. (12) Ibid., 159. (13) Ibid., 160–1. (14) Schmitt, Chronicle, lines 4330, 4425, 4662; Chronique de Morée (1204–1303), ed. J. Longnon (Paris: Société de l’Histoire de France, 1911), lines 207, 218, 317, 326, 502, 762. (15) J. M. Wagstaff, ‘Further observations on the location of Grand Magne’, Dumbarton Oaks Papers, 45 (1991), 141–8. (16) Lurier, Crusaders as Conquerors, 19, 197 n. 79. (17) Wagstaff, ‘Further observations’. (18) P. S. Katsafados, ‘On the quest for the Castle of Grand Magne in Mani’, www.pskatsafados.gr. (2004); J. van Leuven, ‘Blank spots and grey zones: medieval mapping in southern Greece’, in G. Tolias and D. Loupis (eds.), East Mediterranean Geographies (Athens: Institute for Neohellenic Research, 2004), 243–62. (19) Katsafados, ‘On the quest for the castle of Grand Magne’. (20) P. Hetherington, Byzantine and Medieval Greece: Churches, Castles and Art (London: John Murray, 1991), 138–9. (21) Ibid., but see Wagstaff, ‘Further observations’ for my reasons for rejecting this identification. (22) The Mediterranean Pilot, 55. See also H. M. Denham, The Aegean: A Sea-Guide to its Coasts and Islands (London: John Murray, 1963), 9–10. (23) Eulogy by Demetrius Chrysoloras in S. Lambros, Palaiologeia kai Peloponnisiaka (Athens, 1929), 3: 242; Z. Kougeas, Peri ton Meligkon tou Taygetou es Aformis Avekdotou Bizantinis Epigrafis tis Lakonias (Athens: Pragmateia tes Akademeias Athenon, 1950). (24) C. Sathas, Documents inédits relatifs à l’histiore de la Grèce au Moyen Age (Paris, 1883), 5: 30–9.

Page 14 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Evliya Çelebi, the Mani and the Fortress of Kelefa (25) R. Traquair, ‘Laconia 1: medieval fortresses’, Annual of the British School at Athens, 12 (1905–6), 261–70; A. Bon, La Morée franque: recherches historiques, topographiques et archéologiques sur la Principauté d’Achaïe (Paris: Boccard, 1969), 3: 502–8; K. Andrews, Castles of the Morea (Princeton: Princeton University Press, 2006), 49–75. (26) William Martin Leake, Travels in the Morea (London, 1830), 1: 312–13. A. P. Fourikis, Paratiriseis eis ta Toponimika to Chronikon Moreos (Athens, 1928), 12: 29–43 misquotes Leake to the effect that Kelefa was built on the remains of a more ancient castle. (27) G. M. Thomas, Diplomatarium Veneto-Levantinum sive Acta et Diplomata res Venetas, Graecas atque Levantinas illustrantia, a. 1300–1350, I (Venice: Deputazione Venta di Storia Patria, 1880), no. 72, 127; G. Morgan, ‘The Venetian claims commission of 1278’, Byzantinische Zeitschrift, 69 (1976), 411–38. (28) Bon, La Morée, Appendix A, 692. (29) S. B. Kougea, ‘Simvoli eis tin Istopian tis BD. Manis’, Ellinika, 18 (1933); A. Daskalakis, I Mani kai i Othomaniki Autokratia, 1453–1821 (Athens, 1928), 90–1. (30) J. Spon, Voyage d’Italie, de Dalmatie, de Grèce et du Levant fait en années 1675 et 1676, 2nd edn. (Amsterdam, 1679), T. 1, 122–3; G. Wheler, A Voyage into Greece. By G. W. Esq., in Company with Dr. Spon of Lyons (London: William Cademan et al., 1682), 47. (31) Wheler, Voyage. (32) R. F. Kreutel, ‘Neues zur Evliya-Çelebi-Forschung’, Der Islam, 48 (1971), 269–98; Evliya Çelebi in Diyarbekir, ed. and trans. M. van Bruinessen and H. Boeschaten (Leiden: Brill, 1988), 5–8. (33) R. Dankoff, An Ottoman Mentality: The World of Evliya Çelebi (Leiden: Brill, 2004), 7. (34) Ibid., 9–21. (35) Suraiya Faroqhi, Approaching Ottoman History: An Introduction to the Sources (Cambridge: Cambridge University Press, 1999), 132–3, 160–1. (36) Dankoff, An Ottoman Mentality, 157–8. (37) Evliya Çelebi in Diyarbekir; Evliya Çelebi in Bitlis, ed. and trans. R. Dankoff (Leiden: Brill, 1990). (38) Evliya Çelebi in Diyarbekir, ed. and trans. M. van Bruinessen and H. Boeschaten, 7. (39) Ibid. (40) Evliya, Seyahatname, 8: 256–70; Evlia Tselempi: Odoiporiko stin Ellada (1668–1671): Peloponnisos—Nisia Ioniou—Kriti—Kyklades—Nisia Anatolikou Aigaiou, trans. and ed. D. D. Loupis (Athens: Ekdoseis Ekati, 1994), 263–88. (41) J. M. Wagstaff, ‘Settlements in the south-central Pelopónnisos c.1618’, in F. W. Carter (ed.), An Historical Geography of the Balkans (London: Academic Press, 1977), 197–238.

Page 15 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Evliya Çelebi, the Mani and the Fortress of Kelefa (42) Suraiya Faroqhi, ‘Foreword’ to Dankoff, An Ottoman Mentality, vii–xviii. (43) Silahdar Findikli Mehmed Ağa, Silahdar Tarihi (Istanbul: Devlet Matbaasi, 1928), 1: 525–9. The reference was kindly provided by Professor Victor Ménage. (44) Evliya, Seyahatname, 8: 260; Loupis (ed.), Evlia Tselempi, 269. (45) Evliya, Seyahatname, 8: 260; Loupis (ed.), Evlia Tselempi, 268. (46) Evliya, Seyahatname, 8: 263; Loupis (ed.), Evlia Tselempi, loc. cit. (47) Varousi is the local name for the village of Stavropigion, just below the fortress. (48) Evliya, Seyataname, 8: 267; Loupis (ed.), Evlia Tselempi, 282. (49) Loupis (ed.), Evlia Tselempi, 340. (50) Evliya, Seyahatname, 8: 268; Loupis (ed.), Evlia Tselempi, 284. (51) A. Locatelli, Historia della Veneta Guerra in Levante contro L’Impero Ottomano (Colonia, 1705), 161–70. (52) Coronelli, An Historical and Geographical Account of the Morea. (53) Andrews, Castles of the Morea, 8–10. (54) K. D. Mertziou, ‘Pote kai pos epesei i Mani eis cheiras ton Tourkon to 1715’, Peloponnisiaka, 3–4 (1958–9), 276–87. (55) M. Esposito, ‘The pilgrimage of Symon Semeonis: a contribution to the history of medieval travel’, Geographical Journal, 50 (1917), 335–52; Itinerarium Symonis Semeonis ab Hybernia ad Terram Sanctam, ed. M. Esposito (Dublin: Dublin Institute for Advanced Studies, 1960), 41. (56) A stadium is equal to about 194 yards (P. Harvey (ed.), The Oxford Companion to Classical Literature (Oxford: Clarendon Press, 1955), 406), making the estimated distance about 820 yards. (57) Cyriac of Ancona, Later Travels, ed. and trans. E. W. Bodnar and C. Foss (Cambridge, Mass.: Harvard University Press, 2003), 310–11. (58) D. V. Vagiakakos, ‘Bizantina onomata kai eponima ek Manis’, Peloponnisiaka, 3–4 (1958–9), 185–221. (59) D. G. Wright, ‘Bartolomeo Minio, Venetian administrator in 15th century Nauplion’, Electronic Journal of Oriental Studies, 3/5 (2000), Appendix 5, 203–4. (60) Querini-Stampalia C1. IV, Cod. XCIII, n. 1347, Distinti Ragualij delle Fortezze presse nel Regno della Morea soto il Comando dell’Eccellentissimo Kavaliere Procuratore Capitan Generale Francesco Morosini, 79–94; E. D. Diata and K. G. Tsiknakis, Me tin Armada sto Moria, 1684–85: Anekfoto imerologio schedia (Athens: Olkos, 1998), 77–81.

Page 16 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Evliya Çelebi, the Mani and the Fortress of Kelefa (61) S. Stavridou, ‘I chartografisi tis Lakonias apo ta Bizantina cheiografa tou 13ou eos ton 19on aiona’, in Saïtis (ed.), Máni, 39–51; A. Tavtoulos, ‘I entipi chartografisi tis Manis (1605–1805 ai.). Istorikis kai geografikes plirofories’, in ibid., 52–71. (62) P. S. Katsafados, Ta Kastra tis Mainis (Athens: author, 1992), 287–90. (63) The Mediterranean Pilot, 3: 55. (64) K. Kretschemer, Die italienischen Portolane des Mittelalters (Hildesheim: Georg Olms, 1962). (65) A. Delatte (ed.), Les portulans grecs (Liège: Faculté de Philosophie et Lettres; Paris: E. Droz, 1947), 57. (66) D. Dimitrakou, Mega Lexikon tis Ellinikis Glossis (Athens: Dimitrakou, 1951); E. A. Sophocles, Greek Lexicon of the Roman and Byzantine Periods (New York: Charles Scribner’s Sons; Leipzig: Otto Harrassowitz, 1888). (67) Piri Reis, Kitab-i Bahriye (Istanbul: Historical Research Foundation, Istanbul Research Center, 1988), 2: 643–51. (68) Ibid., n. 164. (69) A. da Mosto, L’Archivio di Stato di Venezia: Indice Generale, Storico, Descrittivo ed Antitico, Part III (Rome, 1940), 20. (70) T. Gritsopolou, ‘To en Venetia archaion Grimani kath’ oson afora eis tin Peloponnison’, Peloponnisiaka, 7 (1969–70), 396–9. (71) Andrews, Castles of the Morea, pl. 7. (72) G. Albrizzi, Esatta Notitia del Peloponneso, Volgarmente Peninsola della Morea (Venice, 1687). (73) Olga Katsiardis-Hering, ‘Venezianische Karten als Grundlage der historischen Geographie des griechischen Siedlungsraumes (ende 17. und 18 Jahrh.)’, Mitteilungen des Österreichischen Staatsarchivs, 43 (1993), 281–316. (74) Evliya, Seyahatname, 8: 270; Loupis (ed.), Evlia Tselempi, 288.

Access brought to you by:

Page 17 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project VICTOR OSTAPCHUK SVITLANA BILYAYEVA

DOI:10.5871/bacad/9780197264423.003.0007

Abstract and Keywords The northern frontiers of the Ottoman Empire lay across a swathe of lands between Hungary and Iran, arcing through the principalities of Wallachia and Moldavia, then north of the Black Sea through the steppes of southern Ukraine, and finally proceeding further east along the Caucasus Mountains as far as the Caspian Sea. In a frontier region such as the one on the northern Black Sea, where environment, human geography and historical traditions made the steppe an alien place that did not readily yield to control and assimilation, the fortress was indispensable for maintaining the centre's presence. As imperial presence in such an area was anchored at and emanated from the fortress, the fortress can be seen as a prime target of a strategy aimed at learning about this frontier of the Ottoman world. Keywords:   Black Sea, Akkerman, archaeology, Ottoman Empire, imperial fortress

THE NORTHERN FRONTIERS OF THE OTTOMAN EMPIRE lay across a swathe of lands between Hungary and Iran, arcing through the principalities of Wallachia and Moldavia, then north of the Black Sea through the steppes of southern Ukraine (the western end of the great Eurasian steppe), and finally proceeding further east along the Caucasus Mountains as far as the Caspian Sea (Figure 7.1). The central portion of this arc, roughly from the lower Danube to the Kuban and Don Rivers, was predominantly a steppeland of both flat and rolling plains, overlaid by a network of rivers great and small, as well as ravines, gullies and forests. To the heedless, even if powerful, outsider the vast Black Sea steppe could seem a deceptively accessible and passable region. Instead, it was usually a formidable obstacle for both armies and merchants who were unaccustomed to its conditions and without proper contacts with its nomadic inhabitants. Some of its main rivers facilitated, albeit did not guarantee, passage whether coming from the Ottoman south or Slavic north.1

Page 1 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project With the fall of Constantinople in 1453, Mehmed II moved quickly to establish Ottoman suzerainty, though not total control, over most of the shores of the Black Sea. Probably the greatest Ottoman achievement in this region was the relationship with the newly formed Crimean Khanate, a Tatar successor state of the Golden Horde that would survive until the end of the eighteenth century. Though nominally a subject of the Porte from 1475, the Khanate had a unique relationship with Istanbul, receiving regular and official Ottoman gifts which counted as tribute in the eyes of the khans, who were, thanks to their Chinggisid descent, of a more prestigious line than the Ottomans. Until the end of the seventeenth century, they maintained a virtually independent foreign and military policy. The Ottomans benefited immensely from this connection, for they managed to have a sufficient degree of control over most of the khans. This obviated the need for outright conquest of the Black Sea steppes—something which, as experience would show, they could hardly accomplish anyway. Nor, because of its low sedentary population, did they need outright possession for the benefit of their tax collecting and military machines. From the hands of the Crimean Tatars and their pure nomad brothers in the steppe, the Nogays, the Ottomans received a constant flow of slaves captured in Ukraine, southern Muscovy and the Caucasus region, which was a mainstay of both the Tatar and Ottoman economies. (p.138) (p.139) With the conquest in 1484 of Kili (Kilia) and Akkerman (BilhorodDnistrovs’kyi), the two remaining major fortresses (and ports) not yet in Ottoman hands, the Black Sea became in effect an Ottoman possession. That summer Sultan Bayezid II led the first major campaign of his reign and succeeded where eight years earlier his great conquering father had failed. The location of these ports on the mouths of the Danube and Dniester Rivers respectively meant that whoever possessed them could control the important trade routes which connected the Black Sea with Figure 7.1. The Black sea region central and north-east central Europe. For the Principality of Moldavia, led by the formidable and independent-minded Stephan the Great (r. 1457–1504), the Ottoman takeover of Kili and Akkerman meant an effective end to his struggle to resist Ottoman suzerainty, and in 1486 Moldavia became an Ottoman tributary. Kili, Akkerman and the southern part of the Bucak (Bessarabia) were attached to the sancak of Silistre (Silistra), part of the beylerbeyilik of Rumeli. However, the conquest of the region was not finally completed until 1538 when Süleyman the Magnificent took Tighinia/Bender, the remainder of the Bucak and the stretch of coast between the Dniester and the Dnieper. This is sometimes taken to mark the beginning of the Black Sea as an ‘Ottoman Lake’ as it is commonly known in the literature.2 Subsequently the sancak of Akkerman (also known as Bender and Akkerman) was formed, consisting of the kazas of Akkerman, Kili, Bender and Cankerman (also known as Özi, modern Ochakiv).3

Page 2 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project The northern rim of the Black Sea region remained part of a broad and deep steppe frontier and in large part a sparsely populated wilderness zone. Hence names that denote ‘Deserted Plains’ (Lat. Campi Deserti) or ‘Wild Fields’ (Pol. Dzikie Pole) were applied to parts of it. Yet it contained entire societies, such as the nomadic Little Nogay Horde and the sedentary Zaporozhian Cossacks, who, isolated in their distant lower Dnieper island and marsh refugia (the Zaporozhia, see Figure 7.1), were insulated from (p.140) control or attack by, for example, Poland-Lithuania or the Ottomans, yet were able to seize opportune moments to strike at their neighbours by land or by water. The Crimean Khanate was able to project Tatar cavalry power throughout this steppe frontier and beyond to its Slavic neighbours in the north, towards the Danube and Hungary in the west, and east into the Caucasus. All these players derived their security vis-à-vis the surrounding powers to a greater or lesser extent from the natural features of this steppe frontier. Because there were relatively few towns and a sparse nomadic and even sparser rural sedentary population, fortresses were particularly important nodes, with sufficient and reliable manpower and firepower to protect the Ottoman Black Sea and with indispensable administrative capacity to support and regulate trade between the empire and the northern countries of Poland– Lithuania and Muscovy. They were usually located along or at the mouths of rivers, especially on the great ones—Kili on the Danube (also of importance for east–west trade along this river), Akkerman and Bender on the Dniester, Özi and Kılburun (Kinburn) on the Dnieper, and Azak (Azov) on the Don, as well as KerŞ (Kerch) (plus later Yenikale) and Taman on the Straits of Kerch where the Sea of Azov flowed into the Black Sea. Our view of the frontier here is primarily from the vantage point of the Akkerman Fortress Project,4 a study of one of the best preserved and documented fortifications on the Black Sea coast. While the pre-Ottoman, mostly Moldavian period of the fortress is of interest,5 the main focus is on the Ottoman period, which lasted from 1484 to 1806. Even though this was the longest stretch of Akkerman’s history, it has been the most neglected, mainly for reasons of antiTurkish bias and ignorance of or lack of access to the Ottoman sources. However, substantial Ottoman archival records concerning various aspects of the fortress, particularly its repair and reconstruction, have survived. The historical source base for the project includes both documentary and narrative records in languages as diverse as Polish, Romanian, Russian, Ukrainian, as well as English, French and Italian, but for Ottoman Akkerman the Ottoman sources are not only more numerous and detailed, but for our purposes mostly of higher quality than the others. As a fully-fledged historical and archaeological project the Akkerman Fortress Project is still very much work in progress, and we ask the reader to bear in mind the tentative nature of many of our results.6

(p.141) Akkerman in its Frontier Setting The Black Sea, with its natural resources and commerce, was considered vital for the well-being and growth of the state and its great metropolis, Istanbul. Thus the sea was also seen as a heartland of the realm or even a private dominion (in the sources occasionally called a harem) of the Sublime Porte. During the Moldavian and Ottoman periods Akkerman was a crossroads of trade and military routes by both land and sea. These included a coastal route from the Danube to the Dnieper that had to cross or go around the many long estuaries (liman) along the way, another along the Dniester River, and a sea route that straddled the Rumelian coast from Istanbul to Özi and the Crimea. The commercial importance of the coastal land route is not well known, though the sea route was of great importance for the military and trade. The route along the Dniester, the so-called ‘Moldavian Route’, connected the important city of Lviv (Lwów,

Page 3 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project Lemberg) with the Black Sea and from there with Istanbul, Bursa, Sinop, and other points in the Black Sea basin. This was a rich artery, run largely by Armenian merchants, transporting silk, spices, wine, wax and hides, as well as foodstuffs. Akkerman and its region became an important source of food for Ottoman markets, particularly Istanbul—huge annual quotas of sheep were to be sent to the capital, and grain and fish products were also significant.7 In an attempt to better exploit the rich soil of the region, the Ottoman state promoted colonisation of nomadic lands, encouraging the formation of large farms or ranches (hence known as a type of çiftlik) on their so-called ‘abandoned’ or ‘empty’ lands.8 However, despite attempts to colonise and sedentarise the Bucak, the Ottomans were unable to settle completely or otherwise eliminate its nomads. Instead, there were periods of nomadic resurgence. For example, in the first decade of the seventeenth century, a branch of the Nogays left the Volga–Don-Kuban region, crossed Crimean-controlled territory and established themselves in the Bucak steppe. There they were even better placed than the Crimean Tatars for slaving raids, being much closer to well-populated areas, such as the more westerly Ukrainian lands of the Kingdom of Poland (Podolia, so-called Red Ruthenia (the Lviv region), and Volhynia), and to the occasional rich target of opportunity of Moldavia. In the 1620s an intense rivalry arose between the Bucak Horde and the Crimean Khanate that would involve the Ottomans, Poland and the Zaporozhian Cossacks. The rivalry was not only a military and political one: with the success of the Bucak Horde in the slaving business, Akkerman challenged Kefe’s position as the key Black Sea slave port. (p.142) Moldavia, at the far western end of the Black Sea frontier, by virtue of its natural features alone—considerable uplands and extensive forests—should be considered a transition zone rather than a fully-fledged part of the steppe frontier. Sandwiched between the lands of two large empires—regular Ottoman territories and those of Poland–Lithuania—it was largely a buffer state. Its geopolitical position subjected it to various, often disruptive, influences from all sides—Ottoman, Polish, Tatar and Cossack. Given this array of actors, can we even speak of the frontier at Akkerman as a simple stretch of the Ottoman northern Black Sea frontier that separated the Ottoman Empire from the northern countries of Poland–Lithuania and Muscovy? For in fact on this territory there were simultaneously several frontiers. It was not even, as where the Hapsburg, Venetian and Ottoman empires met, a triplex confinium, but rather a kind of ‘multiplex confinium’—the Ottomans versus Poland–Lithuania and the quasi-independent Ukrainian Cossacks, Moldavia versus the Tatars of the Bucak and of the Crimea, the Bucak Horde versus the Crimean Khanate. To the west of Akkerman was the Ottoman-Moldavianearlier Lithuanian/later Polish frontier continuum.

The Akkerman Fortress Project The Akkerman fortress is a large complex with a circumference of more than two kilometres, an area of more than nine hectares, thirty-four towers and bastions, and six gates (Figure 7.2). On its landward sides there is an imposing tall and thick wall and on the exterior a smaller half-wall rising out of the scarp of a deep and wide ditch (on average 20 X 14 metres). The complex has three main subdivisions—the citadel, the garrison yard and the civil yard (Figure 7.3a); to the north, and below the main fortress is another section, on a level with the Dniester estuary, known as the port yard. The origins of Akkerman fortress are not well known. It was built on the site of Tyras, an ancient Greek colony of Miletus. The citadel came first—there have been versions of Ruthenian, Genoese and Moldavian origin dating to anywhere from the thirteenth to the early fifteenth century. During the century prior to the Ottoman takeover it was held by the Principality of Moldavia and the three main sections of the fortress were constructed.9 What

Page 4 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project stands today is a rare, largely intact survivor of the great fortresses that once defended this frontier. In appearance it preserves much of the form of a highwalled late medieval fortress, though the thickness of the walls indicates that they were designed to face fifteenth-century cannon fire. Prior to 2006, archaeological work concentrated on the port yard, the most wholly Ottoman portion of the complex, where two structures were unearthed—an enigmatic (p.143) originally late fifteenth-century barbican-like structure jutting out of the shore wall towards the estuary (Figure 7.3a–b, Figure 7.4), and a late fifteenth or early sixteenthcentury Ottoman bath-house, or hamam (Figure 7.5).10 In addition to excavation, a GPS survey of the entire fortress complex commenced in 2006, and in 2007 a resistivity survey was begun. Examination and analysis by a fortress specialist was also carried out in 2006 and 2007.11 As far as the written primary sources are concerned, the historical team has so far directed its main efforts towards the Ottoman archives in the Prime Ministry Ottoman Archive and the Topkapı Palace Archive. Approximately 150 archival units have been located so far, and the search for relevant material is ongoing. The units consist of both single sheet documents, files of such documents, and bound registers (defters), and amount to about one thousand often largeformat pages of text. Most of the fortress plans are in Russian archives and a good number have been presented in the works of Mariana Şlapac, in particular in her seminal historical architectural monograph on Akkerman fortress.12

Akkerman in the Fifteenth to Seventeenth Centuries We divide our survey of Ottoman possession and maintenance of Akkerman in the frontier context into two periods, an early (1484–1699) and a late (1700–1807) one. There are few documents on the fortress itself from the early period. The near silence in the documents and narrative sources on details of repair and construction suggests that in this period little or no major work was done to reconstruct the main parts of the fortress complex—the citadel, the garrison yard and the civil yard.13 While there is record of some repairs to the existing structure, there is as yet no evidence of any planned reconstruction projects.14 Typical for documents from this period are blanket statements about a given repair. For example, in 1608 an order sent to Akkerman and Bender states only that urgent repairs to Akkerman’s ‘upper tower’ (yukaru kule) are to be carried out, without providing any further details.15 However, civil buildings in and around the fortress were constructed during this period. The great Ottoman traveller Evliya Çelebi noted that in the suburb (varos, the civil yard?) there was a Bayezid II mosque with a single minaret (today assumed to have been where a single minaret survives on the foundation of a church),16 as well as a Sultan Selim17 mosque and medrese, other mosques and smaller mescids (neighbourhood mosques), and other buildings.18 In one sixteenth-century survey register (tahrir defteri) there is evidence of about a dozen quarters19 and, for example, a register from 1570 notes at least seventeen hans and caravanserais, seven grain depots, and 190 vaklf shops.20 Being one of the largest and best fortified sites that the Ottomans inherited in the northern Black Sea region, it seems that its late medieval form was sufficient. This allowed it to withstand threats from traditional foes in the region, as well as to project power over the nomadic and settled inhabitants of the surrounding steppe and also protect the town of Akkerman outside the fortress as an emporium of Black Sea and Dniester River trade. (p.144)

Page 5 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project (p.145) (p.146) (p.147)

Figures 7.2a and b. Views of Akkerman fortress. (a) Panorama from the liman. Left to right: citadel, garrison yard north-west wall, civil yard (Fireplace Tower and Water Gate, nos. 23 and 22 in Figure 7.3a); port yard is below the garrison and civil yard wall (from the restored round tower, no. 25, to the Water Gate, and beyond to no. 27, not in the photograph); remains of barbican in front of Fireplace Tower. (b) Citadel and garrison yard walls (right to left: Millet and Maiden Towers, Pasha Bastion, Fisher Tower (no. 30)). Photographs by Ihor Zhuk.

Figure 7.2c and d. (c) Main wall with Guardian Tower (no. 3) and part of Wheel Tower (no. 4), ditch scarp with topped halfwall, half-tower. (d) The ditch at Storeyed Tower (no. 11). Photographs by Ihor Zhuk.

Figures 7.3a and b. (a) Akkerman with locations identified from Ottoman documents (italics indicate tentative identities). Plan based on the survey made in 1955, Mariana Şapac, Belgorod-Dnestrovs’kaia krepost’, 90. (b) Akkerman fortress (satellite image). Google Earth.

Page 6 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project (p.148) The Port Yard and Associated Structures At the very end of the Moldavian or at the beginning of the Ottoman period, a new walled section, the port yard, was for some reason added to the original formidable complex. Architectural analysis of the surviving shore walls, which are much lower and thinner than the main walls of the fortress above them, and of their barbicanlike protrusion, is still inconclusive. However, this suggests a late fifteenthcentury origin, and excavated artefacts are overwhelmingly Ottoman, with little trace of Moldavian presence. Since both prior to and after the fall of Constantinople the Ottomans made several attempts to take Akkerman, fortification of the area below the main fortress and next to the estuary by the Moldavians would have made sense, especially if there were buildings or activity carried out there that were worthy of defending. No written sources known to us refer to a Moldavian port yard. Yet we do have an undated, but very early postconquest Ottoman document referring to

Figure 7.4. Barbican viewed from top of civil yard wall, next to Fireplace Tower (no. 23). Photograph by Victor Ostapchuk.

Figure 7.5. Port yard hamam (between Fireplace Tower and Water Gate). Photograph by Caroline Finkel.

complete reconstruction of a low wall along the shore.21 It is still unclear whether or not there was actually a port in the so-called port yard—certainly today the estuary here is very shallow and we know that in the eighteenth century there was a port or landing place (iskele) a few hundred metres downriver (the Istanbul iskelesi).22 However, it is possible that river craft or boats launched from galleys anchored offshore would beach at the port yard.23 But what was the port yard intended to defend? Perhaps it served as a kind of extended barbican for the Water Gate (see Figure 7.3a, no. 22) and to hinder access to the sea walls where there was no ditch, but could it also have been built to protect structures and/or activities here on the shore of the estuary itself? So far, neither the very numerous artefacts nor the few documents of the period allow a sufficiently exact sequencing of the structures in the early port yard, and it is possible that the shore wall and barbican came into being before the Ottoman hamam. If so, it is impossible to tell whether it was built in anticipation of hamam construction or to protect some other object or activity. Traces of ceramic kilns and the remains of a cobblestone road, both dating to the late fifteenth to sixteenth centuries, have been excavated, which indicate activity behind the new sea wall. In any event it is likely that a hamam was built in the port yard some time during Bayezid II’s sultanate. Writing in the mid-seventeenth century, Evliya Çelebi twice refers to this, as he calls it, ‘little hamam’, making it clear that it was ‘below, in the outwork [port yard]’ (.…ve asagı hisar becede bir hamamcığı var) near the estuary and named after Bayezid II.24 A coin from the period of this same sultan was found during excavation of the port yard hamam, which suggests such a dating of the structure. According to a vakıf register from 1565– 70 there were at least (p.149) two hamams at Akkerman, one in the port yard, ‘the Old Hamam in the outwork [port yard] of the fortress of Akkerman (hamam-ı atik der hisar bece-i kale-i Page 7 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project Akkerman)’,25 and ‘the New Hamam in the suburb of Akkerman (hamam-ı cedid der varoŞ-i Akkerman)’.26 By this time both hamams belonged to the vakıf endowed by Sultan Selim I (r. 1512–20).27 Both by design and by construction the port yard bathhouse is indeed a late fifteenth/early sixteenth-century Ottoman hamam. It had at least two domes and a total area of approximately 8 by 20 metres. It was located just under the north-western wall of the civil yard and to the right when exiting through the Water Gate (see Figures 7.3a and 7.5). Between 1999 and 2006 the Turkish and Ukrainian archaeologists revealed parts of its sub-structural walls (hypocaust) and other walls such as those with heating vents. Thanks to the excavations, the spatial distribution of the disrobing space (soyun-malık), cold space (frigidarium or soğukluk), warm space (tepidarium or ılıklık), hot space (caldarium or sıcaklık), furnace (praefurnia or külhan), and the water reservoir have become evident; revealed also is a complex hydraulic system of water pipes, drainage and sewers. The reason for the location of the hamam is not yet fully clear; it may have been because of water supply or drainage considerations or because the hamam was a component of the quarantine function of the port yard.28 As we will see below, initially there was no hamam either in the garrison yard or perhaps even the civil yard. On the basis of material finds we do know that by some time in the seventeenth century this hamam, which already in mid-century Evliya Çelebi called an old-fashioned, dirty and odd structure,29 changed function and became a workshop, where according to finds lead articles such as bullets were produced. What we have tentatively called ‘the barbican’, with its two almost parallel walls (see Figure 7.3a, no. 28 and Figure 7.4),30 plus an internal wall, a rear wall, and a round tower at its nose, is more of an enigma than the hamam. By the end of the twentieth century it was largely covered with earth, the tower having mostly collapsed since the (p.150) early decades of that century. Its function is not yet clear and it certainly seems to have evolved over time, as we will see below (pp.158–60). The barbican’s tower may have been an observation point or a lighthouse, and the entire structure may have defended the low shore walls by forward and flanking fire; alternatively, as a proper barbican, it may have stood in front of and defended an entryway into the port yard. For the first two centuries of Ottoman rule we have almost no unambiguous written sources on this structure. One difficulty is that the term hisar bece (lit., ‘fortress baby’) can refer to a barbican, an outwork, and, as our experience in the later documents shows, also apparently a half-wall, such as that on top of the inner wall of the ditch. In the case of the port yard, it is often difficult to detect whether the entire outwork/shore wall is being referred to or only this barbican-like structure extending towards the estuary; for at least the early period it seems that the former is the case. While documentary evidence is weak for this period, the excavations have shown that there was a gate on the south-western wall of the barbican connecting the estuary and the narrow space between the south-western and north-eastern walls (the Small Water Gate, near the no longer extant Tower 28 in Figure 7.3a), and another, larger gateway at the rear (south-east) of the barbican leading into the port yard via the cobblestone road. This would suggest that the structure was indeed a barbican in that it protected a larger gate leading into the yard beyond. In addition a doorway leading into the round tower, but lockable from the interior of the tower, has been revealed.31

Page 8 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project The story of the barbican does not end here. We shall see that in the eighteenth century, what is normally considered a typical medieval structure, which should have been obsolete at least by the end of the first two Ottoman centuries at Akkerman, not only continued to function in the third, but underwent a transformation at the end of the Ottoman period. This element was very likely unique, at least on this stretch of the Black Sea coast—a sign suggestive of improvisation in defending a unique location on this frontier.32 Documentary Evidence for Decision Making in Fortress Works The first true test of Akkerman as a defensive structure under the Ottomans came in the second half of the sixteenth and lasted well into the seventeenth century with the massive Cossack onslaughts on the Black Sea. In the heyday of their naval raids in swift and highly seaworthy longboats that descended the Dnieper and Don Rivers, the Ukrainian Zaporozhian and Russian Don Cossacks managed to ravage the shores of (p.151) Rumelia and Anatolia and even enter the Bosphorus, as well as cripple Ottoman commercial shipping while putting great pressure on the imperial fleet. Settlements on the northern seashore were doubly vulnerable as they were liable to attack also by land expeditions. Our record of raids is not complete, nor often of sufficient specificity—the sources speak of ‘attacks’, ‘burnings’, ‘destructions’, ‘sacks’ and ‘plunderings’. We know of raids at Akkerman in 1574, 1577, 1583, 1587, 1594, 1595, 1602, 1606, 1609, 1610, 1614, 1616, 1625 and 1634, so within these years at least one raid occurred every 3.5 years. Probably it is no coincidence that our first series of documents concerned with security of the fortress and repairs is from the 1570s. Almost all of them are concerned with strengthening the port yard; it would seem that this sector, with its relatively low walls, was the part most vulnerable to the Cossacks, though there is no explicit articulation of this in the sources. But we do know that while the Cossacks were good with artillery on land battles, in their naval incursions into the Black Sea they relied on surprise and stealth and lighter weapons in taking fortresses (the most famous examples are the sackings of Sinop and Kefe in 1614 and 1616).33 A low and narrow wall would have been vulnerable to scaling and surely had to be in a decent state of repair. Despite the scarcity and less than explicit nature of the documentation of the development of Akkerman fortress prior to the eighteenth century—that is, the virtual lack of details on actual construction, such as on specific types of materials, measurements and techniques—what we have so far does supply information on aspects of the organisational infrastructure that underpinned construction and maintenance, such as the importation of labour and construction material,34 funding,35 urban development,36 and aspects of planning and decision making. Here we give two specific examples of the latter. Our first example of planning and decision making from the early period pertains to mostly local participation in such a process. The first step of a construction project was to assess the work that needed to be done, in both its qualitative and quantitative aspects, the latter including not only materials and dimensions, but also costs. This process was referred to by the word keŞf (lit., ‘estimate, appraisal [of the condition or cost of something]’). Prior to the eighteenth century, occurrences with details of the (p.152) keşf processes are rare, and only occur towards the end of the seventeenth century, when a reform appears to have been made regarding the way construction projects were documented (though we can never be totally sure that previous records have simply not survived).37 Our first example is documented in a kadı’s certificate (hüccet). The previous destruction of the roofs of several towers by lightning strikes required that they be rebuilt, which was carried out in 1673 and 1674. Local experts in construction and roofing (ebniye ve sukuf ahvaline vukuf-i tamm olanlar), carpenters (neccarlar—both Muslims

Page 9 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project and non-Muslims are listed by name in the document), and provincial notables (ayan-ı vilayet), were summoned to a meeting of the kadı’s court and contracted to carry out the necessary work. There, upon their testimony, some of the costs for the required work were registered. In a related document the amounts and costs of the necessary wood are recorded.38 In addition to local authorities such as the kadı, the centre was both directly and indirectly involved. Present at the procedure in the kadı’s court was a court official (müteferrika-ı dergah-ı ali) appointed by the Porte as the ‘executor’ (mübasir) of this repair; the hüccet that is the source on this case ended up in Istanbul as it was probably drawn up to inform and certify that local forces and funds were contracted for the work at hand. Our second example for the early period pertains to decision making at the centre as opposed to on-site. It does not relate to construction works per se, but rather to the defence of the port yard during the years of the first known Cossack attacks on Akkerman. Because it is a good example of the prerogative of the centre in running this fortress, we present it in some detail. An order from the Porte dated June 1576 summarises a request for permission by the kadı and fortress warden (dizdar) of Akkerman that the gate leading into the port yard (hisar bece) from the garrison yard (the Water Gate, Figure 7.3a, no. 22) be opened by the time of the voluntary prayer between midnight and dawn (temcid), and certainly by the time of the first, daybreak prayer (sabah namazı). This was necessary, so the petitioners asserted, because there was no bathhouse inside the fortress proper where the congregation could bathe (istihmam) prior to prayer (presumably perform the ritual ablutions?), and the only place for this was the Sultan Selim hamam (i.e., probably the original Bayezid hamam described by Evliya Çelebi, now under the vakıf of Selim I, see above).39 In the dispositio of the document, giving the order to be implemented, it is somewhat equivocally decided that indeed the gate into the port yard may be opened before daybreak during the season of the temcid (the months of Receb, Saban and Ramadan) and in the winter as long as the security of the castle is not compromised. Aside from conveying a clear sense of the danger (p.153) prevalent during this period, and of the vulnerability of the port yard which prompted such caution, this document is interesting for suggesting a high degree of centralisation in decision making. Did a change in the status of a gate in a distant frontier fortress require clearance at the centre, or was this a case of a mere conflict of the garrison and inhabitants with the dizdar?40

Akkerman in the Eighteenth and Nineteenth Centuries For the late period (eighteenth to beginning of nineteenth century) of Ottoman presence at Akkerman we currently have close to ten times as many Ottoman documents as for the first two centuries. From what we have observed on-site and on the basis of the surviving documentary record, as well as references in narrative and other, non-Ottoman, sources, the eighteenth century was not only a period of much more intense repair and maintenance activity, but also a period when projects of redesign and reconstruction were undertaken. Presently we still have only a rudimentary sense of the precise building periods. As can be expected, building activity tended to occur on the eve of, during, or in the aftermath of the wars or serious diplomatic conflicts between the Ottoman and Russian Empires: 1710–11, 1735–9, 1768–74, 1783 (Russian annexation of the Crimea), 1787–92 and 1806–12. The documentation relating to this period’s more significant repair and reconstruction activity has clustered most heavily within the following years: 1707–9, 1737, 1756–7 (no obvious war connection), 1775–8, 1783–4, 1788–9, 1792–8, 1803–6.41 The relatively great abundance of eighteenth- and early nineteenth-century Ottoman archival material is not only due to the greater preservation of documents from more recent centuries. It seems that by the eighteenth century the Ottoman bureaucracy, and in particular the financial administration (maliye) that generated or acted as repository of most of Page 10 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project the construction-related documentation, began to keep more detailed inventories of the items for which costs were incurred. So it recorded explicit descriptions and discussions of projects, rather than only keeping accounts of costs of building materials, transport and labour without specific details. Often this data was recorded in special registers (defters) with marginal notes explaining the figures and items registered and in addition, with copies of reports (arz/arzadaş¸t) or memoranda (kaime) sent by officials and experts in the provinces where the (p.154) construction was carried out, decisions and orders (ferman/hükm) issued at the centre, and various other types of documentation (hüccet, tezkere). As these defters frequently contain estimates and appraisals, they are often referred to as keşf or tahmin defteri (‘appraisal/estimate register’) or, for example, keşf ve tamirat defteri (‘appraisal and repair register’). The typology of these registers has yet to be worked out, but for now we simply refer to them as ‘building registers’ (for an example see Figure 7.6).42 We have so far located twelve such building registers for Akkerman, covering 1709–10, 1737, 1756–7, 1775–8, 1784, 1794–6 and 1800. The majority of our Ottoman documentation from this period consists of unbound, usually singlesheet documents, indeed, the component parts of the building registers—reports, orders, inventories and other types of documents. Often these have been preserved together, in effect as files devoted to a given construction project, and thus practically amount to building registers themselves—after all, the building registers themselves are largely ‘scrap books’ consisting of copies of various documents. In any event, eighteenth-century Ottoman building documentation is both more numerous and more informative than that for earlier periods, for it allows glimpses if not views of the various stages of a building project, from initial discussion and planning to financing and execution. What does the extensive detail in the Ottoman building registers and separate documents concerning eighteenth-century construction and repair work on the fortress tell us about the Akkerman segment of the Black Sea frontier? Practically every pertinent passage in our documents helps us to reconstruct our frontier region with more or less significant facts— certainly they have more data than we could hope to present here. But what does information on construction and repair, for example, the anatomy of the walls, towers and ditch of a fortress, construction techniques, building materials, and problems and their solution tell us about our frontier qua frontier? Given our current knowledge of Ottoman fortifications, probably very little —how did construction at Akkerman differ from construction in other frontier regions, such as at Özi or Azak, or further afield, at Fethülislam (Kladovo) opposite the Hapsburg frontier, or on the Iranian frontier at Kars or Basra? What about on an internal frontier, such as at Niğbolu (Nikopol) or Rusçuk (Ruse) vis-à-vis the Principality of Wallachia? Easier to answer would be how work at Akkerman would compare with upkeep of non-frontier fortresses, such as Kastamonu or Ankara, fortresses which no longer played vital roles in defence against serious foes and were left behind in time. Perhaps many basic building techniques were the same everywhere and changed little over time, though without evidence, be it written, drawn or material, we cannot say much about techniques employed and materials used at non-frontier fortresses. But one can deduce and even observe at surviving sites that fortresses away from the frontier that did not have to face modern armies developed little and kept their largely medieval appearance, and thus repair techniques and materials may have continued with little or no change. This cannot be said of eighteenth-century frontier fortresses influenced by new challenges and consequently new techniques, often emanating from Europe. (p.155)

Page 11 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project (p.156) The Ottoman Response to Advances in Fortification Technology By the end of the seventeenth century the Ottomans had begun to construct fortresses with thick earthen ramparts and solid Figure 7.6. Sample pages from a building bastions capable of withstanding cannon register including works carried out at Akkerman in 1709 (details on procurement fire, as well as, by means of forward and and quantities of materials, such as wood, flanking fire from the bastions and with nails, lime, for repairs of tower roofs and handgun fire from parapets and covered other portions, as well as on labour imported ways, protecting such fortresses from from other regions). BOA, MAD 3882, 200–1. mining and other forms of close attack. This relates to the original sixteenth-century trace italienne, which progressively developed into complex systems combining elements such as the bastion, ravelin, tenaille, ditch and glacis, pioneered most famously by the Frenchman Vauban (1633–1707). Like the Hapsburgs, because of the great expense, the Ottomans were selective in the fortresses they chose for such modernisation and only the most strategic fortresses were rebuilt in the new style, and even then usually not in pure ‘Vaubanian’ form, but only applying elements considered indispensable.43 For example, at the fortress complex of ÖziKılburun which straddled the mouth of the Dnieper, at the end of the first decade of the eighteenth century, the first plan was to cut down the high towers of the medieval-style Özi fortresses and fill them with earth so as to turn them into bastions of the sort on which heavier artillery could be mounted. Instead, outside of the old castle a new ring of ramparts with still rather rudimentary bastions and a ditch and glacis were constructed. The old castle inside this ring remained until at least the 1730s (it is depicted in a plan from 1737): it may have been left to serve as a possible last line of defence. However, we do know that the towers were used to store ordnance and perhaps other supplies.44 There was much improvisation in Özi’s ‘Vaubanisation’, probably for reasons of cost and, more significantly, because of tactical needs. Rather than imitate without reflection, the Ottomans seem to have been applying western technology in a practical and intelligent manner.45 While Özi was of great strategic significance, as it guarded entry to the Black Sea against the Cossacks of Ukraine in the sixteenth and seventeenth centuries and against the forces of the Russian Empire in the eighteenth century, Akkerman lost much of its importance once Moldavia was more or less subdued in the sixteenth century and still more when the Cossack sea raids waned by the second half of the seventeenth century. When the Russian Empire started threatening and making incursions into Ottoman possessions north of the Black Sea and into the Danubian region, Akkerman was (p.157) usually ignored. Bender, situated at an easier crossing of the Dniester upriver from Akkerman, was more important for defending the approaches to Rumeli. Consequently, in the first half of the eighteenth century we see a less determined effort to modernise the fortress at Akkerman than at Özi or Bender. There was another factor that shaped the development of Akkerman fortress. Unlike Özi, where sandiness of the soil led its original builders to forgo digging a moat or ditch,46 in Akkerman, as already stated, a deep and wide ditch (approximately 20 X 14 metres) had been dug along the landward sides of the fortress even prior to the Ottoman takeover. This impressive and formidable obstacle combined with the wide walls meant that when it came time to consider modernising it in the eighteenth century, it would have made little sense to fill and re-dig the ditch further out and dismantle or rebuild the walls to accommodate a proper modern trace with bastions and outworks such as ravelins and redoubts in place of or beyond the old trace. To do

Page 12 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project so would have cost more in terms of both manpower and money than the strategic importance of the fortress merited. In other words, having such a formidable edifice, the Ottomans chose to save their resources, and, in order to face modern artillery, to strengthen rather than radically rebuild it. So they kept the ditch and instead strengthened the walls and towers by several methods. The exact chronology of their efforts is still uncertain, but it is clear that they piled up earth behind the old walls thereby reinforcing them and providing platforms for cannons. They made sure that most of the originally hollow towers were sufficiently filled with earth to withstand cannon fire. In at least three places, where the trace made an angular turn—between the Great Gate and Tower 7 (for the numbered towers and other features see Figure 7.3a), at the Storeyed Tower (Katlı Kule, Tower 11) and at the White Tower (Ak Kule, Tower 17)—they created what can be called ‘internal bastions’ by packing earth into the interior side of these corners and probably reinforcing them with timber and other fill. Each of these bastions included a stone casemate. Thereby the Agha Bastion (Ağa Tabyası, linking Towers 6 (= Great Gate) and 8), the Storeyed Bastion (Katlı Tabya, linking Towers 10 and 12), and the White Tower Bastion (Ak Kule Tabyası, linking Towers 16 and 18) were created.47 Another bastion, the Pasha Bastion (PaŞa Tabyası, between the north-west tower of the citadel and point no. 1 in Figure 7.3a, tentatively identified as the Millet Tower), which faced the estuary to the north, was formed by building up earth behind the wall and placing a row of large cannon embrasures on top of the original stone wall. Though called a bastion (tabya), this structure was more of a wide gun platform.48 The earth that had been piled up to thicken the straight stretches of wall (p.158) and create the internal bastions and sea-side gun platform was bulldozed away some time after 1955.49 The term tabya may be rendered as ‘bastion’ (the meaning of the original Arabic in non-Ottoman contexts includes pisé (rammed earth), tower, and bastion), but if the rendering as bastion is meant only to refer to standard triangular redoubts or pentagonal bastions, it fails to describe the range of usages in the Ottoman context. While tabya may indeed denote such bastions as can be seen in Ottoman translations of Vauban’s treatises,50 this term was also applied in the documents to the improvised internal earthen structures built to strengthen the fortress as well as to the gun platform at the PaŞa Tabyası. On the other hand, we have not encountered this term denoting a simple straight earth-reinforced wall. This suggests that in modifying the fortress at this time, the Vaubanian concept of bastion was probably operative as an archetype— otherwise the simple earth-reinforced walls would probably also be referred to as tabya. Before eighteenth-century fortification architecture with its key element, the bastion, was incorporated into Ottoman fortress construction, the term tabya seems rarely to have been used, which further suggests that tabya denoted modern bastions and bastion-like structures. After the 1789–92 war against the Russian Empire, the Ottomans brought in the Alsatian engineer François Kauffer to propose a radical redesign of Akkerman fortress, among a number of other fortresses in the region. In the Ottoman archives we have only Ottoman translations of Kauffer’s proposals; the plans he drew and refers to have been lost or misplaced. However, perhaps because he was at the same time allegedly an informant for the Russians, he also supplied copies of his plans to them and today they are preserved in Moscow archives. One such, dating from 1793, shows three possible new traces involving triangular redoubts and pentagonal (external) bastions (Figure 7.7), all extending beyond the old main ditch, with no new ditch, only a glacis. On the basis of a post-conquest plan made for the Russians in 1807 by E. H. Förster, an engineer and major-general in the tsar’s army, by then only one set of four triangular redoubts (two large and two small ones) either had been made or was being planned (Figure 7.8). With the help of geophysical survey and further scrutiny of the documents and plans we hope to be able to determine as precisely as possible what sort of trace or traces were actually installed Page 13 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project outside the ditch. Certainly a glacis was built up. Its remnants can still be observed today outside the south wall of the civil yard. The curious barbican in the port yard is another matter, for the eighteenth-century Ottoman documents retrieved to date offer little information on this structure and only archaeology combined with visual documentation allows us to note its development in the eighteenth century. During the 2007 season, excavation revealed the original floor of the structure in several places, including the three entryways already mentioned above—between tower and walled corridor, into the fortress from the estuary (on the south-west wall), and into the port yard in the rear (on the south-east wall). It should be noted that by the nineteenth century this entire structure was of two storeys (we cannot be certain if this was so previously). In addition, a fireplace and a heatingsystem with ventilation passing inside the walls, discovered in 2006, have been further investigated, as well as gun embrasures that were cut into the side walls. A Russian coin from 1754 found in situ near the top of one of the surviving wall remnants shows the approximate date of a re-planning and rebuilding of the barbican. A shorter internal wall, parallel to the two main walls near the back, has also been discovered. It reveals more of the complexity of this structure, though we have not yet been able to suggest the exact function of this last feature. Most interesting is that some time between 1793, the date of one of Kauffer’s plans, and 1807, just after the Russian conquest, the side gate was blocked and in fact there appears to have been no entrance into or out of the shore of the estuary via the barbican.51 As another Russian plan dated 1807 indicates, there were four gun ports on the side walls providing fire forward of and along the shore wall in both directions.52 Probably at the same time that these ports were installed, a new opening was broken through the shore wall (or had been opened by enemy cannon fire?) and a gate was put there, today filled in but clearly visible. Guarding the wall, and the new gate, the barbican ceased to function as barbican, that is, a structure blocking access to a main gate, and instead, functioned more as a kind of bastion or redoubt with a round tower (the latter may have continued as an observation post or a lighthouse). As we cannot be sure whether this last alteration to the barbican occurred before or after the conquest, we can only say that the improvisational nature of this structure continued to the very end of the Ottoman period or even beyond. (p.159)

Figure 7.7. A plan by Kauffer (1793). Russian State Military-Historical Archive (RGVIA), fond 846, inventory 16, file 21610.

Figure 7.8. Förster plan (1807). Şlapac, Belgorod-Dnestrovs’kaia krepost’, 73 (Figure 32). Reproduced with permission.

Page 14 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project (p.160) The Infrastructure of Maintenance and Construction For the eighteenth century we are in a much better position to gauge how decisions on maintenance and the building of new Figure 7.9. Detail of a plan by Kauffer structures were influenced by perceived (1793). RGVIA, fond 846, inventory 16, file 21608. threats and priorities arising in the frontier context. The documents are revealing of frontier realities relating to organisational infrastructure, giving access to a wide spectrum of data concerning information gathering, planning, decision making, and resource allocation. Here we can only give brief characterisations of these organisational aspects along with a few examples. Let us start with those involving the command and control or the (p.161) management of works at a fortress such as Akkerman. At the top were the chiefs of state acting in the name of the sultan, such as the grand and other viziers and the chief of the imperial finance administration (baş defterdar), while at the bottom, the local sancakbeyi and the kadı (along with the naibs, his assistants, or substitute judges). Perhaps surprisingly, the kadı (and often the naib) is encountered much more frequently in the construction-related documentation than the sancakbeyi. His role may appear greater than it really was, thanks to his function in registering documents, such as orders (hükm/ferman) that arrived from the centre or in issuing certificates (hüccet). However, he was also active in local affairs beyond recording or issuing state documents and other functions such as adjudication or registration of marriages and inheritance transactions. We have already seen the case of personnel, costs and materials for a major repair of tower roofs after lightning strikes in the early 1670s being authorised in the Akkerman kadı’s court. Finally, aside from their duties in matters of taxation, both inside and outside the court (for instance, as inspectors, müfettis, of tax farming operations), they often acted in matters concerning the financing of fortress garrisons and infrastructure. A key role in works on fortification was commonly played by the bina emini, or chief of construction (lit. ‘construction trustee’), who was positioned between officials in the centre and those in the provinces. This was an ad hoc appointee, often one of the royal architects (mimar-ı hassa). The best-documented functions of the bina emini are connected with the informationgathering and planning activity carried out before the commencement of work. The inspection of the site and appraisal of work to be done, including an estimate of the costs and necessary materials—which often resulted in a keşf defteri—would typically be carried out under his supervision. When a keşf was drawn up, it served as the basic document for the project in question—other orders and memos would refer to what was laid out in it.53 On occasion we see other officials at Akkerman entrusted with supervision of the initial on-site planning and of the drawing up of the keşf defteri—sancakbeyis, kadıs, even naibs.54 For the researcher, the apparent lack of consistency in the choices of profession for such a key role can be somewhat baffling. However, such practices may simply be a manifestation of Ottoman adaptability to situations when optimal personnel were unavailable, as well as something akin to the modern business practice of appointing non-specialist managers. (p.162) Unfortunately in such cases the documents refrain from referring to specialist assistants. What was most important seems to have been to record who was responsible for a given process.55 Aside from the keşf defteri, simple reports from Akkerman to the centre, which were often the impetus for the initiation of actual construction work, contain information on the

Page 15 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project site that was used in the planning process. Indeed, as already stated, a significant portion of the building registers consist of copies of reports, orders (which usually reiterated the reports to which they were responding), and other documents, in addition to the quantitative data concerning measurements, personnel, materials and costs. The fact that they were often if not usually drawn up in the centre, and certainly preserved there, is testimony to the key role played by officials at the centre. While Akkerman was a nexus for trade and the surrounding steppe supported herding and some agriculture, all of which generated local wealth and consequently tax revenues that could be designated for maintenance and construction of the fortress, as was the case at other fortified points on the northern Black Sea coast, there were insufficient non-fiscal resources (and often fiscal resources as well) for such works. In practically every construction project at Akkerman for which the documentation is extensive enough to cover labour and materials, there is an indication that much had to be imported from neighbouring regions. Both skilled (e.g. carpenters, neccar) and unskilled (ırgad) workers were brought, usually by ship or boat, though sometimes by wagon as well, from near and distant places on the Rumelian Black Sea coast, on the Danube, in Bulgarian hinterlands, in Moldavia and Wallachia, and even from as far away as Albania.56 As to materials, while brick, stone and lime are mentioned, the main imported material was timber—logs and lumber cut to various specifications. Wood almost always came from Moldavia.57 When funds levied from local tax revenues to cover the materials, labour and transport costs did not suffice, they would be transferred from those revenues levied in neighbouring areas (the havale system). Typically these were customs revenues or, when from Moldavia or Wallachia, cizye.58 The active participation of the centre in organising, approving and enforcing the transfer of human, material and financial resources to construction work at Akkerman is always evident, and its control over information on these resources is impressive. At times the building register is almost a kind of spreadsheet listing what large and small items and quantities—be they wood and lime, instruments, means of transport—can be found in Rumelian provinces or along the Black Sea frontier. Major and even many (p.163) seemingly minor decisions needed to be made or at least approved in the capital. Moreover, the degree of micromanagement of fortress affairs over such long distances is at times striking. An example is the case of replacing the locks of the gates at Akkerman. We have a petition from Akkerman and an order to its kadı and dizdar dated to 1766 in which there is a rather detailed discussion of where to come by ten locks that were needed to replace old and worn out locks in the fortress for the Great, Middle, Water, Postern and other gates. The locks are described as great and mediumsized, fortress locks (kale kilidi) and hanging locks (asma kilidi). What was available in the armoury (cebehane) at Akkerman turned out to be insufficient and in the end an order was issued that if the old locks cannot be repaired, the required number must be released from the imperial armoury for transfer to Akkerman. In the process a slew of officials, some of rather high status (such as the baŞ defterdar), were involved in the decision-making and execution process.59 Perhaps fortress locks were a matter of high security and not trifles, and such centralised micromanagement was not unjustified?

Archaeological Evidence for Frontier Life at Akkerman So far we have sought to show how historical and historico-archaeological study of the development of the fortress at Akkerman can inform us regarding the Ottoman frontier at Akkerman. However, the archaeology has produced a large volume of material not directly related to the fabric of the fortress, but rather to life in and around it that has implications for

Page 16 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project our understanding of the frontier in ways that our written sources so far do not. By the end of the 2007 excavation season more than 17,000 artefacts had been unearthed in the port yard, consisting of different groups of ceramics (pottery and pipes), metal, glass and others (for percentages see Table 7.1). These artefacts already include both imported and locally made goods.60 All types of ceramics, which make up 85.3 per cent of the finds, include relatively high frequencies for some prestige and other items (for percentages see Table 7.2). In the thirteenth and fourteenth centuries sgraffito pottery was the most commonly used glazed ware in the Ottoman Balkan and Mediterranean regions. Sgraffito ware was found on numerous sites in Özi (both on the territory of the old fortress and in the town), as well as not far from Özi at the settlement of Dniprovske-2.61 Finds of this type also in Akkerman (Figure 7.10) strongly suggest that the northern Black Sea region also participated in the development of this ware. Though sgraffito could be found as a quotidian ceramic in the Balkan–Mediterranean-Black Sea area, it primarily served as a prestige pottery until the appearance of Iznik quartz-frit pottery at the end of the fifteenth century62 Interesting are examples of local production found in the Crimea and other northern Black Sea areas, including Akkerman, from the end of the thirteenth and the fourteenth century—furnaces with fragments of sgraffito and unfinished examples were found in the Golden Horde layer at Akkerman from the beginning of the fourteenth century63 Our investigation of the port yard revealed unfinished sgraffito ware produced near the estuary, which may date from the Golden Horde or from the pre-Ottoman or early Ottoman period. Besides the sgraffito style, another type of glazed ceramic, known as Miletus ware, was imported from Anatolia. (p.164) Table 7.1. Percentages of finds from the port yard at Akkerman fortress. Items

%

glazed ware

35.0

unglazed and partly glazed ware

30.2

quartz-frit pottery

8.8

smoking pipes

8.0

metal artefacts

7.3

glass

5.6

sgraffito ware

1.5

Miletus ware

1.0

chinaware

0.8

others

1.8

Table 7.2. Percentages of ceramic wares from the port yard at Akkerman fortress. Items

%

unglazed ware

39.0

monochrome glazed ware (green, yellow, brown)

38.4

Page 17 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project Items

%

Kütahya quartz-frit pottery

8.0

polychrome glazed ware (marble)

6.7

Iznik quartz-frit pottery

3.4

sgraffito ware

2.0

Miletus ware

1.3

chinaware

1.1

(p.165) Though most groups of this ware appeared in the Mediterranean in the thirteenth and fourteenth centuries, imitations were also produced in the northern Black Sea region. Here we might note that other, earlier, Golden Horde and Moldavian period finds—considerable amounts of metal (e.g. belt plates and buckles), glass and other objects, with decorative elements such as ‘three lives’, ‘tree of life’, and other typical Turkic motifs—also show a large influence from Anatolia, as well from Central Asia.64 The pre-Ottoman Anatolian and Mediterranean ceramic wares found at Akkerman and in the northern Black Sea in general reveal long-term contacts between these regions. That is to say, during the Golden Horde period there was strong cultural influence not only from the east, but also from the Anatolian and Mediterranean south. The coming of the Ottomans to the northern Black Sea did not bring an alien material culture— earlier penetration served to pave the way for their cultural presence on the coast and even further north in ethnic Ukrainian territories. When Iznik pottery became a prestige ware in Anatolia and other regions from the end of the fifteenth to the beginning of the seventeenth century, so too did it occupy a place of prestige on the northern frontier. Kütahya ware, produced from the end of the seventeenth century until the end of the eighteenth century, also acquired a significant presence in the north. Excavations at Özi and especially at Akkerman suggest that there must have been significant importation of these wares.65 All phases of Iznik ware are represented at Akkerman and Özi, and in the former, Iznik ceramics are represented by more numerous finds than in the latter, especially c. 1480 to 1530 (Figure 7.11). In her seminal study of medieval and post-medieval ceramics in the Aegean, Joanita Vroom writes that groups such as Iznik ware, polychrome marbled ware and Kütahya ware are rare in late fifteenth- to eighteenth-century sites in Boeotia, where the more prestigious wares from Iznik and Kütahya have a frequency of 0.2–0.3 per cent (Figure 7.12).66 By contrast in the port yard at Akkerman the percentage of Iznik and Kütahya ceramics is 3.5– 4.0 per cent! At the same time that Iznik and Kütahya wares declined as prestigious wares came a challenge by Chinese porcelain in the eighteenth century. (p.166)

Page 18 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project

Figure 7.10. Sgrafitto shards excavated in the port yard. Photograph by Valentyna Kesar.

Figure 7.11. Iznik shards excavated in the port yard. Photograph by Valentyna Kesar.

Figure 7.12. Kütahya coffee cup excavated in the port yard. Photograph by Valentyna Kesar.

Page 19 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project (p.167) This occurred not only in the metropolises of the Ottoman Empire, but also on the frontier. In such regions, including Akkerman, Özi and other Ottoman northern centres, dishes for tea and coffee Figure 7.13. China shards excavated in the port yard. Photograph by Valentyna Kesar. with ornamental figures have been found (Figure 7.13). Of fifteen groups of ceramics from the late Byzantine/Frankish period (thirteenth to mid-fifteenth century) established by Vroom for Boeotia, we have found nine, or perhaps ten groups in Akkerman; of sixteen recognised groups of ceramics (including pipes) from the Ottoman–Venetian period (late fifteenth to eighteenth century), we have found fifteen groups in Akkerman.67 The port yard at Akkerman, like Özi, has yielded a very large collection of mostly Ottoman (and also some Ukrainian Cossack) tobacco pipes from the end of the seventeenth and the eighteenth century—so far we have more than one thousand pipes in total (Figure 7.14). By comparison, while the pipes from Vroom’s Boeotian sites reach a 2 per cent frequency, at Akkerman we have up to 8 per cent. While we are not yet sure what proportion of the pipes was produced locally, clearly the vast majority were made of Anatolian red clay and imported to Akkerman. The large number found at the Akkerman and Özi fortresses suggests that they were popular with Ottoman garrison troops, or were perhaps brought there for sale or as personal items. By virtue of their popularity in Ottoman territories on the northern Black Sea, they spread throughout Ukraine.68 During the height of the Cossack era—the sixteenth to eighteenth centuries—pipe smoking, like other facets of Tatar and Ottoman material culture, became widespread among these frontier warriors. Ottoman and Ottoman-style pipe finds on the northern seaboard and throughout Ukraine promise to provide a better qualitative and quantitative notion of the degree of Turkish cultural penetration in our region. The development and spread of tobacco pipe production in Ukraine, with its clear Turkish stylistic influences, is a brilliant example of the integration of Turkish and Ukrainian features in this kind of material culture. As we can see from Table 7.1 (percentages of finds, which are mostly from the Ottoman period), non-ceramic items such as metal, glass and others (including stone) represent 14.7 per cent of the various artefacts. Metal items, which belong to various periods from the Golden Horde through the Moldavian and Ottoman, include buckles, belt plates, adornments of bronze and silver, and coins. In the course of excavation of the port yard nearly 300 items of silver and bronze were found, and one of gold. Of the coins, 85 per cent are Ottoman. The rest are mainly ancient, Golden Horde, Crimean Tatar, Moldavian, Polish and Russian. This suggests the degree to which Akkerman was connected to the Ottoman economy. (p.168) At Akkerman, as elsewhere, archaeological artefacts provide evidence for longue durée processes, including cultural interactions that go against some of the long-held stereotypes of necessary and constant conflict, and suggest a more balanced view of cultural contacts and exchange. On the level of material culture, the Ottoman

Figure 7.14. Pipes excavated in the port yard. Photograph by Svitlana Bilyayeva.

Page 20 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project presence on the frontier at Akkerman and influence beyond it were surely very great. In the case of ceramics (including pipes), it is possible and perhaps likely that Akkerman was part of an extensive web of commercial and cultural exchange. Given that at Akkerman there is such a wealth of ceramic material, both in terms of numbers of periods represented and frequencies of some types (the latter apparently being higher than in some hinterland regions of the empire), the question arises of why this is so. And why are so-called prestige items so common? At this point it is too early to say for certain. Such wealth may parallel the wealth of Akkerman as an entrepôt of trade. The high frequency of prestige items in the barbican may suggest that a considerable portion of them were possessed by members of the garrison, which has implications regarding their daily life. Of course, the high frequency of items may also be an indication of their more common use and lesser prestige. In any event, the implications are very intriguing and call for further investigation at this site as well as at other Black Sea sites. Another question that arises is, if prestige ceramics in particular were of such apparent commercial importance, should they not have been recorded in the Ottoman customs registers or elsewhere? A large and rare early sixteenth-century customs register that includes Akkerman and Kili has so far yielded no precise references to (p.169) ceramics.69 None are registered in the Kefe register of customs arrears from 1487 to 1490.70 The large disproportion between the huge presence of ceramics and their rarity in the Ottoman documents could be due to a lack of sufficient representative samples of the latter. Alternatively, the discrepancy between the archaeological and documentary record may be due to the ways in which these items were disseminated and used and to attitudes towards them. For example, the prestige items may have been brought as personal possessions rather than imported as commercial ware (in other words, while they were items of substantial value, technically speaking they were not always commercial ware), while simpler pottery that served as containers for foodstuffs might have escaped or simply not qualified for recording in the customs registers.71

Conclusion In a frontier region such as the one on the northern Black Sea, where environment, human geography and historical traditions made the steppe an alien place that did not readily yield to control and assimilation, the fortress was indispensable for maintaining the centre’s presence. As imperial presence in such an area was anchored at and emanated from the fortress, be it as base for urban life, commercial activity or military prowess, the fortress can be seen as a prime target of a strategy aimed at learning about this frontier of the Ottoman world. Analysis of the development of the Akkerman fortress is helpful for understanding some of the challenges in occupying and protecting the area; the type and extent of construction activity can be a gauge of the real or perceived threats from within the steppe and from beyond it. The Ottoman documents on this fortress inform us better than any previous written sources on its development, including specifics on the types of work done in repairing or reconstructing it. This type of information will have greater meaning and significance once we learn more about analogous activities at other fortresses on the same frontier and at those on other Ottoman frontiers. For the time being, the data on the organisation and infrastructure of work at Akkerman, organisation and infrastructure that connects both the centre and lateral regions with Akkerman, reveals many details about the structure and workings of the Ottoman Black Sea frontier in general, and the Akkerman portion of it in (p.170) particular. Though more of a sampling than survey could be provided here, we now have a view of the command and control activities of officials in the centre and at the frontier, the relation to this fortress of human,

Page 21 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project material and financial resources of Moldavia in particular, but also of other areas both on this frontier and in the Rumelian hinterland. So far, at this early stage of the Akkerman Fortress Project, the documents and architectural observation of the standing parts of the fortress have contributed much towards furthering an understanding of its fabric and features, while the archaeology has mostly complemented the historical and architectural analysis. When dealing with a frontier region such as ours, thanks to frequent instability and disruptions, any source, be it written or material, is a welcome prize. Ironically, but probably because of the vagaries of survival and preservation, there is so far little in the Ottoman archives on two outstanding features of the port yard, the haman and the barbican—the first of which is completely Ottoman, and the second of which may have originally been Moldavian, but was certainly several times redesigned and rebuilt by the Ottomans. Without the complementary contribution of the archaeology, these two edifices, with both their traditional and original aspects, would hardly figure in our understanding of this frontier fortress. As the project continues, it is hoped that other areas of Akkerman fortress will open themselves up to what is in some sense an even greater prize, that is, a supplementary one: spots where documents and artefacts overlap more fully, thereby allowing not only a more multidimensional view of a feature and its development, but also providing a chance to better assess the nature and value of the documents and artefacts working in combination. Finding overlap depends on the vagaries of preservation and permission for access, but not only these. The difficulty of overlap is also a function of the vastness of the historical process—even in a relatively compact and discrete site such as a fortress, human activity is so varied that there may be spheres that no source base covers, or when different source types do, overlap is a less frequent case. Realistically, asymmetry and lack of overlap in the contribution of written and material sources will remain the norm in such projects. While the value of the written sources, particularly the Ottoman documents, for gaining a better view of the development of Akkerman has been relatively easy to demonstrate, historians, who typically see the written sources as primary, can only be humbled by the disproportionate contribution of the artefacts on trade and commerce, not to mention daily life, in this frontier site. Note. We would like to thank Caroline Finkel for her comments and in particular for her substantial work with the Ottoman documentary basis of this study. Support for the research for this study was granted by the British Institute at Ankara and the Max van Berchem Foundation. Notes: Proceedings of the British Academy 156, 137–170. © The British Academy 2009. (1) This aperçu of the Ottoman Black Sea frontier is primarily based on Victor Ostapchuk, ‘The human landscape of the Ottoman Black Sea in the face of the Cossack naval raids’, Oriente Moderno, n.s. 20 (2001) (Special Issue, The Ottomans and the Sea, ed. Kate Fleet), 23–95. (2) Though, as we have demonstrated elsewhere, during Cossack ascendancy on the sea in the first half of the seventeenth century it was not quite a (secure and prospering) Ottoman lake. See Ostapchuk, ‘Human landscape’, esp. 37–43. (3) Mihnea Berindei and Gilles Veinstein, ‘Les possessions ottomanes entre Bas-Danube et BasDniepr: réglements fiscaux et fiscalité de la province de Bender-Aqkerman, 1570’, Cahiers du Monde russe et soviétique, 22 (1981), 251–328, esp. 252.

Page 22 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project (4) See www.akkerman.org. In 1999 archaeologists from Turkey and Ukraine began excavations at Akkerman, and by 2006 were joined by historians and archaeologists from Canada, the United Kingdom and the United States. (5) It was known by the equivalent of ‘White City/Fortress’, in various languages in addition to Turkish—a name which dates back to foundation by the late fourteenth/early fifteenth century (Belgorod, Bialogrod, Bilhorod in Slavic languages and Çetatea Alba in Romanian). (6) The results of the project will be published in detail after its conclusion in 2009–10. (7) Halil İnalcık, ‘The Ottoman state: economy and society, 1300–1600’, in Halil Inalcık with Donald Quataert (eds.), An Economic and Social History of the Ottoman Empire, 1300–1914 (Cambridge: Cambridge University Press, 1994), 285–95. (8) ibid., 169–71; Gilles Veinstein, ‘Les “çiftlik” de colonisation dans les steppes du nord de la Mer noire au XVIe siècle’, Istanbul Universitesi Iktisat Fakultesi Mecmuası, 41 (1982–3), 177– 210. (9) For the pre-Ottoman history of Akkerman fortress see Mariana Şlapac, BelgorodDnestrovs’kaia krepost’: Issledovanie srednevekovogo oboronnogo zodchestva (Chişnău: Editura ARC, 2001), 51–61. (10) Excavated by the Turkish team headed by Bozkurt Ersoy along with the Ukrainian team headed by Svitlana Bilyayeva. (11) See James R. Mathieu, ‘An architectural assessment of the Fortress of Akkerman’ (2006–7), www.akkermanfortress.org. (12) Şlapac, Belgorod-Dnestrovs’kaia krepost’. (13) Cf. the statement in a building register from 1737: ‘since the time of the conquest effort was not expended to repair the fortress as necessary’, BOA, Maliyeden Müdevver Defterleri (MAD) 3609, p. 227. (14) Nor, contrary to what Mariana Şlapac says, is there any indication in Evliya Çelebi’s description of Akkerman that his patron and kinsman Melek Ahmed Pasha, refortified the complex in 1657–8. Rather, he reports that Melek Ahmed Pasha ordered the ditch to be cleaned out, although this task may have been substantial, for apparently a huge amount of dirt and refuse—‘500,000 cartloads’—was removed and dumped into the estuary. Further, unspecified repairs were to be undertaken at various fortresses in the area including Akkerman. Earlier he makes it clear that a huge ditch (‘70 [sic] fathoms [kulaç] deep’) was there when Bayezid II arrived to conquer the fortress (Evliya Çelebi, Seyahatname, 5: 60, 63, 90). (15) BOA, Mühimme Defteri (MD) 81, p. 184, no. 407 (Safar 1017/17 May-14 June 1608). (16) If so, here Evliya’s varos may indeed be the civil yard, though the term could of course also be applied to quarters outside all the walls. (17) Evliya Çelebi does not specify whether the mosque and medrese are of a vaklf endowed by Selim I (r. 1512–20) or Selim II (r. 1566–74), but note that Selim I stopped at Akkerman on his way from the Crimea to seize the Ottoman throne from Bayezid II in 1511.

Page 23 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project (18) Evliya Çelebi, Seyahatname, 5: 63. (19) Cf. ibid.; BOA, Tapu Tahrir (TT) 483. (20) İ lhan Şahin, ‘XVI. Yüzyılda Akkerman’ın Demografik ve Sosyal Durumu’, Güneydo ğ u Avrupa Araştırmaları Dergisi, 12 (1982–98), 319–23. (21) TSA, E.6237. (22) Şlapac, Belgorod-Dnestrovs’kaia krepost’, 69 (Kauffer’s general map of Akkerman town, 1793). (23) Kauffer describes it as shallow here in the 1790s: BOA, Hatt-ı Hümayun (HAT) 56956. (24) Evliya Çelebi, Seyahatname, 5: 62, 63. (25) Evliya Çelebi’s statement that the Bayezid Hamam was down below in the hisar bece combined with this phrase suggests that the port yard was referred to as a hisar bece, though this does not mean that the port yard was the only hisar bece at Akkerman (another one seems to be the half-wall extending up from the inner wall of the great ditch—of course there could be no hamam there). For the meanings of hisar bece see below. (26) Here varos could have referred to the civil yard or else to the town outside the entire fortress. (27) BOA, TT 542, p. 40. While Selim I is not specifically mentioned as opposed to Selim II, as this defter was drawn up during the latter’s reign and refers to the ‘evkaf of the late Sultan Selim’ there is no doubt that the original Selim endowments belonged to the former. (28) Cf. two hamams outside the fortress at Seddülbahir on the Dardanelles: see Chapter 9 by Lucienne Thys-Şenocak et al. in this volume and Lucienne Thys-Senocak, Ottoman Women Builders: The Architectural Patronage of Hadice Turhan Sultan (Aldershot: Ashgate, 2006), 146, fig. 4.15. (29) Evliya Çelebi, Seyahatname, 5: 63. (30) In Figure 7.3a (based on the 1955 survey) near no. 28 only the two almost parallel walls, what we refer to as the north-east and south-west walls, are extant; since then we have excavated the rear wall, which is visible in Figure 7.4a and a very small part of the round tower, no. 28; the barbican with this round tower can be clearly seen below in plans of Kauffer and Förster (Figures 7.7 and 7.8). (31) Caroline Finkel et al., ‘Historical-archaeological investigation of Akkerman fortress, Ukraine 2007’, Anatolian Archaeology, 13 (2007), 11–14, esp. 12, 13. (32) According to Şlapac no other stone barbican has been found on the territory of the Principality of Moldavia. Şlapac, Belgorod-Dnestrovs’kaia krepost’, 135. (33) Ostapchuk, ‘Human landscape’, 44–7.

Page 24 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project (34) For the sixteenth century, most often workers and timber from Moldavia: BOA, MAD 55, fols. 76a-77a; MD 26, p. 277, no. 798 (983/1575–76); MD 29, p. 142, no. 349 (14 Zilkade 984/2 February 1577). (35) Mostly transfer of tax revenues collected locally or in nearby areas: BOA, Ibnülemin Askeri 1400. (36) The restriction and when necessary removal of houses and shops built too close to the walls so as to avoid endangering the defence of the fortress: BOA, MD 16, p. 199, no. 386 (979/1571– 2); MD zeyl 3, p. 72 (983/1575–6). Complaints by garrison troops that owners of houses inside the fortress have been selling their property to out siders (non-troops and/or non-residents of the fortress), thereby leaving the troops who are to reside inside the fortress with nowhere to live and making it difficult to bring replacements when their service at the fortress is over: BOA, MD 73, p. 375, no. 825 (1003/1594–5). (37) Caroline Finkel and Victor Ostapchuk, Outpost of empire: an appraisal of Ottoman building registers as sources for the archaeology and construction history of the Black Sea fortress of Özi’, Muqarnas, 22 (2005), 154. (38) BOA, Cevdet Askeri (CA) 1622 (hüccet); CA 26925, p. 1 (defter). (39) Here either the petitioners are misrepresenting, for by this time, as we saw above, there was at least one other hamam, or perhaps this hamam, in the varoŞ (i.e. the civil yard, suburb or town), was, from the perspective of the garrison troops, located outside their fortress, i.e. the citadel and garrison yard. (40) BOA, MD 28, p. 114, no. 273 (20 Ramadan 984/11 December 1576). (41) We note that taking into account that the last quarter of the eighteenth/first years of the nineteenth century was a time of constant war or diplomatic crises with the Russian Empire, the correlation may be even stronger than the figures suggest; this does not rule out non-crisisrelated repairs—more sorting out of the building data and military-diplomatic data is necessary. Cf. Chapter 9 by Lucienne Thys-Şenocak et al. in this volume, where repairs to Seddülbahir were made in response to damage by natural causes rather than military action. Might it have been that this fortress on the Dardanelles was essentially strong enough for the real or potential threats it encountered? Clearly Akkerman in the eighteenth century was no longer, otherwise there would be no rational explanation for the expense and effort incurred beyond major corruption (for which we so far have no evidence). (42) Finkel and Ostapchuk, Outpost of empire’, 154–5. (43) Rhoads Murphey, Ottoman Warfare, 1500–1700 (New Brunswick: Rutgers University Press, 1999), 111–14; Finkel and Ostapchuk, Outpost of empire’, 161 ff. (44) Finkel and Ostapchuk, Outpost of empire’, 164, 180. (45) ibid., 177. (46) In the eighteenth century a faced ditch was put in after installing a new trace. Finkel and Ostapchuk, Outpost of empire’, 166, 168.

Page 25 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project (47) BOA, Bab-ı Defteri Baş Muhasebe Bina Emini (D.B޸M.BNE) 16070, pp. 5, 15, 16. (48) Mathieu developed the notion of the improvised modernisation of Akkerman (including identification and interpretation of the internal bastions) in his Architectural assessment’ (see n. 11 above). (49) The piled-up earth can be seen in old photographs and it is recorded by contour lines in the 1955 survey of Akkerman. ޸lapac, Belgorod-Dnestrovs’kaia krepost’, 90. (50) e.g. Sebastien Le Prestre de Vaubon, Fenni Muhasara, trans. Konstantin İpsilanti (Istanbul: Ibrahim Müteferrika, 1209/1794–5), 33 ff. (51) The exit and entry to the port yard via the barbican is still visible in one of Kauffer’s plans, where we can clearly see the road meandering through the port yard, from the Water Gate, into the rear of the barbican, through the barbican, and turning left out to the estuary by the side gate (Figure 7.9). Note that in the prospective plan of Kauffer (Figure 7.7), the barbican is drawn without any gate: this either indicates a planned alteration or is more likely a mere simplification of a feature irrelevant to his planned bastions. (52) Şlapac, Belgorod-Dnestrovs’kaia krepost’, 135, fig. 84a. (53) e.g. on bina emini see BOA, MAD 3162, p. 130; D.BŞM.BNE 15923, p. 2; D.BŞM.BNE 16/36 etc.; on it being drawn up by one of the royal architects (.…ma‘rifeti ve ma‘rifet-i ser‘ ve mi‘mar halifelerinden Seyid Ömer Halife ma‘rifetle keşf) see CA 49171; for an example of the chief gatekeeper (ser-i bevvabin) as bina emini, see D.BŞM.BNE 20/25. (54) BOA, MAD 3882, p. 199: bera-yi keşf ve tahmin-i kale-i Akkerman ki ba‘ zı mahaleş m‘nhedim ve ta‘ mireş muhtac bude ve bi-ma‘rifet-i mubaşir ve vezir-i mukerrem Yusuf Paşa valii Özi ve ma‘rifet-i şer keşf ve tahrir ve defter ve ‘arz ve i‘lam şude fermude ber-muceb-i i‘lam ve vezir-i mukerrem Yusuf Paş¸a vali-i Özi ve der-kenar ve telhis ve ferman- i ‘ali fi 19 Receb sene 1121 (24 September 1709) emr dade f i 20 Receb-i m (25 September 1709); see also MAD 3162, p. 130; MAD 9917, p. 9. (55) Hence, there is frequent use of the phrase common in Ottoman documents relating to many spheres, bima‘rifet-i…., which in this context means not ‘by the knowledge of but rather ‘by the action of, with the implication being ‘with the responsibility of. (56) e.g. from the kazas of Babadağı, Mangalya, Karaharman, Niğbolu, Ziştovi, Rusçuk, İsmail, Şümni, Isakcı, Hezargrad, Ternova, Ivraca, Plevna, Eski Cuma, Lofça. BOA, CA 26067; CA 47302; CA 40292; CA 52691. (57) BOA, MAD 3882, p. 200; CA 16049; CA 49185; CA 48020; CA 52778. (58) BOA, MAD 3882, p. 200; CA 47302; CA 48020; Cevdet Mütevelli 851. (59) BOA, CA 49354. (60) For example, numerous groups of unglazed and glazed ware were produced in Akkerman; the same holds for several items of metal found in the port yard. For the thirteenth and fourteenth centuries, A. A. Kravchenko found nine kilns for ceramic production outside the Akkerman fortress (A. A. Kravchenko, Srednevekovyi Belgorod na Dnestre (konets XIII–XIV vv.)

Page 26 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project (Kiev: Naukova Dumka, 1986, 38–46)). In the course of our excavations in the port yard several traces of ceramic (e.g. tripods for glazed ware production), metal and bone article production were established; we reiterate that we have also found evidence that the hamam structure in the port yard was used in metal craft production no later than the second half of the seventeenth to eighteenth centuries. (61) A. V. Burakov, ‘Polyv’’ianyi posud z horodyshcha Dniprovs’ke-2’, Arkheolohiia, 4 (1991), 105– 9, esp. 109. (62) V. Volkova-Nesheva, ‘Sgraffito keramika v bolgarskom iskusstve XIII–XIV vv.’, Polivanaia keramika Sredizemomor’ia i Prichornomor’ia X–XVIII vv. (Kiev: Natsional’na akademiia nauk Ukrainy, Instytut arkheolohii, Kryms’kyi viddil, 2005), 55–61. (63) Kravchenko, Srednevekovyi Belgorod, 44. (64) S. A. Beliaieva, Iu. V. Boltryk, E. E. Fialko, ‘Iuvelirnye izdeliia iz raskopok Portovogo dvora Akkermanskoi kreposti’, First International Congress of Eurasian Archaeology (Izmir, 2007), 20– 2 (abstract). (65) While some experts consider Iznik ware as having been a non-prestigious ware as compared with other table ware, we agree with, for example, Lynda Carroll: ‘Iznik was the most important production center of high quality ceramics intended primarily for elite use until the seventeenth century…These vessels were produced mainly for an elite market—being either commissioned by the court…or other elites.’ Lynda Carroll, ‘Toward an archaeology of non-elite consumption in late Ottoman Anatolia’, in Uzi Baram and Lynda Carroll (eds.), A Historical Archaeology of the Ottoman Empire: Breaking New Ground (New York: Kluwer Academic, 2000), 171. Aside from Akkerman being a centre of trade, the fortress’s Ottoman administrative elite would have used this pottery, as would such elite groups in other parts of the empire (and in Europe as well). (66) Joanita Vroom, After Antiquity: Ceramics and Society in the Aegean from the 7th to the 20th Century A.C. (Leiden: Faculty of Archaeology, University of Leiden, 2003), 193–4. (67) Joanita Vroom, Byzantine to Modern Pottery in the Aegean (Utrecht: Bijlevald Press, 2005), 107, 139. (68) For example, even as late as the nineteenth century, imitation Ottoman pipes, known as‘red Turkish pipes’, were produced in a craft shop in Kiev, R. Chaika, ‘Keramichni liul’ky iz Zhovkivshchyny’, Arkheolohichni doslidzhennia L’vivs’koho universytetu im. I. Franka, 7 (2004), 169–77. (69) BOA, MAD 6. However, according to a list of imports and exports at Kili for MarchSeptember 1505 compiled by Halil İnalcik on the basis of this defter, there is only one entry with nondescript ‘drinking cups’ and two entries with ‘Rus [Ruthenian/Ukrainian or Muscovite/ Russian] cups’. Halil İnalcık, Sources and Studies on the Ottoman Black Sea, I: The Customs Register of Caffa, 1487–1490 (Cambridge, Mass.: Harvard Ukrainian Research Institute, 1996), 135–6. (70) nalcık, Customs Register of Caffa.

Page 27 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Ottoman Northern Black Sea Frontier at Akkerman Fortress: The View from a Historical and Archaeological Project (71) For another example in which the artefact and documentary record so far do not correlate well see J. M. Rogers, Archaeology vs. archives: some recent approaches to the Ottoman pottery of Iznik’, in Çigdem Balım-Harding and Colin Imber (eds.), The Balance of Truth: Essays in Honour of Professor Geoffrey Lewis (Istanbul: Isis, 2000), 275–92.

Access brought to you by:

Page 28 of 28 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Palanka Forts and Construction Activity in the Late Ottoman Balkans

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Palanka Forts and Construction Activity in the Late Ottoman Balkans BURCU ÖZGÜVEN

DOI:10.5871/bacad/9780197264423.003.0008

Abstract and Keywords This chapter examines military building activity in the region in the light of Ottoman sources preserved in the Prime Ministry Ottoman Archive in Istanbul and memoirs written by the senior bureaucrats of the Empire. It aims to assess whether the military building programme of the sixteenth and seventeenth centuries continued in later periods in the same spirit as in the earlier time of conquests and expansion, or if the empire only supported repairs of existing strongholds. The issue was noted by numerous Ottoman writers as early as the Koçi Bey Risalesi in the seventeenth century. This chapter examines four frontier areas of the Ottoman Empire: the Hapsburg borderland in Croatia; the frontier between Montenegro and southern Herzegovina; the fortress line on the banks of the Danube in Wallachia; and the Danube Delta region near the Black Sea. Keywords:   Prime Ministry Ottoman Archive, Istanbul, Black Sea, Koçi Bey Risalesi, Hapsburg borderland, Danube

ONE OF THE MOST IMPORTANT TYPES OF MILITARY CONSTRUCTION in the Ottoman Balkans was the palanka (Figure 8.1). Palankas are small palisaded earthwork fortifications, built from timber,1 and are particular to eastern Europe, where the Ottoman frontier seems to have exhibited certain continuities with the Roman limes system.2 There had often been fortifications on the sites of palankas in the pre-Ottoman period, but after the conquests they were rebuilt with recognisable Ottoman features. A typical palanka was surrounded by a ditch, had an entrance guarded by a watch tower and was connected to the outside world by a bridge over the ditch. Within the palanka there were houses or soldiers’ barracks and buildings where munitions and food were stored. They were constructed along the road network in eastern Europe in order to secure effective communication between strategic areas. Because of their temporary character few palankas survive today, but historical and archaeological research shows that the Ottomans frequently used such structures from the fourteenth to the late nineteenth century.3 (p.172)

Page 1 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Palanka Forts and Construction Activity in the Late Ottoman Balkans In this chapter, military building activity in the region will be examined in the light of Ottoman sources preserved in the Prime Ministry Ottoman Archive in Istanbul and memoirs written by the senior bureaucrats of the Empire. It aims to assess whether the military building programme of the sixteenth and seventeenth centuries continued in later periods in the same spirit as in the earlier time of conquests and expansion, or if the empire only supported Figure 8.1. Palanka in L. F. [Luigi repairs of existing strongholds. The Ferdinando Marsigli], Stato militare background to this question is the longdell’Impero Ottomanno (The Hague, 1732). prevailing paradigm of ‘decline’ of the empire in its later centuries. The issue was noted by numerous Ottoman writers as early as the Koçi Bey Risalesi in the seventeenth century, and by the nineteenth century the Ottoman Empire’s general failures in the fields of diplomacy, the military and the economy were widely criticised from within.4 This chapter examines four frontier areas of the Ottoman Empire: firstly, the Hapsburg borderland in Croatia; secondly, the frontier between Montenegro and southern Herzegovina; thirdly, the fortress line on the banks of the Danube in Wallachia; and fourthly the Danube Delta region near the Black Sea.5

(p.173) Geography Along the Ottoman–Hapsburg frontier in Croatia, the Una and Kupa Rivers delineate the area where the Dinaric Alps form the geographical boundary between the interior and the Adriatic coast (Figure 8.2). Bihke (modern Bihać) was the most important city in the region. Today’s northern Montenegro, on the other hand, was a part of the sancak of Hersek (Herzegovina). The eyalet of Bosna covered territory from the sancak of Pojega and sancak of Bihke in the north to the frontier of the Republic of Ragusa (Dubrovnik) in the south-west, including Montenegro in the far south, near İşkodra (Shkodër). Montenegrin territory was delineated by the settlements on the Adriatic coast in the west and by the River Lim, a branch of the Drina, in the east. Important towns in this border area were Kolaşin (Kolašin), Nikşik (Nikšić), Trebin (Trebinje) and the coastal town of Bar. Çetince (Cetinje) was the religious and administrative centre of the Montenegrins. The region was long disputed between the Venetians and the Ottomans. However, after the Dalmatian lands came under the control of the Venetian Republic, the Montenegrins slowly gained in strength. In the course of the eighteenth century they became a formidable foe, with their constant raiding parties making southern Herzegovina unsafe. The third part of the Ottoman frontier which will be examined here is the part of Wallachia that touches the banks of Danube. This great river is a strategic watercourse coming from Belgrade, passing the fortress of Semendire (modern Smederevo), the Iron Gates (Ott. Demir Kapu), Orşova, and then after Fethülislam (today Kladovo) and Turnu Severin bending southward to Vidin, from where it flows for hundreds of kilometres due east to the Dobrudja and from there to the Black Sea. Inland from Vidin lay the large fort of Belgradcık (Belogradchik, known in earlier Ottoman sources as Belgrad). Today this region is divided between three countries, Serbia, Romania and Bulgaria. From antiquity it was considered to be a region of great strategic importance, as it has many places where the great river can be crossed relatively easily. The

Page 2 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Palanka Forts and Construction Activity in the Late Ottoman Balkans Ottoman fortifications at Belgradcık and Fethülislam are situated directly on the foundations of Roman forerunners.6 At the end of the eighteenth century the Russian frontier became another strategically important area. The Danube flows to the Black Sea through three main branches. The chain of fortification against Russia included the towns of Silistra (Silistre), Hırşova (Hiršovo), İ brail (Braila),7 İsakçı (Isakce), Tulça (Tulcea), İsmail (Izmail) and Kili (Kilia), as well as Akkerman (BilhorodDnistrovs’kyi)8 and Özi (Ochakiv)9 further east, in today’s southern Ukraine. (p.174)

(p.175) Late Eighteenth-Century Historical Background After the treaty of Karlowitz in 1699 the Ottoman Empire lost important territories. Parts of Croatia, Transylvania, all of Slavonia and Syrmia fell under Hapsburg control, as, later, would the Banat of Temeşvar and the Bucovina. In the mideighteenth century conditions on the Belgrade–Vidin part of the Danube line remained fairly peaceful. However, in 1770, during the Russo-Turkish War of 1768–74, the Russians seized the most important strategic points on the banks of the Danube

Figure 8.2. Map of the Ottoman Balkans c. 1800 showing the main places mentioned in the text.

such as Kili, Akkerman, İsmail, Bender and İbrail.10 The construction of chains of strongholds in the sixteenth and seventeenth centuries continued in the eighteenth century. The long wars against the Hapsburgs (1683–99, 1737–9) caused the Ottomans to strengthen the frontier areas of Croatia and Montenegro. A great number of palankas were built between 1701 and the 1720s, twenty-eight forts and eight palankas being constructed on the banks of the Sava and Una rivers.11 In 1734, Muşinzade Abdullah Pasha surveyed military structures in the region and wrote a report on the most urgently needed repairs.12 (p.176) Another major military building programme began in 1742, concluding towards the end of the eighteenth century. The Ottoman central administration saw new military construction as crucial for the defence of the Ottoman–Hapsburg frontier.13 Halil Hamid Pasha, one of the viziers of Abdülhamid I between the years 1781 and 1785, paid particular attention to frontier fortifications, bringing in French military experts.14 This expensive programme relied heavily on local resources, and local subject peoples were coerced into compulsory labour on the fortifications. On the Montenegrin part of the frontier a chain of forts was built controlling strategic points such as İşpozi (Spuž), Nikşik and Trebin.15 In Wallachia too military construction on the border took place.16 The Hapsburgs, in turn, strengthened their frontier, and in 1783 the Hofkriegsrat (Austrian High Command) decided to build new fortifications on the Croatian border.17 After the death of Joseph II (1790), the Hapsburgs, now under the new Emperor Leopold II, agreed to make peace with the Ottomans. In 1791 at the Treaty of Svištov (Sistova) on the Lower Danube the frontier was revised.18 The Prussians played a key role in the treaties between the Page 3 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Palanka Forts and Construction Activity in the Late Ottoman Balkans Hapsburgs and the Ottomans.19 In the Treaty of Svištov the Croatian border was graphically defined on a map which was prepared by the Prussian diplomats. Dating to AH 1208 (1793–4), this map shows strategic points and the frontier delineated with a bold line20 (Figure 8.3). It was drawn by an Austrian engineer who probably specialised in military geography, and covers territory from Sternitsa and Kulen-Vakuf in the south to Glina in the north and to Gradiška in the east. The Dinaric Alps, with the Una and the Sava Rivers, form natural boundaries. (p.177) Important strongholds in the area were Bihke, Gradişka and Dubica, within Ottoman territory, and Cetin Grad on the Hapsburg side. Among the thirteen localities mentioned in the annexe of the map are three military buildings: the palanka of Dreznik (Drežnik Grad), and the fortresses of Cetingrad and Aşurviç. The rest comprised small villages and farms. So the area under dispute was not restricted to military strongholds, but also comprised civil settlements as well as farms which supplied provisions to the military and contributed to the economic survival of the frontier territory.

(p.178) Late Military Buildings in the Balkans The Hapsburg Frontier The memoirs of Cevdet Pasha (1822–95), an able bureaucrat and eminent historian, who was appointed by Fuat Pasha21 to serve in the Balkans, highlight the extent to which the repeated attacks of the Montenegrins constituted a major problem for the people

Figure 8.3. Austrian map depicting the Ottoman–Hapsburg frontier in Croatia (BOA, HAT 241/13506. Date AH 1208).

of southern Herzegovina.22 When in the mid-nineteenth century unrest began in Montenegro, the Ottomans decided in 1860–1 to reinforce their control over the region with an effective system of military strongholds.23 The Foreign Minister, Âli Pasha, wrote a report which Cevdet quotes, describing the struggles between the Christian populations: ‘the local Muslim people— who are in a desperate situation in Belgrade because of constant Serbian raids—should be transported to other settlements, such as Vidin, Nišand Sarajevo.’ He mentioned a map showing houses in the Muslim quarter of Belgrade which ‘should be expropriated and destroyed, to create an open area around the fort’.24 Cevdet Pasha, who inspected the area, stated that there were two ways by which the eyalet of Bosna25 could be saved: through a thorough overhaul of the existing strongholds and through the provision of munitions. He suggested demolishing military buildings which did not meet contemporary requirements and served no useful purpose. On the other hand, quoting Âli Pasha’s report, Cevdet stressed that ‘fortifications such as Fethülislam, Semendire, Böğürdelen (Šabac), Sokol and Öziçe formed the boundary of the empire, and they should remain intact.’26 Forts in the eyalet were generally in poor condition, often half-demolished and lacking essential equipment. Thus a budget and energetic official supervisors who could make effective decisions and direct construction works were required. When Cevdet Pasha visited the region, he assessed the current requirements, both civilian and military, of the border territories, in much the same way as Muşinzade Abdullah Pasha had done in the eighteenth century. (p.179) On the basis of Cevdet Pasha’s surveys and recommendations various construction projects were initiated. For instance, next to the Montenegrin border around İşkodra (Shkodër) there was a small island called Gramacor (or perhaps Grand Major) where there was a ruined tower in need of rebuilding. A plan was prepared to construct a tower with two ramparts. The local military administrator of Krajina, Murteza Agha, was appointed to undertake the

Page 4 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Palanka Forts and Construction Activity in the Late Ottoman Balkans construction,27 and the tower was soon completed.28 Another settlement on the Montenegrin coast, the town of Bar29 (now Stari Bar), was inspected, and responsibility for construction of a defensive wall on the shore was given to Arab Mehmed Pasha in 1863–4.30 In Liyubuşka (Ljubuški, 30 kilometres south-west of Mostar), the frontier against Hapsburg Dalmatia required the construction of towers, which was undertaken by Hafiz Efendi. The budget was about 15,000 guruş31 which was provided by local civilians and provincial resources.32 Military buildings here were constructed to safeguard the inhabitants against raids from across the Dalmatian border. In accounts of the work at Liyubuşka, we encounter a figure named Linardovich, a German engineer who prepared some reports and took part in the construction work. He devised a plan to deal with the problem of transporting timber from the dense forest around the River Lim, which required the removal of great rocks blocking the River Drina to allow its shipment. He also participated in the construction of a bridge over the Drina. Cevdet Pasha adds that a han in the vicinity could be converted into a barracks for the soldiers defending the bridge and the village. The task was again given to Linardovich.33 Works, in particular the construction of bridges, were undertaken at Nikşik, one of the most important towns of southern Herzegovina (becoming part of Montenegro in 1878). Lying 45 kilometres east of the fortress of Trebin, Nikşik was situated on a rocky outcrop in the middle of the plain of the same name on the site of the ancient settlement of Anagastum and medieval Onogost.34 The bridges built by the Ottomans allowed the safe transportation of the army during the Montenegrin revolt in 1875–6.35 Bihke, the centre of the sancak of the same name, one of seven sancaks in Bosnia, was (p.180) situated on the River Una very near to the Hapsburg frontier.36 The fortress was repaired by the Governor of Bosnia, Mehmed Pasha, in 1763. Another project was the repair of Kolaşin Palanka. Kolaşin (Kolašin) is situated on the River Moraca in eastern Montenegro and in Ottoman times (till 1878) it belonged to the sancak of Novi Pazar. Upper Kolaşin (Kolaşin-i Bala) was the strategic centre of the area. Towards the midnineteenth century the inhabitants of the district were forced to emigrate because of the constant raids of Montenegrin bandits. However, Ottoman officials encouraged them to return to their homeland on the borders, and decided to fortify the main settlement, the village of Kolaşin. Miralay Selim Bey and Yüzbaşı Hüseyin Efendi were appointed to organise this. They built bridges over the River Tara and various towers on the Montenegrin border. The towers were named after the leading politicians of the day, Fuat Pasha and Cevdet Pasha. Selim Bey and Hüseyin Efendi also had a role in building Kolaşin Palanka, which was considered a defence project of first-rate importance.37 A document of 1865 written to a Cevdet Efendi (probably Cevdet Pasha himself) gives details concerning the construction project at Kolaşin Palanka. A committee chaired by Münif Pasha was in charge not just of construction but also of the resettlement of refugees. The fortifications built by Hüseyin Efendi, who had earlier built towers and bridges over the Tara, were considered a success.38 Other palankas were built or restored in the same region, and the same Hüseyin Efendi supervised works at Aziziye Palanka (1864–5) near ‘Zaklebun’ (unidentified). In Korita near Bileke (Bileća) in southern Herzegovina another palanka was built.39 Cevdet Pasha also placed a wing of the fortress overlooking the great city of Prizren on his list of useless buildings to be demolished. Prizren’s importance as a civilian centre had greatly expanded in the nineteenth century, although the roots of the settlement are said to go back to the Roman town of Therand, and in the Middle Ages it had been the capital of the Serbian Tsar Stefan Dušan. A document from 1852 proposes to the commander-in-chief of the Ottoman army

Page 5 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Palanka Forts and Construction Activity in the Late Ottoman Balkans in Rumeli that the castle should be converted to a new barracks, with the buildings designed by Ottoman military engineers, but the actual construction undertaken by civilian entrepreneurs, who would bid for the contract at auction.40 At the same time another barracks was planned for the Muslim Albanian town Debre-i Bala (Debar) in the west of Macedonia. (p.181) The letter suggests that the plan for the construction of this second barracks should be dropped if Prizren barracks was to be realised. The Danube Frontier A palanka project in Belgradcık which was proposed to the Tophane-i Amire (Imperial Cannon Foundry in Istanbul) in 1861–2 may also have been connected with this programme of construction. Belgradcık belonged to the sancak of Vidin.41 The castle was rebuilt and greatly extended during the age of Mahmud II in 1837–8, the work being executed in ashlar masonry. On the western wing a large extension, a palanka, was constructed, and all the walls were reinforced with additional ramparts. The archival document notes that to the left of the existing castle there was an excellent position for cannons.42 Kladovo, known as Fethülislam to the Ottomans, was another strategic point in the sancak of Vidin. It was situated on the banks of the Danube and belonged to the same border region as Semendire and Böğürdelen. The fort of Fethülislam was first built in 1524. In the midseventeenth century Evliya Çelebi relates that the building had four double walls and a palisade next to the river bank.43 The fortress was occupied by the Austrian army in 1717–39, and again later for short periods. It was rebuilt by the Ottomans in 1813 during the reign of Mahmud II. A report in the Ottoman archives, dated 1795–6, records a project for the reconstruction of the Fethülislam Palanka and its surroundings.44 It mentions that the annual tax revenue of Fethülislam was originally transferred to the foundation of the Harameyn (Mecca and Medina) as well as to the Aya Sofya Library in Istanbul. The document also urges that steps be taken to repair palankas on main roads. It states that the wood for the construction of palisades should be requested from the Wallachian Mountains. The Voyvode of Wallachia, a vassal of the sultan, was ordered to send the wood, as well as 100 construction workers. The construction itself was to be organised by Vizier Mehmed Pasha with the help of local experts. The cost was about 4,900 gurus, and the work took around fifty to sixty days. Further evidence for Fethülislam is offered by a commemorative key depicting the fortress. The geometrically shaped enceinte was reinforced by triangular bastions on the corners and by outer ramparts. Inner buildings, probably administrative blocks and (p.182) barracks for the use of the garrison, can also be observed.45 Cevdet Pasha also quotes a report by Âli Pasha that the castles of Sokol and Öziçe (Užice, known from 1950 to 1990 as Titovo Užice) were in a very bad condition, could no longer be defended and had to be completely demolished.46 The Black Sea Frontier One of the important strongholds in the Black Sea area was Akkerman, which was repaired between the years 1792 and 1801.47 Near Akkerman, Yanık Palanka on the River Dniester was first built in 1616 during the reign of Ahmed I, and a report of 1784 urged that it be repaired. As a result one İsmail Pasha was appointed to oversee construction and the official architect surveyed the existing building and prepared a feasibility report.48 The fortress town of İsmail (Izmail) in the sancak of Silistre was a crucial strategic point on the Danube Delta. Alongside Kili, İbrail and Akkerman, İsmail was one of the main strongholds guarding the Russian frontier.49 The Ottomans realised its strategic significance, and they built a large and well-equipped fortress here in 1768–9.50 This was occupied by the Russians in 1790,

Page 6 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Palanka Forts and Construction Activity in the Late Ottoman Balkans but returned to the Ottomans in 1792, when Selim III immediately started to reconstruct it. Two archival documents written by the construction manager (bina emini) of İsmail fortress, Esseyid Numan Bey, in 1793–4 survive. In the first letter he requests a building plan for the stronghold and technical details concerning the construction procedure. In the second letter, presumably written after the completion of the building works, he writes that he is presenting the Ottoman court with a silver model of İsmail fortress.51 Numan Bey states that a depiction on a map (probably meaning a situation plan) would not represent the fortress accurately, so a threedimensional silver model, made to scale, would best show the building’s inner and outer contours.52 A schematic drawing illustrates the fortress’s situation on the banks of the Danube, where it forms a triangular enceinte with triangular platforms (p.183) projecting towards the outer side supporting the defence. Inside the enceinte, roads and barracks are depicted.53 Architectural models seem to have been widely used. Another large model, made by an English engineer, was on display at Kaĝıthane, on the banks of the Golden Horn in Istanbul, in 1790–1. Based on a drawing, the model included the location of cannons and other weapons. Sultan Selim III was invited to the opening ceremony, and replied to the letter of invitation with a brief note that he would ride out to see the model as soon as the weather was suitable.54 Thus fortification designs were best represented as architectural models, which provided basic information on military strategy.

Late Civil Buildings in the Balkans In the classical age of the empire, conquests were accompanied by a process of Ottomanisation of formerly Christian towns. In the Balkans and Hungary, as well as in Cyprus, the transformation of newly conquered territories began with a tahrir survey, after which viziers and senior commanders would establish civilian settlements (mahalle). Apart from the conversion of the main church into the Friday mosque, they would found charitable foundations (vakıf), including mosques, medreses and schools, often accompanied by a public bath.55 Accommodation around the vakıf complex was then offered to migrants from elsewhere in the empire, who, whether they came voluntarily or by force, played an important part in the town’s economic life.56 Ottoman towns were made up of specific neighbourhoods designated on the basis of religion. In the classical age, areas close to the frontier might be guarded by local troops, as happened at Osijek. Civilians who were listed in the army registers might join the army in times of unrest or war. In the nineteenth century, however, the underlying principles of settlement administration shifted. According to the Cevdet Pasha, civilians in frontier settlements were still expected to support the army in wartime. However, because of rising nationalistic tendencies, their loyalty to the Ottomans could no longer (p.184) be taken for granted, and many participated in paramilitary organisations or civil disobedience.57 They might co-operate with overseas powers such as Russia or France which openly asserted their interests in Ottoman territory through diplomatic channels. However, the Ottomans needed the collaboration of the local people in Bosnia, Herzegovina and Montenegro, to ensure the continued prosperity of local inhabitants and to ensure that the tax revenues collected by the Treasury remained constant. As Cevdet frequently notes, this depended on the Ottomans being able to maintain security in the region. Control had to be asserted through the constant repair and construction of military buildings. In addition, however, the state undertook the construction of numerous civil buildings as a way of enhancing the Ottomans’ image among locals and improving relations between the state and its subjects.

Page 7 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Palanka Forts and Construction Activity in the Late Ottoman Balkans The growing interest on the part of the Ottoman state in its Christian population is illustrated by the map Cevdet Pasha prepared for the Sublime Porte which included information on the geography and demography of the Montenegrin frontier.58 In some towns, the state erected new buildings for Christians. In Mostar in Bosnia, for instance, children of the Catholic community studied in the newly built Rüşdiye School, which was commonly known as the Latin Mektebi thanks to the religious persuasion of its pupils.59 Before the town’s occupation by the Austrian army in 1878, the local administration in Bihke also constructed public buildings to enhance the area’s development. One bridge was built in 1869, followed by another in 1873 over the Una. A new government office (hükümet konağı) was built in 1870 and a Catholic church in 1869.60 As an able supervisor of construction works, Hüseyin Efendi (previously encountered at Kolaşin) was sent to the district of ‘Korianiç’ (unidentified) to repair houses in that area that had been damaged during civil strife. He made plans for a completely new main road between Mostar and Koniça (Konjic), following the course of the Neretva River instead of making the difficult crossing over the flanks of Mt Prenj, as the old road did.61 In the Vidin region, public buildings were erected in Belgradcık after the completion of the artillery fort. New settlers were brought into the district, and a school (1865) and churches (1871–2) were constructed.62 (p.185) Another striking example of this policy comes from İşkodra, where schools for younger children and a mosque were erected. Cevdet states in a letter of 1863–4 that some Christian pupils who had finished these schools converted to Islam and were circumcised. Graduates of the schools in İşkodra together with some local inhabitants formed together a military troop of 5,173 soldiers.63 Apart from maintaining good relations with the local people, this shows that the state also aimed at establishing local troops who had been trained in the newly built schools in civil settlements and who tended be consciously pro-Ottoman. In some areas, the Ottomans clearly thought that training local troops was much more important than building new frontier strongholds. This is a clear shift from the earlier system of the sixteenth and seventeenth centuries. Until the eighteenth century, frontier villages and local peoples were listed in the army register, both Muslims as part-time soldiers (mirili or eyalet askeri) and Christians such as voynuks in Bulgaria or the vlachs and derbendcis in Romania.64 These people were exempted from taxes, but they were without proper military training. Military building activity was often linked in some way with civilian economic activity. To safeguard the salt trade around İşkodra, Cevdet ordered the building of a tower on the hill around the harbour which guarded the two small settlements of Mişik (which was Ottoman) and Kastellesua around the Castle of Alexandre65 in Austrian territory66 These were apparently dependent on each other. Some military buildings were constructed with the collaboration of civilians. A Yusuf Agha helped to restore the fort of Vranina, where both Muslim and Christian civilians were encouraged to live.67 To give another example, estate owners around Nikşik supplied grain to local inhabitants as well as to the Ottoman army. However, Montenegrin assaults damaged the agricultural base of the region, resulting in a crisis that extended to urban areas. As a result, Cevdet urged that the attacks should be stopped at once, and the civil settlements should be revitalised through an immediate programme of repairs.68 Military buildings here would serve both to protect the territory and to ensure its economic survival.

Conclusion In the fifteenth and sixteenth centuries the regional defence system in the western part of the empire was essentially designed against Venetian attacks. In order to protect the (p.186) main military arteries into the heart of Europe, the Ottomans constructed a network of strongholds. Strategic locations such as harbours, main roads (such as the Via Militaria and Via Aegnatia)

Page 8 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Palanka Forts and Construction Activity in the Late Ottoman Balkans and the banks of the Danube were fortified, including a significant number of palanka forts. In the sixteenth and seventeenth centuries, however, military construction strategy was intended mainly to defend the frontier against Hapsburg attacks and to build bases for cross-border raids. The international situation forced the Ottomans periodically to reform and strengthen their fortification system. As we have seen, in 1734 Muşinzade Abdullah Pasha inspected the strongholds in the Balkans and wrote a report calling for their repair. Towards the turn of the eighteenth century sultans like Abdülhamid I and Selim III took a personal interest in the construction and restoration of fortifications. Strongholds on the Austrian and Russian frontiers, especially those near the Black Sea, were repaired or built from scratch. This pride in constructing new strongholds was demonstrated by Selim III’s enthusiasm to examine the fortification model at Kaĝıthane, or the presentation to the Sublime Porte of the precious silver model of İsmail fort. In the eighteenth century, policy emphasised less expansion than sustaining Ottoman power in Rumelia with its substantial population of Muslim immigrants and its important economic links with both east and west. A further building programme on the frontier started in the mid-nineteenth century in the wake of nationalist revolts in the Ottoman Balkans. Economic and diplomatic difficulties had reduced the empire to being known as ‘the sick man of Europe’, but a shift in the Ottoman attitude to the provinces can be observed with the arrival of Cevdet Pasha in the Balkans. He wrote reports on the strongholds, towers, bridges and other military buildings to be repaired or demolished, and he also established new procedures for building in the region. Instead of prestigious military projects as in the eighteenth century, it seems that construction activity in the mid-nineteenth century was limited to relatively small projects and additions to previously built infrastructure. Palankas, on the other hand, were among the most vital elements in the fortification network. Erected rapidly and cheaply, they remained significant from the sixteenth century onwards. Although in Europe Vaubanesque fortification design dominated large military projects after the end of the seventeenth century, the basic architectural characteristics of palankas rarely altered. This indicates that the basic plan of the palanka still met the requirements of war technology. Construction of civil buildings changed over this period. In the sixteenth and seventeenth centuries public structures in the Balkans and Hungary were generally erected by the pashas who ruled in the newly conquered regions. This building activity, and the creation of pious foundations for their maintenance, was one of the most important aspects of the Ottomans’ cultural policies in their newly won lands. In the mid-nineteenth century, however, the Ottomans eventually recognised that their civilian population were not merely statistics who had volunteered for military service to avoid paying taxes, but individuals who might be influenced by the new nationalist (p.187) ideas prevalent in the region. Frequent rebellions led the administrators to view settlements not only in terms of their military value, but also from an economic and social perspective. On the other hand, in the nineteenth century the newly organised central bureaucracy had an evident impact on public works, while at the same time the system of pious foundations was no longer effective in the empire. Priorities for public works were generally decided by official experts appointed by Istanbul, one of them being Cevdet Pasha. The public building campaigns in the 1880s, such as the construction of schools, governmental houses and clock towers, were part of the project of modernisation conducted in Ottoman cities in all provinces. It seems that the late Ottomans intended civil construction to serve several purposes. Firstly, military control over the region would protect economic development. Secondly, the state tried

Page 9 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Palanka Forts and Construction Activity in the Late Ottoman Balkans to restore its legitimacy among its subjects through programmes of public works, especially the construction of churches. According to Cevdet Pasha, the state-sponsored construction programmes in southern Herzegovina and Ottoman Montenegro were intended to cement good relations with the local population and in particular to help immigrants to the region. In addition, the state still hoped to recruit its pro-Ottoman Christian subjects to the military. Thus the empire initiated civil building projects to improve the social welfare of the civilians whose younger generations were to be trained in modern barracks and who were expected to join units of local troops along the frontier. Notes: Proceedings of the British Academy 156, 171–187. © The British Academy 2009. (1) Palanka is also a technical term indicating a method of construction. Evliya Çelebi describes palanka masonry as timber walls, usually made from oak piers, where the vertical layers were filled with mortar:…meşe aĝaçları direklerinden sandık sandık her direk birbirine geçme sandıklar çatılıp şeddadi tuğla duvar üzere çatma sandıkları koyuphorasan ve kireç ve alçı ve cibis ile rıhtım edüp kalayı dahi yüksek etmişler…dolma rıhtımpalanka duvar…(Evliya Çelebi, Seyahatname, 6: 105). There were also stone palankas, such as Bela Palanka (1637) on the road between Niš and Sofia, where the antique ruins of the Roman town Remesiana were re-used (Ireçek, Belgrad, 21, 84, 119 (see n. 3 below)). However, Evliya describes the building as Musa Pasha Fortress, not as a palanka: Seyahatname, 5:188. Palanka forts may also be built as an extension of the main fortress, such as Belgradcık (Belogradchik). (2) On the Hungarian limes see Sándor Soproni, Der spätrömische Limes zwischen Esztergom und Szentendre (Budapest: Akadémiai Kiadó, 1978); Zsolt Visy, Der pannonische Limes in Ungarn (Budapest: Corvina, 1988), 10–20, esp. map 15 shows features very similar to the Ottoman palanka network. (3) In addition to archival materials, several literary sources discuss palankas: Evliya Çelebi’s seyahatname mentions numerous palanka forts in the Balkans, Hungary and Black Sea coast, and Heinrich Ottendorf included illustrations of palankas on the Danube in his manuscript Der Weg von Ofen auf Griechisch-Weißenburg (Vienna, 1665; Österreichische Nationalbibliothek, Vienna, MS Cod. 8481). In the nineteenth century, Constantin Josef Jireček noted palankas on the Via Militaria in his Die Heerstraße von Belgrad nach Constantinopel und die Balkanpässe (Prague, 1877); Turkish translation: Konstantin Yosif İreçek, Belgrad İstanbul Roma Askeri Yolu, trans. A. K. Balkanlı (Ankara: Kültür Bakanlıĝı, 1990). (4) For example, Hüseyin Kazım Kadri, a senior politician in the late nineteenth and early twentieth centuries, openly criticised the empire’s administrative failures in the Balkan provinces: Hüseyin Kazım Kadri, Balkanlardan Hicaz’a İmparatorluğun Tasfiyesi, ed. K. Büyükcoşkun (Istanbul: Pınar Yayınları, 1991). On the historiography of the ‘decline’ see Stanford J. Shaw, History of the Ottoman Empire and Modern Turkey, II: Reform, Revolution and Republic. The Rise of Modern Turkey 1808–1975 (Cambridge: Cambridge University Press, 2002); Cemal Kafadar, ‘The question of Ottoman decline’, Harvard Middle Eastern and Islamic Review, 4 (1997–8), 30–71; Donald Quataert, Ottoman history writing and changing attitudes towards the notion of “decline”’, History Compass, 1/1 (2003). (5) On the Ottoman palanka network in Hungary, see Burcu Özgüven, ‘The palanka: a characteristic building type of the Ottoman fortification network in Hungary’, Electronic Journal of Oriental Studies, 4 (2001) (= (= M. Kiel, N. Landman and H. Theunissen (eds.), Proceedings of Page 10 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Palanka Forts and Construction Activity in the Late Ottoman Balkans the 11th International Congress of Turkish Art, Utrecht, The Netherlands, August 23–28, 1999), 1–12, and eadem, ‘Characteristics of Turkish and Hungarian palanka-protected settlements along the River Danube’, in Ibolya Gerelyes and Györgyi Kovács (eds.), Archaeology of the Ottoman Period in Hungary (Budapest: Hungarian National Museum, 2003), 155–60. For work by Hungarian scholars, see for example, Attila Gaal, ‘Török Palánkvárak a Buda-Eszéki ut Tolna Megyei Szakaszán’, in Magyar és Török Végvárak (1663–1684), Studia Agrensia, 5 (1985), 55–88. For an example from the Hungarian town of Parkan see Lajos Fekete, ‘Osmanisch Parkan’, Körösi Csoma Archivum, 1 (1921–5), 384–8. (6) İreçek, Belgrad, 152 suggests that the crossing at Fethülislam/Kladovo had maintained its importance from the Roman period to the nineteenth century. (7) Quoting a Polish document of 1540, Maxim notes that a stone fortress (arcem muro) in Braila was built by the Ottomans: Mihai Maxim, ‘Osmanlı Döneminde Bir Tuna Liman Kenti: İbrayil (Braila)’, İstanbul Üniversitesi Edebiyat Fakültesi Güney-Doğu Avrupa Arastırmaları Dergisi, 12 (1998), 173–87. (8) On Akkerman, see Chapter 7 above, Victor Ostapchuk and Svitlana Bilyayeva, ‘The Ottoman northern Black Sea frontier at Akkerman fortress: the view from a historical and archaeological project’. (9) On Özi, see Caroline Finkel and Victor Ostapchuk, ‘Outpost of empire: an appraisal of Ottoman building registers as sources for the archaeology and construction history of the Black Sea fortress of Özi’, Muqarnas, 22 (2005), 150–88; further on the archival material Muzaffer Erdoğan, ‘Mehmet Tahir Ağa Hayatı ve Mesleki Faaliyetleri (3)’, İstanbul Üniversitesi Tarih Dergisi, 9 (1958), 164–8. (10) Barbara Jelavich, History of the Balkans: Eighteenth and Nineteenth Centuries (Cambridge: Cambridge University Press, 1983), 69. (11) On the Russian frontier too extensive repairs to the fortresses were required. At Acu-su Boğazı near Azak (Azov) a small palanka, barracks and a building complex inside were renovated in 1728–9. The chief architect, Mehmet Agha, was sent to Acu for a preliminary survey of the buildings’ repair needs. See Nejat Göyünç, ‘The procurement of labor and materials in the Ottoman Empire (sixteenth and eighteenth centuries)’, in Jean-Louis Bacqué- Grammont and Paul Dumont (eds.), Économie et Sociétés dans l’Empire Ottomane (fin du XVIIIe-début du XXe siècle). Actes du colloque de Strasbourg, 1er-5 juillet 1980, Congrès d’Histoire Économique et Sociale de la Turquie (Paris: Colloques internationaux du CNRS, 1983), 327–33. (12) Avdo Suceska, ‘The eighteenth-century Austro-Ottoman wars’ economic impact on the population of Bosnia’, in Gunther Rothenberg, Bela Király and Peter Sugar (eds.), War and Society in East Central Europe, II: East Central European Society and War in the Prerevolutionary Eighteenth Century (New York: Brooklyn College, 1982), 339–50. (13) Suceska, ‘The eighteenth-century Austro-Ottoman wars’, 343, citing Hamdija Kreševljaković, ‘Stari bosanski gradovi’, in Naše Starine (Our Antiquities) (Sarajevo: Institute for the Protection of Cultural Monuments of BiH), 1 (1953), 7–45. (14) Jelavich, History of the Balkans, 117.

Page 11 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Palanka Forts and Construction Activity in the Late Ottoman Balkans (15) Wayne S. Vuchinic, ‘Prince-Bishop Danilo and his place in Montenegro history’, in Rothenberg et al. (eds.), War and Society, II: 271–99. Nikşik fort was built on the ruins of the old Roman city of Anagastum. Although there is a medieval fortress from the ninth century around Trenbinje, it was one of the few towns in the region where the Ottomans built a new urban settlement. The town was completely encircled by Ottoman walls in the eighteenth century. See Amir Pašić, Islamic Architecture in Bosnia and Hercegovina (Istanbul: IRCICA, 1994), 15, 19. (16) Florin Constantiniu, ‘Tradition and innovation in the eighteenth-century military structures of the Rumanian lands’, in Rothenberg et al. (eds.), War and Society, II: 391, 395. (17) Gunther E. Rothenberg, The Military Border in Croatia 1740–1881: A Study of an Imperial Institution (Chicago: University of Chicago Press, 1966), 70, 72. There are numerous sources on Hapsburg fortifications against the Ottomans. For example, several manuscripts in the Austrian National Library describe the Hapsburg frontier, such as: Martin Stier, Graz und die Grenzfestungen gegen die Türken (Graz, 1657; MS Cod. 9225); anon., Festungen in Ungarn, Kroatien, Slawonien (Vienna, c. 1660; MS Cod. 8607 and 8609); anon., Lagerund Schlachtpläne des Feldzuges Karls von Lothringen gegen die Türken in Ungarn 1737 (Vienna, 1737; MS Cod. 8286). (18) Rothenberg, The Military Border in Croatia, 79 (19) Mustafa Nuri Paşa, Netayic-ül Vukuat: Kurumları ve Örgütleriyle Osmanlı Tarihi, ed. Neşet Çağatay (Ankara: TTK, 1992), 3–4: 186–9. (20) BOA, Hatt-ı Hümayun (HAT) 241/13506. (21) Fuat Pasha, the prime minister and Âli Pasha (or Mehmet Emin Âli Pasha), the foreign minister of the time, were among the leading figures of the Tanzimat era. They served the Sublime Porte from 1855 to 1871. (22) Cevdet Paşa, Tezakir, ed. Cevdet Baysun (Ankara: TTK, 1991), vols. 13–20 and 21–39. (23) Maurus Reinkowski, ‘Double struggle, no income: Ottoman borderlands in northern Albania’, in Kemal H. Karpat with Robert W. Zens (eds.), Ottoman Borderlands: Issues, Personalities and Political Changes (Madison: University of Wisconsin Press, 2003), 244. (24) Cited in Cevdet, Tezakir, 13–20: 112–26:…sefaletten ve felaketten kurtarılmaları Şıkkında görünmeğin bunlar Vidin, Niş ve Bosna gibi münasib mahallere yerleştirilip elde olan haritadan müstebân olduğu üzere sırf ehl-i islam mahallatında olan haneleri devletçe bi’l-iştira hedm birle kal’aya meydan bırakılması… (25) In the early nineteenth century, Bosna Eyaleti included Herzegovina. (26) Cevdet, Tezakir, 13–20: 117, 122. However Âli Pasha also expresses concerns about Sokol and Öziçe because of their ruined state. (27) Ibid., 13–20: 175–6. (28) Ibid., 13–20: 187. (29) For Bar: ‘There is a large palanka by the swampy lake…’, Evliya Çelebi, Seyahatname, 5: 70– 1. Page 12 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Palanka Forts and Construction Activity in the Late Ottoman Balkans (30) Cevdet, Tezakir, 13–20: 267. Arab Mehmed Pasha was the commanding officer of strongholds (istihkâm mirlivâsı) in Bar. (31) 15,000 guruş was equal to the annual tax paid by Işpozi (Spuž), a small town of 450 houses in Montenegro, in the fiscal year 1861–2. Ibid., 13–20: 191, 211. (32) Ibid., 21–39: 16, 17. (33) Ibid., 21–39: 59. (34) Jugoslawien (München: Polyglott-Verlag, 1990–91, 22nd printing), 307. (35) Ekrem Hakkı Ayverdi, Avrupa’da Osmanlı Mimari Eserleri, II: Yugoslavya (Istanbul: Istanbul Fetih Cemiyeti, 2000), 3: 266. According to Ayverdi, an early bridge in Nikşik was built by a Hacı İsmail from Nikşik in 1807. (36) Cevdet, Tezakir, 21–39: 93; Evliya Çelebi says that the fortress was previously built by the Croatian dynasty of the Zrinyis, and it was conquered by Düş Hasan Pasha, governor of Bosnia during the age of Murad III: Seyahatname, 5: 270–1. (37) Cevdet, Tezakir, 23: 51, 53. (38) BOA, A.MKT.MHM (Sadaret Mektubi Mühimme Kalemi) 307/60. Hüseyin Efendi is mentioned in the document as being promoted to the rank of Binbası, probably because of his achievements here. (39) Cevdet, Tezakir, 23: 55. (40) BOA, A.MKT.MVL (Sadaret Mektubi Meclis-i Vâlâ) 53/86. (41) Evliya Çelebi does not mention the fortress, so it may be concluded that the building must be later than the 1660s. On the sancak of Vidin see Chapter 17 below, Rossitsa Gradeva, ‘Between hinterland and frontier: Ottoman Vidin, fifteenth to eighteenth centuries’. (42) BOA, A.MKT.NZD (Sadaret Mektubi Kalemi Nezaretler evrakı) 393/60:…Belgradcık Kalesinin cenahı yesarında bulunan tabiyyeye mail olan tepe üzerine bir palanga insasına ve teferruatına dair olan tezkire-i aliyye-i asafhanelerinin arz ve takdimine… (43) Evliya Çelebi, Seyahatname, 7: 175–6. (44) BOA, HAT 110/4387. (45) Zeki Sönmez, ‘Balkanlar ve Karadeniz Çevresindeki Osmanlı Kale Mimarisi Hakkında Bazı Görüsler’, in Balkanlarda Kültürel Etkileşim ve Türk Mimarisi: Uluslararası Sempozyum Bildirileri (Ankara: Atatürk Kültür Merkezi Yayınları, 2001), 2: 649–66. (46) Cevdet, Tezakir, 13–20: 117 (47) Pars Tuĝlacı, Osmanlı Şehirleri (Istanbul: Milliyet, 1985), 313.

Page 13 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Palanka Forts and Construction Activity in the Late Ottoman Balkans (48) BOA, HAT 442/H. Evliya mentions that Yanık Hisar was a fortress around Akkerman and the Dniester on a strategic highway. Its tower with a wooden top, he wrote, was like the Tower of Galata. There was also another palanka in the vicinity: Seyahatname, 5: 64. (49) On the survey in İzmail in 1784–5: Erdoğan, ‘Mehmet Tahir Ağa’, 163. (50) Mustafa Nuri Paşa, Netayic-ül Vukuat, 2–4: 186. (51) BOA, HAT 239/13302. (52) Ibid.,…mücessem simi mezkurdan başka sureti dahi bir kıt’a haritaya vasi ve haremi ve haddi kasrı beyan olunarak mikyası üzere tersim olunduğunda malumu aliyyeleri buyurulmak için takdim olundu. (53) Sönmez, ‘Balkanlar’, 649–66. (54) BOA, HAT 255/14605: Resmi üzerine mühendis İngiltere beyzadesinin Ka g ıthane’de inşa buyurduğu resmi kaleolup hitam bulup… (55) Barkan’s classic essay discusses the early phase of Ottoman colonisation: Ömer Lütfi Barkan, ‘Osmanlı İmparatorlugu’nda Bir İskan ve Kolonizasyon Metodu olarak Vakıflar ve Temlikler I- Istila Devirlerinin Kolonizatör Türk Dervişleri ve Zaviyeler’, Vakıflar Mecmuası, 2 (1942), 279–304. On Hungary see Burcu Özgüven, Osmanlı Macaristan’ında Kentler, Kaleler (Istanbul: Ege, 2001), 162–70; on Cyprus: eadem, ‘From the Ottoman province to the colony: late Ottoman educational buildings in Nicosia’, METU Journal of the Faculty of Architecture, 21 (2004), 33–66. (56) The policy of colonising frontier areas was not an Ottoman invention. It had long been practised by the Byzantine Empire. During the reign of Constantine Copronymous (741–75) colonies were brought from Anatolia and parts of Armenia to the Balkans, to reinforce the defences there. See George Ostrogorsky, ‘Byzantine cities in the Middle Ages’, Dumbarton Oaks Papers, 13 (1959), 62 ff. (57) On the nationalistic movements and revolts in the Balkans between 1804 and 1887 see Jelavich, History of the Balkans, 171 ff. For an eyewitness account of Ottoman Bosnia during the Croatian revolt in 1839–40 see Matija Mažuranić, A Glance into Ottoman Bosnia or a Short Journey into that Land by a Native in 1839–40, trans. Branka Magas (London: The Bosnian Institute, 2007). (58) Cevdet, Tezakir, 13–20: 189–90,…bir aralık İşkodra sancağının bir kıt’a haritasını tanzim ettirerek ahval-i coğrafyasını ve Karadağ hududuna dair malumat-ı lazimeyi…Babıaliye takdim eyledim. (59) Ibid., 18–19: 168, 213–14. (60) Tuĝlacı, Osmanlı Şehirleri, 328; Ayverdi, Avrupa’da, 3: 59 ff. Recent work in the Ottoman archives has discovered masses of documents pertaining to the construction of new churches in the Ottoman Balkans, churches that were partly or wholly financed by the Ottoman state, as well as letters from the church authorities thanking the sultan. (61) Cevdet, Tezakir, 21–39: 17–18.

Page 14 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Palanka Forts and Construction Activity in the Late Ottoman Balkans (62) Tuĝlacı, Osmanlı, 326. (63) Ibid., 213–14. (64) For the soldiers who joined the provincial troops as eyalet askeri: Yaşar Yücel and Ali Sevim, Türkiye Tarihi, IV: Osmanlı Dönemi (1730–1861) (Ankara: TTK, 1992), 97–8. For voynuks, vlachs and derbendcis see Machiel Kiel, Art and Society of Bulgaria in the Turkish Period (Assen/ Maastricht: Van Gorcum, 1985), 74–7, 86–101. (65) On the Castle of Alexandre: Evliya, Seyahatname, 6: 56. (66) Cevdet, Tezakir, 13–20: 185, 186. (67) Ibid., 13–20: 176. (68) Ibid., 13–20: 275.

Access brought to you by:

Page 15 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey LUCIENNE THYS-ŞENOCAK RAHMI NURHAN ÇELİK ARZU ÖZSAVAŞÇI GÜLSÜN TANYELİ

DOI:10.5871/bacad/9780197264423.003.0009

Abstract and Keywords The Ottoman fortress of Seddülbahir on the European shores of the Dardanelles and Kumkale, its sister fortress on the opposite side of the Straits, were both built in 1658 by Hadice Turhan Sultan, the queen mother or valide sultan of Sultan Mehmed IV. The Seddülbahir restoration project illustrates that the type of information that can be extracted from the Ottoman building and repair records is invaluable for guiding decisions concerning potential excavation sites. Along with the non-invasive techniques that are increasingly a part of pre-excavation archaeological planning, a thorough investigation of the extant physical remains, and the visual records provided in engravings and other representational sources, an examination of the building and repair records in the Ottoman archives should be standard methodological practice for any Ottoman era archaeological or restoration project. Keywords:   Seddülbahir restoration, pre-excavation planning, Ottoman archives, Hadice Turhan Sultan

THE OTTOMAN FORTRESS OF SEDDÜLBAHİR on the European shores of the Dardanelles and Kumkale, its sister fortress on the opposite side of the Straits, were both built in 1658 by Hadice Turhan Sultan, the queen mother or valide sultan of Sultan Mehmed IV (Figure 9.1). Located at the entrance to the famed Hellespont, the construction of these two fortresses served to demarcate the Ottoman possession of the Aegean frontier lands of European Thrace and the Asiatic Troad, and signalled a reassertion of the Sublime Porte’s control of the vital waterway that led into the empire’s capital in Istanbul. It is acknowledged by scholars who work on Ottoman sea frontiers that while the ‘borders and the frontiers in the eastern Ottoman Mediterranean were porous, shadowy, and uncertain’1 and a ‘frontier hovered along every

Page 1 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey expanse of coastline and encircled every island in the eastern Mediterranean’, the fortress, for the Ottomans as well as their enemies, served as a tangible symbol of territorial control.2 (p.190) (p.191) Prior to the seventeenth century, the Ottomans had fought the Venetians repeatedly, beginning in 1463 with Mehmed II’s campaigns to take Crete and other island strongholds.3 A renewed threat to Ottoman hegemony of the eastern Mediterranean came from the Venetians in 1645 during the reign of Mehmed IV’s father Sultan Ibrahim, and continued until 1669 when the Ottomans captured Candia and the island of Crete. The fortresses of Seddülbahir and Kumkale were completed during the course of this second campaign against Venice.

Figure 9.1. In the foreground are Turhan Sultan’s fortresses of Seddülbahir (Chateau Neuf d’Europe) and Kumkale (Chateau Neufd’Asie), Veue des Dardanelles de Constantinople, Nicholas de Fer (map maker) copperplate, 23 × 34 cm, late seventeenth century. Engraving property of Lucienne Thys-Şenocak.

The vakfiye, or foundation deed, in which Turhan Sultan’s fortresses are described and endowed, provides information about the types of structures which were to be built within the fortification walls, as well as the reasons why Seddülbahir and Kumkale were constructed. According to this document, now in the Süleymaniye Library in Istanbul, Turhan Sultan’s intent was to provide protection to devout Muslim pilgrims and other Ottoman subjects travelling by sea. The problems were caused by the accursed infidels who habitually come up from the shores of the Mediterranean in their ships and drop anchor in the place known as Old Istanbul [Alexandria Troas], outside Boğazhisar [Kumkale], and wait there and hinder pilgrims and others wishing to go to Egypt and other countries via the sea route, preventing their passage, causing them losses, treating them cruelly, repeatedly assaulting and looting the vessels of the Muslims, torturing them, causing them to drown in a sea of loss, enslaving the Muslims they capture, and burdening them with chains.4 To remedy the situation the valide, in addition to endowing the fortresses themselves, ‘ordered a grand mosque to be built for the fortresses’ guards and wardens within each fortress along with a fine school, a clean public bath and numerous houses, shelters, shops and markets unlike those found anywhere else’.5 Many contemporary Ottoman chroniclers also reported that there was a need for more security on the Aegean frontier. In 1659, shortly after the fortresses were completed, the scribe and boon companion to Mehmed IV, Abdurrahman Pasha, wrote the poem Tarih-i Kale-i Seddülbahir to commemorate a state visit that was made by the sultan and his mother to inspect the fortresses and to celebrate the patronage undertaken by Turhan Sultan on the Dardanelles. He describes the decision made by the valide to construct the fortresses of Seddülbahir and Kumkale as

Page 2 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey monuments which would restore the Ottoman claim to the western frontier zone in the Aegean and secure the entrance to the Dardanelles. (p.192) Behold the image of the Ottoman dynasty They continue to labour hard to do charitable work Especially the mother of Sultan Mehmed Han the gazi That is to say, the valide sultan, the source of benevolence and charity. With sincerity, she intended [to build] this work of charity. From the saints came spiritual guidance. Behold the charitable work she has undertaken, look at these two fortresses, Each of them an honourable guard of the west [Aegean border of the empire]. By building these two fortresses on both sides [of the Dardanelles], The villages of the believers are made secure from the enemies.6 Evliya Çelebi, who claimed to have been at both of Turhan Sultan’s new fortresses during the inaugural visit, concentrated his observations upon Kumkale, which he referred to as Sultaniye. In his description of the visit he recounted both the difficulties that the Ottomans were having prior to the construction of Turhan Sultan’s new fortresses, and how the problem had been solved by the valide’s patronage: The ships of the enemies would encounter the Ottoman navy and defeat them. This happened repeatedly, and our navy was defeated; the valide sultan had the fortress of Sultaniye [Kumkale] built and had large cannons placed on it. Thereafter the enemy’s navy could not pass through the straits of the Dardanelles and the Ottoman navy could travel to whichever city or island it chose.7 From the account of the early eighteenth-century Ottoman chronicler Naima, it is clear that by the mid-seventeenth century, control of the Dardanelles and this part of the Aegean frontier of the Ottoman Empire was under serious threat; the existing fortresses, Kilitbahir and Kale-i Sultaniye, which had been constructed in the 1470s by Mehmed II 30 kilometres inside the Dardanelles, were now seen as inadequate and in need of reinforcement. New fortresses at the entrance to the Dardanelles had to be built to signal to those who ventured into these waters, from the moment they entered the straits, that the Ottomans had complete control and sovereignty over this important maritime corridor. As Naima explains, The miserable infidels [Venetians] would arrive with thirty to forty ships and take refuge around the two aforementioned fortresses [Kilitbahir and Kale-i Sultaniye in Çanakkale]. It would be impossible to take a ship out of the city, and the ships coming into the city would be caught unawares and could easily be targeted.8 (p.193) While Ottoman chroniclers and the archival record show that Turhan Sultan’s imperial project to build two new fortresses at the entrance to the Dardanelles was driven by an urgent and practical need to bolster the Ottoman presence on the Aegean frontier, it is also important to consider the attraction that building on the famed straits of the Dardanelles may have had for Turhan Sultan as her first act of architectural patronage as valide. As a main artery connecting the Mediterranean Sea to Istanbul, control over the Dardanelles had been vital for the Ottomans and their predecessors. Ancient heroes such as Alexander the Great, the Byzantine Emperor Justinian, and illustrious Ottoman sultans such as Orhan, Bayezid, Mehmed the Conqueror and Süleyman the Magnificent had all worked to consolidate dynastic lands on either side of the Dardanelles. Often they had accomplished this by constructing fortifications.9 By the Page 3 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey seventeenth century a long line of Ottoman sultans had patronised military architecture along the Dardanelles. Most notable among them was Mehmed II, whose project to unite Ottoman territory from the Black Sea to the Mediterranean by the construction of fortresses was recorded by the fifteenth-century chronicler Kritovoulos: ‘He [Mehmed II] did this by placing fortresses like gates on either side [of the Straits] and by planting stone-shooting cannon.’10 Both the strategic need to secure the vulnerable Aegean frontier of the Ottoman Empire and the attraction of participating in the noble Ottoman tradition of sultanic patronage of fortifications along the Dardanelles were factors in the location of Turhan Sultan’s fortresses at Seddülbahir and Kumkale. As icons of imperial power, the fortresses and their location on the Dardanelles brought additional prestige to the valide as they tied her and her son to a long legacy of earlier conquerors from both the mythical and actual past who had conducted campaigns or built fortifications along the shores of this famed waterway.11 By undertaking this fortress-building campaign at the Dardanelles, and shoring up the defences that had been constructed here earlier by her son’s renowned ancestors, Turhan Sultan’s act of patronage would not only protect Ottoman subjects and support the empire’s navy, but confirm the valide’s newly acquired status as mother of the reigning sultan. Turhan Sultan would increase her power and political legitimacy by casting herself as a defender of the empire, a role which had been assigned traditionally to the sultan.12

(p.194) The Survey and Restoration Projects at the Fortress of Seddülbahir Since the restoration project at the Ottoman fortress of Seddülbahir was initiated in 2005,13 several areas of the upper level of the fortress have been excavated, and the geodetic surveying of the entire site has been completed along with the post-survey processing of a comprehensive 3D laser scanning of all the extant walls and buildings that comprise the fortification. Additionally, we have created restitution plans for three different eras of the fortress’s past, and finalised the restoration proposal for Seddülbahir. One of the aims of the project was to correlate archival material with the survey data so that physical changes at the site and the structures at Seddülbahir could be better understood. We hoped that by bringing together these different types of resources we could make more informed decisions about where to excavate, ensure greater accuracy in the restitution drawings done for the different historical phases of the site, and perfect the plans we were to submit to the Monument Preservation Board in Turkey for the restoration of the fortress.14 Architectural historians have for several decades been examining the archival records that were generated from the process of commissioning, constructing and repairing works of architecture in the Ottoman Empire.15 Scholars involved in restoration projects have also long since been aware of the potential of archival records to shed light on a wide range of restoration questions, from mortar composition to the origins of building materials such as wood and stone.16 Archival records related to the (p.195) building and repair processes have played a central role in recent investigations of the Ottoman Black Sea fortresses of Akkerman and Özi. By examining the so-called ‘new style Ottoman building registers’, questions about the physical changes of the extant structures and the archaeology of the sites have been addressed, and an attempt has been made to clarify the techniques and terminology used by the Ottomans for military architecture.17 Not only can these records shed light on the evolution of the physical aspects of the historical structure as it was adapted to changing needs, building techniques, and military and/or natural threats over time, they can also give us clues as to how effectively the central administration of Page 4 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey the empire was able to organise and carry out complicated projects such as the construction and repair of fortresses on its frontiers. For example, in the sixteenth century, under the direction of the head royal architect, Mimar Sinan, the corps of royal Ottoman architects had been highly centralised.18 According to Necipoğlu, the corps, based ‘in Istanbul, became an empire-wide network with tentacles reaching out into distant provinces’.19 Necipoğlu cites several examples during the sixteenth century of architects moving between Istanbul and the border regions of the empire bringing plans of fortresses or cost estimates with them.20 The practice of appointing provincial city architects to supervise projects far from the capital had begun in the sixteenth century and continued through the seventeenth, resulting in the increasing autonomy of projects and a gradual decentralisation of the corps of royal architects.21 But from the large number of repair records in the ‘new building registers’ from the seventeenth and eighteenth centuries that show how information continued to circulate between Istanbul and key frontier regions such as the Dardanelles and the Black Sea coast, it appears that decentralisation was a very gradual process. By March 2005, when planning for the restoration project at Seddülbahir began, the team had collected an extensive array of historical documentation from many different archives. Examined in the following pages are some examples of how these archival records, particularly the Ottoman building and repair documents, informed (p.196) decisions about the excavation, restitution project and restoration planning process at Seddülbahir.22

Representational Records Engravings As an example of early modern fortification architecture Turhan Sultan’s European castle is not as impressive as other fortifications along the Dardanelles, such as Mehmed II’s Kilitbahir fortress at the Çanakkale Straits with its unusual clover-leaf plan.23 But because Seddülbahir, along with Kumkale, directly across the Straits, were the first structures encountered by the traveller when entering the Dardanelles, they were frequently recorded in drawings or engravings by those en route to Istanbul. In the early modern pre-photographic era, depictions of Seddülbahir by European travellers were often highly schematic and impressionistic, but they do give some indication of the plan of the fortress and the types of structures that were located within its walls. The information that can be gleaned from a late seventeenth-century drawing of Seddülbahir, now in the Correr Museum in Venice, is a good example (Figure 9.2). Depicted here are the original and still extant levels of the fortress, the upper and lower seaside divisions of Seddülbahir. Seven towers are shown in the drawing, and while only five of these now remain, the round and polygonal plans of the towers along the south sea wall and west wall, respectively, are shown in the drawing—albeit in a very basic way—but as they are today, although now in a much poorer state of repair. The small mosque with its single minaret situated above a division between the upper and lower fortress is no longer extant, but its existence in the late seventeenth century is again confirmed in the 1680 engraving by the French cartographer Nicolas de Fer in his Veue des Dardanelles de Constantinople which illustrated one of the editions of travel narratives by Grelot (see Figure 9.1). In the later English edition of this book, Grelot describes the interior of Seddülbahir as having certain houses of the Agha and other officers, with a mosque of which the Duomo and Steeple appear very plain to be seen, as well as other edifices, as being generally seated in the highest parts of the fortification, from whence you descend by large steps to the

Page 5 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey platforms from where the guns are planted, which lie equal with the surface of the water.24 (p.197) De Fer’s engraving places the mosque closer to the south wall of the upper fortress’s interior than the draughtsman of the Correr Museum engraving had, but the former artist had also placed this structure in the upper fortress adjacent to several other unidentifiable buildings. From these early European sources we begin to get at least a basic idea of the existence and general location of some of the interior architecture once within the walls of Seddülbahir. In the early eighteenth century another engraving of Seddülbahir, in Tournefort’s travel account, provides a view of the curtain walls of Seddülbahir, the prominent circular seaside towers and several buildings in the interior of the fortification (Figure 9.3).25 In this engraving the mosque is again located in the upper fortress but appears to be (p.198) situated in a central bailey between the upper and the lower

Figure 9.2. Drawing of Seddülbahir in Museo Correr, MS Cicogna 1971 (MCCCXLVIII) Venice, 47.5 × 28 cm, fol. 210.

fortress.26 Seddülbahir appears in other engravings and plans from the eighteenth and nineteenth centuries located in various European archives; notable is a plan dated 1 October 1895 produced by the French military officers responsible for the renovation of Seddülbahir. This plan shows how certain modifications were intended to be made to the fortress’s plan, such as the two additional towers along the western wall, but which, from the present-day remains of this wall, we know were never implemented (Figure 9.4). The French plan of Seddülbahir serves as a reminder that these kinds of architectural records, generated by engineers or architects who were commissioned by the sultan to modernise the Ottoman defences, may reflect what was extant at the site at that time; but they may also be records of unrealised plans for a structure, as was the case of the extra west wall towers at Seddülbahir. (p.199)

Figure 9.3. Engraving of Seddülbahir in J. P. de Tournefort, Relation d’un voyage du Levant (Paris, 1717). Engraving property of Lucienne Thys-Şenocak.

Page 6 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey Historical Photographs Historical photographs of Seddülbahir and Kumkale have also proved to be very useful archival resources for the restitution work conducted during the project.27 By the late Figure 9.4. French plan showing proposed changes to be made at the fortress of nineteenth century, the centralised Seddülbahir. Fortifications des Dardanelles, administrative mechanisms of the empire côte d’Europe. Article 14 P. E. Turquie, which had organised, recorded and carton 3, 1 Octobre, 1895. Service generated documentation about the repairs Historique de l’Armée de Terre, Château de Vincennes, France. and building activities conducted at its frontier fortresses were deteriorating. As a result, there are fewer building and repair records to be found in the archives as we move into these later centuries. By the late nineteenth and early twentieth centuries the photographic image becomes increasingly valuable to understand certain aspects of the later phase of (p.200) Seddülbahir’s past. Among the most important historic photographs of the two fortresses are the images that appear in a large photograph album created by Sultan Abdülhamid II between 1880 and 1893.28 A photograph of the south-west and southeast sea walls of the Seddülbahir fortress, taken from the entrance to the Dardanelles, appears in this album (Figure 9.5); the bunkers in the north section of Kumkale were also included (Figure 9.6). Perhaps also intended for the album, but never included in any of the extant copies, was a photograph of the façade of a large stone military barracks in the upper section of the Seddülbahir fortress (Figure 9.7). The entire west wall of Seddülbahir was well documented in another late Ottoman photograph from the nineteenth century and shows the three main towers that stretched along this side of the fortress, still complete at that time with crenellations (Figure 9.8). (p.201)

Figure 9.5. View of south section of Seddülbahir as seen from the entrance to the Dardanelles. From the photograph album of Sultan Abdülhamid II (Istanbul Naval Museum).

Figure 9.6. View of the bunkers in the north section of Kumkale. From the photograph album of Sultan Abdülhamid II (Istanbul Naval Museum).

Page 7 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey (p.202) As Seddülbahir was the site of one of the major Allied landings on 25 and 26 April 1915 during the Gallipoli campaign, and was then used by the French troops for the duration of the Allied occupation of the peninsula, there are numerous photographic records available of various sections of the fortress after it was heavily bombarded. Images of Seddülbahir’s west wall and towers allow us to see the extent of the damage that happened to these areas in 1915 and serve as some of our most important documentation for restitution work in this section of the fortress (Figure 9.9).

Figure 9.7. The newly completed nineteenth-century Ottoman officer barracks in the upper interior section of Seddülbahir prior to its destruction in the First World War. Photograph property of Lucienne ThysŞenocak.

Archival Documents To date no Ottoman architectural plans have come to light for Seddülbahir. There are archival sources from the sixteenth century which mention that architectural plans (resm) were being used for the construction of fortresses, but none of these actual plans remain.29 Whether they were lost when the imperial architectural workshops and palace archives in Vefa disappeared, or were destroyed intentionally after the completion

Figure 9.8. Pre-First World War view of the long west wall of Seddülbahir. From the Fahreddin Pasha collection, FP01 2916, IRCICA, Istanbul.

of the work to be executed, is difficult to ascertain at the present time.30 But the resulting absence of building plans for most works of Ottoman architecture, including fortresses, makes the use of other types of archival records about this architecture even more essential. (p.203) Much more detailed information about the cost of labour, procurement and transportation procedures, building materials brought from the forests and quarries for oak beams and limestone-based mortar, as well as the names, general locations and basic functions of different parts of the fortresses can be extracted from the building and repair records for

Figure 9.9. This photograph shows the results of the bombardment that occurred at the West Tower (referred to in eighteenthcentury Ottoman records as the Başkule, or Principal Tower) during the Allied attacks of April 1915. Photograph courtesy of the Imperial War Museum, London (Q 13233).

Turhan Sultan’s fortresses.31 The information from these types of documents was important for the restitution plans we created to illustrate the major phases of architectural change that occurred at Seddülbahir.

The information in the building and repair records for Seddülbahir covers roughly two and a half centuries of building and repair activities at the fortress, from the mideighteenth to the early twentieth century. This allowed us to start mapping, both in time and at the actual site, various components of the Seddülbahir fortress, particularly the locations of the towers and gates. In

Page 8 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey several cases we were also able to determine the contemporary Ottoman names that had been used for different components of the fortress. Eventually this allowed us to understand the changing functions of many parts of the fortress through the centuries. For example, the South Tower at Seddülbahir, the name we had initially chosen for this section of the fortress because of its geographical location, was referred to in the Ottoman records from 1764 to 1784 as the Tophane or Cannon Arsenal Tower. Housed in the complex of rooms that was attached to this tower’s south-west side was an ammunition depot and two watch towers, one made of stone and the other of wood. Neither the exact location nor dimensions of these additions can be determined from the archival sources, but we were able to confirm their existence and general location, as indicated in our restitution plan for the eighteenth century (Figure 9.10). In the Ottoman records that requested repairs to be made in October 1766 to the fortress after a series of earthquakes which began on 22 May and continued until 7 August 1766, we were able to determine the names of several other sections of the fortress: the south tower at the water’s edge was the Lodos or South Wind Tower; the east tower near the harbour was referred to as the Poyraz Limanı Kulesi, or the Northeast Wind Harbour Tower; the tower closest to the double bath was referred to as the Hamam Kulesi, or the Bath Tower; the north-west middle tower was the Bab-ı Sagir Tower, or the Secondary Entrance Tower; the west tower was the Baş or Principal Tower.32 Outside the walls of Seddülbahir, 26.5 metres to the north of Seddülbahir’s Hamam Kulesi are the remains of a double-domed hamam, which date from the time of the fortress’s initial period of construction and is mentioned in Turhan Sultan’s foundation charter. From the repair records we have found to date, we know that this structure was repaired in 1759 and again in 1838, 1859 and 1872, so it is clear that the bath was in use at least until the late nineteenth century.33 (p.204) (p.205) By using building and repair documents we were able to determine the dates of destruction and often the original functions of many sections of the fortress which have now disappeared. For example, we learned that the main entrance to the fortress in the mid-eighteenth century was referred to as the Bab-ı Kebir, the Great Gate.34 It was located along the north-east wall of the fortress and had three vaulted chambers which housed an armoury and prison; today none of this tower complex remains. Likewise, the Cezayir, or Algerian, Tower, which existed on the south-west wall of Seddülbahir is first mentioned in the late eighteenth-century repair records and continued to be an area where building and repairs were undertaken at least until 1814, the last time that this tower is mentioned in the archival records we have found.35 Also in the upper fortress, by the late eighteenth century, accommodation was built for the

Figure 9.10. Restitution drawing of Seddülbahir for the late eighteenth-century phase of the fortress, showing the location and scale of structures reconstructed from

Page 9 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey notables in charge of administering the fortress, the beylik and dizdar ağalar, as

the restitution data. (M2 refers to metres squared.)

well as a school or sıbyan mektebi.36 Much later, in the 14 December 1834 repair documents, we learn of a plan to build a military barracks between the mosque and the west middle tower.37 While we do not know the exact location of this earlier military barracks, it is very likely that the J14 sondage of the 2005 archaeological excavation revealed one corner of this barracks and that the later Abdülhamid II period barracks was built on top of this earlier barracks (see Figure 9.7). Our future excavations will be now be concentrated in this region of the upper fortress, as we hope to reveal more of the earlier barracks that was discovered through the archival record. A comparison of the earliest representational documentation from the late seventeenth-century European sources discussed above with the extant remains of (p.206) the actual structure today suggests that the basic plan of the Seddülbahir fortress does not seem to have changed significantly since its foundation. However, the archival records reveal a much more complex picture of how, over the centuries, and since the initial construction of Seddülbahir, several sections of the outer walls, the gates, and the interior structures have been destroyed, repaired and rebuilt, and given different names and functions. Today the only interior buildings in the upper fortress are the concrete abandoned shells of barracks built in the 1960s by the Turkish military.38 The existence of a mosque in the upper fortress, a view of which first appears in the aforementioned Venetian drawing dating from the late seventeenth century and in other European engravings, is confirmed by a 1759 request for the repair to the roof of the cami-i şerif using new lead.39 Not only do we learn from this document that there was still a mosque functioning in the interior of the fortress at this time, but also that it had a sloped rather than domed roof. Much later, in 1893, a decision was made to construct a hospital just outside the fortress walls because the mosque in the fortress was being used as a place for the medical treatment of soldiers and was proving to be an inadequate space.40 From this document, as well as recollections by the Seddülbahir village elders gathered during our oral history research, we were able to identify the location and identity of the late Ottoman era hospital that appears in a photograph of the hospital ruins from the Imperial War Museum (Figure 9.11). By examining the building and repair records for Seddülbahir and correlating the dates of repairs with events which could have had destructive effects upon the fortress, such as natural disasters or wars, we were also frequently able to determine the reasons why repairs and reinforcements were conducted at Seddülbahir.41 In this respect it was interesting to note that most repairs made to Seddülbahir in the eighteenth and nineteenth centuries seem to have been prompted by a response to earthquake and weather damage rather than bombardment or any increasing pressure of foreign naval forces in the western region of the Ottoman Empire.42 The surprise appearance of the Russian navy and the damage caused to the Ottoman fleet at Çeşme in the Aegean on 6 July 1770 seems to have prompted little response from the Ottoman administration to refortify the Dardanelles fortresses, as there are no major repairs undertaken at either Seddülbahir or Kumkale from 1770 to 1784, almost fifteen years later.43

Page 10 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey (p.207) Conclusion

Figure 9.11. The ruins of the military The Seddülbahir restoration project hospital just outside the north-west wall of the Seddülbahir fortress. Photograph illustrates that the type of information that courtesy of the Imperial War Museum, can be extracted from the Ottoman building London. and repair records is invaluable for guiding decisions concerning potential excavation sites. Along with the non-invasive techniques that are increasingly a part of pre-excavation archaeological planning, a thorough investigation of the extant physical remains, and the visual records provided in engravings and other representational sources, an examination of the building and repair records in the Ottoman archives should be standard methodological practice for any Ottoman era archaeological or restoration project. While building and repair records from the Ottoman archives often provide information about the cost, quantity and intended location of repairs or building that was conducted at a site, the more specific architectural information about the dimensions of structures, measured plans, elevations, interior and exterior decoration—information which is essential before restoration— is not always possible to extract from these types of historical records. However, these records are invaluable for the restitution work that must accompany a restoration project. They remind us that these buildings have long and diverse pasts, and that the life of a building is always a complex phenomenon and ongoing process. When a decision is made to intervene in a structure by restoring it, even when there is highly accurate information supporting this decision and sound justification for doing so, there is also, at that moment, a choice made which has (p.208) inevitably, by the very nature of the physical intervention required in the act of restoration, prioritised a particular phase of that structure’s past over another moment in time. A rich restitution study, which reflects a thorough analysis and appreciation of the different eras of a building’s past is one of the most important ways that we can both preserve and communicate the complexity of that past. With the rapid development of 3D visualisation techniques, the possibilities for displaying the multiple pasts of a historical structure are increasing, thereby decreasing public pressure for reconstruction.44 The value of the Ottoman building and repair records for the restoration of Ottoman structures is in no way mitigated by the fact that the results of the research may not always lead to precise architectural information. As we have found at Seddülbahir, there are surprises and discoveries to be made in these documents, such as buildings which would have never been found, and functions which would have remained unknown had we not unearthed them first in the archives. Notes: Proceedings of the British Academy 156, 189–208. © The British Academy 2009. (1) Virginia Aksan and Daniel Goffman, ‘Introduction: situating the early modern Ottoman world’, in Virginia H. Aksan and Daniel Goffman (eds.), The Early Modern Ottomans: Remapping the Empire (Cambridge: Cambridge University Press, 2007), 7; see also Oriente Moderno, n.s. 20 (2001) (Special Issue, The Ottomans and the Sea, ed. Kate Fleet). (2) Molly Greene, ‘The Ottomans in the Mediterranean’, in Aksan and Goffman (eds.), The Early Modern Ottomans, 113; see also Chapter 2 in this volume by Palmira Brummett and eadem, ‘Imagining the early Ottoman space: from world history to Piri Reis’, in Aksan and Goffman (eds.), The Early Modern Ottomans, 15–58. Brummett (ibid., 25) states that the Ottoman administrative documents, tax records and oral testimony were used to determine boundaries,

Page 11 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey many of which were designated according to natural features of the landscape or extant structures such as fortresses: ‘Territory was described primarily in terms of castles and the agricultural lands attached to them.’ (3) Following that war Bayezid II also faced the Venetians as a naval power from 1499 to 1503; later, Süleyman I fought with the Venetians in this region from 1537 to 1540. (4) Süleymaniye Kütüphanesi, Istanbul, MS 150, fols. 13b-14a. (5) Ibid., fol. 15a. (6) TSA, E.2477. For the entire Ottoman text of this poem see Lucienne Thys-Şenocak, Ottoman Women Builders: The Architectural Patronage of Hadice Turhan Sultan (Aldershot, Ashgate Press, 2007), 111–13 and Appendix 4.1. (7) Evliya Çelebi, Seyahatname, 5: 157. (8) Naima, Tarih-i Naima (Istanbul: Matbaa-i Amire, 1864–6), 6: 408. (9) See Halil İnalcık, ‘Gelibolu’, EI2, 2: 983–7; Besim Darkot, ‘Çanakkale’, İA, 3: 331–51; H. Burcu Özgüven, ‘Barut ve Tabya: Rönesans Mimarsı Bağlamında Fatih Sultan Mehmed Kaleleri’ (Unpublished Ph.D. dissertation, İstanbul Teknik Üniversitesi, Sosyal Bilimler Enstitüsü, 1997) for information on the history of fortifications along the shores of the Dardanelles prior to the seventeenth century. (10) Kritovoulos, History of Mehmed the Conqueror, trans. C. T. Riggs (Princeton: Princeton University Press, 1954), 197–8. (11) Thys-Şenocak, Ottoman Women Builders, 125, 180. (12) Mehmed II, along with many other sultans, is referred to as gazi, or warrior, of Islam. See Julian Raby, ‘El Gran Turco: Mehmed the Conqueror as a Patron of the Arts of Christendom’ (Unpublished Ph.D. dissertation, Oxford University, 1980), 203, n. 69. (13) The project is jointly directed by Rahmi Nurhan Çelik and Lucienne Thys-Şenocak. Supported by the Vehbi Koç Foundation and administered by the Turkish Forestry Ministry, the restoration project at Seddülbahir builds upon a series of earlier survey and documentation campaigns that the team conducted from 1997 to 2002 at both Seddülbahir and Kumkale. Information about these earlier survey projects and the current restoration project can be found on the project website, www.seddulbahir-kumkale.org. (14) We follow here the definitions of the terms ‘conservation’ and ‘restoration’ used in the International Charter for the Conservation and Restoration of Monuments and Sites (The Venice Charter, 1964) (see articles 4–13 for the full definitions of conservation and restoration: www.international.icomos.org/charters.pdf). The term restoration is used here to refer to the entire process, which includes restitution studies as well as conservation. The term ‘restitution’ is not specified in the Venice Charter but can be defined as the process by which architectural drawings are generated which show the hypothetical reconstruction of a site, or various historical features of a site that may no longer exist. These drawings are generally based upon archival, archaeological, and oral or other types of research, and scientific evidence. The important point is that restitution drawings are generally intended as conceptual and

Page 12 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey hypothetical and are frequently used to show the different historical phases that a structure has gone through over time. In most cases they should not be considered as guidelines for more invasive restoration activities. (15) See, for example, M. Erdoğan, ‘Osmanlı Mimari Tarihinin Arşiv Kaynakları’, Tarih Dergisi, 3 (1953), 95–122; idem, ‘Osmanlı Mimarisi Tarihinin Otantik Yazma Kaynakları’, Vakıflar Dergisi, 6 (1965), 111–37, and Ömer Lutfi Barkan, whose seminal work on the construction of the Süleymaniye mosque complex in Istanbul provided valuable insights into several aspects of the organisation of construction, including labour costs and materials: Ömer Lütfi Barkan, Süleymaniye Cami ve İmareti İnşaatı (1550–1557) (Ankara: TTK, 1972). (16) See, for example, Z. (Nayır) Ahunbay, Osmanlı Mimarlığında Sultan Ahmed Külliyesi ve Sonrası (1609–1690) Istanbul: İstanbul Teknik Üniversitesi Mimarlık Fakültesi, 1975) and D. Mazlum, ‘Fethiye Camii’nin 18. Yüzyıl Onarımları’, Sanat Tarihi Defter 8: Metin Ahunbay’a Armağan Bizans Mimárısı Üzerine Yazılar (Istanbul: Ece Yayınları, 2004), 167–83. (17) See Chapter 7 above, Victor Ostapchuk and Svitlana Bilyayeva, ‘The Ottoman northern Black Sea frontier at Akkerman fortress: the view from a historical and archaeological project’, and Caroline Finkel and Victor Ostapchuk, ‘Outpost of empire: an appraisal of Ottoman building registers as sources for the archaeology and construction history of the Black Sea fortress of Özi’, Muqarnas, 22 (2005), 150–88. The term ‘new building records’ describes a variety of archival records: expenditure books (masraf defterleri), account books (muhasebe defterleri), construction books (inşaat defterleri), repair records (tamirat defterleri), appraisal records (keşf defterleri), appraisal and repair books (keşfve tamirat defterleri), see ibid., 154. (18) For a detailed discussion of the changing role of the Ottoman corps of royal architects (hassa mimarları) see Gülrü Necipoğlu, The Age of Sinan: The Architectural Culture in the Ottoman Empire (London: Reaktion Books, 2005), 153–86. (19) Ibid., 160. (20) Ibid., 166–7. (21) Ibid., 160–1. (22) There is also a rich oral history of Seddülbahir and Kumkale, which we discovered during the two projects we conducted in these villages as part of the larger documentation work. These will not be discussed here but are published in I. C. Cenker and L. Thys-Şenocak, ‘Moving beyond the walls: local and national memories of Turkish fortresses’, in Paula Hamilton and Linda Shopes (eds.), Oral History/Public Memories (Philadelphia, Pa.: Temple University Press, 2008), 65–86. (23) See Raby, ‘El Gran Turco’ and Özgüven, ‘Barut ve Tabya’ for more information on the Dardanelles fortifications built by Mehmed II. (24) William (Guillaume) Grelot, A Late Voyage to Constantinople (London: J. Playford, 1683), 15, 16. (25) J. P. de Tournefort, Relation d’un voyage du Levant (Paris, 1717).

Page 13 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey (26) This is a particularly interesting depiction of the location of the mosque because in the repair records from 14 December 1834 there is mention of a section in the fortress called the central fortress or orta hisar. The location of the mosque is cited as being in this area of the fortress near the Sancak Kulesi, or the Flag Tower, so it does seem to support the visual information provided in the earlier engraving (Kale-i mezkur derununda Orta Hisar balasında olan cami’-i şerif ittisalinde Sancak Kulesi kapusu…): BOA, Bab-ı Defteri Baş Muhasebe Bina Emini (D.BŞM.BNE) 16419, p. 11. (27) While archival photographs are very useful documents to have, it is not generally accepted practice to restore a structure based only upon a photographic image, as it is not always possible to determine from a photograph, particularly one from the nineteenth century when photogrammetric methods were not being employed, the accurate dimensions and plan of a building which no longer exists. Further, in the case of the photographs that exist of Seddülbahir from the late nineteenth century, there are no photographic records of the interior or other façades of the upper barracks, apart from the south façade. (28) For more information about this photograph album see William Allen, ‘The Abdülhamid II collection’, History of Photography, 8 (1984), 119–45; C. Gavin et al., ‘Imperial self-portrait: the Ottoman Empire as revealed in the Sultan Abdul Hamid II’s photograph album’, Journal of Turkish Studies, 12 (1988). (29) Necipoğlu (The Age of Sinan, 167) cites a few examples of these projects, among them the plans that Lala Mustafa Pasha sent to Istanbul after the 1570 conquest of Cyprus for the renovations at the castle of Baf. (30) Ibid., 171. For a detailed discussion of Ottoman building plans, see Gülrü Necipoğlu, ‘Plans and models in 15th and 16th century Ottoman architectural practice’, Journal of the Society for Architectural Historians, 45 (1986), 224–43. (31) The dates of the repair records that we have found so far for Seddülbahir range from 1759 to 1906. (32) BOA, Maliyeden Müdevver Defterleri (MAD) 3160, 608–11. (33) The first reference to this area is in the 1764 repair records BOA, MAD 3160, where there is mention of repairs to the Hamam Tabyası. Later, in the 1805 repair records (BOA, D.BŞM.BNE 16141) and those for 1814 (BOA, D.BŞM.BNE 16353), the name Hamam Kulesi, or Bath Tower, is used. In the repair records dating to 1759 both the mosque inside the fortress and the doubledomed bath underwent extensive repairs (BOA, Evkaf-ı Hümayun Muhasebe Defteri (EV.HMH.D) 1–1). The 1838 records are from BOA, İrade-i Dahiliye (İ.DH) 2905; the 1859 records are BOA, Sadaret Mektubi Kalemi Belgeleri (A.MKTMHM) nr. 28–45; the 1872 records are from BOA, Zabtiye Nezareti Evrakı (ZB) 4/76 (2 May 1872), nr. 4–76. (34) BOA, MAD 3160, 608–11. (35) BOA, D.BŞM.BNE 16034 from 18 December 1784 mentions the existence and need for the repair of the barbican and embrasures of the Cezayir Tabyası (Algerian Bastion). BOA, D.BŞM.BNE 16141, p. 9, dating to 5 April 1810, mentions a Cezayir Kulesi or Algerian Tower which was to have repairs made to the existing masonry.

Page 14 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Understanding Archaeology and Architecture through Archival Records: The Restoration Project of the Ottoman Fortress of Seddülbahir on the Gallipoli Peninsula of Turkey (36) BOA, MAD 3160, 608–11 (Derûn-ı kalede vâki‘Dizdâr Ağalara mahsûs Beylik ve Dizdâr Ağalar Konağı); BOA, D.BŞM.BNE 16034 mentions the sibyan mektebi (Derun-ı kalede talim-i sıbyana mahsus iki bab mekteb-i şerifler). (37) BOA, D.BŞM.BNE 16119, pp. 2, 9–12. (38) One of these abandoned military structures was rehabilitated by our team and used as a conservation laboratory and excavation house during the 2005 and 2006 seasons. (39) BOA, EV.HMH.D 1–1 /5192, fols. 1b-4b (28 July 1759). (40) BOA, İrade-i Tophane-i Amire (İTPH) 1. (41) Finkel and Ostapchuk, ‘Outpost of empire’, 15. (42) Apart from the repairs requested for Seddülbahir in December 1834 (BOA, D.BŞM.BNE 16419), which followed the signing of the Hunkar İskelesi Treaty between the Ottomans and Russia in July of that year, there are few examples of the Ottoman government responding immediately to military threats in the Dardanelles by repairing Turhan Sultan’s fortresses. In 1807 after the British forced the Dardanelles in response to suspicions regarding an alliance that Sultan Selim III had made with the French, there is a request to repair cannons at Seddülbahir and to erect new batteries along the shores of the Straits, not to repair the existing fortress: Hatt-ı Hümayun (HAT) 7598. (43) Caroline Finkel and Victor Ostapchuk (‘Outpost of empire’, 157) have commented upon how the repair and modernisation programme for the Dardanelles fortresses differed from that of the fortress of Özi on the northern Black Sea frontier of the empire, in that the efforts to repair the latter fortress reflected a more sustained commitment of the Ottomans to defending that part of the empire from Russian encroachment. (44) As an example of some of the research being conducted on this subject, see the project at Ename, Belgium: www.ename974.org.

Access brought to you by:

Page 15 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century TOM SINCLAIR

DOI:10.5871/bacad/9780197264423.003.0010

Abstract and Keywords This chapter compares the administration of the Lake Van region under the Ottomans and that of the same region in the late Middle Ages. This required a well-defined region whose boundaries would suggest a certain degree of internal coherence and a certain degree of isolation from surrounding regions; and mountain ridges seemed the best natural boundaries to take for this purpose. Defining such a region meant that one could analyse the manner in which successive overlords dealt with the same political entities on the ground, in this case the local Kurdish principalities. Given that the territory of the Safavid empire lay immediately adjacent to the east, this chapter examines the effect the position of the border exercises on the methods of control on the administration and the internal relationships in question. Keywords:   Kurdish principalities, Lake Van region, Safavid empire, Ottoman Empire, late Middle Ages

THE PROJECT OUT OF WHICH THIS CHAPTER ARISES is a comparison between the administration of the Lake Van region under the Ottomans and that of the same region in the late Middle Ages. In each case I wished to view the administration as something in motion: not merely what the relevant entities were, but how they interacted; not merely what the system was designed to be, but how it worked. This required a well-defined region whose boundaries would suggest a certain degree of internal coherence and a certain degree of isolation from surrounding regions; and mountain ridges seemed the best natural boundaries to take for this purpose. Defining such a region meant that one could analyse the manner in which successive overlords dealt with the same political entities on the ground, in this case the local Kurdish principalities. To the north the boundaries chosen run along a range of heights which delimit from the north the great plain of the Lower Euphrates: the river itself rises near the plain of Eleşkirt, from which it then flows, in a south-westerly direction, into the great plain north of the lake (Figure 10.1). To the east the range continues in the higher massifs of Ala Dag and Tendürek Dag; and

Page 1 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century here it makes contact with the north–south chain over which runs the present border ridge between Turkey and Iran. To the west the range overhanging the Lower Euphrates from the north includes the extinct volcano of Bingöl Dağ, with its renowned pastures; and further west the heights overlook the plain of Muş. A shoulder extends southwards to close off the west end of the plain of Muş and joins the main range of the Taurus. In this sector the latter consists of a series of rounded heights forming ideal summer pastures. The Taurus progresses in a south-easterly direction and passes to the south of Lake Van, broadening and rising in height so as eventually to become a series of mountains from which deep and sharply cut valleys descend towards the lake. Despite the incut nature of this sector of the Taurus, certain areas of even ground are to be found, such as the valley-plain of Hoşap to the south of the lake, the Albak region—the plain of Başkale—further east, and parts of the Zab valley to the southeast. The Taurus range now joins the border ridge, which forms the eastern boundary of the Van region. The border ridge was also the Ottoman–Safavid border, with the exception of one Safavid enclave west of the ridge, from the acknowledgement of Ottoman suzerainty by the Kurdish principalities in 1515 to 1578, when the Ottoman conquests in Azerbaijan and Georgia temporarily pushed the border further east. The Safavid enclave mentioned above was the Mahmudi summer grazing-grounds in the Erçek valley, east of the lake; the Mahmudi principality, as will be explained below, remained loyal to the Safavid empire until 1554. (p.212) (p.213) The border ridge, a clear watershed distinguished by its height above the plains extending westward as far as the lake and by its even greater height above the plains to the east (around Lake Urmia), consists mainly of a series of rounded heights over which tribes and their flocks could move quite easily in summer. Moreover, the several passages through the range, such as the Kotor corridor, do not require a significant ascent, some none at all, above the several plains lying between the border ridge and the lake. The border ridge, despite being quite clear as a geographical feature, was nevertheless relatively easy to cross, and so as a natural frontier was somewhat open.

Figure 10.1. The topography of the Lake Van region, shown in relation to modern political boundaries.

Our aim in this paper is to consider the Van region as an issue in Ottoman administration, but as an issue in administration inevitably influenced by the border. Given the region as defined, how was it controlled and administered; at the same time, what were the internal relationships? These are issues which could arise in any region, but given that the territory of the Safavid empire lay immediately adjacent to the east, what effect did the position of the border exercise on the methods of control (we shall emphasise particularly the role of fortification), on the administration and the internal relationships in question?

Page 2 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century The principalities lay in the mountains or indented territory to the south, west and east of the lake. The flatter land north of the lake was not occupied by any local political entity, with the exception of Hınıs (on which see below) and was always administered directly. The principalities were each based on an agglomeration of tribes, which would be welded into a coherent entity by a single tribe or tribal section. The prince was elected from within the family which led the paramount tribe or tribal section. Among the principalities three were pre-eminent. That of the Hakkari was centred on the upper Zab valley to the south-east of the lake. It included, however, the high valley-plain of Albak, which bordered with Iran, and the district of Vostan by the lake shore, south-west of the city of Van. That of Bitlis lay west and south-west of the lake. It was centred on Bitlis and the Bitlis pass, but embraced also territory east of the city among the outliers of the Taurus by the lake shore, the plain of Muş west of Bitlis, with adjacent pastures, and that of Hınıs further north, with pastures on Bingöl Dag and elsewhere. And that of the Mahmudi lay south of the lake, not far from (p.214) the city of Van, with summer grazinggrounds in flatter country east of the lake.1 These three apart, a number of small principalities lay in certain valleys south of the lake, between the Bitlis principality on the one hand and the Mahmudi and Hakkari on the other.2 The Kurdish principalities of the Van region switched their allegiance from the Safavid Empire, under which they had been restless and rebellious, to the Ottoman Empire in 1515. At this stage the princes were simply vassals: they had no obligation other than to recognise the Ottoman sultan as overlord and not to deal directly with any other potential overlord.3 There was one principality, however, which maintained its loyalty to the Safavid Empire until 1554, six years after the Ottoman capture of the city of Van and in the final year of the full Safavid campaigns in the Van region which had started in 1550. The Mahmudis’ refusal to change sides until 1554 is probably explicable by the proximity of the city of Van, which was in Safavid hands, it has to be remembered, until 1548, to the principality’s territory around Hoşap and by the proximity of its other grazing-grounds east of the lake both to Van and to the first solid stretch of Safavidcontrolled territory in Azerbaijan on the far side of the border ridge.4 The nature of the Ottomans’ regime in the Van region changed somewhat in 1535, when (if we simplify events to arrive at the truth) the reigning prince of Bitlis was expelled and crossed to the Safavid Empire, residing first in Isfahan, then in Qazvin. The territory of the principality was now incorporated fully into the sancak system. Villages were converted into timars, mostly for incoming Ottoman sipahis, but sometimes for tribal chiefs, in other words elements of the previous dispensation under the Bitlis prince. The territory was termed first an eyalet, then a sancak. The whole of it was put under one bey, an appointee from outside the region; the frequency of reappointment thereafter was high.5 In 1538 the district of Hınıs, to which belonged exceptional pastures on Bingöl Dag and elsewhere, was transferred out of the Bitlis sancak, which was then subject to the province of Diyarbakır, and into the province of Erzurum, where, combined with the district of Malazgirt in the great plain of the Euphrates, it became a sancak on its own. This would seem to have been a move (p.215) deliberately planned to interrupt the commuting patterns of tribes within the former principality. The effect would be to make inroads into the coherence of the principality’s tribal agglomeration and so to make it hard for the principality to be reconstituted.6 The loss of Hınıs apart, the territory was divided into two sancaks, those of Bitlis and Muş, probably at a date shortly before 1559–60, and certainly by 1564–5, in any case after the formation of the province of Van in 1548.7 These two measures, designed to prevent the reconstitution of the principality, were also undertaken on the premise that it would not be reconstituted. But as matters turned out, the sultan was forced in 1578 to

Page 3 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century recall the son of the prince expelled in 1535 in order to counter possible moves by Kurdish tribes during the Ottoman advance towards Nakhchevan.8 Turning to the other principalities in the Van region, matters changed across the board with the Ottoman capture of the city of Van in 1548 and the institution of the beylerbeyilik of Van immediately afterwards.9 This meant a change in status and responsibilities for the Kurdish principalities of the region, all of which were included in the new province. For the Ottoman government it meant the task of securing the province against Safavid attacks by military means and the finding of a formula to secure the continued loyalty of the Kurdish princes. In respect of the deployment of military defensive means, certain factors have to be remembered. Despite the capture of Van in 1548 the whole of the lake shore remained a theatre of war in the years 1550 to 1554 and certain districts to the east and north of the lake, particularly Erciş, remained essentially in Safavid hands until 1554.10 Even after the Treaty of Amasya between the Ottomans and Safavids in 1555, which assigned the region to the Ottomans, the lake shore was subject to incursions by Safavid forces. On the other hand, an important trade route led from Khoy and ultimately Tabriz, and from Julfa and Nakhchevan, over passes to Bargiri (now Muradiye) and along the north shore of the lake, then down the Bitlis pass to Amid, centre of the province of Diyarbakır, and ultimately to Aleppo. It appears to have been the Treaty of Amasya which allowed a sudden increase of movement, particularly of Persian silk, along this route.11 The east, north and west shores of the lake were now under direct Ottoman administration, and had been divided into a series of sancaks, all subject to the (p.216) beylerbeyi of Van. Each sancak capital was carefully fortified or had its fortifications repaired. Let us start with the city of Van itself. Here the walled city on the coastal plain was attached to the south side of the formidable citadel rock which rose directly out of the plain and ran for around one kilometre from end to end.12 Soon after the Ottoman capture of Van an eight-month campaign of reconstruction of the citadel and city walls was instituted. Gravestones were uprooted; animalloads of gold were brought to pay wages; grain was commandeered from other sancaks.13 The city’s Armenian inhabitants were made responsible for watching the gates and for clearing snow off the walls’ vulnerable inner face, which had not been clad in stone.14 A steady investment in mosques, medreses and other buildings by the beylerbeyis took place within the city walls.15 The old Urartian canal bringing water to the city was re-opened in 1561: this again required a fourmonth campaign in which 8,000 workers from the city and others from other districts were involved.16 In 1571 the walls’ inner face was at last clad in ashlar, relieving the Armenians of one of their duties.17 However, the Armenians, who lived outside the walls, were still insecure, and in 1582 paid for an extension of the walled area towards the lake (i.e. westwards); but this again required an order for workers from surrounding sancaks.18 Sulphur and saltpetre were refined, and gunpowder produced, in factories here. The sulphur and gunpowder were then sent, via the lake’s east shore, to Trabzon, and from there shipped to Istanbul.19 The hans within the walls were built mostly for the purposes of trade within the city; but a single caravanserai for merchants from elsewhere, and zaviyes (shelters for travellers, among other things) both in the city and in nearby villages, are known.20 The trade in question would have taken place along the Kotor corridor with Khoy and with cities (p.217) beyond, but this seems to have been a minor avenue of movement compared with the line along the lake’s north shore. The customs point was not here, but at Amid (modern Diyarbakır) in Upper Mesopotamia.21

Page 4 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century Coming now to this shore itself, the most easterly of the three fortified settlements maintained there by the Ottoman administration was Erciş. This lay on the shore plain, not far from the lake’s north-easterly tip. The small town, contracted since Ilkhanid days, was protected by a rectangular wall and deep ditch on a considerably reduced alignment.22 The flat ground made it a weak site, and by this stage the town had probably lost most of its civilian population.23 Nevertheless Erciş was the centre of a sancak, and its position was critical. It was the first fortified settlement on the east–west line capable of holding a large garrison. It was also the point of junction with the line of the main tracks from Erzurum via Hınıs and Malazgirt. In the previous twenty years the town had been besieged and changed hands a number of times. In 1535, at the approach of the Safavid shah, the Janissary garrison had abandoned the town and driven the civilian population to Adilcevaz, the next town westward.24 At some point between 1546 and 1550 Erciş was repossessed by Ottoman forces.25 But it was again besieged by Safavid forces for three months in 1552, and its fortifications pulled down.26 The Ottoman reconstruction of the citadel and town wall in 1565–6 took a year and a half. Workmen and animals were collected from all sancaks of the Van eyalet.27 Four years later a pipe was constructed to bring water from a village nine or ten kilometres to the north.28 Besides its defensive role Erciş contained factories for sulphur and saltpetre refinement and for the production of gunpowder. In this role it cannot have been (p.218) much less important than Van. Most of the products (whether sulphur, saltpetre or gunpowder) went to Istanbul via Trabzon; as suggested above, Erciş was also a stagingpost for sulphur and gunpowder being transported from Van to the same destination.29 The sources, the terrain and the memory of so many bitter sieges would suggest that despite all the investment by the Ottoman authorities the civilian population was small and the life of the soldiery here was bleak. The garrison, two years after the great reconstruction of 1565–6, preferred to live outside the walled area.30 In 1577 the soldiers had again to be told to live inside the walls.31 Only one mosque (cami), besides several mescids (neighbourhood mosques), is known.32 Erciş was the starting-point for the road to Erzurum, as mentioned above. First stop after Erciş was the former city of Malazgirt, in the broad basin of the Lower Euphrates. Malazgirt’s history in the mid-sixteenth century illustrates very usefully the exigencies of administration in an open district not far from what we have described as a somewhat open border. The nahiye of Malazgirt lay at first with the sancak of Hınıs, which had certainly been created by 1539.33 In a survey of 1543 it was discovered that sixty of the nahiye’s villages were abandoned and in decay. Another thirteen were uninhabited, though some agricultural production took place on their soil.34 A later survey (1556) revealed that the most frequent population size of the inhabited villages was two households, the next most frequent was one and the greatest size was four. The walled area of the former medieval city contained a population of just twenty-two households.35 A first attempt at repair of the walled area, or possibly just the citadel, was made in 1550.36 In 1559 an Imam Kulu repaired the fortifications. Within a year a separate sancak of Malazgirt was created, of which İmam Kulu was appointed sancakbeyi.37 (p.219) However, this İmam Kulu, we learn from a government order written a year later, was chief of a tribe, which certainly had brought life and cultivation to a number of villages; but another tribe, or more truthfully tribal federation, the Boz Ulus, had been specifically excluded from its old pastures and dwellings in the nahiye of Malazgirt; and in 1564 the Boz Ulus, having submitted a petition, were allowed back, on condition that they revitalise the district. In the same year İmam Kulu was dismissed as sancakbeyi.38

Page 5 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century From Malazgirt, Erzurum could be reached in a variety of ways. The longest track, but one certainly in use at the time, followed the Lower Euphrates upstream to the plain of Eleşkirt. From here the long valley-plain of Pasin, whose west extremity lay not far from the walls of Erzurum, could be reached via the Tahir pass. A more direct way led into the Karayazı Düzü, the watershed-plain in the heart of the massif bordering the Lower Euphrates plain to the north; and from the Karayazı Düzü the track led down to the Pasin district via Avnik. However the straightest way from Malazgirt to Erzurum lay downstream in the broad plain of the Lower Euphrates, then northward through the vale of Hınıs, after which Erzurum was normally reached along tracks passing through high pastureland. Hınıs itself was the chief town of a sancak essentially comprising the Lower Euphrates valley and its northern affluents as far as the plain of Muş. But for the Ottoman administration it also had importance as a strongpoint generally and as a fort guarding the way towards Erzurum. It consisted of a small citadel on a high rock, from which the town walls descended and enclosed a limited area on both sides of the ravine cut by the river.39 We know that by 1568 the citadel was undergoing a complete reconstruction; the town walls were under reconstruction in 1571.40 At Adilcevaz, further west, the walled area, on a steep apron of rock cut away in cliffs to either side, was limited in extent, while the citadel at the top of the rock was tiny. Nevertheless the position was secure, much more so than at Erciş and more secure even than in the case of the following fortress, that of Ahlat. A suburb extended on the plain east of the fortified rock; the place’s importance was increased by the harbour directly in front of the suburb.41 Because of the relative strength of the walled area and citadel, the town had been attacked less frequently than others such as Erciş. The city had been attacked by Safavid forces in 1548, soon after the Ottoman capture of Van. But in this single attack the city had not been taken, and the damage had no doubt been (p.220) slight.42 Repairs here were carried out relatively late: one repair is known, which started in 1574.43 At Ahlat the area of the extensive civilian settlement, which had been dominant in the region during the Ilkhanid period (thirteenth to fourteenth centuries), and which lay on a plateau overlooking the lake, and the walled area above a ravine, had been subjected to systematic destruction during an attack by the Safavid Shah Tahmasp’s troops in 1552.44 It was probably after this attack, and at any rate by the time of a fiscal survey carried out in 1556, that the nahiye of Ahlat was transferred out of the Bitlis sancak and into that of Adilcevaz.45 But as a critical point on the east–west route Ahlat needed to be fortified and garrisoned, and straight after the Treaty of Amasya a small walled area, for military use only, was built on land sloping down from the plateau to the lake. To this a larger walled area, to shelter civilians, was added in 1568.46 Within the walled area gunpowder was manufactured, as at Van and Erciş. Supplies of sulphur and saltpetre were brought from elsewhere; wood was sometimes brought across the lake on boats.47 Although the new walled area allowed the population to grow somewhat, Ahlat remained basically a military installation with a small town attached: beside the civilian enclosure of the fortress, the civilian population probably inhabited isolated patches of the plateau, where among the ruinous mosques and medreses of the former city a few zaviyes still functioned and the giant mausolea of the Ilkhanid period still stood.48 When the north and west shores of the lake had been traversed, the track started the descent down the Bitlis pass. Here for the first time caravanserais, less than a day’s journey apart, were encountered: they guarded against robbers—caravans were more vulnerable on the incline—and were a refuge in case of sudden snowfalls. Particularly lavish sums were invested by one of the

Page 6 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century Ottoman beylerbeyis of Van, Hüsrev Pasha (1566–74), during the interval of direct rule before the Bitlis prince returned, in an enormous double caravanserai at Rahva, by the saddle at the head of the pass. There had in fact been a ribat, perhaps a fortified resthouse, on the same spot in the late Middle Ages.49 In the city of Bitlis itself a formidable citadel stood on a rock at the end (p.221) of the promontory between two rivers. The limited walled area lay on the nose of the promontory itself: walls descended from the citadel to each of the two rivers and then turned along the river bank.50 This left the bulk of the residential area unprotected. Great damage had been done to the area around the Gök Meydan (‘Sky Square’), a second focus of the city at a good distance from the walled area, in an attack by Safavid forces in 1548.51 But the citadel and walled area, probably because of their great strength, appear to have been left undamaged. There is no evidence for Ottoman repairs to the fortifications. However, Ottoman investment in the city was not lacking. Hüsrev Pasha built two caravanserais and 100 shops in the city52 This is without mentioning the great building projects of the princes themselves.53 The fortification work just described, apart from safeguarding the route and providing secure bases, principally against Safavid attack, for the beylerbeyi and certain sancakbeyis, also laid the foundations for the beylerbeyi’s relations with the Kurdish principalities, and we must now turn again to these principalities in their new status within the Van beylerbeyilik. The net of the sancak system had been thrown over them. They were now sancakbeyis, with new responsibilities. Among these responsibilities the most salient was that of bringing troops to the beylerbeyi’s standard in time of war. On the other hand in each principality the appointment of the prince remained within the ruling family, and, with the important exception of Bitlis, the principalities’ territories were not subject to the tahrir. The new rules in effect envisaged that the princes would remain vassals, but with some of the obligations of the sancakbeyi.54 Subtle ways of controlling the princes were found, in particular when the beylerbeyi was able to intervene in disputes over territory or the princeship, as will appear below. In some of the principalities at least the kadıs were appointed by the sultan and, as in the case of the kadı of Moks (south of Lake Van), expected to investigate and bring to justice, on behalf of the sultan, even the prince himself55 (p.222) Naturally the actual impact on, and application of the system to, given principalities varied and depended in part on the actions of the beylerbeyi at Van. The Hakkari principality was split into four sancaks by 1555, and these sancaks were simply former family apanages in disguise, and, of course, in the hands of family members. The principality was prone to division between opposing groups of tribes, to competition between members of the ruling family for the position of prince, and to splits between pro-Ottoman and pro-Safavid factions. The beylerbeyi would act to resolve the disputes between family members and to impose his own choice, especially if the latter had requested help from him. However the princeship always remained within the family56 The Mahmudi principality, once it had passed to Ottoman sovereignty in 1554, presented no problems of loyalty or of competition within the ruling family, and this may be connected with the fact that it was based (as the limiting case) on a single tribe.57 However, the tiny principalities further west, in the valleys between the Mahmudi and Hakkari on the one hand (to the east) and the Bitlis principality, or at least its territory, on the other (to the west), were in a state of upheaval, where infighting between members of the respective ruling families resulted in petitions to the Ottoman beylerbeyi and the breaking-off of further small principalities. Despite the small size, both of the parent principalities and of the new, broken-off ones, each one was a sancak, sometimes no larger than the equivalent of five zeamets. It is worth noting that the

Page 7 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century Hakkari and Mahmudi princes were sometimes involved, either as peacekeepers or as allies brought in by one or another faction in order to claim the leadership of one of these tiny principalities.58 When the prince of Bitlis, Sharaf (son of the fugitive Shams al-Din), came back to his principality in 1578, he exacted from the authorities a few territorial concessions, in reality former possessions of the Bitlis principality. These included Ahlat, on the lake’s north-west shore, and part of the lake’s southern shore. He took over the hass of the former sancakbeyis both of Bitlis and of Muş, and to this hass more land was added. Sharaf derived income, therefore, from estates whose status was now formalised and which were cast in the frame of Ottoman imperial and administrative categories. The prince’s estates lay in three loose concentrations among standard timars and hass-i hümayûn domains. Sharaf also re-established his relations with the tribes. The prince thus regained his principality, the income from certain estates and his tribal base. However, with the timars and the hass-i hümayûn lands dispersed over his territory, a greater degree of control could be exercised over him than in the pre-1535 state of (p.223) vassalage.59 It is worth remembering that in Bitlis itself, even after the return of Sharaf in 1578, there was an Ottoman garrison.60 To sum up, after 1548 Ottoman administration in the border region was based in the first place on a line of heavily fortified towns and small cities around the lake. Those on the north and west shores (Erciş, Adilcevaz, Ahlat and Bitlis) also guarded a track which after the institution of the beylerbeyilik of Van and the Treaty of Amasya acquired considerable importance as a trade route between Iran and Aleppo. We have emphasised the importance attached to securing, even by means of intensive campaigns involving labour recruited from the whole eyalet, the most critical places, such as Van and Erciş; and these were the ones most seriously damaged by the warfare of the years 1548 to 1554. We have also shown how, one by one, the fortifications of the strategically less important and less seriously damaged settlements, such as Adilcevaz and Hınıs, were repaired—but later in time, when Van, Erciş and Ahlat had been secured. Once the network of strongpoints had been set up, it was exploited for the manufacture of gunpowder and sulphur, most of which was transported to Istanbul to meet the demands of the Ottoman army generally. The fortified towns and cities did not merely secure the directly administered areas but also provided a firm base for dealing with the Kurdish principalities. Without such fortresses it would have been impossible to intervene in the principalities’ internal disputes or to insist on the arrangements worked out for them, to counter rebellions or face the principalities with the threat of their suppression. The formal arrangements were an adaptation of the state of simple vassalage which had obtained before 1548. The princes, and sometimes their relations, were now sancakbeyis. But the principalities, at least formally, kept control of all their internal affairs; in particular, succession was kept within the ruling family, and, with the explicable but notable exception of Bitlis, the principalities’ territory was not subject to the tahrir. But in practice, owing to the struggles over the princeship between members of the ruling families and to factionalism between the tribes, the beylerbeyi at Van was able to intervene on one side or the other, thus securing his own candidate or neutralising any pro-Safavid faction, as in the case of the Hakkari principality. Thus a greater degree of control was exercised over the principalities than was strictly provided for in the formal arrangements. During the second half of the sixteenth century the arrangements worked, grosso modo. The princes did bring troops to the banner of the beylerbeyi in time of war and, during the

Page 8 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century campaigns of 1578–90, helped in new conquests. However, the system began to fail in the following century, and there took place a series of rebellions and (p.224) defections.61 Moreover, the Ottoman Empire never succeeded in what had probably been its original aim, as adumbrated by the fate of the Bitlis principality. This probable aim was to expel the princely dynasties, subject their territory to the tahrir, settle this territory with sipahis, unravel the relationships between the constituent tribes, and ultimately dissolve the tribes themselves. Notes: Proceedings of the British Academy 156, 211–224. © The British Academy 2009. (1) On the territory of all three, see Tom Sinclair, ‘The Ottoman arrangements for the tribal principalities of the Lake Van region of the sixteenth century’, in Kemal H. Karpat with Robert W. Zens, Ottoman Borderlands: Issues, Personalities and Political Changes (Madison: University of Wisconsin Press, 2003), 119–20. In the present chapter, for reasons of economy, we shall be summarising various sections of the previous article and refer the reader to it for sources. Fortification work will instead be given prominence, with adequate documentation: this material has not been published elsewhere. (2) Ibid., 121. (3) Ibid., 121–2. (4) Ibid., 123–4. (5) Ibid., 122–3, 124–7. For a comment on one aspect of these arrangements, Mehmet Öz, ‘Ottoman provincial administration in eastern and southeastern Anatolia: the case of Bidlis in the sixteenth century’, in Karpat and Zens (eds.), Ottoman Borderlands, 152. (6) On Hınıs, Sinclair, ‘Ottoman arrangements’, 125. (7) Ibid. (8) Ibid., 129–30. (9) For references on the capture of Van, ibid., 128, and M. Fahrettin Kırzıoğlu, Osmanlılar’ın Kafkas-Elleri’ni Fethi, 1451–1590 (Ankara: TTK, 1993), 187–8. (10) On the fighting in the above years, Kırzıoğlu, Osmanlılar’ın Kafkas-Elleri’ni Fethi, 185–8, 190–2, 211–16, 218–19, 225. (11) On this increase, Edmund Herzig, ‘The volume of Iranian raw silk exports in the Safavid period’, Iranian Studies, 25 (1992), 63–5. On the significance of the Aleppo route, Halil İnalcık with Donald Quataert (eds.), An Economic and Social History of the Ottoman Empire, 1300–1914 (Cambridge: Cambridge University Press, 1994), 243–4, 340, 343–4. (12) For an account of the whole site, Thomas A. Sinclair, Eastern Turkey: An Architectural and Archaeological Survey (London: Pindar Press, 1987–90), 1: 180–8; for general remarks on the Ottoman occupation of it, Orhan Kılıç, XVI ve XVII. Yüzyıllarda Van (1548–1648) (Van: Belediye Başkanlığı Kültür ve Sosyal İşler Müdürlügü, 1997), 201–7; for references to some recent

Page 9 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century archaeological work, ibid., 212–26, passim. On the officials and soldiers stationed here, ibid., 143–98, 315–69. (13) Manvel K. Zulalyan, Arevmtyan Hayastanë XVI–XVIII dd. (Erevan: Haykakan SSH GA Hratarakchutyun, 1980), 41, quoting Matenadaran MS 5273, pp. 438–40. The reconstruction is mentioned in passing by Ibrahim Peçevi, Tarih (Istanbul: Matbaa-i Amire, 1283), 1: 275. See also Kırzıoglu, Osmanlılar’ın Kafkas-Elleri’ni Fethi, 188. (14) BOA, Mühimme Defteri (MD) 6, no. 1003: in İsmet Binark, Necati Aktaş et al. (eds.), 6. Numaralı Mühimme Defteri (972/1564–1565) (Ankara: T. C. Basbakanlık Devlet Arşivleri Genel Müdürlüğü. Osmanlı Arşivi Daire Baskanlıgı, 1995), 2: 102. (15) Kılıç, Van, 212–26, passim. (16) Zulalyan, Arevmtean Hayastanë, 292, quoting Matenadaran MS 6273, p. 466. (17) BOA, MD 10, p. 333, no. 546, and p. 334, no. 547; MD 14, p. 1031, no. 1525. (18) BOA, MD 46, p. 226, no. 493 (1581); MD 47, p. 63, no. 159 (1582). (19) Among others, MD 14, p. 1070, nos. 1575, 1576; MD 23, p. 172, no. 364. See also Turgut Işıksal, ‘Gunpowder in Ottoman documents of the last half of the sixteenth century’, International Journal of Turkish Studies, 2 (1981–2), 89–90; Gábor Ágoston, Guns for the Sultan: Military Power and the Weapons Industry in the Ottoman Empire (Cambridge: Cambridge University Press, 2005), 99, 106–7, 112. (20) Kılıç, Van, 230–42. (21) BOA, MD 21, p. 33, no. 92. Some merchants came to Van via the Albak district in the Hakkari principality. (22) For a general description of the whole site, which is now under water, see H. F. B. Lynch, Armenia: Travels and Studies (London, 1901; repr. Beirut: Khayats, 1965), 2: 27–9, and others cited in Sinclair, Eastern Turkey, 1: 284. For the shape of the walled town, see Evliya Çelebi, Seyahatname, 4: 101; Nasuhü’s Silahi (Matrakçı), Beyān-ı menāzil-i sefer-i ‘Irākeyn-i Süleymān Hān, ed. H. G. Yurdaydın (Ankara: TTK, 1976), pl. 25b. (23) In the mid-seventeenth century Evliya Çelebi found 1,050 houses within the walls, and from the description it is clear that more could have fitted in (Seyahatname, 4: 101), but many of the houses were occupied by members of the garrison (below, nn. 30, 31). (24) Manr Žamanakragut ’yunner, XIII-XVIII DD., ed. V. A. Hakobyan (Yerevan, 1951, 1956), 2: 231 (Hovhannes of Archesh). (25) Coins in Shah Tahmasp’s name were minted in 949/1542–3 and 953/1546–7 (see the collection of the Forschungsstelle für Islamische Numismatik, Tübingen University, nos. 91-14-25, 91-1-30; and undated coins, IA7 C1, C5). An Ottoman repair of the fortress was ordered in 1550: BOA, Kamil Kepeci Tasnifi 209, p. 110, discussed by Orhan Kılıç, XVI. Yüzyılda Adilcevaz ve Ahlat (1534–1605) (Ankara: Tamga Yayıncılık, 1999), 12–13.

Page 10 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century (26) Hasan-i Rumlu, A Chronicle of the Early Safawīs, being the Ahsanu’t-Tawarikh of Hasan-i Rumlu, ed. C. N. Seddon (Baroda: Oriental Institute, 1931), 1: 367–70; Peçevi, Tarih, 1: 296–7; Bitlisi, Schéréfnameh; ou, Histoire des Kourdes [the Sharafnama], ed. V. Véliaminoff-Zerkov (St Petersburg, 1860–2), 2: 205. (27) Zulalyan, Arevmtean Hayastanë, 292, quoting the colophon of Matenadaran MS 4470, p. 406; BOA, MD 6, no. 1029, published in 6. Numaralı Mühimme Defteri, 2: 116–18; see also MD 6, no. 588, ibid. 2: 323. (28) BOA, MD 14, no. 596. Another restoration of the fortifications took place in 1577 or shortly before: MD 31, p. 31, no. 78. (29) BOA, MD 10, p. 28, no. 39, and see Ágoston, Guns for the Sultan, 101, 102; MD 14, p. 24, no. 37, and p. 740, no. 1063. Saltpetre: among others, MD 9, p. 86, no. 224; MD 15, p. 182, no. 1525; MD 24, p. 8, no. 21. Gunpowder: among others, MD 14, p. 210, no. 296, and p. 1070, no. 1576; MD 18, p. 104, no. 233. (30) BOA, MD 7, no. 1636, in İsmet Binark, Necati Aktaş et al. (eds.), 7. Numaralı Mühimme Defteri (975–976/1567–1569) (Ankara: T. C. Basbakanlık Devlet Arşivleri Genel Müdürlüğü. Osmanlı Arşivi Daire Başkanlığı, 1999), 2: 223: date 1568. (31) BOA, MD 29, p. 223, no. 511. Note that the reluctance to dwell within the walls on the part of the soldiery cannot be put down, at least in 1568, to an aversion to living near gunpowder or its materials, as we have no evidence of its production or of the working of sulphur or saltpetre in 1568. (32) Evliya Çelebi, Seyahatname, 4: 101. Only one mosque is shown in Matrākçī, Beyān-i menāzili sefer-i Irākeyn, pl. 25b. Note the disused medrese mentioned in 1560: BOA, MD3, no. 1411, in 3. Numaralı Mühimme Defteri, 1: 613. (33) Dündar (Aydın), Erzurum Beylerbeyliği ve Teşkilatı. Kuruluş ve Genişleme Devri (1535–1566) (Ankara: TTK, 1998), 252. (34) BOA, Maliyeden Müdevver Tasnifi (MAD) 22171, discussed in Yusuf Sarınay et al. (eds.), 294 Numaralı Hınıs Livası Mufassal Tahrīr Defteri (963/1556) (Ankara: T. C. Basbakanlık Devlet Arşivleri Genel Müdürlüğü. Osmanlı Arşivi Daire Başkanlığı, 2000), 2–3. (35) 294 Numaralı Hınıs Livası Tahrīr Defteri, 78–84. The only known public building in the town was a medrese: ibid., 88. On the extent and layout of the medieval city, Lynch, Armenia, 2: 271–4; summary in Sinclair, Eastern Turkey, 1: 286–7, 288. (36) BOA, Kamil Kepeci 209, p. 155; discussed in Kılıç, Adilcevaz, 12–13. (37) Dündar, Erzurum, 285–6, with references. (38) On İmam Kulu’s aşiret or tribe, esp. BOA, MD 3, no. 1073, in Binark, Aktaş et al. (eds.), 3. Numaralı Mühimme Defteri, 1: 476. Otherwise Dündar, Erzurum, 287–8. (39) For descriptions of Hınıs before the First World War and a summary description of the site and fortification, see bibliography in Sinclair, Eastern Turkey, 1: 289, 290; Ełišē Y. Melik‘ean, Hark‘-Xnus (Antilyas, 1964), 19–20.

Page 11 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century (40) BOA, MD 7, no. 1421, in 7. Numaralı Mühimme Defteri, 2: 115–16 (1568); MD 15, p. 77, no. 660 (1571). (41) Tom Sinclair, ‘The city of Adilcevaz in the late Middle Ages and early Ottoman period’, in Colin Imber and Keiko Kiyotaki (eds.), Frontiers of Ottoman Studies, I: State, Province, and the West (London: I. B. Tauris, 2005), 110–11, 113–17. For the military personnel stationed at Adilcevaz, Kılıç, Adilcevaz, 47–74. (42) Peçevi, Tarih, 1: 274. (43) MD 24, p. 332, no. 910. (44) Tom Sinclair, ‘Ahlat Şehrinin Geç Ortaçağ ve Erken Osmanlı Çağı Tarihi’, in CIÉPO XIV. Sempozyumu bildirileri, 18–22 Eylül 2000, Çesme (Ankara: TTK, 2004), 680–3 on state under Ilkhans, 688 on attack of 1552, references in n. 44. (45) BOA, Tahrir Defter (TD) 297, pp. 24–7; see Faruk Sümer, Selçuklular Devrinde Doğu Anadolu’da Türk Beylikleri (Ankara: TTK, 1990), 63. (46) Sinclair, Ahlat’, 689 and references. (47) Kılıç, Adilcevaz, 207; see also BOA, MD 14, p. 1070, no. 1576. (48) Sinclair, Ahlat’, 688–70. (49) On the caravanserais near Bitlis, see bibliography in Sinclair, Eastern Turkey, 1: 311; and Abdüsselâm Uluçam, Ortaçağ ve Sonrasında Van Gölü Çevresi Mimarlığı, II: Bitlis (Ankara: Kültür Bakanlığı, 2002), 127–32. Before the construction of the Rahva caravanserai, local families were sent to guard the pass against robbers: BOA, MD 14, p. 878, no. 1285. The Sharafnama itself mentions the danger from snow: Bitlisi, Sharafnama, 1: 349. On the Rahva caravanserai, M. Oluş Arık, Bitlis Yapılarında Selçuklu Rönensansı (Ankara: Selçuklu Tarih ve Medeniyeti Enstitüsü Yayınları, 1971), 86–90. On Hüsrev Pasha’s dates, Kılıç, Van, 147–8. On the ribat, Fadlullah b. Ruzbihan Khunji-Isfahani, Tarikh-i ‘Alam-Ara-yi Amini, ed. John E. Woods with abridged trans. by Vladimir Minorsky (London: Royal Asiatic Society, 1992), 23. The former ribat in fact stood in a semi-ruined state even in the late sixteenth century: Bitlisi, Sharafnama, 1: 349–50, though the author saw the ruins as a semi-finished building, never used. (50) See the illustration from Matrakçı reproduced in Uluçam, Van Gölü Çevresi Mimarlığı, 15. Evliya Çelebi, Seyahatname, 4: 66: ‘aşağı varoş kalesi’. (51) Bitlisi, Sharafnama, 2: 200. (52) Ibid., 1: 348. (53) On these, see Arık, Bitlis, 19–26, 47–51; Uluçam, Van Gölü Çevresi Mimarlığı, 25–42, 96–100. (54) Sinclair, ‘Ottoman arrangements’, 138–40. (55) Ibid., 140, n. 94; for Moks, BOA, MD 6, nos. 1100 and 1179, in 6. Numaralı Mühimme Defteri, 2: 152, 204. For a discussion of the kadı of Bitlis in the mid-seventeenth century, see Öz,

Page 12 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Administration and Fortification in the Van Region under Ottoman Rule in the Sixteenth Century ‘Ottoman provincial administration’, 147–8, and compare that of Palu, also in a hükûmet, ibid., 154. (56) Sinclair, ‘Ottoman arrangements’, 131–4. (57) Ibid., 134–5. (58) Ibid., 135–8. (59) Ibid., 130–1. (60) MD 36, p. 140, no. 392; cf. Sinclair, ‘Ottoman arrangements’, 131. For a discussion of the mid-seventeenth-century Janissary commander and the harac ağası, both appointed by the sultan, in Bitlis, see Öz, ‘Ottoman provincial administration’, 147–8. (61) Sinclair, ‘Ottoman arrangements’, 140–2.

Access brought to you by:

Page 13 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Frontier Policies in North-East Africa, 1517–1914

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Ottoman Frontier Policies in North-East Africa, 1517–1914 JOHN ALEXANDER

DOI:10.5871/bacad/9780197264423.003.0011

Abstract and Keywords This chapter suggests that insufficient attention has been paid in accounts of north-east African history to the role of the Ottoman Turks. With the capture of Egypt from its Mamluk rulers in 1517, the Ottomans established their first foothold in Africa. However, several factors drew them further into the region. First, there was a threat presented by the Portuguese, who sought to establish a monopoly on the valuable Indian Ocean trade and who challenged Ottoman control of the Red Sea and the pilgrimage routes to Mecca and Medina. Second, the Ottomans wished to secure control over Africa's valuable exports, slaves and gold. Third, in accordance with the sultans' quest for legitimacy as rulers of an Islamic empire, their long-term aim was the inclusion of all north-east Africa into Ottoman territory and hence the Dar al-Islam. Keywords:   Ottoman Turks, Islamic empire, Dar al-Islam, pilgrimage routs, Indian Ocean trade, Red Sea

THIS CHAPTER WILL SUGGEST THAT INSUFFICIENT ATTENTION has been paid in accounts of north-east African history to the role of the Ottoman Turks. With the capture of Egypt from its Mamluk rulers in 1517, the Ottomans established their first foothold in Africa. However, several factors would draw them further into the region. Firstly, there was threat presented by the Portuguese, who sought to establish a monopoly on the valuable Indian Ocean trade and who challenged Ottoman control of the Red Sea and hence the pilgrimage routes to Mecca and Medina. Secondly, the Ottomans wished to secure control over Africa’s valuable exports, slaves and gold.1 Thirdly, in accordance with the sultans’ quest for legitimacy as rulers of an Islamic empire, their long-term aim was the inclusion of all north-east Africa into Ottoman territory and hence the Dar al-Islam.2 Control of the Red Sea was secured in the sixteenth century and ended the Portuguese threat, but the longer-term aim, although pursued into the nineteenth century, failed. The means by which the Ottomans intended to expand their empire were, firstly, to transform Egypt into a fully integrated province, and this was achieved in the sixteenth century; and in the longer term, to conquer Christian Ethiopia and the Upper Nile Valley and incorporate them into

Page 1 of 9 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Frontier Policies in North-East Africa, 1517–1914 the empire, an endeavour in which they failed. In the sixteenth century, the Ottomans set up two new frontier sancaks of Habeş and Ibrim on the Red Sea coast and the Middle Nile Valley respectively, subsequently enlarging them into two new eyalets of the same name. When attacks on Ethiopia failed in the late sixteenth century, around the same time as expansion up the Nile was brought to halt, the Istanbul government turned its attentions to conquests in Asia and Europe, but returned to the attack in the nineteenth century. The ways in which changes in imperial policy affected north-east Africa and its indigenous populations on both sides of the frontiers are best considered in four phases: 1517–55, 1555–89, 1589–1820 and 1820–1914.

(p.226) Period One: 1517–55 The key to expansion in north-east Africa was Egypt, which controlled all the main points of access into the region. The defeat of the last Mamluk, Tumanbey, by Sultan Selim I outside Cairo in 1517, was followed by the establishment of Egypt as an Ottoman province.3 The province had four frontier zones, in each of which there were specific problems and opportunities: two coastal ones (Mediterranean and Red Sea) and two inland ones (Nile Valley and Western Desert). On the Mediterranean frontier, soldiers and resources from Egypt were used in expeditions such as that against Malta in 1565. In the Red Sea, defence of the trade routes and holy places of the Hijaz against the Portuguese was the priority, along with maintaining Ottoman control of Yemen. This too depended heavily on resources from Egypt. On the Western Desert frontier the empire continued the passive policy of its Mamluk predecessors, which was to prevent the nomadic pastoralists of the desert from raiding the agriculturalists of the Delta and Nile valley, and to control and protect the commercial and pilgrim caravans making their way to Egypt from the sub-Saharan savannahs. Initially, the Ottomans relied on local chiefs (mashayikh al-‘urban) to achieve this for them, but in the seventeenth century the oases were incorporated into the sancak of al-Sa‘id (Upper Egypt).4 Likewise, the Istanbul government was content to control its southern frontier, as its Mamluk predecessors had done, through the semi-nomadic confederation of the Banu ‘Umar which had its centre at Aswan at the beginning of the First Cataract.5 More than 800 kilometres to the south lay the Funj Sultanate, newly created from the ruins of the last Christian Kingdom in the Nile Valley, Alwa, with which relations were largely limited to passing trading caravans, mainly by way of the Red Sea ports.6 In the early sixteenth century, the Red Sea frontier presented an urgent problem for the Ottoman sultans, indeed so much so that they had been prepared to assist their Mamluk rivals, with an Ottoman fleet setting sail from Suez in 1515 to do battle with the Portuguese.7 The latter’s naval supremacy in the Indian Ocean and the Red Sea had destroyed the lucrative spice trade with south-east Asia and was threatening not only Egypt but also the Holy Cities of Mecca and Medina. The threat was removed by the creation of an Ottoman Red Sea fleet and the occupation and garrisoning of all the major ports on both coasts. On the African coast these were Suakin, Massawa and (p.227) Zayla,8 of which Suakin on the pilgrim route linking Africa with the Hijaz was to prove the most important. The other problem in this region was the Empire of Ethiopia, the only surviving Christian state in north-east Africa. South of it lay the Sultanates of Adal and Harar, where by the 1530s raids on Ethiopia had developed into a jihad led by Imam Ahmad Gran with the intention of creating a new sultanate. The Ottoman Empire took no formal part in this but aided it with reinforcements, firearms and supplies. By 1547 the jihad’s complete success was only frustrated by the death of Imam Ahmad Gran and Portuguese aid to Ethiopia.9 In 1555, under Süleyman the Magnificent, the Ottomans created the eyalet of Habeş¸ on the Red Sea coast of Africa with its capital at Suakin under Özdemir Pasha, the experienced beylerbeyi of

Page 2 of 9 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Frontier Policies in North-East Africa, 1517–1914 the eyalet of Yemen, accompanied by an army some 3,000–5,000 strong, marking the start of Period Two in the region.10

Period Two: 1555–89 In the Nile Valley in this period the frontier of Egypt advanced 180 kilometres from Aswan to Wadi Halfa at the entrance to the Second Cataract, and a new frontier sancak of Ibrim was created. This relatively minor alteration may have been thought necessary because of the advance northwards of the Funj Sultanate causing a possible threat to Egypt; or to parallel and aid the contemporary advances in Habeş.11 The sancak capital was fixed at al-Dirr and nearby the ancient hilltop fortress of Qasr Ibrim was extensively repaired and garrisoned.12 Only the slender frontier region of Sukkot (around Sai) now lay between Ottoman and Funj territory. The establishment in 1555 of the eyalet of Habeş was in preparation for the attempts to destroy the Empire of Ethiopia.13 There can be little doubt that the government in Istanbul was aware of the increasing weakness of the Christian empire in the 1550–80s under the fierce attacks from the south-west of the Galla (Omoro) tribes of transhumant pastoralists, and that this inspired a major attempt to conquer it.14 In the twenty years after the establishment of the eyalet of Habeş, six attacks were launched, achieving temporary successes but finally failing, Özdemir Pasha himself dying in 1560. Some of his successors like Ahmed Pasha (1577–9) continued the struggle, occupying (p.228) the northern provinces of Ethiopia before he and his local ally were defeated and killed.15 The creation of the eyalet of Habeş was decided in Istanbul and, as stated above, in 1584 it was paralleled in the Nile Valley by the creation of a new eyalet of Ibrim, another central government decision (Figure 11.1).16 The sancak of Ibrim had been briefly attached to the eyalet of Habeş,17 but the new eyalet of Ibrim covered most of Egypt south of Qina up to Sukkot. This opened the way to the conquest of the Funj Sultanate and would have further isolated the Christian empire. Perhaps even before Ibrim was created an eyalet, around 1582–3, an expeditionary force assembled to conquer Dongola, probably at Wadi Halfa, where boats were collected for the long and dangerous passage of the Second Cataract. A near-contemporary account states that only one Ottoman boat returned, the others being lost on the rocks at some point south of Wadi Halfa.18 According to the traditions of the ‘Abdallab tribal confederation, the Ottomans were defeated at Hannek near the entrance to the Third Cataract by a Funj army19 Once again it must have been a central government decision that, after a few months, abolished the new eyalet of Ibrim, made peace with the Funj and returned all the sancaks to Egypt. A new fortress, Qal‘at Sai, somewhat to the north of the new frontier at Hannek, was created from an existing ruined one and given artillery and a Janissary garrison.20 Qasr Ibrim now became a reserve base. Although there continued to be raids on Christian Ethiopia there were no more attempts at conquest; the ‘window of opportunity’ had closed.

Period Three: 1589–1820 In the long period between 1589 and 1820 the Ottoman frontier in the Nile Valley remained largely unchanged. On the Red Sea, the diminishing of Portuguese power and the end of Ottoman attempts to conquer Ethiopia led to a reduction in the status of Habeş, which was joined to Jeddah by the early eighteenth century21 In any event, since the seventeenth century Habeş had been increasingly constricted to the region round Suakin itself, with outlying portions at Massawa and Zayla.22 Although effective control of Egypt passed to local mamluk factions,23 the Ottoman Empire still held all the main points of access into north-east Africa. However, it had little impact on either the Funj Sultanate or Ethiopia, and both neighbours developed in near isolation. The most important Ottoman influence on the region in this and the next period Page 3 of 9 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Frontier Policies in North-East Africa, 1517–1914 seems to have been religious, and a succession of learned migrant holy men (fakis and ulema) from Arabia and Egypt reached the Nile valley and were welcomed and given land.24 They founded zawiyyas, mosques and religious brotherhoods, which were important in all later periods.25 Otherwise, relations with the Funj remained largely limited to trade through Suakin, and consisted mainly of slaves, gum and gold bartered for firearms, but no artillery, cloth and luxury goods.26 (p.229) (p.230) The two frontier sancaks in this period changed in very different ways. In the sancak of Ibrim, the Nubian population seems to have sunk into deep poverty and although there were no mamluk estates, land near the two main fortresses of Sai and Ibrim came into the hands of garrison families, which by the end of the period formed local landowning elites. At Kulubnarti, the only other excavated settlement in the sancak, there is no evidence for any alteration in economic conditions.27 There were some developments in agriculture, new crops of coffee, tobacco and cotton being introduced and a substantial increase in saqiyya (waterwheel) irrigation. At Qasr Ibrim many documents of the period have been excavated and show that changes in Janissary regulations allowed soldiers to marry and build houses in the fortress early in the seventeenth century and their sons to

Figure 11.1. Map of north-east Africa in the late sixteenth century.

join the army.28 They held saqiyya-irrigated land, and in the eighteenth century had a profitable trade in dates to Aswan. The documents show they were losing their knowledge of Turkish but, as Janissaries, they were still being paid from Cairo into the 1790s. Although documentary evidence is lacking, Qal‘at Sai’s garrison probably developed in a similar pattern and even today families whose origins are from both places still claim to be descended from the garrisons. Suakin in Habeş developed rather differently. The garrison was supplied from Jeddah and by the end of the seventeenth century appears to have held little land outside (p.231) the island, which was occupied mainly by merchants, including some from India and the Yemen. The mainland here was controlled by Beja tribes who, as in the Nile Valley, controlled the desert trade routes. In contrast to the isolation of Ibrim and Sai, and despite its diminution in status with the abolition of the eyalet of Habeş, Suakin remained linked to global trade systems, and visitors in the eighteenth and nineteenth centuries noted its trading connections spread as far away as India and even south-east Asia.29 The long peace was broken by the French occupation of Egypt in 1798, but since the invading troops did not penetrate south of Aswan or to the Red Sea coast, the two frontier sancaks

Page 4 of 9 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Frontier Policies in North-East Africa, 1517–1914 remained within the Ottoman Empire. In the sancak of Ibrim, resistance to the French was continued by troops and mamluks retreating from Aswan and volunteer fighters from Arabia.

Period Four: 1820–1914 Although Mehmed Ali’s appointment as governor of Egypt in 1804, when he was thirty-five years old, gave rise to nearly a century of alternating obedience to and rebellion against the Sultans, Egypt itself nonetheless remained during this period part of the Ottoman Empire, both politically and in terms of the culture of the ruling elite of Egypt.30 Mehmed Ali’s expansionist ambitions, which had seen him occupy not just Yemen and the Hijaz but also Syria and Palestine, and his armies advance into Anatolia, were checked by the intervention of European powers in 1833 and 1840.31 Nonetheless, his aim was not to destroy the Ottoman Empire, to which he remained closely linked and which granted him his title of khedive, although historians of the Nile Valley have preferred to emphasise his acceptance of European techniques and personnel, a policy continued by his dynasty.32 More lasting than his other conquests was his empire in Africa. As a vassal of the Ottoman sultan, his campaigns in 1820–4 and those of his heirs in 1849–77 took the Ottoman frontier some 2,000 kilometres further south. The conquest in 1820–4 of the Funj Sultanate was launched from Wadi Halfa by the route taken in 1582.33 It was subsequently authorised by a ferman from Istanbul (p.232) which stated that the conquered territories of Nubia, Kordofan and Sinnar were to be held separately from Egypt for Mehmed Ali’s lifetime only and were to pay a separate tribute.34 Massawa and Suakin were granted to Mehmed Ali on an annual lease in 1846, although this arrangement lasted only three years.35 Although Darfur was included in the ferman, in reality that region only fell under Egyptian control in 1874.36 In any event, a new Ottoman frontier was established in sub-Saharan Africa that stretched from the Red Sea to the borders of Darfur. Turco-Egyptian administration was headed by a hükümdar (governor-general), a new title specific to the Sudan,37 but its organisation was otherwise based on Ottoman models, with officers such as kaşifs and kaymakams. Turkish remained the language of command in the Egyptian army in the Sudan until the late nineteenth century.38 When Mehmed Ali died in 1847, control over the Sudan was retained by his successors. The strength of Turkish influences is reflected in the name traditionally given to this Egyptian empire in the Sudan (1820–81), the Turkiyya. Although the way to south Sudan had been opened as early as Mehmed Ali’s reign with the Nile expeditions of the Turkish sailor Selim Kapudan, further expansion of Turco-Egyptian rule had to wait until the reign of Mehmed Ali’s grandson, the Khedive Isma‘il (1863–79). Isma‘il’s expansionist designs were supported by the Istanbul government, which granted him control of most Ottoman territories on the African Red Sea coast in 1865. These included Suakin and Massawa, although Zayla remained under direct Ottoman sovereignty rather longer.39 This allowed İsmail’s forces to conquer Harar in 1874 and attack Ethiopia from the north and the south.40 In parallel with these attacks, other forces advanced southwards and westwards from the Egyptian Sudan.41 In the south in 1874–5, Egyptian forces under the command of General Charles Gordon advanced some 700 kilometres into the Greater Lakes region of central Africa, setting up a capital of the new province of Equatoria at Gondokoro,42 while in the west Darfur was conquered at last. A naval expedition attempted to establish a post at the mouth of the Juba River opposite Zanzibar.43 By 1879, when Isma‘il was forced to abdicate and retire to Istanbul, Egyptian expansion had reached its greatest extent, bringing vast areas of eastern Africa under nominal Ottoman suzerainty, but this was to be destroyed by the Mahdiyya.

Page 5 of 9 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Frontier Policies in North-East Africa, 1517–1914 (p.233) The successful revolt of Ahmad Muhammad, the self-announced Mahdi (the expected one) was ultimately to result in greater European influence. Between 1885 and 1898 the Mahdist state ruled all of Egypt’s Sudanese territories except for Suakin Island, Equatoria and Darfur. In Darfur, an independent sultanate was revived, which lasted until 1916. Mahdist jihads against Egypt and Ethiopia were launched, both ending in failure. In 1898 the Mahdist state was defeated and destroyed, being replaced by the Anglo-Egyptian Condominium, which also included Suakin and a stretch of the Red Sea coast. The Condominium remained, in name at least, part of the Ottoman Empire until Britain declared a protectorate over Egypt with the outbreak of the First World War in 1914. The ruler of Egypt and the Sudan, Husayn Kamil, successor to Khedive ‘Abbas Hilmi, who was deposed for his Ottoman sympathies, now himself assumed the title of sultan, symbolising his rejection of Ottoman sovereignty.44 Thus the nineteenth century witnessed the dilution of the influence of the Ottoman Empire on the native inhabitants of north-east Africa, especially in the Lower Nile Valley, by European concepts and technology. However, since the Ottoman Empire or its vassals, the khedives of Egypt, retained control over all the major coastal ports of access to the interior, and its frontiers were advanced so far to the south, Muslim influence was actually stronger than it had been since the sixteenth century. Particularly important in this respect was the Egyptian establishment of a class of ulema in the Sudan, and the appointment of kadıs in a region where Islam had generally been spread by more informal means such as Sufi shaykhs and fakis.45 However, in the more distant regions such as Equatoria, which had been conquered by Egyptian troops under European command, Islam did not penetrate, and the administrative systems established diverged significantly from Ottoman and Egyptian norms.46 In these regions conquest had in any event been preceded by trade. In particular, ivory-hunting and the slave trade were the mainstay of these frontier areas, newly incorporated into the Dar al-Islam, for the pagan negroid populations continued to be enslaved into the reign of Isma‘il despite efforts to abolish the trade.47 Beyond the Ottoman–Egyptian frontier east of the Blue Nile valley lay Ethiopia, which had preserved its independence into this century. Although it relied on Muslim merchants and ports for trade, it proved adaptable and accepted many European innovations, which helped it to defeat Ottoman, Egyptian, Mahdist and Italian attempts at conquering it.48

(p.234) Conclusions Historians of north-east Africa, especially those interested in the development of modern Egypt and the Republic of the Sudan, have given special attention to the changes brought about by the introduction of European ideas in the nineteenth century. In doing so, the durability of the Ottoman Empire, seen by them as in terminal decline, has not always been given sufficient attention. Throughout the early modern period, the Ottomans were a constant presence in northeast Africa, with periods of significant expansion in the sixteenth and nineteenth centuries as a result of policies pursued both by the central government in Istanbul and the provincial one in Egypt. Ottoman successes and failures at the African frontier of the Empire profoundly influenced events there and were as important as the new non-Muslim concepts which have led to present-day political, religious and economic patterns in the region. The Ottoman failure to conquer Ethiopia and its long tolerance of independent sultanates in Adal, Funjistan and Darfur allowed the development of idiosyncratic traditions in what are now the independent states of Somalia, Ethiopia, Eritrea and the Sudan.49 One particular decision in the mid-nineteenth century which, for the first time, placed Suakin and a stretch of Red Sea coast under a government in Khartoum, has given the present Republic of the Sudan its own outlet to the Red Sea and so to the world. Finally, it should be noted that, had it not been for the First World War,

Page 6 of 9 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Frontier Policies in North-East Africa, 1517–1914 the Ottoman Empire’s southern frontier in north-east Africa would have continued to be in what is now Uganda. Note. I am grateful to Professor Victor Ménage for help with the Turkish sources. Notes: Proceedings of the British Academy 156, 225–234. © The British Academy 2009. (1) Cengiz Orhonlu, Osmanlı İmparatorluğu’nun Güney Siyaseti: Habeş Eyaleti (Ankara: TTK, 1996), 31–3. (2) On the Ottomans and Islamic legitimacy, see the Introduction to the present volume. (3) Stanford J. Shaw, The Financial and Administrative Organization and Development of Ottoman Egypt, 1517–1798 (Princeton: Princeton University Press, 1962), 12–19. (4) P. M. Holt, ‘The Beylicate in Ottoman Egypt during the seventeenth century’, Bulletin of the School of Oriental and African Studies, 24 (1961), 220–1. (5) Victor L. Ménage, ‘The Ottomans and Nubia in the sixteenth century’, Annales Islamologiques, 24 (1988), 139–40. (6) Giovanni Vantini, Oriental Sources Concerning Nubia (Heidelberg: Akademie der Wissenschaften, 1975), 745–96; Gabriel Warburg, ‘A note on David Ha-Reuveni’s visit to the Funj Sultan in 1523’, Sudan Studies, 34 (2006), 20–31. (7) Palmira Brummett, Ottoman Seapower and Levantine Diplomacy in the Age of Discovery (Albany: State University of New York, 1994), 111–21. (8) P. M. Holt, A Modern History of the Sudan: From the Funj Sultanate to the Present Day (London, Weidenfeld & Nicolson, 1961), 25. (9) Mordechai Abir, Ethiopia and the Red Sea (London: Cass, 1980), 85–92, 123–30; Orhonlu, Habeş Eyaleti, 22–30. (10) Orhonlu, Habeş Eyaleti, 37–9. (11) Ménage, ‘The Ottomans and Nubia’, 145–6. (12) John Alexander, ‘The Turks on the Middle Nile’, Archéologie du Nil Moyen, 5 (1996), 15–35. (13) Ibid., 31–2. (14) Orhonlu, Habeş Eyaleti, 70–2. (15) Orhonlu, Habeş Eyaleti, 56–60; T. Tamrat, ‘Ethiopia, the Red Sea and the Horn’, in Roland Oliver (ed.), The Cambridge History of Africa, III: c.1050–1600 (Cambridge: Cambridge University Press, 1977), 98–182. (16) Ménage, ‘The Ottomans and Nubia’, 152–3. (17) Ibid., 147–8.

Page 7 of 9 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Frontier Policies in North-East Africa, 1517–1914 (18) Voyages en Égypte des années 1589, 1590 et 1591, ed. and trans. Carla Burri et al. (Cairo, IFAO, 1971), 146–9. (19) R. S. O’Fahey and J. L. Spaulding, Kingdoms of the Sudan (London: Methuen & Co., 1974), 35. (20) David N. Edward, ‘The potential for historical archaeology in the Sudan’, Azania, 39 (2004), 16–30; John Alexander, ‘Qalat Sai: the most southerly Ottoman fortress in Africa’, Sudan and Nubia, 1 (1997), 16–19; cf. Chapter 19 by Intisar Elzein in this volume, and eadem, ‘The Ottomans and the Mahas in the Third Cataract region’, Azania, 39 (2004), 50–7. (21) Orhonlu, Habeş Eyaleti, 129–34. (22) See Orhonlu, Habeş Eyaleti, 85 ff. on the contraction of Habeş. (23) Daniel Crecelius, ‘Egypt in the eighteenth century’, in M. W. Daly (ed.), The Cambridge History of Egypt, II: Modern Egypt, from 1517 to the End of the Twentieth Century (Cambridge: Cambridge University Press, 1999), 59–70. (24) J. Spencer Trimingham, Islam in the Sudan (London: Cass, 1949), 102–6. (25) Intisar Soghayroun Elzein, Islamic Archaeology in the Sudan (Oxford: British Archaeological Reports, 2004), 18–19, 121–6. (26) Orhonlu, Habeş Eyaleti, 74–5, 98–9. (27) William Y. Adams, Kulubnarti I (Lexington: University of Kentucky Press, 1994); William Y. Adams and N. K. Adams, Kulubnarti II (London: Sudan Archaeological Research Society, 1998). (28) Martin Hinds and Hamdi Sakkut, Arabic Documents from the Ottoman Period from Qasr Ibrim: Texts from Excavations (London: Egypt Exploration Society, 1986); Martin Hinds and Victor Ménage, Qasr Ibrim in the Ottoman Period: Turkish and Further Arabic Documents (London: Egypt Exploration Society, 1991). (29) See Chapter 24 below, Michael Mallinson et al., ‘Ottoman Suakin 1541–1865: lost and found’. (30) Ehud R. Toledano, State and Society in Mid-Nineteenth-Century Egypt (Cambridge: Cambridge University Press, 1990), esp. 77–87. (31) On this see Muhammed H. Kutluoğlu, The Egyptian Question (1831–1841): The Expansionist Policy of Mehmed Ali Paşa in Syria and the Reaction of the Sublime Porte (Istanbul: Eren, 1998). (32) Khaled Fahmy, All the Pasha’s Men: Mehmed Ali, his Army and the Making of Modern Egypt (Cambridge: Cambridge University Press, 1997), passim. (33) On the conquest, see Richard Hill, Egypt in the Sudan, 1820–1881 (London: Oxford University Press, 1959), 7–21. (34) Orhonlu, Habeş Eyaleti, 146; Hill, Egypt in the Sudan, 75. (35) Holt, A Modern History of the Sudan, 58.

Page 8 of 9 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Frontier Policies in North-East Africa, 1517–1914 (36) Ibid., 70; Hill, Egypt in the Sudan, 137. (37) Holt, A Modern History of the Sudan, 55; Hill, Egypt in the Sudan, 22. (38) Hill, Egypt in the Sudan, 23, 45–7. (39) Orhonlu, Habeş Eyaleti, 148, 156–8. (40) J. S. Trimingham, Islam in Ethiopia (London: Frank Cass, 1952), 120–1. (41) Hill, Egypt in the Sudan, 136–40. (42) John Pollock, Gordon (London: Constable, 1993), 138–42; Mekki Shibeika, ‘The expansionist movement of the Khedive İsmail to the Great Lakes’, in Yusud Fadl Hasan (ed.), Sudan in Africa (Khartoum: Khartoum University Press, 1971), 142–55. (43) Hill, Egypt in the Sudan, 141–2. (44) M. W. Daly, ‘The British occupation, 1882–1922’, in Daly (ed.), The Cambridge History of Egypt, II, 245–6. (45) Hill, Egypt in the Sudan, 43, 125–6. (46) Ibid., 139–40. (47) See in Chapter 26 below, Paul Lane and Douglas Johnson, ‘The archaeology and history of slavery in South Sudan in the nineteenth century’. (48) Mervyn Hiskett, The Course of Islam in Africa (Edinburgh: Edinburgh University Press, 1994), 71–3. (49) For a recent analysis which summarises the traditional ‘European’ interpretation of nineteenth-century events in north-east Africa but which fails to discuss the significance of Ottoman support for, and attempts to control, Mehmed Ali and Isma‘il Khedive’s extensions of the empire’s frontiers there, see Efraim Karsh, Islamic Imperialism (New Haven: Yale University Press, 2006), 99–103.

Access brought to you by:

Page 9 of 9 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Continuities in Ottoman Centre–Periphery Relations, 1787–1915

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Continuities in Ottoman Centre–Periphery Relations, 1787–1915 FREDERICK ANSCOMBE

DOI:10.5871/bacad/9780197264423.003.0012

Abstract and Keywords In the political history of the Ottoman Empire, the long nineteenth century (1789–1915) stands out as a period of far-reaching, rapid change in the nature of the state. While the persistence of old practices should not be assumed along all frontiers of the empire, where it was applied the mutual support arrangement worked reasonably well at both ends of the nineteenth century. The two cases examined in this chapter illustrate this in a surprising fashion. The parallels are unexpected because among the notables involved, Tepedelenli Ali Pasha (1787–1820) in Epirus (Greece and Albania) and the Al Sabah and Al Thani shaykhs (1870–1915) in eastern Arabia carry reputations as unwilling subjects who rebelled against the sultan. It was largely due to the centre's failure to continue to uphold its part of the mutual support arrangement. Keywords:   Albania, Greece, Arabia, Tepedelenli Ali Pasha, Al Thani shaykhs, Ottoman Empire, long nineteenth century

IN THE POLITICAL HISTORY OF THE OTTOMAN EMPIRE, the long nineteenth century (1789– 1915) stands out as a period of far-reaching, rapid change in the very nature of the state. At the beginning of the period, Ottoman government was still pre-modern in nature, limited largely to the traditional tasks of waging war abroad, keeping the peace at home, administering justice and collecting taxes. A bureaucracy of perhaps 2,000 men at the centre managed state affairs. By 1908, the number of civil officials had risen to 35–50,000.1 The numbers had increased as the range of centralised state functions rose: Istanbul added responsibility for public services such as urban and communications infrastructure, education, health, and economic development to its established functions. The relationship between state and subject also changed, with the individual’s ability to avoid contact with the state becoming ever more restricted. At the end of this century of transformation, Istanbul was the dominant centre of the empire in a manner unimaginable in 1789, not only in its expansion beyond traditional fields but also in its ability to enforce its will. Yet despite these transformations, a comparison of administration of the important Albanian frontier early in the century (1787–1820) with the equally sensitive Arabian

Page 1 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Continuities in Ottoman Centre–Periphery Relations, 1787–1915 borderlands at the close of the period (1871–1915) shows surprising continuities in practice across Ottoman space and time. Along these and other parts of the periphery, the durability of features in the administrative model reflected a basic steadiness of purpose in policy-making throughout the century. The empire did not undertake its crash course in reform out of free choice: territorial losses and the sharp rise in European influence forced Istanbul to do whatever was necessary to limit opportunities for clashes with greater powers and, should conflict nevertheless occur, maximise the speed and efficiency with which it could tap the empire’s resources to mount an effective defence. This fundamental state interest in security was laid bare along the physical front lines, the borderlands. Except in frontier provinces most vulnerable to foreign attack or intercommunal strife, the (p.236) centre was content to forgo some of its modern practices applied in provinces closer to Istanbul, in order to boost the state’s bedrock interest as much as possible. In such areas the centre worked closely with local elites, nurturing loyalty to the sultan by stressing the bonds of faith and by the promise of mutual support: as long as Istanbul’s most basic interests were safeguarded, it would back the notables in their local interests, limiting its interference in local affairs.2 While the persistence of old practices should not be assumed along all frontiers of the empire, where it was applied the mutual support arrangement worked reasonably well at both ends of the nineteenth century. The two cases examined here illustrate this in surprising fashion. The parallels are unexpected because among the notables involved, Tepedelenli Ali Pasha (1787– 1820) in Epirus (Greece and Albania) and the Al Sabah and Al Thani shaykhs (1870–1915) in eastern Arabia carry reputations as unwilling subjects who rebelled against the sultan. Insofar as these leaders eventually grew disaffected, it was largely due to the centre’s failure to continue to uphold its part of the mutual support arrangement. Also surprising is the appearance in Arabia of some of the practices seen earlier in Albania and which the extensive programme of Ottoman reform was supposed to have abolished. The military, administrative, legal and cultural standardisation of the reform era touched the frontier only lightly.

Albania, 1787–1820 After being appointed governor of Ioannina in 1787, Ali Pasha came to dominate a long stretch of the Ottoman coastal frontier with Europe. His power base in Epirus was historically a conflict zone whenever the empire was at war with Venice, possessor of the Ionian Islands and four enclaves on the Epirote coast, Butrint, Parga, Preveza and Vonitsa. After Napoleon conquered Venice in 1797, the islands and coastal enclaves became an arena for Great Power competition.3 Only in 1819 did the struggle end, with (p.237) the islands formed into the United States of the Ionian Islands under British protection and the mainland ports under Ottoman control. Ali, as governor, was the main agent acting for the sultan. He also was defender of much of the rest of the Greek coast: at the peak of his career in 1812, he and his sons governed southern Albania, the Morea and much of mainland Greece (Figure 12.1). In Albania and northern Greece, the expansion of his authority resulted partly from victories in violent rivalries with other provincial leaders. As derbendat başbuğu, he controlled communications through highland passes of the Balkans. He held at times other very high titles, including the rank of vizier, the governorship of Rumeli and military commander (serasker) of the Balkans. He was a most noteworthy figure in Ottoman Europe. From then to today, Ali Pasha, ‘the Lion of Ioannina’ and ‘the Muslim Bonaparte’, has aroused more constant interest in Europe than any other Ottoman provincial notable except his contemporary (and fellow Albanian), Mehmed Ali, ‘the Founder of Modern Egypt’.4 As in the Page 2 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Continuities in Ottoman Centre–Periphery Relations, 1787–1915 case of Mehmed Ali, however, interest in Tepedelenli Ali derived not from his importance in Ottoman history but from his romantic image as an exotic strongman, able to rise above his decadent surroundings to defy the sultan, and looking to Europe for inspiration; indeed, he appears in noted works by leading Romantic authors, including Byron, Goethe, Dumas, Hugo and Balzac.5 Yet this fascination has fed almost exclusively on western literary sources and the views of scholars interested in fixing Ali’s place in post-Ottoman national history. (p.238) Given his position in a region straddling the Greek and Albanian borders today, Ali features prominently in multiple national and imperial histories. Scholars interested in Albanian history have portrayed him as the creator of a semiindependent state that included half the Albanian population (up to one-third of the residents of mainland Greece are reputed to have been of Albanian origin), a precursor of independent Albania.6 A significant detail is his seizure of power, ratified post factum by central authorities powerless to influence the course of events in Albania. His frequent contacts with foreign powers are taken as further evidence of his freedom from Istanbul’s control. Ottomanists are less certain of the degree to which Ali sought independence, but they tend to stress his regional dominance and violent campaigns as clear signs of the collapse of Istanbul’s authority7 It is Ali’s role in Greek history, however, that has consistently stirred the strongest interest.

Figure 12.1. Greece and Albania, showing the main places mentioned in the text in relation to modern political borders.

Most European visitors to Ali’s domains could be termed philhellenes, and curiosity about him since their time has fed on the bipolar nature of his reputed treatment of the Greeks.8 On the one hand, he was a cruel leader who readily used brute force against (p.239) not only external opponents but also those among his provinces’ population who challenged or displeased him; given the extent of his family’s authority, most Greeks lived under his draconian rule. Discontent over abuses, exemplified by his destruction of the Orthodox district of Suli, in addition to killings and extortion perpetrated on individual Greeks, fed the urge to fight for independence. On the other hand, Ali also is credited with wittingly or unwittingly aiding the way to Greek independence. Greek traders flourished, and Greek culture blossomed in the security provided by Ali’s absolute rule. Ioannina became the educational centre of Greece.9 Ali also encouraged the growth of demotic Greek as an administrative language. By using Greek as the language of his court, he connected his Greek subjects to issues of governance from which Greeks in other provinces were cut off10 Since Ali attacked Muslims as well as Christians, he is oddly admired for not caring about religion, except for ‘misguided dabblings’ in Bektaşi Sufism—a point also stressed by Albanian historians.11 Perhaps most importantly, Ali is recognised as an instigator of

Page 3 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Continuities in Ottoman Centre–Periphery Relations, 1787–1915 the Greek revolt.12 In so doing, he initiated the collapse of Ottoman rule in Europe and of the ancien régime throughout the empire. In effect, he was an anti-Ottoman revolutionary. What the majority of these various views of Ali share is acceptance of a received reputation that rests upon only a thin base of sound evidence.13 Elements of that reputation are probably accurate: he could be ruthless and an implacable foe, he was politically adept, and his ultimate interests were selfish (hardly a peculiar trait). Yet other images are more implausible. Ali has always been considered a man who cared little for religion, for example, but research in Ottoman records has shown that he was not ‘indifferent’ in matters of religion, and indeed was closely linked to several Sufi brotherhoods.14 Such a finding suggests that other elements of the standard accounts (p.240) emphasising his role in shaping the birth of independent states, his involvement in European diplomacy, and certainly his status as the ‘Muslim Bonaparte’, misjudge his place in the Ottoman milieu, including the nature of his often symbiotic relationship with the imperial centre. Ali undoubtedly caused frequent exasperation among officers of the empire, but he also took care to avoid causing any real rupture with Istanbul. When he ignored directives from the centre, he appears to have judged by the risk of damage to his position in Epirus. The common view is that Ali’s control over his domains was absolute, but this assumption underestimates the challenges he faced. Given his reputation for ferocity, he was not loved by the people of provinces that he controlled, but he and his exactions were tolerated because of the relative stability and security he provided, as well as the fear generated by that reputation. He nevertheless faced periodic armed challenges in a historically turbulent area, both within his territory and from neighbouring provinces, and banditry could only be controlled, not eradicated. Maintenance of security and retention of the loyalty of his network of supporters required constant expense, which Ali supported primarily by asserting control over land, tax farms (iltizam and mukataa) and trade monopolies. The nature of his financial affairs was reflected in the seizures made by the Ottoman authorities after his execution: his land holdings encompassed almost 1,000 çiftliks (privately-owned land) but his storied treasury held relatively little money15 Those directives from Istanbul that he ignored partially or fully were those that threatened his ability to repel any local challenge to his always insecure position; he remained fundamentally loyal to Istanbul, however, by refusing to endanger the state’s main interests in his provinces. That Ali Pasha remained responsibly Ottoman through most of his career in office is indicated by the lack of any Ottoman attempt to remove him from Epirus before 1820. This was not merely the result of the central state’s weakness; Istanbul did launch substantial military assaults upon several powerful provincial governors who did act against the fundamental interests of the empire. Buşatlı Kara Mahmud Pasha of İşkodra (Shkodër, northern Albania) and Osman Pasvanoğlu Pasha of Vidin (north-western Bulgaria) were noted governors who drew punitive military expeditions because of the damage they inflicted upon vital state frontier interests. Kara Mahmud interfered with revenue due to Istanbul, including income from properties of the royal family, but just as critical was the turmoil he promoted along the Montenegrin and Bosnian borders by his repeated raids.16 The violence he initiated harmed security in areas exposed to Austrian and Venetian forces and, through allied Montenegrin clans, Russia. Osman Pasha also threatened stability along the most important Ottoman (p.241) frontier, the Danube, by his own clashes with neighbouring notables and by supporting the rebellious Janissaries of Belgrade. He became the champion of the Janissary rank-and-file in opposition to Sultan Selim III’s Nizam-ı Cedid, the military corps developed on the European model and the array of taxes levied to

Page 4 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Continuities in Ottoman Centre–Periphery Relations, 1787–1915 finance it.17 It is important to note that even these ‘rebellious’ governors considered themselves loyal to the Muslim state and Ottoman royal family, but they did not uphold their part of the compact governing relations between the centre and the peripheries of the empire.18 Ali, by contrast, remained true. Ali’s loyalty and sense of duty was tried on occasion but not to the point of breaking, much as other Ottoman officers felt annoyance over his actions but did not move to oust him from Epirus. Perhaps most surprising was his and his sons’ acceptance of their progressive loss of control over parts of Greece, including the governorships of the Morea, Larissa (Ott. Yenişehir) and Trikala (Ott. Tırhala), beginning in 1812.19 When Istanbul dispatched its forces against Ali in 1820, they faced little real resistance from the pasha’s troops outside Ioannina, and his sons surrendered immediately. Even in 1821, besieged by Sultan Mahmud II’s army, Ali kept alive the hope of reconciliation with his sovereign and permission to retain his accustomed position. That he had encouraged the outbreak of the Greek revolt in no way negates this picture of his wish to continue as before within the Ottoman fold. While besieged in Ioannina, Ali issued an emotionally powerful proclamation to his fellow Muslims of the Balkans and, indirectly, to the sultan. The proclamation reveals much about his style of politics throughout his career, as well as his strategy in the final crisis of his long life. He opened with a stirring description of the threats facing not only him, but all Muslims, following the massacres perpetrated by the Christians in the Morea and elsewhere caught up in the Greek revolt: For so long now so many of our brothers in religion have camped on the stones of Ioannina, now totally stripped of all trace of care and cultivation—all for the sake of overthrowing me. I am grieved by this, especially at such a time when the unbelievers and followers of the Nazarene, the enemies of our religion, have thrown off all ties of obedience, risen up in rebellion and drawn their perfidious swords against the people of Islam. Their attacks upon the Islamic lands increase by the day. If our partners in religion are not grieved by the afflictions brought about by the Christians, if you are not saddened by the insults visited upon holy mosques, as I have seen, then you have no fear for the survival of the Islamic community. He asserted that everything he ever did was for the protection of Islam and its frontiers, conquering lands for the religion and never giving up anything in return. His first wish (p.242) was to serve the religion—his forefathers found martyrdom against enemies of Islam, and he was ready to do the same. He declared that he had served faithfully five sultans, including now the great lord over all Ottoman Muslims, Mahmud II. Ali closed by saying that he had always been generously forgiven for any indiscretions he might have committed, and that now he only asked for the same, and to be allowed to fight for the faith.20 Each of the points raised in the proclamation had enough basis in known fact to have been plausible.21 It also sheds light on Ali’s so-called rebellion. His intention in encouraging the Greek revolt was simply to create a diversion serious enough that Istanbul would restore him to his old responsibilities for the region. The tactic was not novel: Pasvanoğlu Osman probably acted in similar fashion in stoking rebellion among the Janissaries of Belgrade.22 Ali knew that the spectre of Christians killing Muslims would draw a strong reaction from Istanbul and Muslims across the empire, as the prosecution of Serbs who took up arms against the Janissaries in 1804 had shown. Unfortunately for him, however, the Serbian case was replayed insofar as Istanbul’s military turned on the Christians only after defeating the Muslims who had fallen from official

Page 5 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Continuities in Ottoman Centre–Periphery Relations, 1787–1915 grace. Ali was not forgiven and allowed to fight for the faith: although he surrendered on condition that his life be spared, he was executed on Mahmud II’s orders.23 Had Ali been turned loose upon the rebels, he surely would have attacked them quite ruthlessly. As noted above, he had the sensibilities of an active Muslim. European visitors who wrote of his ‘indifference’ to religion pointed to daily interaction between members of different faiths in Ioannina and at his court, and to the presence of synagogues and churches in the city—yet in a city supposed to have been overwhelmingly ‘Greek’, there were reported to have been nineteen mosques, five ‘grand (Sufi) tekkes’, two synagogues, but only six churches.24 Ali did not flout the sharia restrictions on construction of new churches in Muslim territory, and he would have also taken the orthodox Ottoman Muslim attitude towards Christian rebels. Indeed, Christians who (p.243) attacked Muslims in his provinces had suffered very harsh retribution.25 Ali would hardly have claimed in his proclamation that everything he had done was for the sake of Islam, had there not been some evidence to make it remotely plausible. Other details of his proclamation give some indication of what he felt it had been incumbent on him to deliver in his service to Islam and his sultan. Most important was the protection of the frontier, which he singled out for mention citing his recovery of Parga as evidence. Parga had passed into Ottoman possession in 1819 through negotiation with Britain, the power protecting the United States of the Ionian Islands, and Ali had been deeply involved in the proceedings. Ali also funded the property compensation to residents who quit Parga rather than submit to his control. But Parga was only the crowning achievement of his long-running campaign to gain control over at least all points on the coast, and preferably over the Ionian Islands as well. He took Butrint, Preveza and Vonitsa, first from the French in 1798, then from the Russians in 1806. He attempted an assault upon one of the islands, Levkas, but failed to take it; thereafter he built several forts to secure the coast opposite Corfu and the other islands. His ultimately successful campaigns against the inhabitants of Suli and Himara also firmed his control of the coast. Most of his diplomatic contacts with European powers aimed to achieve by policy what he could not gain militarily. While much has been made of those contacts, he never gave anything beyond words to any Christian power.26 Istanbul was aware of not only his efforts to secure the coast, including his attacks on Suli, but also his contacts with representatives of foreign powers.27 Ali was responsible for the frontier, after all, and as with any governor of a frontier far from the centre, Istanbul expected him to act as the situation demanded rather than to wait for detailed orders from the centre. In addition to his standing responsibility for guarding the frontier in Albania and Greece, Ali and his sons led large numbers of troops to other parts of the Balkans whenever Istanbul deemed them to be threatened. This included participation in campaigns against recalcitrant governors such as Kara Mahmud and Pasvanoğlu Osman. Ali had been first appointed to Ioannina in 1787 on condition of raising and leading troops to fight against Kara Mahmud.28 He routed Pasvanoğlu in 1797, dealing the governor of Vidin a serious blow but one that he survived. Ali also took part in the long siege of Vidin in 1798 before being recalled to Ioannina to address new security concerns (p.244) arising along the coast.29 In cases of conflict pitting the state against rebels, as in Belgrade province or against the Muslim ‘mountain bandits’ who plundered much of the Balkans in the reign of Selim III, or against external enemies, notably Russia, Ali and his sons could be counted upon to contribute forces to the Ottoman military effort.30 This service towards the state was as important as Ali’s care of the frontier under his charge.

Page 6 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Continuities in Ottoman Centre–Periphery Relations, 1787–1915 To explain why Ali’s services in supplying troops were so valuable, it is necessary to keep in mind the vital importance to the empire of provincial militias in the decades of the Janissary corps’ decrepitude. Albanians were much in demand and could be found throughout the empire in governors’ retinues and manning garrisons, in addition to fighting in the main army on active war fronts. Groups with similar social and topographic backgrounds on frontiers of the empire were also in demand, including Kurds, Laz, Turks of Dobrudja (straddling the eastern Romania– Bulgaria border) and Maghribis. Discipline among such soldiery was often minimal, but under their own commanders they could fight well. The reputation of Albanian soldiers is illustrated in the comments of an officer battling the mountain bandits. Writing at the end of the eighteenth century, he noted the particular value of three groups’ fighting skills. He praised soldiers from the regions of Canik and Trabzon (the Black Sea coast of Anatolia) and Deliorman (Dobrudja) (not coincidentally, all areas under the control of influential local notables), although the latter were difficult for their officers to control. ‘In the ways and stratagems of war, in bravery and steadiness, the Albanian soldier is above these other two. Using them is a difficult task, however. According to those who have commanded [them], learning their practices of service is a skill needed to gain their special talents.’31 The statement makes plain the tremendous potential value of Albanian soldiers to the empire at a time when it was fighting for its very survival. It also echoes the rationale for the wellestablished Ottoman practice of appointing Albanian officers to govern districts in Albania and nearby areas.32 Ali’s running feuds with some of his rival Albanian governors posed a threat to this key military recruitment system because of the combatants’ need to keep military resources adequate for their own defence, but (p.245) the centre’s usual tactic was to try to broker a truce between rivals when the empire faced a military crisis. This worked on occasion, albeit rarely for very long.33 Istanbul could at least tolerate this situation because Ali, having taken control of a province, was more reliable in sending to Istanbul both the taxes and the troops due from the territory than a less secure, minor notable usually was.34 The figures of troops expected from Ali’s domains thus rose over time, from 1,000 to 4,000 in the early years of his tenure in Ioannina to 10,000 to 40,000 at the peak of his power (1810–13). Ali’s career as governor of Ioannina lasted for thirty-three years because periphery and centre supported each other in the few matters considered crucial to the interests of each. Istanbul was frequently annoyed by, but overlooked, Ali’s oppression of rivals and subordinates, as long as frontier security and the rendition of troops and revenues were not lastingly imperilled. Ali knew, and accepted faithfully, the terms of the relationship. Unfortunately for him, it changed when the nature of Istanbul’s needs changed. Ali gained control over Parga, the last important coastal district still under foreign protection, in 1819, and the empire faced rebuilding rather than active conflict with enemies abroad and at home. Thus began the programme of centralising reform that would improve Istanbul’s ability to undertake the expensive project of military modernisation. Not recognising the changed circumstances, Ali continued his heavyhanded rule and harassment of districts not yet under his control.35 Unlike that other great provincial governor, Mehmed Ali of Egypt, who was extending Ottoman control in Wahhabi Arabia and the Sudan, Tepedelenli Ali no longer served interests of the centre that could compensate for the local disruptions he caused. After extended debate in the imperial divan, he was declared a rebel for disobeying orders, stripped of office, captured and killed. Centralisation was intended to make sure that his ilk could never rise to power in the provinces again.

Page 7 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Continuities in Ottoman Centre–Periphery Relations, 1787–1915 Eastern Arabia, 1870–1915 Eastern Arabia (Figure 12.2) was first under Ottoman administration from 1550 to 1670, but it was the second period under Istanbul’s control that provides the unexpected parallels with Ali’s Epirus. Midhat Pasha, the governor of Baghdad, reasserted Ottoman rights in al-Hasa (Ott. Lahsa, the Eastern Province of Saudi Arabia), Kuwait and Qatar in 1870–1. It could only be expected that such a leading figure of the Tanzimat would introduce a modern administration as described in the introduction to this chapter. The case of Ali in Epirus, where a local notable took charge of domestic security and frontier defence with the aid of irregular troops, and managed local affairs and revenue collection through local appointees, should not have been repeated after decades of modernisation. Yet in eastern Arabia, these ‘pre-modern’ features recurred, at least to some degree.36 (p.246) In Kuwait the relations between centre and periphery came closest to mimicry of the Epirus case. Istanbul was content to work through the leading local notable, the shaykh of the Sabah family, to achieve imperial security goals. Ottoman policy in Kuwait and the rest of eastern Arabia rested primarily upon one of the pillars supporting (p.247) Ali’s position in Epirus, concern for defence of the empire’s borders.37 British India had constructed a network of treaties with shaykhs on the coasts of the Gulf, binding them to promises to keep the maritime peace. Citing those treaties, India intervened forcefully in Bahrain and the nearby mainland in the late 1860s, Figure 12.2. The Gulf, showing the main eventually installing a shaykh in Bahrain places mentioned in the text in relation to favourable to its interests. Fearing the modern political borders. further extension of British influence, in 1870 Midhat asserted Ottoman rights to the best Gulf port remaining outside British control, Kuwait, but he gained Kuwaiti support for this through negotiation. In return for the Al Sabah’s promise that the Ottoman flag would fly over Kuwait and on Kuwaiti ships, and that the shaykh would uphold Ottoman rights in the territory under his influence, the empire would name him kaymakam and leave internal affairs in his hands. Kuwait would also be exempted from the usual responsibilities of an Ottoman district in the modernisation era, including taxes and customs, military conscription and the application of Ottoman law38 Despite tensions, this arrangement persisted with little alteration until the outbreak of the First World War. Between 1870 and 1896, the Al Sabah shaykhs upheld the Ottoman claim to Kuwait and on several occasions even contributed Bedouin forces and boats to military expeditions in eastern Arabia and southern Iraq. The tensions between centre and periphery occurred primarily after May 1896, when Mubarak al-Sabah killed the shaykh-kaymakam, his half-brother. Mubarak tried to persuade Istanbul to recognise him as the new kaymakam, but the centre dithered before upholding its part of the old agreement. Ottoman indecision delayed the

Page 8 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Continuities in Ottoman Centre–Periphery Relations, 1787–1915 appointment until the end of 1897, because some officials willing to accept the fait accompli faced others who wished to see the murderer of an Ottoman appointee punished. Indecision at the centre led to insecurity of tenure for Mubarak, and he consequently accepted a British offer of limited protection in 1899, when a rumoured project for a railway that would connect Kuwait to Russia alarmed India. In spite of the increasingly open secret of Mubarak’s relationship with Britain, however, the shaykh never fully reneged on the agreement negotiated by Midhat. The Ottoman flag flew over Kuwait until 1914, and in an agreement with London signed in 1913, Istanbul recognised Britain’s special relationship with Mubarak in return for Britain’s recognition that Kuwait was indeed Ottoman territory—the very goal of Midhat’s agreement of 1870.39 This arrangement, loose even by the standards operating in Tepedelenli Ali’s era, can be explained by the changes in international law that coincided with the Ottoman modernisation programme of the nineteenth century. The 1856 Peace of Paris, certifying the victory of the Anglo-French-Ottoman alliance in the Crimean War, (p.248) recognised the Ottoman Empire as a full member of the Concert of Europe and as an equal of the Christian European states in international law. According to that body of law, and as the signatory powers promised to enforce, Ottoman territorial integrity was thereafter guaranteed. The flag and the presence of a governor appointed by Istanbul marked Kuwait as Ottoman in international law, potentially the most effective defence mechanism available to the empire, at least in an area where no significant Christian population could spark an incident triggering European intervention. Similar aims of legally securing the remainder of the Gulf frontier caused Midhat in 1871 to bring the Ottoman flag to the remainder of the coast not yet part of India’s treaty system. In Qatar he established a relationship with Qasim, son of the elderly shaykh of the Al Thani family, that resembled his arrangements with the Al Sabah. The Ottoman tie with Qatar continued until 1915, although the relationship with Qasim eventually soured. Unlike in Kuwait, the empire later stationed a garrison in Qatar, which was subject to attack by shaykhs in Bahrain and Abu Dhabi who enjoyed British protection. Russia’s 1870 unilateral abrogation of the Peace of Paris clause mandating demilitarisation of the Black Sea also made Istanbul more wary of trusting the strength of legal symbols lacking armed guards. Unwilling to bear the cost of the garrison unaided, the state demanded tax revenues from Qasim, thereby starting a downward spiral of relations that accelerated when the garrison offered only limited support against Abu Dhabi. The low point came in 1892–3, when Ottoman attempts to resolve differences with Qasim triggered a battle with his supporters, a clash that the Ottoman army lost. Reconciliation followed, but thereafter the garrison remained to ensure that Qatar stayed under the Ottoman flag, and Qasim delegated contacts with Ottoman authorities to his son.40 In the territory lying between Qatar and Kuwait, Ottoman administration came closer to the reform era norm than in either of the shaykhdoms. Midhat designed an administration for alHasa that had features of provincial government in the Ottoman interior, although only the canonically accepted tithe and zakat were to be levied and no conscription attempted.41 His plans were never fully implemented, however, and practices seen in Epirus quickly resurfaced. By 1874 Istanbul withdrew from direct involvement in Hasawi affairs, appointing Shaykh Bazi‘ of the Bani Khalid tribe to the position of regional governor. To police his authority Bazi‘ built a military force of the kind predominating in Tepedelenli Ali’s time, a militia of Bedouin Arabs and other Muslims from tribal areas, including Kurds, Afghans and Baluchis—the kind of tough but undisciplined force that the centre had tried to eradicate from the time of Sultan (p.249) Mahmud II. A significant difference between Ali and Bazi‘, however, was the Bani Khalid

Page 9 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Continuities in Ottoman Centre–Periphery Relations, 1787–1915 shaykh’s weaker position in local society. He faced an uprising in 1875 that he could not quell without assistance from his brother-in-law, Shaykh Nasir of the Muntafiq tribe and governor of Basra in southern Iraq. Istanbul again attempted to work through a tribal shaykh, appointing Nasir’s son as regional governor in 1875; after he resigned in 1876, however, the only further appointment of a ‘local’ as regional governor came in 1902–4, when Sayyid Talib Pasha, brother of the naqib al-ashraf of Basra, held the position. Lower administrative positions, however, continued to be dominated by Hasawis and Iraqis, and in other respects the administration seemed a throwback to the prereform age. Istanbul maintained only sporadically a garrison of regular troops, leaving the onus of providing security primarily upon a gendarmerie of Kurds and local Arabs. This unit followed the model of the Hamidiye regiments formed in predominantly Kurdish frontier areas during the reign of Sultan Abdülhamid II (1876–1909).42 Also as in other relatively conservative, rural regions such as the Hijaz, local religious law took precedence over standardised (nizami) imperial codes, even though the dominance of Hanbali and Maliki interpretations left little room for the Hanafi school of law adopted ever more exclusively by the empire in the nineteenth century.43 Given the stress laid upon the religious duty of Muslims in the frontier area to maintain allegiance to the sultancaliph in Istanbul, the central regime’s decision not to disrupt local interpretation of sharia made sense.44 In economic matters as well, eastern Arabia was little affected by modernising steps of the reform era. The registration of land in the name of the cultivator and the issuance of title deeds, prescribed by the Land Law of 1858, never progressed far in al-Hasa, despite Midhat’s plans. Excise taxes continued to be collected there on both overland and sea-borne goods traded within the empire, although overland excise taxes were abolished empire-wide in 1874.45 The collection of excise and customs was farmed out as iltizam, another old tax practice nominally abolished in the reform era. The slave trade continued after growing restrictions were placed on it outside of Arabia—not only continued, but saw occasional participation by Ottoman officials.46 Perhaps the most striking indication of the internal autonomy on the frontier allowed (p.250) by Istanbul was the regional paymaster’s practice for some years of keeping records in Hebrew.47 The use of Greek in record-keeping at Tepedelenli Ali’s court seems quite unremarkable in comparison. Istanbul’s loose hand in al-Hasa achieved the centre’s chief goal, at least for several decades. Britain refused to be drawn into involvement in the region’s affairs, despite overtures from ‘Abd al-‘Aziz b. Sa‘ud, the Wahhabi amir of Najd (central Arabia). The end of the Ottoman era in alHasa and Qatar was caused more by fears of European designs on the empire than by British action. As experience in other provinces (the Balkans, eastern Anatolia, Tunisia, Libya, Egypt, Cyprus, Crete, Lebanon) sapped confidence in the legal security won in 1856, the urge to strengthen the Ottoman presence along the coast to demonstrate effective control drove plans for development. This, in turn, required diligent collection of local taxes. Three issues made this a problem. The concern for imperial defence rather than efficient administration caused tax assessments to be reviewed too infrequently. The excise tax inspection regime, maintained primarily to combat firearms smuggling, interfered with the locally vital import–export sector, as did occasional efforts to control the Saudis by halting trade with Najd. Most damaging perhaps was the failure to provide consistently adequate security in the interior. Bedouin raids that the undermanned garrison could not stop threatened both settlements and caravans in the last years of Ottoman rule, and when Ibn Sa‘ud attacked successfully in 1913, much of the local population accepted the change from mild but ineffective administration to more ideologically rigid but

Page 10 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Continuities in Ottoman Centre–Periphery Relations, 1787–1915 secure Wahhabi control. As Tepedelenli Ali knew, provision of security was desired as much by the local population as by the imperial centre.

Conclusion Comparison of the south-western Balkan frontier of 1787–1820 with that of eastern Arabia in 1870–1915 raises caution against overestimating the effects of nineteenth-century reforms. It reveals continuities in Ottoman management of centre–periphery relations: maintaining the security of the realm consistently shaped policies along the frontier, and where that could be done through local intermediaries, Istanbul was content with only the limited interference in local affairs needed to ensure that its security goals were achieved. The ambitious nineteenthcentury reforms carried out in core provinces were thus applied more lightly along some frontiers. Reforms were a vital element of Ottoman security plans, of course, but they targeted two major domestic problems, the efficient mobilisation of resources and the management of inter-communal tensions. Where applied, however, reform-driven change also aroused (p.251) opposition, made manifest in anti-Tanzimat agitation, inter-communal violence and calls for decentralisation. In some border regions, the gains from reforms were likely to be outweighed by the security risk arising from alienation of the population through far-reaching change. Just as the assumption that reforms affected most of the empire equally is misleading, it is necessary not to extrapolate the parallels seen in Albania and Arabia to all frontiers. Where conditions dictated, as in relatively rich border areas, those with significant Christian populations or on land frontiers most vulnerable to attack from a strong enemy, the centre did apply reforms. The Danube, site of Pasha’s most ambitious provincial reform, met all three of these criteria. Where the population was overwhelmingly Muslim and faced less threat of immediate, devastating attack, however, as in eastern and western Arabia, Jordan and parts of eastern Anatolia and Iraq, the frontiers experienced reform later and to a more limited extent. Note. I am grateful to Professor Victor Ménage for help with the Turkish sources. Notes: Proceedings of the British Academy 156, 235–251. © The British Academy 2009. (1) Donald Quataert, The Ottoman Empire, 1700–1922 (Cambridge: Cambridge University Press, 2005), 62. (2) Recent years have seen increasing publication of local historical studies that treat regions as provinces of the Ottoman Empire, rather than as national lands in their pre-independence, Ottoman imperialism’ period. Such studies have benefited the field in several ways, challenging the common assumptions of ‘national’ historians that local populations resented and resisted Ottoman domination and, among Ottomanists, that policy decisions made in Istanbul affected all provinces immediately and equally. The workings of the Ottoman state in its Asian provinces have received the best coverage. See, for example, Jane Hathaway with Karl Barbir, The Arab Lands under Ottoman Rule, 1516–1800 (London: Longman, 2008); Eugene Rogan, Frontiers of the State in the Late Ottoman Empire: Transjordan, 1850–1921 (Cambridge: Cambridge University Press, 1999); Frederick Anscombe, The Ottoman Gulf: The Creation of Kuwait, Saudi Arabia, and Qatar (New York: Columbia University Press, 1997); Jane Hathaway, The Politics of Households in Ottoman Egypt: The Rise of the Qazdaglıs (Cambridge: Cambridge University Press, 1997); Dina Khoury, State and Provincial Society in the Ottoman Empire: Mosul, 1540– 1834 (Cambridge: Cambridge University Press, 1997). For the Balkans, note Michael Hickok, Ottoman Military Administration in Eighteenth-Century Bosnia (Leiden: Brill, 1997).

Page 11 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Continuities in Ottoman Centre–Periphery Relations, 1787–1915 (3) Cf. Chapter 13 by Kahraman S¸akul in this volume. (4) The titles of works on Ali Pasha reflect western fascination: Richard Davenport, The Life of Ali Pacha, of Janina, Vizier of Epirus, Surnamed Aslan, or the Lion (London, 1822); Katherine Fleming, The Muslim Bonaparte: Diplomacy and Orientalism in Ali Pasha’s Greece (Princeton: Princeton University Press, 1999); William Plomer, The Diamond of Jannina: Ali Pasha, 1741– 1822 (London: Cape, 1970; 1st edn. 1936). (5) Dennis Skiotis, ‘From bandit to pasha: first steps in the rise to power of Ali of Tepelen, 1750– 1784’, International Journal of Middle East Studies, 2 (1971), 222. Fleming’s Muslim Bonaparte examines in detail the western fascination with Ali. (6) For the standard view, see Selami Pulaha, ‘Wissenschaftliche Forschungen über die osmanische Periode des Mittelalters in Albanien (15. Jahrhundert bis Anfang des 19. Jahrhunderts)’, in Hans Georg Majer (ed.), Die Staaten Südosteuropas und die Osmanen (Munich: Südosteuropa-Gesellschaft, 1989), 176. The northern Albanians were gathered under another semi-independent lord, Kara Mahmud Pasha of Shkodër. On Albanians in Greece, see Stavri Naçi, ‘Le mouvement autonomiste albanais du début du XIXe siècle et l’insurrection grecque’, Studia Albanica, 2 (1980), 57. (7) Quataert, The Ottoman Empire, 46, 49; Caroline Finkel, Osman’s Dream: The Story of the Ottoman Empire, 1300–1923 (London: John Murray, 2005), 401–3, 429; Robert Mantran, ‘Les débuts de la question d’orient (1774–1839)’, in Robert Mantran (ed.), Histoire de l’empire ottoman (Paris: Fayard, 1989), 430; Stanford Shaw, Between Old and New: The Ottoman Empire under Sultan Selim III, 1789–1807 (Cambridge, Mass.: Harvard University Press, 1971). (8) By far the best detailed study of Ali examines him from the perspective of the Greek revolt of 1821: Dionysios Skiotis, ‘The Lion and the Phoenix: Ali Pasha and the Greek Revolution’ (Unpublished Ph.D. thesis, Harvard University, 1971). The most recent book, Fleming’s Muslim Bonaparte, also looks at him from the modern Greek viewpoint as much as the western European. (9) Fleming, The Muslim Bonaparte, 64–5. (10) Ibid. Ali did keep records in Greek, including defters with financial details; BOA, Cevdet Dahiliye (C.DH) 5310 (1819). Subjects who did not know Turkish nevertheless were never barred from submitting petitions and other documents in their own languages, which were translated by official dragomans. (11) Fleming, The Muslim Bonaparte, 23–4; Skiotis, ‘From bandit to pasha’, 221–2. Fleming, however, later ably discredits the European image of Ali as a bad Muslim (ibid., 145–9), and Skiotis elsewhere (‘The Lion and the Phoenix’, 6) draws a subtle distinction between Ali’s tolerance of non-Muslims and the notion of his being indifferent to religious matters. (12) See in particular Dennis Skiotis, ‘The Greek revolution: Ali Pasha’s last gamble’, in Nikoforos Diamandouros, John Anton, John Petropoulos and Peter Topping (eds.), Hellenism and the First Greek War of Liberation (1821–1830): Continuity and Change (Thessaloniki: Institute for Balkan Studies, 1976), 97–109. (13) Skiotis’s works are the obvious exception.

Page 12 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Continuities in Ottoman Centre–Periphery Relations, 1787–1915 (14) Insofar as Ali is supposed to have had any religious sensibilities, he has been pictured as a Bektaşi. The stereotype of Bektaşism fits the competing ideal histories of Ali, since Albanians characterise it as an Albanian’ sect, while Greeks and western Europeans tend to view it as a heterodox belief, almost as close to Christianity as to Islam. Nathalie Clayer has argued convincingly against Ali as Bektaşi, finding in Ottoman records strong evidence of his close relationship to the Sa‘di, Halveti and even Nakşibendi brotherhoods. Nathalie Clayer, ‘The myth of Ali Pasha and the Bektashis: the construction of an “Albanian Bektashi national history”’, in Stephanie Schwandner-Sievers and Bernd Fischer (eds.), Albanian Identities: Myth and History (Bloomington: Indiana University Press, 2002), 127–33. (15) Skiotis, ‘From bandit to pasha’, 221; Fleming, The Muslim Bonaparte, 46–54; K. W. Arafat, ‘A legacy of Islam in Greece: ‘Ali Pasha and Ioannina’, Bulletin of the British Society for Middle Eastern Studies, 14 (1987), 178. (16) Frederick Anscombe, ‘Albanians and “mountain bandits”’, in Frederick Anscombe (ed.), The Ottoman Balkans, 1750–1830 (Princeton: Markus Wiener, 2006), 100–1; Hickok, Ottoman Military Administration, 153–75. (17) Rossitsa Gradeva, ‘Osman Pazvantoğlu of Vidin: between old and new’, in Anscombe (ed.), The Ottoman Balkans, 115–61; Robert Zens, ‘Pasvanoglu Osman Paşa and the Paşalık of Belgrade, 1791–1807’, International Journal of Turkish Studies, 8 (2002), 89–104. (18) Hickok, Ottoman Military Administration, 174; Zens, ‘Pasvanoğlu Osman Paşa’, 103. (19) Arafat, ‘A legacy of Islam’, 178. (20) BOA, C.DH 14436 (1821). (21) Ali’s grandfather died a martyr while attacking the Venetian stronghold of Corfu in 1715. Skiotis, ‘From bandit to pasha’, 226. (22) Zens, ‘Pasvanoğlu Osman Paşa’, 89. Another example of seeing career opportunities in trouble-spots can be seen in 1768–9 when Ahmed Cezzar, ex-mamluk and future governor of Sidon, tried to use his knowledge of Egyptian affairs to persuade the Porte that he was uniquely qualified to put down mamluk-inspired unrest there. Percy Kemp, An eighteenth-century Turkish intelligence report’, International Journal of Middle East Studies, 16 (1984), 497–506. (23) Vakanüvis Esad Efendi Tarihi, ed. Ziya Yılmazer (Istanbul: Osmanlı Araştırmaları Vakfı, 2000), 711. The description of Ali’s execution appears in the addendum by Bahir Efendi. Ali’s sons and grandsons had surrendered in 1820, and had been pardoned and exiled to Anatolia, but five of them were soon executed for writing ‘secret letters’: BOA, C.DH 8876 (1821). At least one of his other grandsons later became an Ottoman official, however: BOA, C.DH 9910 (1831). (24) Arafat, ‘A legacy of Islam’, 177, quoting Ibrahim Manzur, a French renegade more intimately acquainted with Ali and Ioannina than other European visitors. (25) On the campaign against Suliote ‘rebels’, see BOA, C.DH 5256 (undated). Several leaders of a Christian uprising in 1808–9 were reportedly executed in barbarous fashion by, for example, being beaten to death with sledgehammers.

Page 13 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Continuities in Ottoman Centre–Periphery Relations, 1787–1915 (26) For an example of Ali’s unwillingness to surrender anything in negotiations with Europeans, see C. de Bosset, Proceedings in Parga and the Ionian Islands (Chicago: Obol International, 1976), 78. Ali adroitly refused British demands for justice against Albanian border guards accused of shooting an officer who strayed across the frontier while Parga was still in British hands. (27) See for example BOA, Hatt-ı Hümayun (HAT) 1510 (1806). Regarding attacks on Suli, Ali sent heads of Christian rebels to show the Porte that their revolt was being suppressed: HAT 3414 (1804). (28) BOA, C.DH 3815 (undated, apparently from 1787). (29) Zens, ‘Pasvanoğlu Osman Paşa’, 97–9; C.DH. 204 (1798); BOA, Mühimme Defter 205, 10 no. 116 (1797) for the provincial levies of troops raised for the siege, of which Ali led the very significant Albanian contingents. Contrary to the impression given in studies of Ali (e.g. Fleming, The Muslim Bonaparte, 91–3), his movements to and from Vidin in 1798 were co-ordinated with Istanbul. (30) BOA, C.Askeriye (CAS) 4803 (1790); HAT 6517 (1800); CAS 3828 (1806); HAT 16734 (1810); HAT 16250 (1813). Ali and his son Muhtar mounted the most effective Ottoman campaign against the mountain bandits. Anscombe, ‘Albanians’, 105; Shaw, Between Old and New, 312–13. (31) BOA, C.DH 430 (undated, apparently from 1798–1801). This has been published in Yücel Özkaya, ‘XVIII. Yüzyılın İkinci Yarısına Ait Sosyal Yasantıyı Ortaya Koyan Bir Belge’, Ankara Üniversitesi Osmanlı Tarihi Araştırma ve Uygulama Merkezi Dergisi, 2 (1991), 303–34. (32) Frederick Anscombe, ‘The Ottoman empire in recent international politics, II: the case of Kosovo’, International History Review, 28 (2006), 784–5. (33) BOA, HAT 2058 (1790); HAT 15326 (1800–1); HAT 5218 (1802). (34) The British officer William Leake, who spent some years in Greece, noted that Istanbul benefited financially as well as strategically from Ali’s dominance of the region. Quoted in Fleming, The Muslim Bonaparte, 94. (35) Ali’s troublemaking forays were spreading into northern Albania; see, for example, BOA, HAT 26619 (1819–20). (36) The following discussion of Ottoman eastern Arabia draws significantly upon Anscombe, The Ottoman Gulf. (37) The first Ottoman foray into the region had been similarly shaped by security concerns, focused upon the Portuguese threat. Frederick Anscombe, ‘The Ottoman role in the Gulf’, in Lawrence Potter (ed.), The Persian Gulf in History (New York: Palgrave Macmillan, 2008). (38) For the details of this arrangement, see Frederick Anscombe, ‘The Ottoman Empire in recent international politics, I: the case of Kuwait’, International History Review, 28 (2006), 551–5. (39) Ibid., 557.

Page 14 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Continuities in Ottoman Centre–Periphery Relations, 1787–1915 (40) Anscombe, ‘The Ottoman role in the Gulf. Whereas 1913 saw the death of Qasim and Ibn Sa‘ud’s expulsion of Ottoman troops from Hasa, the garrison in Qatar clung to its position in Doha until 1915, when it surrendered to British Indian forces. (41) His plans were detailed in a multi-part report to Istanbul, BOA, Irade Dahiliye 44930 (1872). These have been published in Yusuf Halaçoğlu, ‘Midhat Paşa’nın Necid ve Havalisi ile İlgili Bir Kaç Lâyihası’, Istanbul Üniversitesi Edebiyat Fakültesi Tarih Enstitüsü Dergisi, 3 (1972), 149–76. (42) Martin van Bruinessen, ‘Kurds, States, and Tribes’. www.let.uu.nl/~Martin.vanBruinessen/ personal/publications/Kurds,%20states,%20tribes.htm (43) At the turn of the century the chief judge in al-Hasa was a member of a prominent Baghdadi family of Hanbali ulema; BOA, Şura-yi Devlet (S¸D) 2184/6, encl. 47 (1899). (44) The support for Shafi‘i law in Kuwait also fits this pattern. (45) BOA, Ayniyat 851/258–9 (1877); Frederick Anscombe, An anational society: Eastern Arabia in the Ottoman period’, in Madawi al-Rasheed (ed.), Transnational Connections and the Arab Gulf (London: Routledge, 2005), 24–5; Donald Quataert, ‘The age of reforms, 1812–1914’, in Halil İnalcık with Donald Quataert (eds.), An Economic and Social History of the Ottoman Empire, 1300–1914 (Cambridge: Cambridge University Press, 1994), 827. (46) Anscombe, ‘An anational Society’, 26–7. (47) BOA, S¸D 2157/22 (1882).

Access brought to you by:

Page 15 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars KAHRAMAN ŞAKUL

DOI:10.5871/bacad/9780197264423.003.0013

Abstract and Keywords This chapter attempts to analyse the shift in the Adriatic policy of the Ottoman Empire in the Napoleonic period. The focus is on the formation of the Republic of the Seven United Islands — the Ionian islands of Corfu, Paxos, Leucada, Cephalonia, Ithaca, Zante and Cythera — through active Ottoman and Russian intervention. Ottoman-Russian quarrels over the status of the Republic as well as the conflict between imperial realities versus local interests are integral to the understanding of the delicacies of running the Adriatic frontier in the Napoleonic period. When the Russo-Ottoman alliance shattered and the Ionian Islands were abandoned to the French together with the four coastal towns, the Ottomans once again resorted to the good old policy of cautious diplomacy; that is to say, the provisioning of the occupying force in Corfu. Keywords:   Ottoman Empire, Napoleonic period, Republic of the Seven United Islands, Russian intervention, Ottoman-Russian quarrels, Adriatic frontier, France, Ionian islands

THIS CHAPTER IS AN ATTEMPT TO ANALYSE the shift in the Adriatic policy of the Ottoman Empire in the Napoleonic period. The focus will be on the formation of the Republic of the Seven United Islands—the Ionian islands of Corfu, Paxos, Leucada, Cephalonia, Ithaca, Zante and Cythera—through active Ottoman and Russian intervention. Ottoman–Russian quarrels over the status of the Republic as well as the conflict between imperial realities versus local interests are integral to the understanding of the delicacies of running the Adriatic frontier in the Napoleonic period. In the eighteenth century this frontier was largely stabilised, as Ottoman warships as well as the Barbary corsairs were banned from entering the Adriatic Sea. The demise of Venice, however, as a result of the French aggression in the region made the Ottomans and the French neighbours in the Adriatic after the latter annexed the Ionian Islands by the Treaty of Campo Formio (17 October 1797) (Figure 13.1). This treaty troubled the Ottomans, for it left to victorious France only the barren Ionian Islands, while the most fertile lands of former Venice went to the Hapsburgs.1 Alarmed by the possibility that the French would use the islands as a stepping-

Page 1 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars stone to invade the Morea or Albania, the Ottomans were pressed to experiment with various frontier policies in response to the fluctuating military and diplomatic situation. Ottoman policies for controlling the Adriatic frontier ranged from cautious diplomacy to armed intervention. The Sublime Porte signed an alliance against Napoleon (1799–1807) with Russia— equally threatened by the French presence on the edge of the Balkans—in Istanbul on 3 January 1799, sending shock waves through France. This alliance underscored the Ottoman decision to take the initiative in the Adriatic frontier. The joint Russo-Ottoman fleet expelled the French from the Ionian Islands and this expedition resulted in the formation of the Republic of the Seven United Islands (1800–7) under Russo-Ottoman protection. While Tsar Paul lost his interest in the region as a result of his growing aversion to his British and Hapsburg allies, his successor, Tsar Alexander, as early as 1802 considered the Ionian Islands as a possible naval base for the coming Third Coalition War that in 1805 would bring together Russia, Britain, Austria and Portugal in an unsuccessful attempt to defeat France. This made it difficult for the Ottomans, who did not join the anti-French coalition, to enjoy their suzerainty over the Republic. With the Treaty of Tilsit (1807), France once again occupied the Ionian Islands, forcing the Ottomans to adopt their former policy of cautious diplomacy, which mainly consisted of turning a blind eye to the provisioning of the occupation force in the Ionian Islands from the Balkans.2 The short-lived Republic, nevertheless, protected the south-western frontier of the Ottoman Empire against France almost for a decade until the Treaty of Tilsit.3 (p.254) (p.255) The French expedition to Egypt has interested historians at the expense of the Adriatic frontier, even though the latter became one of the major preoccupations of Ottoman reformers. In fact, the growing French presence in the region had been the initial cause of alarm to the Porte and Russia even before the French occupation of Malta for the latter or Egypt, for the former.4 The first reaction of the Porte was to pursue a policy of cautious diplomacy towards France. At the request of General Chabot, the commander of the French troops in Corfu, the Porte decided to lend him money and allowed him to buy foodstuffs from Tepedelenli Ali Pasha of Ioannina in order to prevent a rupture with France.5 An Ottoman document suggests that before Napoleon’s Egyptian campaign, the Ottomans had played with the idea of purchasing the Ionian Islands from France so as to nip in the bud the French menace in the Adriatic frontier. This fragment of information is extremely interesting, because the Ottoman stratagem of buying land as a way of expansion had been

Figure 13.1. Map showing the borders of the Ottoman Empire, Venice and Dubrovnik. TSA, E.4004/7 (no date: before 1797?). The legend reads, ‘The borderlines of the Sublime State are shown in green colour. The borderlines of Venice are contoured by yellow colour, and the borderlines of Dubrovnik are contoured by blue colour.’ The

Page 2 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars four Dalmatian towns shown in yellow are abandoned by the early fifteenth century6 Butrinto, Parga, Preveza and Vonitsa from Obviously, events took on an entirely top to bottom. different complexion (p.256) with the French occupation of Egypt, which finally convinced the Ottomans to form an alliance with Russia and Britain. The participants resolved to send a combined Russo-Ottoman fleet under Admiral Ushakov to the Ionian Islands to be provisioned by the Ottomans, while Ali Pasha was charged with the occupation of four towns— Parga, Butrinto, Preveza and Vonitsa—situated along the Dalmatian coast facing Corfu.7 The joint fleet set sail to its destination on 1 October 1798, Russian warships passing through the Straits for the first time in history, marking Russia’s entry into ‘the Napoleonic wars through the Straits’.8 The Russo-Ottoman fleet occupied the islands by March 1799, the siege of Corfu lasting four months.9

How to Run the Adriatic Frontier: Diplomatic Problems Once the islands were occupied, the Ottomans and Russians began negotiations on the political status of the islands. The Russo-Ottoman Convention signed in April 1800 stipulated that the new Republic was to be subject to the Porte in the same manner as the Republic of Ragusa (Dubrovnik), and Russia would guarantee the constitution and the integrity of the new Republic. Four coastal towns on the Dalmatian coastal strip were ceded to the Ottomans with several exemptions and privileges similar to those enjoyed by the Principalities of Wallachia and Moldavia. Following the Ragusan example, the new Republic was to pay 75,000 guruş to the Ottoman Empire once every three years. However, this was a symbolic amount, equalling less than 1 per cent of the senate’s revenue.10 In Ottoman eyes, the Ionian Republic was little different from Dubrovnik and the Principalities— a ‘buffer protectorate’.11 The definition of the power relations between the Porte and the Principalities as suzerainty is a nineteenth-century invention, and notions of tributary, protection, vassalage and autonomy actually reflected different aspects of the same reality in Ottoman juridical and political terminology. The differences (p.257) between the Ottoman tributaries in their relations with the Porte lay above all in the historical, geographical, political, diplomatic and military circumstances of the time.12 The wording of the articles of the RussoOttoman Convention concerning the political status of the Ionians as well as the tribute largely derived from Ottoman-Islamic law and thereby resembles the wording of similar arrangements for other tribute-payers. The procedure of tribute-paying, the grant of an imperial diploma and a banner, and the bestowal of robes of honour on the Ionian delegation all symbolised the subjugation of the Republic to the Porte in Ottoman political culture.13 However, both the Ottomans and the Russians took advantage of the terminological confusion in order to achieve real political goals related to the Ionian Republic.14 From the Western or Russian perspective, the convention made the Ionian Islands virtually an independent state, like Ragusa, as it denied any real authority to the Porte and granted many privileges to the Ionians.15 On the other hand, the Ottomans considered that the status of the Principalities, the Ragusan Republic and the Ionian Republic was regulated by Islamic law according to which the tribute they paid brought about mutual rights and responsibilities in different degrees. Consequently, the definition of the political status of the Ionian Republic as identical with Ragusa signified virtual independence for Russia, but submission for the Porte because of the incompatibility of the Western and Ottoman terminologies. The feudal terms vassal and suzerain were unknown to Ottoman political terminology until the conclusion of the Russo-Ottoman convention.16 The Ottomans considered the term vassal to mean tabi‘ (political subject/

Page 3 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars dependent), mahkum (subject to a judge or ruler; that is, legally bound), and mahmi (guarded/ protected), while they took suzerain to mean metbu‘ (sovereign/liege-lord), hakim (judge/ruler), hami (protector).17 Furthermore, the constitution was a mere nizam-ı dahili (regulations concerning domestic affairs) as defined in (p.258) Article II of the Convention, while the aforementioned tribute was cizye (poll tax) and the Republic was cizye-güzar (poll-tax payer). The Republic should pay the tribute by a special envoy sent to the Porte, which symbolised the acknowledgement of allegiance on the part of the Republic in the same manner as the Principalities and Ragusa. Conversely, Russia considered the tribute a symbolic one, tailored to exempt the Ionian merchants from capitation (harac/cizye) and all other Ottoman taxes.18 Russia was strict about the implementation of the Ragusan model, as it understood it. For instance, it managed to halve the amount of the Ionian tribute, arguing that the tribute of Ragusa had also decreased.19 In late November 1804 the Porte was expecting the payment of 37,500 guruş by the Ionian Republic as tribute in the near future.20 The Ottomans tried to entrench their hold on the islands through several means, the most important of which was the granting of an imperial diploma (menşur/nişan-ı ali) to the Ionian deputy. From the Ottoman point of view the ‘constitution’ was limited to the organisation of the internal administration of the Republic, leaving the regulation of foreign affairs of the Republic to the diploma.21 Alarmed, Russia demanded that the diploma should be identical with the one presented to Ragusa, but the Ottomans rejected the demand on several grounds. Ragusa had obtained its diploma five centuries earlier with certain privileges. It had been an independent republic before the Ottoman rule, unlike the Ionian Islands. Moreover, it was a Catholic state, whereas the Ionians were of the Orthodox faith, implying that the new Republic naturally fell within the orbit of Ottoman sovereignty22 Therefore, the diploma presented to the Ionian Republic laid down obligations, although such diplomas usually only granted privileges. It stated that the Republic was a vassal state of the Ottomans and thus it should not aid the enemies (p. 259) of the Porte in any war. It should not shelter the Greek subjects either. On the other hand, it acknowledged the titles of the Ionian nobles, referring to them as bey and knez.23 Such arrangements concerning the political duties of the tributaries and titulature were common in the relations between the Porte and the tributaries.24 The Russians, nevertheless, considered the diploma a violation of the agreement, causing a rupture with the Porte for five years.25 Russian reservations about the diploma were based on the fact that it gave the Ottomans a legal pretext to interfere with Ionian affairs. When the Russo-Ottoman war broke out in 1806, the Porte demanded the Ionians remain neutral, as stipulated in the diploma. In reaction to the diploma crisis Paul neither ratified nor rejected the Constitution and the diploma. Thus, ‘by the terms of the Convention Emperor Paul was the guarantor of a constitution which he did not ratify’26 The ceding of the four towns to the Ottoman Empire caused consternation on the part of the Ionians, who were dependent for grain on the mainland, as the Ionians produced only a third of the food consumed in the islands.27 Trusting in friendly relations with Zante because of its dependence on the Ottoman mainland, local officials in the Morea hoped to induce the people of Zante to arrest the French garrison on the island even before the joint fleet entered the Adriatic.28 Through occupying these towns, the Porte hoped to extend its authority over the traditional foes, namely, the warlike tribes of the Tsamides (Çam) and the Suliotes. One of the issues the diploma raised was the question of fugitives. From the early Ottoman centuries onwards, these towns had served as a hiding place and a passageway to the Ionian Islands for the unruly inhabitants of Venetian Albania. The diploma strictly prohibited the Ionians from hiding the fugitives, while the Ionians adamantly denied such accusations.29 The agreement reached with the Porte on the towns was actually a favourable one; drawing on the example of Page 4 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars the Principalities, the Convention (Article 8) forbade Muslims to live there, while their inhabitants had full religious freedom.30 Furthermore, in order to prevent outside intervention, the diploma restricted the right of owning property in the four towns to Ottoman subjects, the people of the towns and the subjects of the Republic. Thus, when a Russian general (Micaniko?) attempted to buy the fishing enclosure and the land in its vicinity from its former owner, Ali Pasha, the (p.260) latter, who managed the enclosure as a tax-farm, strongly opposed this, basing his arguments on the diploma.31 The granting of a banner (Figure 13.2) was another symbol of the Porte’s relationship with its tributaries. The Ionian flag was designed in Istanbul with the consent of St Petersburg in such a way that it represented the subject status of the new Republic. The Ottomans stipulated that the flag would be a modified version of the lion of St Mark, which was the symbol of Venice. It should not feature any symbol of the French Republic in the French flag, such as a pillar and bouquet. Thus the flag featured seven arrows instead of pillars, representing the seven islands. Furthermore, in order to symbolise Ottoman status as the protector, a red strip contouring the margins was added to the flag and the hijri date 1214 (1800) was put on the upper left as the date of liberation of the islands from the French rule.32 Paul approved the design of the flag and proposed that the hijri date should be written against a background in red to increase the degree of symbolism, which became subject of mockery among the Ionians.33 (p.261) The Ottomans attached great importance to the Treaty of Amiens (27 March 1802), which they hoped would give international acknowledgement to the subjugation of the Ionian Republic to the Porte. However, the recognition of the Ionian Republic in the Treaty of Amiens was a mixed blessing for the Porte because of the wording of Article 9, which failed to mention the Ottoman suzerainty (metbuiyet) and the Ottoman annexation of the four coastal towns.34 The Porte urged, to no avail, that the article should be rephrased to

Figure 13.2. The Flag of the Republic of the take account of this.35 Nevertheless, it Seven United Islands. Başbakanlık Osmanlı Arşivi, Hatt-ı Hümayun Kataloğu 175/7600. never ceased to argue that its recognition of the islands as a republic in the treaty was linked to their subjugation to the Porte, as the Ionian Islands had been neither independent nor a state prior to the conclusion of the RussoOttoman Convention.36 The question of Ottoman suzerainty continued throughout the Napoleonic wars, as the Ionian Islands changed hands from one power to another. After the Treaty of Tilsit, Talleyrand promised the Ottoman ambassador that he would leave the Ionian Republic to the Ottoman Empire.37 By 1812, however, Corfu was under French control, while Zante and Cephalonia were occupied by Britain. Rumours concerning the possibility that the two powers would give the Ionian Islands to Russia alarmed the Porte, for Russia was likely to provoke the Greeks of the Morea and Anatolia against Ottoman rule, taking advantage of their shared Orthodox faith and the oppression of Ali Pasha of Ioannina in Albania.38 The Ottoman recognition of the British annexation of the Ionian Republic in 1815 came only in 1820, when the latter gave back Parga to the Ottomans.39 Even then, the Ottomans considered Britain and

Page 5 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars Russia usurpers, for they had taken control of the Ionian Republic without the consent of the Porte, its rightful protector, from the Third Coalition Wars onwards.40

(p.262) How to Run a Frontier: Local Interests versus Imperial Priorities As usual with frontier zones, imperial interests and local realities were in conflict, as can best be observed in the preparation of the constitution. For the ordinary populace, the Ottomans were barbarous infidels whereas Russia was the champion of Orthodoxy. They hoped that Paul I would keep the Ottomans away and liberate them from both the French and the old nobility, whom they regarded as collaborators with oppressive Venetian rule.41 Yet Paul was against territorial aggrandisement and favoured the restoration of the status quo ante in the region under Russian protection and Ottoman suzerainty. However, Russia failed to present a common front against the Porte. Poor communications between St Petersburg and its agents in the Ionian Islands forced the latter to take the initiative. The Porte, on the other hand, was prudent enough to send a diplomatic representative to Corfu, Mahmud Raif Efendi, who knew French and English, while the Russian admiral Ushakov was left alone with no knowledge of diplomacy and no skills in foreign languages.42 Thus Ushakov often contradicted the official Russian policy of favouring nobles since he realised that the exclusive rule of the hated nobility could not maintain order in the Republic, and he supported the limited participation of the notables in parliament.43 Nevertheless, Istanbul and St Petersburg considered the nobility their natural allies, while suspecting that the ordinary populace was affected by revolutionary ideas. Consequently, Ushakov’s provisional constitution (May 1799), which favoured the notables over the nobles, was considered scandalous by the imperial centres.44 The nobility also found Ushakov’s provisional constitution too liberal and turned to the Porte. Two Ionian deputations were to be sent to Istanbul and St Petersburg for the ratification of the Ushakov constitution. Ushakov chose the deputies from among his own supporters, while Abdülkadir Bey, the commander of the Ottoman fleet in the Adriatic, sent pro-Ottoman deputies to the Porte. Backed by the British, the two allies agreed to prepare another constitution in Istanbul with the active involvement of the pro-Ottoman deputies sent to the Porte. The new constitution restoring the legacy of old Venetian rule was prepared in March/April 1800.45 The Ionian peasants, notables, and the Russian officials in the Ionian Islands were totally frustrated with the new constitution and pejoratively called it ‘the Byzantine Constitution’.46 It is worth noting (p.263) that Ottomans systematically portrayed the supporters of Ushakov’s provisional constitution as ‘Jacobins’ to manipulate St Petersburg,47 and sought the participation of Russia in the preparation of the new constitution in order to prevent any Russian objection in the future.48 Both the new constitution and the aforementioned diploma became a major concern to the Russian officers as well as the commoners, causing various riots and disorders, especially in Zante and Cephalonia. Ottoman officials in Corfu suspected that the Russian officers were the main instigators of these upheavals.49 Tomara, the Russian ambassador to Istanbul, represented official Russian policy as opposed to Ushakov’s improvisations. In a letter addressing the Ionian senate, he asserted that the disorders stemmed from the populace’s ignorance about the RussoOttoman Convention. Assuring the senate that the Republic was fully recognised as a state in the Convention, he urged the senators not to be concerned about the diploma since the international status of the Republic was guaranteed by all the European powers. He tried to appease the Ionians by claiming that the political status of the Republic was far superior to that of Ragusa, while the four towns had more privileges than the free towns in Germany50

Page 6 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars The security situation deteriorated such that at one point, the Ottomans had to invite British ships to put down an uprising in Zante, where the mob flew the British flag in protest at Ottoman suzerainty51 The Ottomans hoped the presence of the British navy would tip the balance in favour of the Porte in the power struggle with the Russians, who, the Porte suspected, backed the ‘mob of Jacobins’. Furthermore, British troops could also be deployed in the Morea after suppressing the rebellion in the Ionian Islands against the menacing French.52 The Porte took a number of measures to end the disorders, ranging from a general amnesty to using the services of the Orthodox Patriarchate to pacify the islanders.53 The rebels handed back the senate to the senators after negotiations in which Mustafa Reşid Efendi, the superintendent of Corfu, took an active role.54 Nevertheless, two years after its ratification, the Ionians modified the constitution so as to secure a limited participation of the notables. The Ottomans were attentive to winning the hearts of the Ionians, as they were well aware of the military support they had provided to Russia against the Ottoman Empire throughout the eighteenth century. Over 4,000 Greeks and Albanians served in the Ionian units between 1799 and 1807, many of which fought against the Ottomans after (p.264) 1806 at Rusçuk (Ruse) and Vidin.55 Officially, St Petersburg and Istanbul wanted to present themselves as ‘liberators’ of the Ionians from the hated French. The Ottomans enlisted the support of the Greek Patriarchate to reconcile the Ionians with the Ottoman cause. In order to secure the co-operation of the Ionians, the Porte printed and distributed 400 copies of the Orthodox Patriarch Gregory’s Greeklanguage proclamation that incited Orthodox fervour against the Catholic French.56 The election of the archbishop was another battleground between the Ottomans and the Russians where the Patriarchate was actively involved in favour of the Porte. A Protopappas had been in charge of the religious affairs of Corfu for the last five centuries, but Ottoman suzerainty brought the Republic within the Orthodox Patriarchate of Istanbul. At the request of the Corfiotes, Patriarch Gregory raised Corfu to an archbishopric with the encouragement of the Porte. An election was held in June 1799, and the Protopappas, Giorgio Calichiopoulo-Manzaro, was elected Metropolitan. In this election, Abdülkadir Bey managed to prevent the appointment of the Russian candidate as the Metropolitan. Despite Tomara’s opposition, the Metropolitan of Ioannina interfered with the process of election on the orders of the Porte.57 Manzaro’s sudden death necessitated another election, in which Ushakov secured the election of his candidate (Count Pietro Bulgari). The Patriarch, however, annulled his election allegedly on canonical grounds. In a fresh election on 8 February 1800, Ushakov’s candidate was defeated.58 Ottoman documents show how the Porte made use of the agents of the Patriarchate for informationgathering and rallying local support in the islands.59 Attempts to rally the Ionians to the Ottoman cause had limited success. Local officials of both sides presented the Corfiotes as their own sympathisers to their respective courts, and tensions were never lacking between them, especially over the control of the city. After the fall of Corfu, both the Ottoman and Russian flags flew over the citadel to symbolise the equal status of the allies, who agreed to deploy the same number of troops in the town until reaching a decision on the political status of the Ionians. Russian officers, however, were opposed to the Ottoman military presence within the city walls, on the pretext of protecting the Corfiotes from potential Ottoman aggression. Those persecuted by Ali Pasha during the occupation of the four towns from the (p.265) French had taken shelter in the islands. Their earlier campaign against the Albanians made it difficult to deploy Ottoman marines in Corfu. For the time being, the Ottomans appointed patrona Şeremet Mehmed Bey to command the 300 Ottoman troops in the town in compliance with the petition of the Ionians, who considered him reaya-perver, that is, well disposed towards the Christians.60 Mahmud Raif Efendi, the Ottoman diplomatic Page 7 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars representative in Corfu, strongly recommended the appointment of an able fortress commander to Corfu with a number of disciplined troops under his command. This officer should also be capable of ‘attracting the people to the Ottoman side and reconciling them with the Ottoman rule’, which, in his opinion, was of the utmost necessity61 The Porte accordingly sent to Corfu al-Hajj Mustafa Ağa as the fortress commander with roughly 700 troops. The Russian commander Borozdin, however, did not allow him to enter the fortress in January 1800. He eventually moved into the citadel with his troops and his retinue of fifty people as a result of the strong protests of the Corfiotes against Borozdin.62 In addition to a fortress commander, the Ottomans sent Mustafa Reşid Efendi to Corfu with the official title of the Superintendent of Corfu (Korfa Nazırı) in September 1801.63 The Porte charged him with collecting intelligence on and around the Ionian Islands regarding the French and the Russians, while his official task was seemingly to oversee the storage and distribution of the foodstuffs sent to the Russian troops in Corfu by the Ottomans.64 Apart from friction with the Russian agents and the Ionian notables, the Ottomans also faced financial problems. The fortress commander Mustafa Ağa, the superintendent Mustafa Reşid Efendi, and the fleet commander were the main Ottoman agents in the Republic. While the Ottoman fleet in Corfu was maintained by the Porte, the expenses of the fortress commander and the superintendent were to be met by the Ionian senate.65 Nevertheless, the Ionians procrastinated about their financial obligations and alienated both of the Ottoman representatives in Corfu. In June 1801 the senate had offered to replace Ottoman and Russian troops in the town with native (p.266) troops, as it was beyond the Republic’s means to provision the foreign troops. As a result, Mustafa Ağa was required to disband the Ottoman fortress guards and to continue to stay in the town.66 Financial problems and the language barrier haunted both representatives, as their correspondence suggests. The aforementioned disorders that broke out in the Republic made it impossible to raise money from the Ionians to finance the Ottoman representatives. The Porte therefore undertook to pay Mustafa Reşid’s salary, which was often in arrears.67 Mustafa Reşid Efendi seems to have stayed in Corfu until at least February 1807, although he felt abandoned by the Porte throughout his term of office.68 Under these circumstances, the performance of the Ottoman representatives in Corfu is unclear. Nevertheless, the sources at our disposal suggest that they were not mere symbolic figureheads. For instance, the fortress commander Mustafa Ağa was in touch with Tomara, often contradicting Admiral Ushakov, while Mahmud Raif regularly corresponded with Nelson to keep the Russians in check.69 Tomara at one point advised the Porte to use Mustafa Reşid to restore order in the Republic during the rebellion on the grounds that he had amicable relations with the Corfiotes.70 A major problem between the Ottomans and the Ionians arose over Ottoman debts to Ionian merchants. The Ottomans undertook the transportation of 2,300 French prisoners-of-war from Corfu to Toulon and Ancona, and the cost of carriage was borrowed from Ionian merchants through the mediation of Ushakov71 The Republic’s envoy72 in Istanbul sent numerous petitions to the Porte for payment of the debt. Apparently the Porte could not pay this off until the end of 1804.73 The Ionian delegation sent to Istanbul had several goals in mind in addition to the establishing of the rule of the nobility in the Republic and the annexation of Venetian Albania. The expansion of Ionian commercial links into the Levant by obtaining commercial privileges as well as protection against Barbary corsairs and Ali Pasha were indispensable to the welfare of the burgeoning

Page 8 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars Republic.74 From the perspective of the Porte these were legitimate demands of a tributary state. In return, it considered (p.267) the repatriation of refugees from the mainland and the prevention of recruitment of the Ionians by foreign powers, that is, Russia, as prerequisites to the stabilisation of this frontier zone. Recruitment of Christian Albanians as well as Ionians into the Russian navy throughout the Napoleonic wars embittered the Ottomans, who kept a close eye on such Russian activities through Ali Pasha and other Ottoman local governors. As the Third Coalition was formed against Napoleon, Russian military preparations in the region increased to an alarming level for the Porte. Ottoman intelligence scrutinised the growing Russian military presence in Corfu supported by thousands of Ionians and Suliotes, who held Russia in high esteem for its military victories. To the dismay of the Ottomans, Russia offered each recruit twice as much the Porte.75 The Ottomans unofficially made known their reservations over the recruitment issue to the Russian ambassador.76 The status of the Ionian refugees in the Ottoman Empire constituted another dimension of Ottoman–Ionian relations. Many Ionians had moved to the Ottoman realms during the French invasion. France declared the Ionians French citizens after occupying the Ionian Islands. However, the Porte treated the Ionian refugees who fled the French as former Venetian subjects. The Porte required these refugees to pay capitation tax and thus become Ottoman subjects. Appreciating their value as able seamen, it also encouraged them to serve in the Ottoman fleet sent against the French.77 After the foundation of the Republic, the Porte exempted those refugees who had not already been listed in the head-tax registers from paying capitation.78 However, the local officials usually ignored such orders and demanded its payment from all the refugees.79 The refugees became a bone of contention between the Republic and the Porte, since the Ionian consulates allowed them to fly the Ionian flag by distributing patents as though they were Ionian subjects. The Porte, however, prohibited the issuing of patents to those refugees who had moved to the Ottoman realms before the date of the granting of the imperial diploma, and ordered local officials to protect the immigrants from the demands of the Ionian consulates.80 The Porte and the Republic, however, co-operated against those Ionians who flew the Ionian flag without the patent, as they caused trouble for both sides.81 (p.268) Ottoman suzerainty offered new prospects for the expansion of the Ionian trade thanks to their commercial privileges. The opening of the Black Sea to international trade was one of the major concerns of the Ottoman reformers in this era.82 The Ottomans allowed the merchants of a number of minor European states to enter the Black Sea with their own flag to prevent them from flying the Russian flag.83 On 22 August 1799 Ottomans and Russians agreed upon a new tariff tailored to increase the custom revenues of the Porte.84 The new Republic would initially pay 6 to 9 per cent customs to the Ottomans, like Ragusa, and had the right to participate in the Black Sea trade. By October 1801, however, the Porte expanded the Russian tariff so as to include the Republic, reducing the customs duties to 3 per cent for exports and imports and 5 per cent for the goods they bought and sold within the Ottoman borders.85 The figures suggest that the Ionian merchants benefited immensely from the new arrangement, after which the number of Ionian ships in the Black Sea rose to 100 (in 1804) from 37 (in 1803).86 Within a couple of years, the Republic had opened consulates throughout the Mediterranean islands as well as in important commercial towns such as Izmir, Salonica, Aleppo, Acre and Alexandria.87 However, the local Ottoman officials resisted the extension of the privileges of Russian merchants to the Ionians, for it would cause a loss in their revenues. The

Page 9 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars Ionian merchants not only complained about the arbitrariness of local officials but also argued that the new tariff allowed them to participate in domestic maritime trade without having to pay the higher internal customs.88 This, coupled with the demand for capitation by force despite the Porte’s prohibition, was the most common problem the Ionians faced in the Ottoman realms. The diploma required the Ottomans to protect the Ionian merchants from the raids of Barbary corsairs, who resumed their operations in the Adriatic as the ban on conducting piracy in the region was virtually lifted after the fall of Venice.89 Notably, orders sent to the Deys of Algiers and Tunis induced them to respect the Ionian flag as stipulated by the diploma. Nevertheless, piracy continued to be a problem in this (p.269) vulnerable frontier zone, causing protests from the Ionian envoy and the Russian ambassador.90

Conclusion Ottoman frontier policy in this region was based on creating a buffer-protectorate to keep in check French aggression. Such a protectorate seemed viable because of the lack of a common Ionian policy in Russia and Paul’s general lack of interest in Ionian affairs. Nevertheless, the Tsar’s assassination on 24 March 1801 meant a shift in Russian policies. Appreciating the strategic value of the Islands for the Third Coalition Wars, the new tsar Alexander I heavily reinforced Corfu with around 10,000 Russian troops before the conclusion of the Treaty of Tilsit. Thus, as we saw in the active Russian encroachment on the revision of commercial privileges, of the constitution, and of the tribute, Russia increasingly became the dominant power in the Ionian Republic by 1802. The resilience of an empire which was believed to be on the verge of disintegration is striking, although it was a costly policy. The Ottomans had agreed to furnish the necessary provisions to the Russian fleet in the Mediterranean including wine and rakı—the Turkish vodka—the cost of which was met by the Campaign Fund, which was created to finance jihad.91 The Ottoman decision to remain neutral in the Third Coalition Wars did not impress Russia, which still expected the Porte to continue to supply the Russian Mediterranean fleet with provisions. Questioning the benefits of this policy, the chief accountant pointed out that it had ruined the Ottoman finances in the last four years.92 In 1804 the Porte complained to the Ionian envoy that the Adriatic frontier cost the Porte several thousand purses of guruş annually.93 This was a huge cost for a state whose treasury revenues did not exceed 18,000,000 guruş,94 and amply illustrates the primary reason for the growing Ottoman resentment over Russian entanglement in the Mediterranean. With the Russo-Ottoman alliance shattered and the Ionian Islands abandoned to the French together with the four coastal towns, the Ottomans once again resorted to the good old policy of cautious diplomacy; that is, the provisioning of the occupying force in Corfu—first France and then Britain after (p.270) 1812—from the Ottoman mainland through the services of Ali Pasha, which it found cheaper and politically more expedient. Note. I would like to express my gratitude to the anonymous reader and Gerasimos M. Vlachos for their detailed comments and insightful suggestions on earlier drafts. I am fully responsible for all the remaining errors in fact and interpretation. I am also indebted to Andrew Peacock for his editorial suggestions. Notes: Proceedings of the British Academy 156, 253–270. © The British Academy 2009.

Page 10 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars (1) BOA, Hatt-ı Hümayun (HAT) 168/7123 (26 June 1797 to 14 June 1798): the report of Hasan Pasha, the Governor of the Morea, to the Porte; E. Z. Karal, ‘Yunan Adalarının Fransızlar Tarafından İşgali ve Osmanlı-Rus Münasebatı, 1797–1798’, Tarih Semineri Dergisi, 1 (1937), 111; B. Lewis, ‘The impact of the French Revolution on Turkey’, Journal of World History, 11 (1953), 120. (2) Ottoman policy of provisioning as a means of diplomacy was carried out through Ali Pasha of Ioannina, on whom there is a vast literature; see Chapter 12 by Frederick Anscombe in this volume. Historians usually misinterpret the contractual nature of the centre-periphery relations and the bargaining process involved in it as open defiance of the Sublime Porte by the local power-brokers: see Virginia Aksan, Ottoman Wars 1700–1870: An Empire Besieged (Harlow: Longman, 2007), 214–59. The Porte’s interests coincided with Ali Pasha’s ambitions in the Napoleonic period, who supplied most of the troops against France in the Adriatic and alternated his favour between France and Britain after the Treaty of Tilsit on the clear orders of the Porte, emerging as the major powerbroker in the region in the 1810s: see H. Sezer, ‘Tepedelenli Ali Paşa İsyanı’ (Unpublished Ph.D dissertation, Ankara University, 1995). (3) James L. McKnight, ‘Admiral Ushakov and the Ionian Republic: The Genesis of Russia’s First Balkan Satellite’ (unpublished Ph.D dissertation, University of Wisconsin, 1965), 3, 51–2. (4) Norman E. Saul, Russia and the Mediterranean 1797–1807 (Chicago: University of Chicago Press, 1970), ix, 59–62, 73–4; McKnight, ‘Admiral Ushakov’, 33. (5) BOA, HAT 176/7673; HAT 220/12179; HAT 171/7331 (late July 1797); HAT 1192/46903-C (26 July 1797). (6) BOA, HAT 176/7665 (1802?). (7) BOA, HAT 270/15756 (10 September 1798), minutes of the negotiations with Britain and Russia in the Bebek Conference held on 10 September 1798; McKnight, ‘Admiral Ushakov’, 32–5; the texts of the treaties are in Cevdet Paşa, Tarih-i Cevdet (Istanbul: Dersaadet Matbaası, 1309), 7: 304–11; BOA, Cevdet Hariciye Kataloğu (C. Hariciye) 4232 (3 January 1799), the treaty with Russia. (8) Saul, Russia and the Mediterranean, 69. (9) The troops of Ali Pasha were not allowed to set foot on the island to ‘win the hearts of the people’ (ahalinin celb-i kulubu), BOA, HAT 176/7676–A (1798–99), but the final victory came only after their participation. (10) McKnight, ‘Admiral Ushakov’, 223–4; the sum equalled 5,000 pounds sterling (18,750 francs). (11) This term is suggested by Panaite in order to avoid the confusion created by the uncritical application of Western terminology: Viorel Panaite, ‘Wallachia and Moldavia from the Ottoman juridical and political viewpoint, 1774–1829’, in A. Anastasopoulos and E. Kolovos (eds.), Ottoman Rule and the Balkans, 1760–1850: Conflict, Transformation, Adaptation (Rethymno: University of Crete, 2007), 43. For the case of Ragusa see Viorel Panaite, The Ottoman Law of War and Peace: The Ottoman Empire and Tribute Payers (New York: Columbia University Press, 2000), 154–5.

Page 11 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars (12) Panaite, Ottoman Law of War and Peace, 472–3. (13) Ibid., 356; Panaite does not deal with the Ionian Republic. (14) One of Selim’s counsellors favoured the implementation of the beylik model (the Principalities) over the Ragusan model in the Ionian Islands, assuming that Corfu was conquered by the Ottoman troops rather than by the Russian fleet: see TSA, E.4004/4 (n.d.) unsigned letter. This shows that the power struggle between the allies cannot be reduced to the field of terminological differences, albeit they provided a great leverage in diplomacy. (15) McKnight, ‘Admiral Ushakov’, 224. (16) The Russo-Ottoman Convention serves as a correction to Panaite’s statement that the Western term ‘suzerainty’ entered the Ottoman documents by the Treaty of Adrianople in 1829 in a study on the terminological problems in the literature on the Principalities: see Panaite, ‘Wallachia and Moldavia’ and Ottoman Law of War and Peace, 472–3. For the relevant article of the Convention refer to n. 17. (17) [Devlet-i Aliye] Cumhur-ı mezkurun sujereni yani hakim ve hami ve metbuı ve Cumhur-ı mezkur dahi Devlet-i Aliye’nin vasalı yani tabii ve mahkum ve mahmisi olmakdan naşi işbu himaye merasimi canib-i Devlet-i Aliye’den tamamıyla Cumhur-ı mezkur hakkına icra oluna. Article 1 of the Convention: BOA, HAT 176/7677 (3 April 1800); another reference to suzerainty is…vücuh ve asil ve erkan ve sergerdelerinin kadr ve itibarları tezayüdü içün lazım gelen vasallık istikmali emirlerine müsaade-i seniye…, C.Hariciye 2050 (25 October 1800), order sent to the Polltax Bureau [Cizye Kalemi]; Cevdet Pasha also defines the power relation between the Porte and the Ionian Republic as a suzerainty: see Tarih-i Cevdet, 7: 51. (18) Cumhur-ı mezkur canib-i Saltanat-ı Seniye’ye tabiiyetini işaren ve Devlet-i Aliye’nin metbuiyet ve himayetini itirafen tezayüd ve noksan kabul itmemek şartıyla ber-vech-i maktu üç senede bir kere 75.000 guruş cizye edasına muteahhid olub işbu virgü Dubrovniklünün virgüsi teslim olundukça vaki olan rüsum gibi Cumhur-ı mezkur tarafından dahi resmen elçiler irsaliyle Hazine-i Amire’ye teslim oluna. Ve meblağ-ı merkumdan mada Cumhur-ı mezkur ahar bir gune virgü veda itmeye. Ve reayası Dubrovnik Cumhuru reayası misillü Memalik-i Mahruse-yi Hakaniye’de cizye ve tekalif-i sairenin mecmuundan muaf ve müsellem olmağla…Article 4 of the Convention: BOA, HAT 176/7677; for the Russian perspective see McKnight, ‘Admiral Ushakov’, 224. On the ceremony of tribute-paying as an acknowledgement of allegiance, see Panaite, Ottoman Law of War and Peace, 208–10. (19) BOA, HAT 175/7655 (23 April 1803 to 10 April 1804). (20) C.Hariciye 2052 (27 November 1804); the Ionian envoy submitted the tribute in the presence of Selim III on 28 November 1804: M. A. Beyhan, Saray Günlüğü (1802–1809) (Istanbul: Doğu Kütüphanesi, 2007), 160. According to the correspondence of the Ottoman consul in Corfu, the Ionians were requesting a further reduction in the tribute as late as 3 November 1807; that is, after the Treaty of Tilsit (8 July 1807): see H. Baha Öztunç, ‘Yedi Ada Cumhuriyeti’ (unpublished MA thesis, Gazi Osman Paşa Üniversitesi, 2007), 39, 78. (21) McKnight contends that the Sublime Porte presented the imperial diploma when the convention and constitution were ratified in November 1800 (McKnight, ‘Admiral Ushakov’, 226), but the Ottoman documents suggest that it was appended to the Russo-Ottoman Convention together with the constitution on 3 April 1800: see BOA, HAT 176/7671 (1800–1), Page 12 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars BOA, Bab-ı Asafi Defterleri Kataloğu, Yedi Adalar Ahkam ve Berat Defteri (A.DVN.DVE 105/1), no. 4, pp. 3–5; C.Hariciye 1988 (3 April 1800); C.Hariciye 4691 (3 April 1800); C.Hariciye 2050 (25 October 1800) all contain the text of the imperial diploma (nişan-ı şerif-i alişan). (22) BOA, HAT 175/7599 (13 May 1801). (23) BOA, C.Hariciye 2050; A.DVN.DVE 105/1, no. 4, pp. 3–5. (24) Panaite, Ottoman Law of War and Peace, 340–51, 378. (25) McKnight, ‘Admiral Ushakov’, 221, 225–7. (26) Ibid., 37. (27) Ibid., 95, 175, 225. (28) BOA, C.Hariciye 2025 (22 August to 1 September 1799). (29) For Ali Pasha’s persecution of these fugitives, see P. Stathis, ‘From klepths and armatoloi to revolutionaries’, in Anastasopoulos and Kolovos (eds.), Ottoman Rule and the Balkans, 167–79. (30) BOA, HAT 176/7677. (31) BOA, Cevdet Maliye (C.Maliye) 22780 (5 May 1804); the Porte charged Ali Pasha with the running of the tax-farms in these towns in recognition of his services in the campaign. (32) Saul, Russia and the Mediterranean, 96–104; BOA, HAT 175/7600 (25 May 1800 to 13 May 1801). The Ionian flag was first published in İsmail H. Uzunçarşılıoğlu, ‘Arşiv Vesikalarına Göre Yedi Ada Cümhuriyeti’, Belleten, 1/34 (1937), 640. See also HAT 165/6904 (28 November 1800). (33) McKnight, ‘Admiral Ushakov’, 241–2; BOA, HAT 176/7662 (5 June 1799 to 24 May 1800); the stamp of the Ionian ambassador features only the lion of St Mark with seven arrows, see C.Hariciye 5178 (10 August 1803), C.Hariciye 1244 (12 August 1804). (34) BOA, HAT 949/40822–A (12 December 1814), HAT 950/40827 (17 June 1802): the texts of the treaty. The latter contains the internal correspondence of the British and French governments concerning the question of Malta. This contradicts the conventional view that the Porte was ill-informed about European diplomacy of the time. (35) BOA, HAT 169/7218–A (n.d. spring 1802): recommendations of Galip Efendi, plenipotentiary sent to Paris, and Seyid Ali Efendi, ambassador to Paris, on the relevant articles of the Treaty of Amiens. (36) BOA, HAT176/7675 (n.d., after Amiens). (37) BOA, HAT 143/5946 (23 July 1806). (38) BOA, HAT 966/41292 (5 February 1812). (39) BOA, C.Hariciye 742 (4 February 1818); BOA, Bab-ı Asafi Amedi Dosyaları (A.AMD) 42/32 (1820). (40) BOA, HAT 1239/48176-P (1818–20?). Page 13 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars (41) Clara J. Tucker, ‘The Foreign Policy of Tsar Paul I’ (Unpublished Ph.D dissertation, Syracuse University, 1965), 40–5, 316–20, 331; Roderick E. McGrew, Paul I of Russia, 1754–1801 (Oxford: Clarendon Press, 1992), 320–1. (42) McKnight, ‘Admiral Ushakov’, 211–12, 235. (43) Ibid., 123, 211–12, 235; Tucker, ‘Foreign Policy’, 225–6, 231. (44) McKnight, ‘Admiral Ushakov’, 66–8, 144–9. (45) Ibid., 150–8, 193–4; Count Antonio Capodistria, father of Ioannis Capodistria, elected governor of Greece (1828–31), was the most prominent pro-Ottoman deputy, see ibid., 174; Uzunçarşılıoğlu, ‘Arşiv Vesikalarına Göre Yedi Ada Cümhuriyeti’, 635–6. (46) McKnight, ‘Admiral Ushakov’, 223–4. (47) Ibid., 169–73, 194–9. (48) BOA, HAT 176/7671 (1800–1); HAT 175/7601 (1800–1). (49) BOA, HAT 156/6520–B (24 June 1800). (50) BOA, HAT 175/7602 (4 August 1800). (51) BOA, HAT 144/6033 (11 January 1802); HAT 176/7711 (1801–2). (52) BOA, HAT 261/15092 (1801–2); HAT 176/7712 (13 February 1802). (53) BOA, HAT 176/7716 (19 January 1802). (54) BOA, HAT 176/7675 (14 April 1802), Mustafa Reşid describes the inhabitants of Zante as untrustworthy. (55) N. Pappas, Greeks in Russian Service in the Late Eighteenth and Early Nineteenth Centuries (Thessalonica: Balkan Institute, 1991), 325–37; at least 6,500 ‘Greeks’ (Orthodox Christians) served in the armies of Russia, France and the Kingdom of Naples in 1807. Many more served in the French and British armies in the 1810s. For their contribution to the Greek Revolution see Stathis, ‘From klepths and armatoloi to revolutionaries’. (56) Similar proclamations (nasihatname varakları) were printed in Turkish, Arabic and French to rally Muslims against the French and to demoralise the French troops in Egypt: see K. Beydilli, Mühendishane ve Üsküdar Matbaalarında Basılan Kitapların Listesi ve bir Katalog (Istanbul: Eren, 1997), 16. (57) BOA, HAT 172/7378 (2 July 1799); also see Uzunçarşılıoğlu, ‘Arşiv Vesikalarına Göre Yedi Ada Cümhuriyeti’, 635–6. (58) McKnight, ‘Admiral Ushakov’, 218. (59) BOA, HAT 238/13221 (5 June 1799 to 24 May 1800); HAT 176/7661 (5 June 1799 to 24 May 1800).

Page 14 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars (60) BOA, HAT 162/6746 (5 June 1799); HAT 157/6536 (10 May 1799): Abdülkadir Bey’s correspondence with the Porte on the approval of the good conduct of the Ottoman marines by the Corfiotes; McKnight, ‘Admiral Ushakov’, 124–5. (61) BOA, HAT 158/6577-B (13 May 1799):…celb ve telif-i ahali ve beraya maddesi cümleden akdem idüği. (62) McKnight, ‘Admiral Ushakov’, 206–10; the Russian commander of Corfu refused to receive the Ottoman marines returning from Palermo, HAT 155/6513-M (4 October 1799). Mustafa Ağa was the former director of the registered property of Trikala/Tirhala (defter nazırı), BOA, C.Askeriye 25418 (2 February 1801). (63) BOA, C.Maliye 200/8283 (6 September 1801); C.Hariciye 1277 (27 November 1800). (64) BOA, HAT 172/7432 (11 February 1805). (65) The Ottomans maintained a fleet of three battleships with a crew of 700 men to patrol the Adriatic. Kürd Mehmed Kapudan was the commander of the first fleet that served in the region from the autumn of 1799 to spring 1801. A second fleet of the same size under patrona Şeremet Mehmed Bey replaced the first fleet and stayed in the Adriatic until spring 1804. Another fleet under riyale Hüseyin Bey took its place and stayed in the region until the end of 1805, BOA, HAT 163/6774 (10 October 1799), C.Bahriye 9494 (25 June 1801); C.Bahriye 1747 (29 December 1805). (66) BOA, C.Hariciye 1770 (13 June to 11 July 1801); his daily allowance was 27 guruş when he needed at least 100 guruş, C.Hariciye 648; later he asked to be excused from the post, C.Hariciye 643. (67) BOA, C.Maliye 655/26814 (30 May 1802), C.Hariciye 9196 (12 April 1804). (68) BOA, C.Maliye 19/887 (14 July 1807); C.Hariciye 3845 (7 February 1805). (69) McKnight, ‘Admiral Ushakov’, 125, 220. (70) BOA, HAT 259/14932 (6 April 1802). (71) BOA, C.Hariciye 1277 (27 November 1800); 24,055 real (Spanish?) (78,178.5 guruş 30 para), C.Hariciye 3365 (10 April 1802). (72) The Convention allowed the Republic to appoint consuls charged with overseeing commercial activities of the Ionians in the Ottoman Empire. Antonio Tomaso Lefcochilo, who was within the Ionian delegacy sent to Istanbul, was nominated officially as extraordinary envoy to the Sublime Porte (Inviato Estraordinario presso la Sublima Porta) with the duties of ambassador but without the official title. The Ottomans used the word elçi both for envoy and ambassador. (73) BOA, C.Hariciye 2052 (27 November 1804). (74) BOA, HAT 258/14875 (7 September 1799): Ali Pasha’s attack on two Ionian vessels. (75) BOA, HAT 159/6635 (15 August 1805) and HAT 175/7610 (15 August 1805): the correspondence of Ibrahim Pasha of Avlonya on the Russian offer of 20 to 25 guruş a month. Page 15 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars (76) BOA, HAT 151/6372 (16 September 1805): the report of Hüseyin Kapudan from Corfu; HAT 176/7680. (77) BOA, C.Hariciye 367 (12–21 September 1798); C.Hariciye 1491 (December 1798); C.Hariciye 2320 (2 November 1798). (78) BOA, C.Hariciye 4836 (20 January 1799). (79) BOA, C.Hariciye 1313 (28 March 1801). (80) BOA, C.Hariciye 3067 (mid-October 1801). (81) BOA, C. Hariciye 7774 (27 February 1804). (82) İdris Bostan, ‘Rusya’nın Karadeniz’de Ticarete Başlaması ve Osmanlı İmparatorluğu, 1700– 1787’, in İdris Bostan (ed.), Beylikten İmparatorluğa Osmanlı Denizciliği (Istanbul: Kitap Yayınevi, 2006), 285–325. (83) Kemal Beydilli, ‘Karadeniz’in Kapalılığı Karşısında Avrupa Küçük Devletleri ve “Miri Ticaret” Teşebbüsü’, Belleten 55/214 (1991), 687–755. (84) BOA, C.Hariciye 5963 (22 August 1799). (85) BOA, A.DVN.DVE 105/1, p. 93, no. 709; C.Hariciye 4554 (27 October 1801), the tariff agreement with the Republic. C.Maliye 64/2912 (13–23 April 1802); C.Hariciye 4384. (86) BOA, HAT 176/7749 (1803–4). (87) BOA, A.DVN.DVE 105/1, no. 709. (88) The taxation rates in customhouses ranged between 6.6 per cent and 20–25 per cent instead of the official 3 per cent. BOA, C.Hariciye 4384; C.Hariciye 5118 (15 April 1803); C.Hariciye 1244 (12 August 1804), C.Hariciye 1252 (25 October to 4 November 1804); C.Hariciye 5178 (10 August 1803). (89) McKnight, ‘Admiral Ushakov’, 232–3. (90) For similar cases, see BOA, A.DVN.DVE 105/1; C.Hariciye 3428 (20 July 1804); C.Hariciye 4986 (7–16 September 1804). (91) The Porte also delivered to the Russian fleet 600,000 guruş a year. It decreased to 120,000 guruş after Paul withdrew from the Mediterranean, BOA, C.Hariciye 7916 (27 March 799), C.Hariciye 2010 (20 August 1800), C.Hariciye 1586 (9 May 1801), C.Hariciye 1303 (15 August 801); the Ottomans procured olive oil, butter, biscuit, green beans, pounded wheat, chickpeas, salt, vinegar, wine, arak/rakı and firewood. (92) C.Hariciye 7863 (17 April to 11 July 1802), C.Hariciye 4708 (9 February to 18 April 1802); the Porte’s debt to Russia amounted to 389,351 guruş in 1806, C.Hariciye 1862 (28 January 1806). (93) HAT 175/7655 (1803–4); a purse contains 500 guruş.

Page 16 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Attempts to Control the Adriatic Frontier in the Napoleonic Wars (94) Y. Cezar, Osmanlı Maliyesinde Bunalım ve Değişim Dönemi (Istanbul: Alan Yayınları, 1986), 236–7.

Access brought to you by:

Page 17 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century GÖKHAN ÇETİNSAYA

DOI:10.5871/bacad/9780197264423.003.0014

Abstract and Keywords Throughout its four centuries of Ottoman rule, Iraq remained a frontier of geographical, tribal, religious, economic and imperial boundaries. Iraq was an outlying region; it had a large Shi'i population; it remained a tribal and economically poor country; as a frontier region, it was vulnerable to invasion and peaceful penetration by foreign powers, Iran and Britain. The Ottoman central government expended considerable effort to overcoming these challenges, but proved unable to resolve them completely, and as a result, both its authority and reform efforts were undermined. These obstacles to the Ottoman administration of the Iraqi provinces, due to Iraq's location in a frontier region, compounded by the government's inability to satisfy the conflicting desires and interests of those involved, presented a dilemma to the empire which it was unable to transcend. Keywords:   Iraq, Istanbul, Shi'I Iran, India, Britain, Ottoman administration

THROUGHOUT ITS FOUR CENTURIES OF OTTOMAN RULE, Iraq remained a frontier of geographical, tribal, religious, economic and imperial boundaries. Iraq was an outlying region; it had a large Shi‘i population; it remained a tribal and economically poor country; as a frontier region, it was vulnerable to invasion and peaceful penetration by foreign powers, Iran and Britain. These characteristics remained constant until the end of the Ottoman rule in Iraq, and they would prove to be significant challenges to the efforts of the Ottoman government to assert its rule over the Iraqi provinces. In the nineteenth century the main agenda of the Ottoman administration of Iraq was socioeconomic development and administrative reform. Implementation of this agenda was hindered by several factors which will be discussed in this chapter. Specifically, the administrative difficulties which became institutionalised at the provincial level because of the bureaucratic regulations and practices of the Tanzimat period will be explored. These regulations and practices led to further structural division among administrative bodies, and faced a precarious balance of local notables and Ottoman officials. Furthermore, tribal rivalries frustrated land and tax reforms initiated by the government,

Page 1 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century deepening the political and economic power struggle occurring at multiple levels in the Iraqi provinces. Compounding the challenges facing the Ottoman administration of Iraq was its location on the boundary of Shi‘i and Sunni sects, as well as its proximity to regional powers: Shi‘i Iran, and the British in India and the Gulf (Figure 14.1). These religious and imperial interests transcended Iraq’s political border: Shi‘i shrines and schools of jurisprudence in Iraqi cities; a growing Shi‘i population, which was interpreted as a threat to the Ottoman rule; and British India’s financial interference in Shi‘i religious sites, as well as British economic and political penetration. These threats to Ottoman administration, thanks to Iraq’s status on the frontier of varying local, regional and international interests, presented a significant and complex set of challenges to the Ottoman authorities in Istanbul. (p.272) Prior to the Ottoman conquest of Iraq in the first half of the sixteenth century, Iraq had already lost its importance as a source of wealth, as its irrigation system had been damaged during the Mongol invasion in the thirteenth century, and a great part of the country had passed under the control of pastoral tribes and tribal confederations. From the start of the Ottoman era, Baghdad was placed under the direct control of Istanbul. It was regarded as a centre for the defence of the frontier against Iran, and an important Ottoman garrison was stationed there. Under the supervision of an (p.273) Ottoman governor-general (beylerbeyi) at Mosul, Kurdish families around Kirkuk (Şehrizor) and Sulaymaniya were appointed as local governors or tax collectors, in Figure 14.1. Ottoman Iraq in the late nineteenth century. return for protecting the Iranian frontier. Basra kept its importance as a naval base temporarily, until the Portuguese and Dutch threats in the Gulf receded by the end of the seventeenth century. Basra also gradually declined as a centre of trade and commerce after the trade of the Indian Ocean moved from the Gulf to the Red Sea.1 Like other outlying provinces which joined the empire late, such as Egypt and the Yemen, Iraq was never fully integrated into the classical Ottoman administrative and land system, and the Porte did not maintain all-embracing political control there. Its control was further weakened by periodic warher weakened by periodic wars with Iran, which did not finally end until the early nineteenth century, and also by periodic Iranian occupations. Mosul alone was subjected to the timar system; Baghdad and Basra were administrated as salyane provinces. Notwithstanding the absence of timars, the bulk of the land in Baghdad and Basra was regarded as miri (state land); nonetheless, the land regime was complicated by old Islamic customs, including the extensive use of vakıfs, and also by the destabilising influences of widespread tribalism and endless wars with Iran. The provincial division of the region varied. Iraq was originally divided into three eyalets: Baghdad, Basra and Mosul; however, the province of Mosul at times lost territory to

Page 2 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century Diyarbakır and Şehrizor, and a separate province, called the eyalet of al-Hasa (Lahsa), was formed in Najd in the second half of the sixteenth century.2 The main cities of Iraq differed in their economic and social orientation, depending on their geographical position. The ties of Mosul were with Syria and Anatolia, Baghdad was linked to Iran and the western and southwestern deserts, and Basra mainly to the Gulf and India. In the eighteenth century, central authority grew weaker, and the vacuum of power was filled by local potentates, all owing allegiance to Istanbul: Georgian mamluks in Baghdad, and the Jalilis, a local family, in Mosul. For about a century, from 1747 to 1831, Baghdad, with Basra, was ruled by successive mamluk pashas. The Jalilis took control of Mosul from 1726 onwards, but power in Sulaymaniya was in the hands of the Kurdish Baban family, while other Kurdish districts fell under the control of local amirs.3 Not until the reign of Sultan Mahmud II (1808–39) did central government set (p.274) out to restore its authority over the provinces, and to produce a reformed and centralised system of provincial administration. The process was gradual: Baghdad was restored to central authority in 1831 and Mosul in 1834. However, the subordination of the Kurdish amirates around Diyarbakır and Rawanduz took several years more, the Babans of Sulaymaniya submitting as late as 1850.4 In the years immediately following the restoration of central authority in Baghdad, questions of security took precedence over reform. The authorities faced a number of local uprisings, and a standing threat from Mehmed Ali Pasha of Egypt, who had wrested control of Syria from the sultan between 1831 and 1839.5 As proclaimed in 1839, the Tanzimat reform programme promised a serious attempt to introduce reforms into Iraq. Even so, the Tanzimat reforms could not be introduced throughout the empire simultaneously, and the process in Iraq, like many other provinces, was gradual.6 For all the difficulties and deficiencies in the implementation of governmentsponsored reforms, it is clear that the Tanzimat era initiated a process of social and economic change in Iraq, which continued through the twentieth century. In so far as stronger central control produced better law and order, it served as a stimulus to trade and agricultural production. So did the partial settlement of the tribes, and the associated distribution of state lands commenced during Midhat Pasha’s time as governorgeneral of Baghdad (1869–72). Other important stimuli were the growth of Iraq’s foreign trade, particularly through the Gulf, and the development of modern communications, including telegraph lines, and steam navigation in the Gulf and on the Tigris.7

(p.275) Dilemmas of Reform in the Iraqi Provinces In the second half of the nineteenth and the early twentieth century, Ottoman officials, whether at the centre or in the Iraqi provinces, had reached a broad consensus with regard to the question of reform and development in Iraq. It was agreed in their numerous reports that the Iraqi provinces of Baghdad and Basra had considerable potential for agricultural development. The key to unlocking this potential lay partly in irrigation and marshland drainage, partly in better river and rail communications, and partly in the settlement of the tribal population, who should be encouraged to become peaceful cultivators. The process of settlement, it was further agreed, was in part a question of stronger administrative and military control, and in part a question of continuing the policy initiated in the 1870s of releasing state lands to the local population. More controversially, some officials argued that the success of these reform and development proposals would be enhanced by a fundamental administrative restructuring, with much greater powers devolved to governors-general, and with the whole of Iraq established as a single province.8

Page 3 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century Yet, for most of the period, the Sublime Porte failed to adopt these reform proposals. In the case of the ambitious plans for railway construction and irrigation, the problem was finance. Political considerations frustrated the proposals about administrative decentralisation and provincial autonomy, and the policy of distributing state lands. Furthermore, Iraq’s reform process confronted four major challenges in the second half of the nineteenth century: serious conflicts in provincial administration; continuous tribal unrest; and threats of Iranian (Shi‘i) and British penetration.

Chronic Conflicts in Provincial Administration As far as the provincial administration in Iraq was concerned, there were chronic conflicts among civil, military and financial high-ranking officials in Iraq in the second half of the nineteenth century. The persistence of these conflicts is best explained by the structure of the provincial administration system, as responsibility was divided: the vali (governor-general) was only responsible for civil administration; the local troops were under the command of the müşir (Field Marshal); and the finances of the vilayet were controlled by the defterdar (provincial director of finances). Furthermore, these three officials were responsible to different departments at the Porte: the vali to the Ministry of the Interior, the müşir to the Ministry of War, the defterdar to the Ministry of Finance. After the legal reforms of 1879, the judiciary also became separated from the (p.276) civil administration. The result was a permanent potential for conflict. Two more factors exacerbated this situation. In the first place, the Ottoman administration of the Iraqi provinces was based upon a delicate balance between civil and military officials and local notables. In all three provinces, the chief notable families, both Arab and Kurdish, enjoyed significant local influence. Local notables played a variety of roles in provincial administration, stretching from membership of the Administrative Councils to the holding of subordinate posts, like that of sub-governor. Inevitably, the notables brought their own differences and rivalries into the administration. As a result, senior officials found themselves drawn into local quarrels, sometimes on opposing sides. In the second place, the Palace entourage played a key role, and certain notables had connections and protection at the Palace, which strengthened their position vis-à-vis provincial officials, introducing a further element of conflict.9

Continuous Tribal Unrest in the Iraqi Provinces The second challenge was the continuous tribal unrest in Iraq. The majority of Iraq’s population had a tribal organisation. This tribal population included nomadic, semisettled and settled elements. The Ottoman authorities regarded the presence of this large tribal population as a major obstacle to good government in Iraq, and to economic and social progress. The tribes were seen as a standing menace to law and order, as recalcitrant to taxation and conscription, and as an obstacle to the establishment of effective governmental authority. At the root of these problems lay the tribes’ status as semi-autonomous political, social and even military units, ruled by their own chiefs, bound by their own internal loyalties, and governed by their own customary laws.10 From 1831 onwards, the Ottomans aimed to break the power of the tribal confederations in Iraq. During the early years of the Tanzimat period, in the absence of a powerful army, Ottoman tribal policy mainly relied on the principle of ‘divide and rule’, fostering or exploiting rivalries within the ruling families of the confederations, or between shaykhs of constituent tribes, or between the latter and the ruling families, or ignoring the shaykhs altogether and dealing directly with the chiefs of tribal sections. In the pursuit of these policies, the Ottoman officials used land and tax as useful weapons both before and after the application of the 1858 Land Law. The beginning

Page 4 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century of (p.277) the application of the 1858 Land Law in the early 1870s, accompanied by the development of navigation and communication, and the emergence of a market-oriented agricultural economy, brought new developments in terms of the tribal issue. Under the Land Law, the usufructs of state lands were granted sometimes to the cultivators, though more often to the paramount tribal shaykhs, city merchants and notables, for an appropriate fee (the tapu system). In some places, like Diyala, Karbala and the neighbourhood of Baghdad, this led to a rapid erosion of the tribal system. In other places, like the Hilla–Diwaniya region of the middle Euphrates, the tribal system persisted, but had been significantly modified, as the authority of the paramount chiefs of the old tribal confederations declined, while that of the lesser shaykhs and sarkals (rural foremen) of individual tribes grew stronger. In Muntafiq and ‘Amara, however, the tribal system remained strong.11 In Kurdish areas, on the other hand, where the power of the big Kurdish amirates was destroyed after 1831, power passed to the beys, and ağas, and to the shaykhs of the tariqas (a Sufi order), and following the application of the Land Law, the bulk of tapu lands passed into their hands. As a result, agrarian relations remained quite similar to the period before the tapu system, with those who had previously acquired wealth and authority as tax farmers and moneylenders now acquiring it as landowners. Since the Ottoman government was not able to collect taxes directly from the cultivators, tax farming continued to provide an important power base for the notables. The result was a chronic political and economic power struggle among shaykhly families, beys and ağas, as the development of commercial agriculture and regional trade increased the importance of tax farming and possession of lands.12 The principal source of conflicts in the Iraqi provinces, whether between notables or tribes, or between them and the government, was land. The state lands were periodically leased to the tribes, and the struggle for the control of these leases was a major source of conflict between tribes, and also between tribes and the government officials who awarded the leases. A further cause of tribal disturbances in Iraq was the substantial arrears in tax payments, especially in Basra, and the government’s periodic efforts to reclaim them by military force. Both land and tax issues often led to armed clashes, but matters gradually escalated when the tribes of Iraq began to arm themselves with modern weapons, enabling them to defy the government troops. Thus the government became concerned with the illegal arms-trade in the region. After about (p. 278) 1900, Basra, and later Mosul, faced constant and gradually worsening tribal outbreaks, and the local troops proved ineffective. Though the Porte responded by initiating a series of reform measures, these depended upon the establishment of law and order, and this was never achieved.13

The Iranian (Shi‘i) Challenge Iraq bordered on Iran, a Shi‘i state which had historical and religious claims there, and which, in the course of centuries, had fought numerous wars with the Ottoman Empire for the possession of Iraq. Since the establishment of the Safavid state in Iran during the early sixteenth century, Ottoman–Iranian relations had been characterised by continuous struggle. Power struggles took place over eastern Anatolia–Azerbaijan and Iraq–western Iran. The Ottomans’ focus of interest was Azerbaijan and the Caucasus region, while the Iranians were concerned with Iraq, which contains the holiest sites of Shi‘i Islam. This prolonged struggle continued in varying degrees until the end of the First World War.14 Ottoman Iraq faced three major challenges regarding Iran in the nineteenth century. First, following the 1821–3 war between the two countries, and despite Iran’s relative military inferiority, Ottoman statesmen considered Iran a potential military threat, particularly in the event of a Russian invasion of Anatolia. For this reason, policy towards Iran remained a central

Page 5 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century preoccupation for the Ottoman government during wars with Russia, especially in the case of both the Crimean War of 1853–6 and the Russo–Turkish War of 1877–8. In times of peace, Iranian troop movements on the Iraqi border always alarmed the Ottoman authorities in Baghdad and Istanbul.15 Second, the question of the delimitation of the Ottoman–Iranian border remained a constant source of tension in bilateral relations.16 Several seminomadic Kurdish tribes inhabited both sides of the Iraq–Iran frontier and did not recognise any border; the reciprocal cross-border movements of the tribes caused a series of problems. Both states competed to gain the tribes’ loyalty and to establish patronage over each other’s (p.279) Kurds. As the peace treaty of 1823 failed to solve the problem and as the process of Ottoman centralisation began, the boundary question once more came to the fore. As a result of Russian and British intervention, an agreement was finally reached in 1847 stipulating that the entire border be surveyed by a mixed commission, whose work was finally completed in 1865, and the frontier was confirmed by a convention signed in Istanbul in 1869. But this step did not finally resolve the issue: as Iran proved useful to Russia during the Russo–Turkish war of 1877–8, Russia inserted Article 60 into the Berlin treaty ending the conflict, giving the disputed territory of Kotor (near Van) to Iran.17 From September 1905 onwards, Ottoman troops occupied a series of disputed territories on the Iranian border, from Bayazid south to Vazne, and remained there until 1913.18 When the Committee of Union and Progress (CUP) came to power in 1908, it was expected that the Ottoman army would soon withdraw from the occupied territories over the border in Iran. This did not occur, however, as the CUP regime was inclined to persevere with the occupation in order to give material support to Iranian (Azeri) constitutionalists, to help bring about Russian withdrawal from Iranian Azerbaijan, and to gain advantage and strengthen Turkish claims during the expected border-delimitation negotiations with Iran. But as time went on, the Porte proved unable to resist Russian and British pressure for an end to occupation and a swift delimitation of the border. When bilateral negotiations produced little result, Britain and Russia intervened and convinced the Porte to concede a more favourable line to Iran. As a result, in the midst of the Balkan Wars of 1912–13, these quadripartite negotiations resulted in the signing of the Constantinople Protocol (November 1913), in which the boundary line was described in considerable detail. A new delimitation commission, composed of representatives of all four governments, was charged with confirming the exact demarcation. The work of the commission lasted from January to October 1914; the result, however, was never ratified in a formal treaty, because of the outbreak of the First World War.19 The third major challenge was the ‘Shi‘i threat’ in Ottoman Iraq, which emerged in the nineteenth century. The Iraqi provinces were home to a substantial population of Arabicspeaking Shi‘i Muslims. They constituted an absolute majority of the population, especially in Baghdad and Basra. Furthermore, throughout the nineteenth century, there appears to have been a growth in this Shi‘i population at the expense of the Sunni sect, as the former expanded through conversion. To the Ottoman authorities, the presence of a large and growing Shi‘i population in Iraq represented a serious political (p.280) problem. The Ottoman Empire was a Sunni state, with which its Shi‘i subjects could not be trusted to identify. Nor, in principle, did Shi‘i Muslims recognise the Ottoman claim to possession of the great Islamic Caliphate, a claim which Ottoman sultans repeatedly emphasised in an effort to give religious legitimacy to their regime. In short, the Shi‘is were regarded as potentially disloyal and as a natural ally of Iran in any future conflict.20

Page 6 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century The Shi‘i problem in Iraq also had international dimensions: Iran and Britain. There was constant communication between Iraq and Iran. Iraq contained the most sacred Shi‘i shrines, located at Najaf, Karbala, Kazimayn and Samarra, and collectively known as the ‘Atabat. In the eighteenth century, the ‘Atabat became a centre for the Usuli school of Shi‘i jurisprudence, which argued for a political role for the ulema; and the Atabat retained their primacy as a centre of religious authority throughout the nineteenth century. Most of the important Shi‘i mujtahids (jurisconsults) either resided and taught there, or studied there for a time before returning to Iran. Together with these mujtahids, a large number of mullahs, akhunds (Shi‘i mullahs who wandered among the tribal population), and students resided in the ‘Atabat. First in the Tobacco Protest of 1891–2, and later from 1902 onwards, and especially, during the years of the Iranian Constitutional Revolution (1905–11), the mujtahids of the ‘Atabat became actively involved in Iranian politics. In addition, there were many Iranian subjects at the ‘Atabat: religious students, merchants and pilgrims. Every year an important number of people, from Iran and India, visited the shrine cities of Iraq, or brought the remains of their relatives to bury at the ‘Atabat. The Iranian government made annual grants to the shrines at Karbala, Najaf and Kazimayn.21 The British, too, had links with the ‘Atabat. Not only did numerous Shi‘is from British India visit and reside at the ‘Atabat, but the British government had direct links with the mujtahids through the Oudh Bequest. The bequest had been established by the King of Oudh in India, and provided for the annual distribution of charitable money at the ‘Atabat. Following Oudh’s annexation by the British government of India in the 1850s, control over the bequest had passed into British hands, and the annual distribution of funds was conducted by the British consul-general at Baghdad, through two selected mujtahids, one at Najaf and one at Karbala. For these favoured mujtahids, the bequest was a major source of local influence and prestige, and indirectly it was a potential channel for British influence, too. After 1902 the British began to see the mujtahids as a factor in their own policy towards Iran, as the implications of the Great (p.281) Game became more complicated thanks to new developments in Iranian internal politics.22 Since the Shi‘i ulema enjoyed great prosperity and wealth through the Oudh bequest and other donations, they exercised much influence in Iraq, especially among the tribes. It appears that through well-established medreses in the cities, and through akhunds, Shi‘ism expanded in the region. Given the fact that mujtahids distributed an important amount of money to religious students and the poor, it comes as no surprise that some of the tribesmen, especially the newly settled ones, and also some small town-dwellers were attracted to Shi‘ism.23 The growth of Shi‘ism among the tribal population was known to the Porte in the 1860s, though it was not regarded with the same seriousness as it would later be under Abdülhamid II (1876– 1909). During Midhat Pasha’s governor-generalship, for example, the extent of the problem was clearly seen, provoking serious concern on the part of the Ottoman authorities.24 It appears that this concern soon subsided, and for about fifteen years the Ottoman government paid little attention to the issue. From 1885 onwards, however, their attitude changed, and official reports persistently emphasised the growth of Shi‘ism in the region. All these reports alarmed the Hamidian regime in the late 1880s and early 1890s, and prompted the Palace to embark upon a serious consideration of the Shi‘i issue.25 In order to forestall the growth of the Shi‘i sect, different steps were taken: several commissions were sent to the region; local officials were asked to write detailed reports on the subject; and some measures were taken in the field of education. But these steps resulted in little change, as Ottoman officials, in general, seemed to

Page 7 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century take a naive view of the causes and remedies of the ‘problem’, believing that they could easily ‘convert’ people from their ‘superstitious belief through preaching and education.26 In the early 1890s, however, after a period of consideration and consultation, Sultan Abdülhamid came to the conclusion that a policy of Sunni-Shi‘i unity would be the best long-term solution to the ‘Shi‘i problem’ in Iraq, and decided to promote a radical (p.282) programme to secure a union or religious rapprochement between Shi‘i and Sunni Islam, and even the establishment of a form of confederal relationship with Iran. His chosen tool was Jamal al-Din al-Afghani, the Iranian-born political activist and philosopher, who arrived in Istanbul in the late summer of 1892.27 The implementation of the project for the creation of Islamic alliance and unity began in early 1894. A working-group was set up under Afghani which sent hundreds of letters to prominent Shi‘i ulema all over the Islamic world. However, when Shah Nasir al-Din of Iran learned about the correspondence between Afghani’s Istanbul circle and the Shi‘i ulema, he quickly responded. Tehran demanded the deportation of Afghani and his disciples, but the Iranian authorities also used this issue to bring the Armenian question’ to the fore, putting pressure on the Ottoman sultan by giving a free hand to Armenian revolutionaries, both inside Iran and on the border. Under the pressure of the Armenian crises both in Anatolia and in Istanbul, and because of this Iranian support for Armenian revolutionaries in eastern Anatolia, Abdülhamid appears to have been forced to give up his scheme for Sunni-Shi‘i unity. The project was abandoned, Afghani’s relations with Abdülhamid deteriorated, and Afghani remained in Istanbul as a virtual prisoner until his death in March 1897. The extent of Ottoman concern at the Shi‘i problem was primarily determined by the general state of relations between Iran and the Ottoman Empire: the more relations deteriorated, the more the Porte became concerned with the Shi‘i population of Iraq. From the early 1900s onwards, the Ottoman authorities grew more concerned as internal developments in Iran began to affect the Atabat. In the early 1900s, this concern led the sultan to continue his efforts to solve the Shi‘i problem in Iraq, as well as to gain support among the Shi‘i ulema in the ‘Atabat. In the end, no ‘solution’ was ever found.28

The British Penetration of Iraq and the Gulf Throughout the nineteenth century, Ottoman statesmen were preoccupied by their empire’s vulnerability to the European Great Powers.29 It was not simply that they feared military attack, and knew that their chances of resisting it were slim; they also feared that the Powers might undermine the empire’s independence and integrity from (p.283) within, through techniques of ‘peaceful penetration’. The latter fear was grounded in historical experience. Since the 1830s, European Powers had succeeded in penetrating the Ottoman Empire to a considerable degree, interfering in its internal affairs, and recruiting networks of clients among the empire’s own subjects. A number of factors had facilitated this penetration. The Powers had acquired certain legal rights of interference in Ottoman internal affairs, through the reform provisions of the treaties of Paris (1856) and Berlin (1878), through the capitulations, which gave their subjects legal and fiscal privileges within the Ottoman Empire, and through the religious protectorates which particular Powers asserted over particular groups of Ottoman Christians. In addition, the considerable expansion of the Ottoman Empire’s trade with the European Powers, and the various economic concessions, including ports, railways, mines and river navigation, which had been awarded by the Ottoman government to European enterprises, had enabled the Powers to build up local commercial clienteles, particularly in the major ports and trading centres. This commercial influence was accompanied by cultural influence, promoted by missionaries and educational institutions. Finally, the omnipresence of European political influence was assured

Page 8 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century through chains of consuls, established in almost every important provincial centre throughout the empire. To the Ottoman statesmen, the danger seemed obvious: if left unchecked, European influence would eventually undermine the political authority of the Ottoman government, leading to the establishment of ‘zones of influence’ and to ultimate partition. The examples of Egypt and India were not encouraging.30 In Iraq, the obvious source of danger was Britain. Since the 1830s, the British had acquired a virtual monopoly of European influence in Iraq and the adjoining regions of Arabia and the Gulf. British warships regularly patrolled the Gulf, where many of the local shaykhdoms had concluded ‘trucial’ protective agreements with representatives of the government of British India. A large proportion of the trade of Iraq and the Gulf was done with British India, and British and British Indian vessels dominated Gulf merchant shipping. A British enterprise, the Lynch company, held a concession for steam navigation on the Euphrates and the Tigris. Since 1862 a British mail service had run between Iraq and India, and British-constructed telegraph lines linked Baghdad to India, Istanbul and Tehran.31 The British Indian pilgrims and students who flocked to the Shi‘i shrines of southern Iraq were a further channel for British influence, (p. 284) and as already explained, the Oudh Bequest brought the British directly into contact with Shi‘i religious leaders in the region. During the Tanzimat era, when Anglo-Ottoman relations had generally been close, the growth of paramount British influence in the Gulf region had caused the Porte serious concern, leading it to strengthen the Ottoman naval presence in the region in the 1860s, and to bring Najd and the Arabian coast of the Gulf down to Qatar under its direct control in the 1870s.32 Soon after Sultan Abdülhamid’s accession, however, and in particular after the disastrous Ottoman defeat in the 1877–8 war with Russia, Anglo-Ottoman relations took a lasting turn for the worse. Abdülhamid was convinced that Britain’s failure to protect the Ottoman Empire during the war with Russia, and still more, her willingness to join the other Powers in truncating the empire’s territories at the subsequent Congress of Berlin, were clear proof that the British had abandoned their former policy of preserving the empire, and that their ultimate aim was to partition it.33 In particular, Abdülhamid suspected them of harbouring designs upon his Arab provinces, and of actively promoting the notions of Arab political independence and an Arab Caliphate’.34 Yet in practice, Sultan Abdülhamid had few means of defence against the British in Iraq, and in the adjoining regions of Arabia and the Gulf. In other regions, like, for example, Syria, he could play upon the Powers’ mutual rivalries, exploiting them to hold all in check; but in Iraq and the Gulf, the British had no rivals, at least until the early 1900s, when the Baghdad railway project brought the Germans into the picture. The empire’s local military resources were minimal: the naval flotilla at Basra was no match for the British, and the 6th Army of Iraq was chronically under-strength and short of funds. Ottoman administrative control, too, was weak, and permanently beset by tribal rebellions and disturbances, and by a potentially disloyal Shi‘i population. Proper vigilance might at least ensure that the British were deprived of opportunities for extending their influence further into Iraq, but the sultan lacked the means to embark upon an aggressive or ‘forward’ policy, except on a small-scale, opportunistic basis. His safest course was to avoid direct confrontations.35 (p.285) The British, too, generally chose to eschew aggressive policies in the region. Sultan Abdülhamid was broadly correct in his view that the British had lost faith in his empire, and that they expected it to collapse at some unspecified date in the future. He was also correct in his view that they regarded regions like Arabia, Iraq and the Gulf as lying within their own

Page 9 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century particular ‘sphere of interest’, and as due to fall to themselves, in some sense, when the Ottoman Empire’s inevitable demise occurred. Yet apart from exceptional circumstances, like the Armenian crisis of 1894–6, successive British governments had no wish actively to hasten the empire’s demise. They were broadly willing to tolerate the status quo, particularly where, as in Iraq and the Gulf, it seemed obviously favourable to themselves.36 Yet if neither side was ready to embark upon an aggressive policy, neither was fully in control of the situation, particularly on the shores of the Gulf and the adjoining regions of Arabia. In these regions, the actions of quasi-independent local actors could potentially bring Britain and the Ottoman Empire into conflict. Since the 1830s the British had gradually expanded their influence in these regions, annexing no territory, but concluding trucial arrangements with the local rulers of Qatar, Bahrain, Muscat and the Trucial Coast, some of whom were regarded by the Porte as Ottoman subjects.37 In 1871, however, Midhat Pasha, as vali of Baghdad, had set out to reassert Ottoman authority to the southward, persuading the ruler of Kuwait to acknowledge Ottoman suzerainty, and sending an expedition to Najd which re-established Ottoman authority throughout the region.38 Before long, Ottoman officials had also persuaded the ruler of Qatar to acknowledge Ottoman suzerainty, and to receive Ottoman garrisons. The British raised no protest, though without clearly acknowledging Ottoman claims to sovereignty. The potential danger lay not only in the uncertainty of claimed Ottoman and British rights, but in the shakiness of Ottoman physical control. Even in Kuwait and Qatar, this control was more nominal than real, while in Najd, the chronic struggle for overlordship between the rival families of Ibn Rashid of the Jabal Shammar confederation, and of Saud, was a further potential source of local instability (p.286) and complications into which both Ottoman and British governments might easily be drawn.39 Notwithstanding the growth of these mutual suspicions, the early years of Abdülhamid’s regime saw no open conflict between Britain and the Ottoman Empire in Iraq and the Gulf. The British were content with the status quo, and the sultan lacked the strength to challenge it, except in minor ways. Within Iraq, the Ottoman authorities did take some measures to limit the spread of British influence. These included the abolition of the British postal service between Baghdad and Damascus and its replacement by an Ottoman postal service (1881–6); the registration of British citizens and British-protected persons in Baghdad (1881–4); efforts to limit contacts between British consuls and local tribes (1882–5); opposition to British missionaries’ efforts to establish a missionary school at Baghdad (1883–9); the encouragement of an Ottoman rival to the Lynch steamship company on the Tigris (1885–8); and attempts to keep foreign warships out of the Shatt al-‘Arab.40 In the Gulf, however, the Ottoman government was less assertive. In 1879, in response to British complaints of a revival of piracy, the Porte ordered a small, though apparently effective, reinforcement of its Basra flotilla.41 In 1880, and again in 1882, the Porte considered a more substantial strengthening of its naval presence in the Gulf, with the purchase of new ships, and the construction of new shore facilities at Basra; but on both occasions, it appears, lack of funds prevented the implementation of these plans.42 Not until the second half of the 1880s did the Ottoman government make a serious attempt to strengthen its military position in the Gulf, by constructing a large fort at Faw. This led to a confrontation with Britain.43 The protracted dispute over Faw contributed to a growing Ottoman sense of unease on the subject of the Gulf, Najd and Basra, an unease reflected in a number of reports submitted by local Ottoman officials. These reports identified Britain as the primary threat, but they also drew attention to other

Page 10 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century dangers, including the ambitions of local Arabian rulers like Ibn Rashid of Najd, and the flourishing arms-trade in the Gulf.44 (p.287) Even though mutual suspicions gained a certain momentum with the construction of the Faw fortifications, there was no open conflict between Britain and the Ottoman Empire in Iraq and the Gulf in the later years of Abdülhamid’s regime. Although the Porte took steps, whenever the opportunity arose, to strengthen the empire’s claims to sovereignty over the Arabian shores of the Gulf, in particular over Kuwait, Qatar and Bahrain, it strove to avoid any direct challenge to the British. The late 1890s and early 1900s saw the emergence of potential threats to the regional status quo in Kuwait and Najd, and also in the Baghdad railway project.45 Still, both sides in the last resort wanted to preserve the status quo and avoid conflict, even though apprehension on the part of the local Ottoman authorities was growing steadily. In general, relations between the Ottoman Empire and Britain in the region were influenced by the overall state of their relations, and by other mutual problems in neighbouring regions, such as the Yemen or the Red Sea.46 The Ottoman administration of the Iraqi provinces straddled several complex frontiers and presented significant challenges to the ability of the government to assert its authority and carry out the reform process. The Ottoman central government expended considerable effort to overcome these challenges, but proved unable to resolve them completely, and as a result, both its authority and reform efforts were undermined. These obstacles to Ottoman administration of the Iraqi provinces, due to Iraq’s location in a frontier region, compounded by the government’s inability to satisfy the conflicting desires and interests of those involved, presented a dilemma to the empire which it was unable to transcend. Notes: Proceedings of the British Academy 156, 271–287. © The British Academy 2009. (1) For Iraqi provinces in the ‘classical period’, see Stephen H. Longrigg, Four Centuries of Modern Iraq (Oxford: Clarendon Press, 1925); Dina R. Khoury, State and Provincial Society in the Ottoman Empire: Mosul, 1540–1834 (Cambridge: Cambridge University Press, 1997); Ahmet Gündüz, Osmanlı İdaresinde Musul, 1523–1639 (Elazıg: Fırat Üniversitesi, 2003). (2) See Halil Sahillioğlu, ‘Osmanlı Döneminde Irak’ın İdari Taksimatı’, Belleten, 211 (1990), 1233–57; Halil İnalcık, The Ottoman Empire: The Classical Age, 1300–1600 (London: Weidenfeld & Nicolson, 1973), 104 ff. (3) For the Mamluk period, see Albert Hourani, ‘The Fertile Crescent in the eighteenth century’, Studia Islamica, 8 (1957), 91–118; Tom Nieuwenhuis, Politics and Society in Early Modern Iraq: Mamluk Pashas, Tribal Shaykhs and Local Rule between 1802 and 1831 (The Hague: Martinius Nijhoff, 1982); Thabit A. J. Abdullah, Merchants, Mamluks, and Murder: The Political Economy of Trade in Eighteenth-Century Basra (Albany, NY: State University of New York, 2001); Thomas Lier, Haushalte und Haushaltspolitik in Baghdad, 1704–1831 (Würzburg: Ergon Verlag, 2004). (4) See Longrigg, Four Centuries, 267–76; J. G. Lorimer, Gazetteer of the Persian Gulf, Oman, and Central Arabia (Calcutta: Superintendent Government Printing, 1908–15), I/1B: 1313 ff. (5) See Lorimer, Gazetteer, I/1B: 1314–19, 1349 ff.; Longrigg, Four Centuries, 277 ff.; J. B. Kelly, Britain and the Persian Gulf, 1795–1880 (Oxford: Oxford University Press, 1968), 270–6, 838–9.

Page 11 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century (6) See Ebubekir Ceylan, ‘Ottoman Centralization and Modernization in the Province of Baghdad, 1831–1872’ (Unpublished Ph.D. thesis, Bogaziçi University, 2006); Keiko Kiyotaki, ‘Ottoman Land Policies in the Province of Baghdad, 1831–1881’ (Unpublished Ph.D. thesis, University of Wisconsin, 1997). (7) For a general picture of social and economic change in Iraq in the Tanzimat period, see Roger Owen, The Middle East in the World Economy, 1800–1914 (London: Methuen, 1987), 82, 180 ff., 273 ff.; Charles Issawi, Economic History of the Middle East and North Africa (New York: Columbia University Press, 1982), 127 ff.; idem, The Fertile Crescent, 1800–1914: A Documentary Economic History (New York: Oxford University Press, 1988); Hanna Batatu, The Old Social Classes and Revolutionary Movements of Iraq: A Study of Iraq’s Old Landed and Commercial Classes and of its Communists, Bathists, and Free Officers (Princeton: Princeton University Press, 1978), 13–361; Ghassan R. Atiyyah, Iraq, 1908–1921: A Political Study (Beirut: Arab Institute for Research and Publishing, 1973), 19–50; Stephen H. Longrigg, Four Centuries, 298–324; idem, Iraq, 1900 to 1950: A Political, Social, and Economic History (London: Oxford University Press, 1953), 1–40; Hala Fattah, The Politics of Regional Trade in Iraq, Arabia, and the Gulf 1745–1900 (Albany, NY: State University of New York, 1997). (8) For details, see Gökhan Çetinsaya, Ottoman Administration of Iraq, 1890–1908 (London: Routledge, 2006), 24–48; idem, ‘The politics of reform in Iraq under Abdülhamid II, 1878–1908’, Turkish Journal of Islamic Studies, 3 (1999), 41–72. (9) For details, see Çetinsaya, Ottoman Administration of Iraq, 49–71; idem, ‘Sultan Abdulhamid II’s officials: the case of Nusret Paşa at Baghdad, 1888–1896’, The Journal of Ottoman Studies, 21 (2001), 257–67; idem, ‘The Caliph and the shaykhs: Abdülhamid II’s policy towards the Qadiriyya of Mosul’, in Itzchak Weismann and Fruma Zachs (eds.), Ottoman Reform and Muslim Regeneration: Studies in Honor of Butrus Abu-Manneh (London: I. B. Tauris, 2005), 97–107. (10) For the characteristics of Iraq as a tribal society, see Albertine Jwaideh, ‘Tribalism and modern society: Iraq, a case study’, in R. M. Savory (ed.), Introduction to Islamic Civilisation (Cambridge: Cambridge University Press, 1976), 160–7; Batatu, Old Social Classes, 13 ff., 37 ff., 63 ff., 152–69. (11) See Kiyotaki, Ottoman Land Policies’; Albertine Jwadieh, ‘Midhat Pasha and the land system of lower Iraq’, St. Anthony’s Papers: Middle Eastern Affairs, No. 3 (London, 1963), 105–36; idem, ‘Aspects of land tenure and social change in lower Iraq during late Ottoman times’, in Tarif Khalidi (ed.), Land Tenure and Social Transformation in the Middle East (Beirut: American University of Beirut, 1984), 333–56; Marion Farouk Sluglett and Peter Sluglett, ‘The transformation of land tenure and rural social structure in central and southern Iraq, c.1870– 1958’, International Journal of Middle East Studies, 15 (1983), 491–505. (12) See Martin van Bruinessen, Agha, Shaikh and State: On the Social and Political Organization of Kurdistan (Utrecht, 1978), 275 ff.; David McDowall, A Modern History of the Kurds (London: I. B. Tauris, 1996), 38 ff.; Sinan Marufoğlu, Osmanlı Döneminde Kuzey Irak, 1831–1914 (Istanbul: Eren, 1998); Sarah D. Shields, Mosul Before Iraq: Like Bees Making Five-Sided Cells (Albany, NY: State University of New York, 2000). (13) For details, see Çetinsaya, Ottoman Administration of Iraq, 72–98.

Page 12 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century (14) For the wars with Iran and the Iranian occupations in Iraq, see Longrigg, Four Centuries; R. K. Ramazani, The Foreign Policy of Iran, 1500–1941 (Charlottesville: University Press of Virginia, 1966); Stanford J. Shaw, ‘Iranian relations with the Ottoman Empire in the eighteenth and nineteenth centuries’, in Peter Avery et al. (eds.), The Cambridge History of Iran, VII: From Nadir Shah to the Islamic Republic (Cambridge: Cambridge University Press, 1991), 297–313. (15) For Ottoman–Iranian relations in the nineteenth century, see Mohammed Reza Nasiri, Nasireddin Şah Zamanında Osmanlı-İran Münasebetleri, 1848–1896 (Tokyo: University of Foreign Studies, 1991); Gökhan Çetinsaya, ‘Tanzimat’tan Birinci Dünya Savaşı’na Osmanlı-Iran Iliskileri’, KÖK Araştırmalar, Osmanlı Özel Sayısı (2000), 11–23; idem, ‘Essential friends and natural enemies: the historic roots of Turkish-Iranian relations’, Middle East Review of International Affairs, 7 (2003), 116–32. (16) See Richard Schofield (ed.), The Iran-Iraq Border, 1840–1958 (Archive Editions, 1989), vols. 1–5. (17) Ibid., vols. 1–3. (18) Ibid., vol. 4; Sinan Kuneralp, ‘The Ottoman Drang Nach Osten: the Turco-Persian border problem in Azerbaican, 1905–1912’, in idem (ed.), Studies on Ottoman Diplomatic History IV (Istanbul: Isis, 1990), 71–6. (19) Schofield (ed.), The Iran-Iraq Border, vol. 5; Abdolghafour Baghdadi, ‘Muhammed Ali Şah Devrinde OsmanlıIran Siyasi Ilişkileri ve Iran’daki Iç Olaylar, 1907–1909’ (Unpublished Ph.D. thesis, University of Istanbul, 1982); Burcu Kurt, ‘Şattü’l-Arab Sorunu’ (Unpublished MA thesis, Marmara University, 2006). (20) See Çetinsaya, Ottoman Administration of Iraq, 99–126; idem, ‘The Caliph and mujtahids: Ottoman policy towards the Shi‘i community of Iraq in the late nineteenth century’, Middle Eastern Studies, 41 (July 2005), 561–74; Selim Deringil, ‘The struggle against Shiism in Hamidian Iraq: a study in Ottoman counter-propaganda’, Die Welt des Islams, 30 (1990), 45–62. (21) See Hamid Algar, ‘Atabat’, EIr, 2: 902–3; Meir Litvak, Shi‘i Scholars of Nineteenth-Century Iraq: The ‘Ulama’ of Najaf and Karbala (Cambridge: Cambridge University Press, 1998); Yitzhak Nakash, The Shi‘is of Iraq (Princeton: Princeton University Press, 1994), 163–201, 238 ff.; Lorimer, Gazetteer, I/2, 2357–63. (22) See Lorimer, Gazetteer, I/1B, 1409–15, 1477–84, 1598–1616; Juan R. I. Cole, ‘“Indian money” and the Shi‘i shrine cities of Iraq, 1786–1850’, Middle Eastern Studies, 22 (1986), 461–80; Meir Litvak, ‘Money, religion, and politics: the Oudh Bequest in Najaf and Karbala, 1850–1903’, International Journal of Middle East Studies, 33 (2001), 1–21; idem, ‘A failed manipulation: the British, the Oudh Bequest and the Shii ulama of Najaf and Karbala’, British Journal of Middle Eastern Studies, 27 (2000), 69–90; Nakash, Shi is of Iraq, 205 ff. (23) See Yitzhak Nakash, ‘The conversion of Iraq’s tribes to Shi‘ism’, International Journal of Middle East Studies, 26 (1994), 443–63; idem, Shi is of Iraq, 25–48; Gertrude L. Bell, Review of the Civil Administration of Iraq (London: HMSO, 1920), 27. (24) See Ahmet Nuri Sinaplı, Seyhül Vüzera, Serasker Mehmet Namık Paşa (Istanbul: Yenilik Basımevi, 1987), 196, 259–60; Ali Haydar Midhat, The Life of Midhat Pasha (London: Murray, 1903), 61–2; Lorimer, Gazetteer, I/1B, 1417–22; Nasiri, Nasireddin Şah, 151–6; Edhem Eldem, Page 13 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century ‘Quelques lettres d’Osman Hamdi Bey à son père lors deson séjour en Irak (1869–1870)’, Anatolia Moderna, 1 (1990), 126–7. (25) See Çetinsaya, Ottoman Administration of Iraq, 99–126; idem, ‘The Caliph and mujtahids’, 561–74. (26) See Gökhan Çetinsaya, ‘The Ottoman view of the Shiite community of Iraq in the late nineteenth century’, in A. Mosutti et al. (eds.), The Other Shiites: From the Mediterranean to Central Asia (Bern: Peter Lang, 2007). (27) For Afghani, see Nikki R. Keddie, ‘Afgani, Jamal-al-Din’, EIr, 1: 481–6; idem, Sayyid Jamal adDin ‘al-Afghani’: A Political Biography (Berkeley: University of California Press, 1972). (28) Ibid. (29) See M. S. Anderson, The Eastern Question, 1774–1923 (London: Macmillan, 1966); Roderic H. Davison, Reform in the Ottoman Empire, 1856–1876 (New York: Gordian Press, 1973); Charles Issawi, The Economic History of the Middle East, 1800–1914: A Book of Readings (Chicago: University of Chicago Press, 1966); idem, The Economic History of Turkey (Chicago: University of Chicago Press, 1980); Donald Quataert, Social Disintegration and Popular Resistance in the Ottoman Empire, 1881–1908: Reactions to European Economic Penetration (New York: New York University Press, 1983). (30) See F. A. K Yasamee, ‘Abdülhamid II and the Ottoman defence problem’, Diplomacy and Statecraft, 4 (1993), 20–36; idem, Ottoman Diplomacy: Abdülhamid II and the Great Powers, 1878–1888 (Istanbul: ISIS, 1996); idem, Ottoman diplomacy in the era of Abdülhamid II, 1878– 1908’, Çagdas Türk Diplomasisi: 200 Yıllık Süreç (Ankara: TTK, 1999), 223–32; idem, ‘The Ottoman Empire and European alliances, 1815–1914’, in The Great Ottoman-Turkish Civilization: Politics (Ankara: Yeni Türkiye Yayınları, 2000), 409–17. (31) See Foreign Office, Mesopotamia (London: HMSO, 1920); Kelly, Britain and the Persian Gulf; Malcom E. Yapp, ‘The nineteenth and twentieth centuries’ and ‘British policy in the Persian Gulf, in Alvin J. Cottrell (ed.), The Persian Gulf: A General Survey (Baltimore: Johns Hopkins University Press, 1981), 41–69, 70–100. (32) See Sinaplı, Mehmet Namık Paşa, 210 ff.; Eldem, ‘Quelques lettres d’Osman Hamdi Bey’, 127; Midhat, Midhat Pasha, 55–61; Yusuf Halaçoğlu, ‘Midhat Paşa’nın Necid ve Havalisi ile İlgili Bir Kaç Lâyihası’, Tarih Enstitüsü Dergisi, 3 (1972), 149–76. (33) See Yasamee, Ottoman Diplomacy; idem, ‘Ottoman defence problem’, 20–36. (34) See Selim Deringil, ‘The Ottoman response to the Egyptian crisis of 1881–1882’, Middle Eastern Studies, 24 (1988), 3–24; F. A. K. Yasamee, ‘The Ottoman Empire, the Sudan and the Red Sea coast, 1883–1889’, in Selim Deringil and Sinan Kuneralp (eds.), Studies on Ottoman Diplomatic History, V (Istanbul: ISIS, 1990), 87–102; Ş. Tufan Buzpınar, ‘The Hijaz, Abdulhamid II and Amir Hussein’s secret dealings with the British’, Middle Eastern Studies 31 (1995), 99– 123; idem, ‘Opposition to the Ottoman Caliphate in the early years of Abdülhamid II, 1877– 1882’, Die Welt des Islams, 36 (1996), 59–89; idem, ‘The repercussions of the British occupation of Egypt on Syria, 1882–1883’, Middle Eastern Studies, 36 (2000), 82–91; idem, ‘Vying for power

Page 14 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century and influence in the Hijaz: Ottoman rule, the last emirate of Abdulmuttalib and the British (1880–1882)’, The Muslim World, 95 (2005), 1–22. (35) Yasamee, Ottoman Diplomacy; idem, ‘Ottoman defence problem’, 20–36. (36) See Anderson, Eastern Question; Keith M. Wilson, ‘Constantinople or Cairo: Lord Salisbury and the partition of the Ottoman Empire, 1886–1897’, in Keith M. Wilson (ed.), Empire and Continent: Studies in British Foreign Policy from the 1880s to the First World War (London: Mansell, 1987), 1–30; Stuart A. Cohen, British Policy in Mesopotamia, 1903–1914 (London: Ithaca, 1976). (37) See Yapp, ‘British policy in the Persian Gulf; Kelly, Britain and the Persian Gulf; idem, ‘The legal and his torical basis of the British position in the Persian Gulf, St. Anthony’s Papers, No. 4: Middle Eastern Affairs (London, 1959), 119–40. (38) See Kelly, Britain and the Persian Gulf 717 ff.; Ravinder Kumar, India and the Persian Gulf Region, 1858–1907: A Study in British Imperial Policy (London: Asia Publishing House, 1965), 113 ff.; Frederick F. Anscombe, The Ottoman Gulf: The Creation of Kuwait, Saudi Arabia, and Qatar (New York: Columbia University Press, 1997), 16–53; Zekeriya Kurşun, Necid ve Ahsa’da Osmanlı Hakimiyeti: Vehhabî Hareketi ve Suud Devleti’nin Ortaya Çıkış ı (Ankara: TTK, 1998), 79–122. (39) See Anscombe, The Ottoman Gulf; Kurşun, Necid ve Ahsa; idem, The Ottomans in Qatar: A History of Anglo- Ottoman Conflicts in the Persian Gulf (Istanbul: Isis, 2002). (40) See Çetinsaya, Ottoman Administration of Iraq, 127–46; Lorimer, Gazetteer, I/2, Appendix K: Mail Communications and the Indian Post Office in the Persian Gulf, 2439 ff.; I/1B, 1534 ff., 1548–66; 1877–83. (41) Ibid.; cf. Kumar, India, 119–20. (42) Ibid.; for other political and strategic considerations of the time, see Hilmi Kamil Bayur, Sadrazam Kamil Paşa: Siyasi Hayatı (Ankara: Sanat Basımevi, 1954), 86–8; Ş. Tufan Buzpınar, ’Abdülhamid II and Sayyid Fadl Pasha of Hadramawt’, Osmanlı Arastırmaları, 13 (1993), 227–39; İdris Bostan, ‘Basra Körfezinin Güney Kesimi veOsmanlılar, 1876–1908’, Osmanlı Arastırmaları, 9 (1989), 311–22. (43) Çetinsaya, Ottoman Administration of Iraq, 130–2; Schofield (ed.), The Iran-Iraq Border, 3: 533 ff, 741 ff. (44) For the reports, see Çetinsaya, ‘The Ottoman view of British presence in Iraq and the Gulf: the era of Abdulhamid II’, Middle Eastern Studies, 39 (2003), 194–203; for the Palace’s concern over arms-trade in Syria, Arabia and the Gulf, see Tahsin Paşa, Abdülhamid: Yıldız Hatıraları (Istanbul: M. A. Halit Kitabhanesi, 1931), 345–7; for illegal arms-trade in Iraq and the Gulf, see Lorimer, Gazetteer, I/2, Appendix N: The Arms and Ammunition Traffic in the Gulfs of Persia and Oman, 2556 ff. (45) For details see Çetinsaya, Ottoman Administration of Iraq, 136–46; Anscombe, The Ottoman Gulf; B. C. Busch, Britain and the Persian Gulf 1894–1914 (Berkeley: University of California Press, 1967); J. B. Kelly, ‘Salisbury, Curzon and the Kuwait Agreement of 1899’, in K. Bourne and D. C. Watt (eds.), Studies in International History: Essays Presented to W. Norton Medlicott Page 15 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Challenges of a Frontier Region: The Case of Ottoman Iraq in the Nineteenth Century (London: Longmans, 1967), 249–90; Richard Schofield, Kuwait and Iraq: Historical Claims and Territorial Disputes (London: RIIA, 1991); Alan Rush (ed.), The Records of Kuwait: 1899–1961 (Archive Editions, 1989); Robin Bidwell (ed.), The Affairs of Kuwait: 1896–1905 (London: Frank Cass, 1971); idem (ed.), The Affairs of Arabia, 1905–1906 (London: Frank Cass, 1971). (46) On relations between Britain and the Ottomans in Yemen, see Chapter 15 by Isa Blumi in this volume.

Access brought to you by:

Page 16 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914 ISA BLUMI

DOI:10.5871/bacad/9780197264423.003.0015

Abstract and Keywords The negotiations with the Ottomans over how exactly to define the boundary separating each party's domain were largely confused by a completely different set of criteria. The Ottomans constantly argued that the areas they claimed (large areas of which the British contended existed within Dali territory) had historically and thus always formed part of the Ottoman territory. They installed troops in the areas in dispute and actually started to collect taxes, in part thanks to Muqbil's aggressive alliance-building. The longer this physical presence was maintained, the more difficult it was for the British to argue that these areas were actually Dali. It was largely the growing insurgency in Ottoman Yemen, in some respects a product of British machinations, that ultimately led to the 1903 capitulation by the Ottoman authorities to British demands for formal control of the Dali plateau. Keywords:   Yemen, imperial statecraft, Ottoman, Britain, Muqbil, Dali plateau

BETWEEN 1914 AND 1990 SOUTH-WEST ARABIA WAS DIVIDED IN TWO BY A BORDER, the first 138 miles of which were drawn by the Ottoman and British Empires in negotiations that lasted from 1902 to 9 March 1914.1 With the construction of the Suez Canal in 1869, the southern Arabian peninsula, located on the route to India, had become of great strategic value to the industrial world, above all the British. In 1872 the Ottomans regained control of the Yemeni highlands—Yemen had been incorporated into the Ottoman Empire in the sixteenth century, but Ottoman troops were expelled by 1630—seeking both to forestall European expansion and ensure their own economic advantage. Curiously, despite the region’s importance to the British and the Ottomans, both sets of rulers administered central Yemen from 1872 until 1914, often in direct competition, without finalising a border. Indeed, both powers lacked even decent maps of the area until the end of the nineteenth century. This lack of a boundary went against the general trend in the rest of the world at the time. Along with a standing army, schools and a modern bureaucracy, a political border was a critical tool in

Page 1 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914 inter-state diplomacy, commerce and social development.2 Events in southern Arabia thus shed new light on modern imperialism, for local initiatives, driven by political and economic interests often beyond the reach of imperial powers’ influence, account for the history of the Yemeni boundaries far more than the (p.290) machinations of the Ottomans and the British. Testimony to the so-called ‘subaltern’ agency that lies at the heart of this study is the fact that both powers eventually adopted a new policy in Arabia that resorted to drawing boundaries (as already done in the Balkans in 1878 and Africa in 1886) along the lines dictated by local factors.3

A Context for Modern Imperial Statecraft From the very outset, indigenous politics played a central role in shaping imperial interests in South Arabia. In the highland regions of Yemen, economic partnerships with foreign merchants, especially British and Egyptian, reinforced Yemen’s connection to the global economy. Coffee, honey and the indigenous narcotic, qat, harvested exclusively in the mountains, became important commodities in the early nineteenth century. For some, like the Fadl family of ‘Abdali just outside the southern port of Aden, such partnerships helped them to establish themselves as local dynasties that evolved into pseudo-states by the late nineteenth century. As a result, most of the region maintained political and economic autonomy, which is the key to understanding its relations with outside powers in the last quarter of the nineteenth century. In 1839, when the British seized Aden, the ‘Abdali region became the centre of British commercial and eventually administrative expansion into South Arabia.4 The British were concerned mainly with securing the port and provisioning its population of mostly imported South Asian labourers and local commercial agents. In return for assisting with this, families in ‘Abdali could expect a privileged and exclusive relationship with the Government of India, which administered Aden. As long as the water and camel caravans came into Aden, and silver, gold and products from the outside world went out, there was little reason for direct British investment in the region beyond the city itself. These partnerships with locals in the hinterland (based on small monthly cash payments which were periodically renegotiated) for more than a century gave power to ‘Abdali families, from whom came the Sultan of Lahj, always a member of the Fadl b. Muhsin extended family. Through this sultanate (acknowledged by the British through formal ceremonies and treaties), the Aden administrators maintained relationships with tribal communities such as the Fadli, ‘Amiri, Yafi‘i, Hawshabi and ‘Alawi who lived in outlying areas (Figure 15.1). Since Aden’s first administration negotiated a treaty with the ‘Abdali sultan in the 1840s, Britain had assumed that their political and commercial connections in the highlands (p.291) assured their place in the region’s commercial activity.5 However, the situation changed with the arrival of the Ottomans in the 1860s, followed by France and Italy joining the competition for influence in Yemen, after establishing colonies on the opposite shore of the Red Sea (France took Djibouti in 1885, and Italy Eritrea in the same year). As a result, the entire region between the highland city of Ta‘izz and Aden would become a battleground of imperial agents seeking to secure the loyalties of local community leaders.

Page 2 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914 The Ottomans started to re-establish their presence in Yemen in the wake of British interest in the port of Mukha, reconquering the Red Sea coastal region in 1849–72.6 From the 1860s, the Ottomans started to cultivate alliances among the coffeeproducing communities of the highlands.7 For much of Ottoman rule in Arabia, the region was (p.292) plagued by the fragmentation of power,8 as communities struggled for control of shrinking resources, Figure 15.1. Yemen in the late nineteenth which drew the Ottomans further into century, showing the ‘tribes’ under British Yemen. The authority of the Zaydi imams, protection. the Shi‘i rulers of San‘a’ and much of the highlands, was challenged by their own relatives, resulting in turmoil among the tribesmen. Ottoman intervention in highland Yemen in 1872 came at the invitation of the Imam Ghalib b. Muhammad, who needed outside support to maintain his tenuous position.9 Although the Ottomans did encounter constant resistance in the highlands, there were also enthusiastic supporters of Ottoman rule, many of whom openly clashed with other communities and even their own leaders, who had often earlier signed treaties with British-supported amirs.10 In this way, unrest spread to Zaydi areas as well as those along the hinterland of the Indian Ocean.11 The Ottomans responded to the political breakdown in the highlands by attempting to fill the power vacuum themselves. However, Ottoman strategic concerns were not solely opportunistic. The opening of the Suez Canal in 1869 transformed the Sublime Porte’s reading of the situation in the eastern Mediterranean; the increased significance of Britain’s modest position in Aden and its subsequent occupation of Egypt in 1882 in particular lent new importance to local requests for assistance in usurping British-supported families in the Arabian Peninsula. Not only would new links to Yemen help thwart British expansion, they could help secure agricultural lands that promised large increases in tax revenue for the central coffers as well as adding upwards of four million people to the empire.12 Thus one ambitious Ottoman commander, Ahmed Raşid, went beyond simply establishing a foothold in commercial centres, planning to rule all of the highlands at the expense of the still limited British presence in Aden.13 Likewise, Ahmed Muhtar Pasha, commander of the army that conquered the highlands and established an administration, already in January 1873 envisioned full integration into the Ottoman Empire of the new Yemeni province that would be no different from the others.14 Within two years, a wellrun legal and fiscal bureaucracy was in operation, (p.293) with courts adjudicating Ottoman imperial laws and a regular tax being collected by local allies.15 The British response to these dramatic changes was contradictory at best. The Foreign Office was convinced that the diplomatic partnership with Istanbul developed since the Crimean War in the 1850s had secured a mutual understanding that Ottoman actions would not affect areas under British influence. Events on the ground, however, proved more complicated. The correspondence circulating between Aden, Bombay and London demonstrates that many in southern Yemen, including the Resident and the local allies of the British administration in Aden, were not so sanguine about the long-term consequences of Ottoman ascendancy in areas just north of Aden.16 Within weeks of the first Ottoman incursions into areas south of Ta‘izz in 1872, authorities in Aden (and the British media in London) had voiced fears of future Ottoman Page 3 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914 political and economic domination of the region. As the reports stated, Ottoman ‘interference’ in Yemen threatened to undermine the ability of local ‘sovereigns’ (i.e. the Sultan of Lahj) allied with British Aden to control their subordinates, thereby threatening lucrative commercial relations. Ottoman activity is not the only factor, however; British tactics towards locals also proved crucial. Aden had established in 1869 a system in which the British Resident managed directly relations with locals. This appears to have been a momentary—but significant—attempt by the British to marginalise the Sultan of Lahj.17 In a panic, the Sultan, Fadl b. Muhsin, repeatedly emphasised to the British authorities his responsibility for securing access to the trade British merchants needed in southern Arabia. He also tried to exploit the growing Ottoman military campaign to his north. In a series of letters sent to the British Resident in Aden warning of political and commercial turmoil in the region, he linked Ottoman successes in the highlands to threats of instability in and around Aden in an attempt to secure his position with the British. In one of Fadl’s letters, the sultan’s disquiet is evident as the British had not responded in a manner that had given confidence to neighbouring leaders. The Ottoman commander Muhtar Pasha was aggressively offering alliances as well as asserting that although the Sultan in Istanbul had ultimate sovereignty, he was a benevolent ruler wishing to see Yemen prosper. Fadl, having heard nothing back from the Aden authorities, openly pleaded for some sign of action: I am the more afraid as you gave me no assurance not to fear. If you had done so, I would have nothing to fear; nor did you write what answer to give to the Imam [Zaydi Imam Sayyid ‘Ali b. Ahmad] in case you think that an answer should be sent; if so, tell me what reply I am to send. (p.294) If you tell me to remain quiet I will do so. You know the power of this [Ottoman] army; you know also my condition, the state of my country and people. In comparison with them I am as a drop in the ocean. If I had no connection with Aden like other hill tribes of Yemen, I would act according to the dictates of my reason. I do not want to be under the control of any other Government save yours. If a person has got authority, and has orders to do a thing, he will do it, whether willing or not. Lahj is the garden [breadbasket] of Aden, and it contains also a Government garden. I feel sure you will not desert me without giving me assurance of a nature to set my fears at rest. If you will give me confidence, I will not give a reply [to offers of alliance with Ottomans]; otherwise I will make my own arrangement about the safety of my property, of my blood and of my honor. I will then say that I was constrained to do so, and I did so. This is a real hardship to me, as I cannot resist them, owning to my feebleness. It is equally hard not to be able to make any reply without your orders, as I am under your protection.18 British authorities in Aden, far from ignoring Fadl this time, appeared to share his concerns. In an early report to Bombay, Major-General Tremenhere reminded his superiors that the Ottomans were making territorial claims at British expense. For instance, the Ottomans claimed areas just outside Aden on the basis of their sixteenth-century rule over Yemen, arguing that the sultan in Istanbul was the only legitimate sovereign of the area. Tremenhere warned the Government of India that Sultan Fadl of Lahj was no longer able to maintain his authority over many of his allies, who were increasingly attracted by Ottoman overtures.19 Many of Fadl’s subordinates were also concerned that their access to the coffee fields was being lost as Ibb and Ta‘izz redirected their goods to the Ottoman-controlled Red Sea ports of Hudayda and Mukha. Economic considerations alone posed long-term threats to the sustainability of Aden as a commercially viable enterprise.20

Page 4 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914 Six months later, letters from Aden were still not receiving much interest in India or London. Brigadier-General Schneider, the new Resident in Aden, wrote that as Resident, he was ‘in ignorance of the policy proposed to be followed as regards the overtures that were certain to be made, sooner or later, by the Pasha of Yemen himself to the chieftains of the principalities bordering on Aden for the transfer of their adherence to the Turkish Government.’ Schneider is clearly suggesting the Government of India stand up to the Ottomans, if for no other reason than Britain’s responsibility to (p.295) the many communities formerly linked to Aden: ‘I need scarcely refer…to the feeling with which they [local allies] will regard us, if they are coerced by the Turks against their will, while we, as they would suppose, looked on with indifference or with no power to prevent such a proceeding.’21 Schneider, still not receiving instructions as to how to deal with a problem he saw as undermining British control of Aden, resorted to recruiting a well-established merchant to bring the issue to the British government’s attention. Sir Bartle Frere (1815–84) warned of a threat to commercial interests in India and China and wrote to Sir Henry Elliot, ambassador in Istanbul, in unequivocal terms. Citing the role of the Sultan of Lahj, ‘chief of ‘Abdali’, as the primary supplier of food for Aden, he reminded Sir Henry that Fadl b. Muhsin received a stipend: ‘he and his [Sultan of Lahj] people are as much our servants as your French cook or Italian gardener are yours, and the Porte has by right or prescription even less authority over them, for it has absolutely none at all.’ In light of Muhtar Pasha’s attempts to recruit the services of the Shaykh of ‘Abdali, Frere continues: ‘It is difficult to imagine any parallel aggression on us, by so friendly a power…I cannot explain the fact that the Resident here, after reporting this and several aggressive moves of the [Ottoman regime]…is still without instructions, save an injunction to his predecessor on one of his earlier reports some months ago to be careful not to commit himself, without further orders, to any counter-action.’22 The Foreign Secretary, Earl Granville, wrote to Sir Henry stating, apparently in response to Frere’s hysteria, that ‘Her Majesty’s Government’ is confident that ‘no operations have been or will be commenced against the Sultan of Lahej without orders from the Porte, which will do nothing without communicating with [London].’23 Such confidence in diplomacy seemed misplaced to those on the ground. Schneider and the Sultan of Lahj pointed out that subordinates like ‘Ali b. Mani‘, chief of the Hawshabi, were dangerously close to joining the Ottomans, a move that could lead to an invasion of the suburbs of Aden itself as a large expeditionary force had been organised in Ta‘izz in January 1873. The letters sent between Aden and the British authorities in London and India stressed the link between community leaders nominally under of Fadl’s ‘authority’ seeking Ottoman patronage and a direct military threat to Aden’s security.24 The Ottomans established a regional administration at Qa‘taba by May 1873. The proximity of this garrison posed an immediate challenge to the authority of regional (p.296) allies of the British like Fadl. Others astutely played the growing tensions to their financial advantage. ‘Ali b. Muqbil b. ‘Abd al-Hadi, for example, exploited the opportunities presented to him by a new Ottoman administration in Qa‘taba by accepting the offer to collect on behalf of the Ottoman authorities 800 dollars a year in local taxes in return for a 40-dollar monthly salary.25 Fadl, long assuming ‘Ali b. Muqbil was a subordinate, had little recourse but to solicit the assistance of the British Resident in Aden. Schneider reportedly tried to remind ‘Ali b. Muqbil of his responsibilities to the British.26 Fortunately for Fadl and his Aden-based British sympathisers,

Page 5 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914 some officials in Bombay had begun to recognise that these developments constituted a serious problem.

A Contingent British Response In February 1873 the Government of India responded to the complaints from Aden. Its answer was based on the notion that the so-called Sultan of Lahj was an independent sovereign under no other authority than his own—in sharp contrast to Bartle Frere’s considering him a ‘servant’ no different than Elliot’s Italian gardener—and was thus immune from Ottoman pressure as long as he maintained a diplomatic alliance with Britain via a treaty that had originally been signed in 1802: We cannot too strongly express our opinion that any interference on the part of the Turkish Government with those Chiefs who have Treaties with, or receive stipends from us, is most dangerous to our interests at Aden, and should be most peremptorily prevented. The assurance which the Turkish Government has given respecting the Sultan of Lahej is not sufficient, and we are strongly of opinion that nothing will meet the circumstances of the case short of immediate and distinct instructions to the Turkish authorities in Yemen to abstain from all interference, direct or indirect, which has for its object to induce any of the Chiefs who have relations with us to acknowledge the supremacy of the Ottoman Porte. We have called upon the Resident at Aden [Schneider] to submit a list of such Chiefs, which we will forward when received. In the meantime, we have authorized the Resident to inform the Turkish officials in Taizz that perseverance in their present proceedings will be considered an unfriendly act towards the British government. The Resident will also inform the Chiefs who are threatened of the views entertained by us, and that the proceedings of the Turkish officers are believed not to be authorized by the Government of the Sultan. In the event of these measures proving insufficient to check further encroachments, we shall be prepared,…to take the Chiefs directly under our protection and to engage to defend them from aggressions [which] are destructive to the peace of the country and most injurious to British interests in Aden.27 (p.297) Clearly Schneider’s persistent lobbying had finally attracted Bombay’s attention. In rapid order, Bombay approved a plan for Schneider to send out an expeditionary force to reassure wavering ‘chiefs’ that the British were to be taken seriously and that they were both firmly behind the Sultan of Lahj and able to commit to long-term formal alliances with individual communities willing to promise their allegiance.28 This quickly produced results: the British authorities in Aden were able to secure nominal alliances with nine separate community leaders. On 10 February 1873 Schneider communicated to the Ottoman authorities in Ta‘izz the list of local shaykhs whom Aden considered to be under treaty with the British. The tribes under this ‘Protectorate’—‘Abdali (Lahj), Fadli, ‘Aqrabi, Hawshabi, ‘Alawi, ‘Amiri, Subayhi, Yafi‘i, ‘Awlaqi— were to be treated as formally independent entities that would be firmly under British protection.29 Interestingly, the formal declaration of the Protectorate did not include territorial boundaries, as the geography of the region was more or less unknown.30 This stopgap effort to stem the rapid deterioration of Aden’s position among its vulnerable allies seemed to put an immediate halt to Ottoman overtures in some areas. In the case of Fadl, further controversy would arise because of locals themselves refusing to respect these British claims of protection.31 Thus ‘Ali b. Mani‘, the Hawshabi leader, was repeatedly pressurised to stop accepting Ottoman money. His response, however, proves a crucial reminder that local calculations contributed far more to this process than previously acknowledged by historians:

Page 6 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914 As to the Turks you say cut them off from entering my land. I cannot hinder them, they are the soldiers of the Government. The land is God’s and the Sultan’s of Rum. The whole district could not hinder them. As to my pay, if it is cut off on account of the fault of the Turks [then] for me my own judgment [will remain to be acted on]. And concerning the roads, if my pay is cut off, it will cut off the roads; and I have a wild country.32 This constituted a threat to both the Sultan of Lahj and his British allies: unless adequately rewarded for abandoning the Ottomans, who had historical legitimacy, communities like the Hawshabi could undermine Aden’s ability to secure the region. The Ottomans seemed to penetrate to the very heart of British alliances. Sultan Fadl’s younger brother ‘Abdallah, for example, accepted Ottoman overtures even as late as a year after Schneider’s aggressive actions. The governor of Ta‘izz promptly rewarded ‘Abdallah with a public commendation and a position in the Ottoman (p.298) administration. In no time, he became a valuable ally for the Ottomans and adeptly subverted British influence in and around Lahj, openly challenging his sibling’s authority. In mid-1874, the British tried in vain to convince the Ottomans that they had no ‘legal’ right to challenge the ‘rightful’ sovereign leader of Lahj by backing his younger brother. Unfortunately for the British, Ottoman ascendancy to the north completely changed the local dynamics in and around Lahj, compelling them to take further measures to ensure that nominal British allies would not follow the path ‘Abdallah had taken.33 These and other challenges to British presumptions of hegemony in the region pose the question of what exactly attracted locals to the Ottoman administration. It was not purely a result of the Ottomans’ ability to project power, conjure up sentiments of religious solidarity or offer more than the British in terms of stipends. Rather, the Ottoman state rewarded pliant locals seeking opportunities to circumvent trade networks dominated by established regional leaders. Not only did new trade possibilities arise with the arrival of Ottoman authorities—garrisons needed to be fed, roads protected, taxes collected—but an alliance with the Ottoman administration demonstrated to many locals sitting on the sidelines that previously subordinate elements of southern Yemen’s hierarchical society could find new political or economic niches in the region.34 Ottoman successes highlighted the weaknesses of the British and their allies in the region. Britain’s failure to understand the situation convinced the Aden authorities that they needed to be better informed about events in the highlands and to co-opt more directly communities living in areas vulnerable to Ottoman overtures.35 In time, the need to gather better local intelligence would transform the way the British administered Aden and dramatically raise the stakes for local potentates.36 However, the imposition of a border would take many years.

Harnessing Imperial Patronage It was not until 1880 that any formal treaty was actually signed with any of these local allies, a full six years after this initial scramble to secure Aden’s role in the hinterland. This suggests the limited extent to which both powers influenced the day-to-day affairs of the region. The treaty signed in 1880 between Aden officials and the Amir of Dali‘, (p.299) a subdistrict located deep in the highlands of what was previously Lahj territory, is suggestive of the complexities on the ground. Both the Ottomans (who did not object to the formal treaty between Aden and Lahj) and the British were concerned by recent attempts by locals to infiltrate and undermine neighbouring districts. The push for greater access to the rich agricultural land belonging to villages located in the Qa‘taba valley, a region firmly under the Ottomans’ protection, risked infringements on the spheres of interest of Ottoman clients based in Qa‘taba. Local inhabitants needed to be incorporated into a formal treaty in order to avoid any violence that could spill over Page 7 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914 into the southern highlands. The treaty of 1880 may in fact have been an attempt to give more direct authority to the Amir of Dali‘ to rein in his subordinates.37 Meanwhile, the British strove to survey at last the regions in which they had an indirect political and economic stake. The surveys, initiated in 1886, would continue for six years, leading to a set of reports.38 For the most part, this surveying of the region did not result in any dramatic confrontation between the Ottomans and the British.39 The problem was more local communities pitted against each other in smuggling goods in and out of Ottoman-controlled areas. Thus the British tried to strengthen ties locally in the hope of containing better the region’s commercial activity. In the course of these surveys, Aden’s officials signed agreements with the purported ‘shaykhs’ of Baysi, ‘Amiri, ‘Alawi and Hawshabi. Over the next ten years, ten other ‘shaykhs’ of the ambiguously defined frontier zone also signed Protectorate treaties with Aden, assuring future British claims of a sphere of influence in hotly contested areas. Against British intentions, these treaties (and the subsequent distribution of stipends and highly sought-after rifles) set off another scramble for influence among locals. Over 1885–1900, the treaties that were intended to strengthen local and state collaboration in fact resulted in social disruption on the ground. Increasingly nervous about the intermittent clashes occurring as officials attempted to control trade from the Ta‘izz district down into Dali‘, authorities in Istanbul began to interpret such treaties as signs of British aggression, not as stopgap reactions to consistent efforts by locals to threaten the stability both the British and the Ottomans wanted.40 Poor communication and the aggressive activities of some of these locals transformed a relatively successful example of imperial cohabitation. Ultimately, it was Muhammad Nasir Muqbil of Humar whose ascent as a regional power destroyed the possibility of maintaining order without boundaries and resulted in their imposition. Between 1898 and 1901, as an agent of the Ottomans, Muqbil (p.300) recruited local allies at the expense of his rivals (including the British surrogate based in Dali‘) by distributing Ottoman money and weapons. In return, Muqbil was able to restrict access to key trade routes to his allies, linking merchants in isolated inland communities to traders based in the ports of Mukha and Hudayda and as a consequence excluding Aden’s merchants from an increasingly lucrative regional trade network. For a while the British authorities dealt with Muqbil by belittling him in their reports. Insults only went so far, however, as he increasingly played a leading role in local politics.41 By the late 1890s, Muqbil had created a cluster of alliances that paralleled those of a state and he proved capable of disrupting the balance of power in the region. For Muqbil, any attempt to deny him access to trade, ostensibly any regulatory measure initiated by the British, was a possible cause for open confrontation. This occurred when the British decided to reinforce the territorial reach of their new ally in the region, Muhsin b. ‘Ali b. Mani‘, son of the same ‘Ali b. Mani‘ who had proved elusive in 1873, and a rival of Muqbil’s. The British formally recognised Muhsin b. ‘Ali b. Mani‘ as the Hawshabi Sultan in August 1895. This territory encompassed much of the region that Muqbil actually controlled. Muhsin b. ‘Ali, as signatory of a ‘protectorate treaty’ with the authorities in Aden, was not only afforded ‘sovereign rights’ over Hawshabi, but also important supplies of weapons and a financial subsidy in order to enforce it. According to the map drawn by the British officer Colonel Wahab during the survey of 1886–92, Muhsin b. ‘Ali was ‘shaykh’ of the entire territory designated as Dali‘, a territorial claim that apparently did not correspond with the actual political and economic realities on the ground. When Aden declared that Muhsin b. ‘Ali was indeed sovereign over these important Page 8 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914 lands, Muqbil simply occupied Darija village and constructed a fortress. Because the dispute between Muqbil and British allies was territorial, the maps produced by Colonel Wahab (the only ones of their kind at the time) had to be dusted off and used to reaffirm the claims of the Amir of Dali‘, Muqbil’s rival and a British client. The contested fort lay within a zone the British deemed to be part of the extended territory of the Dali‘ amirate by way of Wahab’s map.42 In response to Muqbil and early Ottoman counter-claims to the territory, officials in Aden and India ordered even more studies to prove their position. One of these ad hoc survey teams was attacked by Muqbil’s allies, while British allies from the Hawashib, armed with a British cannon, retaliated by infiltrating Ottoman-administered territories. Matters were quickly spiralling out of control. (p.301) Muqbil thus exposed the underlying weakness of the British position in Arabia as well as dragging the Ottoman Empire into a larger diplomatic crisis it could ill afford at the time. The British decided to cage Muqbil in and by doing so ensure that Aden maintained capacity to run its affairs through local proxies. The only way they could do this, however, was to change fundamentally how they operated in Yemen.

Territorialising Yemen In response to Muqbil’s actions, a new form of delimitation of territory was needed, a formal frontier, beyond mere treaties and polite correspondence about British alliances with one shaykh or another. However, the British had first to convince the Ottoman authorities that a boundary was needed, as they tried at the Berlin Congresses in 1878 and 1885–6. The easiest way to exert pressure on the Ottomans was to support the smuggling of weapons and goods into Ottoman Yemen. This led to the influx of, among other things, modern rifles which, in time, would threaten the British and Ottoman monopoly of modern military technology in the region.43 Furthermore, Italian and French manipulation of smuggling and piracy influenced the situation on the ground by the late 1890s.44 Eventually, the two powers agreed upon a joint commission to delimit a boundary and establish a working frontier that could then be policed by the British to help the Ottomans with their smuggling problem.45 From the commission’s first meeting in February 1902 until April 1905, the machinations of influence peddling, claims and counterclaims, exactly what the British had so dearly hoped to avoid during the 1880–1900 period, were in evidence.46 Competing communities, even factions within the same family, lobbied both the Ottoman and British border commissions for support. The Ottomans, in particular, proved adept at preempting any British claim to areas by erecting symbols of direct (p.302) administration, either by placing troops in villages that welcomed them or erecting forts on the land of allies. A great deal of ambiguity existed as to the Amir of Dali‘’s claims. The more the British authorities learned about the precarious nature of the Amir’s formal claim to larger areas, including the Sha‘iri, Bayda and Bayhan, the more the Ottomans exploited the local activism of communities in Lakmat and Kharaba.47 In time, the British authorities realised that they needed to clearly designate areas on a map. However, the only available map was that produced by Colonel Wahab’s survey. As both parties quickly learned, the maps drawn by Colonel Wahab were out-dated and ambiguous. Wahab himself, a veteran of the region and Aden’s chief negotiator, complained that the India Office misunderstood faraway Yemen through ‘misinterpretations’ of his map.48 For Colonel Wahab, the British needed to investigate ‘actual circumstances’ on the ground (determined by the Amir of Dali‘’s present alliances) and reinforce those political realities with documentary evidence that the Ottomans would have to accept, even if they undermined an already weak Amir of Dali‘.49

Page 9 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914 The Ottomans understood the direction of Wahab’s thinking and were confident they could counter any claim. They had officials on the ground collecting the kind of evidence that would reinforce Ottoman claims that the Sultan of Lahj and his subordinate, the Amir of Dali‘, were in fact marginalised among the Hawashab.50 Ultimately, the ambiguity of the map first introduced by the British, as well as documented assertions that the Humaydi and Ahmadi areas, for example, were not under British allied control, supported Ottoman counter-claims.51 Wahab often reported with great frustration that his Ottoman interlocutors ignored the geographic logic he was trying to use.52 With a complex local and international situation and some elements of the British government willing to settle with the Ottomans with as little confrontation and eventual cost as possible, Wahab’s tactics were in vain.53 That is not to say that the local dynamics were entirely working against the British commission. On the contrary, the Amir of Dali‘, long allied with Aden, began to produce (p.303) the kind of documentation needed to reinforce territorial adjustments to the eventual boundary. It appears that he encouraged his local allies to pay visits to the commission for formal interviews that helped determine under whose authority they operated, and a stream of shaykhs visited the encamped Border Commission in February and then April, August, September and October of 1902. All those who came either declared previously held agreements with the Amir of Dali‘ or even requested British protection, in other words, to be incorporated into the scheme of the Protectorate originally designed as the ‘nine tribes’.54 Wahab thus recorded in his reports and presumably shared his impressions with his Ottoman counterparts that a number of local shaykhs owed allegiance to the Amir in direct contradiction to Ottoman claims.55 The result of these initial meetings was a map which Wahab would draw to delineate the Amir’s authority, taking into account both historical and current political alliances. The map, enclosed with a memorandum outlining the case Wahab was making on behalf of the British government, showed that the ‘autonomous tribes’ all acknowledged the Amir’s supremacy, and in most cases paid him tribute. At the same time, Wahab reported that ‘all important matters affecting the tribes are referred to him [the Amir], and his influence and authority are constantly exercised.’56 It is important to note the role locals played in reaffirming the extent to which British or Ottoman claims would be justified. This constitutes a remarkable example of locally applied power. In the end, however, it worked: the Ottomans agreed on terms of negotiation that would eventually lead to a treaty.57

Conclusion The negotiations with the Ottomans over how exactly to define the boundary separating each party’s domain were largely confused by a completely different set of criteria. The Ottomans would constantly argue that the areas they claimed (large areas of which the British contended existed within ‘Dali‘’ territory) had historically and thus always formed part of Ottoman territory. They installed troops in the areas in dispute and actually started to collect taxes, in part thanks to Muqbil’s aggressive alliance-building. The longer this physical presence was maintained, the more difficult it would be for the (p.304) British to argue that these areas were actually Dali‘.58 A serious impediment for the British was the confusion within the cartographic material they were using to make their claims.59 It was largely the growing insurgency in Ottoman Yemen, in some respects a product of British machinations, that ultimately led to the 1903 capitulation by the Ottoman authorities to British demands for formal control of all of the Dali‘ plateau, which could not have been agreed upon before.60 It is clear that efforts to delineate the frontiers of Yemen had consequences for all parties involved. For some locals, a new set of opportunities opened up with the arrival of a Page 10 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914 dynamic increasingly mirroring those found in eastern Africa and Asia. In the eyes of potential European or Ottoman state patrons, the less-than-servile activism of local communities meant that the political and economic parameters of empire were constantly being negotiated. The kinds of transactions the two imperial states wanted to regulate and ultimately draw revenue from were thus constantly being undermined by local political realities. These local factors, by way of direct interaction with each other and by extension, the Ottoman and British imperial regime, prove to be the roots of the modern frontiers of southern Arabia. Thus the borders of the empire in the end were the result of bureaucratic reactions to indigenous agency. It is especially instructive, in this regard, that the two powers discussed here tried to avoid confrontation and until 1901 did not countenance drawing a border. Thus the hesitation in laying down formal boundaries in Yemen is suggestive and could be the source of serious theoretical rethinking of the entire paradigm of modern imperialism. Notes: Proceedings of the British Academy 156, 289–304. © The British Academy 2009. (1) For a copy of the final meetings which sealed the actual delimitation of the border that would eventually be codified by a treaty signed between the two states in March 1914, see TNA, Foreign Office files (FO) 371/1805, ‘Minutes of meeting at Shaykh Said signed by Colonel Mustafa (Ottoman Commissioner) and G. H. Fitzmaurice (British Commissioner)’, dated 20 April 1920, and for a copy of the signed convention, see TNA 372/573, ‘Signing of the Convention between the United Kingdom and Ottoman Government relative to Aden boundary’, dated 4 and 9 March 1914. (2) For two recent studies of these transformations in the region, see Keith d. Watenpaugh, Being Modern in the Middle East: Revolution, Nationalism, Colonialism, and the Arab Middle Class (Princeton: Princeton University Press, 2006) and Jens Hanssen, Fin de Siècle Beirut: The Making of an Ottoman Provincial Capital (Oxford: Oxford University Press, 2005). (3) For a sophisticated discussion on how subaltern studies has influenced our study of empire see Frederick Cooper, Colonialism in Question: Theory, Knowledge, History (Berkeley: University of California Press, 2005). (4) On the British takeover of Aden see R. J. Gavin, Aden under British Rule, 1839–1967 (London: C. Hurst, 1975), 1–38; C. E. Farah, The Sultans Yemen: Nineteenth-Century Challenges to Ottoman Rule (London: I. B. Tauris, 2002), 14–29, 120–6. (5) For a comprehensive account of the process of incorporating British interests with the Sultans of Lahj see Robert Playfair, A History of Arabia Felix or Yemen (Amsterdam: Philo Press, 1970), 159–64. (6) Farah, The Sultan’s Yemen, 1–13, 58–81. (7) See among others, ‘Abdullah al-Jirafi, al Muqtataf min Ta‘rikh al-Yaman (Beirut: Manshurat al-‘Asr al-Hadith, 1951), 205–6 and ‘Abdullah b. Muhammad al-Hibshi, Hawliyat Yamaniya min sanat 1225 h. ila sanat 1316 h. (San‘a’: Wizarat al-I‘lam wa-l-Thaqafa, 1980), 296. (8) I have argued elsewhere that such a dynamic should not be surprising: Isa Blumi, ‘The Ottoman Empire and Yemeni politics in the sancaq of Ta‘izz, 1911–1918’, in J. Hanssen, T. Philipp

Page 11 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914 and S. Weber (eds.), The Empire in the City: Arab Provincial Capitals in the Late Ottoman Empire (Beirut: Beiruter Texte und Studien, 2002), 349–67. (9) Farah, The Sultan’s Yemen, 77–87. (10) See Muhammad b. Muhammad Zabara, A’immat al-Yaman bi-l-qarn al-rabi‘ ‘ashar (Cairo: alMatba‘a al-Salafiyya, 1956) 2: 96; al-Hibshi, Hawliyat Yamaniya, 347–350, and ‘Abd al-Wasi‘ alWasi‘i, Ta’rikh al-Yaman (Cairo: al-Matba‘a al-Salafiyya, 1928), 122–5. (11) Husayn b. Ahmad al-Sayaghi, Safahat Majhula min Ta’rikh al-Yaman (Sana‘a’: Markaz alDirasat al-Yamaniyya, 1978), 125. (12) Ahmad Fadl b. ‘Ali Muhsin al-‘Abdali, Hadiyat al-Zaman fi Akhbar Muluk Lahj wa ‘Adan (Cairo: al-Matba‘a al-Salafiyya, 1932), 34–41. (13) For an understanding of how Yemen was seen in strategic terms by the 1872–3 campaign’s principal administrative architect, see Ahmed Raşid, Tarih-i Yemen ve San’a (Istanbul: Basiret Matbaası, 1291), 2: 258–9, 351–5. (14) BOA, İrade-Meclis-i Mahsus 1922, Ahmed Muhtar Pasha’s report to the Grand Vizier (22 January 1873), document 1–2. (15) On the extent to which courts were in operation by 1875–6, see BOA, Sadaret Mektubi Mühimme Kalemi (A.MKT.MHM) 486/18, Mustafa ‘Asim Pasha to Grand Vizier (27 July 1876). (16) The power struggle among local communities and the empires making their separate claims to the region is reflected in letters saluting in unequivocal terms the role of the Ottoman sultan in bringing stability and harmony. See BOA, Hariciye Nezareti Siyasi (HR.SYS) 87/47, open letter of support to the sultan signed by twenty-three village leaders in the Lahj region (2 December 1872). (17) Gavin, Aden under British Rule, 127–9. (18) TNA, FO 424/32, Translation of letter from Sultan Fadl of Lahj to Major-General Tremenhere received 17 May 1872, enclosed in Major-General Tremenhere to Gonne at Bombay Castle, dated Aden, 21 May 1872, file no. 4. (19) Indeed, the Ottomans were making all kinds of promises. In one instance, a letter from Shaykh Amin b. Qasim Husayn to Sultan Fadl b. Muhsin testifies that Ahmed Muhtar Pasha was very persuasive while making historical arguments as well as respectful of Yemeni customs. He treated everyone with dignity and offered lucrative positions, appointing Amin b. Qasim Husayn as shaykh of his home district. British sources reported that there was talk of Sultan Fadl communicating with Muhtar Pasha and Amin b. Qasim, who were both soliciting his allegiance in return for similar political recognition. For commentary on the circular of Pasha sent to all communities see TNA, FO 424/32, Notice circulated by the Pasha, Governor-General of Yemen [Muhtar Pasha] dated San‘a’, 18 Safar 1289 [26 April 1872], document no. 4, enclosure 21. (20) TNA, FO 424/32, Major-General Tremenhere to Gonne at Bombay Castle, dated Aden, 7 May 1872, file no. 4, enclosure 5.

Page 12 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914 (21) TNA, FO 424/32, Brigadier-General Schneider to Gonne at Bombay Castle, dated Aden, 26 October 1872, file no. 5. (22) TNA, FO 424/32, enclosed copy of private letter sent by Sir B. Frere to Sir H. Elliot, dated Aden, 2 January 1873, file no. 14. (23) TNA, FO 424/32, official correspondence no. 15, Earl Granville to Sir H. Elliot, dated Foreign Office, 23 January 1873. (24) TNA, FO 424/32, copy of letter from Brigadier-General Schneider to Mr. Gonne in Bombay, dated Aden, 4 January 1873, file no. 21. Interestingly, Schneider, apparently losing all faith in his superiors, sent a copy to the Duke of Argyll, the Secretary of State for India, in an attempt to involve others in this debate as a means to pressurise Bombay and London into action. (25) TNA, FO 424/32, Brigadier-General Schneider to Mr. Gonne in Bombay, dated Aden, 23 January 1873, file no. 30B, enclosed letter 5. (26) TNA, FO 424/32, Brigadier-General Schneider to Mr. Gonne in Bombay, dated Aden, 12 May 1873, file no. 52, enclosed letter 3. (27) TNA, FO 424/32, The Governor-General of India in Council to the Duke of Argyll, dated Fort William, 14 February 1873, file no. 30B. (28) See Harold Ingrams, ‘The exploration of the Aden protectorate’, Geographical Review, 28 (1938), 634–9. (29) TNA, FO 424/32, Brigadier-General Schneider to Mr. Gonne in Bombay, dated Aden, 10 February 1873, file no. 30B, enclosed letter 15. (30) British Library, India Office Records (henceforth IOR) R/20/E/96, No. 38, Declaration of Protectorate, Government of India (Foreign Department) to Duke of Argyll, dated Fort William, 11 April 1873. (31) TNA, FO 424/32, Brigadier-General Schneider to the Duke of Argyll, dated Aden, 27 May 1873, file no. 42. (32) TNA, FO 424/32, Brigadier-General Schneider to Mr. Gonne, dated Aden, 18 May 1873, file no. 52, enclosed letter 5. (33) BOA, HR.SYS 90/3, correspondence no. 5516/166, Embassy to Sublime Porte, dated London, 13 June 1874. (34) An informative first-hand account of this process of securing British control over ‘Abdali may be found in Harold Jacob, Perfumes of Araby: Silhouettes of Al Yemen (London: Martin Secker, 1915), 213–25. (35) TNA, FO 424/32, Brigadier-General Schneider to the Duke of Argyll, dated Aden, 27 May 1873, file no. 42. (36) F. M. Hunter, and C. W. H. Sealy, An Account of the Arab Tribes in the Vicinity of Aden (1909) (London: Darf Publishers, 1968), 4–6.

Page 13 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914 (37) BOA, HR.SYS 90/14, report 34, copy of letter from the mutessarif of Ta‘izz to Ahmed Izzet Pasha, governor of Yemen, dated 13 July 1884. See also Gavin, Aden under British Rule, 200–2. (38) The outcome of these forays into the Aden hinterland is reproduced in Hunter and Sealey, An Account of the Arab Tribes (1909), 235–321. (39) See examples of how Ottoman officials were opposed to such activities, BOA, Meclis-i Vükelâ Mazbatakları 106/89 (25 July 1903). (40) BOA, HR.SYS 90/7, London Embassy to Said Pasha, no. 15212/111, dated London, 14 May 1892. (41) The initial reports on Muqbil’s occupation of a key trading post in Darija reveal this inability to take Muqbil seriously, a position the Ottoman authorities could not afford to adopt themselves. BOA, Yıldız Tasnifi: Sadâret Hususî Mâruzât Evrakı 418/65 (2 August 1901). (42) For a clear copy of the Wahab map that represented the boundaries of Dali‘, see TNA, FO 925/171007, ‘Section from Aden Frontier Sketch’, by Colonel R. A. Wahab, 1903. (43) For information on the smuggling of guns into Yemen at the height of the border crisis discussed below and how that was identified as a central cause of the disorder in the region see BOA, Yıldız Mütennevia (Y.MTV) 238/58 (12 January 1903). On how the Ottomans often failed to make distinctions between ‘criminal’ and normal local behaviour in the context of lower Yemen and the issues surrounding the smuggling of weapons see BOA, Y.MTV 238/69 (13 January 1903). (44) Events over the period of 1900 to 1911 hint at Italy’s long-term strategy to gain influence in Mukha by supporting individuals like ‘Umar Salim, a local merchant and political deal-maker. BOA, HR.SYS 1568/2. For a detailed report on the French connection in Yemen’s weapons trade that clearly gave them political leverage in Mukha see BOA, Y.MTV 238/82 (15 January 1903). (45) IOR, Political and Secret Records Department (henceforth L/P&S) 10/63, P. J. Maitland, The Political Resident, Aden to Sir W. Lee-Warner, dated Aden, 6 November 1901 and enclosures. (46) All too often, the diplomatic exchanges were forged around a rhetoric that denigrated local actors, calling their actions ‘illegal’. These so-called illegal activities involved a number of actors who were operating in a field of logic that carried a great deal of importance on the ground. For a detailed report see BOA, Yıldız Sadaret Hususi Mazuret Evrak (YA.HUS) 418/65 (2 August 1901). Further disturbances caused a rise in tensions resulting in the dispatch of British troops towards the unmarked frontier, see BOA, YA.HUS 418/66 (3 August 1901). (47) IOR, R/20/E/233, Wahab to Secretary of Governor of India, dated Aden, 7 April 1902. (48) IOR, R/20/E/233, Wahab to Secretary of Governor of India, dated Aden, 16 May 1902. (49) IOR, R/20/E/233, Wahab to Governor of India, Foreign Department, dated Dali‘, 7 April 1902. (50) BOA, Yıldız Perâkende Evrakı, Umum Vilayetler Tahriratı 22/93, Mehmed Ziya to Palace, dated 8 June 1891.

Page 14 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Frontier as a Measure of Modern Power: Local Limits to Empire in Yemen, 1872–1914 (51) IOR, R/20/E/234, Wahab to Secretary of Governor of India, dated 13 June 1902. To this charge Wahab produced documents claiming to prove that the ‘Amiri leader had gone into some of the contested areas (that were not marked on his 1892 map as he did not survey those areas) and actually collected taxes as late as 1901, thus contradicting Ottoman claims that their troops were stationed and operating in authority there. The actual documents to make Wahab’s argument about the Amir’s tax collection may be found in IOR, R/20/A/1198, Wahab to Governor of India, dated Dali‘, 25 May 1903. (52) For extensive survey of the actual Aden delimitation process, consult IOR L/P&S/18/B137, P. J. Maitland, Brigadier-General, Resident at Aden, ‘Aden Delimitation: The History of the Question, and the present situation as regards the Territories of the Amir of Dthali’, pages 1–10, dated India Office, 28 July 1902. (53) IOR, R/20/E/233, Political Resident to Bombay, dated Aden, 7 March 1902. (54) IOR, L/P&S/10/64, No. 236A, From Secretary of State to Viceroy, 25 November 1902. (55) For a list of local leaders coming to assert their association with the Amir of Dali‘, see Wahab’s diary entry for February, including IOR, R/20/E/233, Political Resident to Bombay, dated Aden, 6 February 1902. (56) See enclosed memorandum on ‘Amiri frontier in IOR, R/20/E/234, Wahab to Foreign Department, India Office, dated 2 December 1902. For a copy of the map to which Wahab refers, see Doreen Ingrams and Leila Ingrams (eds.), Records of Yemen, 1798–1960 (London: Archive Edition, 1993), 5: 97. (57) For details of these conversations and overtures to the British by locals, see IOR, R/20/E/ 234, Wahab to Dane, Secretary to the Governor of India, dated Aden, 2 December 1902. (58) Indeed, as reported by Colonel Wahab, the chief liaison for the Government of India in the negotiations, Ottoman troops and their local allies were collecting taxes in key districts— Humada, ‘Ammara and Wadi Safiyya—that constituted a significant threat to the road network leading to Aden. IOR, R/20/E/233, Wahab to Secretary to the Governor of India, dated 23 April 1902 and summary of Wahab’s reports in IOR, R/20/E/233, Political Resident to Political Department in Bombay, dated Aden, 24 April 1902. (59) IOR, R/20/E/234, Curzon to Secretary of State for India, dated London, 13 June 1902. (60) IOR, L/P&S/10/65, No.1, ‘Memorandum on the Boundary of the British Protectorate near Aden’, dated Foreign Office, 3 January 1903; IOR L/P&S/10/66, No. 1, R. A. Wahab, Commissioner, Aden Boundary Commission to Government of India, dated Aden, 20 March 1904; and for the 37-page report on the nearly complete delineation from the south-west frontier at Shaykh Sa‘id to the Yafi‘i regions in the north-east, more than 130 miles of boundary, see IOR L/ P&S/10/66, P. J. Maitland, Major-General, Resident at Aden, ‘Report on the demarcation of the frontier between the tribes in the Protectorate of Aden and the Turkish Province of Yemen’, dated Aden, 12 June 1904.

Access brought to you by: Page 15 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 SARA NUR YILDIZ

DOI:10.5871/bacad/9780197264423.003.0016

Abstract and Keywords This chapter examines the initial Ottoman campaign of conquest of the Karamanid principality with the assault on Gevele, a mountain fortress lying 11 kilometres to the west of Konya, and the seizure of Konya in 1468. It also looks at the establishment of administrative and military control of the fertile agricultural plains as well as over the rich steppe grasslands of the region's northern highlands by the Ottomans on the eve of their conquest of the Karamanid principality in order to reveal an important factor in the conquest of Karaman: Ottoman political involvement in Karamanid internal affairs before the conquest. By interfering in Karaman succession politics, the Ottomans laid the political groundwork by which they were able to establish an administrative and military presence in the hinterland of the more vulnerable stretches of Karamanid territory prior to the Ottoman military conquest. Keywords:   Karamanid frontier, Karaman succession politics, Gevele, Konya, Ottoman conquest, hinterland

Bu Karamanoğulları’nın hikayesi uzundur. (The story of the Karamanids is a long one.) Aşık Paşazade WHILE SURVEYING THE MEDITERRANEAN COAST OF ANATOLIA for the British Admiralty in 1811–12, Captain Francis Beaufort was struck by the absence of Ottoman government over the rocky central coast of Karaman lying to the immediate north of Cyprus: ‘Sheltered from all effectual control of the Porte by the great barrier of the Mount Taurus, the half-independent and turbulent Pashas, amongst whom they are parcelled…engaged in constant petty hostilities with each other, so that their respective frontiers change with the issue of every skirmish.’1 Just as the early nineteenth-century Ottoman state experienced difficulties in asserting central control over parts of south-central Anatolia, especially along its rugged coastland and mountainous interior, Roman imperial rulers centuries earlier were confronted with rebellion and unrest in Page 1 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 the region. Indeed, its inhabitants had a reputation for being difficult to rule. The Isaurians, the indigenous inhabitants of the Roman (p.308) south-central Anatolian province of the same name,2 have left their mark in history as bandits and rebels.3 South-central Anatolia during the Turkish era under the local rulers of the Karamanids (c.1256– 1483) offers a striking case of historico-geographical continuity with the Roman period.4 The Karamanids and their Türkmen supporters defiantly resisted Ottoman hegemony and rule throughout the fifteenth century5 Indeed, of the many Anatolian Turkish polities absorbed by the Ottoman Empire, the Karamanids proved to be the most recalcitrant. Even after the conquest of the principality in the 1480s, the Ottomans continued to face difficulties in maintaining control over this region endemic with political unrest.6 Military campaigns against the Karamanids continued until the death of the last Karamanid prince in 1483–4. Pacification of the region, however, did not immediately follow the conquest of the Karamanid dynasty. The former supporters of the Karamanids, primarily the Turgut and Varsak tribesmen, allied themselves with the Mamluks during the Ottoman-Mamluk border war of the late 1480s, leading to continued political instability in the mountainous interior of İçel (p.309) and reprisals on the part of the Ottomans. The last bid for power on the part of a Karamanid descendant was witnessed with the rebellion in İçel led by a certain Mustafa in the winter of 1500–1. Having lived in exile in Iran all his life, Mustafa claimed to be a grandson of Karamanoğlu İbrahim II. His attempt to revive the Karamanid throne with the support of commanders from the Varsak and the Turgut Türkmen tribes while Bayezid II was distracted in the west with the conquest of the Morea proved unsuccessful.7 During Bayezid II’s reign and into that of Selim I, the Turgut participated in the vast Safavid network of agents involved in intelligence and propaganda operations throughout Ottoman-controlled Anatolia.8 Geographical factors lay behind much of the difficulty faced by distant imperial powers in the pre-modern period in ruling the region. South-central Anatolia stretching to the southern Mediterranean coastline opposite Cyprus consisted of a rugged mountainous interior core flanked by northern highlands comprising river basin plains and open steppeland (Figure 16.1). In his study of the Isaurians of the Taurus mountains under Roman rule, Brent D. Shaw noted the importance of the dynamic between the plain and the mountains, and its relation with domination and autonomy, describing it as ‘an oscillating play between physically dominant mountain ranges, and politically and culturally prevalent plains’.9 Control of the plains and steppes lying to the north of the central Taurus mountains continued to play a key role in the ability of subsequent imperial powers such as the Ottomans in establishing domination over the mountainous interior of south-central Anatolia. Taking into consideration the dynamic between the vulnerable plains and the relatively inaccessible interior core of mountains, this chapter examines the initial Ottoman campaign of conquest of the Karamanid principality with the assault on Gevele, a mountain fortress lying 11 kilometres to the west of Konya,10 and the seizure of Konya in 1468.11 The establishment of administrative and military control of the fertile agricultural plains as well as over the rich steppe grasslands of the region’s northern highlands by the Ottomans on the eve of their conquest of the Karamanid principality is likewise examined in order to reveal an important factor in the conquest of Karaman: Ottoman political involvement in Karamanid internal affairs before the conquest. By interfering in Karaman succession politics, the Ottomans laid the political groundwork by which they were able to establish an administrative and military presence in the hinterland of the more vulnerable stretches of Karamanid territory prior to the Ottoman military conquest. Thus in addition to cannon-fire and gunpowder assault weapons, the

Page 2 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 presence of Ottoman administration in nearby areas facilitated the conquest of Gevele and Konya. This development may also have precipitated the 1467 rebellion of the Karamanid prince Pir Ahmed, officially a vassal of the Ottomans. Perhaps in reaction to the loss of the vast economic and military resources provided by the hinterland of Akşehir, Beyşehir and Konya as the Ottomans ‘peacefully’ wrested control of the region from the Karamanids beginning in 1466, Pir Ahmed rebelled against his Ottoman supporters in 1467, thus provoking Mehmed II to conquer and destroy the Karamanid dynastic house the following year and incorporate its territory directly within the borders of his growing empire. (p.310)

(p.311) Geography of the Karamanid Frontier The territory under the control of the Karamanid dynasty was commonly referred to as the region of Karaman, which, as indicated above, roughly corresponds to the classical geographical designations of Lycaonia, Isauria and Cilicia Trachea, or Rough Cilicia. At the height of its power in the late fourteenth century, the Karamanid principality ruled (p.312) over the former core Seljuk centres of Konya, Aksaray and Ereğli, as well as Niğde and Kayseri in Figure 16.1. Relief map of Anatolia, showing Cappadocia, with their traditional political the principal places mentioned in the text. centre at Larende (today’s town of Karaman) lying close to the northern slopes of the Taurus mountain range 40 kilometres south-east of Konya. Karamanid territory was traditionally referred to as the Diyar-i Yunan (‘the Greek lands’) in the sources.12 The term Karaman came to be used following the Ottoman reorganisation of the territory as an Ottoman vilayet. The Ottomans thus transformed the core Karamanid lands into the Ottoman vilayet of Karaman, subdividing it into eight sancaks, with Konya at the centre of Ottoman administration, as described in several tahrir surveys. The other seven sancaks were centred around the most important cities of the region: Akşehir, Aksaray, Niğde, Beyşehir, Kayseri and Kırşehir, in addition to the sancak of İçel centered at Silifke/ Seleucia. İçel remained a Karamanid sancak, or sub-province, until the conquest of Cyprus in 1571, when it was attached to the latter province.13 The Karamanid principality constituted a contentious frontier lying to the southeast periphery of the expanding Ottoman state throughout much of the fifteenth century. Complicating the Ottoman-Karamanid frontier, however, was the existence of an additional internal frontier within the Karamanid realm along the Taurus mountain range. Thus while the external Karamanid frontier with the Ottomans consisted of a vaguely defined stretch of plateau, beginning in the vicinity of Akşehir and Ilgın to the west and extending south-east to Beyşehir and Seydişehir (Figure 16.2), the internal frontier was the result of a natural geographical mountainous barrier. The spurs of the western Taurus mountain range, sweeping in a semicircle to the west and extending southwards to the Mediterranean coast, divided the fifteenth-century Karamanid principality into two different spheres, with the first consisting of a steppe basin lying above the mountains and stretching into the Anatolian plateau, and the second characterised by rough

Page 3 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 mountainous terrain extending from the main spurs of the Taurus and reaching as far south as the Mediterranean coast.14 The rise and fall of the Karamanid principality was shaped by the dynamics of these vastly different frontiers: the external steppe basin, economically and culturally centred at Konya and vulnerable to invasion because of the lack of geographical barriers; and the internal frontier of the mountainous regions of Isauria and Rough Cilicia, the penetration of which posed difficulties for invading armies. (p.313) (p.314) The principal core of Karamanid power was the town of Larende,15 which straddled the frontier transition zone between mountains and plains, and controlled the mountain passes through Rough Cilicia as well as the Roman road connecting Konya to the eastern end of the trade route passing through Cilicia en route to Syria. Despite its peripheral location in Anatolia, particularly in relation to an imperial power ruling from the western end of the peninsula, Larende was located in a strategic transition zone between the Anatolian plateau and the central Taurus mountain spurs. Larende and its hinterland formed the principal corridor of movement from western Anatolia to Syria via Cilicia as well as to Cyprus. Thus, as a peripheral area in interaction with multiple cores,16 the Karamanid principality presented a Figure 16.2. Konya and the surrounding region. potentially serious hindrance to further eastward extension of Ottoman hegemony. It was this contact with other cores that made control of Karaman imperative for the success of future Ottoman imperial ventures over neighbouring powers. Indeed, it was in the context of consolidating his control over Anatolia and the Anatolian-Cilician border, a buffer zone with Mamluk Syria, that the Ottoman Sultan Mehmed II embarked on the conquest of the Karamanid principality as a necessary step in securing the potentially destabilising frontier.

Ottoman-Karamanid Relations in the Fifteenth Century and the Karamanid Succession Struggle, 1464–8 The Karamanids spent most of the fifteenth century in conflict with the Ottomans. While recognising Ottoman sovereignty for much of this period in a vassal-like relationship, the Karamanid rulers nevertheless sought opportunities to encroach on territories claimed by the Ottomans, as well as to form alliances with Ottoman enemies in order to undermine Ottoman expansion in Anatolia. Timur’s defeat of Bayezid I and his conquests in Anatolia at the beginning of the fifteenth century set back Ottoman territorial ambitions in Anatolia for decades. Timur’s policy of re-establishing Turkish principalities as vassal states recognising his sovereignty in effect challenged Ottoman political claims in Anatolia. It also gave new life to various Turkish states subjugated (p.315) by the Ottoman conqueror Bayezid I. Thus the Karamanid dynasty

Page 4 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 was able to reconstitute its power when the Karamanid princes were freed from Ottoman captivity following Timur’s army’s seizure of the Ottoman capital at Bursa in August 1402. Seated on his throne by Timur, Karamanoğlu (‘the Karamanid’) Mehmed II ruled over his patrimony including Konya, and Aksaray, as well as the more distant Antalya and Akşehir, recognising Timurid sovereignty17 His brother Ali ruled in Niğde, subordinate to him.18 After independent rule for fifteen years, Karamanoglu Mehmed Bey II was forced to accept Ottoman suzerainty in 1417 following defeat in battle by Sultan Mehmed I. Soon afterwards, in 1418–19, Karamanoğlu Mehmed II was ousted from the throne and replaced by his brother Ali with Mamluk support.19 In an effort to counterbalance Mamluk interference in Karamanid dynastic politics, and to seize the Karamanid throne from his uncle Ali Bey, İbrahim II, the son of Karamanoğlu Mehmed Bey II, sought Ottoman assistance. Following a succession struggle with his brother, Karamanoğlu İbrahim Bey seized Konya and the Karamanid throne.20 By diplomacy (p.316) and statesmanship, the Karamanid prince İbrahim II (1423–64) managed to hold his principality together over the next forty years despite periodic Ottoman encroachments. Thus, while recognising Ottoman sovereignty, the Karamanids under İbrahim Bey nevertheless ruled over their ancestral lands with little interference from the Ottomans. İbrahim Bey’s primary duty as a vassal of the Ottomans was the contribution of troops for campaigns, as he dutifully provided for Mehmed II during the conquest of Trebizond in 1461. İbrahim Bey’s death in 1464 and the subsequent succession struggle over the Karamanid throne among his sons disrupted the equilibrium between the two powers.

Ottoman Interference in Karamanid Succession Politics Determining the succession to the Karamanid throne in 1464 was key for the Ottomans in their attempt to control the region of Karaman. The Ottomans sought to enthrone a Karamanid prince who would seek their backing and accept their hegemony. The natural choice for Ottoman support was Pir Ahmed, who, as İbrahim Bey’s son from an Ottoman princess, was assumed to have vested interests with the Ottomans.21 The succession struggle over the Karamanid throne in 1464 introduced another regional actor in the struggle over dominance of south-central Anatolia and the Anatolian-Syrian borderlands. The Akkoyunlu Türkmen leader Uzun Hasan,22 based in Tabriz, immediately countered Ottoman attempts to control Karamanid internal succession politics by supporting the opposing candidate to the throne, İshak Bey, rival to his brother Pir Ahmed. Indeed, Akkoyunlu support was the primary means by which the Karamanid princes, completely dispossessed of their core hereditary territories by 1473, were able to continue to oppose Ottoman rule in their land. Uzun Hasan’s continued interference in Karaman, as well as his support of the Mamluk sultan over the Ottomans, likewise led to military clashes with Mehmed II, culminating in the Akkoyunlu defeat in 1473 at the Battle of Başkent/Otlukbeli in eastern Anatolia. The eldest son of the Karamanid ruler İbrahim II (1426–64), İshak Bey was born of a slave (cariye). Ottoman sources consider the lowly birth of İshak Bey as the underlying reason his father chose him as successor to the Karamanid throne: having no (p.317) blood ties to the Ottomans like his younger brothers born of an Ottoman princess, he seemed less likely to be influenced by Ottoman interests or bow to Ottoman domination.23 The younger brothers, however, did not react favourably to İshak Bey’s designation as heir apparent. According to Ottoman narrative sources, Karamanoğlu İbrahim Bey feared for the safety of İshak Bey because of the hostility he faced from his other brothers. His fears proved justified, for when the other sons, with the exception of the two youngest who fled to their Ottoman relatives,24 learned of their father’s attempt to deprive them of their patrimony, they marched on Konya, stormed the

Page 5 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 city, seized their father, and proceeded to take him to the nearby castle of Gevele for detention. The Karamanid bey İbrahim II, however, appears to have passed away just as he was being carried off to the dungeons.25 Either having escaped his brothers, or, perhaps, installed earlier there in anticipation of conflict, İshak reigned from the safety of the fortress of Silifke, while his half-brother Pir Ahmed was installed at Konya with Ottoman approval.26 İshak Bey, the Karamanid prince in exile at Silifke, sought and received the support of Uzun Hasan. He likewise sought the support of the Mamluk Sultan al-Zahir Khushqadam (1461–7), offering recognition of sovereignty of the Mamluks over that of the Ottomans.27 In September 1464 Uzun Hasan sent Akkoyunlu forces into Karamanid territory in order to restore İshak Bey to the Karamanid throne. The pretext was military aid in response to a raid into Karaman by the Dulkadirid Türkmen.28 Posing as the protector of Karamanid territories, Uzun Hasan’s Türkmen forces passed through westward to Konya, Beyşehir and Akşehir, and back east through Develi Karahisar and Kayseri, expelling Pir Ahmed as well as İshak Bey’s other half-brothers, the sons of the Ottoman princess.29 For a brief period, İshak Bey reigned from Konya, where, upon his entrance, he had the Friday sermon read in the name of the Mamluk sultan. John E. Woods views this act as an outward expression of homage to the Mamluk sultan, echoing the Karamanids’ patron Uzun Hasan’s homage to Khushqadam.30 In addition to gaining support from the Mamluks, Uzun Hasan (p.318) sought alliances with western powers, in particular the Venetians, to whom he sent an ambassador while Venice was in the midst of a war with the Ottomans.31 In the midst of this conflict, the Akkoyunlu-supported candidate İshak Bey, now in possession of the Karamanid throne, and reigning from Konya, attempted to negotiate peace with the Ottomans, offering Akşehir and Beyşehir as an inducement to the sultan. Mehmed II rejected the proposal, pointing out that the Ottomans had these territories by right thanks to previous agreements with his father İbrahim II.32 Determined to control the Karamanid throne through his own candidate rather than make peace with that of his Akkoyunlu enemy, the Ottoman sultan sent an army under the command of the sancakbeyi of Antalya, Köse Hamza Bey, together with the sancakbeyi of Karahisar, to reinstall Pir Ahmed in Konya.33 Faced with invading Ottoman forces, İshak Bey fled Konya for the mountains of Rough Cilicia, where he was defeated in battle by his brother at Ermenek. Defeated and dispossessed of his patrimony, İshak Bey fled east and took refuge at the Akkoyunlu court.34 Reinstalled on the throne by Ottoman armies, Pir Ahmed once again reigned over the Karamanid realm in 1465 as a vassal of Mehmed II, as indicated by an inscription at Kayseri dated 869/1466.35 The Karamanid prince İshak Bey, Pir Ahmed’s brother and rival, died in September 146636 while in the retinue of Uzun Hasan, with all his treasure passing into the hands of the Akkoyunlu leader. The Karamanid legacy was now left uncontested to his half-brother Pir Ahmed. The pretext for the Ottoman conquest was the insubordination of the Karamanid prince Pir Ahmed. Despite his Ottoman blood, and the vows of loyalty taken while visiting Istanbul,37 Pir Ahmed proved to be a less than reliable vassal. Not only did his communications with the Venetians put his loyalty to the Ottomans under suspicion,38 but he also failed to live up to his military obligations. The particular pretext for (p.319) the Ottoman invasion of the Karamanid principality was Pir Ahmed’s refusal in 1467 to join a proposed Ottoman campaign against the Mamluks,39 when Mehmed II hoped to take advantage of the tumult in the Mamluk realm after a series of usurpations in late 1467 and early 1468, ending with Qaitbay’s rise to the throne.

Page 6 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 Karamanid non-cooperation, however, prevented the campaign. Explaining the Karamanid conquest as a prelude to the war against the Mamluks, Tursun Bey mentions that Pir Ahmed came to be at odds with Mehmed because, although he was ordered to join the campaign and provide the service of guiding the Ottoman troops through the countryside (asker-i zafer rehbere kulavuz düşe), he refused to appear. Thus, by breaking his terms of vassalage (nahz-i ahd idüp gelmedi) with the Ottomans, Pir Ahmed brought the sultan’s wrath upon him (padişahun ateş gazabı heycan buldı). Since Karamanid territories comprised the frontier between the Ottoman and Mamluk lands, Mehmed II decided to first seize Karaman, and drive Pir Ahmed out of power, thus initiating the introductory step in the conquest of Egypt and Syria: ve bunı mukaddime-i feth-i Misr u şam kıla.40

Gevele and Konya In 1468, following the Ottoman assault on Albania in 1466–6 and the destruction of the power of the Albanian lord Scanderbeg, strengthening the Ottoman position vis-à-vis the Venetians in the course of their ongoing war,41 Mehmed II turned his attention eastwards. However, rather than move against the Mamluks as planned, Mehmed II broke camp at Karahisar,42 where the Ottoman forces had gathered, and marched in the direction of Konya with the intention of removing from power his disloyal vassal, the Karamanid bey Pir Ahmed.43 Passing the town of Beyşehir and moving into the Konya hinterland, the Ottoman army came upon the fortress of Gevele, located above the town of Sille (Su-diremi). The fortress, which was possibly a Roman structure later (p.320) rebuilt by the Byzantines, was situated high on the summit of an inactive volcano, Takkeli Mountain, and overlooked the original Roman road coming westwards from Beyşehir towards Konya. While the fortress of Gevele no longer stands today, its traces have been identified from among the remains of eight towers as well as the rubble of dispersed cut stone. The published results of the investigation of the site by the local amateur historian, M. Ferid Uğur, provide us with basic information.44 Along the road leading to the fortress, near the foot of Takkeli Mountain, lies the village Sarayköyü, ‘the palace village’, perhaps associated with a summer palace, which Ferid Uğur claims existed on the top of the summit, although no trace of it can be found. Following the road moving north, one runs upon the remains of stables and cisterns. The road comes to an end at the foot of the mountain, thus requiring at least an hour to make the ascent by foot up to the remains of the fortress.45 A century earlier, William J. Hamilton, travelling through the region in 1836 in search of the Greek village ‘Zillieh’ (Sille),46 an important source of the reddish volcanic or trachytic rock used in construction in Konya, first came across the very same Sarayköyü, which he identified as being located ‘at the eastern foot of the conical hill of Kara Burca, which had all the appearance of being volcanic, with a ruined castle at the top…’.47 Unaware that this castle was Gevele, Hamilton describes the different layers of the fortress’s ruins: ‘On the summit a small part of the ruins near the entrance appear to be Cyclopian, consisting of large irregular blocks without cement. The rest of the walls were more modern, and extended all over the summit of the rock.’48 Hamilton observed that the part of the ruins of Cyclopean construction49 dates back to ancient times, possibly anywhere from the Hittite to the Hellenistic age. (p.321) Difficult to reach with the steep ascent up the mountain slope, the fortress of Gevele served as a detention centre as much as a defensive stronghold. During the Seljuk period, rebellious commanders and Seljuk family members were detained in Gevele in order to prevent them from stirring up trouble.50 Sultan Rukn al-Din Kılıç Arslan IV likewise used the fortress to imprison Bunsuz, the son of Karaman, the eponymous founder of the Karamanid dynasty, some Page 7 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 time in the 1260s.51 In a similar vein, it served as a prison for İbrahim Bey II in 1464 just before his death when his Ottoman-backed sons seized him in order to prevent the succession of İshak Bey. Given the inaccessible nature of the fortress’s location situated at the top of a steep mountain summit, one wonders how the Ottomans managed to transport and array artillery around the fortification in order to bombard it. The Ottomans apparently were to repeat this feat in their siege of other Karamanid fortresses in Rough Cilicia, with the best example being Gedik Ahmed Pasha’s assault a few years later on Mennan, located approximately 16 kilometres from Ermenek on a particularly high precipice. Although the stronghold of Gevele was considered impregnable by the standards of the time, shortly after beginning the artillery assault on the fortress, the castle was surrendered by its defenders.52 Gevele had been in a position to defend the city of Konya; indeed of all the cities under Karamanid rule, Konya, together with its hinterland of large tracts of cultivated land, was not only the most populous, but also the most vulnerable to attack thanks to its location on a flat, open plain.53 With the surrender of Gevele, the conquest of Konya was assured. The fortifications of Konya, including the outer walls as well as the inner citadel, had been either destroyed or left in great disrepair some time in the Karamanid period, possibly during the reign of İbrahim II.54 After Konya, abandoned by the Karamanids and bereft of fortifications, was secured, orders were given to raze Gevele to the ground. The question remains why the Ottomans had no need for this fortress. First of all, with the annexation of the (p.322) northern region of Karaman to the rest of Anatolia under Ottoman rule, Gevele lost its frontier defensive function along the old Roman road55 leading from the western cities of Akşehir and Beyşehir to Konya. Furthermore, its location on a mountain summit made it impractical as a garrison fortress. Thus one may presume that the Ottoman rationale for razing the fortress was to prevent it from being used by rebellious elements before the pacification of the rest of Karaman had been completed. Immediately after razing Gevele, Mehmed II ordered the rebuilding of the inner citadel of Konya as well as strengthening the outer fortifications of the city56 The Ottomans did not hesitate to advertise their success at Gevele. Soon afterwards the Ottoman chancery sent announcements to European powers describing their victory over the hasanat u rasanat (impregnable and solid) fortress of Gevele.57 Such official correspondence may have aimed at creating an image of Ottoman invincibility as well as sought to warn others of the superiority of Ottoman artillery, weaponry and fighting techniques. Although the fortress no longer stands, its magnitude may be imagined through descriptions provided by contemporary and near-contemporary Ottoman authors. In his history Tevarih-i Al-i Osman, İbn Kemal extols the fortress’s impregnable walls (sur-i meni‘), soaring towers reaching the Pleiades in the heavens (burc-i refi‘i evc-i Süreyya irmişti), and firm barricades reminiscent of Alexander’s Rampart (sedd-i İskender). İbn Kemal’s metaphoric imagery does more than testify to the strong defences of the fortress; more importantly, it celebrates the might of the victorious Ottoman forces which seized the magnificent fortress.58 Likewise, the contemporary Ottoman poet Hamidi reveals how the Ottoman army seized the stronghold of Gevele so swiftly: witnessing the death of some of its defenders by the impact of cannon balls (top-i firangi) bombarding the stronghold, ‘the people, out of fear, surrendered the fortress.’59 Hamidi’s statement illustrates the psychological dimension of cannon warfare in this period: fear imparted by the terrific force of cannonfire played a great role in its success. Realising the futility of opposing armies employing cannons (p.323) in siege warfare, even against fortresses considered impregnable, besieged populations demanded that their defenders capitulate rather than resist.60 Hamidi’s

Page 8 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 observation resonates with accounts of similar capitulations in the narration of the Ottoman conquest of Karaman in contemporary Ottoman historical works.

The Importance of the Hinterland Ottoman propaganda in the form of correspondence and verse compositions regarding the conquest of Gevele obscures an important factor in the initial Ottoman campaign of conquest of the Karamanid principality. This was the establishment of Ottoman administration of the northern and western plains and plateau between Akşehir and Beyşehir as far as the surrounding hinterland of Konya, and the implementation of the timar system in the region before the campaign. The price that Pir Ahmed had paid for Ottoman support in ousting his brother İshak Bey from the Karamanid throne was the loss of control of this hinterland. After marching into Konya in order to put Pir Ahmed back on the throne, the Ottomans did not vacate the region. Rather, they proceeded to put the entire sweep of land extending from the western hinterland of Konya to Akşehir under direct Ottoman administration, some time around in 871/1466, as a tax register compiled around that time indicates.61 They also made their rule felt in the steppe immediately north of Konya, as well as that in between Aksaray and Konya. The land register produced at this time thus covers not only the western regions of Akşehir, Beyşehir and Ilgın, but also Zengicek and İnsuyu,62 situated to the immediate north of Konya. In effect, the Ottomans succeeded in putting a stranglehold on Konya without ever firing a cannon. Dependent on Ottoman military aid, Pir Ahmed was in no position to object. The subsequent conquest of Konya must have been accomplished with relative ease, with only the fortress of Gevele standing in the way.

The Displacement of the Turgutoğlu from the Hinterland The Ottoman appropriation of the Konya hinterland involved granting timar properties in the vicinity to cavalry commanders, or sipahis, from elsewhere. Beldiceanu-Steinherr’s (p.324) careful study of the land register (BOA, Maliyyeden Müdevver 241) demonstrates how certain timar villages were taken away from Karamanid Turgutoğlu commanders by imperial command.63 For instance, five villages in the subdistricts of İnsuyu64 and Zengicek65 to the north and north-east of Konya, as well as villages attached to the sub-district of Göçü66 in the prosperous river valley in the northern river basin of Beyşehir, were taken from the Karamanid commander Turgutoğlu Hüseyin and given to Ottoman commanders. The village of Horta in İnsuyu became the timar of the Albanian Hamza, and the village of Kirmir/Giymir67 attached to the subdistrict Zengicek, was granted to İsa of İstrumca (Strumitza) on 2 September 1466.68 It is interesting that the individuals appointed to timars by the Ottomans in the region were of Balkan or Rumeli origins. Mehmet Akif Erdoğru’s study of Ottoman administration of the Beyşehir sancak, or district, of Karaman confirms this trend. According to Erdoğru, the first timar holders in Beyşehir during the reign of Mehmed II were sipahis of devşirme origins who had served the Ottomans in Rumeli.69 Incorporating the region into the Ottoman administrative system resulted in the displacement of the very group who provided the military backbone of the Karamanid forces and steadfast support for the Karamanid dynasty—the Turgutoğlu and their horsemen.70 The Turgutoğlu supplied both commanders and troops for the Karamanids.71 It is, however, difficult to trace how the two groups originally became associated. Scattered references to the Turgutoğlu appear in the Ottoman chronicles as well as in the semi-historical text Karamanoğlu Tarihi (The History of the (p.325) Karamanids) by Şikari, an author of unknown identity.72 Turgut, the eponymous ancestor of the Turgutoğlu, appears in the entourage of Nureddin, the mid-thirteenth-century Karamanid ancestor in Şikari’s account, which consists of early fourteenth-century Page 9 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 developments projected back a century earlier.73 One may guess that the clan of the Turgutoğlu subordinated themselves to the Karamanids upon the breakdown of Mongol rule on the Anatolian central plateau in the 1330s. Their rise to prominence under the Karamanids may have occurred as a result of their unwavering support to the Karamanid ruler Alaeddin Bey (ruled c. 1358–98) against his main rivals the Eretnids.74 Akin to an elite cavalry force,75 the Turgutoğlu were also distinguished by being the preferred marriage partners of the female members of the Karamanid dynasty. In fact the name Turgut may derive from the Turco-Mongolian term törgüt, which, according to B. Y. Vladimirtsov, means ‘relative through one’s spouse’.76 During the period of Karamanid administration, the Turgutoğlu clans appear to have had great influence as timar holders in the region directly north of Konya on the steppe in between Akşehir Gölü and the mountain range of the Emir Dagları in the west, and the Salt Lake (Tuz Gölü) in the east. (p.326) The Ottomans chose Turgut as a name for a kaza, or subdistrict, with its centre lying immediately north of Ilgın on the north-east slope of the Gölçük Mountains, which was an important stopping point on the northern road from Akşehir to Konya, testimony to the clans’ lasting importance. Furthermore, a great many endowments and public monuments associated with the Turgutoğlu in the urban centre of Konya and its hinterland attest their enormous wealth as well as economic and political power during the Karamanid period.77 Turgutoğlu patronage of public works and monumental construction appears to have even surpassed that of the Karamanids in Konya and its northern and western districts. Whereas the Karamanids favoured Larende as the home for their türbes and other monumental construction, the Turgutoğlu appear to have preferred Konya. For instance, the Karamanid commander-in-chief (al-amir al-kabir) Turgutoğlu Pir Hüseyin, son of Emirşah Bey, constructed a türbe (mausoleum) for himself in Konya in 1427 or 1432.78 Turgutoğlu Pir Hüseyin Bey’s sway extended to the north-west of Konya, as seen in his construction activity in the region. In 1423 Turgutoğlu Pir Hüseyin Bey constructed a mosque at Ilgın. He likewise built an early fifteenth-century mosque at Sarayönü.79 The elite of the Turgutoğlu clan had a concentration of vakıf in the vicinity of the subdistrict of Zengicek in the immediate hinterland of Konya. Turgutoğlu Pir Hüseyin’s son Ahmed Bey tied the income of several endowed properties in the vicinity of Zengicek to the upkeep of his father’s mausoleum in Konya. Ahmed Bey’s son Turgutoğlu Sırrı Bey continued as the mütevelli (overseer) of the pious endowment. Other family members likewise contributed income to the mausoleum (türbe) by attaching to it other endowed property80 According to an endowment deed drawn up in the name of a certain Sultan Hatun, the daughter of the Turgutoğlu Yusuf Şah Bey, half of the income of the nearby village of Zıvarık (Altınekin) lying almost immediately to the north of the road passing by the subdistrict centre of Zengicek, was set aside for her pious endowment of vakıf. Indeed, it has been speculated that the nearby caravanserai of Zıvarık Han was constructed by her.81 Another branch of the Turgutoğlu clan likewise must have asserted influence in the southwestern point of the Konya hinterland at the town of Seydişehir. Rüstem Bey (d. 1439), the son of the Turgutoğlu commander Halil Bey, built a family mausoleum in the vicinity of the Seyyid Harun Veli’s mosque complex in Seydişehir. The mausoleum contains the graves of several family members, including that of Rüstem Bey’s (p.327) son Ali Bey, who died the same year as his father, Rüstem Bey’s daughter Dürrühand Hatun (d. 1428/9) as well as a certain Sultan

Page 10 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 Hatun (d. 1422), daughter of the Turgutoğlu Emirşah, and the sister of the above Turgutoğlu Pir Hüseyin Bey82 The Turgutoğlu clan not only played an important military role for the Karamanids by providing top commanders and son-in-laws for the Karamanids. Its elite likewise exerted much influence in Konya as well as in its south-western, north-western and northern hinterland, as seen in extensive monumental building and vakıf properties. Unfortunately we have no narrative sources on the reaction of the Turgutoğlu as the Ottomans moved into their traditional seats of power and removed them from their timars, although their role as Safavid agents suggests that their opposition to the Ottoman dynasty remained unabated into the sixteenth century83 Nevertheless one can assume that the removal of the Turgutoğlu elite from the Konya hinterland must have adversely affected their ability to muster military forces. Thus it is not surprising that the Karamanids and their Turgutoğlu generals fled upon the approach of the Ottoman army upon Konya rather than fight on the open plain in 1468. As the Ottoman army proceeded to secure the vicinity around Konya, probably moving south towards Larende, Pir Ahmed escaped to the mountains of Karataş (başın alup Kara-taş sarpına sıgındı)84 in the south-eastern part of Rough Cilicia. This was a remote and unpopulated region situated in the mountainous interior close to the Mediterranean coast north-east of Silifke and south-west of the river the Lamas Çayı.85 According to Aşık Paşazade, however, Pir Ahmed first fled to nearby Larende, 40 kilometres south-east of Konya, with Mahmud Pasha, the Ottoman grand vizier, in pursuit. Although Mahmud Pasha was able to catch up with Pir Ahmed and engage him in battle, the Karamanid bey escaped and disappeared in the rugged mountainous countryside.86 Although he failed to seize Pir Ahmed, Mahmud Pasha moved north-east to the vicinity of the Bulgar Mountain near the Cilician Gates. He had been ordered to fight the Turgutoğlu who, according to military intelligence, had fled in that direction. After procuring quite a few captives from among the Turgutoğlu, Mahmud Pasha returned to Konya to (p.328) organise the deportations of the urban population in the region, while şehzade Mustafa took up his post as Ottoman governor of Konya. Thus concluded the first set of Ottoman operations against the Karamanid dynasty.87

Conclusion It has been pointed out that conquest in the pre-modern period occurred not according to ‘contiguous blocks of territory’, but rather ‘in terms of routes defended and fortresses garrisoned’, as Palmira Brummett puts it.88 According to Brummett, Ottoman conquest took the pattern of occupying towns, setting up of customs posts and the appointing of central authorities.89 While control of routes and the garrisoning of fortresses was essential for establishing military control, Ottoman methods of conquest were also accompanied by other military strategies, as the case of Karaman indicates. Thus the role of non-military means of establishing political influence and domination in a territory before attempting to conquer it with force should not be overlooked; nor should the strategy of taking control of the hinterland before moving on against fortresses be forgotten. If we consider how the Ottomans approached the conquest of Karaman, it becomes clear that decades of Ottoman intervention in the succession and internal politics of the ruling dynasty had their desired effect. Finally, in 1464–6, as a result of their pivotal role in the outcome of the Karamanid succession struggle, the Ottomans were able to intensify their control of the region by appropriating the hinterland of Konya and implementing direct Ottoman administration, which appears to have involved dispensing lands as timar to their cavalry commanders. This must have been accompanied by the removal of the Turgutoğlu. With the Turgutoğlu forces no longer guarding the frontier with the

Page 11 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 Ottomans in the hinterland, both Gevele, only kilometres away from Ottoman-administered territories, and nearby Konya were left vulnerable to Ottoman assault. The Ottoman victory over the fortress of Gevele as depicted by Ottoman authors was a less remarkable feat than the Ottomans would have had their enemies believe. Once they overcame the topographical challenge of moving artillery up steep mountain slopes to reach the fortress on the summit, the Ottoman forces undertook what must have been a rather brief siege. Furthermore, there was no question of besieging Konya: lacking fortifications, it lay defenceless before the invading Ottoman armies. Perry Anderson’s comment that ‘the development of bronze-cast cannon made gunpowder (p.329) for the first time the decisive arm of warfare, rendering baronial castellar defences anachronistic’90 may be a fitting description for the fate of fortresses like Gevele. The razing of Gevele points to its obsolescence: it was no longer useful to fortify rural mountain summits any more, especially when the frontier no longer existed. The Ottomans preferred garrisoned urban fortifications, as the rebuilding of the citadel and walls of Konya on the orders of Mehmed II indicate.91 Notes: Proceedings of the British Academy 156, 307–329. © The British Academy 2009. (1) Francis Beaufort, Karamania, or A Brief Description of the South Coast of Asia Minor and the Remains of Antiquity with Plans, Views, &c. (London: R. Hunter, 1817), p. iii. (2) A rather amorphous geographical designation, Isauria comprised the mountainous region of Anatolia adjacent to Rough Cilicia (Cilicia Tracheia). In the second century AD, Isauria proper was joined with Cilicia Tracheia and the southern part of Lycaonia, as well as the eastern region of Pisidia around Lake Suğla into a new province, the boundaries of which have a striking resemblance to the core regions of the Turkish principality of Karaman of the fourteenth and fifteenth centuries AD. As a result, the designation Isauria popularly referred to the entireregion. See A. S. Hall, ‘Notes and inscriptions from eastern Pisidia’, Anatolian Studies, 18 (1968), 60; ‘Isauria’, The Oxford Dictionary of Byzantium (Oxford: Oxford University Press, 1991), 2: 1014; Stephen Hill, The Early Byzantine Churches of Cilicia and Isauria (Aldershot: Variorum, 1996), 4; Sviatoslav Dmitriev, ‘Observations on the historical geography of Roman Lycaonia’, Greek, Roman and Byzantine Studies, 41 (2000), 355. (3) Brent D. Shaw, ‘Bandits in the Roman Empire’, Past & Present, 105 (1984), 52; idem, ‘Bandit highlands and lowland peace: the mountains of Isauria-Cilicia’, Journal of the Economic and Social History of the Orient, 33 (1990), 199–233; 33/3 (1990), 237–70; Noel Lensky, ‘Assimilation and revolt in the territory of Isauria: From the 1st century to the 6th century AD’, Journal of the Economic and Social History of the Orient, 42 (1999), 413–65; idem, ‘Basil and the Isaurian uprising of A.D. 375’, Phoenix, 53 (1999), 308–29. (4) An additional continuity of the region of Ottoman-dominated Karaman with the Roman period is the vulnerability of the plains north of the mountains to the attacks and raids of the tribes of the mountainous interior. As Stephen Mitchell points out, when the Roman emperor Augustus absorbed the kingdom of Amyntas in 25 BC, he found his new territory under attack by the tribes of the Isaurian and Pisidian Taurus who had for decades waged warfare against the king Amytas in the plains. In particular the hostile tribe of the Homonadenses (or Homonadeis as Strabo referred to them) continued to pose a threat to the security of the region and the safety of the trade and communication routes passing to Syria. The Roman solution to the problem of controlling the region of Isauria was the establishment of garrisoned colonies in the areas Page 12 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 dominated by hostile tribes. See Stephen Mitchell, ‘Legio VII and the garrison of Augustan Galatia’, The Classical Quarterly, n.s. 26 (1976), 298; Sir Ronald Syme, ‘Isauria in Pliny’, Anatolian Studies, 36 (1986), 159. (5) Unlike the Karamanid dynasty, whose power bases were established in the mountainous interior of Rough Cilicia, the Seljuks ruling from the plains of Konya during the thirteenth century had little success in controlling Rough Cilicia. See Sara Nur Yıldız, ‘Reconceptualizing the Seljuk-Cilician frontier: Armenians, Latins and Turks in conflict and alliance during the early thirteenth century’, in Florin Curta (ed.), Borders, Barriers and Ethnogenesis: Frontiers in Late Antiquity and the Middle Ages (Turnhout: Brepols, 2005), 91–120. (6) See İsmail Hakkı Uzunçarşılı, Anadolu Beylikleri ve Akkoyunlu, Karakoyunlu Devletleri (Ankara: TTK, 1937), 35; M. C. şihabeddin Tekindağ, ‘Karamanlılar’, İA, 6: 327; Faruk Sümer, ‘Karaman-oghulları’, EI2, 4: 624. (7) See Müneccimbaşı, Saha’if al-Ahbar fi Vaka’i al-A’sar, ed. İ. Erünsal (Istanbul: Tercüman, 1974), 412–14; Sümer, ‘Karaman-oghulları’, 624; Selahattin Tansel, Sultan II. Bayezit’in Siyasi Hayatı (Istanbul: Milli Eğitim Basımevi, 1966), 123. (8) A Turkish letter from Shah Isma‘il to Musa Turgutoğlu, dated May 1512, provides evidence of a well-organised cadre of Safavid agents, such as Ahmad Ağa Karamanlu, presumably a Karaman emigré to Safavid Iran, in cooperation with the Turgutoğlu commanders in Karaman: M. C. şahabeddin Tekindağ, ‘Yeni Kaynak ve Vesikaların Işıgı Altında Yavuz Sultan Selim’in İran Seferi’, Tarih Dergisi, 17/22 (1967), 52; Adel Allouche, The Origins and Development of the Ottoman Safavid Conflict, 906–962/1500–1555 (Berlin: Schwartz Verlag, 1983), Turkish trans. Ahmed Emin Dağ, Osmanlı-Safevi İlişkiler Kökenlerli ve Gelişimi (Istanbul, 2001), 183–5. (9) Shaw, ‘Bandit highlands and lowland peace’, 200. (10) İbrahim Hakkı Konyalı, Konya Tarihi (1964; reprint, Konya: Enes Kitap Sarayı and Ankara: Burak Matbaası, 1997), 121. The site is identified with Byzantine Caballa: William M. Ramsay, Historical Geography of Asia Minor (London: John Murray, 1890), 359; Franz Babinger, ‘Kavalla’, Der Islam, 29 (1950), 301. (11) There has been great confusion in the scholarly literature regarding the date of the conquest of Konya. In his study of the political history of the reign of Mehmed II which provides the most comprehensive account of these events, Selahettin Tansel neglects to give a date for the seizure of Konya. See Selahettin Tansel, Fatih Sultan Mehmed’in Siyasi ve Askeri Faaliyeti (Istanbul: Milli Eğitim Bakanlığı, 1971), 287–98. Besim Darkot’s date of 1465 is too early (Besim Darkot, ‘Konya’, İA, 6: 849). Much of Turkish scholarship follows İ. H. Uzunçarşılı’s erroneous date of 1466, based on the date AH 871 provided by Hoca Sadeddin’s Tacü’t-Tevarih. The date of AH 871 is not limited to 1466, but rather extends over the period approximately between 12 and 13 August 1466 and 1 and 2 August 1467. See İsmail Hakkı Uzunçarşılı, Osmanlı Tarihi (Ankara: TTK, 1943), 2: 90; idem, Anadolu Beylikleri, 32. On the other hand the date of AH 872 is cited by almost all the contemporary fifteenth-century Ottoman sources, in particular, the histories by Nişancı Karamani Mehmed Pasha, Tawarikh al-Salatin al-‘Uthmaniyya, trans. İbrahim Hakkı Konyalı as ‘Osmanlı Sultanları Tarihi’, in N. Atsız (ed.), Osmanlı Tarihleri (Istanbul: Türkiye Yayınevi, 1949), 1: 355–56; Tursun Bey, Tarih-i Ebü’l-Feth, ed. Mertol Tulum (Istanbul: Baha Matbaası, 1977), 146; Aşık Paşazade, Tevarih-i al-i osman, ed. Kemal Yavuz and M. A. Yekta Saraç as Osmanoğulları Tarihi (Istanbul: MAS Matbaacılık, 2003), 253, 520–1; as well as Neşri, Page 13 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 Kitab-i Cihan-nüma: Neşri Tarihi, ed. Faik Reşit Unat and Mehmed A. Köymen (Ankara: TTK, 1957), 2: 784–5. Since AH 872 approximately falls between the dates 2 and 3 August 1467 and 20 and 21 July 1468, and if one takes into account that Mehmed II was battling in Albania against the forces of Scanderbeg in the spring of 1467, and resident in Istanbul in the winter of 1467–8 (Colin Imber, The Ottoman Empire, 1300–1481 (Istanbul: Isis, 1990), 196–8), his campaign against Konya must have occurred during the spring of 1468, that is, towards the end of AH 872. Further, the initial plans to invade the Mamluk sultanate following the death of Khushqadam on 10 October 1467 (Imber, The Ottoman Empire, 198), a campaign that was diverted by the assault on Konya, it seems quite evident that the campaign took place in the spring of 1468. (12) A term frequently used by the author of a popular history of the Karamanid dynasty, Şikari, Tarih-i al-i Karaman, ed. M. Mesud Koman and M. Ferid Uğur, Şikari’nin Karamanogulları Tarihi (Konya: Yeni Kitap Basımevi, 1946), passim; an anonymous Ottoman chronicle entitled Zikr-i Mülk-i Al-i Osman ve Gazevat-i Işan specifically points out that Yunan vilayeti was the former name for the Karaman vilayet: Yunan vilayeti kim şimdi Karaman vilayetidür (MS Istanbul, Topkapı Library, Revan no. 700, fol. 7). (13) Suraiya Faroqhi, ‘Sixteenth-century periodic markets in various Anatolian sancaks: Içel, Hamid, Karahisar-i Sahib, Kütahya, Aydın, and Mentese’, Journal of the Economic and Social History of the Orient, 22 (1979), 36–7; Mehmet Akif Erdoğru, ‘Beyşehir Sancagı İcmal Defteri’, Belgeler, 13/17 (1988), 117. (14) Barrington Atlas of the Greek and Roman World, II: Map by Map Directory, ed. Richard J. A. Talbert (Princeton and Oxford: Princeton University Press, 2000), 1013. (15) While Konya had prospered as the capital of the Anatolian Seljuk empire of the thirteenth century, under the Karamanid rulers ‘it sank to a city of the second rank’, J. H. Mordtmann, ‘Ereğli’, EI2, 2: 705. The late Ilkhanid financial officer Mustawfi, writing in approximately 1340, noted the decline of Konya under the Karamanids, pointing out that in his day much of Konya was in ruins, although the area immediately below the citadel had a large population: G. Le Strange, The Lands of the Eastern Caliphate: Mesopotamia, Persia, and Central Asia from the Moslem conquest to the time of Timur (London: Frank Cass, 1905), 148. (16) The ability of peripheral areas to interact with multiple cores has been noted by theorists of ancient archaeology: P. Kohl, ‘The ancient economy, transferable technologies and the Bronze Age world system: a view from the northeastern frontier of the ancient Near East’, in M. Rowlands, M. T. Larsen and K. Kristiansen (eds.), Centre and Periphery in the Ancient World (Cambridge: Cambridge University Press, 1987), 13–24. (17) During his siege of Bursa, Timur’s grandson Mehmed Mirza had released from Ottoman captivity both Mehmed and ‘Bengi’ Ali, the two sons of the Karamanid ruler Alaeddin Bey, who had been defeated by Bayezid II in 1398. Mehmed Mirza brought the heirs to the Karamanid throne before Timur who, in addition to their ancestral lands, granted them the cities of Beypazarı, Kayseri, Kırşehir and Sivrihisar. Karamanoğlu Mehmed Bey II struck coins in the name of Timur in Konya, Larende, Kayseri and Eğridir. See Uzunçarşılı, Anadolu Beylikleri, 17; Abu al-Mahasin Ibn Taghri Birdi, History of Egypt, 1382–1469: Translated from the Arabic Annals of Abu l’Mahasin Ibn Taghri Birdi. Part II: 1399–1411, ed. and trans. William Popper (Berkeley: University of California Press, 1954), 62.

Page 14 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 (18) Ali’s rule in Niğde and subordination to the Ottoman ruler Mehmed I is confirmed by an inscription at the Ak Medrese in Niğde dating to 812/1409: I. H. Uzunçarşılı, ‘Nigde’de Karamanoğlu Ali Bey Vakfiyesi’, Vakıflar Dergisi, 2 (1942), 65. (19) Imber, The Ottoman Empire, 78, 87–8. (20) Ibn Kemal, Tevarih-i Al-i Osman, Book 7, ed. Şerafettin Turan (Ankara: TTK, 1957), 183. Karamanoğlu İbrahim Bey’s father was Mehmed Bey II, who, during an attack on the fortress of Antalya, died because of an unfortunate incident with a cannon which blew him to smithereens (top kolayına geldiğinleyin top dokunup, kazayla pare pare oldu). According to Neşri, İbrahim owed his succession to the Karamanid throne following his father’s death to Murad II who ‘granted him his sancak and girded him with the sword of sovereignty (kılıç kuşattı)’. Thus, according to the Ottoman chronicle tradition, it was through Ottoman support that İbrahim II was able to possess the throne in the face of rivalry from his two brothers İsa and Alaeddin as well as his uncle Bengi ‘Ali Bey, and owed his sovereignty to Murad II (Neşri, Kitab-ı Cihan-nüma, 2: 591). Nevertheless, this did not prevent İbrahim II from attempting to throw off Ottoman domination when the opportunity presented itself, and he made various alliances throughout his reign with both eastern and western enemies of the Ottomans. In an attempt to reimpose Karamanid recognition of Ottoman sovereignty, Murad II forced the Karamanid ruler to sign a peace treaty in 1444, in which he swore not to ally himself with enemies of the Ottomans nor attack Ottoman possessions: Yahya bin Mehmed el-Katib, Menahicü’l-Insa: The Earliest Ottoman Chancery Manual, ed. Şinasi Tekin (Roxbury, Mass.: Community Art Workshop, 1971), 31–2; Alaaddin Aköz, ‘Karamanoglu II. İbrahim Beyin Osmanlı Sultanı II. Murad’a Vermis Olduğu Ahidname’, Türk Arastırmaları Dergisi, 24/38 (2005), 72–92. Recognition of Ottoman sovereignty by the Karamanids had to be re-established following the death of Murad II and Mehmed II’s accession to the Ottoman sultanate in 1451. Mehmed II succeeded in imposing his personal sovereignty over Karamanoğlu İbrahim II by military force following a Karamanid invasion of Ottoman possessions in Anatolia. Only when he was confronted with Mehmed II’s greater military forces did İbrahim II agree to peace terms and have his troops withdraw: Anonim Osmanlı Kroniği (1299–1512), ed. Necdet Öztürk (Istanbul: Türk Dünyası Araştırmaları Vakfı, 2000), 85; Aşık Paşazade gives a more detailed account of these events: Aşık Paşazade, Osmanoğulları Tarihi, 216–17, 486–7. (21) İbn Kemal states that İbrahim Bey’s Ottoman wife was Sultan Hatun, daughter of Bayezid I and sister of Murat II, thus the maternal aunt of Mehmed II: İbn Kemal, Tevarih, 237. Neşri, on the other hand, claims that Sultan Hatun was the daughter of Mehmed I rather than Bayezid I: Neşri, Kitab-ı Cihan-nüma, 2: 770–1. (22) Uzun Hasan was the charismatic leader of the Akkoyunlu tribal confederacy who took power following a five- year revolt against his brother and kinsmen in 1452–7. The Akkoyunlus consisted of a tribal confederation of Oğuz Türkmen who ruled in eastern Anatolia, northern Mesopotamia and western Iran until the Safavid conquest in 1501. Under the leadership of Uzun Hasan in the 1450s and 1460s, the Akkoyunlu confederation rapidly expanded and came into conflict with the Ottomans: see John E. Woods, The Aqquyunlu: Clan, Confederation, Empire (Salt Lake City: University of Utah Press, 1999), 72, 83, 85 ff.; R. Quiring-Zoche, ‘Aq Qoyunlu’, EIr, 2: 163, 165. (23) Neşri, and İbn Kemal, whose information resembles to a certain extent but not completely that of Neşri, list the seven sons as İshak, Pir Ahmed, Kasım, Ali, Süleyman, Karaman and NureSufi. Şikari’s list of the seven sons varies considerably: İshak, Pir Ahmed, Kasım, Alaeddin, Halil, Page 15 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 Yakub and Küçük Mustafa: Neşri, Kitab-ı Cihan-nüma, 2: 771–2; İbn Kemal, Tevarih, 237; Şikari, Karamanoğulları Tarihi, 190–1. (24) These two youngest sons were Süleyman and Nure-Sufi: Neşri, Kitab-ı Cihan-nüma, 2: 772–3. (25) Ibid., 2: 772–3. Drawing on the later seventeenth-century Ottoman historical tradition of Cenabi and Müneccimbaşı, J. H. Kramers writes that İbrahim II died while besieged at the fortress of Gevele, presumably by his sons, thus providing a different twist to this story: J. H. Kramers, ‘Karaman-oghlu’, EI1, 4: 751. (26) During this time he minted coinage: Cüneyt Ölçer, Karaman Oğulları Beyliği Madeni Paraları (Istanbul: Yeni Basımevi, 1982), 107. (27) Imber, The Ottoman Empire, 192. (28) Ibid. (29) Woods, The Aqquyunlu, 93. (30) Ibid., 93, 95. (31) Imber, The Ottoman Empire, 193. Venice found herself in her first long war (1463–79) with the Ottomans, which erupted over control of the Morea. During the course of this war, Mehmed II captured the island of Negroponte (Euboea, Eğriboz), one of the most important Venetian commercial colonies: M. E. Mallet and J. R. Hale, The Military Organization of a Renaissance State: Venice c.1400 to 1617 (Cambridge: Cambridge University Press, 1984), 43–5. (32) Imber, The Ottoman Empire, 194. After Mehmed II seized Akşehir and set out for Konya in 1451 as a reprisal against Karamanoğlu İbrahim Bey II’s attack on Ottoman territory, İbrahim Bey immediately capitulated and agreed to hand Akşehir, Beyşehir and Seydişehir to Mehmed II as part of the conditions of the peace treaty. (33) Aşık Paşazade, Osmanoğulları Tarihi, 250, 520–1; Neşri, Kitab-ı Cihan-nüma, 2: 772–5. (34) Aşık Paşazade, Osmanoğulları Tarihi, 250, 520–1. İshak Bey appears to have been in Konya long enough to have struck coinage in his name from its mint: Ölçer, Karaman Oğulları Beyliği Madeni Paraları, 106. (35) Kramers, ‘Karaman-oghlu’, 751. (36) The date cited by Tekindağ is AH Muharram 870, or June 1466 based on Ibn Taghribirdi’s Hawadith: M. C. şehabettin Tekindağ, ‘Son Osmanlı-Karaman Münasebetleri Hakkında Arastırmalar’, Tarih Dergisi, 17–18 (1963), 52. (37) Aşık Paşazade, Osmanoğulları Tarihi, 251–2; 520–1. (38) Pir Ahmed’s diplomatic ventures are known from Venetian documents. For Pir Ahmed’s correspondence see Comte Pierre-Antoine-Noel-Bruno Daru’s Histoire de Venise (Brussels: Société typographique belge, 1838), 1: 259; Arduino Cremonesi, La sfida turca contro gli Asburgo e Venezio (Udine: Arti Grafiche Friulane, 1976), 118.

Page 16 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 (39) Tekindağ, ‘Son Osmanlı-Karaman Münasebetleri’, 326. Tursun Bey, Tarih, 56. İnalcık points out that tensions began to rise between Mehmed II and the Mamluks in 1466 during the reign of Khushqadam (1461–7) related to Mehmed’s II claim of being the superior power in the Muslim world: Halil İnalcık, ‘Mehmed II’, İA, 6: 514. For a study of the earliest signs of deterioration of Ottoman-Mamluk relations during the reign of Mehmed II, as early as 1453, over Mehmed’s attempt to establish his authority over the Orthodox community, see Ralph S. Hattox, ‘Mehmed the Conqueror, the Patriarch of Jerusalem, and Mamluk authority’, Studia Islamica, 90 (2000), 105–24. (40) Tursun Bey, Tarih, 145–6. (41) See Imber, The Ottoman Empire, 195–6. (42) Karahisar-ı Sahip or Afyonkarahisar (Afyon), to the west of Akşehir (Metin Tuncel, ‘Karahisar’, TDVİA, 416–17) was located on the main route from Istanbul passing through Bursa and Kütahya, and leading to Akşehir and Konya. Karahisar-ı Sahip was a major Ottoman military headquarters in Anatolia, fortified by a large citadel on the summit of a formidable mass of black volcanic rock. (43) İbn Kemal, Tevarih, 237. (44) Mehmet Ferid Uğur (1881–1942), an amateur historian and archaeologist, did much to advance the local efforts of writing the history of his hometown Konya and of the Karamanids: M. Mesut Koman, ‘Merhum Mehmet Ferit Bey Üstadımız’, Konya, 6/41 (1942), 54. (45) M. Ferid Uğur, ‘Takkeli Dağ’, Konya Mecmuası, 5 (1937), 300–1. (46) Hamilton describes the location of Zillieh, a Greek village in the mountains approximately two hours from Konya, in ‘a deep gorge in the trachytic hills.’ Zillieh, Hamilton points out, was the source of the red trachytic stone commonly used in the buildings, pavement and graveyards of Konya: William J. Hamilton, Researches in Asia Minor, Pontus and Armenia; with some account of their antiquities and geology (London: John Murray, 1842), 2: 208–9). Adnan Baysal points out that this source of trachyte dates to prehistoric times, going as far back as the Chalcolithic and Bronze Age. The famed Sille stone ranges from grey to pink and reddish tones. A large quarry is still in operation producing volcanic trachyte for commercial use: Adnan Baysal, ‘Ground stone report’, www.catalhoyuk.com/archive_reports/2004/ar04_28.html. (47) Hamilton, Researches in Asia Minor, 2: 208–9. (48) Ibid. (49) By ‘Cyclopian’, Hamilton presumably means the massive stones of irregular shape and size reminiscent of a primitive style of masonry characteristic of prehistoric construction. The term was commonly used by nineteenth- century classicists to refer to the Hellenistic and earlier periods; Hittite construction is likewise characterised by the use of Cyclopean blocks. See Charles Waldstein, ‘The earliest hellenic art and civilization and the Argive Heraeum’, American Journal of Archaeology, 4 (1900), 53 ff.; Ronald M. Burrows, ‘Pylos and Sphacteria’, Journal of Hellenic Studies, 16 (1896), 55 ff.; Jeanny Vorys Canby, ‘Hittite art’, The Biblical Archaeologist, 52 (1989), 109 ff. For the latest evaluation of Cyclopean masonry, consult John Paul Zielinski, ‘Cyclopean Architecture in Minoan Bronze Age Crete: A Study in the Social Organization of a

Page 17 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 Complex Society’, (Unpublished Ph.D. dissertation, State University of New York at Buffalo, 1998). (50) Ibn Bibi, El-avamirü’l-‘Ala‘iyye fil-Umuri’l-Ala‘iyye, ed. Adnan Sadık Erzi (Ankara: TTK, 1956), 88. For instance, Ibn Bibi refers to how Sultan Ghiyath al-Din Kaykhusraw I in 1205 detained his nephew Kılıç Arslan III ibn Sulaymanshah, a potential rival for the throne, at Gevele fortress until consolidating his power over the realm. According to later sources the young Kılıç Arslan III never left the fortress alive. See Osman Turan, Selçuklular Zamanında Türkiye (Istanbul: Istanbul Matbaası, 1971), 274. (51) Ibn Bibi, El-avamir, 687. (52) Konya appears to have been vulnerable to armies and conquerors throughout its history. For instance, in 1069 a Seljuk army sacked the city after having routed the Byzantine army under Romanus Diogenes: J. G. C. Anderson, ‘The road-system of eastern Asia Minor with the evidence of Byzantine campaigns’, Journal of Hellenic Studies, 17 (1897), 39. (53) Konyalı, Konya Tarihi, 167; Babinger, ‘Kavalla’, 301; M. Zeki Oral, ‘Fatih Sultan Mehmed’in Gevale Kalesi ile Karaman İllerini Fethi ve Hamidi’nin Terci-i Bendi’, Vakıflar Dergisi, 4 (1958), 83. (54) Konyalı, Konya Tarihi, 322. (55) The Roman road known as the Augustan Via Sebaste was an important communications route linking the western Meander valley to the Cilician gates in the east of Anatolia. Konya could be approached from the west by the Via Sebaste, which linked Konya with the western province of Pisidia. From the Konya plain the road threaded its way westwards, passing over the Aladağ mountains to the north and the Kızılören mountains to the south and arriving at the Pisidian town of Yunuslar (Papa-Tiberiopolis). Following the road one could proceed westwards towards Yalvaç (Pisidian Antioch). Or one could turn south towards Beyşehir at the juncture of Yunusler with the Barsak Çay river valley. This branch of the Via Sebaste linked Beyşehir via the Irmak valley eastwards to the Çarşamba River valley passing through the mountains to the south of Konya: see Hall, ‘Notes and inscriptions’, 60–2; Stephen Mitchell, ‘R.E.C.A.M. Notes and studies no. 3: a Latin inscription from Galatia’, Anatolian Studies, 28 (1978), 95–6. (56) An Arabic inscription dated AH 872 commemorating Mehmed II’s rebuilding of the city’s fortifications is housed at the Konya Museum (inventory no. 872): Mehmet Önder, Mevlana Şehri Konya: Tarihi Kılavuz (Konya: Yeni Kitap Basımevi, 1962), 52, n. 2. (57) Tansel, Fatih Sultan Mehmed’in Siyasi ve Askeri Faaliyeti, 288, n. 47. (58) İbn Kemal, Tevarih, 247. (59) Konyalı, Konya Tarihi (Istanbul: Fatih Matbaası, 1970), 290. (60) An emigré from Khurasan, Hamidi (d. 1485) spent twenty years at Mehmed II’s court as poet and companion (nedim). As a member of Mehmed II’s retinue, Hamidi apparently was present at the campaign and seizure of the fortress of Gevele: Külliyat-ı Divan-ı Mevlana Hamidi, ed. İsmail Hikmet Ertalay (Istanbul: Milli Eğitim Basımevi, 1949), 12; Oral, ‘Fatih Sultan Mehmed’in Gevale Kalesi’, 88, n. 29.

Page 18 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 (61) BOA, Maliyyeden Müdevver 241. See Beldiceanu-Steinherr, ‘A propos des tribus Atçeken (Xve–XVIe siècles)’, Journal of the Economic and Social History of the Orient, 30 (1987), 121–95. (62) İnsuyu is now a village west of modern Cihanbeyli. (63) Beldiceanu-Steinherr, ‘A propos des tribus Atceken’, 121–95. (64) İnsuyu is located at the classical site of Pillitokome in northern Lycaonia, in the valley of the İnsi Deresi, lying 7.5 kilometres west and slightly north of Cihanbeyli, and 95 kilometres northeast of Konya: Klaus Belke, Tabula Imperii Byzantini, IV: Galatien und Lykaonien (Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 1984), 179. (65) The village of Zengicek, the centre of the subdistrict, has been renamed Kontkasa/Koçkaya. It lies 57 kilometres north of Konya. The ruins of a castle, known as the Zengicek Kalesi, lies 2 kilometres to the west of the settlement: Belke, Tabula Imperii Byzantini, 245. (66) Another village by the name of Göçü exists to the immediate north-east of Konya. The relative importance of the district of Göçü in the hinterland of Beyşehir, however, makes it a greater possibility that the timar here refers to the Göçü of Beyşehir. Göçü of Beyşehir comprised a subdistrict of fifty-four villages in 1507 and was the greatest producer of grain among the nahiyes or subdistricts of Beyşehir in the sixteenth century. See Mehmet Akif Erdoğru, Osmanlı Yönetiminde Beyşehir Sancağı (1522–1584) (Istanbul: IQ Kültür Sanat Yayıncılık, 2006), 345. (67) Giymir has been definitively identified as the classical site of Perta. It lies 58 kilometres north of Konya on the skirt of the Ballık mountain: Belke, Tabula Imperii Byzantini, 213. (68) Beldiceanu-Steinherr, ‘A propos des tribus Atçeken’, 158. (69) Erdoğru, Osmanlı Yönetiminde Beysehir Sancagı, 302. (70) Faruk Sümer, ‘Turgut-eli’, İA, 12: 120; Lindner points out that the nomads of Turgud, together with the Bayburd clan, controlled the highlands between Akşehir and Salt Lake by the early fourteenth century: Rudi Paul Lindner, Nomads and Ottomans in Medieval Anatolia (Bloomington: Indiana University, 1983), 91. (71) Gary Leiser, ‘Torghud’, EI2, 10: 570. Leiser points out that the Turgut were a major element, if not the most powerful one, of the Karamanid polity. (72) Despite its questionable historicity, characterised by large amounts of tendentious, fabricated, chronologically absurd and anachronistic material, Şikari’s Karamanoğulları Tarihi is nevertheless an important source as the only history of the Karamanid principality, especially in light of the paucity of works covering the period composed outside of the Ottoman tradition. Since few works of dissenting voices have survived Ottoman ideological domination of the production of historical works, Şikari’s Karamanoğulları Tarihi is significant for its proKaramanid perspective. The work’s tendentiousness may be contextualised as a counter-attack to Ottoman polemics against the Karamanids, who posed a serious threat to Ottoman rule in Anatolia during the fourteenth and fifteenth centuries. In addition to its anonymous authorship and suspect historical information, the work contains no dates, including that of its composition. Although next to nothing is known about the circumstances under which this text was composed, internal textual clues point to a few generalities regarding its production. One may surmise that Page 19 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 the final form of the history dates to the mid-sixteenth century, as the work’s reference to Shah İsmail’s battles with the Uzbeks in 1511 indicate. That the author lived under Ottoman rule during the sixteenth century and was well versed in Ottoman literary and political culture becomes clear from the contents of the history. See S. N. Yıldız, ‘Şikari’, TDVİA, forthcoming. (73) Şikari, Karamanoğulları Tarihi, 10–11. (74) The best source for Alaeddin’s reign is Şikari’s history, half of which is devoted to the life and times of the Karamanid ruler. Indeed the most historical part of Şikari’s work is the long reign of Alaeddin, whose reign represents the golden age of Karamanid power in the text. (75) Aşık Paşazade, Osmanoğulları Tarihi, 252, 520–1. Turgutoğlu Ömer Bey was Karamanoğlu Bey’s chief commander, or beylerbeyi: Faruk Sümer, ‘Turgutlular’, 121. (76) B. Y. Vladimirtsov, Moğolların İçtimai Teşkilatı: Moğol Göçebe Feodalizmi, trans. Abdülkadir İnan (Ankara: TTK, 1995), 60. Togan believes that the Turgut were either originally a Tatar tribe or a Turkish tribe that came to the region of Karaman with the Mongols: Zeki Velidi Togan, Umumi Türk Tarihine Giriş (Istanbul: Enderun Kitabevi, 1981), 318. Sümer, on the other hand, regards them as members of the Oğuz Yıva tribe, which came to Anatolia during the Seljuk period. The Yıva constituted an important tribal group in İçel during the Ottoman period: Faruk Sümer, Oğuzlar (Türkmenler). Tarihleri, Boy Teşkilatı, Destanları (Istanbul: Türk Dünyası Araştırmaları Vakfı, 1992), 261, 267. (77) For a detailed description of endowed villages and monuments of the Turgutogulları, see Zeki Oral, ‘Turgut Ogulları, Eserleri-Vakfiyeleri’, Vakıflar Dergisi, 3 (1956), 31–64, and idem, ‘Turgut Ogulları’, in IV. Türk Tarih Kongresi. Ankara 10–14 Kasım 1948. Kongreye Sunulan Tebliğler (Ankara: TTK, 1952), 141–57. (78) Osman Nuri Dülgerler, Karamanoğulları Dönemi Mimarisi (Ankara: TTK, 2006), 174. (79) Near Said-ili/Kadınhanı and immediately north of Ladik, a settlement on the Akşehir-Ilgın road before it bends around the mountains south to Konya: Dülgerler, Karamanoğulları Dönemi Mimarisi, 43, 50. (80) Oral, ‘Turgut Ogulları’, 140–2. (81) Dülgerler, Karamanoğulları Dönemi Mimarisi, 193, n. 513. (82) Ibid., 171. (83) See p. 309 above and n. 8. (84) Tursun Bey, Tarih, 146. (85) In the early sixteenth century under Ottoman administration, Karataş formed a kaza of the sancak, or liva dis trict of İçel: 387 Numaralı Muhasebe-i Vilayet-i Karaman, 34. This particularly inhospitable part of Rough Cilicia had few towns or villages when the 1530 defter was compiled. The situation seems to have changed little when J. Theodore Bent travelled through the region in 1891, which he found inhabited mostly by nomads: J. Theodore Bent, ‘Explorations in Cilicia Tracheia’, Proceedings of the Royal Geographical Society and Monthly Record of Geography, New Monthly Series, 12/8 (1890), 445 ff.; Ensar Köse and Doğan Atlay, Mut, Claudiopolis (Mut:

Page 20 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Razing Gevele and Fortifying Konya: The Beginning of the Ottoman Conquest of the Karamanid Principality in South-Central Anatolia, 1468 Mut Belediyesi Kültür Yayınları, 2005), 547. The Karataş of İçel is not to be confused with the well-known Neolithic site Karataş in the hinterland of Elmalı of Lycia excavated by the Bryn Mawr team in the 1960s: Machteld J. Mellink, ‘Excavations at Karatas-Semayük in Lycia, 1963’, American Journal of Archaeology, 68 (1963), 269 ff.; Machteld J. Mellink and J. Lawrence Angel, ‘Excavations at Karatas-Semayük in Lycia, 1967’, American Journal of Archaeology, 72 (1968), 243 ff. (86) Aşık Paşazade, Osmanoğulları Tarihi, 252, 520–1. (87) Aşık Paşazade, Osmanoğulları Tarihi, 252, 520–1. (88) Palmira Brummett, Ottoman Seapower and Levantine Diplomacy in the Age of Discovery (Albany, NY: State University of New York, 1994), 13. (89) Ibid. (90) Perry Anderson, Lineages of the Absolutist State (London: Verso, 1979), 22. (91) The Karamanid mountain fortresses of Ermenek and Mennan suffered a similar fate to Gevele. Nevertheless, several mountain fortresses remained in operation in the highly fortified Ottoman province of Karaman, which in 1530 had a total of fifteen castles, eight of which were in İçel. One such example is Mavga, a relatively small fortress, more akin to a watch tower than a fully fledged garrison, which was located on a particularly precipitous craggy outcrop 10 kilometres north of Mut. It never played a role in the Ottoman conquest, which may explain its survival. In 1530 it was garrisoned with thirty-four individuals, including six gunners (topçu) and possessed an assortment of artillery: Facsimile text, 387 No’lu Muhasebe-i Vilayet-i Karaman, 289, transcription, 35.

Access brought to you by:

Page 21 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries ROSSITSA GRADEVA

DOI:10.5871/bacad/9780197264423.003.0017

Abstract and Keywords This chapter examines how the frontier and the changing fate of the region influenced the military system as well as the provincial administration and agrarian regime. It focuses on Vidin in the period of the wars which made it a frontier outpost again, being either directly occupied (by the Holy League) or under immediate threat (in 1715–18). As in many other frontier areas, military activity affected relations between Muslims and Christians in the province and the town of Vidin. This chapter contributes to a better understanding of the complex impact of the wars during the expansion and contraction of the Ottoman state. The discussion is based mainly on documents from the series of kadi sicils, the records of the sharia court in the town, preserved from the last years of the seventeenth century onwards and kept in the Sofia National Library. Keywords:   Holy League, Vidin history, kadi sicils, Ottoman Vidin, Sofia National Library, Muslim-Christian conflict

A STRONGHOLD AND A PORT ON THE DANUBE FROM ROMAN TIMES, and the capital city of one of the Bulgarian states (1371–96), Vidin was incorporated into the expanding Ottoman principality shortly after the Battle of Nicopolis in 1396 as a punishment for its ruler’s cooperation with the crusaders. The battle marked the end of the Second Bulgarian Empire. From then until the mid-sixteenth century the town and the surrounding sancak constituted an important part of the Ottoman frontier against Serbia, Wallachia and Hungary. It was besieged and captured several times, while the region was laid waste during neighbouring states’ raids. However, the fall of Belgrade (1521) and especially the conquest of Temeşvar (Temesvár, Timişoara) in the 1550s left Vidin in the second and even the third line of Ottoman defence. The unstable Wallachian affiliation to the empire, however, justified the need to retain a solid chain of fortresses on the Danube. Vidin remained an important component of the Ottoman fortified line, a military as well as a civil port on the south bank of the river. Yet, until the war with the Holy League (Austria, Venice, Poland and Russia, 1683–99), despite several attacks from the north, the region remained free from actual fighting. It was only then, in 1688–90 in particular, that Hapsburg forces penetrated deep into Ottoman territory, capturing the fortress of Vidin (13

Page 1 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries October 1689). This event marked a turning point in the history of the town and the region, bringing them back to the front line, which was now against the Hapsburgs rather than the Wallachians, the latter coming more directly under Ottoman authority1 The importance of Vidin was enhanced after the war of 1715–18 when the Hapsburgs occupied and kept for thirty years Belgrade, much of northern Serbia, and even two districts of the Vidin sancak. During the war of 1737–9 (p.332) Vidin was again besieged by the Hapsburgs.2 After the Treaty of Belgrade (1739), Ottoman control over the region was restored. However, in the last decades of the eighteenth century and the first of the nineteenth century, there occurred one of the most serious attempts to secede in the history of the Ottoman state, that of Osman Pasvanoğlu3 and the Serbian Uprisings. In fact, it was only after the 1820s that Ottoman authority was finally restored in Vidin. In this chapter I examine how the frontier and the changing fate of the region influenced the military system as well as the provincial administration and agrarian regime. I focus on Vidin in the period of the wars which made it a frontier outpost again, being either directly occupied (by the Holy League) or under immediate threat (in 1715–18). As in many other frontier areas, military activity affected relations between Muslims and Christians in the province and the town of Vidin.4 This chapter is not intended to be a comprehensive study of Ottoman Vidin, but rather to contribute to a better understanding of the complex impact of the wars during the expansion and contraction of the Ottoman state.5 The chapter is based mainly on documents from the series of kadı sicil s, the records of the sharia court in the town, preserved from the last years of the seventeenth century onwards and kept in the Sofia National Library; tapu tahrir defteris and single documents related to the functioning of the timar system in the region; and the results of relevant archaeological research, as well as published research on Vidin’s history6

(p.333) Material Structures During the fifteenth century the main strongholds on the lower course of the Danube were Vidin, Niğbolu (modern Nikopol) and Silistre.7 Around them were clustered smaller forts on the river and deeper into the territory of modern northern Bulgaria, and several bridgeheads on the Wallachian side which were probably inherited from the Bulgarian tsardoms.8 These three fortresses constituted the backbone of the Ottoman frontier in the region until Bulgarian Liberation (1878) with one major shift. Starting from the seventeenth century Rusçuk (Ruse) and Yergöği (Giurgiu), which was on the opposite side of the river, gradually replaced Niğbolu and its twin, Holıvnik (Turnu Măgurele). The location of Rusçuk, safe from Hapsburg attacks, seems to have justified its choice as the base of the Ottoman fleet on the river and the seat of the kapudan paşa (admiral). During the seventeenth century all four places functioned as land fortifications, as bases of the respective sections of the Ottoman flotilla patrolling and guarding the river, and as important stores of munitions and provisions. What made Vidin particularly important was its proximity to the Hungarian front, which forced the Ottomans to preserve its more complex defensive system, including a constellation of satellite fortifications. Elsewhere on the Danube, after the fifteenth century and the establishment of Ottoman control over the region, the forts on the second and third lines of defence were abandoned and only a string on the river banks were kept, mainly to control the traffic across and along the waterway. In the fifteenth century the sancak of Vidin, probably more an uc9 than a regular administrative unit, was defended by the fortress in Vidin and the forts of Belgrad (later Belgradcık, modern Belogradchik), Filurdin (Florentin), İsfirlik (Svrljig), and Bane (Soko Banja). Later, Fethülislam (Kladovo) was added to them,10 then Severin and Orşova. The latter two remained in the sancak Page 2 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries until at least 1586, but in the seventeenth century were no longer part of it.11 Towards the end of the war with the Holy League and immediately after it, fortifications were built, Brza Palanka in modern Serbia, and Archar, Lom (Polomiye) and Kutlofçe (Montana) in modern Bulgaria.12 Six of the fortifications in the sancak—Fethülislam, Brza Palanka, Filurdin, Vidin, Archar and Lom— are on the Danube, while four are in the hinterland; only Archar and Lom are east of Vidin. This clearly shows that the sancak’s defences were orientated against the north-west, with the river considered the weak point. An order from 1701 concerning stocktaking of munitions in the serhad (frontier) lists Vidin, Lom, Belgrad, Filurdin, İsfirlik and Archar as fortresses (kale) in the sancak.13 What had happened to the rest is difficult to judge—they may have been destroyed during the war or, more likely, transferred to another administrative unit as part of the reorganisation triggered by the contraction of the Ottoman borders after the Treaty of Karlowitz.14 Some of the satellite forts are called interchangeably kale and palanka15 (Lom, Archar, Filurdin, Belgrad).16 At the beginning of the eighteenth century, units of the Janissary corps and of the local troops (yerlü) were stationed in all of them. There were detachments of the elite Janissary cavalry, the Six Regiments (altı bölük), in Vidin, Kutlofçe and Berkovitsa,17 the latter two being deep in the territory of the province, and probably included in it only after 1699. I have not yet discovered any evidence of a fort in Berkovitsa. (p.334) (p.335) Originally the fortifications of Vidin, the focal point in the system and the seat of the governors, consisted of a citadel and fortress walls surrounding the outer city.18 During the reign of Bayezid II (1481– 1512) the defences were reconstructed to accommodate artillery and firearms. At an unknown date the walls around the outer town were dismantled, perhaps after the conquest of Hungary, because they were no longer needed or, more likely, after the devastating raids of Michael the Brave, Prince of Wallachia and Transylvania, between 1595 and 1599, when Vidin, Filurdin and Fethülislam, along with Rusçuk, Silistre, Niğbolu and even towns deep within the Danube plain were plundered and set on fire, and their inhabitants killed, enslaved or forced to

Figure 17.1. The Ottomans’ Danube frontier, showing the principal places mentioned in the text.

move across the river.19 The citadel of Vidin, however, survived this violent (p.336) period. It continued to house the garrison, the residence of the governor, and stores of munitions, water and provisions.20 Its primary function was to repel eventual incursions of Wallachians and those of bandits. Changes were triggered by the Hapsburg con-quest of the region during the war with the Holy League. On capturing the town, the Austrians started to restore the fortifications of Vidin in line with Vauban’s system. This was completed by the Ottomans in 1722–3, after the Treaty of Passarowitz with the Austrians and Venetians (1718), when the fortress came to be directly on the frontier with the Hapsburgs.21 The citadel was modernised and a new semicircular fortress was constructed that protected some neighbourhoods from the river’s spring tide and military threats. Improvements to the Page 3 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries fortifications continued during the eighteenth century. Osman Pasvanoğlu renovated and strengthened them further, turning Vidin into an invincible stronghold, at least against the sultan’s army. The final enlargement and alterations to the fortifications were undertaken in the nineteenth century as part of the modernisation of the Ottoman military, and because of the region’s role as an outpost against the growingly independent Serbia and Wallachia, and the Russian menace. Also inherited from the medieval Balkan states was the use of river men-of-war.22 It seems that the earliest organisation of the fleet on the Danube included units of boats attached to the main ports which patrolled strictly defined stretches. From the early fifteenth century there had been a captain (kapudan) in Vidin.23 In the 1660s he was in command of ten boats with 300 soldiers whose main task was to persecute brigands who crossed the river or infested the islands.24 It is not clear whether before the seventeenth century the Vidin detachment and the others in the sancak had formed an independent unit under the general command of the sancakbeyi or had been part of a separate structure of the river fleet. From the beginning of the seventeenth century, (p.337) when the Tuna kapudanlık (Danube Admiralty) was formed and the seat of the admiral was established in Rusçuk, until the mid-eighteenth century, all small captainships on the lower Danube, including the one in Vidin, were under the command of this high officer, who at the end of the seventeenth and probably at the beginning of the eighteenth century held the rank of a provincial governor (mirimiran).25 Unfortunately, the organisation of the Ottoman military fleet on the river for the following century has been little researched. According to a couple of documents in the sicils of the kadı of Vidin, it was the admiral of the Danube fleet who initiated the appointment of the new garrison commander (dizdar) of the palanka of Archar in 1705,26 but it is not clear why. Neither do we know whether he had control over all the fortified ports on the Lower Danube or if the dizdar was also the commander of the local unit of ships. Under Osman Pasvanoğlu the defence of the Vidin fortress was supported by ships, which are mentioned in the context of the sieges by the imperial army. Until the end of the sixteenth century ships were constructed at several places along the Danube, including at Vidin. The local shipbuilding works produced boats for the last campaign of Süleyman the Magnificent in Hungary in 1566, and in 1573–4 after the destruction of the Ottoman fleet at Lepanto.27 It is difficult to tell when this activity ceased at Vidin, but during the second half of the seventeenth century the dockyard (tersane) at Rusçuk/Yergögi, under the direct supervision of the kapudan paşa, became the main basis for construction, repairs and equipment of the river stateowned ships. During the seventeenth and eighteenth centuries some workshops at Vidin were probably maintained for smaller repairs.

Human Resources As elsewhere in the Balkans, after the conquest and the pacification of the region, the Ottomans introduced to Vidin the ‘classical’ timar system which supported the provincial heavy cavalry (sipahi) troops, the fortress guards (mustahfızan), the highest-ranking officers on the boats, and commanders of some of the paramilitary corps. In the mid-fifteenth century the sancakbeyi of Vidin led a force of 292 men, including timarlı sipahis (122, of whom two were zaims), their servants (gulam, 46) and the auxiliary armed men (cebelü, 124) they had to bring.28 Around 18 per cent of the sipahis were (p.338) Christians, and 3 per cent were recent converts.29 In 1520–30, there were 10 zaims and 225 sipahis,30 that is a fighting force of probably double the number, or even more, if the proportion of auxiliaries seen in the previous registration remained the same. Around 1560 there were 12 zeamet and 190 timars of sipahis in the sancak but the register does not provide information about the nature of their holders’ military obligations and

Page 4 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries the number of the auxiliary soldiers. As elsewhere in the central Balkan provinces, by the middle of the sixteenth century, Christian sipahis had practically disappeared from the fighting force of the sancak.31 Unfortunately, the existing documentation concerning the timar holdings there from the last decades of the sixteenth century has, with a few exceptions,32 remained largely inaccessible to me.33 Even what is available, however, allows me to agree with the conclusion of Kayapınar that during the fifteenth to sixteenth century the number of the hasses, zeamets and timars in the region grew, and so did the number of their holders, with a noticeable tendency for the revenues of one timar to be shared by more than one sipahi.34 (p.339) Sources for the seventeenth century are scarce. There are no registers that reveal the full size of the sipahi forces. Some information may be drawn from the Law-book of Ayni Ali (1609–10), who records 12 zeamets and 195 timars for the sancak of Vidin.35 Given doubts about the dating of the figures provided by him and the fact that the numbers are very close to the late sixteenth-century ones,36 I am reluctant to draw any final conclusions based on it alone. On the other hand, while the sipahis from Vidin do not figure in the 1605–6 and 1607–8 roll-calls (yoklama defteri),37 they did participate in the war, only on another front, supporting the antiHapsburg revolt of István Bocskai in Transylvania.38 Evliya Çelebi, who visited the town in the 1660s, records 12 zeamet and 65 timars in the sancak. This would make about one-fourth to onethird of the cavalrymen supported by the timar system only a century earlier. However, according to the same author, the sancakbeyi, who at that time held a hass of 330,000 akçes, actually produced a force of 2,000 soldiers, which included the private troops of the pasha and another 300 furnished by the nazır (overseer) of the mukataas (tax units) in the kaza.39 As with Ayni Ali, suspicions about the reliability of Evliya’s data are widespread. My work so far, however, suggests that while occasionally the numbers may be inflated and some of his stories sound rather fabulous, most of the information he provides about places he had personally visited does stand comparison with the documentary evidence.40 Despite the remoteness of the Hapsburg frontier, during the first three centuries of Ottoman rule, the number of the fortress guards increased steadily. In Vidin alone their number grew from thirteen in 1454–5 to eightythree in 1530, including the commander of the garrison and his deputy, and then to ninety-three in 1560,41 and around 150 in the 1660s.42 (p.340) Thus for the seventeenth century the available sources43 suggest that the timarlı sipahi troops still existed in the sancak, but were clearly in decline; timars were being lost in favour of the expanding hasses of the sultan and the grand vizier.44 After the war with the Holy League the timar system and the cavalrymen and garrison troops supported by it seem to have disappeared from the landscape in this region. This situation, while reflecting a general trend of decline of the timarlı troops, shows that in the region of Vidin the process had developed faster than in the majority of the Balkans, where sipahis lingered at least in name until the mid-nineteenth century.45 After the war at the end of the seventeenth century Vidin emerged as a highly militarised town with a large number of Janissaries.46 Although precise numbers are unclear, we know that at the turn of the eighteenth century there were detachments of the elite Janissary cavalry (probably around sixty men on the territory of the sancak) and of the regular Janissary infantry as well as corps of the ‘local troops’ (yerlü) such as beşlüyan, farisan and so on.47 Half a century later, in 1750, according to Ahmed Cevad Pasha, the Janissaries’ number was an impressive 5,440 men, the largest contingent in the Balkans,48 probably including ‘local troops’. It seems that during the eighteenth century, as is well attested in other parts of the empire, ‘regular’ Janissaries in Vidin stopped rotating and started taking roots locally. At the same time enterprising (p.341)

Page 5 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries locals infiltrated their ranks. An exemplary case is that of the family of Osman Pasvanoğlu. In the 1760s his father was the commander (aga)49 of the 31st Janissary orta, and later Osman too was to become one of its commanders, implying the existence of hereditary affiliation to the unit. Another interesting trait of the Vidin Janissaries in the eighteenth century is the fact that a significant number among them bore Abdullah as their patronym. This may indicate that in Vidin converts, though not levied through the devşirme, may have served as a major source to fill the ranks of the corps even in the eighteenth century, serving alongside Janissaries who had migrated to Vidin from the territories lost after Karlowitz and Passarowitz, and local Muslims.50 An important feature of the region, which can be traced back to the fifteenth century, is the very close links between the Muslim inhabitants of the town (and the whole sancak) and the military corps. In the late seventeenth and eighteenth centuries these groups to a great extent overlapped, as is demonstrated by the way in which names of members of military units are often linked with professions, such as grocer, butcher, barber, dervish, tanner and tailor. At the same time almost every single inventory of the estates of deceased Muslim men in the eighteenth-century Vidin sicils includes varieties of swords and rifles, revealing a society that lived with war at the doorstep, always ready to fight. This impression becomes even stronger after the war of 1715–18.51 None of these traits was unique to Vidin, but there they seem to have been particularly acutely expressed. It is difficult to say when and how exactly the transformation from a timarlı to a Janissary region took place but it may be related to seventeenth-century changes in the concept of warfare which had made the sipahi troops obsolete. Of no less importance was the impact of the frontier and of the wars of 1683–99 and 1715–18. *** During the early period and as late as the second half of the sixteenth century, the majority, if not all, Muslim reaya (tax-paying subjects) in the sancak seem to have been engaged in paramilitary activities. Some, especially those living in its western parts, (p.342) served as irregular cavalry (akıncı)52 whose major tasks included ravaging territories under attack, scaring the local population and clearing the way before the main body of the army, and taking part in defence in the event of enemy raids. In return they received part of the spoils and occasionally tax relief. Towards the end of the sixteenth century akıncıs were still found in the towns of Bane, İsfirlik, Belgrad, and several villages in the sancak. While all of them were Muslims, there is evidence that the corps contained converts.53 Muslims were also involved in the defence of the forts, although direct evidence for this in Vidin does not exist.54 However, in nearby Niğbolu under Mehmed II and Süleyman I, the Muslim reaya were obliged to serve in defending the fortress there and its counterpart across the Danube in lieu of the so-called ‘extraordinary taxes’.55 These arrangements were confirmed several times, including in 1697–8. Among the Muslim reaya in Niğbolu there were also serhadlis, irregular soldiers on the border within the so-called serdengeçti corps.56 The latter were conscripted by the state as volunteers from among the Janissaries and served as a vanguard entrusted with the most difficult tasks. The same arrangements were in place at Silistre, where the Muslim reaya also supported the local travel station and expenses for state messengers.57 The information about Niğbolu and Silistre, and the indirect evidence for Vidin in the seventeenth century, suggests that many, if not all, Muslims in Vidin and the province continued to be directly involved in warfare (let us not forget the 2,000 men Evliya Çelebi mentions) as volunteers or in auxiliary services, and that the boundaries between Muslim reaya and the askeri

Page 6 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries had become even more blurred than in earlier (p.343) centuries.58 It is not surprising that Vidin is one of the places where the division between ‘Muslims’ and ‘reaya’, in which reaya stands for Christians, appears in local documentation rather early, from the first decade of the eighteenth century.59 Apart from its probably derogatory meaning, this may have reflected a reality in which all Muslims could serve as military and the Christians were completely excluded from it. In the fifteenth and sixteenth centuries, however, the Christian reaya in the sancak of Vidin had also performed paramilitary and auxiliary services for the Ottoman military system60— as derbendcis (mountain pass guards),61 martolosan,62 voynuks,63 müselleman,64 and filurciyan.65 Christians were also involved in the maintenance of the fortresses, mainly as craftsmen—carpenters, builders, rope-makers, caulkers, smiths—but also as producers of arrows, shields, guns, cannons.66 Most of these men were exempted from the ‘extraordinary taxes’, and sometimes also from the ispence or paid it at a reduced rate (the derbendcis, martolosan, filurciyan, müselleman), the cizye (the martolosan, filurciyan, müselleman, voynuks), the tithes (the filurciyan). Some received a baştina, a tax-exempt farm (voynuks, martolosan, filurciyan); others were salaried (p.344) (especially the craftsmen). While these institutions were staffed as a rule by Christians, their commanders, who often received timars, were exclusively Muslims. Towards the end of the sixteenth century, after the frontier moved west, some of the corps were disbanded (such as the filurciyan), others were preserved, but their nature changed and they had less and less to do with military obligations. Their members were reduced to the status of ordinary reaya. Unfortunately, there are no reliable sources about the Vidin Christians’ participation in defence and military activities for the seventeenth century and until the war of 1715–18. Christian participation in Ottoman military forces was not confined to Vidin. A glimpse at the other fortresses on the Danube may give some ideas as to how things might have developed in the century and a half between the late sixteenth century and 1718. During the fifteenth and sixteenth centuries a significant number of Christians in Niğbolu were engaged in real military service, defending the frontier and the fortress under the command of the sancakbeyi, as archers, gunners, martolosan, and in maintenance, as smiths, carpenters, arrow-producers, builders, in return for substantial tax exemptions. Some served also in the fortresses of Holıvnik (Turnu Măgurele) and Yergöği (Giurgiu) across the Danube. Another group operated under the command of the Niğbolu kapudan, employed both as a fighting force and to undertake maintenance. During the seventeenth century the numbers of all these groups dropped significantly, but more importantly, now the majority were auxiliaries, serving on the flotilla and as craftsmen, rather than having paramilitary or strictly military functions.67 Until the end of the sixteenth century only a small group of Christians were engaged as craftsmen, mainly smiths and carpenters in the defence of the fortress of Silistre. As with Vidin and Niğbolu there were also voynuks, who in the seventeenth century were no longer a paramilitary corps but were responsible for taking care of the sultan’s horses. The real change occurred after the Long War (1593–1606), when the local Christians fought bravely against the Wallachian invasion. The Ottomans appreciated their loyalty and granted them considerable tax exemptions. In exchange the Christians were obliged to equip and maintain two military boats, which were reduced to one at the beginning of the eighteenth century.68 Vidin seems to have been closer to the pattern of Niğbolu than of Silistre. What makes it distinct, however, is the fact that after the Treaty of Passarowitz (1718), at which the Ottomans lost substantial Balkan territories to Austria, local Janissaries applied to the central authorities to sanction their demand for the banishment of the Christians from the newly fortified part of the town, referring to the kanun-i serhad (the law of the frontier). The Christians were ordered Page 7 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries to sell their houses to Muslims and build new ones of wood, not stone, outside the walls so that these could not serve as a (p.345) shelter for an enemy in the case of attack. Their two churches were preserved where they were but were surrounded with high walls. Practising Christianity became very difficult, especially since Christians were not allowed to stay in the fortified part of the town after sunset. Jews remained in the walled part of the town, as a clear indication that they were the more trusted of the non-Muslim groups.69 The spatial division of faiths remained in force until the end of the Ottoman domination in the region, although some neighbourhoods outside the walls were mixed. The reasons behind this spectrum of attitudes to non-Muslims in the three strongholds on the Danube were varied. The decline in the use of non-Muslims in the military establishment was contemporaneous with the increasing ‘sunnification’ of the Ottoman Empire and in line with one of the limitations imposed by Islamic legal theory on the status of the so-called ‘protected’ (zimmi) non-Muslims in the Islamic state which prohibits their possession of weapons. It was accompanied by a steady growth of the Muslim population, which became sufficiently large to meet the requirements of both offensive and defensive warfare on its own. Another important factor was the expanded Janissary corps, which increasingly became an instrument for Muslim reaya to change their status.70 The curtailment of non-Muslims in regular military service was also rooted in the political climate, for the Ottoman authority was becoming increasingly aware of the shift in their Christian subjects’ sentiments towards Ottoman enemies. Normally, conquered non-Muslims— when they submitted voluntarily to the Muslims—would be left in their residential areas, even in the citadel of a fortress, as several examples from the Ottoman conquest show. However, any subversive act could upset this arrangement.71 So far I have not come across any evidence of any acts by the Vidin Christians in support of the Hapsburgs. However, there was (probably) in 1688 a rising of Bulgarian Catholics in north-western Bulgaria as well as several scattered rebellions and conspiracies in northern Bulgaria among the Orthodox.72 During the war with the Holy League, Orthodox Christians in Serbia widely supported the Austrian army. At (p.346) its withdrawal thousands of them followed, led by Patriarch Arsenije III of the Peć Patriarchate. Yet none of these events, which took place very close to Vidin, seems to have upset relations between Christians and Muslims there, at least not immediately. After the war Christians in the district of Vidin applied for permission to reconstruct their churches, which had obviously suffered during the war.73 In 1696 and 1698–9 they were relieved from paying cizye on the grounds of poverty and devastation caused by military action.74 The list of the Ottoman authorities’ acts showing that they were dedicated to preserving good relations with their Christian subjects could be extended. However, the Austrian occupation brought about not only significant displacement of Vidin’s population,75 but caused considerable problems for local Muslims. More than twenty years later the latter still referred to the ‘infidels’ invasion’ (kefire-i istilâ) as the reason for the flight of many local people, for the destruction and desecration of places of worship, and for the loss of the registers of the kadı court, which required recording many documents anew.76 This shock, reinforced by the even more disastrous war of 1715–18, when the district once again became the theatre of military hostilities, and not last, the fact that many of the Muslims now living in Vidin had come from the lost territories, must have laid a lasting imprint on the attitude of the Muslim inhabitants of the town vis-à-vis Christians. A final blow to the relations between the Ottomans/Muslims and the local Christian reaya must have been dealt by the growth of banditry during the war of 1715–18, which prompted the local

Page 8 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries authorities to demand the signing of a collective liability declaration from the local Christian reaya that they would not support the hayduds.77

(p.347) Administrative System and Agrarian Regime The territory of Vidin sancak remained relatively stable through the centuries. For most of the period under study until the end of the seventeenth century Vidin belonged to the eyalet of Rumeli. The new geopolitical situation after the Treaty of Karlowitz forced the Ottomans to restructure the administrative units in the region. Vidin was then included in the eyalet of Temeşvar and the frontier. After 1718 it went back to Rumeli, this time as the immediate serhad. Significant fluctuations in its boundaries can be observed during the fifteenth century, related to the Ottoman expansion westward, and then after the end of the seventeenth century, reflecting the reorganisation of the Ottoman frontier and the loss of territories to the neighbours. During the fifteenth and sixteenth centuries the timar system and the corps based on it underlay the administrative system in the region. Indeed, one of the main features of the landholding there, with a direct bearing on the development of the administrative system, was the introduction of the miri system. The imperial hass in the sancak comprising 498 villages in 1560 was, according to D. Bojanić-Lukać, the largest one in the Balkans.78 Other hasses were held by the sancakbeyi, and in the late seventeenth century by the grand vizier.79 The rest of the taxyielding sources were assigned to timar- holders and to Muslim and Christian reaya who provided various services to the state, mainly related to warfare. This was paralleled by the absence in Vidin of agrarian mülk properties and vakıfs, that is, no significant plots of land were held as quasi-private property even by the powerful lords from the time of the conquest.80 According to a synoptic defter of 1454–5, the sancak consisted of the vilayet of Vidin, and the nahiyes of İsfirlik, Bane and Belgrad, as well as the unspecified units of Timok (along the river of the same name), Gelviye (along the upper course of the river Crni Timok), Veleşniçe, Çerna Reka (along the middle and lower course of the river Crna Reka) and Zagoriye. All reflect the subunits within the troops coming out of the province. Around the mid-sixteenth century the administrative structure seems to have attained its optimal shape, including the nahiyes of Vidin, İsfirlik, Bane, Timok, Zagoriye, Çerna Reka, Fethülislam, Krivina, Lom and Belgrad.81 However, there was also a parallel administrative system based on judgeships, the kazas of Vidin, Bane, ˙ sfirlik and Fethülislam. In 1699 at least two districts (Berkovitsa82 and (p.348) Kutlofçe83) formerly in the Pasha sancak had been added to the Vidin one, while others (Fethülislam and Bane) were no longer part of it. Between 1718 and 1739 Fethülislam and Krajna, the northwesternmost parts of the sancak, were in Hapsburg possession, but then were joined again to the Vidin province, where they stayed until the beginning of the nineteenth century. The evolution of internal divisions reflected changes in military and political structures, in the first place the disappearance in the course of the seventeenth century of the timars and of the tax-exempt holdings of reaya groups with paramilitary functions at the expense of the imperial hasses. The Vidin nezaret, founded on the hass lands during the seventeenth century, expanded after the war with the Holy League to include the territory of the whole sancak, which was divided into several territorial and/or tax-based mukataas.84 The parallel division into judgeships remained, at the end of the seventeenth century comprising Vidin, Fethülislam, İsfirlik, Bane, Timok85 and Lom.86 However, they seem to have ceased to function as the basis of the Ottoman administration in the province. During the eighteenth century the sancak of Vidin had turned into a nezaret, and the division into mukataas had replaced the classical one of the sancak-kazanahiye type in the province, although occasionally there is indication that mukataa X belonged to (tabi’) kaza Y.87 Their revenues were allocated as ocaklık88 for the salaries of the paid troops in Page 9 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries the sancak and in several other provinces, such as Temeşvar and Isakçea, and of captains and crews of the river ships stationed up the Danube.89 (p.349) This process, combined with the massive migrations and the displacement of villagers during the wars at the end of the seventeenth and in the first half of the eighteenth century, led to the emergence of a specific system of land ownership in the area (gospodarlıks90). The title deeds (tapu) were no longer in the possession of the peasants cultivating the land, but in the hands of new agents who intervened between the nazırs, acting as representatives of the sultan, and the direct cultivators. The tapus were bought by local notables, mainly Janissaries, citizens of Vidin. This deprived the villagers of their legal possession rights to the land, for they were now treated as tenants. Towards the end of the eighteenth century this process was given additional impetus by another series of wars, and by the devastation of the countryside by massive brigandage, by plagues, and especially by the regime of Osman Pasvanoğlu. The peasants no longer paid the traditional taxes, which were replaced by a single tax to the owner of the tapu and one to Pasvanoğlu (in the Vidin area) who had acquired, among other positions, the combined prerogatives of the former superintendents of the imperial hasses and of the nazır.91 As late as the first quarter of the nineteenth century the Janissaries in Vidin claimed that, in compliance with the kanun-i serhad, the law of olden times, property rights on land in the frontier areas belonged exclusively to the Muslim soldiers from these fortresses.92 The transformation of the villages into gospodarlıks or agas’ villages, as they were also called, corresponds to the process of čitlučenje taking place in the adjacent Belgrade pasalık, where a similar agrarian regime established by the Janissaries was among the direct causes for the uprising of 1804, leading to the emergence of the Serbian Principality, the first successor state to the Ottoman Empire in the Balkans.93 This agrarian relations problem seems to have been tinged by religious confrontation, because at that time the Muslim population was concentrated in urban centres while the peasants were exclusively Christians.94 The regime (p.350) remained in force until well into the nineteenth century, after the introduction of the agrarian reforms in most of the Ottoman Balkans, probably because the central authorities were reluctant to confront openly local Muslim society in this strategic region. It was only a series of riots and especially the uprising of 1850 that convinced it that reforms could no longer be avoided. These were finally introduced only in 1863.95

Conclusion When assessing the impact of war on the Vidin region one must bear in mind the similarities but, more importantly, the distinctions in the two periods when it was directly on the Ottoman frontier. During the fifteenth and sixteenth centuries, the wars were waged by a rapidly expanding state; in the late seventeenth and eighteenth centuries, despite some reversals in the trend, wars led to retreat and contraction. Both periods put a great strain on the entire local society. The different circumstances required different responses, and, naturally, had different consequences. Warfare demanded the involvement of all the available manpower and infrastructure, and flexibility on the part of the authorities. After the Crusade of 1443–4, and with the exception of the raids of Michael the Brave, Prince of Wallachia, in the late 1590s, no other serious disruption of the order had taken place in the region. However, from 1688, the Ottoman territories again became a theatre of war. This required changes in the administrative and defensive systems, the construction of new forts, especially along the Danube, forming a constellation around the modernised and enlarged fortress of Vidin. In the first period the Ottomans took advantage of the existing infrastructure and human potential and adapted it to their needs, introducing at the same time the kadı court and the timar

Page 10 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries system, the two most important institutions in provincial administration. Although occupying the lowest level in the military corps, Christians were conspicuously present in them, and they were widely used in paramilitary and auxiliary corps. Over the centuries local circumstances in combination with the changed realities led to the abandonment of the traditional timarlı troops and the transformation of Vidin into a Janissary town and to the militarisation of local (Muslim) society, a society in which Muslims were the soldiers, and Christians were already entirely excluded from warfare, even from auxiliary functions. While reducing the participation of the non-Muslims in the armed forces had been a general trend, it seems that the attitude to them varied from place to place, and the closer to the border the less acceptable they were in (p.351) defence, though they still had a role in maintenance. Thus, in Vidin, on the basis of the ‘law of the frontier’ (which, interestingly, does not appear in sources from the earlier period!) Christians were excluded not only from service but also had to leave their residences in the fortified part of the town. The same law was used by the Vidin Janissaries to give legal sanction to the deprivation of Christian peasants of land ownership. Similar events occurred around the same time, in the 1720s, in Niš, Sofia, and several other places that had changed hands during the wars or had been under direct threat, and had remained in frontier areas. In some cases the justification was the kanun-i serhad, in others it was different, but the result was the same—the expulsion of Christians from the central/fortified parts of the town. This seems to have been initiated by local Muslim communities, while the central authorities had to accept it more or less as a fait accompli. The reasons for their demands are not necessarily rooted in concrete acts of the local Christians but rather in an atmosphere in which Muslims felt more and more vulnerable and exposed to danger from outside. ‘The infidels’ invasion’ was definitely an important memory for the Vidin Muslims, who attributed many of their misfortunes to it, as did Jews. Probably Christians did too, but Muslims saw in them potential danger to the Ottoman rule. This memory was reinforced by the Hapsburg expansion, which brought the frontier very close to Vidin. These events contributed to the emergence of a frontier spirit, very different from that of the expanding Ottoman Empire. It did not imply a pervasive hostility between the faiths, and there were even friendships between Muslims and Christians across the border in the tradition of the old epics, but it made the Muslims very sensitive to the position of Christians, and of course gave them opportunity to profit economically. Notes: Proceedings of the British Academy 156, 331–351. © The British Academy 2009. (1) On relations between Wallachia and the Ottomans during the period, see Mihai Maxim, L’Empire ottoman au nord du Danube et l’autonomie des Principautés Roumaines au XVIe siècle: Études et documents (Istanbul: Isis, 1999); Viorel Panaite, The Ottoman Law of War and Peace: The Ottoman Empire and Tribute Payers (New York: Columbia University Press, 2000). (2) On the contemporary military and diplomatic developments and their effect on the region see Ivan Parvev, Habsburgs and Ottomans between Vienna and Belgrade (1683–1739) (New York: Columbia University Press, 1995), 75–115, 163–91,211–46. (3) See on him Rossitsa Gradeva, ‘Osman Pazvantoğlu of Vidin: between old and new’, in Frederick Anscombe (ed.), The Ottoman Balkans, 1750–1830 (Princeton: Markus Wiener, 2006), 115–61, and the bibliography referred to there. (4) For space considerations the effect of these events on the Jewish community will only be mentioned in passing.

Page 11 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries (5) Space precludes discussing here the very important issue of the migrations towards and from Vidin throughout the centuries. (6) There is vast literature on Vidin and its region, mainly for the early Ottoman centuries and for the late eighteenth and nineteenth centuries but much less for the seventeenth and eighteenth centuries. In addition to the works referred to in this article, see the following important studies: Bistra Tsvetkova, Prouchvaniya na gradskoto stopanstvo prez XV–XVI vek (Sofia: Nauka i izkustvo, 1972), including law-books for Vidin and the sancak, 164–6, 168–78; eadem, ‘Za etnicheskiya i demografski oblik na Vidin prez XVI vek’, Izvestija na Etnografskiya Institut i Muzey 7 (1964), 11–21; Vera Mutafchieva, L’Anarchie dans les Balkans à la fin du XVIIIe siècle (Istanbul: Isis, 2005), chapters 2–7, passim, on Osman Pasvanoğlu and the situation in the region of Vidin during the last decades of the eighteenth and the beginning of nineteenth century; Halil İnalcık, Tanzimat ve Bulgar Meselesi, Doktora Tezi’nin 50. Yılı 1942–1992 (Istanbul: Eren, 1992), who discusses at length the Niš and Vidin Uprisings of the midnineteenth century, and the gospodarlık regime in Vidin, 45–67, 83–107; Machiel Kiel, ‘Urban development in Bulgaria in the Turkish period: the place of Turkish architecture in the process’, International Journal of Turkish Studies, 4 (1989), 101–2. (7) Immediately upon their conquest they were made administrative centres of the newly established units stretching along the Danube from the Iron Gate to the Black Sea. (8) See Aleksandăr Kuzev and Vassil Gyuzelev (eds.), Bălgarskisrednovekovni gradove ikreposti, I: Gradove i kreposti po Dunav i Cherno more (Varna: Izdatelstvo ‘G. Bakalov’, 1981), 94–200, passim, on the medieval defence system including information on the Ottoman period, the latter based mainly on archaeological research and narrative sources. (9) See Rossitsa Gradeva, ‘Administrative system and provincial government in the Central Balkan territories of the Ottoman empire, 15th century’, in eadem, Rumeli under the Ottomans, 15th-18th Centuries: Institutions and Communities (Istanbul: Isis, 2004), 26–8. (10) 370 Numaralı Muhasebe-i Vilâyet-i Rum-ili Defteri (937/1530) (Ankara: Basbakanlık Devlet Arşivleri Genel Müdürlüğü, 2002), 2: 594–7, 604, 607. Fethülislam appears for the first time as a fortified place in this register. (11) Mihnea Berindei, Marielle Kalus-Martin and Gilles Veinstein, ‘Actes de Murad III sur la région de Vidin et remarques sur les qanuns ottomans’, Südost-Forschungen, 35 (1976), 15. (12) For all of them we find references as ‘newly constructed palanka’—Saints Cyril and Methodius National Library-Sofia (hereafter NBKM), Oriental Department, S 13, fol. 30a, doc. II (the establishment of Brza was clearly still in progress since the order is about staffing its garrison with soldiers from the palanka of Lom); fol. 27b, doc. II (Lom); fol. 26b, doc. II (Kutlofçe); fol. 16b, doc. II (Archar), of 1698–9. (13) NBKM, Or. Dept., S 14, fol. 31b, doc. II. (14) Bane, Brza Palanka and Fethülislam were the westernmost forts in the Vidin system. (15) See Chapter 8 by Burcu Özgüven in this volume, and Behija Zlatar, ‘Tipologija gradskih naselja na Balkanu u XVI vijeku’, in Verena Han (ed.), Gradska kultura na Balkanu (XV–XIX vek) (Belgrade: Institut des études balkaniques, 1988), 65.

Page 12 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries (16) See for example for Belgrad, an old fortress in the province, NBKM, Or. Dept., S 13, fol. 23a, doc. I (as a ‘newly constructed palanka’, probably reconstructed and strengthened after the withdrawal of the Hapsburgs, 1699), and S 14, fol. 40b, doc. I (as kale, 1701). (17) NBKM, Or. Dept., S 13, fol. 29b, doc. I, of 1699. (18) See in more detail on its fortifications, Kuzev and Gyuzelev (eds.), Bălgarski srednovekovni gradove, 98–115; Svetlana Ivanova, ‘Widin’, in EI2, 9: 206–8; Rossitsa Gradeva, ‘War and peace along the Danube: Vidin at the end of the seventeenth century’, in eadem, Rumeli, 107–19; Nikolay Tuleshkov, ‘Krepostnoto stroitelstvo po vreme na osmanskoto vladichestvo’, in Stefan Boyadzhiev et al., Krepostnoto stroitelstvo po bălgarskite zemi (Sofia: Stampa, 2000), 289–300. (19) See descriptions of the events left by Christian sources in Konstantin Veliki, ‘Pohodite na Mihai Viteazul na yug ot Dunav’, Istoricheski Pregled, 29 (1973), 65–7, 70–1; Boyan Beshevliev and Pavlina Boycheva, ‘Istorikogeografski svedeniya v tri dokumenta, posveteni na Mihai Hrabri (kraya na XVI-nachaloto na XVII vek)’, in Axinia Dzhurova, Georgi Bakalov et al. (eds.), Obshtoto i Spetsifichnoto v Balkanskite Kulturi do Kraya na XIX Vek: Sbornik v chest na Prof. V. TăpkovaZaimova (Sofia: Gutenberg, 1999), 256–64. See a detailed account of the effect of the Wallachian attack on Silistre in particular based on foreign and Ottoman sources in: Stefka Părveva, ‘Bălgari na sluzhba v osmanskata armiya: voenni i voennopomoshtni zadălzheniya na gradskoto naselenie v Nikopol i Silistra prez XVII vek’, in Elena Grozdanova, Olga Todorova, Stefka Părveva, Yoanna Spisarevska, Stefan Andreev, Katerina Venedikova, Kontrasti i Konflikti ‘zad kadăr’ v bălgarskoto obshtestvo prez XV–XVII vek (Sofia: Gutenberg, 2003), 240–2. (20) In the seventeenth century the other fortified places along the Danube also had just one citadel, where the garrison, the stores for provisions and munitions, the seat of the commander, and a mosque would be located. See for Niğbolu (an exception, with two, one at the port and another on a hill dominating the area) and Silistre, Părveva, ‘Bălgari na sluzhba’, 227–8; for Rusçuk, Teodora Bakărdjieva and Stoyan Yordanov, Ruse, Prostranstvo i Istoriya (kraya na XIV v.-70-te godini na XIX v.): Gradoustroystvo, infrastruktura, obekti (Ruse: Avangard Print, 2001), 63–86. (21) Three inscriptions of 1723 have been preserved, commemorating the work of the Ottoman architect who, upon the sultan’s order, had completed the fortifications. They praise in particular the wall along the river fortified by the Austrians: Vera Mutafchieva (ed.), Rumeliyski delnitsi i praznitsi ot XVIII vek (Sofia: OF, 1978), 71. (22) On the organisation of the Ottoman fleet on the Danube in the first centuries of Ottoman rule, see Rossitsa Gradeva, ‘Shipping along the lower course of the Danube (end of the seventeenth century)’, in Elizabeth Zachariadou (ed.), The Kapudan Pasha: His Office and His Domain (Rethymno: Crete University Press, 2002), 301–23. (23) Dušanka Bojanić-Lukać, Vidin i Vidinskiyat sandzhak prez XV–XVI vek. Dokumenti ot Tsarigradskite arhivi, ed. Vera Mutafchieva and Mihaila Staynova (Sofia: Nauka i Izkustvo, 1975), 73: the timar of Ahi Ali, kapudan in Vidin in 1454–5. A note in the margin of the register explains that he had served in this position for more than ten years, but had earlier received a salary in cash. (24) Evliya Çelebi, Seyahatname, 6: 97–8.

Page 13 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries (25) See on the institution Svetlana Ivanova, ‘Ali Pasha: sketches from the life of a Kapudan Pasha on the Danube’, in Zachariadou (ed.), The Kapudan Pasha, 325–45; Gradeva, ‘Shipping’, 307–13. (26) NBKM, Or. Dept., S 38, fol. 13b, doc. II, fol. 14a, doc. I (the berat), of 1705. (27) Colin Imber, ‘The navy of Süleyman the Magnificent’, in idem, Studies in Ottoman History and Law (Istanbul: Isis, 1996), 62; Olga Zirojević, Tursko vojno uređenje u Srbiji, 1459–1683 (Belgrade: Istorijski Institut, 1974), 232. (28) Bojanić-Lukać, Vidin i Vidinskiyat sandzhak, 55–94: sixteen timars were held jointly by more than one timariot but this does not change the number of the horsemen serving on campaigns. See also Vera Mutafchieva, ‘Vidin i Vidinsko prez XV-XVI vek’, included as an introduction to Bojanić-Lukać, Vidin i Vidinskiyat, 22–9; Ayşe Kayapınar, ‘Le sancak ottoman de Vidin du XVe à la fin du XVIe siècle’ (Unpublished Ph.D. thesis, École des hautes études en sciences sociales, Paris, 2004), 214–19 also gives an overview of the timar system and the sipahi and mustahfız troops supported by it during the fifteenth and sixteenth centuries, without distinguishing between the two groups. (29) See on them Halil Inalcık, ‘Stefan Dusan’dan Osmanlı İmparatorluguna: XV. Asırda Hıristyan Sipahiler ve Menseleri’, in Mélanges Fuad Köprülü—60 Doğum yılı münasebetiyle Fuad Köprülü Armağanı (Istanbul: Ankara Üniversitesi Dil ve Tarih-Coğrafya Fakültesi, 1953), 207–48; for Vidin, 224, 233. (30) 370 Numaralı Muhasebe-i Vilâyet-i Rum-ili Defteri, 608. (31) Bojanić-Lukać, Vidin i Vidinskiyat, 97–160; Mutafchieva, ‘Vidin i Vidinsko’, 35–6. Unfortunately the other published timar registers for the intermediary period are only fragmentary and do not allow conclusions about the exact military force provided by the sancak. See Dušanka Bojanić, ‘Fragmenti zbirnog popisa Vidinskog sanđakaiz 1466 godine’, in Mešovita Građa (Belgrade, 1973), 2: 5–77; eadem, ‘Fragmenti opširnog popisa Vidinskog sanđaka iz 1478– 81’, in ibid., 79–177; ‘Timari văv Vidinsko, Berkovsko, Belogradchishko i porechieto na Timok (1479/80 g.)’, three fragments of an icmal defter, published in Nikolai Todorov and Boris Nedkov (eds.), Izvori za bălgarskata istoriya: Seriya XV-XVI vek (Sofia: BAN, 1966), 13: 103–59; ‘Otkăs ot podroben registăr na timari v sandzhak Vidin ot vtorata polovina na XVI vek’, in Bistra Tsvetkova and Asen Razbojnikov (eds.), Izvori za bălgarskata istoriya (Sofia: BAN, 1972), 16: 496–523: some of the timars in the latter belonged to sipahis, others to fortress guards. (32) See NBKM, Or. Dept., F 26 A, archival unit (a.u.) 1997, fol. 2b, doc. II, about a timar in the nahiye of Zagoriye, of 1584; F 1, a.u. 14690, 221–2: documents about the movement of timars in the nahiyes of Vidin, Timok and Bane, of 1588/89, and D 15, fol. 12b, and fol. 36a: respectively timars in Vidin and Fethülislam, of the same year. (33) See the list in Kayapınar, ‘Le sancak ottoman de Vidin’, 113–14: BOA, Maliyyeden Müdevver (MAD) 123 (of timars), of 1574/5; Tapu Tahrir (TT) 664 (icmal), of 1574–95; MAD 15428 (of timars), of 1584; MAD 15278 (of timars), of 1591; Tapu ve Kadastro Genel Müdürlüğü, Ankara (TK) 219 (of timars), of 1579—according to Kayapınar it includes data on six hasses, 16 zeamets and 465 timars, those of the fortress guards included; TK 57 (mufassal), of 1586. Undated information about the sixteenth century evaluates the military based on the timar system in the sancak at 201 sipahis (without the cebelüs) and 268 fortress guards, see Evgeni Radushev,

Page 14 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries Agrarnite institutsii v Osmanskata imperiya prez XVII-XVIII vek (Sofia: Akademichno izdatelstvo ‘Marin Drinov’, 1995), 78. (34) Kayapınar, ‘Le sancak ottoman de Vidin’, 214. See also Strashimir Dimitrov, ‘Politikata na upravliavashtata vărhushka v Turtsiya spriamo spahiystvoto prez vtorata polovina na XVIII vek’, Istoricheski Pregled, 18/5 (1962), 36, who explains that this tendency continued also in the seventeenth century combined with a growing number of the cebelüs the sipahis were obliged to bring on campaigns. (35) Gălăb Gălăbov, Turski izvori za istoriyata na pravoto po bălgarskite zemi (Sofia: BAN, 1961), 1: 107. (36) See for example Caroline Finkel, The Administration of Warfare: The Ottoman Military Campaigns in Hungary, 1593–1606 (Vienna: VWGÖ, 1988), 56–7. Hezarfen’s treatise of 1669–70 (Hezarfen Hüseyin Efendi, Telhisü’l-beyan fi kavanin-i Al-i Osman, ed. Sevim İlgürel (Ankara: TTK, 1998), 117, quoted from Kayapınar, ‘Le sancak ottoman de Vidin’, 219, n. 1016) gives the same numbers but this is not surprising bearing in mind that to a large extent he relied on the data provided exactly by Ayni Ali, Lütfi Pasha and Kâtib Çelebi, see Victor Ménage, ‘Husayn Hezarfenn’, EI2, 3: 623b-624a. (37) Vera Mutafčieva, ‘Sur l’état du système des timars au cours de la première décade du XVIIe s. d’après les yoklamas datants de 1014 et 1016 de l’Hégire (1605–1606 et 1607–1608 A.D.)’, in Vera Mutafčieva and Strašimir Dimitrov, Sur l’état du système des timars des XVIIe–XVIIIe ss. (Sofia: BAN, 1968), 10–11. (38) Finkel, The Administration of Warfare, 59. Unfortunately the mobilisation order does not specify numbers. (39) Evliya Çelebi, Seyahatname, 6: 97–8. (40) See also Robert Dankoff, ‘Evliya Çelebi’, in Historians of the Ottoman Empire, Cemal Kafadar, Hasan Karateke and Cornell Fleischer (eds.), www.ottomanhistorians.com, and cf. Chapter 6 above, Malcolm Wagstaff, ‘Evliya Çelebi, the Mani and the fortress of Kelefa’. (41) Bojanić-Lukać, Vidin i Vidinskiyat, 73–5, 137–45; 370 Numaralı Muhasebe-i Vilâyet-i Rum-ili Defteri, 594. Some data about the fortress guards can be drawn from the above-cited fragmentary registers but they are not sufficient to evaluate fully their numbers in Vidin. The garrison troops included also artillerymen as well as auxiliaries, which I shall not discuss here. (42) Evliya Çelebi, Seyahatname, 6: 97–8. (43) The existence of timars and sipahis in the region even after Evliya’s visit to the region is confirmed by, for example, NBKM, Or. Dept., F. 26, a.u. 1976, of 1678, about a timar in the nahiye of Timok, in the sancak of Vidin. (44) There are occasional references to sipahis and zaims in the sancak also from the very end of the seventeenth century, listed among the local administrators who are warned against interfering with the collection of some taxes (see for example NBKM, Or. Dept., S 13, fol. 13b, doc. I, of 1698). I tend to agree with Hristo Gandev that these incidences should rather be considered a cliché, as part of an already accepted format, without actually meaning that they really existed. See Hristo Gandev, ‘Zarazhdane na kapitalisticheski otnosheniya v chiflishkoto Page 15 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries stopanstvo na Severozapadna Bălgariya prez XVIII vek’, in idem, Problemi na Bălgarskoto Văzrazhdane (Sofia: Nauka i izkustvo, 1976), 287. In the rare cases of identification of individuals as sipahis, there is a very high probability that the persons in question were members of the sipahi regiments of the Janissary corps. See for example NBKM, Or. Dept., S 38, fol. 6b, doc. II: Medine-i Vidin sukkânından olub kalâ kethüdası Osman sipahi… (45) On the later history of the sipahilik, see Dimitrov, ‘Politikata na upravliavashtata vărhushka v Turtsiya’, 32–60, as well as NBKM, Or. Dept., F. 26, a.u. 303, 313, 316, and many others in the same collection about timars in the neighbouring districts of Oriahovo, Lovech, Niğbolu, Tarnovo, Sevlievo, Vratsa and Şehirköy (Pirot, Serbia) from the 1830s and 1840s. (46) One of the interesting peculiarities in this region is the presence of a unit of twenty-six yeniçeri oğlanları in the nahiye of Bane in 1586 who served as martolosan and were exempt from the extraordinary taxes (avariz-i divaniye ve tekalif-i örfiye) and other services. See Berindei, Kalus-Martin, and Veinstein, ‘Actes de Murad III sur la région de Vidin’, 58–60. This, to my knowledge unique, occurrence of employment of ‘young Janissaries’, in such a capacity may be an indication of the sources of conscription for the corps at the time and of the range of their functions. (47) See details in Gradeva, ‘War and peace’, passim. (48) Ahmed Djevad, L’état militaire ottoman depuis la fondation de l’Empire jusqu’à nos jours, trans. George Macridas (Constantinople, 1882), 167. On the expansion of the Janissary corps in Vidin after the fall of Temeşvar in 1716, see Evgeni Radushev, ‘Osmanlı Ordusu ve Balkan Halkları’, in Hans-Georg Majer and Raoul Motika (eds.), Türkische Wirtschafts- und Sozialgeschichte von 1071 bis 1920, Akten des IV. Internationalen Kongresses (Wiesbaden: Harrassowitz, 1995), 275–6, and later, Virginia Aksan, ‘Whose territory and whose peasants? Ottoman boundaries on the Danube in the 1760s’, in Anscombe (ed.), The Ottoman Balkans, 67. (49) In some sources, the standard-bearer. (50) The issue of the voluntary conversion to Islam with the intention of joining the Janissary corps is approached by Anton Minkov, Conversion to Islam in the Balkans. Kisve Bahası Petitions and Ottoman Social Life, 1670–1730 (Leiden: Brill, 2004), 75–7, and Evgeni Radushev, Pomatsite. Hristiyanstvo i isliam v Zapadnite Rodopi s dolinata na r. Mesta, XV-30te godini na XVIII vek (Sofia: NBKM, 2005), 224–30. The latter also discusses the phenomenon of the ‘rural’ Janissaries in the Rhodopes and elsewhere in the Balkans in connection with ‘personal applications’ for conversion. The two authors call this phenomenon ‘voluntary devsirme’ and regard it as one of the reasons for the abrogation of the ‘original’ devşirme in the late seventeenth century, a hypothesis that I do not find very strongly substantiated. (51) For a snapshot of the structure of the military, the militarisation of local society, and the interaction between peaceful professions and military obligations in Vidin at the turn of the eighteenth century see Gradeva, ‘War and peace’, 113–19. (52) ‘Otkăs ot podroben registăr na timari v sandzhak Vidin’, 499–502. See on them Aurel Decei, ‘Akindji’, EI2, 1: 340. For much of the second half of the fifteenth century the sancakbeyi of Vidin was Ali Bey Mihaloğlu, of the family of Gazi Mihal who were hereditary leaders of the akıncıs and led their incursions against Wallachia and Hungary. In 1527 sancakbeyi with a hass worth 400,000 akçe was Yahşi Bey, also of the Mihaloğlus. See Metin Kunt, The Sultan’s Servants: The

Page 16 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries Transformation of Ottoman Provincial Government, 1550–1650 (New York: Columbia University Press, 1983), 105. Members of other akıncı bey families also held the position, a clear indication about Vidin’s importance as one of their bases. On the Mihaloğlus see Franz Babinger, ‘Mikhaloghlu’, EI2, 7: 34. (53) Several of them have Abdullah as a patronym. This is not surprising bearing in mind that non-Muslims could also be enrolled as akıncıs. See Boris Nedkov, Osmanoturska diplomatika i paleografiya, II: Dokumenti i rechnik (Sofia: Nauka i Izkustvo, 1972), 175–7, defter of the akıncıs, of 1472. The order preceding it explicitly states that non-Muslims were also eligible. (54) ‘Otkăs ot podroben registăr na timari v sandzhak Vidin’, 499–502. (55) This is the more or less standard arrangement with the categories of reaya with special obligations including those with military functions. Sometimes they would be exempt also from other services and taxes. For an overview see Elena Grozdanova, ‘Die Privilegien—‘Trojanisches Pferd oder Achillesferse’—als Element der Innenpolitik des Osmanishen Reiches auf dem Balkan während des 15.-18. Jh.’, Bulgarian Historical Review, 22/3–4 (1994), 19–36. (56) Părveva, ‘Bălgari na sluzhba’, 229–31. I have not found the latter in Vidin, at least in the documentation I have worked with, which is rather strange, because they were part of the garrisons in all eighteenth-century Ottoman towns in these parts, including Rusçuk: Gradeva, ‘War and peace’, 114. (57) Părveva, ‘Bălgari na sluzhba’, 231–2. (58) See the observations of Hülya Canbakal on the basis of Ayntab at the end of the seventeenth century, Society and Politics in an Ottoman Town: Ayntab in the Seventeenth Century (Leiden: Brill, 2007), 85–8. (59) NBKM, Or. Dept., S 38, passim. Even earlier this division appears in a kadı sicil from Rusçuk, 1663–4, History Museum-Ruse, B 2919, passim, contemporaneous with another war which ended with the defeat of the Ottomans at the Battle of St Gotthard, but with the preservation of the border status quo at the Vasvár Peace (1664). It had become a standard formula only from the mid-eighteenth century onwards. See on the term and its meanings through the centuries Joseph Kabrda, ‘Raia’, Izvestiya na Istoricheskoto druzhestvo v Sofiia, 14–15 (1937), 176–85; Suraiya Faroqhi, ‘Raiyya’, EI2, 8: 404. (60) See in more detail the survey of the paramilitary and auxiliary institutions on the territory of the sancak in the fifteenth and sixteenth centuries in Kayapınar, ‘Le sancak ottoman de Vidin’, 242–79; eadem, ‘Les filorici dans la région timoko-danubienne à l’époque ottomane (XVe–XVIe siècles)’, in Faruk Bilici, Ionel Cândea and Anca Popescu (eds.), Enjeux politiques, économiques et militaires en Mer Noire (XIVe—XXIe siècles). Études à la mémoire de Mihail Guboglu (Braïla: Éditions Istros, 2007), 243–88, including also a survey of the martolosan and voynuks. (61) Depending on the importance of the place for the maintenance of security and order, sometimes all the villagers would be serving (especially on the banks of the Danube and the Morava rivers and close to mountains), but in others just some of them, while the rest were ordinary reaya. The institution was progressively expanding on the territory of the sancak, including towards the end of the sixteenth century nearly 4,000 households.

Page 17 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries (62) They were employed as a policing force fighting against banditry in the mountainous areas in the nahiyes of Çerna Reka and Bane, while those in the fortresses of Vidin and Fethülislam served on the state boats on the Danube, also ensuring security in strategic places. (63) Voynuks originally (fifteenth-century) had exclusively military functions and participated in campaigns but from the beginning of the sixteenth century became more and more an auxiliary institution, involved, for example, in pasturing the sultan’s horses. In the late fifteenth century there were over 1,500 men registered as voynuks in the sancak of Vidin. (64) Vidin was one of the few places where the müselleman were Christians, serving on boats on the Danube and controlling the river banks. (65) The filurcis, conscripted mainly among the semi-nomad Vlachs, were one of the specific corps in this province and the western parts of the neighbouring Niğbolu sancak. They served as the advance guard, spying and showing the routes to the main body of the army. (66) Bojanić-Lukać, Vidin i Vidinskiyat sandzhak, 55–94. (67) Părveva, ‘Bălgari na sluzhba’, 232–9. For the earlier period see Evgeni Radušev, Ottoman border periphery (serhad) in the Nikopol Vilayet, first half of the sixteenth century’, Études balkaniques, 31 (1995), 140–60. (68) Părveva, ‘Bălgari na sluzhba’, 239–44. (69) NBKM, Or. Dept., Vidin, muhafız, N 2534; Ivanova, ‘Widin’, 206; Mitko Lachev, ‘Kratka istoriya na hrama “Sv. Nikolay Mirlikiyski Chudotvorets” v gr. Vidin’, Duhovna kultura, 70/11 (1990), 20–30. (70) Părveva, ‘Bălgari na sluzhba’, 246–7. See on these developments in general Halil İnalcık, ‘Military and fiscal transformation in the Ottoman Empire, 1600–1700’, Archivum Ottomanicum, 6 (1980), 283–337; and on the settlement of Janissaries in the provinces, Tsvetana Georgieva, Enicharite v bălgarskite zemi (Sofia: Nauka i izkustvo, 1988), 116–72. (71) This is what happened in Ioannina in 1611 when local Christians led by the metropolitan rose in an uprising and massacred their Muslim co-citizens. After its suppression Christians who had been granted special privileges as a prize for their voluntary submission to the Ottomans at the time of the conquest, the right to remain in their residences in the fortified part of the town and exemption from devşirme, lost both. (72) There are many questions around the so-called Chiprovtsi events. See in more detail Yoanna Spisarevska, ‘Chiprovskoto văstanie—mit i realnost’, in 300 godini Chiprovsko văstanie. Prinos kam istoriyata na bălgarite (Sofia: BAN, 1988), 180–203; on contemporaneous developments among the Orthodox see Strašimir Dimitrov, ‘Mouvements de libération en Bulgarie orientale pendant les années 80 du XVIIe siècle’, Études balkaniques, 28 (1992), 235–48. (73) S 14, fol. 4b, doc. I, of 1699. (74) Boris Nedkov, ‘Pogolovniyat danăk v Osmanskata imperiya s ogled na Bălgariya’, Istoricheski Pregled, 1 (1945–6), 25.

Page 18 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries (75) Muslims, as sicil material from the beginning of the eighteenth century shows, had dispersed to more or less distant places—from Bursa to nearby Vratsa. Some of them returned, others chose to sell their property and leave for good. The same had happened to Jews. Rossitsa Gradeva, ‘Jews and Ottoman authority in the Balkans: the cases of Sofia, Vidin and Rusçuk, 15th-17th centuries’, in eadem, Rumeli, 245–6. (76) See numerous references in the following kadı sicil s: NBKM, S 345, of 1696; S 13, of 1698– 99; S 14, of 1699–1702; and especially in S 38, of 1705–13. (77) In 1711 the Ottomans discovered a conspiracy involving Orthodox Christians in several places in northern Bulgaria (but not in or near Vidin) who were in connection with Tsar Peter the Great and planned an uprising in support of his campaign. Rossitsa Gradeva, ‘Villagers in international trade: the case of Chervena Voda, seven-teenth to the beginning of eighteenth century’ Oriente Moderno, 25 n.s. 1 (86) (2006) (Special Issue, The Ottomans and Trade, ed. Ebru Boyar and Kate Fleet), 20. I do not believe that it could have had any contribution to the events after 1718, though. A much more probable direct cause for the demand of the Vidin Janissaries may have been the bandit activities of the Bulgarian haydud Papazoğlu who in 1715 disrupted traffic on the Danube and crossed to the south bank from Wallachia. In 1716/17, according to an order to the overseer (nazır) of Vidin, the three ‘captains’, Georgi, Dimitri and Filimon, crossed the Danube with a numerous band (and their families) and encamped in a place eight to nine hours away from Vidin, between Vidin and Lom, trying to raise the local reaya against the Ottomans. However 200 ‘brave Muslims’ from Vidin attacked and routed them. See Bistra Cvetkova, ‘Un document turc inédit concernant un mouvement de résistance en Bulgarie du Nord-Ouest au XVIIIe siècle’, Rocznik Orientalistyczny, 38 (1976), 96, 99–100. (78) Bojanić-Lukać, Vidin i Vidinskiyat, 13. (79) Strashimir Dimitrov, Văstanieto ot 1850 g. v Bălgariia (Sofia: BAN, 1972), 18. (80) The Mihaloğlus had large estates around Pleven, Ihtiman, and probably also Filibe/Plovdiv, in Bulgaria, as well as around Edirne, Turkey, but none is registered in the Vidin sancak. See Babinger, ‘Mikhal-oghlu’, 34. (81) Gradeva, ‘Administrative system’, 36–7; Kayapınar, ‘Le sancak ottoman de Vidin’, 117–29; NBKM, Or. Dept., S 14, fol. 14b, doc. II. (82) There is conflicting information about Berkovitsa. While the document we have referred to above speaks of the election of the kethüda yeri (their local leader and representative in front of all authorities) by the commanders (aga) of the six detachments of the altı bölüks stationed in the sancak of Vidin, there is evidence from the same year that Berkovitsa was a nahiye in Paşa sancak: Dokumenti za b ălgarskata istoriya, III: Dokumenti iz turskite dărzhavni arhivi, trans. and ed. Pancho Dorev (Sofia: Pridvorna Pechatnitsa, 1940), doc. 73, 72–3, also from 1699. The two documents might be reflecting a transitional period or some confusion in the Ottoman central records. (83) Both districts are attested as part of the sancak until the late eighteenth century. (84) In the documents appear the territorial mukataas of Sahra, Timok, Krivina, Belgrad, Polomiye, Krayna, Archar and others. See for example, NBKM, Or. Dept., S 14, fol. 14b, doc. I, S

Page 19 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries 38, fol. 20a, doc. I; or the mukataas of the revenues from the ports of Vidin and Filurdin, the varos of Vidin, etc., ibid., S 38, fol. 78b, doc. I. See also Radushev, Agrarnite institutsii, 57–102. (85) NBKM, S 345, fol. 4b, doc. I, of 1697. (86) NBKM, S 14, fol. 9b, docs II and III, of 1700. (87) This tendency is not unique to Vidin. In the first decades of the eighteenth century the timar holdings were abolished and added to the sultan’s hasses (for example on the islands of Mytilene and Crete). Even earlier, in the late seventeenth century, timars were transformed into mukataas in the sancaks of Diyarbakır, Tokat, Aleppo, Amasya and others. See Radushev, Agrarnite institutsii, 71–2. Sometimes this was done as the result of yoklama (muster rolls), Dimitrov, ‘Politikata na upravliavashtata vărhushka v Turtsiya’, 37 (in 1715 on the basis of yoklama were confiscated 2,199 timars in Erzurum, of which 1,517 were vacant and 602 belonged to deserters; they were all included in the sultan’s hasses). (88) On the ocaklık as a system of payment and administration see Rhoads Murphey, ‘The Functioning of the Ottoman Army under Murad IV (1623–1639/ 1032–1042): Key to the Understanding of the Relationship between the Center and Periphery in Seventeenth-Century Turkey’ (Unpublished Ph.D. dissertation, Chicago, 1979), 187–202; Michael Hickok, Ottoman Military Administration in Eighteenth-Century Bosnia (Leiden: Brill, 1997), 79–98; Radushev, Agrarnite institutsii, 80–9. On the use of mukataas for direct salary assignments see also Haim Gerber, ‘Mukataa’, EI2, 7: 508. (89) NBKM, Or. dept., S 13, fol. 25a, doc. I, fol. 25b, doc. I, both of 1699; S 14, fol. 36a, doc. I, of 1700; S 13, fol. 21b, doc. II, of 1699. (90) The term originates from the Slavic word gospodar, that is, master. (91) Dimitrov, Văstanieto ot 1850 g, 13–34. See also Gandev, ‘Zarazhdane na kapitalisticheski otnosheniya’, 271–393. Bruce McGowan based his discussion of the developments in the region of Vidin on Gandev’s works: Economic Life in Ottoman Europe: Taxation, Trade, and the Struggle for Land, 1600–1800 (Cambridge and Paris: Cambridge University Press/Éditions de la Maison des Sciences de l’Homme, 1981), passim. (92) Dimitrov, Văstanieto ot 1850 g, 32. (93) While čitlučenje comes from the term çiftlik, in essence it represents a process similar to the above-described emergence of gospodarlık villages in which the tapu holders were usually members of the local military elite. In this regime the peasants rented the land, sometimes even their houses, and paid not only the usual taxes to the state but also a land rent to the holder of the tapu, that is the Janissaries, and had the status of tenants. However, in Serbia there existed another power centre supported by the central authority, the sipahis, Dimitrov, Văstanieto ot 1850 g., 21–2. On Serbia and Bosnia, see also Bruce McGowan, ‘The age of the ayans, 1699– 1812’, in Halil İnalcık with Donald Quataert (eds.), An Economic and Social History of the Ottoman Empire, 1300–1914 (Cambridge: Cambridge University Press, 1994), 685–6; on Kiustendil and on the distinction between gospodarlık and çiftlik lands see Slavka Draganova, Kiustendilskiyat Region, 1864–1919 (Sofia: Akademichno izdatelstvo ‘Marin Drinov’, 1996), 26– 38.

Page 20 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Between Hinterland and Frontier: Ottoman Vidin, Fifteenth to Eighteenth Centuries (94) For an exception of a powerful Christian who had agricultural estates of a similar nature in the nearby Vratsa region see Svetla Ianeva, ‘Dimitraki Tsenov HADZHITOSHEV, in Iliya Todev (ed.), Koi koi e sred bălgarite, XV–XIX v. (Sofia: Anubis, 2000), 281–3. (95) See for example the lists of the ‘agalar villages’, in the nahiye of Polomiye/Lom from around 1835 and prior to 1850 in Simeon Damianov, Lomskiyat kray prez Văzrazhdaneto. Ikonomicheski zhivot i politicheski borbi (Sofia, 1967), 324–38. In many cases there are direct indications that the villages in question had been transformed into çiftliks by or during the rule of Pasvanoğlu.

Access brought to you by:

Page 21 of 21 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Garrison and its Hinterland in the Ottoman East, 1578–1605

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

The Garrison and its Hinterland in the Ottoman East, 1578–1605 RHOADS MURPHEY

DOI:10.5871/bacad/9780197264423.003.0018

Abstract and Keywords This chapter gives an overview and analysis of the consequences and localized effects of the period of endemic war between the Ottomans and Safavids along an extended front in eastern Anatolia over the thirteen years between 1578 and 1590. It assesses how the creation of a new militarised zone where large and very resource-demanding garrisons, constituting a key element of Ottoman military strategy, contributed to the stability or otherwise of the frontier regions in the post-war period. In particular, it evaluates the impact of Ottoman military investment, deployment of resources and local cost-sharing and manpower provision arrangements on social relations, both in cities where the major garrisons were situated and in the surrounding countryside whose resources contributed to their maintenance. Keywords:   endemic war, eastern Anatolia, Ottoman military, Ottoman-Safavid war, military investment

THIS CHAPTER GIVES AN OVERVIEW AND ANALYSIS of the consequences and localised effects of the period of endemic war between the Ottomans and Safavids along an extended front in eastern Anatolia over the thirteen years between 1578 and 1590. It assesses how the creation of a new militarised zone where large and very resource-demanding garrisons, constituting a key element of Ottoman military strategy, contributed to the stability or otherwise of the frontier regions in the post-war period. In particular we will be interested in evaluating the impact of Ottoman military investment, deployment of resources and local cost-sharing and manpower provision arrangements on social relations, both in cities where the major garrisons were situated and in the surrounding countryside whose resources contributed to their maintenance. A key figure in these periods of activation and deactivation, militarisation and attempted demilitarisation in the east was the eunuch, Cafer Pasha, whose two terms as governor-general of the newly created province of Tabriz, the first between 1585 and 1592, immediately after its capture, and the second lasting from 1597 to 1599, account for eleven of the eighteen years of Ottoman military administration in the former Safavid capital. Comparing the shifts in policy and

Page 1 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Garrison and its Hinterland in the Ottoman East, 1578–1605 tactics associated with his two terms as governor, alternating between the application of hard power and recourse to a punitive approach in relations both with the civilians in Tabriz and his own soldiery during the active war phase and experimentation, with various forms of soft power and consensus-building techniques in the post-war period, allows us to bring into clear focus some of the typical dilemmas faced by Ottoman administrators during the transition from war to peace. During wartime, major garrisons on active military fronts always placed colossal demands on local economies, dominating the landscape and distorting normal economic relations, but in the territorially compact provinces of the east such phenomena were intensified and their effects more pervasive and deep-reaching. Before entering into a detailed discussion of conditions along the frontier in the Ottoman east and an account of the life and times of Cafer Pasha, it may help to compare the character and makeup of Ottoman frontiers in other sectors and at other (p.354) times, to assess how frontier formation in the east conformed to or diverged from typical patterns encountered elsewhere in the empire. Ever the pragmatists, the Ottomans allowed form to follow function and there were fundamental reasons why conformity to so-called ‘classical’ modes and administrative models proved unworkable in the east. For one thing, imposing the burden of supporting a locally resident timariot class in areas where the land was less productive was likely to produce not stability, but resistance to Ottoman rule.

Front Formation and Maintaining Fortified Frontiers in the Balkans and North Africa Ottoman expansion in the early part of the sixteenth century on other fronts had been prefaced with propaganda promising an improved standard of living and comfort for the general population after the establishment of Ottoman rule. The vehicle for achieving this, the postconquest cadastral surveys, carefully allocated local resources to support local improvements and were intended also to convey the impression that the Ottoman overlords had no confiscatory intentions, but planned instead to contribute more to the annexed region than the empire would extract by way of treasury revenues. In particular, military subsidies for garrisons designed to support its military presence compensated in part for the fact that it was civilian populations closest to a permanently garrisoned and intermittently active military front that bore the main brunt of the disruption caused by military activity. In the case of the militarised zone created in the middle Danube region after 1540, tax exemptions for those performing auxiliary services in aid of army activity and cash infusions from the treasuries of Egypt and the imperial centre were all part of the package offered to the inhabitants of central Hungary to lighten the burden of Ottoman rule. The Ottomans’ Balkan conquests were typically followed by settlement and the phased implementation of a plan designed to win the active support of the local population in the conquered territories. Ottoman administrators, especially the governors of provinces inside the militarised zone, pursued a strategy that recognised that the long-term viability of Ottoman rule would be contingent on short-term investment in economic recovery after the conquest, paired with delivery on its pre-war promises of perceptible improvements under Ottoman rule. Apart from financial aid provided from inland regions of the imperial heartlands, locals were promised legal protection from excessive claims and exactions on the part of Ottoman administrators and from unruly or oppressive behaviour on the part of other representatives of sultanic authority such as the garrison soldiers stationed in their midst. Likewise, attempts by embezzlers and treasury defrauders to subvert Ottoman administration or divert funds earmarked for border defence were subject to strict monitoring and control by local inspectorates. All these

Page 2 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Garrison and its Hinterland in the Ottoman East, 1578–1605 guarantees, accompanied by powers and procedures for enforcement, hastened the return to normality and stability in the post-conflict phase. (p.355) The more remote the front, the greater the temptations as well as opportunities for usurpation of power and profiteering by government agents and representatives, but while the Bosnian–Croatian–Hungarian frontier was by no means immune from the danger of such abuse, it was unequivocally a better-managed and better-regulated provincial environment than any part of the ‘Wild East’ during the Ottoman–Safavid wars of the late sixteenth century. When abuses occurred in the ‘Settled West’ of the Hungaro-Bosnian frontier in the same period, mechanisms for discovering, reporting and taking action to punish the perpetrators were in place, thus acting as at least a partial deterrent against the commission of the most egregious breaches of the Ottoman social contract. For example, we learn from the social critic Mustafa Ali of the inflated reporting of fortification and garrisoning expenses by a certain Ferhad Bey during his time as lieutenant-governor of Kilis on the contested Bosnian-Croatian frontier, reported as follows: ‘In the castle which he put down for 300 men, there was not even room for 100 and when it was inspected there were never more than 40 or 50.’1 Examples of such attempts to defraud the treasury are abundant in Ottoman documentation, and the practice of using counterfeited pay-tickets (esames) to collect the pay of non-existent soldiers was seemingly widespread, not just in border districts but even in the central payroll registers maintained in Istanbul. The difference between such irregularities as occurred in the Balkans and Istanbul and what occurred in the ‘Wild East’ is that in the west not only were these lapses discovered and reported, but punitive and corrective action could usually be taken, whereas the administrative infrastructure to achieve this regulatory intervention of the state effectively ceased to exist east of the Tigris river and the shores of Lake Van. Another common approach employed by the Ottomans in the creation of a new frontier was the self-financing, self-regulating mode, applied with greatest effect in the maritime sphere and used most consistently in the middle decades of the sixteenth century, when rapid mobilisation of the Ottomans’ North African allies and proxy warriors to confront Spain was achieved by adopting a permissive attitude towards pillaging by the Barbary corsairs, so long as their depredations were aimed against the enemy and a significant proportion of the proceeds were used to sustain the anti-Spanish war effort. This relieved the Ottoman treasury of the need to provide war subsidies or to invest in costly preparations for naval confrontation in the Mediterranean. Furthermore, since the principal source of livelihood of the pirate kingdoms derived from the sea, the disruptive effects and resource demands posed by the presence of fortified coastal cities such as Algiers fell rather lightly on the inhabitants of its continental African hinterlands. The attempt to apply this model of the self-financing and self-sustaining frontier, derived from a maritime context, to a land-based environment associated with the (p.356) activation of a new military frontier in eastern Anatolia in the 1580s and 90s raised a series of thorny issues relating to social conditions and military–civilian relations. These issues remained largely unresolved during the prolonged Ottoman attempt to consolidate rule in Azerbaijan and neighbouring Karabagh over the three decades between 1578 and 1605. The cost of erecting fortifications, bastions and focal points for projecting Ottoman military power across this region was borne during the early stages of the conflict with the Safavids by the Ottoman state with monies injected into the region from outside sources, but over time the crushing financial burden as well as transport- and distance-related problems and supply vulnerabilities gave rise to a variety of cost-sharing and service-sharing arrangements with local populations. Gradually, but perceptibly, the burden of maintaining the Ottoman military presence shifted away from central

Page 3 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Garrison and its Hinterland in the Ottoman East, 1578–1605 provision to localised arrangements conceived, managed and monitored at the local level. Local populations were not only asked to bear the burden of continuing insecurity associated with the prolongation of war, but also faced raised expectations about their share of material support for the war effort itself. Under such conditions, far from being able to trumpet credible claims concerning the advantages of the Pax Ottomanica, the most Ottoman officials in the eastern border provinces could promise was to minimise and contain the worst excesses of the soldiery against the civilian population. In the absence of sufficient and sustained support in the form of wages, and provisions for the troops from the centre, provincial governors were powerless to hold military personnel in absolute check, but at the same time maintaining law and order was the essence of their role and responsibility as governors. Maintaining the ideal balance between permissiveness and discipline was never easy. Restraining soldiers driven by hunger and privation and the psychological extremes associated with combat and military blockade posed the most serious challenge, but military–civilian friction could equally arise under peacetime conditions, when the continued stationing of large garrison forces charged with monitoring the peace imposed particularly onerous burdens on garrison cities positioned at a slight remove from the frontier. To gain a fully rounded picture and reach a balanced assessment of conditions in garrison cities in the Ottoman east in the late sixteenth century, we need to examine a wide sample, including garrison towns in newly occupied territory in Azerbaijan and Shirvan on the one hand, and near-border cities such as Erzurum and towns of the immediate frontier zone such as Kars on the other. Despite its longer association with Ottoman administration dating from the mid-sixteenth-century conquests of Süleyman the Magnificent, Kars experienced a period of renewed uncertainty and instability during the reactivation of the eastern front after 1578. Paradoxically, it was during peacetime after 1590 and in the city of Erzurum, furthest removed from the former military front, that some of the most dramatic recorded incidents of disturbances between garrison forces and civilians took place. The townspeople’s toleration for the demands of the swollen military regiments, stationed near the front so as to be rapidly deployable in case of need, reached its breaking point not (p.357) under active threat of attack but during a lull in fighting. A four-part micro-analysis of the specific developments and localised conditions encountered in the eastern militarised zone during the late sixteenth century will convey a sense of the diverse reactions to the Ottomans’ heightened presence in this least accessible and least developed of all their imperial frontiers.

Page 4 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Garrison and its Hinterland in the Ottoman East, 1578–1605 (p.358) Case One: Shirvan Province

Figure 18.1. The Ottoman–Safavid border in Shirvan and its main garrison town, Eastern Anatolia and the Caucasus c.1612. Derbend, had only one reasonably reliable route of access and supply, and that was via Crimea, the Straits of Kerch and overland across the steppe of Kuban. This access route presumed both stability and a high level of co-operation within the Crimean khanate. In actual fact, it was largely due to raised expectations for Tatar help in opening new Ottoman frontiers in the Caucasus that a relationship of mutual trust and common interest formed during joint operations in Eastern Europe was put under unprecedented strain and fatally compromised. Expecting the Tatars to serve on a second front in the east, where the rigours of campaign were especially severe, the distances daunting2 and prospects for material reward or compensation through war booty slim, was bound to put strains on the Tatar-Ottoman alliance. In the event, one of the first casualties of war was the cooperation and loyalty of the Tatar allies. The 1584 rebellion of the Khan Mehmed Giray II (1577–84) was related to his resistance to Ottoman war demands and its suppression was only achieved by means of an Ottoman armed intervention. The fallout from the loss of unquestioning Tatar support was particularly serious in the case of Shirvan, since unlike offensive campaigning in Eastern Europe where Tatar support was helpful but not essential, maintaining the Ottomans’ military presence in the Caucasian lands was highly dependent on the Crimean access route. The only alternative for supplying Ottoman strongpoints and fortresses in Caucasia, particularly those situated in the territories of the Ottomans’ new-found Georgian allies in the former kingdom of Kakhetia, was Kars, whose position itself was not wholly secure and whose remoteness and separation by difficult terrain and the hostility to Ottoman expansionist aims of the populations in the intervening territories made its use impracticable. Geography, climate and the complex politics of the Armeno-Georgian world all mitigated against Kars fulfilling its intended strategic role and purpose. Because of the insecurity of the overland road links via Ottoman territory and the compromising of the Ottoman–Tatar co-operative relationship in the early years of the war with the Safavids, the Ottomans were placed in the position of having (p. 359) to rely more heavily on their local allies, in particular Alexander II (1574–1605), whom the Ottomans retained in place after the fall of Tiflis, Shemakha and Derbend in the opening phase of the war. From the very beginning of the relationship, the Ottomans required Alexander to pay a high price for retaining his position, demanding what amounted to a tributary payment by an Ottoman vassal of 10,000 gold coins’ worth of raw silk (approximately 1.2 million akçes calculated by 1587 exchange rates) and substantial assistance in defraying the costs of garrisoning the province.3 In the post-war phase after 1590, when the province was incorporated under nominal Ottoman rule, Alexander Khan’s ‘contribution’ to the cost of Ottoman administration was set, according to one source, at 2.5 million akçes.4 Taking into account the double tribute, the second being paid by the last relic of the Shirvanshahid dynasty, Abu Bakr Mirza, who also hoped to cling on to his sovereign status under Ottoman sponsorship, a further two million was to be delivered to the local treasury of Shirvan. The ultimate source of all these contributions was the labour and productivity of the tenants and peasants who paid dearly for their ancestral overlords’ dreams of dynastic renewal.5 A few years later, towards the end of 1598, Selaniki recorded that the Shirvani silk ‘contribution’ (tribute) was raised from 50 pack-loads to 100, its cash value then amounting to roughly 10 million akçes. This was easily

Page 5 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Garrison and its Hinterland in the Ottoman East, 1578–1605 enough to cover the cost of meeting the quarterly wage instalment of the Janissary corps, for which it was earmarked and on which it was expended as soon as it arrived from the east.6 In effect, more blatantly after than during the war, local power brokers co-opted to the Ottoman imperial cause were placed in the position of having to subsidise the cost of their countries’ military occupation by the Ottoman state. The status of the Ottoman ‘province’ of Shirvan after 1590 was ambiguous, sharing many features of the pay-as-you-go self-financing frontier model of the North African sphere. This intermediate, indeterminate status of Shirvan, somewhere in between tributary vassal state and fully incorporated province, relegated the Ottoman garrison cities in that region to (p.360) the position of disconnected islands of power and safe havens whose relationship with their rural hinterlands was based on economic dependency or even exploitation. As a model of Ottoman protectorate, Shirvan province in the brief period of Ottoman direct rule between 1590 and 1605 arguably represented a case where the roles of protector and client were confused. The presence of several large and resource-demanding fortress cities intruding into a postage-stamp-sized territory distorted the economic landscape to a far greater degree than was the norm in areas such as, for example, Ottoman Hungary, where the several largest Ottoman garrisons were dispersed across a wider territory, drew on a wider supply and support radius, and also benefited substantially from subsidies injected into the local economy from outside sources. The confusion of roles in administrative arrangements for Shirvan province is indicated in the fact that beyond providing support to maintain the Ottomans’ military presence, treasury surpluses built up locally were also used to contribute towards the financing of Ottoman military activism elsewhere. It appears that the positive balances built up during Cafer Pasha’s one and a half year governorship of Shirvan in peacetime between 1593 and 1595 were used not to maintain security in the eastern frontier zone, but as a handy cure for the empire’s financial ills arsing from its mobilisation for war on the Hungarian front after 1592.7 Given the openly exploitative character of the relationship between Ottoman governor and local clients and allies, it is perhaps not surprising that the Kakhetian ruler Alexander II seized the opportunity to redefect to the Shah in March 1604, thus bringing a swift end to Ottoman ambitions for permanent rule in the eastern Caucasus.

Case Two: Kars Kars is an example of the Ottomans’ efforts in good faith, starting from the time of the mutual accommodation enshrined in the Treaty of Amasya of 1555, to stabilise the frontier between the Safavid and Ottoman empires. In the first period, pacification of the border region was sought through settlement and post-war economic revival, relying heavily on the fortress and its garrison to provide a zone of safety to lend support to the return of civilian populations and encourage their resumption of production and normal distribution of goods following the disruptions of war. A retrospective account by the Safavid chronicler Iskandar Beg acknowledges Sultan Süleyman’s efforts to transform Kars into a secure bastion for the projection of Ottoman influence and control further east, linked with the transfer of several thousand settlers and an equal number of cavalry guards to patrol the frontier, but notes that Ottoman efforts to (p.361) pacify the frontier using these traditional methods had met with limited success.8 A sultanic letter dating from the decade following the signing of the Amasya Treaty also confirms the scale and seriousness of Ottoman efforts at settlement. At the same time it refers to sporadic attacks, originating from across the border in territory nominally controlled by the Safavid state, targeting Ottoman settlements, even the fortress city of Kars itself.9

Page 6 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Garrison and its Hinterland in the Ottoman East, 1578–1605 One stage for the launching of raids on Ottoman settlements around Kars that receives attention in the contemporary record was the border marches north of Akhaltsikhe (Ott. Ahıska). The border situation remained quite fluid after the Amasya agreements and it cannot be said that either of the would-be Safavid or Ottoman hege-mons really controlled the areas assigned to their respective jurisdictions. One commonly encountered problem was that once the full-scale Ottoman military presence was withdrawn, the administrative sub-structures were too weak to fill the void, resulting in a quick reversion to pre-existing local power structures once the Ottoman armies had retreated.10 Kars still retained its character as a tentative and vulnerable frontier town three decades after Süleyman’s ‘conquest’ of the region when, in 1579, Lala Mustafa oversaw an ambitious expansion of the existing fortifications at Kars. Realistically speaking, his initiatives constituted not a resumption of earlier efforts but a new beginning. The Amasya Treaty had stipulated at the end of the war in 1555 that ‘Kars should be left in ruins and neither side was to occupy this territory.’11 Even though the Ottomans maintained a token presence in the intervening period, the 1579 phase of development represented a starting over from scratch. Contemporary estimates from the time of the high-profile Ottoman military presence in the region in 1578–9 suggest the size of the newly expanded garrison was possibly as few as 300, and certainly not more than 1,000.12 It seems safe to conclude from the reduced scale of permanent Ottoman military presence in this sector of the frontier that the protection of settlers who populated the rich agricultural zone of Çukur-i Saad and the Aras river valley between Kars and Yerevan remained more a theoretical than an achievable objective. Far from being able to fulfil its intended purpose of providing a safety zone for normal activity such as (p.362) agricultural production to support the garrison’s existence, the unsettled conditions throughout the wider war zone in the 1580s and 1590s relegated even the larger Ottoman garrisons to a status of isolated enclaves of soldiers detached from their agricultural hinterlands. The latter’s populations were being forcibly deported by a deliberate Safavid strategy designed to starve the Ottoman garrisons into submission or provoke large-scale desertions that would weaken the Ottomans’ overall military position. When garrison troops were forced by desperation into raiding their immediate perimeters in search of food for themselves and forage for their animals, this placed them in further difficulties, not just from the standpoint of maintaining their military position, but also in terms of winning the peace and establishing a harmonious working relationship with the civilian populations settled in the surrounding countryside. The larger the garrison, the greater the potential pressure on military–civilian relations in the event of disruptions of normal supply. Finding the right balance between optimal size and strength of military garrisons and the capacity of the towns and their hinterlands to support the border garrisons and—especially during active campaigning—to play host to much larger numbers of armed troops in their midst was a challenge faced by the Ottomans in all sectors of the frontier. There was seemingly no magic formula for achieving uniform success in meeting this challenge. Peacetime conditions where contingency planning could be more precise were seemingly no easier to manage than the fluid and unpredictable conditions associated with active campaigning, when whole sectors of the wider border region became destabilised. The wider and longer-term impact of the general military build-up in the east in the period 1578–90 can be judged from the well-documented case of Erzurum in the mid-1590s. The temporary lull in fighting on the eastern front between 1590 and 1603 witnessed no resumption of harmonious

Page 7 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Garrison and its Hinterland in the Ottoman East, 1578–1605 social relations or return to normal patterns of exchange and interaction, but a serious deterioration in the relationship between townsfolk and garrison forces.

Case Three: Military-Civilian Conflict in Peacetime Erzurum as Judged by the Events of 1592–4 In Ottoman times Erzurum was, as in many ways it still is today, a border town and customs or toll station on the near approaches to the Iranian frontier, yet situated far enough inland to be protected by both geography and the strong presence of state and provincial administrative structures from the spill-over effects of the disorder and attenuated control prevailing in some other sectors of the non-linear, non-continuous, non-uniform and largely notional political frontier dividing the two polities. As the final staging point for Ottoman campaigns, it lent both financial and administrative support to armies operating in this theatre, but the prolongation of the Ottoman–Safavid (p.363) wars after 1578 gave it enhanced stature as a regional base for military operations. For the local populace this meant quartering of large numbers of regular army troops in the winter seasons between campaigns as well as a swollen permanent garrison to meet the needs not just of border patrol and control in peacetime, but potentially also resumption of active offensive campaigning on short notice. As a result of these needs, Erzurum’s military garrison gradually acquired a dominant position in local society, which produced resentments and jealousies most keenly felt in peacetime, when locals were hoping for a relaxation of wartime restrictions and belt-tightening, and found instead new demands placed on them by military requisitioning of supplies. The requisitioning caused marketplace disruptions, localised shortages or even scarcity and price inflation, especially for basic goods and necessities of life sought by average consumers. The breakdown of civilian–military relations in Erzurum during the mid-1590s was related to this very issue of skewed distribution and fierce competition for basic goods and the feeling among the local populace that the military allocated to itself a disproportionate share of available goods without much concern for the general good or respect for principles of fair distribution and equal access enshrined in Ottoman law and market inspection protocols. The events in Ezrurum are particularly richly documented in contemporary narrative sources, because the failure of the government to resolve the running dispute between townspeople and garrison forces led to the resignation of the grand vizier and the downfall of his administration on two separate occasions. The situation as recorded by the historian Selaniki gives voice to the Erzurumites’ side of the dispute as follows: ‘The Janissary mob, having taken up residence and set down roots, now assert the right to interfere and obstruct us in the carrying out of our everyday business, while allocating for their own exclusive use all the food supplies reaching Erzurum from outside the region and allowing us no share.’13 When the civilians’ patience with the Janissaries’ arrogance and the excessiveness of their demands reached breaking point, insults and even physical blows began to be exchanged. When the Janissaries on their part complained to the authorities in Istanbul of the insolence and insubordination of those they judged to be rogue elements and habitual disturbers of public order and called for general sanctions against the ‘evil elements of provincial society’ (vilayet eşirrası), the authorities were left with little option but to side with the military as law enforcers. Hence they authorised the arrest and punishment of the ringleaders of the civilian protesters, whose demonstrations, by taking a violent course, had invalidated their otherwise valid cause. Even these gestures of support were, however, regarded by the Janissaries as inadequate compensation and revenge for the loss and personal humiliation suffered by their Janissary comrades (p.364) stationed in Erzurum, so in the end the government was forced to dispatch a

Page 8 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Garrison and its Hinterland in the Ottoman East, 1578–1605 special prosecutor (yasakçı), who carried out his own independent investigations. These concluded with fourteen executions and the award of substantial sums in monetary compensation.14 Erzurum was by no means unique in witnessing the destabilising effects of heightened military demands on inland districts far removed from the war zone proper. These heightened demands could take the form of material provisions for the military, but sharing the financial burden of new garrisons established at the front line of Ottoman expansion into a new theatre of operations also became routine. The demands imposed by the need to man a large new garrison at Tiflis after 1579 stretched far into the Ottoman heartlands in Anatolia. According to a document preserved in Berlin datable to the later part of Murad III’s reign, upwards of two million akçes was set aside from the revenues of Tokat treasury as their contribution towards meeting the wage bill of the Tiflis garrison. This contribution amounted to approximately one half of all allocations for ‘local’ administrative expenses and special grants totalling four million akçes per annum in that particular district. While this subsidy for Tiflis was sufficient to cover only approximately 20 per cent of a wage bill that would have amounted to approximately 10 million akçes per annum for a garrison numbering a notional 6,000 Janissaries paid a notional daily wage of five akçes (over four pay periods totalling 340 days), its impact on the local finances of the interior and largely peaceable province of the vilayet-i Rum was significant.15

Case Four: An Account of the Change in Community Attitudes and Evolution of Military–Civilian Relations in Tabriz During Times of War (1585–90) and Times of Peace (1590–1603) If civilian–military relations could deteriorate to the degree witnessed in an inland garrison town such as Erzurum during conditions of peace and deactivation of the military frontier, it is not difficult to imagine how competition for basic supplies might affect a war-torn city such as Tabriz, where the demands placed on available resources were infinitely more extreme. The eighteen-year period of Ottoman rule and military occupation of Tabriz, between 1585 and 1603, was characterised by intense periods of siege, counter-siege and blockade. These were followed by tentative and only partially successful attempts at extending Ottoman jurisdiction into the countryside, aimed at restoring some semblance of normality to economic and social relations in peacetime. A brief survey of the ups and downs of Tabriz during war and peace completes our (p.365) four-part glimpse at the character and quality of civilian–military relations in Ottoman garrison towns of the east at the close of the sixteenth century. Tabriz offers a particularly good opportunity for comparative analysis, since, despite the brevity of Ottoman rule, the relevant documentation is relatively rich. To provide the proper context for evaluating Ottoman decision making and administrative choices in Azerbaijan requires that we examine in some detail Ottoman governance during both war and peace. Phase One: Winning the War on the Azeri Front, 1585–90 The Ottoman–Safavid confrontation in Azerbaijan that came to a head in late September 158features associated with Ottoman army deploy features associated with Ottoman army deployments in the eastern frontier zone. One tactic, the strategic retreat, was, at least initially, clearly employed by the Safavids, since what the Ottoman forces encountered on their arrival in the environs of Tabriz was a city defended not by soldiers, but by what one historian characterised as ‘lawless elements and street beggars’.16 The disproportion and incongruity in scale and style between the two opposing forces presented a number of operational challenges to the technically ‘superior’ Ottoman military. The first and most pressing of these challenges was the task of provisioning the full membership of the Ottoman army and their mounts, with

Page 9 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Garrison and its Hinterland in the Ottoman East, 1578–1605 the inexorable approach of the cold season marking the end of easily accessible grazing lands for the cavalry troops’ mounts.17 Several decades earlier, in midsummer 1548, Süleyman had been forced to abandon his occupation of Tabriz after a stay of only three days when faced by a similar dilemma of inadequate provisions for his troops and their mounts.18 It seems perfectly clear that such constraints, influenced particularly by the demands of quartering the whole Ottoman army in Tabriz and its immediate environs during the period beginning in the last week of September 1585, played a significant role in the breakdown in trust and co-operation between members of the Ottoman army and the civilian population of Tabriz, sparking the general massacre which occurred on 16 October, three weeks after the Ottoman occupation of the city on 25 September. This conflict was a primitive battle over access to scarce provisions (p.366) and other basic necessities of life. Asafi’s estimate of the civilian death toll from this single day of violence was between 5,000 and 6,000 souls.19 It is generally confirmed that until this time the Ottoman forces had acted with considerable restraint in the face of provocations, harassment and sustained guerrilla attacks against their ranks. However, just as every soldier has his individual physical as well as psychological breaking point, soldiers in their collective ranks share a basic instinct for survival that comes into play under conditions of exceptional pressure or deprivation. It seems that this threshold point between rational and instinctual behaviour had been reached in the Tabriz campaign of 1585 by mid-October. The timing of the general army retreat on 28 October, less than two weeks after the events of 16 October, seems to confirm that the Ottoman command, led by the veteran campaigner Özdemiroğlu Osman Pasha, had finally come to the belated but inevitable conclusion that both the scale and prolongation of the Ottomans’ military presence in the region had become not just counterproductive, but itself a significant contributor to the deteriorating security conditions in the capital city of their latest ‘conquest’. The high-profile Ottoman military presence in the city, even though its purpose was to complete the repairs to the city’s defences and set the stage for further urban reconstruction, was creating a new set of problems that only its immediate removal could resolve. It seems clear that when the main body of the Ottoman army did eventually begin its retreat from Tabriz in late October 1585, the size of the garrison left to defend the city was relatively modest, consisting according to Asafi’s account of no more than 2,000–3,000 garrison soldiers.20 As the first Ottoman governor of the newly created Ottoman province, Cafer Pasha found that one of his most persistent and vexatious problems in the period up to the signing of peace in the spring of 159021 was calculating the optimal and at the same time sustainable size for the Tabriz city garrison. Conditions during the Safavid counter-siege in the eleven months between November 1585 and October 1586 necessitated the doubling of Osman Pasha’s modestly proportioned garrison of 3,000 men to a force of about 6,000 men,22 but concentration of such numbers in single confined locations was not sustainable for long. Evidence suggests that the garrison size was again reduced, by natural attrition, desertion and deliberate redeployment to other smaller fortresses belonging to Tabriz’s outlying defensive perimeter, such that its strength during the winter of 1587–8 had been reduced to no (p.367) more than 1,500 men.23 Without significant outside financial as well as material support from other Ottoman provinces in the secured zones of inner Anatolia, these numbers were about all that could be mustered from local sources, and when the Safavid counter-siege was reactivated after the accession of Shah ‘Abbas in December 1587, peace was still a good two and a half years off in the distant and unpredictable future.24 After the establishment of peace in 1590, questions about the cost, proportions and means of supporting the Tabriz garrison continued to preoccupy the province’s peacetime governors. The lowering of the garrison’s once dominant profile in the province’s Page 10 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Garrison and its Hinterland in the Ottoman East, 1578–1605 finances allowed far greater scope for Ottoman attempts in the traditional mould aimed at establishing a better rapport and basis for reconciliation with the local populace. This reconciliation was based on the offer of inducements, tax modifications and exemptions and other forms of power-sharing explicitly designed to win their support for Ottoman rule. Phase Two: Establishing Peace in Ottoman Azerbaijan, 1590–1603 During the immediate post-conflict era of the 1590s, Ottoman governors in Tabriz waged a battle for the Tabrizis’ loyalty and support against an equally sustained and determined campaign of persuasion mounted by the newly enthroned ruler Shah ‘Abbas. In this period, the social elites of Tabriz were caught in a two-way bidding war, with concessions offered by their present and former overlords, the Ottomans and the Safavids, both competing to win their affections. The operation of this peacetime dynamic is abundantly clear from documentary evidence, which fortuitously survives from the period of Cafer Pasha’s second governorate. It appears that Cafer Pasha was selected in July 1597 by the sultan to undertake this delicate mission of winning the peace, based on his previous experience and success as a wartime administrator, and as a response to the enthusiastic welcome and vote of confidence expressed by the residents of Tabriz themselves. According to the historian Selaniki, far from resisting the return of the administrator associated with the brutal conditions of the first phase of Ottoman military occupation, they praised him for his ‘even-handedness and sense of justice’.25 They furthermore pledged to provide an annual payment of 3 million akçes from provincial revenues after the costs of local administration and defence had been met.26 (p.368) A document composed in early May 1598 preserves in considerable detail the actual character of the contract drawn up between the governor and the governed in Tabriz, by which both sides agreed to mutually acknowledged terms and conditions to cover a three-year term from May 1598 to May 1601.27 By the terms of this agreement, the Tabrizis committed themselves to increasing the annual treasury contribution to 5 million akçes on condition that the government agreed to continue throughout the full extent of the agreement the subsidies from the treasuries of Van and Diyarbakır for garrisoning Tabriz.28 Other terms required the Ottoman authorities to abolish fees collected for irrigation rights to orchards and gardens, whose assessment, at a time when all the orchards of Tabriz had been chopped down to supply timber for reconstruction projects during the first decade of Ottoman occupation, was an obvious absurdity.29 In recognition of the liberal terms of the Ottomans’ plan for reconciliation, the residents on their part committed themselves to increasing the city’s (province’s) contribution from 5 to 6 million akçes annually if conditions at the end of the three-year contract indicated a steady progression towards the goal of repopulating and revitalising the city. Part of the plan for return to normality required Cafer Pasha to use his powers of enforcement to rid the city, by means of forceful deportation, of seventy-nine named individuals identified as disruptive elements.30 However, the greater part of the document (terms and conditions nos. 13– 31) is devoted to recording Cafer Pasha’s concessions to members of the social elite in Tabriz and the surrounding districts, divided into the five categories of bölük, çavuşluk-i dergah-i ali, muafiyat, zeamet and timar. These corresponded to appointment to low-level military staff positions as garrison forces; appointment to high-ranking imperial service as messengers, gobetweens and diplomatic envoys; tax-exemptions for agricultural land, especially gardens and orchards (bag); and high-income (zeamet) and medium-income (timar) land assignments. They were given in exchange for provision of key services in military campaigning and provincial defence.

Page 11 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Garrison and its Hinterland in the Ottoman East, 1578–1605 The Ottomans were no strangers to the subtle art of acquiring the services of defectors from the Safavid cause among Kurdish tribes of the border regions, as both earlier and later examples amply demonstrate.31 However, the framework for mutual (p.369) co-operation reached between Cafer Pasha and the residents of Tabriz in 1598, based on shared obligations and a real division and distribution of power, was a relatively novel concept. It derived from a new understanding of the character of administration appropriate to the Ottomans’ newest territorial acquisitions in the east, where natural resources were scarce and the goodwill of the governed arguably their most valuable asset. While the sudden end of Ottoman rule a short time later in October 1603 makes the detailed assessment of the restorative potential of Cafer Pasha’s policies impossible to assess, it seems fair to suggest that, if the peace had been preserved for any extended period of time beyond 1603, the pasha would be seen to have identified, after much trial and error and experimentation with unproductive policies based on military force and suppression, the formula for success.

Conclusion and General Findings The question which inevitably arises from any investigation of Ottoman frontier formation in the east is why the Ottoman venture in imperial expansion in Caucasia and Azerbaijan was so shortlived and yielded such limited results compared with their expansionist ventures in the Balkan and Mediterranean spheres. By way of partial explanation, it may be said that in relation to frontier design and planning projections there was a tendency in the east to underestimate the effects of distance, terrain and cost of military operations compared with the west, where transport, terrain and distance constituted more known quantities. On the other hand, they tended to overestimate the level of requirements for military personnel and fortress staffing, forgetting the lessons imparted by the Crimean Khan Sahib Giray, who had offered advice to Süleyman at the time of his eastern campaign of 1547. Sahib Giray had insisted that in the east it was not large armies, which were cumbersome, difficult to deploy and expensive to maintain, but small, flexible, highly mobile and lightly armed and equipped units, which could be rapidly deployed and retreat equally rapidly, that provided the best way of combating the Safavid army, which itself employed a similar strategy with considerable success.32 Planning provision for these small units imposed a fraction of the burden on local supply compared to that incurred by the prolonged presence of large armies and the establishment of large garrisons. Furthermore, by lightening the burden on local populations, the Ottomans removed a source of irritation that lent credence to anti-Ottoman resistance. (p.370) The understanding that in the east bigger was not necessarily better perhaps dawned on the Ottomans too late to reverse the negative course that their experiment in eastward expansion began to take after a decade of trial in the 1580s. The creation of a lean, mean fighting machine, proportioned to suit the particular conditions of the sparsely populated eastern frontier regions of the empire and modelled on the lines suggested in Sahib Giray’s advice to Süleyman in the 1540s, never took place. Neither was the model of provincial administration proposed by Cafer Pasha in his second governorate after 1597, with its lightertouch approach to civilian–military relations, given a fair chance to succeed. With the resumption of military conflict in 1603 and its near-continuous prolongation until the signing of the Treaty of Kasr-ı Şirin in 1639, the Ottomans and Safavids entered into the last and arguably most destructive phase of their troubled relations. The consequences for populations of the borderland districts were unmatched in their intensity and environmental impact by any other Ottoman war zone, whether in the Balkans, trans-Danubia or the North African sphere of operations. By 1638 the zone of conflict in the east, especially the Azeri corridor linking several

Page 12 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Garrison and its Hinterland in the Ottoman East, 1578–1605 potential fronts in the wider conflict zone between Caucasia and Iraq, had experienced the effects of active mobilisation and engagement of Ottoman and Safavid armies for forty of the sixty years between 1578 and 1638. While it is pointless to indulge in speculation about the ‘what ifs’ of history, it is intriguing to wonder what might have become of both empires had they been spared the disruption and displacement, to say nothing of the mutual destruction and waste suffered by both sides, during this period of prolonged conflict. That the eventual demise of the Safavid Empire in 1722 was hastened by the need to divert resources to the defence against Ottoman aggression or threat cannot be questioned, but what of the wider effects on the border populations of both empires of the prolonged instability and uncertainty in political relations between the two dynasties? This is difficult if not impossible to assess, but it seems undeniable that both empires suffered the consequences, in both the shorter and the longer term, of their respective miscalculations and overestimations of the need for, and underestimations of the true costs of, the chronic intermittent activation of military fronts in the period between 1578 and 1638. Notes: Proceedings of the British Academy 156, 353–370. © The British Academy 2009. (1) Mustafa Ali’s Counsel for Sultans of 1581, ed. and trans. Andreas Tietze (Vienna: VÖAW, 1979–82), 1: 62. (2) The one-way transit time of eighty days from Bağçesaray to Derbend is confirmed in Bekir Kütükoğlu, Osmanlı-İran Siyasi Münasebetleri, 1578–1590 (Istanbul: Istanbul Üniversitesi Edebiyat Fakültesi Matbaası, 1962), 124, n. 202. (3) Details of these semi-tributary relations are provided in Selaniki Mustafa Efendi, Tarih-i Selaniki, ed. Mehmed İpşirli (Istanbul: İstanbul Üniversitesi Edebiyat Fakültesi Basımevi, 1989), 1: 193. (4) The 2.5 million figure results from the cash equivalent of fifty camel-loads of silk valued at 50,000 akçes per load; see Selaniki, Tarih, 1: 375, entry dated July 1594. If we accept the average pack-load weighed 160 okkas or approximately 205 kg (see Halil İnalcık, ‘Yük (himl) in Ottoman silk trade, mining and agriculture’, Turcica, 16 (1984), 132) the price (value) per okka (1.2828 kg) works out at about 312 akçes. For general price instability reflected in the dramatic fluctuations in silk prices experienced at this time, see further at n. 6 below. (5) On Abu Bakr Mirza’s contribution, see Gelibolulu Mustafa Ali ve Kunh’ül-ahbarında II. Selim, III. Mur ad ve III. Mehmet Devirleri, ed. Faris Çerçi (Kayseri: Erciyes Üniversitesi Matbaası, 2000), 3: 529 and on his line of descent from the Shirvanshahid ruler Shah Rukh (d. 1535), see Cihangir Zeyneloğlu, Şirvanşahlar Yurdu (Istanbul: Cumhuriyet Kitaphanesi, 1931), 154 and Clifford Edmund Bosworth, The New Islamic Dynasties (Edinburgh: Edinburgh University Press, 2004), 141. (6) Selaniki, Tarih, 2: 793 (entry dated January 1599) records that the silk prices in Istanbul at this time varied between 600 and 800 akçes per okka. Thus the standard pack-load of 160 okkas was now valued not at 50,000 akçes, as previously (see n. 4 above), but roughly double that amount. (7) The entry dated July 1594 (Selaniki, Tarih, 1: 375) includes the expression ihracat-i sefer derdine külli deva oldu. Page 13 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Garrison and its Hinterland in the Ottoman East, 1578–1605 (8) Iskandar Beg Turkman, History of Shah Abbas the Great (Tarikh-e Alamara-ye Abbasi), ed. and trans. Roger M. Savory, (Boulder, Co.: Westview Press, 1978), 1: 119–20, specifies that 4,000–5,000 settlers were transferred from the interior regions of Anatolia and they were supported by an equal number of mounted troops. (9) See Ahmed Feridun, Münşeat-al-selatin (Istanbul: Dar al-Tibaat al-Amire, 1274–5), 1: 626: Kars vilayeti kabl al-sulh defaatla huddam-i vala-makamımdan bazıya sancak tarikiyle sadaka olup, ve ona tabi nice kura ve mezari dahl-i defter-i hümayun olmağın…Içinde evler ve damlar yapıp karar ve aram eylemişler iken… (10) The source of the raids as well as opposition to the Ottoman incorporative strategy is identified in the source as taife-yi haife-yi Gürcüden Varaza nam lain. See Feridun, Münşeat alselatin, 1: 625. (11) Iskandar Beg, History of Shah Abbas, 2: 933. (12) The lower estimate is offered by Iskandar Beg Turkman (History of Shah Abbas, 2:933) and the higher one by Mustafa Ali (Kunh’ül-ahbar, 2: 328). (13) Selaniki, Tarih, 1: 263: Yeniçeri taifesi yurt tutup mekan edinmekle, kar u kisbimize mani olup, ve taşradan gelen mekulat’a bizi dahl ettirmeyip… (14) Selaniki, Tarih, 1: 282. (15) The document, a tax-farm contract for the province of Rum, is preserved in the Preussischer Kulturbesitz (Berlin), MS Or. Oct. 132, among the unpaginated documents at the end of the volume. (16) Ibrahim Peçevi, Tarih (Istanbul: Matbaa-i Amire, 1283), 2: 98, eşkiya-i Tabriz ve tulengiyan. (17) Asafi’s account, Asafı Dal Mehmed Çelebi Şecaatname: Özdemiroğlu Osman Paşa’nın Şark Seferleri, ed. Abdülkadir Özcan, (Istanbul: Çamlıca Basım ve Yayın, 2006), 556, credits the leaders among the Ottoman sipahis with the following insight: kalmadı etrafda dahi alef / atlar oldu telef (‘There remained thereabouts not a blade of grass for consumption / leaving all the army’s mounts to confront certain destruction’). (18) The accounts by Iskandar Beg Turkman, History of Shah Abbas, 1: 119, and the Ottoman historian Lutfi Pasha, Tevarih-i Al-i Osman (Istanbul: Matbaa-i Amire, 1341), 348, both confirm the decisive role played by critical lack of provisions, especially adequate grazing land for the army mounts in forcing a sudden change in Süleyman’s campaign strategy. (19) Asafi, Şecaatname, 541. (20) Ibid., 558. (21) Although the date traditionally associated with the commencement of the period of peace is the Ottoman–Safavid agreement signed in Istanbul on 22 March 1590 (see Kütükoğlu, Siyasi Münasebetleri, 195), the final disposition of the frontier and full-scale demobilisation came only eighteen months later in September 1591 (Selaniki, Tarih, 1: 253). (22) Kütükoğlu, Siyasi Münasebetleri, 167, n. 70.

Page 14 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Garrison and its Hinterland in the Ottoman East, 1578–1605 (23) The figure is supplied in a document dating from January 1588 from BOA, Mühimme Defteri 63, duplicated in docs. 25 and 63, which provides the following explanation: Tebrizde olan askeri ekseri Tebriz etrafında feth olunan yerlere konulmağla, Tebrizde ancak 1,500 mikdarı adam kalmıştır. (24) See note 21 above. (25) Selaniki, Tarih, 2: 703, Cafer Paşa’nın adl u dadından her vechiyle hoşnud ve şukran üzereyiz. Bize onu hakim eylen. (26) The 3 million akçe figure is supplied by Selaniki (Tarih, 2: 703) in an entry dated 14 August 1597, coinciding with Cafer’s arrival in Tabriz after an absence of five years. (27) The document in question, BOA, Maliyeden Müdevver (MAD) 7348, 20–1, specified thirtytwo separate terms and conditions (TC) on whose fulfilment the agreement’s validity was made contingent. (28) BOA, MAD 7348, 20 (TC-10). (29) On the logic supporting a suspension of the mirabiyye dues until a later stage of the postconflict economic recovery in Tabriz, see BOA, MAD 7348, 20 (TC-11). (30) On the forceful removal of these eşkiya-i memleket, see BOA, MAD 7348, 20 (TC-3). (31) For the distribution of favours to reward the defection of the hereditary chief of Baneh Süleyman Bey, whose tribal lands straddled the frontier between Ottoman Şehrizor and Safavid Ardelan, in the opening phase of the 1587 campaign, see BOA, MAD 675, p. 71 which records the distribution of thirty-four timars and twenty-three appointments to regimental rank to the followers and relations (akraba) of Süleyman, and for a later phase of concessionary grants targeting tribal groups of the borderlands after 1603, see Rhoads Murphey, ‘Resumption of Ottoman–Safavid border conflict, 1603–1638: effects of border destabilization on the evolution of state-tribe relations’, in Stefan Leder and Bernard Streck (eds.), Shifts and Drifts in NomadSedentary Relations (Wiesbaden: Ludwig Reichert Verlag, 2005), 307–23. (32) Anon., Tarih-i Sahip Giray Han: Histoire de Sahib Giray, Khan de Crimée de 1532 à 1551, ed. and trans. Özalp Gökbilgin (Ankara: Baylan Matbaası, 1973), 114 (fol. 62a) and the French translation, 246–7.

Access brought to you by:

Page 15 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Archaeology of the Middle Nile Valley in the Sudan

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Ottoman Archaeology of the Middle Nile Valley in the Sudan INTISAR ELZEIN

DOI:10.5871/bacad/9780197264423.003.0019

Abstract and Keywords This chapter aims to provide an outline of the archaeological remains reflecting the Ottoman presence on the Middle Nile, with preliminary interpretation and suggestions for areas in which future research could most profitably concentrate. The Nubian frontier region of the Ottoman Empire is one of its least-known areas. It raises numerous questions relating to both Sudanese and Ottoman history, as well as the nature of relations between the Ottomans and the Funj, in which the Ottoman garrisons on the Middle Nile must have played an intermediary role. While Sai is traditionally thought of as marking the southernmost extremity of Ottoman rule, the evidence of some of the sites above suggests that Ottoman rule extended further south, towards Hannek. Keywords:   Sudan, Berberistan, Ottoman Empire, middle Nile, Nubian frontier, Sai garrison

THE TURKISH CONNECTION WITH THE SUDAN has generally been associated with the empire established by the Ottoman Viceroy of Egypt, Mehmed Ali Pasha, in 1820, known as the Turkiyya.1 Monuments from this period survive in several places in the Sudan, built in a distinctive Turko-Egyptian style. The Nile town of Berber, for example, contains remains of administrative buildings, military towers, a police station and an indigo factory. There are also examples of architecture in the Turkish style in the town of al-Khandaq, notably its mosque.2 In Khartoum, the most conspicuous feature is the domed tombs of certain officials (Figure 19.1). Located in the centre of the city on Baladia Street (formerly ‘Abbas Avenue), the qubbas in which rest the remains of two Circassian governors-general of the Sudan, Ahmad Pasha Abu Adhan (1839–43) and Musa Pasha Hamdi (1862–65) testify to the Sudan’s connections with the wider Ottoman Empire.3 Furthermore, even remote regions such as Darfur maintained links with the Ottomans in the nineteenth and early twentieth centuries. Indeed, during the First World War, the last sultan of Darfur, ‘Ali Dinar, was urged by the Young Turk government in Istanbul to join their jihad against the British.4

Page 1 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Archaeology of the Middle Nile Valley in the Sudan Otherwise, the main Ottoman remains in the region are at Suakin on the Sudan’s Red Sea coast,5 and in Lower Nubia, at the fort of Qasr Ibrim in Egypt.6 In contrast, (p.372) the earlier Ottoman presence on the Nile south of the Second Cataract is much less known. However, recent research by John Alexander has shed new light on this, and his studies have emphasised the range of remains relating to this period.7 He has confirmed an Ottoman imperial presence along the Middle Nile Valley from the sixteenth century, and overturned the previously held associations of the Ottoman arrival in the Sudan with Mehmed Ali’s conquests. Over the mid to late sixteenth century, the frontier was gradually pushed south to the Third Cataract as a result of conflict between the Ottomans and the Funj rulers of Sinnar. A new sancak of the Mahas (the area around the Third Cataract) was established, which lasted for a single year, and led to the foundation of the fortress of Sai (Figure 19.2).8 (p.373) This chapter aims to provide an outline of the archaeological remains reflecting this Ottoman presence on the Middle Nile, with preliminary interpretation and suggestions for areas in which future research could most profitably concentrate.

The Ottomans in the Middle Nile Valley The Ottoman occupation of Nubia coincided with the florescence of the Funj Sultanate, which covered much of the present-day Sudan south of the Third Cataract region.9 Unfortunately, literary sources for their advance are few and far between, the region generally being beyond the notice of Ottoman chroniclers. For a later period, the main evidence is supplied by the account of the Ottoman traveller Evliya Çelebi, who (p.374) purports to have visited the region around 1671–2, although whether he did or

Figure 19.1. Tombs of the nineteenthcentury Circassian governors of the Sudan, Khartoum.

not is a matter of some debate.10 It is supplemented by an Ottoman map of the Nile made around 1685, probably by a member of Evliya’s suite who accompanied him on his journey11 Even the archival evidence is slight, at least in terms of what has come to light hitherto. The Funj historiographical tradition is equally slender, and, as is often the case in Sudanese history, scholars are obliged to

Figure 19.2. Map of the Mahas region.

rely to some extent on oral traditions, recorded much later.12 The account of John-Lewis Burckhardt, who travelled in the region in the early nineteenth century, also provides valuable evidence,13 as do the Arabic and Ottoman documents from Qasr Ibrim relating to the garrison’s life.

Page 2 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Archaeology of the Middle Nile Valley in the Sudan The Nile Valley between the First and Third Cataracts was known to the Ottomans as Berberistan. According to most scholars, Ottoman authority in Berberistan was established in the mid-sixteenth century by Özdemir Pasha, the former beylerbeyi of Yemen. The frontier advanced as far as the First Cataract near Aswan around 1555, and by 1570 had reached Ibrim in Lower Nubia; by these means the sancak of Ibrim was created. In 1582 or 1583 an inconclusive battle took place between the Funj and the Ottomans at Hannek by the Third Cataract, which seems to have represented the furthest point of Ottoman expansion along the Nile.14 The Ottomans certainly encountered considerable resistance to their attempts to dominate Berberistan. The ‘Abdallab, Funj vassals who ruled the northern part of the kingdom, had a tradition recording the defeat of an Ottoman force near ‘the Egyptian border’. Another tradition records the crushing defeat of Funj forces by a ghuzz governor named Ibn Janbalan at Hannek (see below on the term ghuzz). Holt has suggested that Ibn Janbalan may be identified with Sulayman Janbulad, a mamluk bey who around 1620 became the first governor of Upper Egypt and hence the immediate overlord of the kaşif of al-Dirr, the main civilian revenue-collector.15 If so, this would suggest that interest in Berberistan, if not on the part of the Sublime Porte, then at least of the Ottoman administration in Egypt, lasted rather longer than the initial phase of conquest, about which we are better informed. In the sixteenth century, garrisons of Janissaries (traditionally thought to be Bosnians) were established at Aswan, Ibrim and Sai. Revenue collection, however, was (p.375) entrusted to a civilian based in the provincial centre of al-Diwan near al-Dirr,16 known as the kaşif— representing an interesting continuity with Mamluk administrative practice, in which the same title was used. The office of kaşif became hereditary and the elite who held the position formed a group that were known as ghuzz, a term which originally indicated Turkish ethnicity, but in Egypt had been used since the Middle Ages for mamluks irrespective of origin.17 As also happened at Massawa,18 the ‘Bosnian’ troops intermarried locally and their descendants formed a privileged class known as kalecis (men of the fortress) or Osmanlı s, ruled by their own ağas. According to Burckhardt, the Ottoman military were independent of the kaşifs.19 However, like the kaşifs, the Ottoman garrisons were probably subject to only limited regulation by the authorities of Ottoman Egypt, being left instead to maintain themselves by means of intermarriage with the local population. Given the failure of the Ottoman state to pay salaries to such provincial garrisons from the end of the sixteenth century, it is not surprising that, as happened elsewhere, the soldiers were obliged to make a living by other means, as is attested by evidence from Qasr Ibrim.20 The forces stationed in the Nubian garrisons seem to have included Bosnians, Hungarians, Albanians, Turks and Circassians. Additional private forces were maintained by the kaşifs.21

Berberistan: The Funj-Ottoman Frontier Region according to Archaeological Evidence and Evliya Çelebi The study of Ottoman influences in Berberistan is in its infancy. However, in addition to the evidence provided by the literary sources, archaeological remains that may date to the period survive. In this section I seek to compare the evidence of Evliya and the Vatican map for the border region with the material remains. Such an attempt is at this stage in places very tentative, for few of these sites (with the exception of Sai) have been studied by archaeologists. Nonetheless, it is hoped the outline below will shed some light on this forgotten frontier. The sites discussed have been chosen because they appear to date from the post-medieval, Islamic period (between the fifteenth and nineteenth centuries) and can be connected with the places Page 3 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Archaeology of the Middle Nile Valley in the Sudan mentioned in our Ottoman sources.22 The intention is to outline possible areas for further work, as so little primary archaeological evidence is available to date. It should be stressed that this study is still (p.376) very much in its initial stages; excavation and research into oral histories, in addition to meticulous topographical study of the area and its place names, will in the future provide supplementary evidence. The political structure of the frontier region was complex. The final Ottoman outpost was probably always the fortress of Sai. However, as discussed above, in the sixteenth century the frontier with the Funj had been established rather further south, at Hannek, and the short-lived sancak of Mahas was created. It is clear from Evliya Çelebi’s account, confirmed by independent evidence, that in the seventeenth century Mahas was ruled by vassals of the Funj, the meks (kings) of Mahas. Evliya names the mek as Kör Hüseyin.23 Nonetheless, there is some evidence for continuing Ottoman influences in Mahas. It is interesting to note that even on the southern frontier of Mahas, Evliya could encounter a Turkish-speaking local ruler, who had been born in Ottoman-controlled Aswan and presumably received an education in Turkish there.24 Indeed, as far south as Old Dongola, the Funj ruler’s name was followed in the Friday prayers by a mention of the Ottoman sultan in his capacity as the guardian of Mecca and Medina.25 Qal‘at Sai Qal‘at Sai26 (N 20° 44′ 12″ / E 30° 19′ 52″) was the southernmost Ottoman fortress of a significant size. It had been under Ottoman control since at least 1582–3,27 and occupied the site of an earlier fortress—indeed, there had been one here as early as the New Kingdom (1580– 1050 BC). It was strategically located to control the frontier region, trade routes and local agriculture. Estimates for the size of the garrison vary substantially; when Evliya Çelebi visited in the late seventeenth century he found the garrison to be 150 strong, but other evidence puts it as high as 200 during the period 1585–1660.28 At any rate, the garrison at Sai was always at least twice the size of that at Ibrim.29 The Ottoman administration at Sai consisted of a kaymakam and a kadı.30 Substantial modifications were made at Sai (Figure 19.3) by the Ottomans in order to accommodate modern artillery. One bastion was completely rebuilt in mudbrick for cannon, another was largely remodelled. Inside the fort was a Friday mosque, barracks (p.377) and houses that seem rather similar to the design of those at Suakin, some two or three stories high. They had a diwan and harem, and were perhaps inhabited by soldiers and their families as happened at Qasr Ibrim shortly after 1600.31 Thus although Sai is an example of the reuse of an existing fort, it is evident that the Ottomans were prepared to invest significant sums in modernising it.

Page 4 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Archaeology of the Middle Nile Valley in the Sudan Sai’s importance derived from its position near the frontier with the Funj. The Shayqiyya confederation, based south of Dongola, revolted around 1660, which lessened Funj control over the north. The fort may have been captured from the Ottomans in the late seventeenth century, presumably by the mek of Mahas. Certainly, it seems from Evliya’s evidence that the inhabitants of Sai distrusted the inhabitants of the areas to the south of it.32 However, any such occupation must have been temporary, as the fort is mentioned in Ottoman archival documents as late as the end of the eighteenth century.

33

Figure 19.3. Qal‘at Sai. On the left, the reinforced south-west bastion is visible.

(p.378) Tinare Evliya Çelebi reports that Tinare (N 20° 21′ 12″ / E 30° 28′ 58″) had a garrison of 800, making it substantially larger than Sai—apparently it was ‘the chief place of Mahas’.34 It was under the control of a certain Salih, who ruled over 40,000–50,000 Kerrarish nomads.35 The archaeological evidence confirms the importance of Tinare. It is a large structure, made up of two parts, a fort and a stable area. Built of stone, mud brick and mud (jalus), it lies on an outcrop above the river about 100 metres east of Tinare village. To the north and the east lies the river, with hilly areas to the west and south. In addition to the naturally fortified location, the fort is enclosed by high walls with towers. There are openings for muskets, and cannonplatforms at the eastern and the north-western towers. The so-called stable block comprises about five stone-built rooms, along the southern mud wall. The walls have openings for firearms. Small finds from here include pottery sherds of the Christian and Islamic eras, fragments of pipes, and date kernels.36 It is interesting to note the presence of the gun emplacements. Although dating is problematic, they may indicate an Ottoman presence at Tinare as the Ottomans had a monopoly on such weaponry in Nubia. Jabal Sese Jabal Sese (N 20° 07′ 15″ / E 30° 33′ 15″) was, according to Evliya Çelebi, the southernmost point the Ottomans reached, being temporarily occupied by the kaşif of Ibrim, Mehmed, in 1562–3.37 It was opposite Delgo, and three days’ march from Sai. The local lord was the mek of Mahas, Kör Hüseyin Bey. It had a thousand houses of reed, and like Ibrim and Sai, no market. Evliya records that, in contrast to the artillery and cannon used by the Ottomans at Sai, the local weapons of war were ballista, lances and bows and arrows, and soldiers there rode elephants, horses and dromedaries. It was seven stages from here to the boundaries of Habeş. The extant remains at Sese are situated on a prominent hill and comprise a massive stone wall on its crest, with a large gateway on the south side. On its eastern slope are remains of a large settlement of stone houses. There is a small number of larger structures within the higher northern part of the main enclosure. Surface sherds include a wide range of material from the Christian and Islamic eras beside earlier periods. (p.379) According to local tradition the site was associated with the traditional coronation ceremonies of the Mahas kings.38

Page 5 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Archaeology of the Middle Nile Valley in the Sudan Kada Musa Kada Musa (N 20° 03′ 40″ / E 30° 35′ 30″) is a fortified site, located on the east bank of the Nile, north of the cataract, opposite Nanarti Island. This is presumably to be associated with the island fortress of Naznarinte with a garrison of 500 noted by Evliya. The Vatican map records a Narante on the east bank of the Nile, which was under Kör Hüseyin, had a garrison of 300 black soldiers and was defended by guns of elephant bones. On the west bank was Narnarte-i Sagira under Kör Hüseyin’s son.39 The site consists of a substantial enclosed settlement and a cemetery, with a large stone-built jetty. The settlement covers around two hectares, and both the settlement and the cemetery differ substantially from other monuments in the Mahas. The use of large flat green mudbricks is not attested elsewhere, and the only obvious parallels to the upstanding tombs come from near Sai. These unusual features have given rise to the suggestion that the site might be an Ottoman military outpost, constructed in the wake of the Ottoman campaigns of c.1584. In the words of the team that surveyed it, ‘the location of the site, the regular plan and construction of the internal buildings and its otherwise unusual features are certainly compatible with such a hypothesis.’ However, a single AMS sample of the mudbrick gave a date of 470±30 BP (AD 1405–60), rather earlier. The site is certainly in need of further investigation.40 Arduan Island Arduan Island is a large seasonal island, the channels separating it from the mainland drying up during the low Nile period. It contains many forts and settlements that cover different periods of the history of the Sudan. Phonetic similarities give rise to the suspicion that this is the place mentioned by Evliya and the Vatican map as Virdan, comprising a solidly built fortress, seven mosques, a caravanserai and other public buildings. It was also known as Madinat al-Kubra, and was under Kör Hüseyin.41 Alternatively, it may have been Hafir-i Kabir, Kör Hüseyin’s capital, which was a great city and fortress adorned with large and small mosques, a caravanserai, baths and pious foundations and had a garrison 7,000 strong. The Vatican map reports that ivory, (p. 380) rhinoceros horn, lizard skin, ebony and teak could be found on its streets. Hafir-i Kabir at one point at least seems to have been under joint control of the Funj and an Ottoman kaşif.42 Arduan Island contains the remains of several important post-medieval sites that require surveying and excavation. One of these is Amla Diffi, a large and well-preserved diffi (mud-brick or mud-fortified structure) measuring c. 24 × 18.5 m, close to the river bank, north of Amla village. To the east of the diffi are the extensive but poorly preserved remains of other as yet unidentified structures. Although much of the pottery is medieval, some of the remains are likely to be more recent, and according to oral tradition the site was reoccupied during the Mahdiyya.43 Diffi Dunni is at the southern end of the village close to the river, and is a small well-preserved diffi of unusual construction, built of mud-mortared stone foundations and jalus superstructure, notably having rounded corners. Arduan Diffi, Arduan-Sageig and Artindiffi are among the other sites and fortifications associated with the post-medieval period on the island.44 Simit Island Simit Island is a large island approximately five kilometres long, very rocky and barren in the interior, although the edges have agricultural land and date palms. The island may be the Sindi of Evliya and the Vatican map, which marked the boundaries of Kör Hüseyin’s territories and the start of the heart of the direct sovereignty of the Funj.

Page 6 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Archaeology of the Middle Nile Valley in the Sudan On Simit are ruins of many structures; one of them, situated by the river bank, bears the suggestive name Osmanikki. Two principal fortifications from the postmedieval period survive. Shiglindiffi (N 19° 46′ 18″ / E 30° 21′ 82″) is a diffi on the south-eastern edge of the island, with three surviving towers and a gate. It has stone foundations and a mud superstructure. Saabindiffi (N 19° 47′ 46″ / E 30° 19′ 07″) is a large fortified settlement in the north of the island. There is a gateway in the south wall, and a large tower in the centre of the west wall; also possibly four corner towers, all now heavily eroded. The site is possibly a medieval foundation but may be more recent.45 Hannek Hannek (N 19° 42′ 69″ / E 30° 21′ 8″) is well known as the traditional site of the Ottoman-Funj frontier in the sixteenth century, and is mentioned by Evliya as having (p.381) a fortress. The most significant post-medieval building is a large fortified structure just to the south of the present village, a diffi with a standing tower. The building is traditionally considered to be the palace of the kings of Mahas, and may have been the site of a governor’s residence in the postmedieval Kokka kingdom.46

Unidentified sites in Berberistan Above, an attempt has been made to link the places mentioned in Evliya Çelebi and the Vatican map with the archaeological evidence. In some instances, such as Tinare, it seems certain that we are dealing with the same place because the modern and seventeenthcentury place names are the same; in other instances, such as Kada Musa and Simit, the identification is merely a hypothesis on which only future research can shed further light. There are also, of course, a number of post-medieval sites which one might assume played an important role in this frontier region but which cannot readily be linked with the places mentioned by our seventeenth-century Ottoman sources. Among the important archaeological sites is Jazirat Tombus (N 19° 42′ 10″ / E 30° 23′ 40″), an island just south of Hannek. Although it is therefore slightly to the south of the traditional sixteenth-century border region, it was clearly of considerable military importance. On the south-west corner of the island is a massive mud-brick fortified building (diffi) which is a mixture of dates from medieval and post-medieval times. Including later extensions, the building measured 125 m north-south and 45 m west-east. The diffi is one of the largest Islamic sites in the region.47 Kommer (N 20° 14′ 3″ / E 30° 34′ 8″) also contains a substantial castle dating to the medieval and Islamic periods, built largely of mudbrick and stone.48 There are also numerous smaller fortifications such as at Musul (N 19° 48′ 11″ / E 30° 18′ 64″) that cannot be traced in the literary sources. Equally, not everywhere mentioned by Evliya Çelebi or the Vatican map can be readily identified, Hafir-i Kabir and its sister town Hafir-i Sagir being a case in point. Another important fortification mentioned by Evliya which is surrounded by some mystery is Maqraq, Evliya’s next destination after Sai according to the Seyahatname. This had been conquered by the Ottomans and held for forty days in the days of Hadım Süleyman Pasha, governor of Egypt in the 1520s or 1530s. At the time of Evliya’s visit, it was in the hands of a certain Hasan Bey from the tribe of Kelapish. From time to time Hasan Bey would rebel against the Funj and recognise the authority of the Ottoman governor of Egypt, but Maqraq was currently subject to the Funj. The same (p. 382) place is mentioned by Burckhardt as Mekreke, but it is placed by him to the north of Sai.49 Udal has attempted to explain this difficulty by suggesting that Evliya’s chronology is at this point confused, and that he in fact visited Maqraq before Sai.50 This brings one back to the question of how reliable these seventeenth-century Ottoman accounts are. Although at times

Page 7 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Archaeology of the Middle Nile Valley in the Sudan they are clearly prone to exaggeration, and not every place mentioned in them can be identified, they clearly do discuss real places and generally present a feasible north-south itinerary.51 The surviving archaeological remains, especially the substantial fortifications, support the evidence of Evliya Çelebi that the Mahas region was quite strongly militarised, with forts and garrisons in almost every place he stopped. Of course, to some extent such fortifications probably aimed to ensure the autonomy and security of local rulers such as Kör Hüseyin rather than being aimed specifically against the Ottomans. Nonetheless, traditions such as that recording Ibn Janbalan’s defeat at Hannek in the seventeenth century indicate that the region remained a bone of contention and conflict between the Funj and the Ottoman rulers of Egypt long after the end of the sancak of Mahas.

Conclusion The Nubian frontier region of the Ottoman Empire is one of its least-known areas. It raises numerous questions relating to both Sudanese and Ottoman history, as well as the nature of relations between the Ottomans and the Funj, in which the Ottoman garrisons on the Middle Nile must have played an intermediary role. For instance, the Ottoman garrisons—based above all at Sai—must have left a genetic legacy, as was noted by Burckhardt in the early nineteenth century, who commented on the fairskinned kalecis of Bosnian descent.52 The nature of Ottoman rule is also a difficult problem. While Sai is traditionally thought of as marking the southernmost extremity of Ottoman rule, the evidence of some of the sites above suggests that Ottoman rule extended further south, towards Hannek where the great battle with the Funj was fought in the sixteenth century. The presence of openings for flintlocks and platforms for cannon in some of the structures investigated is intriguing, for only the Ottomans were thought to possess this technology in the region. In many ways, the key to understanding the nature of the interactions here is Mahas. Even after the short-lived sancak had disappeared, connections between Funj vassals and the Ottomans appear to have continued, with the presence of kaşifs alongside the meks, the occasional Turkishspeaking (p.383) local potentate, and the Ottoman sultan’s name mentioned in Friday prayers. An interesting feature of the structures associated with Ottoman rule is that almost all those from the first Turkish period seem to be based on earlier sites—in the case of Sai stretching back to the New Kingdom. This reflects both the continuities of the frontier region and (with the exception of Sai, where alterations were clearly extensive) the fairly limited investment by both the Funj and the Ottomans in the region. Both detailed textual and archaeological research are required to shed light on these problems. Our knowledge of the Islamic archaeology of Nubia is in any event limited; terms like qal‘a (fort), qasr (palace) and diffi are often used interchangeably without definitions. Excavation, along with the study and analysis of small finds like tobacco pipes, gun remains and flint bullets, can shed greater light both on the functions of the structures and the cultural orientation of their inhabitants. In terms of documentary research, a more detailed study of the text of Evliya Çelebi as well as the Vatican map than has been undertaken to date is needed, along with a comparison of their information with the actual topography of the area. Philological research, such as the study of place names like Osmanikki that seem to suggest an Ottoman connection, is also required. In addition the rich oral traditions collected by travellers like Burckhardt and ethnographic accounts offer hitherto under-exploited sources of information. A prerequisite for a deeper understanding of the Ottoman presence on the Middle Nile is an interdisciplinary approach to research, bringing together oral histories, ethnoarchaeological studies, material culture, and examination of historical documents in conjunction with an archaeological survey.

Page 8 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Archaeology of the Middle Nile Valley in the Sudan Notes: Proceedings of the British Academy 156, 371–383. © The British Academy 2009. (1) For more on this see Chapter 26 below, Paul Lane and Douglas Johnson, ‘The archaeology and history of slavery in South Sudan in the nineteenth century’. (2) David N. Edwards, The Nubian Past: An Archaeology of the Sudan (London: Routledge, 2004), 277–80; idem, ‘The potential for historical archaeology in the Sudan’, Azania, 39 (2004), 20–6. (3) Andrew McGregor, ‘The Circassian qubba-s of ‘Abbas Avenue, Khartoum: governors and soldiers in the nineteenth century Sudan’, Nordic Journal of African Studies, 10 (2001), 28–40. (4) A. B. Theobald, ‘Ali Dinar, Last Sultan of Darfur 1898–1916 (London: Longman, 1965), 42–3, 152–3, 160, 181. (5) See Chapter 24 below, Michael Mallinson et al., ‘Ottoman Suakin 1541–1865: lost and found’. (6) John Alexander, ‘The Saharan divide in the Nile Valley: the evidence from Qasr Ibrim’, The African Archaeological Review, 6 (1988), 73–90; Martin Hinds and Hamdi Sakkout, Arabic Documents from the Ottoman Period from Qasr Ibrim: Texts from Excavations (London: Egypt Exploration Society, 1986); Martin Hinds and Victor Ménage, Qasr Ibrim in the Ottoman Period: Turkish and Further Arabic Documents (London: Egypt Exploration Society, 1991). (7) John Alexander, ‘The Turks on the Middle Nile’, Archéologie du Nil Moyen, 7 (1996), 15–35; ‘Qalat Sai: the most southerly Ottoman fortress in Africa’, Sudan and Nubia, 1 (1997), 16–20; ‘The archaeology and history of the Ottoman frontier in the Middle Nile Valley 910–1233 AH / 1504–1820 AD’, Adumatu, 1 (2000), 47–61; ‘The real Turkia’, Cahiers de Récherches de l’Institut de Papyrologie et d’Égyptologie de Lille: Mélanges offerts à Francis Geus, 26 (2006), 17–28. (8) Intisar Soghayroun el Zein, ‘The Ottomans and the Mahas in the Third Cataract Region’, Azania, 39 (2004), 50–7; V. L. Ménage, ‘The Ottomans and Nubia in the sixteenth century’, Annales Islamologiques, 24 (1988), 137–51. (9) On the Funj see R. S. O’Fahey and J. L. Spaulding, Kingdoms of the Sudan (London: Methuen, 1974), 15–116. (10) Evliya Çelebi, Seyahatname, 10: 438–48; Italian summary and commentary: Maria Teresa Petti Suma, ‘Il viaggio in Sudan di Evliya Čelebi’, Annali dell’Istituto Orientale di Napoli, 14 (1964), 433–52; English summary: John O. Udal, The Nile in Darkness: Conquest and Exploration, 1504–1862 (Norwich: Michael Russell, 1998), 17–35. (11) Ettore Rossi, ‘A Turkish map of the Nile River, about 1685’, Imago Mundi, 6 (1949), 73–5. (12) O. G. S. Crawford, The Fung Kingdom of Sennar (Gloucester: John Bellows, 1951), 168–71. (13) J.-L. Burckhardt, Travels in Nubia (London: John Murray, 1819), 40–85, 133–48. (14) Ménage, ‘The Ottomans and Nubia’, 142–4; O’Fahey and Spaulding, Kingdoms, 34–5; William Y. Adams, Nubia: Corridor to Africa (Princeton: Princeton University Press, 1977), 609–12; cf. Chapter 11 above, John Alexander, ‘Ottoman frontier policies in north-east Africa, 1517–1914’.

Page 9 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Archaeology of the Middle Nile Valley in the Sudan (15) P. M. Holt, A Modern History of the Sudan from the Funj Sultanate to the Present Day (London: Weidenfeld & Nicolson, 1961), 3, 24–5. (16) On al-Dirr and al-Diwan, see Hinds and Sakkout, Arabic Documents, 2–3. (17) P. M. Holt, ‘Sultan Selim I and the Sudan’, Journal of African History, 8 (1963), 20–1. (18) Hinds and Sakkout, Arabic Documents, 26. (19) Burckhardt, Travels, 33. (20) Hinds and Ménage, Qasr Ibrim in the Ottoman Period, 8. (21) On the Ottoman garrisons, see Adams, Nubia, 611–12. (22) Omitted from discussion here is Nauri, on which see el Zein, ‘The Ottomans and the Mahas’, 536. (23) Alexander, ‘Qalat Sai’, 19; idem, ‘The Turks on the Middle Nile’, 25–6. (24) Udal, The Nile in Darkness, 25. (25) Ibid., 32. (26) This is one of the few sites to have been subject to a detailed archaeological investigation, although as yet little work has concentrated specifically on the Ottoman remains at Sai. See F. Geus, ‘Archaeology and history of Sai Island’, The Sudan Archaeological Society Newsletter, 8 (1995), 27–34. (27) Udal, The Nile in Darkness, 23. (28) Ibid., 18. (29) Ménage, ‘The Ottomans and Nubia’, 137. (30) Udal, The Nile in Darkness, 23–4. (31) Alexander, ‘Qalat Sai’, 17–18. (32) Udal, The Nile in Darkness, 24. (33) Stanford J. Shaw, The Financial and Administrative Organization and Development of Ottoman Egypt 1517–1798 (Princeton: Princeton University Press, 1962), 395. (34) Burckhardt, Travels, 58. (35) Udal, The Nile in Darkness, 24–5. (36) The Mahas Survey, 2005 season, unpublished findings. (37) Udal, The Nile in Darkness, 25.

Page 10 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Archaeology of the Middle Nile Valley in the Sudan (38) David Edwards and Ali Osman Salih, The Mahas Survey 1991: Interim Report and Site Inventory (University of Cambridge, 1992), 97 (available at https://lra.le.ac.uk/handle/2381/498). (39) Petti Suma, ‘Il viaggio in Sudan’, 440–1. (40) Ali Osman and David Edwards, The Mahas Survey Report, 2000: A Preliminary Report, 9–10, 27 (available at https://lra.le.ac.uk/handle/2381/498). (41) Petti Suma, ‘Il viaggio in Sudan’, 442. (42) Petti Suma, ‘Il viaggio in Sudan’, 441, n. 49. (43) Osman and Edwards, Mahas Survey Report, 2000, 11. (44) Ibid., 8–15. (45) Ali Osman and David Edwards, The Mahas Archaeological Survey, 2002: A Preliminary Report, 5 (available at https://lra.le.ac.uk/handle/2381/498). (46) Ibid., 9. On Kokka see A. Osman, ‘The post-medieval Kingdom of Kokka: a means for a better understanding of the administration of the medieval Kingdom of Dongola’, in J. M. Plumely (ed.), Nubian Studies: Proceedings of the Symposium for Nubian Studies (Warminster: Aris and Phillips, 1982), 185–97. (47) Edwards and Osman, Mahas Survey 1991, 23. (48) David Edwards and Ali Osman Salih, The Mahas Survey 1990: Interim Report and Site Inventory (University of Cambridge, 1993), 62 (available at https://lra.le.ac.uk/handle/2381/498). (49) Burckhardt, Travels, 53. (50) Udal, The Nile in Darkness, 24. (51) Cf. Chapter 6 above, Malcolm Wagstaff, ‘Evliya Çelebi, the Mani and the fortress of Kelefa’, 122–3. (52) Burckhardt, Travels, 134–5.

Access brought to you by:

Page 11 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds IBOLYA GERELYES

DOI:10.5871/bacad/9780197264423.003.0020

Abstract and Keywords It is generally accepted in Hungarian scholarly literature that Ottoman soldiers stationed on Hungarian territory did not mix with the local Hungarian population during the Ottoman period in Hungary (1541–1699). However, the written historical sources attest direct contact between Ottoman garrisons and the Hungarian population. This chapter attempts to examine how the picture drawn by the historical sources can be supplemented by archaeological evidence, in particular ceramics. This can be seen in the records of shops owned by the Ottoman Treasury and in customs registers. These sources attest that craftsmen and traders settling in the wake of the soldiers and supplying them were employed in the production and sale of goods that were almost exclusively for the satisfaction of basic everyday needs. Keywords:   Ottoman Hungary, ceramics, Ottoman garrisons, supply hinterlands, Ottoman Treasury

IT IS GENERALLY ACCEPTED IN HUNGARIAN SCHOLARLY LITERATURE that Ottoman soldiers stationed on Hungarian territory did not mix with the local Hungarian population during the Ottoman period in Hungary (1541–1699). Together with the civilians who followed in their wake, they lived, we are told, sequestered in their castles and forts, or in separate districts of the occupied towns. In other words, the two cultures existed alongside each other and in parallel.1 However, the written historical sources attest direct contact between Ottoman garrisons and the Hungarian population. For example, some Hungarians served in the Ottoman military and some, such as the Hungarian scribes and secretaries of the pashas of Buda (Ott. Budun), converted to Islam.2 The number of Christian Hungarians serving with the Ottoman army was relatively small, amounting in the vilayet of Buda to approximately 0.42 per cent of all Ottoman troops there.3 Much more important than this—and more widespread—was the practical connection represented by the involvement of the Hungarian population in the building and maintenance of castles and forts, and in the transportation of the materials necessary for such work.4 Thus we can assume that contact on a day-to-day level did exist. One of the prime forms of material evidence for this is archaeological data, especially handicrafts catering to

Page 1 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds basic needs. This study will attempt to examine how the picture drawn by the historical sources can be supplemented by archaeological evidence, in particular ceramics. (p.386) The incomers attempted to recreate their earlier way of life in their new environment. The most obvious evidence for their lifestyle, which was appreciably different from that of the Hungarians, is to be found in the probate inventories of Turks in Hungary who died without heirs. It is also reflected in the records of shops owned by the Ottoman Treasury and in customs registers.5 These sources attest that craftsmen and traders settling in the wake of the soldiers and supplying them were employed in the production and sale of goods that were almost exclusively for the satisfaction of basic everyday needs. In the forefront stood activity connected with food, drink and clothing.6 The written sources make scant reference to Turkish potters working in occupied Hungary, although two potters are mentioned among the trades of the Muslim population of Baç (Hu. Bács) in 1578.7 We also know that in Buda Castle, in an area still not exactly located, masterpotters inhabited a separate quarter,8 and there are the remains of what seems to be a Turkish potter’s kiln in the Szenttamáshegy district of Esztergom (Ott. Estergon).9

New Ceramic Types in Ottoman Hungary These appeared in the Ottoman-controlled central part of the country during the sixteenth to seventeenth centuries, as part of the material culture of the Turks and those peoples of Southern Slav origin who entered Hungary with them and after them (and even before them). These new types were first noted in the early twentieth century, following excavations at Buda Castle and in one of Buda’s Turkish quarters, a part of the city inhabited by craftsmen engaged in leather production.10 The first comprehensive analysis of this ‘new’ type of ceramic was largely based on these finds.11 Such material is also to be found in the collections of the provincial museums (p.387) that have grown up across the country; analysis of it, however, only started in the late 1950s.12 In excavation and restoration work carried out in the decades following the Second World War, huge quantities of ceramic material from the late Middle Ages and from the Ottoman era were unearthed, primarily at castles. All subsequent excavations have confirmed that types of ceramic vessel appeared in the second half of the sixteenth century that were not just unknown hitherto, but also substantially different in both shape and technique from medieval Hungarian ceramic products. The excavations prove that the production of local Hungarian ceramics and their importation from Hungarian territories not under Turkish sway continued during the period of the Ottoman presence in the country. This chapter contends that ceramic artefacts can indicate ethnicity. If we accept this, the abovementioned ceramic finds may serve as the basis for an investigation relating to the parallelism of the two cultures, and to the influences they exerted on one another. The connection between ceramics use and ethnicity and religion is best explained by the different eating habits of different cultures.13 Among archaeological finds in all periods, new ceramic forms appearing among archaeological finds en masse and without precedent locally have always called attention to a possible change in population. In historical periods (for example, at the time of the Ottoman Empire’s spread in Europe), establishing a change in population is simple, since written sources are available. In Ottoman Hungary, Christian Hungarian inhabitants were replaced in the occupied castles and towns by a population with different ethnic roots. This new population ate foods that were different to those consumed by the local inhabitants, and they prepared and

Page 2 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds served them in a different way. One manifestation of this change is the mass appearance in the archaeological record of new ceramic forms. Ceramic types surviving in small quantities that differ fundamentally from local wares are a different matter. In these cases we are speaking of presents or of the collection of artefacts considered exotic in their new environment. An excellent example here is the Iznik ceramic ware that appears sporadically among pre-Ottoman Hungarian finds, for instance at the court of King John Zapolya (1526–40). This can be linked to the king’s diplomatic connections with the Porte.14 Some ceramics also occur as a result (p.388) of trade. The appearance of special imported ware clearly indicates a change in demand rather than a change in ethnicity. A good example of such a change is the appearance in Hungary of Chinese porcelain cups, which appeared as a result of the spread of coffee-drinking, a change in eating and drinking habits.

Classification of the New Ceramic Types In what follows, three large and distinct groups are examined. The sizes of these groups are then compared to one another, and also to the group consisting of local Hungarian ceramics. Classification is performed partly on the basis of technique used and partly on the basis of shape. First Group This consists of yellow, brown or green lead-glazed vessels: footed bowls and spouted jugs. These were made on a foot-driven fast wheel using good-quality red or brown clay, and were well fired. In addition to these two main types, many other vessels and artefacts for use appear. Examples are two-handled storage vessels (amphoras), jugs with broad brims, candleholders, and inside-glazed stove-elements shaped like rimmed bowls (Figures 20.1 and 20.2). These finds changed neither in form nor in technique during the Ottoman period. (p.389) With regard to shape, the footed bowls are uniform: whether shallow or deep and whether supplied with a low or high foot, their upper parts are mostly semi-spherical. Nor is there any variety in decoration. The vessels are coated either with a monochrome glaze or with bands of glaze each of a different colour. In terms of technique, the characteristics of the spouted jugs and the wide-brimmed jugs without spouts accord with those of the footed bowls. They were made using red or light brown clay. On the upper third of the vessels, on a white engobe base, a monochrome yellow, brown or green glaze was applied. These jugs are also glazed inside.

Figure 20.1. Glazed Turkish ware. Sixteenth to seventeenth century. Hungarian National Museum, Budapest. Photograph by Judit Kardos.

Figure 20.2. Line drawing of glazed Turkish ware. Sixteenth to seventeenth century. Buda Castle.

This type of ceramic is called ‘Turkish’ by Hungarian scholars. On the basis of parallels known from Anatolia and the Balkans,15 such vessels can be linked to Muslim civilians and soldiers, whose eating and drinking habits were different from those of the Hungarian population. Hence Page 3 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds these vessels should characterise the ceramic finds at the castles held by the Ottomans. In fact, as we shall see in the analysis below, there are significant differences in the material recovered from the various sites. One observation, (p.390) though, is valid for the whole of Hungary, namely that so-called Turkish ceramics occur only where there was a Muslim population, i.e. in Ottoman-held castles and palankas and the Muslim town districts connected with these. On the basis of the archaeological record, we can assert that unlike ornamental Turkish pieces, this type of ceramic appears nowhere before the Ottoman occupation.16 Additionally, finds of this kind are unknown in the Hungarian territories beyond the Ottoman-controlled area, for example in material from castles that were never in Ottoman hands.17 Especially interesting is the fact that the Hungarian population of the occupied areas—the inhabitants of Hungarian-populated market-towns and villages of western Hungary and the Great Plain—do not seem to have used these vessels.18 We can suggest that in Ottoman Hungary so-called Turkish ceramics belonging to this group indicate ethnic (Turkish) identity and religious (Muslim) identity at one and the same time.19 Second Group These were fired using the so-called reduction-fired technique and include good-quality grey or black jugs and crocks made similarly to those of the first group. They come to light primarily in castles and palankas that were in Ottoman hands.20 There are direct parallels to the pieces uncovered in Hungary of this group from Serbia, and so their appearance in the archaeological record is at present linked by Hungarian researchers to Southern Slav peoples arriving from the Balkans.21 Some of these vessels are poorerquality unembellished and simple pieces that in form resemble medieval Hungarian village ware; such artefacts appeared on Hungarian territory as early as the 1530s, before the Ottoman occupation.22 Our present knowledge suggests that the Hungarian population adopted the technique (but not the forms) used in this pottery during the Ottoman period, and to some extent even before it. Outside the Ottoman-occupied regions, black pottery appears only during the second half of the seventeenth century, among finds recovered from castles in Hungarian hands (Figure 20.3). (p.391) Third Group The existence of this group was only noticed during the second half of the twentieth century.23 It comprises grey or black thickwalled, unglazed pots fashioned from coarse Figure 20.3. Reduction-fired black jug. Seventeenth century. Hungarian National clay on a hand-turned wheel. Surfaces are Museum, Budapest. Photograph by Judit covered with wavy-line, spiral and Kardos. impressed ornamentation; in the case of larger vessels, ribbing with fingerprints is used. The technique, characteristics and the simple mode of embellishment of these vessels prompted earlier researchers to date them to some centuries before the Ottoman occupation, a (p.392) practice which continued until the contents of assemblages could be dated with certainty. This type of pottery counted as archaic in sixteenth- to seventeenthcentury Hungary, where use of a foot-driven wheel had become general by around 1500.24 Its origin is clearly indicated by its close connections with archaeological and ethnographical material from the Balkans.25 Indeed, based on ethnographical parallels in the twentieth century, some researchers have proposed the term ‘Bosnian’ for pottery of this kind. Material belonging to this pottery group typically occurs in southern Transdanubia (i.e. in south-

Page 4 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds west Hungary), at the smaller palankas first and foremost. It has been found at SzekszárdÚjpalánk (Ott. Yenipalanka), Babócsa (Ott. Bobofça), Márévár (Ott. Malvar), Mecseknádasd (Ott. Nadasd), Segesd (Ott. Şegeş), and Törökkoppány (Ott. Koppan),26 as well as in the places featuring in the analysis below. So far, pottery of this kind has not come to light in the Ottomancontrolled but Hungarian-inhabited villages of Transdanubia and the Great Plain. Moreover, in the Hungarian territories outside the area of Turkish rule it is completely unknown (Figures 20.5 and 20.6, nos. 1 and 2). In my analysis below, I have included in this group baking vessels used for cooking outside over an open fire, and also the bell-shaped covers for these artefacts, as their technique is similar. The picture is, of course, more complicated than this. Among the archaeological finds there also appear luxury ceramics of Turkish origin.27 From the mid-sixteenth century onwards, large quantities of Iznik faience are detectable primarily among the finds recovered at Buda Castle, but also at most other castles that were in Ottoman hands.28 By the early seventeenth century, Iznik material no longer occurs, yielding place to Chinese and Persian ceramic ware.29 Since archaeologists are now examining material from the early eighteenth century, it seems that with the end of the Ottoman occupation in the late seventeenth century, Oriental ceramics disappear from the archaeological record. The present analysis will not extend to faience pieces of Oriental origin as, with a very few exceptions, these appear only in the material associated with Ottoman garrisons, and have no role in the links between these garrisons and the local population.30 However, the above types of Ottoman-era material never occur simply by themselves, but in different proportions depending on time and place, and always mixed with Hungarian ceramics (Figures 20.4 and 20.6, nos. 3–7). The pieces in individual assemblages exhibit significant differences. (p.393) (p.394)

Figure 20.4. Hungarian pottery. Seventeenth century. Hungarian National Museum, Budapest. Photograph by Judit Kardos.

Figure 20.5. Pots made on a hand-turned wheel. Late sixteenth century. Mór Wosinsky Museum, Szekszárd. Photograph by Judit Kardos.

Page 5 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds (p.395) Ottoman Castles and Ceramic Finds Ottoman castles and palankas fall into a number of categories, depending on their geographical, strategic or administrative function (for a map see Figure 3.2). In making the selection on which the current analysis is based, I have endeavoured to present as many different types as possible. Of course, the possibilities are limited by the sites excavated and by the proportion of their finds processed so far. In the analysis in Table 20.1, I have used material recovered during my own excavations at Békés (Ott. Bekeş)31 and Ozora (Ott. Ozora),32 as well as hitherto unstudied artefacts found in the course of earlier digs at Buda,33 Visegrád (Ott. Vişegrad)34 and Érd-Hamzsabég (Ott. Hamzabey);35 I have also utilised material already published relating to Eger (Ott. Eğri),36 Bátaszék (Ott. Batasek)37 and Barcs (Ott. Barç).38 The samples are based on material of closed Figure 20.6. Late sixteenth-century pots from Ozora Castle. 1–2. Pots made on a assemblages recovered from cisterns, pits hand-turned wheel. 3–7. Hungarian pottery. and filled-in cellars. In the case of smaller finds, entire assemblages could be processed, but in those from the larger castles such as Buda or Eger, the processing of all the material recovered is not practicable. (p.396)

Page 6 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds

Table 20.1. Ceramic finds in Ottoman castles of sixteenth- to seventeenth-century Hungary. %of finds Site and period of finds

Function

Dates of Ottoman occupation

Size and compositon of

Glazed Turkish ceramics

Reduction-fired black ceramics

Ceramics on a hand-turned wheel

Local Hungarian ceramics

garrison*

Bares (Ott. Bare) palanka

Protected River Dráva crossing and hinterland of Szigetvár (Ott. Sigetvar) and Kanizsa (Ott. Kanije)

1567–1664

193 AMa (1568) 6

3

84

7

Bátaszék (Ott. Batasek) palanka

Protected military road and shipping along the Danube

1550–1686

103 FAMa + 4 unknown number of timar holders (1552) 30 MÜT + timar holders (1565) 69 MÜTFA (1613) 21 TF (1629)

25

36

35

Békés (Ott. Bekeş) palanka. Late 16th cent.

Protected shipping at confluence of two rivers

1566–95

248 HMüFAMaT 2 + timar holders in unknown number (1591)

0

0

98

Buda Castle Palace. Early 17th cent.

Capital of vilayet of Buda

1541–1686

2,361 HYMüCeTGFAC r (1613) 3,361 HYMüCeTGFAC r (1629)

58.3

8.3

0

33.4

5,697 HYMüCeGACr (1662–77)

33.8

8

13.8

44.4

Buda Castle Palace. Late 17th cent.

Page 7 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds

%of finds Site and period of finds

Function

Dates of Ottoman occupation

Size and compositon of

Glazed Turkish ceramics

Reduction-fired black ceramics

Ceramics on a hand-turned wheel

Local Hungarian ceramics

52.4

7

3.1

39.5

1,965 62.2 HYMüCeTGFAM aCr (1605)

0

0

37.8

43

0

8.6

47.4

garrison*

Buda Castle Palace. Averages Eger (Ott. Egri) Castle. Early 17th cent.

Capital of vilayet of Eger

1596–1697

Eger (Ott. Egri) Castle. Late 17th cent. Hamzabey Sarayi palanka. Mid-17th cent.

Protected military road and shipping along Danube

63 MüFAMa (1613) 75 HMüTFAMa (1629)

48

8

0

44

Ozora Castle. Mid-16th cent.

Stronghold deep 1545–1686 inside Ottoman territory

93MüA(1545) 42 Mü (1568)

2.3

0.5

2.3

94.9

Ozora Castle. Late 17th cent.

1545–1686

27.2

3

22

48

Visegrád (Ott. Protected Visegrád) Lower crossing-point Castle. Late on Danube 16th cent.

1544–1684

342 MüTAMa (1544) 125 MÜTA (1568)

36.5

0

1.5

62

52MüA(1613) 19 A (1629)

33

4.4

8.3

54.3

Vi§egrad Lower Castle. Late 17th cent.

(*) I use the database compiled by Klára Hegyi together with her abbreviations. These are as follows. A: riiesa ve azeban (foot soldiers); Ce: cebeciyan (armourers); Cr: craftsmen; F:farisan (horsemen); G: gönüllüyan (volunteers); H: hademe-i cevami (mosque staff); Hu: Hungarians; Ma: martolosan (raiders); Mü: müstahfizan (elite foot soldiers); T: topfuyan (artillerymen); and Y: yenigeriyan (Janissaries). Klára Hegyi, ‘The

Page 8 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds Ottoman network of fortresses in Hungary’, in Géza Dávid and Pál Fodor (eds.), Ottomans, Hungarians, and Habsburgs in Central Europe: The Military Confines in the Era of Ottoman Conquest (Leiden: Brill, 2000), 174–93.

Page 9 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds (p.397)

(p.398) Conclusions Based on the analysis above, the types of pottery used by garrisons, and the ratios of these types to one another, are determined basically by two factors. The Composition of the Garrison in Question It is no coincidence that better-quality, higher-standard glazed Turkish pottery (assigned to the first group above) occurs in substantially greater proportions in vilayet capitals. This type of ware is found in the greatest proportions in Buda (52.4 per cent) and Eger (52.5 per cent).39 The association of this pottery with elite troops is supported by the fact that where such units were present in smaller numbers or were absent altogether, the proportions of such pottery, too, were smaller (4 per cent at the palanka at Bátaszék and 6 per cent at the one at Barcs).40 However, the presence of Janissaries or other elite Muslim troops did not preclude the use of local pottery in large proportions. Such products could make up almost one half of the whole, as at Eger Castle in the late sixteenth century, or approximately 40 per cent as at Buda Castle and at the palanka of Hamzsabég.41 Ethnicity of the ‘Supply Hinterlands’ Diversity in the ethnic composition of supply hinterlands and changes in this over time are reflected by the variety in the make-up of finds uncovered at the different fortifications. This suggests the existence of a close, everyday link between the Ottoman garrisons and the inhabitants of the settlements nearby, at least on the level of trade. The exclusive or very largescale use of local ware is puzzling primarily in the case of those garrisons where elite troops or mosque staff can be found among the soldiers.42 In such cases, significant assimilation to local conditions may have occurred despite (p.399) strict religious obligations. One sign of this was changes in habits connected with eating and drinking.43 At the same time, vessels produced locally on the basis of Turkish models appear. These vessels were quality footed bowls made from yellowish-white clay and supplied with glazes of different types. They have been found in Buda, Eger and Békés.44 In places where the supply hinterland was Hungarian, local Hungarian pottery appears in significant quantity. At less strategically important palankas deep inside Ottoman territory (for example, those at Békés and Ozora), the portion of local Hungarian wares tends to be greater. Here Turkish ceramics are very rare, perhaps 2 per cent of the whole, as at Békés palanka. In these cases more than 90 per cent of pottery recovered is local ware, independently of the whereabouts in Ottoman Hungary of the stronghold investigated. This is well indicated by a comparison of the palankas, of Békés and Ozora, where the proportions of the different types of pottery recovered are almost identical. In the Békés case, the palanka was situated in an area that was entirely Hungarian: according to the sancak registers of 1567 and 1579, the population of the town of Békés and of the villages around it was Hungarian.45 The same was true for the palanka at Ozora in the second half of the sixteenth century.46 It was these Hungarian-inhabited areas that served as the hinterland from which the soldiers at the two forts obtained supplies necessary for day-to-day life. Despite the occupation, the population of the sancak of Buda likewise remained Hungarian in the middle of the sixteenth century.47 This is reflected in the high proportion of Hungarian finds among the ceramics material recovered at Buda, Visegrád and Hamzsabég.

Page 10 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds Ethnic Changes as Reflected in the Ceramic Corpus Southern Transdanubia experienced ethnic changes from 1570, when transhumant Vlach shepherds first appeared in the sancak of Koppan (Koppány) and, in smaller numbers, in the sancak of Şimontorna (Simontornya). These newcomers settled in abandoned Hungarian villages.48 In the seventeenth century, following the so-called Fifteen Years War (1593–1606), the Rác population of southern Transdanubia increased appreciably, in Tolna county most of all, but also in Fejér and Somogy (p.400) counties.49 At least half of the inhabitants of Tolna and Fejér counties belonged to this grouping. Despite their large number, Rác villages did not form solid blocs but were mixed with Hungarian villages and thus no solid bloc of Hungarians similar to that on the Great Plain was to be found in Transdanubia. Rác settlements stretched all the way to Buda.50 Just south of Buda, the palankas of Hamzsabég, Ercsi (Ott. Erçin), Adony (Ott. Korkmaz), and Pentele (Ottoman place name unrecorded) had links with Hungarian settlements, while further to the south the larger and smaller Ottoman palankas of southern Transdanubia (for example, those of Dunaföldvár (Ott. Fedvar), Paks (Ott. Pakşa) and Bátaszék) had ties with the local Rác inhabitants, whose presence was described by Heinrich Ottendorff in 1663.51 The evidence of the written sources is borne out by archaeological finds. In the ethnically changed area of southern Transdanubia, the proportion of pottery of Balkan origin increased. In Rác-inhabited Tolna, the material at the palanka at Ozora was completely transformed. By the end of the seventeenth century, the use of local Hungarian pottery there had fallen to half, while from an earlier average of 2 per cent the use of Turkish pottery had increased to 27 per cent and that of Southern Slav pottery made on a hand-turned wheel to 22 per cent.52 Other conspicuous examples are the palankas, of Hamzsabég/Hamzabey and Bátaszék/Batasek. Located by the side of the Danube, they performed similar functions. In both fortifications Christian Southern Slav soldiers constituted the greater part of the garrison, but at Hamzsabég Balkan pottery is entirely absent while at Bátaszék it forms the largest group. The difference in the material, which was used by garrisons of identical ethnic composition, clearly reflects the ethnic difference described by Ottendorff, namely the absence of a local Rác population at Hamzsabég and the presence of one at Bátaszék. The proportions of pottery made on a hand-turned wheel (assigned to the third group above) deserve special attention. This type of pottery appears in substantial quantities in precisely those parts of southern Transdanubia from which the Hungarians withdrew northwards, namely those parts settled by Balkan peoples during the seventeenth century. It occurs in very substantial proportions in the finds from Barcs and Bátaszék (both in southern Transdanubia), and by the late seventeenth century in increased proportion at Ozora, too. At Bátaszék and Barcs, the increase in (p.401) the proportion of hand-turned ware is accompanied by a fall in the proportion of glazed Turkish pottery, and at Ozora by a fall in that of local Hungarian ware. The fact that hand-turned pottery does not occur among the finds from Eger, Békés and Hamzsabeg—and only in minimal amounts among those at Buda and Visegrád—supports the above thesis. One type of ware, the black pottery assigned to the second group, is more ambiguous, and a wider selection of material should be studied. Even so, two key observations can be made. Firstly, black pottery occurs almost everywhere except Eger and in practically the same proportions. These are always significantly below the proportions for glazed Turkish ware. There is one glaring exception, Bátaszék palanka, where it reaches 25 per cent. Secondly, its appearance is not characteristic of the sixteenth century.

Page 11 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds The mutual influences of the two cultures on each other in the long run were certainly different. However, there have been no studies to date on the influence of Hungarian culture on the Turks and others withdrawing from Hungary. On the other hand, it is worth repeating what was said in the introduction about the pottery used by the Hungarian population of the time. The Hungarians living in a solid bloc on the Great Plain and in the Hungarian-populated villages of Transdanubia used neither glazed Turkish pottery nor hand-turned Balkan ware. In the period examined, medieval Hungarian handicraft traditions largely lived on in these parts. Despite this, as a final outcome of the 150-year-long occupation, particular Ottoman influences gained a certain acceptance in the local population, with Ottoman elements entering Hungarian culture.53 Notes: Proceedings of the British Academy 156, 385–401. © The British Academy 2009. (1) Klára Hegyi, Török berendezkedés Magyarországon (Budapest: História-MTA Történettudományi Intézete, 1995). (2) Sándor Takáts, ‘A magyar és török íródeákok’, in Kálmán Benda (ed.), Művelődéstörténeti tanulmányok a 16–17. századból (Budapest: Gondolat, 1961), 173. (3) Hegyi, Török berendezkedés, 110. (4) Pál Fodor, ‘Bauarbeiten der Türken an den Burgen in Ungarn im 16–17. Jahrhundert’, Acta Orientalia Academiae Scientiarum Hungaricae, 35 (1981), 55–88. (5) Lajos Fekete, Das Heim eines türkischen Herrn in der Provinz im XVI Jahrhundert (Budapest: Akadémiai, 1960); A. Velics and E. Kammerer, Magyarországi török kincstári defterek (Budapest: A M. Tud. Akadémia Történelmi Bizottsága, 1886–90); Lajos Fekete and Gyula Káldy-Nagy, Rechnungsbücher türkischen Finanzstellen in Buda (Ofen) 1550–1580 (Budapest: Akadémiai, 1962). (6) See Velics and Kammerer, Magyarországi török, 2: 22–4, 120–3, 252–4; Fekete and KáldyNagy, Rechnungsbücher; Gyula Káldy-Nagy, A gyulai szandzsák 1567. és 1579. évi összeírása (Békéscsaba: Békés Megyei Levéltár, 1982), 42–4, 47–51. (7) Klára Hegyi, ‘A török Bács’, in Pál Fodor and Géza Pálffy (eds.), Tanulmányok Szakály Ferenc emlékére (Budapest: n. p., 2002), 199–213. (8) Lajos Fekete et al., Budapest a törökkorban (Budapest: Egyetemi Nyomda, 1944), 84. (9) Géza Fehér and Nándor Parádi, ‘Esztergom-szenttamáshegyi 1956. évi törökkori kutatások’, Esztergom Évlapja, Esztergomi Múzeumok Évkönyve 1 (1960), 35–44. (10) Lajos Nagy, ‘Tabán a régészeti ásatások világában’, Tanulmányok Budapest Múltjából, 4 (1936), 27–8, Plates 14–15. (11) Sándor Garády, ‘Agyagművesség’, in Fekete et al., Budapest a törökkorban, 382–401. (12) See Olivér Soproni, ‘Szolnok török kerámiája’, Jászkunság, 3 (1956), 56–61, 102–6; Géza Fehér, ‘A pécsi Janus Pannonius Múzeum hódoltságkori török emlékei’, Janus Pannonius Múzeum Évkönyve (1959–60), 103–49; Géza Fehér, ‘Adatok Eger török agyagművességéhez’, Az Egri Page 12 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds Múzeum Évkönyve, 10 (1972), 191–214; Gyöngyi Kovács, Török kerámia Szolnokon (Szolnok: Szolnok Megyei Múzeumi Adattár, 1984). (13) For a further detailed discussion of the link between ceramics and religious affiliation in the Balkans, see the findings of the Krajina Project published by Carlton and Rushworth, Chapter 21 below. For a study of the relationship between eating habits and ceramics for the Ottoman period see Joanita Vroom, After Antiquity: Ceramics and Society in the Aegean from the 7th to the 20th Century A.C (Leiden: Faculty of Archaeology, University of Leiden, 2003), 335–57. (14) Gyöngyi Kovács, ‘Iznik pottery in Hungarian archaeological research’, in Ibolya Gerelyes (ed.), Turkish Flowers: Studies on Ottoman Art in Hungary (Budapest: Hungarian National Museum, 2005), 82–3. (15) For direct parallels of the finds in Hungary known from the territory of Serbia and Bosnia see Fehér, ‘A pécsi Janus Pannonius Múzeum’, 103. (16) This is indicated by, for example, finds unearthed at the palace in Buda Castle. For an analysis of the fills of refuse pits from the Ottoman period and earlier see Imre Holl, Fundkomplexe des 15.-17. Jahrhunderts aus dem Burgpalast von Buda (Budapest: Archäologisches Insitut der Ungarischen Akademie der Wissenschaften, 2005), 16–17, 28–9. (17) See for example material from two castles in northern Transdanubia (north-west Hungary), which lay outside the Ottoman Empire: Károly Kozák, ‘A sümegi és a szigligeti vár XVII. század végi kerámiája’, Veszprém Megyei Múzeumok Közleményei, 5 (1966), 81–9. (18) An example from Transdanubia is ceramic material recovered from the market-town of Ete, 10 kilometres from the Ottoman palanka at Szekszárd. In this material there were no such vessels. See Zsuzsa Miklós and Márta Vizi, ‘Válogatás egy késő középkori fazekasműhely leletanyagából’, in Szilvia Andrea Holló and János Szulovszky (eds.), Az agyagművesség évezredei a Kárpát-medencében (Budapest and Veszprém: MTA-VEAB, 2006), 91–9. The same picture is presented by material recovered from Hungarian villages in the Kecskemét area between the Danube and Tisza rivers, a region under Ottoman rule: See Kálmán Szabó, Kulturgeschichtliche Denkmäler der Ungarischen Tiefebene (Budapest: Országos Magyar Történeti Múzeum, 1938). (19) For the ethnic composition of castle garrisons see Hegyi, Török berendezkedés, 100–4. (20) Finds of this type were recovered at Szolnok Castle and Pécs, in addition to the sites discussed below. See Fehér, A pécsi Janus Pannonius Múzeum’, plate 28; Kovács, Török kerámia, plates 24–5. (21) Z. Simić-Milanović, ‘Beograd kroz muzejski material’, Godišnjak Muzeja Grada Beograda, 1 (1954), 17. (22) Nándor Parádi, ‘Magyarországi pénzleletes középkori edények. XI-XVII.sz’, Archaeológiai Értesítő, 90 (1963), 205–51. (23) Győző Gerő, ‘Türkische Keramik in Ungarn. Einheimische und importierte Waren’, in Géza Fehér (ed.), Fifth International Congress of Turkish Art (Budapest: Akadémiai, 1978), 347–61.

Page 13 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds (24) Imre Holl, ‘Adatok a középkori magyar fazekasság munkamódszereihez’, Budapest Régiségei, 7 (1956), 177–96. (25) A. Kalmeta, ‘O seljačkom lončarstvu u zapadnoj Bosni: Über die Töpferei in den Dörfern West- und Mittelbosniens’, Glasnik Zemaljskog Muzeja u Sarajevu, 9 (1954), 131; Cv. D. Popović, ‘Loncarstvo u Bosni i Hercegovini’, Glasnik Zemaljskog Muzeja u Sarajevu, 11 (1956), plate 1; M. Marjanović-Vujović, ‘Kuća iz druge polovine XVII veke otkopana u utvrećenom podgraću Beogradskoj grada-Donjem gradu’, Godišnjak Muzeja Grada Beograda, 20 (1973), plate IV. (26) Tamás Pusztai, ‘The pottery of the Turkish palisade at Bátaszék’, in Ibolya Gerelyes and Gyöngyi Kovács (eds.), Archaeology of the Ottoman Period in Hungary (Budapest: Hungarian National Museum, 2003), 301–10. (27) Gerő, ‘Türkische Keramik’, 348. (28) Kovács, ‘Iznik pottery’, 69–86. (29) In the material from Eger Castle, which fell to the Turks in 1596, there is scarcely any Iznik faience; on the other hand, the amount of Persian faience and Chinese porcelain there is substantial: László Fodor and Károly Kozák, ‘Leletegyüttesek a románkori székesegyház környékéről (Adatok az egri vár XVII-XVIII. századi kerámiájának történetéhez I.), Az Egri Múzeum Évkönyve (Eger), 8–9 (1970–1), 173. (30) For reasons of space, another important area, the occurrence of imported West European ceramics, has to be omitted here. However, so-called Vienna pottery does appear in significant quantities, albeit in differing proportions from garrison to garrison and to all intents and purposes only in sixteenth-century layers. (31) Ibolya Gerelyes, ‘Előzetes jelentés a Békés-kastélyzugi törökkori palánkvár ásatásáról, 1975– 1978’, Archaeologiai Értesítő, 107/1 (1980) 102–12. (32) The archaeological investigation of the site began in 1981. The present author directed the investigations into the Turkish palanka established around the castle. (33) For Buda, only a small part of the vast ceramic corpus has been consulted, namely finds in the 1970s from the area of the Middle Castle (the so-called Orta Hisar) and Arsenal Square (the so-called Tophane meydani) to the north. On the excavations there see László Gerevich, A budai vár feltárása (Budapest: Akadémiai, 1966); Holl, Fundkomplexe. (34) Large-scale archaeological excavations were conducted at Visegrád’s Lower Castle in 1959– 63 and 1966–9. My analysis is based on material recovered during these digs that is now kept at the King Matthias Museum in Visegrád; I also draw on the findings of an earlier paper: Ibolya Gerelyes, ‘Török kerámia a visegrádi Alsóvárból’, Communicationes Archaeologicae Hungariae (1987), 168–78. (35) My analysis was made on the basis of unpublished finds from the period 1962–5 kept at the Hungarian National Museum, Budapest. These were uncovered in the direct vicinity of the cami, the minaret of which still stands, although the building itself was washed away in 1838. Pottery recovered at Hamzsabég differs significantly from that at similar palisaded forts.

Page 14 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds (36) For Eger my analysis is based on material from two pits of different periods published by Fodor and Kozák, ‘Leletegyüttesek’, 147–99. (37) Ilona Valter, ‘Die Ausgrabungen in der ehemaligen Cisterziensen Abtei Cikádor’, Annalecta Cisterziensia, 52 (1997), 251–64. (38) Gyöngyi Kovács, ‘A barcsi török palánkvár kerámialeletei’, Communicationes Archaeologicae Hungariae (1998), 155–80. (39) Use of so-called luxury ceramics was likewise highest in these two places. Of the Iznik faience recovered from the Ottoman-occupied territory of Hungary, 80 per cent was found at Buda, where a similarly high percentage of the Persian faience and Chinese porcelain unearthed in the country was also discovered. Iznik faience from Eger is almost unknown; on the other hand, it is there that the largest amount of Chinese porcelain has been found. (40) In the garrison of the Bátaszék palanka there were, in the first years, faris, azeb and martolos units. Later on elite mustahfiz soldiers also appeared, although their number was modest (twenty-one to sixty-nine men). In the palanka at Barcs only azabs and martaloses served: Klára Hegyi, ‘The Ottoman network of fortresses in Hungary’, in Dávid and Fodor (eds.), Ottomans, Hungarians, and Habsburgs, 183, 186. (41) As well as other elite units, Janissaries served in Buda and Eger. Elite foot and mounted units were stationed at the Hamzsabég palanka: Hegyi, ‘The Ottoman network’, 174, 175, 176, 190. (42) The presence of elite units and mosque staff indicates a higher proportion of Muslim soldiers within a garrison. Velics and Kammerer, Magyarországi török, 1: 378; 2: 244–6, 251, 264; Fekete, Budapest a törökkorban, 86–8. (43) The same is suggested by the discovery of significant numbers of pig bones in the material from Ottoman-held palankas. (44) For the moment, I can only assume—on the basis of examples consistent with Hungarian products—that these vessels were made in workshops of Hungarian potters. (45) Káldy-Nagy, A gyulai szandzsák, 148–274. (46) Géza Dávid, A simontornyai szandzsák a 16. században (Budapest: Akadémiai, 1982), 67, 220–7. (47) Gyula Káldy-Nagy, A budai szandzsák 1559. évi összeírása (Budapest: Pest Megyei Levéltár, 1977). (48) Géza Dávid, ‘The connection between history and archaeology in researching Ottoman rule in Hungary: interrelationships and perspectives’, in Gerelyes and Kovács (eds.), Archaeology of the Ottoman Period in Hungary, 11–16. (49) Rác was the blanket term in Hungarian at this time for Orthodox Serbs and Muslim and Catholic Bosnians. (50) Ferenc Szakály, Magyar adóztatás a török hódoltságban (Budapest: Akadémiai, 1981), 180–1; Hegyi, Török berendezkedés, 195. Page 15 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Garrisons and the Local Population in Ottoman Hungary: The Testimony of the Archaeological Finds (51) Heinrich Ottendorff, Der Weg von Ofen auf Griechisch-Weissenburg (Österreichische Nationalbibliothek, Vienna. MS Cod. 8481). (52) The Békés palanka known to archaeological research burned down in the late sixteenth century. The location of its seventeenth-century successor is not known. This explains the absence of comparisons with material from that time. (53) For example, a certain type of vessel, the spouted jug, and a type of production, black pottery, were adopted by Hungarian village ceramics workshops.

Access brought to you by:

Page 16 of 16 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland RICHARD CARLTON ALAN RUSHWORTH

DOI:10.5871/bacad/9780197264423.003.0021

Abstract and Keywords This chapter summarises the results of the Krajina Project, which was established in 1998 to investigate the archaeological remains, material culture and continuing ethnographic legacy of this distinctive late medieval/early modern frontier society. The project has focused on an area in the north-west corner of Bosnia-Herzegovina, between Kladuŝa and Bihać, known as the Bihaćka Krajina. This was one of the last districts in the region to be conquered by the Ottoman state, not falling to the sultan's forces until the late sixteenth century — a territorial high water mark. The ethnographic evidence provides significant insights into the continuing legacy of the Ottoman-Hapsburg frontier in the nineteenth and twentieth centuries. Keywords:   Bosnia-Herzegovina, Bihaćka Krajina, Ottoman Empire, pottery, Ottoman-Hapsburg frontier

BETWEEN THE SIXTEENTH AND NINETEENTH CENTURIES, much of Croatia and northern Bosnia formed a contested frontier region or krajina, where three great powers—Hapsburg Austria, maritime Venice and the Ottoman Empire—all collided. By the end of the period the Vojna Krajina, or ‘Military Frontier’, was a tightly defined zone of the Hapsburg Empire stretching from Dalmatia to Transylvania, where military settlers with special privileges and restrictions were fostered to provide a defensive buffer against the threat of Ottoman incursion. Earlier arrangements were less formal, but both the Ottoman and Hapsburg states made extensive use of local militia forces, variously labelled Hayduds, Voynuks, Martolos or Grenzer, and the label ‘Krajina’ is thus applied to districts on both sides of the eighteenth- and nineteenth-century Ottoman-Hapsburg frontier, which now forms the border between Bosnia and Croatia. Perhaps the best-known of these groups, following the work of Wendy Bracewell,1 are the piratical Uskoks of Senj, on the northern Adriatic coast. The frontier system was finally formally disbanded in the later nineteenth century, leaving a militarised, marginalised society with traditions and complex identities which may have contributed significantly to recent conflicts.

Page 1 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland This chapter summarises the results of the Krajina Project, which was established in 1998 to investigate the archaeological remains, material culture and continuing ethnographic legacy of this distinctive late medieval/early modern frontier society2 The project has focused on an area in the north-west corner of Bosnia-Herzegovina, between Kladuša and Bihać, known as the Bihaćka Krajina. This was one of the last districts in the region to be conquered by the Ottoman state, not falling to the sultan’s forces until the late sixteenth century—a territorial high water mark. Thereafter it remained under Ottoman control until 1878 when Bosnia was occupied by AustriaHungary, which finally annexed Bosnia in 1908. Prior to the Ottoman conquest, the area had been incorporated in the frontier defences organised by the Hapsburg authorities from the 1520s onwards,3 and only a few kilometres to the north and west, in present-day Croatia, lie territories which were reabsorbed into the Hapsburg empire in 1699 and then subjected to the formalised military administration of the Vojna Krajina, which remained in place until 1881.4 As a result of these repeated fluctuations in the border between the two powers, the Bihaćka Krajina and adjacent districts of Croatia provide an opportunity to study the reaction of settlement and society to a variety of political and military frameworks during the history of the frontier (see Figures 21.1 and 21.2). (p.404) (p.405)

Figure 21.1. Map showing the location of the Bihaćka Krajina within Bosnia and Herzegovina, with successive OttomanHapsburg frontier lines marked.

Page 2 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland (p.406) Fieldwork has involved collaboration between staff of the Archaeological Practice, the University of Newcastle upon Tyne, English Heritage and the Zemaljski Muzej, Sarajevo, and has encompassed several different strands, including the survey and excavation of known fortified centres, the recording of traditional building types and ethnographic study of craft practices. It is important to stress that ethnographic fieldwork was carried out in tandem with the investigation and recording of historic or archaeological remains, and that no clear distinction was made between what is considered ethnographic and archaeological, present or past. Clearly, examples of excavated material culture may be considered to have Figure 21.2. Map showing main places in been derived from an archaeological the Krajina mentioned in the text, with approach, but in many cases the materials modern political borders. derived from such investigations are items discarded within current lifetimes, or at least recognisable as tools used within current lifetimes: the distinction between past and present, then, is blurred. Many of the results and conclusions presented here are necessarily preliminary and tentative, designed to suggest avenues of enquiry and raise questions for debate, rather than present definitive answers to historical problems.

Castles and Fortifications The archaeological component of the project has concentrated on fortification sites as the most tangible surviving remains of the Ottoman frontier. A hierarchy of fortification types can be documented. Large modernised fortresses, furnished with angle bastions and substantial garrisons of regular troops, were relatively rare because they were so expensive to construct and man. Such treatment was reserved for the principal frontier headquarters and logistical bases, such as Karlovac or Slavonski Brod, or the partially modernised Bihać (Ott. Bihke), held by the Ottoman Empire after 1592. Much more common were medieval castles, most of which continued in use in the early modern period. Lower down the scale were simple timber forts or blockhouses known as palanka5—inevitably much harder to detect in the archaeological record— and individual towers, or kula (Ott. kule), some of medieval origin, like Radetina Kula, others of more recent timber construction, like the small wooden turrets built by the Hapsburgs as part of a cordon sanitaire to prevent the spread of cholera from the Ottoman Empire. The most visible of these fortification types in the Bihaćka Krajina are the medieval castles. The detailed recording and investigation of two of these, Pećigrad and Stijena, were undertaken as case studies in 2001 and 2002. The fieldwork involved topographic survey, photographic recording and evaluation trenching. The results of the work at (p.407) Pećigrad are summarised below, together with parallels provided by Stijena and rapid examination of a number of other sites.

Page 3 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland Pećigrad Castle (Ott. Török Mihal Pekleri) (Figures 21.3 and 21.4) occupies a central, strategic location within the Bihaćka Krajina at the head of the Kladuša valley, with commanding outlooks to the north and east especially. To the south the modern route towards Cazin and Bihać climbs to the summit of the watershed at the Skokovi Pass. The castle is first documented in the fourteenth century, when it formed part of (p.408) the demesne of the lordship of Kladuša,6 hence its name in 1334, Kladuška Peć. Subsequently it passed through the hands of various members of the Croatian lesser nobility until 1577, when its garrison of sixteen infantrymen (haramija) was overcome by Ottoman assault.7 The status of Pećigrad over the next half century is unclear, but it seems to have been fully occupied by Ottoman troops during an interval in hostilities, the karlovački mir, in

Figure 21.3. Plan of Pećigrad castle.

1635. This date is also given by Bećirbegović8 for the construction of its mosque. Published photographs,9 oral testimony and the surviving remains of houses suggest that the interior of the castle was more densely occupied at the beginning of the twentieth century than it is today, justifying its local designation of starigrad (‘old town’). Shops and houses were arranged along two main streets on either side of the (p.409) mosque. The settlement supported a weekly market and there was even a small museum for a time, up to around 1920. The castle suffered some damage during the Figure 21.4. View of Pećigrad from the Second World War, but far more severe south-east showing castle and mosque destruction occurred in the immediate within. aftermath, when extensive demolition of the ideologically unfashionable monument was instigated under the new communist regime and its looted stones were used to build ‘houses of culture’ and schools locally. This robbing was halted after c.1947, when responsible governmental authorities became established and were able to intervene. More recently, the strategic location of the castle was underlined when it was used as a strongpoint during the Bosnian War (1992–5) by both the Bosnian Government’s forces based at Bihać and the troops of the breakaway Autonomous Province of Western Bosnia, led by Fikret Abdić and based in Velika Kladuša.10

Page 4 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland The survey plan shows the main elements of the site. The castle occupies a hilltop location bordered to the east by a precipitous, vertical cliff which forms its eastern defences and partly determines the crescent-moon-shaped plan of the inner bailey (Figure 21.3). Only a short stretch of a second, outer line of defences could be traced in 2001. The walls of the inner bailey are constructed with a rubble and mortar core, faced by small, neatly coursed dressed stones. Although extensively robbed, the walls survive reasonably well, particularly in the southern part of the structure, where they remain at their original height complete with facing stones on both internal and external, the south-east example being the urvive, the south-east example being the best preserved, retaining a domed roof over the internal space, albeit with a hole resulting from shelling in the Second World War. The main gateway is situated on the west side of the circuit and is a relatively simple affair, the entrance being arranged at right angles to the line of the curtain and well protected by one of the projecting towers positioned immediately to the south. Much better preserved is the postern located next to the south-east angle tower. The only visible remains of the outer defences are a single dome-vaulted tower and adjoining 15 metre length of curtain wall. The outer tower is almost circular and displays a very different style of construction to that of the inner bailey, the facing made up of undressed rubble with much use of mortar to achieve a relatively smooth finish. No trace of any curtain walling linking the inner and outer defences was noted, but it is likely that the precipitous natural defences rendered this unnecessary. Within the inner enceinte there are a number of more recent buildings, most notably the late seventeenth-century mosque situated in the centre of the castle on the summit of the hill. Since the project’s first visit in 1998 this building has received a new vestibule on the north side, but retains the traditional form adopted by all but the largest of Bosnian mosques, with a pyramidical roof and single minaret. An ornate carved door, mentioned by Bećirbegović,11 no longer survives. Four houses inside the castle are still (p.410) inhabited and a further one, the least modified, stands unoccupied but still intact immediately north-west of the mosque. The remains of several other houses and house platforms can be seen in various states of decay. One house and several outbuildings were destroyed by rocket-propelled grenade fire during the most recent war. The contrasting masonry styles of the inner and outer defences suggest that Pećigrad castle was built in two separate phases. The rough, heavily mortared rubble construction of the surviving tower on the outer defences is typical of medieval fortications in the Bihaćka Krajina and neighbouring districts of Croatia, being encountered at Tržac, Todorovo, Stijena, Velika Kladuša and Slunj, for example. The small, neatly-coursed, dressed stonework employed on the inner enceinte is also paralleled in the region, at Stijena, and at Cetingrad, 17 kilometres to the northwest in Croatia.12 Cetingrad (Ott. Çetin) was an important feudal stronghold of the medieval Hungarian-Croatian kingdom and later formed part of the frontier fortifications of the Ottoman Empire; whatever the precise date of the construction phases of Pećigrad, they must fit within the same context. More archaeological research is required before any precise chronological significance can be attached to either form of masonry. Indeed, both types may have been in use contemporaneously. At Pećigrad, a plausible case can be advanced that the outer defences were a later addition to the inner circuit, but at no point can a direct structural relationship between the two be observed which could confirm that hypothesis, and the alternative possibility, that the inner circuit represents a later citadel or reduced circuit, forming a wholesale rebuilding within an earlier, more extensive fortification, cannot be decisively refuted.

Page 5 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland No evidence of Ottoman defensive refurbishment—such as provision of gun loops or outer bastions—was noted, although the removal of large quantities of facing stone in or around 1947 will have removed much evidence for small-scale repairs and refacing. Furthermore, the possibility cannot be excluded that gun loops were incorporated in stretches of the outer curtain which have now disappeared or were positioned in the towers of the inner circuit at higher levels than now survive. Nevertheless, the principal structural legacy of the Ottoman period is clearly represented by the mosque established in the centre of the castle. An evaluation trench was also excavated in the interior behind the south-west curtain to gain an understanding of the depth and character of deposits. This revealed late nineteenth- to early twentieth-century cobbled roadway and structures overlying a series of deposits dumped down the slope towards the base of the curtain. (p.411) A post hole was uncovered at the bottom of this sequence, suggesting that more extensive excavation might yield significant information regarding the castle’s internal structures. Finds included large quantities of pottery and clay pipe bowls, and flintlock pistol mechanisms—the latter an apt testament to frontier life. It is hoped that further work on the pottery derived from this and future excavations will help to elucidate patterns of contact and trade within the frontier province of the Ottoman Empire and across its borders at various phases in the lifetime of the active military frontier. The clay pipe bowls perhaps offer the best chance for close dating entire assemblages, but much work remains to be done on sourcing and dating these artefacts. At Stijena (Ott. İstineli) there is more evidence of Ottoman alterations. The gateway bears an inscription of 177113 and has evidently been inserted or refashioned contemporary with the inscription (Figure 21.5). Again there is a mosque in the (p. 412) interior. There is also evidence of early modern defensive modifications—two of the towers have been clad in an additional thick masonry skin and instead of crenellations the parapet incorporates a row of loops for muskets or crossbows. These may represent a response to an increasing Ottoman threat in the late fifteenth and sixteenth centuries, rather than Ottoman work. Thus, both at Pećigrad and Stijena, medieval castles were adapted as centres of local Ottoman power, with mosques built inside their circuits. Two questions are worthy of further comment. Firstly, the presence of mosques within both castles is no simple coincidence. It is paralleled elsewhere in the Bihaćka Krajina, at places such as Brekovica, Podzvizd and Todorovo, and at Tržac one stands immediately beside the castle. Was the siting of mosques in such

Figure 21.5. The gateway of Stijena castle surmounted by an inscription of AH 1185 / ad 1771.

Page 6 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland prominent defensible locations a deliberate policy on the part of the Ottoman authorities to impose the symbols of Muslim authority on the principal centres in each district, or a more gradual and ad hoc response by local Muslim communities? If the former, were the mosques at Pećigrad, Stijena and elsewhere built on the sites of earlier churches? A church, Sv. Križa, is certainly recorded at Pećigrad in the late medieval period. The best evidence of such replacement and reuse in the region comes from Bihać, where the former church of Sv. Antuna was converted to become a mosque, the Džamija Fethija, following the Ottoman conquest, with the addition of a minaret and a mihrab carved in the south-east corner. A fine group of sixteenth-century inscribed tombstones has been found during works at the entrance to the mosque, the originals now preserved in the Zemaljski Museum with copies displayed in Bihać Museum. Ranging in date from 1520 to 1565, these provide an excellent cross-section through the elite of Bihać in the period before its fall to the Ottoman Turks. The deceased individuals include the captain of the Bihać military district, Ivan Izačić (1565), plus several noblemen, a judge and a knight.14 Secondly, it might seem strange that fortresses such as Stijena and Pećigrad remained in use well into the age of gunpowder, despite relatively little defensive modernisation. However, contemporary accounts of sieges make it clear that, in the context of a strictly limited campaigning season and predominantly ‘low intensity conflict’ (or mali rat15), such apparently antiquated defences performed an important role, providing protection from mounted raiding parties. On occasion they could prove formidable obstacles when stoutly manned, slowing the enemy’s advance, buying valuable time for defending forces elsewhere. Moreover, even when their military utility had ceased, such sites remained powerful symbols of local power, combining both secular and religious authority in the Ottoman period.

(p.413) Ethnography A full investigation of the ethnographic legacy of the frontier region, or even a small part of the region, would be a monumental task, albeit a worthwhile one, and certainly lies far beyond the scope of the present project. Even works on the scale of Lockwood’s investigation of the Central Bosnian micro-region of Potočari, or Bringa’s investigation of a mixed Bosnian Muslim and Catholic village near Kiseljak, would demand at least a year’s continuous fieldwork.16 Such a study in the Bihaćka Krajina region would indeed be invaluable, since despite recent population movements it remains possible to study micro-regions containing the three main ethno-religious groupings of the area, Bosnian Muslim, Catholic and Orthodox. Given the resources available to the project team, however, ethnographic study has so far been restricted to providing a general ethnographic overview of the area and focusing on discrete areas for more detailed study. A brief overview of the ethnographic context of the area under consideration here is necessary before particular aspects of its material culture are examined in more detail. The Bihaćka Krajina lies at the north-west tip of Bosnia and Herzegovina, and stretches from the southern environs of Bihać northwards to the Croatian border at Velika Kladuša, a distance of some 50 kilometres. At its widest point, between Tržac and Izačić in the west and Otoka and Bosanska Krupa in the east, the district extends over 30 kilometres. In general terms, notwithstanding various subtle intra-regional cultural distinctions, this area, unlike most other parts of the former Ottoman frontier (or present Bosnian border), has an overwhelmingly Bosnian Muslim population.17 The last census recorded the population of north-east Bosnia, including Bihać, Cazin and Kladuša, to be over 97 per cent Muslim. It is thought that the region’s present Muslim population is derived principally from the descendants of indigenous converts to Islam, mostly Catholics who converted gradually in the later sixteenth and seventeenth centuries. The Muslim Page 7 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland population, therefore, shares many inherited cultural characteristics, not least a common language and dialect, with Catholics who remained in the region following the Ottoman advance and with Orthodox Christians who followed in its wake. (p.414) The Muslim population of the Western Krajina, particularly the area between Kladuša and Bihać, forms a discrete population group away from the main Muslim population group in Bosnia which, prior to 1992, extended eastwards from Sanski Most and Prijedor to Sarajevo and beyond to Novi Pazar. While certain population groups within this wider Muslim population display cultural distinctions, none are separated to the same degree as the Western Krajina from the main body of the Muslim population. The geographical and cultural separation of the western Krajina has fostered distinct cultural traditions and an attitude of marginalisation. To what extent this contributed to the Pokrajna Zapadna Bosnia movement of the 1990s is debatable, but the political and economic conditions that enabled the rise of Agro-Commerce, the conglomerate commercial enterprise which sustained the movement, are in part attributable to its geography and political culture. It is also of interest to note that the main towns of the region, Kladuša, Cazin and Bihać, maintain fairly discrete social and economic networks—it is rare, for example, for marriages to occur between individuals from Bihać and Kladuša areas— thus reflecting in microcosm the fragmented modern state of Bosnia-Herzegovina. The Catholic population of the region has gradually declined in numbers during recent decades, but several discrete population areas remain and indications of its former presence remain elsewhere. The principal and most numerous Catholic presence in the region is focused on a number of villages on the north and west sides of Bihać, namely Vrkašić, Kralje, Žegar and others. Other, smaller populations exist close to the Croatian border, notably at Gornji Šiljkovača near Kladuša, and at Velika Kladuša itself, where the Catholic church is the only one between Bihać and the border at Maljevac, north of Kladuša. The latter populations, always rather small, have shrunk since the 1992–5 war and Šiljkovača now contains no more than eight Catholic households. Unlike the Bihać population, it is sustained partly by direct contact with Catholic populations at Cetin on the Croatian side of the border, where its children go to primary school.18 All Catholics in the main Bihać concentration and other, dispersed, populations have joint Croatian and Bosnian citizenship, and many look towards Croatia for employment. Other Catholic populations were present until recently around Pećigrad, but probably disappeared from there during or soon after the Second World War, while place names such as Marjanovac, near Mala Kladuša, indicate that this place also originated as a Christian settlement. Orthodox Christian populations were non-existent in the region until the Ottoman conquest, and have subsequently remained small, in stark contrast with most other parts of the former Hapsburg-Ottoman and (to a lesser extent) Venetian-Ottoman borders, where they have traditionally predominated. Small Orthodox or Greek -Catholic (p.415) populations existed until recently, however, in a rural area close to the border at Gornji Šiljkovača, where they formed a mixed population with Catholics, but have not returned following the 1992–5 war, although in most cases their houses remain intact and their owners visit periodically. Other populations, probably originating during the period of industrialisation and urbanisation in the former Federal Republic of Yugoslavia (1945–91), existed until recently in Velika Kladuša and Bihać. In the latter case discrete Orthodox-populated areas existed on the roads north and south of the town, notably at and in the vicinity of Ripač. Most of this population has now deserted the town, however, leaving only a few older returnees and a glut of cheap property.

Page 8 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland Although Muslim and Christian populations of the Bihaćka Krajina share many common traditions and some, at least, have a common ancestry, there are also a number of clear cultural distinctions which are of potential archaeological interest, other than those directly associated with religion. Principal among these are distinctions in vernacular building traditions. Three types of vernacular building were selected for recording in 2001 and 2002—rural houses, watermills and the stela-like Muslim funerary monuments (nišani)—with interesting results. Rural Houses Different rural house types are identifiable which may be correlated with the ‘ethnic’ or confessional identity of the particular community, the distinction between square Muslim and oblong Christian (Catholic, Orthodox or Greek-Catholic) house types being clearly drawn (compare Figures 21.6 and 21.7). Typical features of both include timber and wattle-and-daub construction; the accommodation of livestock downstairs, with family living quarters located upstairs; the use of an outshot latrine at the rear, above the midden. Larger houses of both Christian and Muslim type are built on stone bases up to first-floor level, while older smaller houses are entirely of wooden construction. The personal recollections of older inhabitants, backed up by what is known from historic descriptions and photographs, suggest that the majority of dwellings in the past were smaller than most of those constructed in the last phase of traditional construction up to c.1970. This suggests that the majority were of the most simple kind, lacking a stone base. Such houses, indeed all houses of the timber-framed type, do not tend to last more than a generation or two, with only one or two in the region recorded up to 100 years old and most falling into disrepair well before fifty years. The apparent relative ephemerality of the housing extends also to its indefensibility and may, arguably, have originated as one of two possible responses to the ever-present threat of raiding and social upheaval, namely that it was deemed worth investing little time and expense in constructing housing which was likely to be destroyed in any case. Such ephemeral structures form a marked contrast to the defensible, towered dwellings evident elsewhere in Bosnia,19 particularly further south, in Herzegovina, which fall within a tradition of stone-built, residential towers encountered throughout the Ottoman Balkans, as far south as the Mani peninsula. Towers of this type may typify the other possible built response to insecurity, that of investing a great deal in a house, fortifying it to render it more defensible and therefore less likely to be lost. The varying intensity of raiding experienced in different regions may be one factor explaining such variations in vernacular building traditions, but, in actuality, the evolution of a particular tradition in a given area is likely to have been the result of a complex interaction between a wide range of influences, of which the degree of local conflict was only one. Geological and environmental factors, such as the ready availability of timber in northern Bosnia compared with the karst limestone landscape of Herzegovina, differences in the structure of local societies and in the pattern of their subsistence bases—the relative status of the occupants, the strength of kinshipbased social structures, or ‘clans’ and the prevalence of pastoralism in the rural economy, for example—are all likely to have exerted significant influence on the building forms adopted. (p.416) (p.417)

Page 9 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland (p.418) Funerary Stelae

Figure 21.6. A house in the Catholic/ Orthodox tradition at Šiljkovača. By Peter F. Distinctive Muslim funerary stelae (nišani) Ryder. (Figure 21.8), incorporating elaborate decorative motifs, are commonly found in cemeteries throughout the district. Examples have been examined at Pećigrad, Tržac, Stijena, Kajtezi and Lučka, with the most detailed record of carved decoration being made at Lučka. Female stelae often Figure 21.7. A typical house in the Muslim contain the symbol of plaited hair, while tradition of the Bihaćka Krajina at Grabovac, with construction detail. By Peter F. Ryder. male stelae are often topped by a turban— large turbans indicate that the deceased had undertaken the hajj. Perhaps most significant is the prominence on male stelae of muskets, scimitars and other antique (and not so antique) weaponry as decorative motifs. These may imply that the warrior status associated with the bearing of arms was a crucial element of male identity in the Bihaćka Krajina during the eighteenth, nineteenth and even twentieth centuries, perhaps attesting the persistence of a martial ideology long after the disbanding of the frontier institutions. The practice of carving of such stelae continues to this day. There are several family groups of stone carvers still working in the region, at least two at Bihać and one in Cazin, with one group at Bihać being able to trace its origins as a family business back to at least 1826. This opens up the potential, through the recording of oral testimony, to understand the symbolism (and changes in symbolic preferences) of these intriguing monuments in much greater detail. (p.419)

Page 10 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland Watermills Several examples of watermills, used for grinding corn into flour, have been examined and recorded at Pećigrad, Marjanovac, Vejnac, Tržac, Gornji Šiljkovača (Figure 21.9) and Otoka. All except the latter feature horizontal wheels set individually or in rows in the water channel, linked directly by vertical drive shafts to corresponding rows of millstones. The mills are of wooden construction, raised on stilts over the water channel, and originally roofed with wooden shingles, or perhaps thatched, but now almost always using iron sheeting. The present-day mills are owned by groups of individuals, or syndicates, and operated on a time-share basis, with each individual allotted a certain length of time for grinding. The time permitted to each individual is sometimes specified by (p. 420) date, but is more commonly a period of one or two days per month to be used flexibly. There is obvious potential to map these syndicates, through interviews with current users, to determine how they correlate with named communities, territorial districts or family groups.

Figure 21.8. Carved motifs on nišani in the cemetery at Lučka near Pećigrad.

In recent decades, due to a combination of factors including the availability of cheap milled products and a reliable source of electricity through a national grid, the use of watermills has reduced and, in some Figure 21.9. Cutaway reconstruction cases, as at Marjanovac, restrictions on drawing of a typical watermill, based on an individual use have been dropped, while example at Gornji Šiljkovača. By Peter F. Ryder. others, as at Vejnac and Mala Kladuša, have fallen into disrepair. However, it is unlikely that mill sites, or the community organisations (p.421) supporting them, will be formally abandoned while the memory of the 1992–5 war and its aftermath remains fresh and economic conditions in the area continue to be relatively poor. The recent conflict led to the renewed use of watermills; some, such as the family-run mill at Šiljkovača, a Christian area on the Croatian border, being equipped to generate household electricity. Pottery Making The ethnographic study of portable material culture and local traditional practices, such as pottery making, has also contributed to the definition of cultural boundaries and interactions within the survey area. Pottery has been studied particularly closely because it offers the best chance to develop a continuous pattern of social and economic interaction through time, linking Page 11 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland the present with the past through material culture. The detail provided in the following account reflects the potential of pottery to inform upon historical patterns of social and economic interaction by combining information derived from ethnography and archaeology. Pottery-making traditions in the west Balkans have developed under the influence of processes including invasion, acculturation and apparently non-causal change into a proliferation of strategies fully reflecting the long and complex history, not to mention present ethnic diversity of the region. The usual way of rationalising this enormous variety is to select for purposes of classification a single, easily observable characteristic, exemplified by Tomić’s division of pottery-making industries in Serbia (later applied to the entire area of the former Yugoslavia) according to their use of hand-wheel, foot-wheel or non-wheel technologies.20 Pottery making using the hand-wheel is practised mainly in central and western BosniaHerzegovina, southern Croatia, and western Serbia. It is characterised by the use of a wheel rotating upon a fixed, upward-pointed pivot and operated entirely, or almost entirely, by hand (Figure 21.10). Fabrics are usually coarse, generally calcitetempered, firing is usually by means of an open fire, sometimes in single- or doublechambered kilns, and quenching is almost universally practised. Types of pottery produced vary from place to place, but are mainly baking, cooking and storage wares (Figure 21.11) rather than decorative or table-wares. Other means of treating vessel surfaces are rare and decoration minimal, though glazing is occasionally reported (notably at Kaluđerovac). Almost all hand-wheel potters are peasants who depend partially upon pottery for their livelihood, but regard themselves primarily as farmers. The term ‘foot-wheel’ normally refers to a wheel comprising an upper working platform and lower fly-wheel which rotates within a fixed bearing below the fly-wheel. It is used, exclusively in some areas, throughout northern Croatia and northern Bosnia, and in Slovenia except the Bela Krajina/Ribnica valley areas.21 In Serbia, foot-wheel technology is usually associated with urban potters who, lacking connections with the land, must derive all of their income from pottery. Migrations of one such group, originating around Pirot in the Serbian-Bulgarian border area, have been reported by several observers.22 Northern Bosnian towns were colonised for the first time by such potters in the early part of the twentieth century, while their maximum range was fixed in 1977 when they established themselves at Vrginmost (now Gvozd) in central Croatia.23 Elsewhere, particularly in west Croatia and Slovenia, pottery making is predominantly rural and follows the potter-farmer model prevalent among handwheel potters (see above). In the Ribnica valley of south Slovenia, a fixed-pivot foot-wheel persisted until recently, which resembled in basic morphology a much enlarged hand-wheel, but operated as a foot-wheel. (p.422)

Page 12 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland (p.423) In the western Balkans, foot-wheel technologies are always associated with the use of kilns, the great majority of which are of the open or enclosed, double-chambered updraught variety. Single-chambered, through-draught tunnels of brick and clay daub were in use up to 2001 at Rastoki, north-west of Karlovac, on the edge of the ‘handwheel zone’,24 and kilns of similar design may also have been in use at Kompolje in the Lika region. The use of double-chambered or through-draft kilns by foot-wheel potters carries a number of implications, including increased costs in time, materials and space, the ability to produce harder, glazed products, and the potential problem of over-firing in association with certain coarse pastes.25 (p.424) The study of pottery as part of the Krajina Project has included three main approaches:

Figure 21.10. The potter Fehim Javurdzija photographed at Demiševci, north-west Bosnia in 1989.

1. The investigation of the kinds of pottery used in the Krajina region 2. The investigation of sources of pottery 3. A focused pilot study into the purchase and use of pottery vessels in various parts of the survey region. Local Production Centres

With regard to sources of pottery, two production centres are known to have operated in the Bihaćka Krajina, at Brezova Kosa/Skokovi, an area of scattered Muslim settlement south of Pećigrad, and at various Catholic villages on the north and west side of Bihać, the populations of which constitute Figure 21.11. Cooking pots from Demiševci, a large part of the surviving, minority Brezova Kosa and Kaluđerovac. Catholic community in what is one of the most ethnically homogenous parts of Bosnia. At Vrkašić, some three kilometres north of Bihać, the last surviving potter, Stjepan (Stipo) Jerković (born 1934), was interviewed on 14 January 1998 in order to confirm and extend the record of Kalmeta.26 The process and organisation of production follows the pattern outlined above, but it is of interest that a variety of fabrics were produced, using crushed calcite, sand, or a mixture of the two, depending on the type of vessels to be made and the quality of vessels required. Firing was done in an open bonfire in the yard or using a brick-built, single-chamber kiln, or peč, following

Page 13 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland which the pots were quenched in a mixture of flour and water.27 Products included open dishes (pršula/tava used for cooking, zdijela for food preparation and presentation), large storage vessels (ćup), jar-shaped cooking pots (rukatka or lonac), portable bread-ovens (peka), stove-tiles (petnjak) and other pots used in construction, such as chimney flues (lula za dimnjak). Other forms, such as the spouted water storage and pouring vessels locally known as fiskija or dugum ibrik, but more commonly throughout Muslim parts of Bosnia as bardak, were known but not produced. Indeed, with the exception of tava, all of the names given above are exclusive to the Catholic population in Bosnia and to parts of Croatia. The pots were sold at weekly markets and seasonal fairs, or by peddling, as noted by Kalmeta.28 For purposes of sale it was made known to prospective buyers that pots in up to three distinct fabrics were available: tempered with calcite, sand or a mixture of the two, which were priced accordingly—those tempered entirely with calcite being the most expensive, thanks to their superior resilience when used in association with (p.425) heat. An alternative means of sale was to merchants who came by horse and cart to buy directly from source and sold on by the merchants to the ‘Muslimani kraj Krajina’, the Muslim part of the Krajina region—principally the Bihaćka Krajina.29 The pottery-making locations of Brezova Kosa and Skokovi are conjoining settlements in an area of relatively densely populated undulating hills north of Cazin, where the population is entirely Bosnian Muslim. Pottery making was described at both Brezova Kosa and Skokovi by Popović,30 at which time both were still active, and one potter, Husejn Pajić was still occasionally active at Brezova Kosa up to about 2000. Pottery making there was similar in most respects to Vrkašić, although it appears that only crushed calcite was used as a tempering material. The forms of pottery made were largely similar to those of Vrkašić and other ‘hand-wheel’ potters in Bosnia and south Croatia, although the range produced, at least latterly, appears to have been more restricted than elsewhere. The names given to the vessels tend to be Bosnian Muslim equivalents of the ‘Catholic’ names given at Vrkašić, although most potters regardless of ethnic or cultural affiliation are aware of the alternative names used (and asked for) by consumers within their area of trade. The only pottery-making centre operating in the Croatian Vojna Krajina was Gornji Lađevac/ Arapovac, 5 kilometres south-east of Slunj in central Croatia, 9 kilometres from the Bosnian border. Pottery making here was in all respects similar to the tradition at Vrkašić and Brezova Kosa but is poorly recorded, having ceased around the Second World War. Nothing is known about the trade of pottery from this centre, but the relatively sparse local population suggests that cross-border trade would have been necessary to support the industry there. Non-local Sources of Pottery

The Bihaćka Krajina lies at the centre of an exceptionally diverse range of pottery-making traditions, a number of which are known to have been active in the area until well into the second half of the twentieth century. Pottery-making traditions of similar character to those described at Vrkašić and Brezova Kosa operated elsewhere in Bosnia, most locally around Sanski Most, some 65 kilometres east of Bihać. These and other potters in Bosnia, however, seem to have had little impact on the Krajina; even the decorated fine-wares originating from Liješevo and Višnjica west of Sarajevo, otherwise widely traded throughout Bosnia, do not seem to have reached the Krajina in significant quantities. Much of the non-local pottery reaching the Krajina, certainly in the nineteenth and twentieth centuries, came in from Croatia.

Page 14 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland (p.426) Numerically the most important of the Croatian ‘hand-wheel’ centres, and the source which continued for the longest period, was the Lika region south-west of Bihać. Here were a number of pottery-making centres in two main concentrations, one at Sinac, nine kilometres south-east of Otočac, the other at and around Kaluđerovac, including the villages of Klanac, Oteš, Gornji Košinj and others some 10 to 15 kilometres north of Gospić. Near the end of the nineteenth century some 193 potters were recorded in official statistics at Kaluđerovac, Gornji Košinj and Klanac,31 but this is likely to have been something of an underestimate. Gavazzi described the industry at Kaluđerovac in his Kaluđerovacki Lončari, and brief accounts also occur in the works of Vojnović, Ostrić and Kašpar.32 It is similar in most respects to Vrkašić, including the use of sand and calcite as an additive to clay, but it is perhaps significant to note that a wider range of pots was produced there and there was limited use of lead glaze by some potters. It is known from various accounts that potters from Kaluđerovac and neighbouring villages travelled extensively to peddle their pots along the Dalmatian coast, northwards to Karlovac and the vicinity of Zagreb, eastwards into Slavonia and northern Bosnia and extensively in north-west Bosnia—where the dense rural population provided a source of high demand well into the modern period. When interviewed in 1994, the potter Grga Župan of Brezova Polje supported Gavazzi’s claim that potters travelled beyond the Krajina as far as Sarajevo in central Bosnia. Other large concentrations of pottery-making villages using essentially the same, hand-wheel/ open firing tradition as Vrkašić and Kaluđerovac existed west of the Bihaćka Krajina on the south-east side of the Kupa river at Liplje and neighbouring villages, including Milani, Johi and Glavica north of Ogulin. Pottery making seems to have ceased in most of these villages by 1900, but continued at Liplje until 1947–8; local information indicates that, until at least the First World War, pots were traded into eastern Slavonia and Bosnia. Thus, although there is no specific evidence that pots from here reached the Bihaćka Krajia, there is a strong possibility that they did. Other potential sources included Oštarije, 6 kilometres south-east of Ogulin in north-west Croatia, where pottery making ceased in the early twentieth century, and the villages of Veselići and neighbouring Goli Vrh, 13 kilometres south-south-west of Metlika on the east bank of the Kupa river, where Kaspar photographed at least four ex-potters in 1985.33 All of the above sources of pottery are known to have been active from at least the midnineteenth century and it is likely that their present distribution extends back at (p.427) least as far as the early eighteenth century, when the borderline stabilised following the Hapsburg conquest of Lika. Whatever their most recent distribution, however, the regional presence of potters working in this tradition is much earlier. In addition to the large number of potters operating out of the centres mentioned above, potters from various ‘foot-wheel’ traditions also penetrated the Krajina for trade purposes. Such traditions are characterised by their use of foot-powered wheels and double-chambered updraught kilns, producing sandy or fine fabrics which are often glazed, but while such pots are harder than the lower-fired products of the hand-wheel/open-firing tradition, they tend to be less durable as cooking pots. Most centres of ‘foot-wheel’ pottery making lie outside the Krajina, and the scale of their involvement with the Bihaćka Krajina remains largely unknown prior to the twentieth century. Thus, it is unknown whether potters from the Zagorje region of Croatia, from Dolenja Vas near Ribnica in Slovenia,34 and at Rastoki near Jastrebarsko traded into the Krajina, but it is highly likely that they did so, albeit perhaps intermittently. Interestingly, however, information gained from Kompolje, the only centre of foot-wheel production within the zone traditionally occupied exclusively by potters using hand-wheels and open (or single-chamber)

Page 15 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland kilns, suggests that while sale had been to Rijeka and Crikvenica on the coast and as far as Josipdol, towards Karlovac, in the north, it did not extend to Bosnia.35 Archaeological evidence is undoubtedly the key to determining the sources of such pottery traded into the region during the Ottoman period. Undoubtedly the most significant of the foot-wheel potters in the post-Second World War period were urban potters in Bosnia and north Croatia working in the Pirot tradition of western Serbia.36 Such potters spread from around the town of Pirot in east Serbia from the end of the nineteenth century to towns throughout Serbia, Bosnia and Croatia, eventually (as late as 1977) extending as far as Gvozd (formerly Vrginmost) in the Croatian Krajina. It seems that such potters formerly existed in the majority of large Bosnian towns and many smaller ones, including Brčko, Prijedor, Sarajevo, Zvornik, Gračanica and Doboj (the last two are still active), and in Croatia they were present in many Slavonian towns, surviving into the present century only at Pleternica. All share essentially similar working practices, including foot-driven or electrically (p.428) powered wheels and double-chambered kilns within developed workshop complexes, suggesting a greater tendency towards full-time professionalism than is exhibited by the traditional country potters of Bosnia and Croatia, whose work was traditionally part-time and seasonal. The assemblage of pots produced at Vrginmost in 1990 (Figure 21.12) was typical in some respects of pottery of the Pirot tradition elsewhere, but had been considerably modified by local demand, an observation also made independently by Kašpar: ‘They brought to our parts their characteristic shapes of pots and became a very strong competition to native potters from Rastoki. There was a mutual taking over of various pottery shapes from both sides.’37 This observation chimes with one of the main conclusions of a pilot study into Figure 21.12. Potter from the Jontić family the purchase and use of pottery vessels in photographed at Vrginmost (now Gvozd), various parts of the survey region, revealing Croatia in 1990. distinct variations in the names given to functional types of vessels by consumers in different areas. Thus, as indicated in the case of Vrkašić (p.429) (above), potters (and other traders) in order to be successful need to be sensitive to the types of pottery used by local consumer groups as well as the locally and ethnically distinctive terminology used by them. In the case of the Vrginmost potters, sales were carried out at widespread regional fairs in Bosnia, Croatia and western Slovenia in the late summer and early autumn period, and some forethought would be required in order to match local demands with products offered for sale.

Conclusions The examination of pottery held by households in the region and found in abandoned houses recorded as part of the vernacular buildings survey reveals a variety of pottery types corresponding to the known range of potential sources on both sides of the present BosnianCroatian border. Similarly, the pottery derived from archaeological excavations carried out at Pećigrad attests to the diversity of pottery traded in the region during the nineteenth and early

Page 16 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland twentieth centuries, with both hand-wheel and foot-wheel traditions apparently represented by several different fabric groups, and ceramic smoking pipes adding to the diversity of products and suggesting a potential avenue for more refined dating of such diverse assemblages. Further work is required to elucidate the pattern of trade represented by these finds. Indeed, archaeology provides the greatest potential to determine the sources of pottery at various phases of the active military frontier and, by implication, to deduce wider patterns of contact and trade within the frontier province of the Ottoman Empire and across its borders. It is hoped also that the extension of research into the purchase and use of pottery vessels across the survey region will allow the marketing trends of producers and consumers to be plotted spatially, thus compared with social groupings or territories suggested from other sources. The ethnographic evidence provides significant insights into the continuing legacy of the Ottoman-Hapsburg frontier in the nineteenth and twentieth centuries. The relative absence of craft manufacturing traditions such as those found in Central Bosnia may reflect the insecurity of the borderlands, as may the portable nature of pottery technology and the ephemeral nature of vernacular structures. While some traditions, such as methods of farming, pottery making and, to a large extent, speech are universal within the region, others selectively signal cultural distinctions. Thus, differing cultural traditions affecting matters as basic as the form of houses betray the existence of a fractured society, populated by different ethnic or confessional groupings. The predominant cultural grouping in the Bihaćka Krajina during recent centuries has been Muslim. A continuing frontier mentality and warrior ethos on the part of the Muslim adult male population can be glimpsed through the carved motifs which decorate the funerary stelae erected to commemorate individual members of that community. The challenge for any future programme of research is both to amplify that picture of life (p.430) in the recent past through further ethnographic study and, by means of archaeological investigation and documentary survey, to chart the various historical processes resulting from the region’s prolonged existence as a frontier zone of the Ottoman Empire, which led to the creation of that ethnographically documented society. Note. Fieldwork was undertaken by researchers from the Archaeological Practice, the University of Newcastle upon Tyne, English Heritage and the Zemaljski Muzej, Sarajevo with the initial assistance of a grant from the University of Newcastle upon Tyne Excavation and Fieldwork Committee. We are grateful for the advice and assistance of staff at the UN Mines Awareness Centre (MAC) office at Bihać. We would also like to thank the staff at Bihać Museum for their advice and Djenana Butorović and Zilka KujundzićVejzagić at the Zemaljski Museum, Sarajevo, for arranging all necessary permissions. Above all we are grateful for the support and continuing assistance of our principal collaborator from the Zemaljski Museum, Sarajevo, Mr Mirsad Sijarić, and we must also record our appreciation of our hosts, Remiza and Sulejman Nezirović of Mala Kladuša, and Mr Asim Rasić and the other inhabitants of Pećigrad for the copious quantities of coffee and detailed information generously provided. Notes: Proceedings of the British Academy 156, 403–429. © The British Academy 2009. (1) Wendy Bracewell, The Uskoks of Senj: Piracy, Banditry and Holy War in the Sixteenth-Century Adriatic (Ithaca and London: Cornell University Press, 1992).

Page 17 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland (2) Richard Carlton and Alan Rushworth, ‘The Croatian and Bosnian Vojna Krajina (Military Frontier)’, Universities of Durham and Newcastle Archaeological Reports, 22 (1998, pub. 1999). (3) Gunther E. Rothenberg, The Austrian Military Border in Croatia, 1522–1747 (Urbana, Illinois: University of Illinois Press, 1960). (4) Gunther E. Rothenberg, The Military Border in Croatia 1740–1881: A Study of an Imperial Institution (Chicago: University of Chicago Press, 1966). (5) For palankas see Chapter 8 above, Burcu Özgüven, ‘Palanka forts and construction activity in the late Ottoman Balkans’, 171–87. (6) The history of Pećigrad castle is summarised by Ćiro Truhelka, Naši Gradovi: Opis Najljepših Sredovječnih Gradova Bosne i Hercegovine (Sarajevo, 1904), 55–8. (7) Ibid., 57; Milan Kruhek, Krajiške Utvrde i Obrana Hrvatskog Krajevstva Tijekom 16. Stoljeća (Zagreb: Institut za Suvremenu Povijest, Biblioteka Hrvatska Povjesnica, 1995), 268. (8) Madžida Bećirbegović, Džamije sa Drevenom Munarom u Bosnii i Hercegovini (Sarajevo: Veselin Maslesa, 1990), 130. (9) Truhelka, Naši Gradovi, 56–7. (10) Brendan O’Shea, Crisis at Bihać: Bosnia’s Bloody Battlefield (Stroud: Sutton Publishing, 1998), 27, 73, 85–6. (11) Bećirbegović, Džamije sa Drevenom Munarom, 170; plates 131, 191 and 192. (12) The castle at Bužim (Ott. Büzin) may provide a useful type site, incorporating several masonry styles on its inner circuit which can be phased. Undressed rubble facing was employed in the lower section of the east curtain wall and associated north-east and south-east corner towers. These were later heightened using coursed squared stonework, composed of the distinctive local and easily workable yellow limestone known as sedra, which, perhaps because of its exceptional lightness as a material, was cut into much larger blocks. The later west curtain, on the other hand, features much smaller coursed stonework, more comparable to that at Pećigrad and Stijena. However, there is no guarantee that a similar chronology is applicable to either of those sites. (13) Mehmed Mujezinović, Islamska Epigrafika Bosne i Hercegovine (Sarajevo: Veselin Masleša, 1982), 3: 78–9. (14) Džafer Mahmutović, Stari Bihać (1260–1940) (Bihać, 2001), 41, reproduces photographs of three examples. (15) For mali rat see Sanja Lazanin and Nataša Štefanec, ‘Habsburg military conscription and changing realities of the Triplex Confinium (sixteenth-eighteenth centuries)’, in Drago Roksandić and Nataša Štefanec (eds.), Constructing Border Societies on the Triplex Confinium (Budapest: Central European University, 2000), 98. (16) William G. Lockwood, European Muslims: Economy and Ethnicity in Western Bosnia (New York, San Francisco and London: Academic Press, 1975); Tone Bringa, Being Muslim in a

Page 18 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland Bosnian Way: Identity and Community in a Central Bosnian Village (Princeton: Princeton University Press, 1995). (17) The terminology adopted for the main ethno-religious groupings here follows the modern nationality and expressed religion of the individuals/groups concerned. Thus Bosnians of Muslim faith are referred to as ‘Bosnian Muslim’. The term ‘Bošnjak’, which some Bosnian Muslims (and Muslim Slavs from the Sanjak of Novi Pazar) apply to themselves, is not used because it is a term which could be, and once was, used for Bosnians in general. Similarly, the terms ‘Croat’ and ‘Serb’ are not used here in place of ‘Bosnian Catholic’ and ‘Bosnian Orthodox’ because they are terms which have come into use only relatively recently and do not allow us to draw a validethnic or ethnographic distinction between peoples of different religion, sharing, among their various attributes, aspects of the long heritage derived from common residency in Bosnia. (18) Muslims living on the Croatian side of the border within the municipality of Cetin, of whom there are currently over 400, travel to Kladuša in Bosnia for schooling, and are joined there by Catholics from Bosnia for secondary schooling. (19) See Amir Pašić, Islamic Architecture in Bosnia and Hercegovina (Istanbul: IRCICA, 1994), 137–8. Some towers served as official residences, for example the Kapitanova Kula in Bihać. (20) Persida Tomić, Grnčarstvo u Srbiji (Belgrade: Etnografski Muzej u Beograd, 1983). (21) For references to foot-wheel potting in northern Croatia see: V. Pinter, ‘Lončarstvo u Jerovcu’, Vjesnik Etnografskog Muzeja u Zagrebu, 1 (1935), 69–78; L. Kačpar, ‘Lončarsko rukotvorstvo u selu Liplje u Gorskom Kotaru’, Vijesti Muzealaca i Konzervatora Hrvatske, 3/4 (1978), 58–62. For Zagreb: idem, ‘Piročanski Pečalbari u Vrginmostu’, Godičnjak Hrvatskog Etnoločkog Društva, Etnoloska Tribuna, 2 (1979), 89–91; M. Randić-Barlek, ‘Prehrana’, in I. Magdalenić, M. Vranesić, and M. Zupančić (eds.), Zumberak: Bastina i Izalovi Buducnosti (StariGrad Zumberak, 1996), 67–98. For northern Bosnia see Richard Carlton, ‘The Pirocanci pottery-making tradition in former Yugoslavia’, The Studio Potter, 27 (1999), 98–9. (22) Kašpar, ‘Piročanski Pečalbari u Vrginmostu’, 89–91; Tomić, Grnčarstvo u Srbiji; Carlton, ‘The Pirocanci potterymaking tradition’, 98–9. (23) Kašpar, ‘Piročanski Pečalbari u Vrginmostu’. (24) A. Muraj, ‘Lončarski centar u Rastokima kraj Karlovca’, Zbornik Gradskog Muzeja u Karlovacu, 1 (1964), 126–41; D. Cvetan, Tradicijski lončarski centri u plešivičkom prigorju (katalog izlozbe) (Jastrebarsko, 1985); idem, ‘The living tradition of arts and crafts in the Jastrebarsko region’, 12th International Conference of Anthropological and Ethnological Sciences (International Union of Anthropological and Ethnological Sciences), Zagreb, Yugoslavia, 24–31 July 1988 (Jastrebarsko: Zavicajni Muzej, 1988), 25–32. (25) See R. J. Carlton, ‘Exploring the role of quenching in the production of calcite-tempered pottery in the western Balkans’, Joint Newsletter of the Prehistoric Ceramic Research Group and the Ceramic Petrology Group, 10 (2002), 1–6; idem, ‘Paste and fire: some comments on the technology of prehistoric pottery in the Western Balkans in the light of ethnoarchaeological research’, in D. Georgiou (ed.), The Archaeology of Fire (London: British Archaeological Reports, 2002), 63–81.

Page 19 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Krajina Project: Exploring the Ottoman–Hapsburg Borderland (26) A. Kalmeta, ‘O seljačkom lončarstvo u srednjoj i zapadnoj Bosni’, Glasnik Zemaljskog Muzeja u Sarajevu (Etnologija), n.s. 9 (1954), 144–8. (27) This may have been done in order to make the fabric less porous, or for the blackening effect, but experimental evidence suggests that the process compensates for partial over-firing. (28) Kalmeta, ‘O seljackom lončarstvo’, 148. (29) Small amounts of pottery from Liješevo in Central Bosnia continue to be sold in the Krajina by this means—such traders were observed near Ostrožac in summer 2007. (30) C. Popović, ‘Lončarstvo u Bosni i Hercegovini II’, Glasnik Zemaljskog Muzeja u Sarajevu (Etnologija), n. s.12 (1957), 32. (31) M. Zoričić Žiteljstvo, Kraljevina Hrvatske i Slavonije po zvanju i zanimanju (RAD Jugoslavenske akademije znanosti i umjetnosti, Kniga CXXV. Filogijsko-historijski i filosofijskojuridički razredi, XLIV, Zagreb, 1896). (32) M. Gavazzi, Kaluđerovacki Lončari, Lički Zbornik, 1 (1978), 25–34; O. Ostrić, Narodno lončarstvo (katalog izlozbe) (Zadar: Narodni Muzej, Etnografski odjel, 1979); L. Kašpar, Lončarstvo Karlovačke Okolice (Karlovac: Gradski Muzej, 1985). (33) Kašpar, Lončarstvo Karlovačke Okolice, 16 and 18. (34) Pedlars from Ribnica, mainly carrying wooden wares but also some pottery, are known to have travelled in the later Middle Ages and well into the modern period to Germany, Italy, Spain and even across the Mediterranean into Africa. A photograph of one such pedlar in Germany is published by F. Mitija, Kočevska: Izgubljena Kultura Dediš ćina Kočevskih Nemcev (Ljubljana, 1993). Trade into Bosnia, then, seems likely. (35) Kraljevine Hrvatske i Slavonija na Tisućgodisnja Zemaljskoj Izložbi kraljevine Ugarske (Kingdom of Croatia and Slavonia on the thousand-year nationhood exhibition of the kingdom of Hungary) (Budapest and Zagreb, 1896), catalogue no. 7407; the potter Tomo Kranjčević exhibited several pots: vrč (jug), padela (pouring dish for milkpreparation), ćup za mast (storage pot for fat), tanjur (plate), lonac za mlijeko (milk pot), zila (bowl), plant pots and an ink pot. (36) Carlton, ‘The Pirocanci pottery-making tradition’. (37) Kašpar, Lončarstvo Karlovačke Okolice, 15.

Access brought to you by:

Page 20 of 20 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18 NEIL FAULKNER NICHOLAS J. SAUNDERS

DOI:10.5871/bacad/9780197264423.003.0022

Abstract and Keywords The Arab Revolt of 1916–18 played a significant part in the military collapse of the Ottoman Empire at the end of the First World War. This chapter argues that archaeological evidence indicates that the revolt's importance was probably substantially greater than has sometimes been acknowledged. The evidence demonstrates the need for a critical re-evaluation of the issue in southern Jordan. The archaeological investigation of sites associated with the Arab Revolt in southern Jordan offers dramatic insights into the material consequences for the Ottoman army of combating the guerrilla tactics of British-backed Arab guerrillas. The aim of the discussion is twofold: to give more precision to the military assessment of the Arab Revolt in the area between Ma'an and Wadi Rutm, and to demonstrate the potential of the new and multidisciplinary sub-discipline of twentieth-century ‘conflict archaeology’. Keywords:   Arab Revolt, Ottoman Empire, First World War, southern Jordan, Wadi Rutm, conflict archaeology, guerrilla tactics

THE ARAB REVOLT OF 1916–18 PLAYED A SIGNIFICANT PART in the military collapse of the Ottoman Empire at the end of the First World War. Just how significant is disputed, though we argue here that archaeological evidence indicates that the revolt’s importance was probably substantially greater than has sometimes been acknowledged. At the very least, our evidence demonstrates the need for a critical re-evaluation of the issue in southern Jordan.1 Embedded in the complex geopolitical prosecution and consequences of the war, the revolt can be seen as having contributed in both subtle and overt ways to shaping the post-war Middle East and generating conflicts that continue in some form to the present day.2 The celebrity cult of T. E. Lawrence (i.e. as Lawrence of Arabia)—a partial construct of the American impresario Lowell Thomas in 19193—has meant that the event has been of popular interest for ninety years. The result has sometimes been a perhaps overly dismissive reaction from historians, many of whom have been unwilling to engage seriously with the revolt and Lawrence’s role in it as valid subjects of historical enquiry4 Military historians and theorists, by Page 1 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18 contrast, have often been more positive, (p.432) sometimes regarding Lawrence as a seminal figure in the development of modern guerrilla warfare and special-forces operations.5 The archaeological investigation of sites associated with the Arab Revolt in southern Jordan adds a new dimension to these debates, and offers sometimes dramatic insights into the material consequences for the Ottoman army of combating the guerrilla tactics of British-backed Arab guerrillas. Our aim here is twofold: to give more precision to the military assessment of the Arab Revolt in the area between Ma‘an and Wadi Rutm, and to demonstrate the potential of the new and multidisciplinary sub-discipline of twentieth-century ‘conflict archaeology’.

The Archaeology of Modern Conflict The Great Arab Revolt Project, in southern Jordan has several aims: (a) to investigate the material implications of the historical and cultural issues referred to above; (b) to extend the boundaries of the emerging sub-discipline of First World War archaeology;6 and (c) to demonstrate its connections to the wider, and equally new, archaeology of twentieth- and twenty-first-century conflict.7 Hitherto, First World War archaeology had focused almost exclusively on the battlefields of the Western Front in northern France and Belgium,8 where it has developed as a distinctive multidisciplinary endeavour, whose methodologies are still in their infancy, but whose potential as an anthropologically informed ‘archaeology of the recent past’ is matched only by its ethical responsibilities to all those who were (and continue to be) affected by the world’s first industrialised global conflict. What has become clear from First World War archaeological investigations along the Western Front and from our work in southern Jordan is that the hitherto widely used term ‘Battlefield Archaeology’ is wholly inappropriate and inadequate. Tainted by the excesses of battlefield scavenging, indiscriminately applied to all battlefields from medieval times to the present, and deployed as a theoretically impoverished adjunct to military history, Battlefield Archaeology is unable either to conceptualise modern (p.433) conflicts, or offer analytical purchase to the many complex issues that confront those who investigate them. First World War archaeology, and conflict archaeology more widely, are hybrids. They draw on Industrial Archaeology—which tends to be rich in mass-produced artefacts—and on Historical Archaeology, defined by access to a wealth of written documents. They also incorporate aspects of Social Archaeology and Public Archaeology, and have been deeply influenced by the renewed anthropological interest in material culture.9 Above all, these new and conjoined endeavours investigate the ‘archaeology of the recent historical past in time of war’, and are characterised equally by cross-cutting interests in the study of landscape, cultural memory, museums, heritage and tourism.10 All of these issues inform the Great Arab Revolt Project, including the results of the initial archaeological reconnaissance and excavation presented here. Conflict archaeology does not seek to confirm or deny previous interpretations, but rather to draw on the insights and expertise of various disciplines (including traditional archaeology), as appropriate, to present a broader and more nuanced appreciation of the course and consequences of modern conflicts. Many of the battle-zones of the Arab Revolt in southern Jordan are uniquely informative in this regard. While they belong to the new age of industrialised war, they are characterised by short-lived, one-off military encounters more akin to pre-1914 conflicts. Such battle-zones represent a unique turning point in the study of modern conflict—how a conventional army seeks to contain the real and perceived threats of guerrilla warfare in the context of a global industrialised war by large-scale transformations of the landscape. These transformations remained largely unknown and unacknowledged until 2006, Page 2 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18 despite their re-use and mythification by subsequent generations of Bedouin, their potentially lethal effects on the unwary, and their potential value for heritage and tourism.11 Such a multidisciplinary approach characterises the Great Arab Revolt Project. Conflict landscapes illustrate the project’s anthropological dimension as they are cultural artefacts tied to a sense of personal and national identity, and are subject to changing attitudes towards war and memory. Developing a modern archaeology of the Arab Revolt will inevitably throw a different light on the events of 1916–18, but it also has the potential, through the analysis of physical remains, to re-shape perceptions of the historical events, the individuals involved, and people’s real and imagined relationships with landscape and war, nationhood and boundaries.12 The Arab Revolt offers both a new dimension in First World War archaeology (and twentieth- and (p.434) twenty-first-century conflict archaeology more widely) and an opportunity to investigate a conflict of great historic importance and contemporary relevance in its own right.

The Great Arab Revolt Project After small-scale reconnaissance in January 2005 and February 2006, intensive field investigations began in November 2006.13 The presence of First World War archaeology was confirmed by field reconnaissance at Guweira, Abu ’l-Lisan, and several Hijaz railway stations, but at only two sites did we carry out systematic fieldwork (Figure 22.1). One of these is the major Hijaz railway hub which lies immediately south-east of Ma‘an, the largest town in southern Jordan, and which shares the same name. Ma‘an station is characterised by a complex of well-preserved late Ottoman buildings, while on the surrounding hills and ridges are extensive surviving trench-lines and other defensive features. The other site is an abandoned Hijaz railway station at Wadi Rutm, 60 kilometres south of Ma‘an, where three late Ottoman station buildings with signs of improvised fortification lie ruined in a stark desert location. (p.435) Our purpose here, on the basis of our preliminary work in the field, is threefold. First, to place the Arab Revolt in the wider context of the terminal crisis of the Ottoman Empire. Second, to identify and define the material traces of the hitand-run tactics employed by the Arabs. Third, to suggest how this kind of archaeology can contribute to an understanding of the collapse of Ottoman control between 1916 and 1918 in a conflict that can be characterised as a ‘war without frontiers’ and a ‘war of detachment’, where the unquantifiable threat of guerrilla attack produced an Ottoman response predicated on the fortification and occupation of landscapes that never became battlefields, but that necessarily involved considerable time and expense, and tied down large numbers of men.

Page 3 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18 The Ottoman Empire Goes to War

Figure 22.1. Map of southern Jordan From 1908 onwards, the Ottoman Empire showing the location of Great Arab Revolt Project sites. was in the throes of a profound political crisis. The ‘Young Turk’ revolution of 1908– 9 overthrew the absolute monarchy and led to the establishment of an authoritarian bourgeois-nationalist regime. This regime was committed to the development of the empire as a capitalist economy and a great power, but was crippled by the relative backwardness of Ottoman industry and infrastructure, and threatened with loss of territory to both foreign aggression and nationalist revolt. The decision of the Young Turk leaders to align themselves with Germany in November 1914 was a response to the empire’s weakness, semi-colonial status and internal instability. The imperatives of both economic development and political independence made an alliance with the greatest industrial and military power in Europe desirable. Germany had no interest in acquiring Ottoman territory, yet had become a major investor in infrastructure projects, notably railway construction, and had been instrumental in the modernisation of the army, providing expertise, training, and equipment.14 By contrast, Britain, France and Russia all had actual or potential designs on (p.436) Ottoman territory, while the British and the French were the empire’s principal creditors and controlled a third of its tax revenues.15 At the same time, the 1908–9 revolution had stirred powerful internal forces. Virtually the whole of the small Ottoman working class had participated in a major wave of strikes, and many of the non-Muslim/non-Turkish subject peoples of the empire, in particular the Christian Armenians, had been involved en masse in tax-strikes, demonstrations and armed clashes.16 The politics of Turkish nationalism and imperialist war offered an ideological mechanism for marginalising these forces, centralising political authority and consolidating the rule of the leaders of the Committee of Union and Progress (CUP). The First World War was primarily an industrialised war between two alliances of great powers. The Ottoman Empire, by contrast, was a predominantly agricultural society ruled by traditional elites in the framework of an essentially pre-modern state bureaucracy; it was changing, but in most sectors modernising reforms were yet to achieve any real social and cultural transformation. Communications across the great distances of the empire, though much improved, were still far below the demands of modern war. Railways were few, roads poor, and the sea-lanes were dominated by hostile fleets. From Istanbul it took more than a month to reach Syria and almost two to reach Mesopotamia.17 The empire’s difficulties were further compounded by strategic mistakes. Pan-Turkish nationalism inspired a disastrous offensive in the Caucasus.18 Determination to uphold the Ottoman Caliphate meant that much effort was expended on retaining the holy city of Medina, and hope of igniting Muslim revolt against British colonial rule led to an attempt on the Suez Canal. These efforts drained scarce resources away from hard defensive struggles elsewhere—in Gallipoli, Mesopotamia and Syria-Palestine. The First World War climax of the Ottoman Empire’s crisis is the political and military context for the Arab Revolt of 1916–18. Sharif Husayn, the Amir of Mecca, had long harboured ambitions to break free of Ottoman rule and make himself the independent ruler of an Arab state. These ambitions were strengthened by the increasing oppressiveness of Ottoman rule in the Arab territories, which suffered from tax rises, military conscription, wartime shortages and political repression under the Young Turk regime. They were also encouraged by promises of British military and financial aid, and the apparent prospect of British support for independent

Page 4 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18 Arab rule in the (p.437) post-war Middle East. The immediate catalyst for revolt was a strengthening of Ottoman security measures in the Hijaz. The course of the revolt fell into four distinct phases. Between June 1916 and January 1917 the revolt remained confined to the Hijaz, trapped in a military impasse, and threatened by a large Ottoman garrison in Medina. The capture of Wajd in January 1917 opened the second phase of the campaign by exposing Medina’s railway communications to repeated attack and destroying its garrison’s strategic mobility. The capture of ‘Aqaba in July 1917 began the third phase, extending operations into Syria, where the war continued until September 1918. This phase consisted, on the one hand, of widespread, endemic, low-level guerrilla activity across much of southern Syria (i.e. modern Jordan), and on the other, usually in tandem with major British offensives in Palestine, occasional deep-penetration raids to attack Ottoman communication lines. These operations involved four distinct elements: (a) camel-borne Bedouin guerrillas, many giving long-term, though erratic, service to the Hashemite cause; (b) Arab regulars of the embryonic Hashemite army; (c) Syrian peasants organised on a tribal and village basis, who could be mobilised locally and occasionally when circumstances were opportune; and (d) various British, Indian and French specialist units, including demolition experts, machine-gunners, mortar crews, mountain-gunners, camel corps, armoured cars and aircraft. It is this long third phase that is our primary focus here. The fourth and final phase of the revolt occurred with the collapse of the Ottoman forces in Palestine and Syria, and the joint British and Arab advance to Damascus and Aleppo in September-October 1918.19

Lawrence’s Conception of Guerrilla Warfare T. E. Lawrence is one of the most controversial British public figures of the last ninety years, the subject of numerous books and articles. Much of the output is popular, romanticised, The way in which different vional. The way in which different versions of the Lawrence legend have been reconstructed and de-constructed in that time is a subject for study in itself20 Here, we are not concerned with this issue, nor with debates about his personality, reliability as a historical source, or leadership role in the Arab Revolt. As our focus is on the archaeological investigation of conflict landscapes in southern Jordan, our concern with Lawrence is his contribution to military theory. This focus is timely, as the recent successes of guerrilla fighters in Afghanistan, Iraq and Lebanon have revived interest in Lawrence in western military academies.21 (p.438) We argue that Lawrence should be considered a seminal figure in developing the theory of modern guerrilla warfare. While its reconfiguration in terms of modern technological warfare is mainly a twentieth-century development, guerrilla warfare in principle is as old as history: it is the instinctive response of rebellious people fighting against a more powerful opponent.22 But, as far as we are aware, there is no surviving substantive piece of theoretical writing outlining the principles of guerrilla warfare that pre-dates the work of Lawrence. Since his time a number of treatises have been written, often by leading practitioners of guerrilla warfare like Mao Tse-Tung, Vo Nguyen Giap and Ernesto Che Guevara,23 as well as a number of important academic studies.24 But there appears to be nothing that is earlier. ‘Suppose’, Lawrence wrote, ‘we were an influence (as we might be), an idea, a thing invulnerable, intangible, without front or back, drifting about like gas? Armies were like plants, immobile as a whole, firm-rooted, nourished through long stems to the head. We might be a vapour, blowing where we listed.…Our war should be a war of detachment: we were to contain the enemy by the silent threat of a vast unknown desert…’25 At the time—the time, that is, of

Page 5 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18 months-long pitched battles like the Somme and Passchendaele on the Western Front—this was an astonishingly radical conception of war to come from a serving British officer. Lawrence’s earliest published ideas on guerrilla warfare appeared on 20 August 1917 in an article entitled ‘Twenty-seven articles’, in The Arab Bulletin, an internal publication produced by wartime British intelligence in Cairo. Only after the war did he publish a full statement of his military principles, first in an article entitled ‘The evolution of a revolt’, in The Army Quarterly (1920), then in chapters 35 and 65 of the 1922 Oxford edition of Seven Pillars of Wisdom (with chapter 35 later becoming chapter 33 of the more widely available 1935 London edition), and finally, jointly with the military historian Basil Liddell Hart, as an entry entitled ‘Science of guerrilla warfare’, in the 1929 edition of the Encyclopaedia Britannica.26 (p.439) Lawrence distinguishes between rebellion and war—or ‘murder war’—arguing that actual fighting in the former is much less important than simply building mass support to isolate and pin down the enemy in their bases; ‘battles’, indeed, ‘were a mistake’. Careful reading allows one to discern thirteen distinct principles of war in Lawrence’s treatises (each of which covers essentially the same ground, often in the same words): 1. Strive above all to win hearts and minds. 2. Remain strategically dispersed. 3. Make maximum use of mobility. 4. Operate mainly in small, local groups. 5. Remain largely detached from the enemy. 6. Do not attempt to hold ground. 7. Operate in depth rather than en face (i.e. not in static lines). 8. Concentrate only to achieve momentary tactical superiority. 9. Never engage in sustained combat. 10. Always have lines of retreat open to safe havens. 11. Aim for perfect intelligence about the enemy. 12. Make war on matériel rather than on men. 13. Make a virtue of the individuality, irregularity and unpredictability of guerrillas. What is striking is how these principles break the paradigm of conventional, western-style, ‘Clausewitzian’ warfare, which, famously, is concerned with seeking out the enemy’s main forces, engaging them in pitched battle, and bringing about their annihilation, such that one can then impose one’s will on the enemy27 Lawrence was not a regular army officer trained in mainstream military theory. He was a wartime amateur, an intellectual, a romantic, and by instinct a rather mischievous maverick: all characteristics that perhaps enabled him to think ‘outside the box’. He had not even undergone basic training, having entered the army sideways to do specialist Middle Eastern intelligence work. Once in Arabia and working as a liaison officer attached to Amir Faysal’s Northern Army (from October 1916 onwards), he used his experience of traditional Arab warfare to turn Clausewitz on his head. Many of Lawrence’s principles are the very opposite of conventional military theory. Against decisive battle to destroy the enemy, he (p.440) urged avoidance of battle; against concentration of force, dispersion; against defence of territory, evasion. Elusive to the point of invisibility, the rebels would offer their enemies no targets to shoot at; yet, by their immanent presence, would leave the Turks in control of no more than the ground on which they stood. The

Page 6 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18 Turks would find that ‘to make war on rebellion is messy and slow, like eating soup with a knife’.28 Our archaeological investigations of the Arab Revolt’s conflict landscapes focused on three questions arising from Lawrence’s military theory. First, would the archaeological remains confirm that Lawrence’s conception of the war was what actually took place? Second, assuming that it was, what level of success in applying the conception would the remains indicate? Third, assuming that the remains conformed to the conception, what would be the archaeological imprint of such a ‘war of detachment’?

The Archaeology of Imperial Collapse 1: A Trench Fortress at Ma‘an Ma‘an and Wadi Rutm between them constitute a convenient initial sample of First World War conflict archaeology in southern Jordan (see Figure 22.1). The former was a major military base and trench fortress, the latter one of the seventy-seven stations located along the railway line between Damascus and Medina. Ma‘an therefore demonstrates the potential for conventional warfare around heavily manned and fortified fixed positions, while Wadi Rutm represents the potential militarisation of the landscape as a whole in relation to a dispersed and unpredictable guerrilla threat. Here we review the archaeological signature at each site and then draw some preliminary conclusions. Ma‘an was a major communications node. It lay on the ancient pilgrim route along the western edge of the desert that linked the cities of Syria to the north (Amman, Dar‘a, Damascus and Aleppo) with the holy cities of Arabia (Medina and Mecca) far to the south. It was the place where this pilgrim route intersected with important east-west communication lines. A major caravan route ran through the eastern desert to Ma‘an via the wells at al-Jafr, and other caravan routes connected Ma‘an with Petra and the King’s Highway in the Jabal al-Batra mountains to the west, and with the Red Sea port of ‘Aqaba to the south-south-west via the pass over the Jabal al-Batra range at Abu ’l-Lisan. Ma‘an was also blessed with a plentiful water supply, thanks to nearby limestone springs. For this reason, a substantial medieval town had developed at Ma‘an, with a strong Bedouin and commercial character, which later became the seat of an Ottoman district governor. When the Hijaz Railway reached Ma‘an in 1904, the town comprised about (p.441) 500 flat-roofed mud houses 1.5 kilometres north of the station, and another 200 or so one kilometre further north again. A large station complex was established, and Ma‘an came to be regarded as the ‘Gateway to Arabia’.29 Control of Ma‘an provided the Ottoman army with a strong base for the accommodation of soldiers, matériel and supplies that could be moved rapidly in any of several directions: by train north or south to support station garrisons under threat; by road westwards into the relatively rich, heavily populated, grain-producing uplands beyond the desert fringe; and again by road south-south-west to guard the route to the Red Sea coast. Between 11 and 20 April 1918, when Ma‘an was attacked by Arab forces, the garrison comprised 4,000 men of the Fourth Army’s Second Corps commanded by Wahid Bey. The attack involved cutting the main garrison’s communications to north and south, forcing the outer defences south-west of Ma‘an on Jabal Samna, and then launching a direct attack on the inner defences around the railway station itself. The attack on the station involved Hashemite regulars, Bedouin irregulars and a French artillery battery. The Arabs captured the station on 17 April, but were soon ejected by Turkish counter-attacks from the north and forced to fall back to Page 7 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18 Jabal Samna. Thereafter the Arabs avoided direct assaults on Ma‘an, which fell only in September 1918, when the garrison was cut off and chose to surrender.30 Many of the trench-works defending Ma‘an were recorded at the time on British airreconnaissance sketches and photographs. Some still survive on the ground today and have been recorded in recent air-photographic surveys, on satellite imagery, and through field reconnaissance and GPS survey, allowing us to map four groups of Ma‘an trench-works on the high ground west of the Hijaz Railway (Figure 22.2): 1. ‘The Hill of the Birds’ group comprises: a firing trench 725 metres long facing northwest; a second firing trench 500 metres long facing south-west; communication trenches serving these; and three hilltop redoubts, designated the Northern, Central and Southern Redoubts, of which the Northern Redoubt is the most substantial. Various other features have been noted on the Hill of the Birds, including four artillery emplacements and two associated command-and-control positions, which lie a short distance south of the Northern Redoubt. 2. ‘The Northern Ridge’ group lies 0.9 kilometres north of the Northern Redoubt and comprises a firing trench 500 metres long facing north-west, and two hilltop redoubts. A third redoubt, separate from the rest of the Northern Ridge trench-work, can be seen on a recent air-photograph immediately to the west. Field reconnaissance in February 2006 revealed that this earthwork had since been destroyed, probably by (p.442) military bulldozers to create dead ground around the perimeter of a modern Jordanian army base. A fourth redoubt and two further stretches of firing trench of 125 metres and 120 metres have been observed on satellite imagery 1.1 kilometres north-east of the Northern Redoubt, but have not yet been examined on the ground. 3. ‘The Southern Ridge’ group lies 1.2 kilometres south-west of the Northern Redoubt and comprises a firing trench 400 metres long facing Figure 22.2. Plan of the area west of Ma‘an south-west, and the layout for a station showing recorded Ottoman defence further trench, surviving as a slight works. © GARP. yet distinct earthwork which was never (p.443) cut. An additional section of trench has been observed on satellite imagery 2 kilometres south-east of the Northern Redoubt. This system faces south-east, is approximately 220 metres long, and is yet to be surveyed. 4. ‘The Western Hillock’ group lies 0.6 kilometres west of the Northern Redoubt and comprises a hilltop redoubt, and a communication trench up the north-east side of the hillock. Instead of continuous trench-lines, the Ottoman defences at Ma‘an comprised a series of upslope strongpoints, with firing trenches that ran just below the upper contours of hills and ridges, and redoubts on the hilltops (situated to combine the advantages of elevation without the risk that soldiers would be silhouetted against the skyline by attackers approaching from below). Individually these trench-works had excellent command of ground, and in some cases, in combination, offered intersecting arcs of fire. They also constituted a system of defence-in-depth —a series of upslope strongpoints radiating outwards from Ma‘an station, each capable of blocking the advance of an attacker. Moreover, the trench-works we have seen are only a

Page 8 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18 sample. We are already aware of other trench-works elsewhere, especially on the western approaches to Ma‘an, and these will be investigated in the future. Further detail was revealed by excavating a sample of the trenches forming the Northern Redoubt and the firing and communication trenches immediately west of it. These revealed two phases of development and an implicit change in tactical conception. First, there was a linear system based on a firing trench extending for about 725 metres along the western slope of the ridge. This was served by communication trenches which extended from the opposite, eastern slope of the ridge. The firing trench appears to have been roughly worked and very shallow, being cut less than 0.70 metres below ground level into deposits of gravel, compacted sand and soft limestone rock. A parapet bank had been piled up in front of it, however, and though this survived to a height of only about 0.20 metres, it may once have been a substantial defence. Even so, deep trenches offer better protection than shallow ones fronted by banks. Moreover, the firing trench was not especially well sited. In places it was overlooked by, and could have been enfiladed from, higher ground. However, it was dug with closely spaced traverses, which would have offered protection against both enfilade and direct blast. Overall, however, this defensive scheme was, at some point, deemed inadequate, and was superseded by a new layout. The nearby Northern Redoubt is defined by a ring-trench around the hilltop and a centrally placed command bunker. Excavation revealed that the main communication trench leading down to the linear firing trench had been blocked when the ring-trench was dug, cutting off access to the original defences, which presumably had become redundant. The new ring-trench was more substantial than the old linear one, in places measuring more than 1.5 metres in depth from the top of the surviving parapet bank to the base of the walkway, necessitating the provision of a 0.35 metre high fire-step in (p.444) the base of the trench. The parapet was solidly constructed of sandbags, mud bricks and large stones, and survived more than one metre high and one metre wide. High, deep traverses were placed on average every 5 to 6 metres between the firebays. As well as providing much more secure protection, without any risk of enfilade, the Northern Redoubt would have given defenders complete command of the slopes extending in a semicircle to the west. In combination with the Central and Southern Redoubts, between which and the Northern Redoubt there would have been intersecting arcs of fire, the whole western approach to the ridge would have been effectively covered. If our interpretation is correct, we have a phase of shallow linear trenching followed by a phase of deeply cut and heavily banked-up ring-trenches forming hilltop redoubts. The latter imply a more sophisticated tactical conception. Behind this may lie a change in the character of the war. Three factors may be pertinent: (a) the increasing threat posed by Arab forces; (b) a growing influx of German officers raising the professionalism of the Ottoman army; and (c) the manpower and logistical problems facing the Ottoman army putting a premium on economy of force.31 The change may, of course, have been a direct response to weaknesses exposed during the First Battle of Ma‘an in April 1918 (see above), though an alternative possibility is the occasion of Ma‘an’s reinforcement following the fall of ‘Aqaba the previous summer. Eugene Rogan reports that ‘the German General von Falkenhayn came to Ma‘an to advise on its reinforcement with 6,000 infantry, a regiment of cavalry and mounted infantry, and air cover provided by a flight of airplanes’.32 Probably the detailed evidence from the trenches on the Hill of the Birds should be read in combination with the results of our landscape surveys and desktop research. Many of the trenches and redoubts observed on high ground further afield may represent increased

Page 9 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18 defensive effort and sophistication in response to a growing military threat from the summer of 1917 onwards.

The Archaeology of Imperial Collapse 2: A Fortified Railway Station at Wadi Rutm Wadi Rutm is the site of an abandoned Hijaz railway station about 60 kilometres south of Ma‘an. The station lies towards the head of wide wadi with high, steep hills on either side. To the southeast the wadi widens and eventually terminates in an open desert plain which extends into north-western Arabia. To the north-west it ends in an escarpment, where the railway, having curled around from the north, descends through a narrow cutting from the wide plateau of southern Jordan. Above the escarpment, (p.445) about 10 kilometres from Wadi Rutm, lies the site of another (now destroyed) station, that of Batn al-Ghul. Our study area was originally limited to the immediate vicinity of the three ruined late Ottoman buildings that represent the site of Wadi Rutm station itself. The evidence for a more extensive militarisation of the landscape led us to expand our search beyond the station itself. A combination of desktop research, field reconnaissance, landscape survey, metal-detector survey, standing-building recording, and sample excavation allows us to summarise the First World War conflict archaeology of Wadi Rutm in the following terms (Figure 22.3): (p.446) 1. A probable traditional place of assembly, camping and trading represented by concentrated finds of 128 coins and various other small metal objects of ancient, medieval and later Figure 22.3. Plan of Wadi Rutm showing recorded late Ottoman sites. © GARP date. The site has been metal-detected and its location and approximate limits defined. 2. An old caravan and pilgrim road—paved and covered with bitumen (now almost disappeared), probably during the 1930s for motorised transport—running parallel with the later railway embankment and the modern desert highway through at least part of the wadi. This road has not been traced along its entire length, but it appears to branch into two as it approaches the traditional campsite. 3. The embankment of the Hijaz railway, which can be seen running the length of the wadi, with a substantial siding and passing place at Wadi Rutm station. All rails and sleepers have been removed, though a rail section, several sleepers, and numerous fishplates, bolts and other railway engineering artefacts have been found on the site. 4. Three late Ottoman railway buildings at the station site itself, which we have numbered 1, 2 and 3 from north-west to south-east. All are ruinous and roofless, though 1 is otherwise fairly well preserved, 2 is largely demolished with only one wall and a corner surviving, and 3 has substantial wall survival. These three buildings stand on what appears to be a siding and passing place on relatively low-lying ground. A record has been made of all three buildings, including a detailed one of Building 1, and various improvised defensive features have been noted, including loopholes hacked through walls, crude stone rooftop parapets, and mud-brick extensions (though none of these features can be dated with confidence). 5. A defended Ottoman army camp on a commanding flat-topped hill 1.1 kilometres north-west of the station. The hill, which is about 150 metres at its widest, dominates the railway to the north-east and the road through the wadi to the southwest. The main archaeological characteristics of the site are: firing dugouts with stone parapets at either end of the hill; at least twelve stone tent-rings distributed across the plateau; a possible Page 10 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18 circular parade-ground marked by boulders and a central post; various other small circular and sub-rectangular stone features; and numerous finds of domestic military artefacts and general debris, including wooden tent pegs, bits of tent canvas and sandbag, uniform buttons, glassware, tin cans, a clay pipe-bowl, grape pips, an apricot stone, and fragments of newspaper. 6. A small defended observation post or machine-gun position on a hillock 200 metres north-west of the station. Most of the hillock has been quarried away, probably for gravel for railway embankments during an abortive modern reconstruction project, but the feature itself, defined by stone parapets, has been respected and has therefore survived. 7. A fortified observation post at the top of the ridge forming the north-east side of the wadi, from which position it has commanding views of the entire wadi and the (p.447) high ground to the west. It comprises a sub-rectangular enclosure formed of dry-stone walling 85 metres long and between 45 and 20 metres wide. Several short trenches have been carved into the high ground that protects the approaches to this hilltop fortification. It seems highly likely that further exploration will reveal more military features and sites within the militarised landscape around Wadi Rutm, especially as our investigations extend northwestwards, towards the head of the wadi, the escarpment and the site of Batn al-Ghul station on the plateau above. While preliminary, the investigations have already revealed an ancient landscape extensively and profoundly militarised in 1916–18. It is important to stress that there is nothing special about Wadi Rutm. To our knowledge, no attacks on this station are recorded. It can therefore be assumed to be fairly typical of the seventy-seven stations on the railway between Damascus and Medina. Nor were railway stations the only strategic assets that the Ottoman army was obliged to defend in 1916–18. Various roads were also garrisoned, as were numerous posts among the civilian population, whose control as a source of tax revenues, grain, conscripts and much else was a military necessity. Wadi Rutm may well be representative of the conflict archaeology—and therefore of the military effort—at up to 200 comparable military posts scattered across the landscape of Jordan.

The Archaeology of Imperial Collapse 3: The First Battle of Ma‘an The archaeology of combat (‘battlefield archaeology’) is much less visible than the archaeology of conflict. Conflict is protracted and involves heavy investment in fortifications and logistics. Occupation and activity are represented by features, structures and militarised landscapes. Combat, by contrast, is formed of single events. Such events, often brief collisions of opposing groups of human beings, may leave few traces. Many such events concentrated in time and space—as on the battlefields of the Western Front—may leave an accumulation of superimposed traces. Because of the frequency and intensity of combat on the Western Front, archaeological sites are saturated with munitions, war debris, and sometimes human remains. In these circumstances, single combat events are rarely distinguishable. Battlefield archaeology of other periods has sometimes been more successful in identifying and defining single combat events,33 but, as far as we are aware, no single (p.448) combat event in the First World War has ever been reconstructed archaeologically. We believe that we may now have achieved this for the First Battle of Ma‘an (11–20 April 1918). Though there were four socalled ‘Battles of Ma‘an’, only the first involved a direct assault on the heavily fortified town and

Page 11 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18 station. Moreover, the actual assault on the inner trench-works, leading to the brief capture of the station itself, took place over just two days (17–18 April 1918).34 (p.449) A systematic gridded metaldetector survey recovered substantial quantities of expended munitions on the Hill of the Birds (Figure 22.4). The survey extended along about 500 metres of the Figure 22.4. Plot of expended munitions recovered in and in front of the north-westmain firing trench on the western slope and facing firing trench on the Hill of the Birds at involved timed scanning and recovery in 18 Ma‘an. © GARP. grid-squares each of about 30 metres. Of the seven main types of munitions, representing either outgoing or incoming fire, the following quantities were collected during the survey: 65 fired 7.65 mm Mauser cartridge cases (outgoing small-arms); 6 Mauser-round cartridge clips (outgoing small-arms); 59 fired .303 bullets (incoming small-arms); 11 fired large-bore Martini lead bullets (incoming small-arms); 257 10 mm lead balls (incoming artillery); 20 shell fragments (incoming artillery); and 8 unidentified shrapnel fragments (incoming artillery). This material was not evenly distributed along the 500 metre extent of the survey: the character and overall intensity of the exchange of fire seems to have varied along the line. Since there is only one recorded battle at the site—that of 17–18 April—it seems highly likely that detailed aspects of that (poorly recorded) clash are actually represented. If a single combat event created this thin spread of munitions, then this contrasts sharply with the archaeological signature of combat on the Western Front, where battlegrounds are palimpsests of many military actions and single combat events are invisible—implying a different kind of war in southern Jordan. This conclusion is strengthened by negative evidence elsewhere: large areas of Wadi Rutm yielded no expended munitions at all. Both at a tactical and at a strategic level, therefore, we are learning about the nature of combat in the Arab Revolt.

Conclusions War without frontiers The archaeological imprints of the Arab Revolt at Ma‘an and Wadi Rutm are representative of this theatre of war as a whole (essentially western Arabia and Jordan). Field reconnaissance and contemporary written sources give every reason for assuming that they are typical of numerous sites, especially along the Hijaz Railway line and major roads. Old settlements, communication nodes, water sources and other key points in the landscape received a new layer of First World War militarisation as they were transformed by fortification into contested strategic assets. On this evidence, the Ottoman strategic defence consisted of dispersed points arranged in depth rather than a line protecting territory securely held in the rear. There were no frontiers in the sense of linear boundaries, because the whole landscape had become contested. The frontiers of the Ottoman Empire had dissolved, and the entire space inhabited by the Arabic-speaking people of western Arabia and Jordan had become the contested space of a battle-zone. In this respect the archaeological imprint seems to confirm the following four of Lawrence’s principles of guerrilla warfare: remain strategically dispersed; make (p.450) maximum use of mobility; operate mainly in small, local groups; and operate in depth rather than en face. The implication is that the scale of threat represented exceeded that implied by characterisations of the Arab

Page 12 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18 Revolt as merely ‘a sideshow of a sideshow’—a conclusion made possible by the wide-focus conflict archaeology approach that has been adopted. War of Detachment In what may be the first-ever archaeological investigation of a guerrilla war, our investigations indicate that the regular/imperial forces are represented by a high degree of archaeological visibility, while the irregular/rebel forces are characterised by near-total archaeological invisibility. The Ottoman army seems omnipresent in the conflict archaeology of southern Jordan, represented by fortified buildings, linear trenches, hilltop redoubts, dugouts and bunkers, stonewall enclosures, military encampments, and artefact assemblages. Their opponents, by contrast, seem mere shadows, usually represented, if at all, by nothing more than a thin layer of incoming bullets and shrapnel. This faint trace of war, moreover, coupled with the evidence for outgoing fire, attests the low intensity of actual fighting. Conflict was endemic and chronic; combat was only occasional, fleeting and small-scale. Here it seems that other principles of Lawrence’s theory of war are illustrated: remain largely detached from the enemy; do not attempt to hold ground; never engage in sustained combat; and make war on matériel rather than on men. The effect implicit in the archaeological evidence is that the threat of potential and unpredictable attack compelled the Ottoman enemy to commit time, manpower and resources into defending an extended landscape, most of which never become an actual battlefield. War of the Flea Robert Taber, in his classic study of guerrilla warfare, War of the Flea (1965), repeatedly stressed the economy of force inherent in situations where small numbers of irregulars tied down and neutralised far larger numbers of regulars.35 A number of commentators have highlighted this aspect of the Arab Revolt. One estimate of overall military strengths in Palestine, Syria and Arabia gives the numbers as 100,000 Ottomans, 340,000 British, and only 23,000 Arabs (15,000 of whom never left Arabia). It is further estimated that more Ottomans were fighting the Arabs east of the River (p.451) Jordan than were engaged against the British to the west of it.36 This would constitute an astonishingly efficient economy of force and a very substantial contribution to the war effort, indirectly weakening the Ottoman line facing the British on the Palestine front to the point where a decisive breakthrough became possible in September 1918—one which led directly to the armistice and the collapse of the Ottoman Empire. The figures are disputed, but the archaeology fully supports these estimates and the conclusions that flow from them. The Great Arab Revolt Project’s investigations have begun to reveal evidence for the hitherto unrecognised scale of landscape transformation and occupation by Ottoman forces in response to a real or imagined threat. Where history may be equivocal, archaeology deals with in situ physical evidence—the material correlates of perceptions, decisions and actions in the past filtered through the vagaries of survival. The evidence of conflict archaeology between Ma‘an and Wadi Rutm demands further investigation and interpretation, yet already suggests that far more is going on locally than had previously been realised, and whose precise regional implications are as yet unclear. It is in the nature of guerrilla warfare that the distinction between perception and reality is blurred—the ultimate success or otherwise of Arab guerrilla tactics is preserved in archaeological traces that are only now coming to light. Note. Our thanks are due to our colleagues in the field, Fizz Altinoluk, Ali Baldry, Angie Cannon, Cat Edwards, Jules Evan-Hart, David Hibbitt, Martin Plummer, David Thorpe,

Page 13 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18 Duncan Ward and John Winterburn; to our Jordanian colleagues Hani Falahat and Muhammad Abu Abilah; and to the sixteen other volunteers who worked with us in November 2006. We are also grateful to Bob Bewley and David Kennedy for the use of their air reconnaissance photographs, and more generally to the Jordanian Department of Antiquities, the Council for British Research in the Levant, and HRH Prince Hassan Bin Talal for their support. Notes: Proceedings of the British Academy 156, 431–451. © The British Academy 2009. (1) Much of the debate has been inextricably associated with the vexed question of the precise role and influence of Colonel T. E. Lawrence (Lawrence of Arabia), with hostile biographers minimising both Lawrence’s contribution to the Arab Revolt and the Arab Revolt’s contribution to the British war effort in Palestine/Syria: e.g. Richard Aldington, Lawrence of Arabia: A Biographical Enquiry (London: Collins, 1955), esp. 196–205. Other biographers have been much more positive, e.g. Michael Asher, Lawrence: The Uncrowned King of Arabia (London: Penguin, 1999). (2) An argument developed in detail in a number of studies, including George Antonius, The Arab Awakening (New York: Capricorn, 1965) and David Fromkin, A Peace To End All Peace: The Fall of the Ottoman Empire and the Creation of the Modern Middle East (London: Phoenix, 2000). (3) Lowell Thomas, With Lawrence in Arabia (London: Hutchinson, 1925). (4) For a recent example, see Caroline Finkel, Osman’s Dream: The Story of the Ottoman Empire, 1300–1923 (London: John Murray, 2005), 532. (5) Malcolm Brown (ed.), T. E. Lawrence in War and Peace: An Anthology of the Military Writings of Lawrence of Arabia (London: Greenhill, 2005), 40–55, is a good example. On the other hand, some military historians are firmly in the detractors’ camp: see, for instance, Matthew Hughes, ‘What did the Arab Revolt contribute to the Palestine campaign? An assessment’, The Journal of the T. E. Lawrence Society, 15 (2006), 75–87. (6) N. J. Saunders, ‘Excavating memories: archaeology and the Great War, 1914–2001’, Antiquity, 76 (2002), 101–8, and idem, Killing Time: Archaeology and the First World War (Stroud: Sutton, 2007). (7) John Schofield, W. G. Johnson and C. Beck (eds.), Matériel Culture: The Archaeology of 20th Century Conflict (London: Routledge, 2001). (8) For a rare Italian exception to this focus on the Western Front, see Marco Balbi, ‘Great War archaeology on the glaciers of the Alps’, in N. J. Saunders and P. Cornish (eds.), Contested Objects: Material Memories of the Great War (London: Routledge, in press). (9) Victor Buchli (ed.), The Material Culture Reader (Oxford: Berg, 2002), and C. Tilley et al., Handbook of Material Culture (London: Sage, 2005). (10) N. J. Saunders, ‘First World War archaeology: between theory and practice’, in Proceedings of the Conference ‘Archeologia della Grande Guerra’, Luserna (Trentino), 23–24 June, 2006 (in press).

Page 14 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18 (11) N. J. Saunders, ‘Matter and memory in the landscapes of conflict: the Western Front, 1914– 1999’, in B. Bender and M. Winer (eds.), Contested Landscapes: Movement, Exile and Place (Oxford: Berg, 2001), 37–53. (12) Saunders, ‘First World War archaeology’. (13) We put a full team of twenty-eight specialists and volunteers in the field for two weeks in late November 2006. The team was led by Neil Faulkner, Nick Saunders and David Thorpe, and the specialists included a landscape archaeologist, two geophysicists, two metal-detectorists, two excavation supervisors, a finds specialist, a recording assistant and a photographer. (14) John Stephenson, ‘Turkey enters the war’, in Purnell’s History of the First World War (London: Purnell for BPC Publishing, 1969–71), 1: 397–403. (15) Cem Uzun, Making the Turkish Revolution (Istanbul: Antikapitalist, 2004), 37. In fact, Germany, also a creditor state, joined the international protest when the Young Turk regime ended debt repayments (Finkel, Osman’s Dream, 528), but wider interests ensured that the alliance between the two states held. The French had definite designs on Ottoman territory, and, though the British were divided, some British statesmen, certain that the Ottoman Empire could not survive, had been toying with partition between the great powers since the 1890s: A. J. P. Taylor, The Struggle for Mastery in Europe, 1848–1918 (Oxford: Oxford University Press, 1971), 359. (16) Uzun, Making the Turkish Revolution, 52–83 and passim, and Finkel, Osman’s Dream, 506– 18. (17) Finkel, Osman’s Dream, 529. (18) W. E. D. Allen and Paul Muratoff, Caucasian Battlefields: A History of the Wars on the TurcoCaucasian Border, 1828–1921 (Nashville: Battery Press, 1999), 249–85. (19) A good recent narrative account is James Barr, Setting the Desert on Fire: T. E. Lawrence and Britain’s Secret War in Arabia, 1916–1918 (London: Bloomsbury, 2006). (20) Among a number of studies that have explored this aspect is Lawrence James, The Golden Warrior: The Life and Legend of Lawrence of Arabia (London: Abacus, 1995). (21) Brown (ed.), T. E. Lawrence in War and Peace, 49–54. (22) Among practitioners of guerrilla warfare against the Roman Empire, for example, are such figures as Viriathus, Jugurtha and Vercingetorix: Philip Matyszak, The Enemies of Rome: From Hannibal to Attila the Hun (London: Thames and Hudson, 2004). (23) Key treatises are Mao Tse-Tung, On Protracted War (El Paso: El Paso Norte Press, 1938), 165–287, Vo Nguyen Giap, People’s War, People’s Army (Honolulu: University Press of the Pacific, 1959), 41–150, and Ernesto Che Guevara, Guerrilla Warfare (New York: Monthly Review Press, 1961). These texts are largely preoccupied with the relationship between guerrilla and regular warfare, and the progression from one to the other that is considered essential to final victory. They have little to add to Lawrence about guerrilla warfare per se. (24) The best-known study is probably Robert Taber, War of the Flea (New York: L. Stuart, 1965).

Page 15 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18 (25) T. E. Lawrence, ‘The evolution of a revolt’ (1920), in Brown (ed.), T. E. Lawrence in War and Peace, 265–6. (26) ‘Twenty-seven articles’, ‘The evolution of a revolt’ and ‘Science of guerrilla warfare’ are all reproduced in Brown (ed.), T. E. Lawrence in War and Peace, 142–7, 261–73, 275–84. The 1922 Oxford edition of Seven Pillars of Wisdom has recently been republished: T. E. Lawrence, Seven Pillars of Wisdom: The Complete 1922 ‘Oxford’ Text (Fordingbridge: J. and N Wilson, 2004). (27) Rupert Smith has recently advanced a strong argument that Clausewitz was the theoretician of a revolutionary new form of aggressive mass warfare developed by Napoleon: Carl von Clausewitz, On War, ed. Anatol Rapoport (Harmondsworth: Penguin, 1968), 29–63. Certainly, notwithstanding any references he may make to other forms of warfare, he is universally recognised as the exponent and advocate of a distinctive paradigm of war based on direct confrontation and head-on battle to seek out and destroy the enemy’s main forces at the earliest opportunity. Others have argued that this ‘western way of war’ is in fact much older, perhaps having its origins in Classical Greece, notably Victor Davis Hanson, The Western Way of War: Infantry Battle in Classical Greece (London: Hodder & Stoughton, 1989). Either way, Lawrence’s conception represents a radical alternative. (28) Brown (ed.), T. E. Lawrence in War and Peace, 265. (29) J. Nicholson, The Hejaz Railway (London: Stacey International, 2005), 37. (30) Syed Ali El-Edroos, The Hashemite Arab Army, 1908–1979 (Amman: The Publishing Committee, 1980), 134–7. Jafar Pasha Al-Askari, A Soldier’s Story: From Ottoman Rule to Independent Iraq (London: Arabian Publishing, 2003), 142–6. (31) Liman von Sanders, Five Years in Turkey (Uckfield: Naval & Military Press, 2005). (32) Eugene Rogan, Frontiers of State in the Late Ottoman Empire: Transjordan, 1850–1921 (Cambridge: Cambridge University Press, 1999), 233. (33) One example from British prehistory—involving exchanges of arrow-shot at Crickley Hill causewayed enclosure in Gloucestershire in the Early Neolithic—is reported in Roger Mercer, ‘The origins of warfare in the British Isles’, in J. Carman and A. Harding, Ancient Warfare: Archaeological Perspectives (Stroud: Sutton, 1999), 143–56. The classic example of this kind of combat archaeology is, however, the investigation of the site of the Battle of the Little Big Horn: Douglas D. Scott et al., Archaeological Perspectives on the Battle of the Little Bighorn (Norman: University of Oklahoma, 1989). (34) El-Edroos, The Hashemite Arab Army; Jafar Pasha Al-Askari, A Soldier’s Story, 142–6. (35) Taber, War of the Flea, passim. (36) El-Edroos, The Hashemite Arab Army, 178–9. The issue of numbers and deployments is far from straightforward, both in relation to the Ottoman forces and even more so the Arabs, most of whom were irregulars, such that the numbers serving at any one time fluctuated wildly. However, the most detailed calculation, that of Syed Ali El-Edroos in 1980, is broadly in line with that of contemporary British official estimates. Lieutenant-Colonel (later Field Marshal) Wavell served as a liaison officer under Allenby and subsequently wrote a history of the Palestine campaign. He gives the British front-line combat strength in Palestine in the late summer of Page 16 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

War without Frontiers: The Archaeology of the Arab Revolt, 1916–18 1918 (on the eve of the last great offensive) as 69,000 men. He puts the opposing Ottoman strength at 17,000 (i.e. front-line combat troops deployed west of the Jordan). But he also has 14,000 Ottoman troops deployed in Syria and on the Hijaz Railway (i.e. east of the Jordan), with a further 3,000 in general reserve; moreover, these figures do not include the Medina garrison, which still numbered 8,000 when it finally surrendered in January 1919. See A. P. Wavell, The Palestine Campaigns (London: Constable, 1931), 195; Cyril Falls, Official History of the War: Military Operations. Egypt and Palestine, II (Nashville: Imperial War Museum and Battery Press, 1996), 624. On Wavell’s figures, therefore, we do appear to have more Ottoman troops deployed against the Arabs in Arabia and Syria than we have facing Allenby in Palestine.

Access brought to you by:

Page 17 of 17 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Military Service and Material Gain on the Ottoman–Hapsburg Frontier

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Military Service and Material Gain on the Ottoman–Hapsburg Frontier MARK STEIN

DOI:10.5871/bacad/9780197264423.003.0023

Abstract and Keywords Just over a century ago, Frederick Jackson Turner presented his famous Frontier Thesis, which continues to be one of the most debated and controversial theories in historical scholarship and has affected all discussions of frontier history worldwide. This chapter explores one aspect of Turner's work that may be applicable to other frontiers — that of the frontier as a zone of economic opportunity. It discusses how military service on the seventeenth-century OttomanHapsburg frontier presented a number of chances for economic advancement to men who were willing to take the risks of living and working along the border. Moving to the frontier offered economic opportunities not found in the interior. In North America, a large part of the opportunity was the potential to acquire land. It is possible that in some cases there was similar opportunity along the seventeenth-century Ottoman-Hapsburg frontier. Keywords:   Frederick Jackson Turner, Frontier Thesis, economic opportunity, Ottoman-Hapsburg frontier, North America

JUST OVER A CENTURY AGO, Frederick Jackson Turner presented his famous Frontier Thesis, which continues to be one of the most debated and controversial theories in historical scholarship and has affected all discussions of frontier history worldwide. My purpose in this chapter is not to add to the debate on the nature of the North American frontier, nor to try broadly to apply Turner’s Thesis to any other part of the world. Rather, I want to explore one aspect of Turner’s work that I believe is applicable to other frontiers—that of the frontier as a zone of economic opportunity. In this paper I will discuss how service on the seventeenthcentury Ottoman-Hapsburg frontier presented a number of chances for economic advancement to men who were willing to take the risks of living and working along the border. A frontier was often a place to begin a new life with new opportunities—particularly the opportunity for economic advancement. As Turner eloquently states, ‘each frontier did indeed furnish a new field of opportunity, a gate of escape from the bondage of the past.’1 Moving to the frontier offered economic opportunities not found in the interior. In North America a large part of the opportunity was the potential to acquire land. Homesteading was promoted, and land Page 1 of 14 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Military Service and Material Gain on the Ottoman–Hapsburg Frontier ownership was open to new migrants.2 It is possible that in some cases there was similar opportunity along the seventeenth-century Ottoman-Hapsburg frontier. Archival records show that some men were drawn to the frontier by the possibility of acquiring farmland in much the same way as immigrants were attracted to the New World. A 1673 report from the defterdar, or financial scribe, at Uyvar (modern Nové Zámky in Slovakia), an important fortress on the Hapsburg frontier, discusses the (p.456) difficulties encountered in trying to carry out a proper land survey3 He finds that local timar-holders tried to keep farmers who were living in the area unregistered to prevent them from being assigned to the imperial hass lands. These unregistered peasants were probably young, unmarried men who had left their home villages and come to the frontier region in search of land to farm. The defterdar reports that there were many peasants living in villages and mezraa,4 abandoned arable land. Because they were new arrivals, they were not listed in the land registers. Thus, the lure of what Turner called ‘open land’ had an effect on the Ottoman-Hapsburg frontier (for maps, see Figures 3.1 and 3.2). It was the military nature of the frontier, with its lines of well-garrisoned fortresses, however, that presented the most opportunity for economic advancement. These opportunities were available both to the askeri military class and the reaya peasant class. They existed both for men sent to the frontier from the centre of the empire and for local troops and workers. For the reaya the greatest economic opportunities came in working to build and maintain the frontier fortresses. This was especially true of skilled craftsmen. Expense reports reveal many details about the work that went in to fortress construction and repairs.5 These records provide many types of information. Detailed financial accounts not only give an idea of the cost to the state of these repairs, but also a breakdown of how the money was spent. Lists of types of workers and their pay, and materials and their costs, also furnish a picture of the actual activities involved in the repairs. Other documents, such as the orders from the central administration authorising repairs, contribute to a clearer understanding of the bureaucratic process of financing and overseeing fortress repairs. These expense reports reveal many details about the work that went in to fortress repairs and the workers who were responsible. The accounts for the repairs to Gradişka (modern Gradiška in Bosnia) in 1677 were kept week by week, and list the expenses for workmen and supplies.6 Construction was done in the late summer and early autumn, and lasted ten weeks. The craftsmen appear to have worked six-day work-weeks. The project was overseen by a group of financial officials acting as supervisors and called mûtemed in the documents. These men were paid quite well, earning 42 akçes, small Ottoman silver coins, per day. An architect, or mimar, was also on-site to direct the (p.457) work. The mimar earned the most of any of the men involved in the construction, as befitted his skills and responsibilities. He earned 50 akçes per day. The last of the men listed who would be included among the supervisors of the project was a çavuş, or messenger, who earned 40 akçes per day. A great number of skilled workmen were needed for the repair work. From the list of workers and supplies, it becomes clear that the fortifications being built were palanka-style.7 Thus it comes as no surprise that the largest number of craftsmen listed are carpenters. These men appear to have been recruited locally, and all the men from the same kaza, or district, worked and were paid together as a unit.8 There were up to four different kaza crews working on the project at any one time. In the early weeks of the repairs around fifty carpenters laboured on the project. As the work progressed, more carpenters were hired. By the end of construction, about 115 carpenters were involved. These men were all paid 25 akçes a day, a relatively high salary, Page 2 of 14 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Military Service and Material Gain on the Ottoman–Hapsburg Frontier especially compared to carpenters in other fortress garrisons who, according to payroll documents, usually earned only 6 to 7 akçes a day9 Blacksmiths were also among the craftsmen working at Gradişka. A master smith, paid 40 akçes a day, was assigned to the fort, and he had a messenger of his own, who earned 20 akçes.10 Presumably, if the master smith needed assistance beyond that of his çavuş he would recruit a crew of unskilled workmen, as there are no other blacksmiths listed in the document. The final group of craftsmen listed were men skilled in the construction of palanka walls. Building the parallel lines of upright logs, bracing them together with crossbeams, and filling the space between with earth were specialised tasks calling for experts in construction. The expense documents show that a crew of these men was called in when needed. As work progressed, the need for these craftsmen waxed and waned. In the early weeks of construction, large numbers of these men were brought in for just one day to establish the layout of the fortress walls. In the middle weeks of the project, a large number of workers were hired, and they worked full weeks. Towards the end of the project fewer palanka wall specialists were needed, and the crew grew smaller. As few as nine men were working during the last week of work. These men earned 25 akçes a day, which was the same as the carpenters. Other recorded expenses help to shed light on the work conditions for the craftsmen on such a job. In the first week 700 akçes were paid out in gifts.11 These bonuses most likely went to the workmen as an extra incentive (p.458) to begin the job. Clearly for skilled workers there was economic opportunity along the frontier. By far the greatest opportunities for economic advancement came to men willing and able to fight and serve in the frontier garrisons. A soldier’s life always had some basic economic appeal, primarily from the promise—often unfulfilled—of a regular income from the state treasury. Military salaries, however, were never high in early modern armies. Thus for all Ottoman frontier troops, Janissaries and irregular levies alike, service in frontier garrisons also offered the opportunity to acquire wealth through raiding for booty. Garrison troops had the skill with arms, proximity to enemy territory, comrades and time to cross over the border and loot towns and villages. States usually turned a blind eye to such raids by their forces unless their depredations were serious enough to provoke a major response from the other side of the frontier. Border troops had to balance carefully their own desire for gain against the risk of attracting the attention of the authorities of both states. Full-scale warfare on the frontier brought new risks to the garrison troops, increasing the likelihood of military action and decreasing the opportunities for raiding and the resultant economic benefit. Border incursions by both Ottoman and Hapsburg garrison troops to rustle livestock and take captives were common during the seventeenth century. It should be noted that a certain low level of raiding was acceptable under the provisions of the treaties between the two empires; small-scale raids were not considered a cause for war. This followed the precedent of Ottoman agreements in the region. Similar provisions appear as early as the 1483 treaty with King Matthias of Hungary12 Opportunity, however, came with risk. Booty raids not only made life along the frontier dangerous but, if large and disruptive enough, also violated the OttomanHapsburg peace treaties.13 Raids into enemy territory could be very lucrative. All kinds of booty were taken, including horses, other livestock, grain, male and female captives, and, when the raid was against a military target, weapons. In his famous travel book, the Seyahatname, Evliya Çelebi describes the raids in which he took part while at Kanije (Hu. Kanizsa), a major Hungarian fort, in 1664.14 Page 3 of 14 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Military Service and Material Gain on the Ottoman–Hapsburg Frontier The booty and captives taken in the raid were sold in the marketplace after the raiders returned. Such sales could last several days. Evliya relates that it took five days to sell the booty from one raid and eight days for another. (p.459) The proceeds from one of these sales totalled 86,000 gurus (large silver coins).15 A comparison with regular salaries shows the significance of this income and how important booty was as a supplement to the amount the garrison was usually paid. Mevacib defterleri (payroll records) for the garrison provide this information. A particularly detailed defter from 1687–8 gives the total daily payroll for Kanije as 18,654 akçes.16 When this amount is computed for the entire year and converted to gurus, the total comes to just over 55,000 gurus. Clearly raiding was a major supplement to a soldier’s normal pay. It was especially welcome when pay from the central administration was late, a common problem for the Ottomans in the seventeenth century. Captives were also an important and valuable commodity along the frontier. Evliya recounts that after one of the raids forty captives were sold for a variety of prices. Five leaders among the captives, whom Evliya calls reis kâfirleri, possibly officers, were sold for 1,000 gold coins each. Five others were sold for 500 gold coins each, and the remainder were sold for between 200 and 300 each. Thus the money raised by selling captives—at least 13,500 gold coins and perhaps as much as 16,500—was a major portion of the total of 18,160 gold coins the sale of booty had generated.17 Some captives were taken for resale to urban slave markets; others were put to work locally. During campaigns many captives were used to dig trenches in Ottoman siege works. Such was the fate of Count Luigi Fernando Marsigli, a Hapsburg officer captured during the 1683 siege of Vienna, who later wrote a book describing the Ottoman state and military.18 Marsigli ultimately arranged his ransom and was released. Ransom of captives, primarily captured officers and soldiers, became a lucrative aspect of frontier raids in the seventeenth century and was a feature of raids by both Ottoman and Hapsburg troops.19 The ransoms paid for captives could amount to much more than the value of the booty a soldier could carry away in the initial raid. Peter Sugar, in his discussion of the institution of ‘professional prisoners’ along the (p.460) Ottoman-Hapsburg frontier, gives information on the amounts of ransoms.20 Citing documents from the Hungarian National Archives, Sugar reports that the Hungarian commander of Körmend, the fortress facing Kanije, paid the following for the release of ninety-one of his men between 1648 and 1650:21 28,220 tallers in cash 44 lengths woollen cloth 100 lengths linen cloth 17 pistols 24 large vats wine 900 kila22 weight honey 29 kila weight pepper 3 kila weight saffron 100 cases butter. Individual ransoms for officers could be large. In 1588 Olay Bey paid the commander at Körmend 1,000 tallers in cash, 6,000 tallers worth of cattle, and two complete sets of horse gear, one decorated in gold and the other in silver.23 Even ransoms for regular soldiers could be substantial. Page 4 of 14 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Military Service and Material Gain on the Ottoman–Hapsburg Frontier By the seventeenth century European armies had grown in size, and the Ottomans underwent this same military transformation. The manpower demands of early modern gunpowder warfare and the economic opportunities for able men on the frontier led to a marked increase in the number of both volunteers and recruits.24 Volunteers, called gönüllüs, moved to the frontier to try to make a living as military men, much as the gazis had during the early years of Ottoman expansion. They hoped through valorous service to be granted a timar land grant or a regular salary like those of the Janissary corps.25 These men were Muslim, and it is probable that they had some previous (p.461) military training. They most likely were the sons of Janissaries, timar-holders, or men in the military retinues of timar-holders. There is also evidence that men who had lost timar rights, or had deserted from the regular army and hoped to regain their previous status, volunteered for service on the frontier.26 Although early references to gönüllü troops state that they did not receive a salary, by the seventeenth century these soldiers were paid by the Treasury. Gönüllü troops served both in the field army and in fortresses. When serving in garrisons they came under the direct command of the fortress commander.27 These soldiers served as both cavalry and infantry and were organised in units similar to those of the Janissaries.28 Vacancies in the salaried units were filled from among experienced volunteers. Gönüllü troops were a significant part of frontier garrisons. Records from Uyvar show that gönüllü units were often the largest in the garrison.29 In peacetime almost 20 per cent of the troops assigned to Uyvar were gönüllüs.30 During wartime the proportion of gönüllü troops in fortresses could rise even higher as other troops were transferred from garrison duty to the field army, as records from the last months of the 1663–4 Ottoman-Hapsburg war demonstrate. With the main army marching to meet the Hapsburgs, the number of gönüllü troops at Uyvar rose, making up one quarter of the garrison. After their defeat at St Gotthard in August 1664, the Ottomans increased the size of garrisons in forts all along the frontier. Uyvar’s non-Janissary garrison troops increased from 634 to 850 men, with gönüllüs comprising the largest single contingent, making up one third of the troops.31 Although the gönüllü volunteers were significant, ongoing manpower demands, especially in wartime, led the Ottomans to recruit more irregular troops to reinforce the standing Janissary army. There were several types of these auxiliary troops, all similar, but each having distinctive features. Unlike the gönüllü, who appear to have been drawn from timariot backgrounds, most of the irregular levies were from the tax-paying reaya class. Thus military service was not just a means of individual economic advancement, but also a potential means of social mobility. The most numerous and important of these types were the azeb. The term azeb (or azab, as it appears in some secondary sources; pl. azeban) derives from the Arabic for ‘bachelor’ and was used to refer to soldiers who fulfilled various roles in the Ottoman military. As early as the thirteenth century, sources report soldiers called azeb serving (p.462) as marine troops in the beylik of Aydin.32 These soldiers have been described as being ‘identical in origin, motivation, and organization to the frontier-ghazis’.33 Similar kinds of troops were later found in the Ottoman navy and were called azeb. A 1474 register from the Ottoman naval base at Gallipoli lists four azeb units serving on a variety of ships, from galleys to horse transports.34 This term, however, was also used for infantry troops and one must differentiate between the infantry azebs, often called yaya azeb, and the naval variety, called both bahriye (navy) azeb and deniz (sea) azeb.35 By the fourteenth century the infantry azebs assigned to garrison duty were considered a separate group called kale (fortress) azebs.36 In the seventeenth century, these

Page 5 of 14 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Military Service and Material Gain on the Ottoman–Hapsburg Frontier azebs played a major role in fortress garrisons, and were listed by Marsigli as one of the five infantry divisions of the Ottoman frontier forces.37 Azebs were found in the Ottoman army from the earliest years of the state. Some sources say that they pre-date the organisation of the Janissaries.38 They were recruited in large numbers in Anatolia and, later, Rumelia, and were organised in units similar to those of the Janissaries. The azebs served as light infantry, armed with bows, swords, and sometimes pikes, and wore a red börk, a felt hat of the same style worn by the Janissaries, but a different colour.39 In battle the azebs acted as archers and took their place in the front line, before the cannon and the Janissaries.40 The procedure for recruiting azeb troops is preserved in the kanunname, or law code, of Sultan Süleyman the Magnificent.41 Troops were raised in each sancak at the local level, under the supervision of the local judge, or kadı. One man was levied from every twenty to thirty households (hane). It is not clear if these groupings were the same as those used in collecting the taxes that were assessed by hane. In the fifteenth and six-teenth centuries, these local groups were responsible for supporting the soldiers sent from their area.42 To support each azeb 300 akçes had to be collected from the assigned households. In time of war, this money took the place of the extraordinary taxes called (p.463) avarız.43 The local community also had to provide someone to act as guarantor for each azeb. If the soldier deserted the army the guarantor was responsible for his training and equipment expenses. The azeb had to be an unmarried man, in good health, strong, and brave in battle. He could have no sons or other dependants, such as elderly or sick people. The soldiers’ names were entered in two yoklama defters (muster rolls), one of which was kept by the kadı, the other forwarded to the central administration. Vacancies in an established garrison azeb unit would be filled by appointing a worthy young man to the post.44 By the seventeenth century an azeb’s position and, most importantly, his place in the garrison pay and muster rolls, called gedik, could be inherited by his son.45 This shows that although the azeb units were originally established to contain single men, in time these men could marry and retain their place in the garrison. This also reflects a way in which military service could be an economic opportunity not just for an individual, but for his family as well. Another important type of levied irregular troop was the sekban, often also called sarıca. These soldiers were initially used in the retinues of local governors, but the need for musket-carrying men led to their use on campaign and in garrisons. Sekbans filled many of the same roles as azeb troops, and the procedure for recruiting them was similar to that used to raise azeb units.46 The sekban served as auxiliary units to the Janissaries, and served under Janissary officers. An officer would be sent out to the provinces to enrol soldiers. Landless men would be taken into the new units, and would be promised pay from the central Treasury. The officer would carry with him an order from the Sultan authorising the enlistment of reaya peasants, as well as a flag (bayrak), which would act as the unit’s standard. This standard, usually red in colour,47 represented the authority granted to the unit by the Sultan, and its revocation was the sign that the troop was to be disbanded. During the sixteenth and seventeenth centuries, some Anatolian sekban units refused to give up their weapons and be demobilised, and turned to brigandage. These itinerant soldiers were often hired by provincial pashas and beys, and were the troops used against the state in what are known as the Celali rebellions.48

Page 6 of 14 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Military Service and Material Gain on the Ottoman–Hapsburg Frontier (p.464) Sekban units were organised into units, ideally of fifty men, with the Janissary who recruited them acting as commander. The standard-bearer of the unit (bayraktar or alemdar) was the second-in-command. Sekban troops were generally used as infantry, although some were used as cavalry49 These soldiers carried firearms, and their recruitment was a response to the increased demand for men with guns experienced throughout Europe during the seventeenth century. The sekban units were not, in fact, the best troops in the Ottoman army. Marsigli considered them to be the lowest of the five types of troops he lists as border forces.50 The standards for the men enlisted as sekbans were lower than those for azebs, who were to be young and brave. Sekban soldiers were recruited for numbers, in times of need, not valour. They were useful during sieges, when manpower in the trenches was important, and as reserves or camp guards on campaigns. Their frequent service in garrisons also indicates that they were inferior troops, as armies usually assign better, more reliable, troops to the field army. It is also notable that the sekban units were recruited in all parts of the empire, and that men of varied backgrounds were enlisted, including Christians. Marsigli writes that ‘Turks, Greeks (i.e. Muslims and Orthodox Christians) and Catholics’ all served as sekban soldiers.51 Ottoman sources also refer to large numbers of Christian sekbans among the siege forces at Vienna in 1683.52 These sources all agree that Muslim sekban units were more trusted than Christian ones, and there is evidence that some units defected to the enemy army53 Payroll information for sekban units in the seventeenth-century garrisons is only found in those documents that list the pay for Janissary troops, underscoring their relationship to the Janissaries.54 They are not listed in the documents that record the other, non-Janissary, soldiers. In the Janissary records sekban units are listed after the Janissary units in the entries for individual garrisons. Sekban troops in Ottoman frontier garrisons earned between 4 and 9 akçes a day. This salary is less than most other garrison troops received, again demonstrating their relative lack of military skill. Despite the rhetoric of both Ottomans and Hapsburgs portraying their conflict as a religious one, there were Christians, such as some of the sekban forces, willing to serve the Ottomans fighting against their co-religionists. Many Christian troops had served the Ottomans during the early conquests of the Balkans.55 Christian irregulars such as (p.465) the voynuks served in the army and Christian craftsmen and troops served in forts. The martolosan (sing. martolos) were the most significant Christian troops found in garrisons in Hungary in the seventeenth century. The Ottoman term martolos most likely derives from the Greek armatolos, meaning ‘armed’ or ‘weapon-carrying’.56 Forms of the word are found in Bulgarian, Serbian and Hungarian. Despite its Greek origin, it is interesting that the term itself is not securely documented in the Greek sources until the eighteenth or early nineteenth century57 Some Ottoman sources refer to troops called martolos who were used as spies as early as the campaigns of Osman Gazi and Orhan.58 By the fifteenth century the term is used for local Christian troops assigned to Ottoman border forts in the Balkans.59 European sources of that period also use the term, usually in reference to Christian sailors manning Ottoman boats on the Danube.60 In the sixteenth and seventeenth centuries, martolos units were found in most Ottoman frontier garrisons. Why would Christians join the Ottoman military? Like their Muslim comrades, the martolosan were drawn by the economic opportunity that military service provided. It may have been that life as a soldier was more attractive than life as a farmer. Farming was hard work, and available Page 7 of 14 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Military Service and Material Gain on the Ottoman–Hapsburg Frontier land was difficult to find. Garrison service was by no means easy, but it was less onerous than working the land. Furthermore, some of the martolosan had served in irregular forces prior to the Ottoman conquest and were thus continuing their military careers. There were also major financial advantages to becoming a soldier. Martolos troops were exempt from many taxes levied on local villagers, including harac (a land tax),61 ispence (a type of personal tax),62 and the various extraordinary wartime taxes that (p.466) came under the rubric avarız.63 These exemptions were similar to those granted to the derbendci, men who guarded roads and mountain passes. Some martolos units, especially those serving in forts, were entered into Ottoman paybooks and given regular salaries. By the sixteenth and seventeenth centuries, payroll records for Hungarian forts show that most martolosan were salaried, and earned between four and eight akçes per day64 Martolos troops also joined their comrades in the garrisons on raids across the frontier, and shared in the proceeds. Although the rank and file of the martolos units were Christian, the commanding officers were Muslims, although squad leaders could be Christian. Records show fathers, sons and brothers serving in the same martolos units.65 These Christian troops played an important part in the frontier defences of the Ottoman Empire, particularly in Hungary.66 As Christians fighting other Christians to serve the ends of their Muslim rulers, they are a prime example of how people living on both sides of the border can form a joint community, and how identity within that community can be transitional. Affiliation could change based on circumstances and individual self-interest. In the case of the martolosan, the economic advantages of garrison duty outweighed any reluctance to fight against fellow Christians. In the seventeenth century the composition of the martolos units changed in a way that even more strongly shows the transitional nature of identity within the frontier community. As seen above, the martolosan were organised as Christian troops serving under Muslim officers. This confessional arrangement was the case from the earliest mention of these troops through the sixteenth century. Detailed payroll records that list each soldier’s name bear this out. Of the 1,077 martolosan recorded in a 1549–50 document for the vilayet of Budin (modern Buda, Hungary), seventeen Muslim names are listed, not counting the commanders.67 Of this handful, six men are called ‘ibn Abdullah’ (‘son of the slave of God’), the name traditionally taken by converts to Islam. The martolosan of this period were overwhelmingly Christian. Similar records for the following century, however, present a different situation. The seventeenth century witnessed an increase in the number of Muslim soldiers in the nominally Christian martolos units. More and more frequently frontier garrison payroll records list members with obviously Muslim names, such as Mehmed, Mustafa and Ahmed.68 As with the earlier, isolated instances of Muslim martolosan, many of these (p.467) men were converts carrying the name ‘ibn Abdullah’. It appears that, for these soldiers, service in Ottoman forts among Muslim comrades-at-arms led to a more formal association with Islam. Records show that as their presence in these units increased, Muslims often served as squad leaders, suggesting that the idea that officers be Muslim may have extended down the ranks. Over time the number of Muslims in martolos units became significant. In some garrisons in Hungary, half the martolosan were Muslim.69 Clearly, what was once an all-Christian unit was now a mixed troop. Archival records show that substantial numbers of irregular and volunteer troops served on the Ottoman-Hapsburg frontier. This was especially true in the early years of Ottoman rule in Hungary. More men were needed to hold the newly conquered territories and men were drawn to serve there in exchange for the pay and tax exemptions offered by the state. After the Page 8 of 14 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Military Service and Material Gain on the Ottoman–Hapsburg Frontier occupation of Budin in 1541, 1,000 of the 3,500 men in the new garrison were martolosan.70 By 1550 there were few martolos units assigned to Budin, but they still served in other garrisons in the province. More than a quarter of the non-Janissary troops in the garrisons in the Hungarian towns of Peşte and Estergon (today Pest and Esztergom, respectively) were martolos. The percentage of martolos troops in smaller garrisons in the province was even higher, reaching as much as one third.71 Records for the same period show azeban to be the most numerous units in the garrisons of the newly acquired territories.72 Indeed, azeb troops often made up the largest single contingent in a fort. In smaller forts and palankas in Hungary the entire garrison was often azeb troops.73 In the seventeenth century, although martolos units were found in many forts, their number in proportion to the entire garrison decreased as other types of troops—both regular Janissary units and other irregulars—were assigned frontier duty. Martolos units remained significant, however. Payroll records from mid-century show that martolosan still comprised 27 per cent of the garrison at Peşte.74 By 1683, records show significant martolos units in the most forward frontier forts, such as Yanık (Győr), where they made up 10 per cent of the garrison, but far fewer stationed at forts deeper within Ottoman territory.75 Large numbers of azeban also continued to be stationed in frontier garrisons. They were the largest component of the garrison at Kanije through the seventeenth century76 Close to half of the garrison in these years were azeb (p.468) troops.77 Throughout the century opportunities for economic advancement through military service continued to be offered on the frontier. The early modern Ottoman-Hapsburg frontier could be a dangerous place to live and work. For some, however, the risks of living there seemed to be worthwhile. Whether askeri or reaya, soldier or workman, there was money to be made, and new opportunities open to men willing to accept those risks. Notes: Proceedings of the British Academy 156, 455–468. © The British Academy 2009. (1) Frederick Jackson Turner, Rereading Frederick Jackson Turner: ‘The Significance of the Frontier in American History’ and Other Essays, ed. John Mack Faragher (New York: Henry Holt and Company, 1994), 59. (2) An interesting discussion of New World migration and land ownership is Walter Nugent, ‘New World frontiers: comparisons and agendas’, in David J. Weber and Jane M. Rausch (eds.), Where Cultures Meet: Frontiers in Latin American History (Wilmington: Scholarly Resources Inc., 1994), 72–85. (3) BOA, Baş Muhasebe Kalemi Uyvar Hazinesi (D.BŞM.UYH) 17083. For an analysis of this report see Mark L. Stein, Ottoman bureaucratic communication: an example from Uyvar, 1673’, Turkish Studies Association Bulletin, 20 (1996), 1–15. (4) See Halil İnalcık, ‘Mazra‘a’, EI2, 5: 959–61. (5) A study of the reconstruction of a fortress that makes great use of the available archival records is Rhoads Murphey, ‘The construction of a fortress at Mosul in 1631: a case study of an important facet of Ottoman military expenditure’, in Osman Okyar and Halil İnalcık (eds.), Social and Economic History of Turkey (1071–1920) (Ankara: Hacettepe Üniversitesi, 1980), 163–77.

Page 9 of 14 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Military Service and Material Gain on the Ottoman–Hapsburg Frontier (6) BOA, Maliyeden Müdevver (MAD) 1370. (7) On palankas, see Chapter 8 by Burcu Özgüven in this volume. (8) Workers at Mosul were also recruited by kaza. See Murphey, ‘Construction of a fortress’. (9) BOA, MAD 5158; Ottoman Garrisons on the Middle Danube: Based on Austrian National Library MS MXT 562 of 956/1549–1550, ed. and trans. Asparuch Velkov and Evgeniy Radushev, introduction by Strashimir Dimitrov (Budapest: Akadémiai Kiadó, 1996). (10) Listed as haddad başı and çavuş-i haddad, respectively. (11) The document uses the term bahşiş. (12) Halil İnalcık, ‘The Ottoman Turks and the Crusades 1451–1522’, in Kenneth M. Setton (ed.), A History of the Crusades, VI: The Impact of the Crusades on Europe (Madison: University of Wisconsin, 1989), 336. (13) Gustav Bayerle has shown that complaints about border raids were frequently the subject of communications between Hapsburg and Ottoman officials. See his Ottoman Diplomacy in Hungary: Letters from the Pashas of Buda 1590–1593 (Bloomington: Indiana University, 1972) and The Hungarian Letters of Ali Pasha of Buda, 1604–1616 (Budapest: Akadémiai Kiadó, 1991). Similar complaints are found in many of the letters preserved in the Esterházy family archive and published as Türkische Schriften aus dem Archive des Palatins Nikolaus Esterházy 1606– 1645, ed. Ludwig Fekete (Budapest, 1932). (14) Evliya Çelebi, Seyahatname, 7: 10–14. See also Halil İnalcık with Donald Quataert (eds.), An Economic and Social History of the Ottoman Empire, 1300–1914 (Cambridge: Cambridge University Press, 1994), 307. (15) Evliya Çelebi, Seyahatname, 7: 10. Gurus in this period were usually worth 120 akçes. For more on currency see Halil Sahillioglu, ‘Kurulustan XVII. Asrın Sonlarına Kadar Osmanlı Para Tarihi Üzerinde bir Deneme’, unpublished (Istanbul, 1958), and ibid., ‘Bir Asırlık Osmanlı Para Tarihi, 1640–1740’, unpublished (Istanbul, 1965). (16) BOA, MAD 3645. (17) Evliya Çelebi, Seyahatname, 7: 14. (18) Luigi Fernando Marsigli, Stato militare dell’Imperio Ottomanno, incremento e decremento del medesimo (The Hague and Amsterdam, 1732), reprinted and edited by M. Kramer and Richard Kreutel (Graz: Akademische, 1972). (19) Discussions of the treatment of prisoners and the raising of their ransom make up a good part of the correspondence between the pashas of Buda and Hapsburg authorities given in Bayerle, Ottoman Diplomacy and Hungarian Letters. For an eighteenth-century episode of ransoming of prisoners, see Osman Ağa of Temeşvar’s account of his own ransoming in Die Autobiographie des Dolmetschers Osman Ağa aus Temeschwar (Cambridge: Gibb Memorial Fund, 1980) or Frederic Hitzel’s discussion of this episode, ‘Osman Ağa, captif ottoman dans l’Empire des Habsbourg à la fin du XVIIe siècle’, Turcica, 33 (2001), 191–216. The recent collection, Géza Dávid and Pál Fodor (eds.), Ransom Slavery along the Ottoman Borders (Leiden:

Page 10 of 14 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Military Service and Material Gain on the Ottoman–Hapsburg Frontier Brill, 2007) contains a variety of articles concerning captives along the frontier from the fifteenth through eighteenth century. (20) Peter F. Sugar, ‘The Ottoman “professional prisoner” on the western borders of the Empire in the sixteenth and seventeenth centuries’, Études Balkaniques, 7 (1971), 82–91. (21) Ibid., 85. (22) What Sugar calls ‘kila’ here is the kile, a measure of capacity for grain. Although in 1579 this measure was mandated to be equal to 30 okka or vukiyye (1 okka = approx. 1.282 kg) in Hungary, it actually varied across the province from anything from 18 to 32 okka. See Halil İnalcık, ‘Introduction to Ottoman metrology’, Turcica, 15 (1983), 311–48. (23) Sugar, ‘The Ottoman “professional prisoner”’, 83. (24) For information on Ottoman garrisons in Hungary in the sixteenth century see Klára Hegyi, ‘The Ottoman mili tary force in Hungary’, in Géza Dávid and Pál Fodor (eds.), HungarianOttoman Military and Diplomatic Relations in the Age of Süleyman the Magnificent (Budapest: ELTE, 1994), 131–48 as well as her ‘The Ottoman network of fortresses in Hungary’, in Géza Dávid and Pál Fodor (eds.), Ottomans, Hungarians, and Habsburgs in Central Europe (Leiden: Brill, 2000), 163–93. For a more in-depth analysis of these garrisons in the seventeenth century see Mark L. Stein, Guarding the Frontier: Ottoman Border Forts and Garrisons in Europe (London: I. B. Tauris, 2007). (25) See Halil İnalcık ‘Gönüllü’, EI2, 2: 1120–1; ‘Mısır Kanunnâmesi’, in Ömer Lütfi Barkan, XV ve XVIıncı Asırlarda Osmanlı İmparatorlugunda Ziraî Ekonominin Hukukî ve Malî Esasları (Istanbul: İÜ Edebiyat Fakültesi, 1943), 355. (26) Pál Fodor, ‘Making a living on the frontiers: volunteers in the sixteenth-century Ottoman army’, in Dávid and Fodor (eds.), Ottomans, Hungarians, and Habsburgs in Central Europe, 229– 63. (27) İsmail Hakkı Uzunçarşılı, Osmanlı Devleti Teskilatından Kapukulu Ocakları (Ankara: TTK, 1984), 1: 3, 330. (28) İnalcık, ‘Gönüllü’, EI2, 2: 1120. (29) BOA, Bab-i Defteri Büyük Kale Kalemi (D.BKL) 32187, MAD 2052, D.BKL 32195. (30) BOA, D.BKL 32195. (31) BOA, D.BKL 32187. (32) Halil İnalcık, ‘The rise of the Turkoman maritime principalities in Anatolia, Byzantium, and Crusades’, Byzantinische Forschungen, 9 (1985), 209; H. Bowen, ‘‘Azab’, EI2, 1: 807; M. Fuad Köprülü, İsmail Hakkı Uzunçarşılı, ‘Azab’, IA, 2: 82. (33) İnalcık, ‘The Ottoman Turks and the Crusades 1329–1451’, in Setton (ed.), A History of the Crusades, VI: The Impact of the Crusades on Europe, 226. (34) İnalcık, ‘Gelibolu’, EI2, 2: 985.

Page 11 of 14 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Military Service and Material Gain on the Ottoman–Hapsburg Frontier (35) Köprülü, Uzunçarşılı, Azab’, 82; Mehmet Zeki Pakalın, Osmanlı Tarih Deyimleri ve Terimleri Sözlüğü (Istanbul: Milli Eğitim Bakanlığı, 1983), 1: 131. (36) Bowen, ‘Azab’, 807; Köprülü, Uzunçarşılı, Azab’, 82. (37) Marsigli, Stato militare, 1: 83. The others, according to Marsigli, were hisar eri, sekban, lağımcı and müsellem. (38) Pakalın, Osmanlı Tarih Deyimleri, 1: 129. (39) Azabs during the beylik period wore the red börk to differentiate them from the immediate retinue of the bey, who wore white börks. See İnalcık, ‘Rise of the Turkoman maritime principalities’, 209. (40) Uzunçarşılı, Osmanlı Devleti Teskilatından, 1: 374; Köprülü, Uzunçarşılı, ‘Azab’, 2: 83. (41) Kanunname-i Al-i Osman also known as the Süleyman Kanunnamesi, ed. M. Arif, supplement to Tarih-i Osmani Encümeni Mecmuası (Istanbul, 1329 [1911]), 59–61. (42) Köprülü, Uzunçarşılı, ‘Azab’, İA, 2: 82. (43) Ibid., 83; Bowen, ‘‘Azab’, 807. (44) Barkan, ‘Mısır Kanunnâmesi’, 358. (45) Bowen, ‘‘Azab’, 807; Köprülü, Uzunçarşılı, ‘Azab’, 83. (46) The following description of recruitment procedures derives from Halil İnalcık, ‘The sociopolitical effects of the diffusion of fire-arms in the Middle East’, in V. J. Parry and M. E. Yapp (eds.), War, Technology and Society in the Middle East (London: Oxford University Press, 1975), 195–217 and Halil İnalcık, ‘Military and fiscal trans- formation in the Ottoman Empire 1600– 1700’, Archivum Ottomanicum, 6 (1980), 283–337. (47) Marsigli, Stato militare, 1: 85. (48) For the Celali revolts, see İnalcık, ‘Military and fiscal transformation’; Mustafa Akdağ, Celali Isyanları (Ankara: Ankara Üniversitesi, 1963); William J. Griswold, The Great Anatolian Rebellion 1000–1020/1591–1611 (Berlin: Schwarz, 1983); Karen Barkey, Bandits and Bureaucrats: The Ottoman Route to State Centralization (Ithaca: Cornell University Press, 1994). (49) İnalcık, ‘The socio-political effects of the diffusion of fire-arms’, 200. (50) Marsigli, Stato militare, 1: 85. (51) Ibid. (52) Pakalın, Osmanlı Tarih Deyimleri, 3: 146 and İA, 10: 326. (53) Pakalın, Osmanlı Tarih Deyimleri, 3: 146. (54) BOA, MAD 6705; MAD 7420; MAD 116; MAD 16398; BOA, Bab-i Defteri Yeniçeri Kalemi (D.YNÇ) 33980; MAD 16742.

Page 12 of 14 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Military Service and Material Gain on the Ottoman–Hapsburg Frontier (55) Halil İnalcık, ‘Stefan Duşan’dan Osmanlı İmparatorluğuna XV. Asırda Rumeli’de Hıristiyan Sipahiler ve Menşeleri’, in Mélanges Fuad Köprülü—60. Doğum Yılı Münasebetiyle Fuad Köprülü Armağanı (Istanbul: Ankara Üniversitesi Dil ve Tarih-Coğrafya Fakültesi, 1953), 207–48. (56) W. J. Griswold ‘Martolos’, EI2, 6: 613; Robert Anhegger, ‘Martolos’, İA, 7: 341; idem, ‘Martoloslar Hakkında’, Türkiyat Mecmuası, 7–8 (1940–2), 284–5. (57) The circumstances of how this term appeared and developed in the Ottoman and Greek contexts bear further study. My thanks to Walter Kaegi for bringing this absence in the Greek sources to my attention. Kaegi suggests it may be significant that no example of the term appears in Emmanouel Kriaras’ exhaustive multi-volume Lexiko tes mesaionikes Hellenikes demodous grammateias 1100–1669 (Thessalonike: Royal Hellenic Research Foundation, 1968-). The term in the nineteenth-century context is outlined in K. E. Fleming, ‘Armatoli’, Encyclopedia of Greece and the Hellenic Tradition, ed. Graham Speake (London: Fitzroy Dearborn, 2000), 169–70. See also Apostolos E. Vakalopoulos, The Origins of the Greek Nation: The Byzantine Period 1204–1461 (New Brunswick: Rutgers University Press, 1970), 157–60. (58) Anhegger, ‘Martolos’, 342; Anhegger, ‘Martoloslar Hakkında’, 285–6. (59) İnalcık, ‘Stefan Dusan’, 222–44; Cengiz Orhonlu, Osmanlı İmparatorluğunda Derbend Teskilatı (Istanbul: ˙ İÜ Edebiyat Fakültesi, 1967), 79–80; Anhegger, ‘Martolos’, 342. See also Milan Vasić, ‘Die Martolosen im osmanische Reich’, Zeitschrift für Balkanologie, 2 (1964), 172– 89; idem, ‘The martoloses in Macedonia’, Macedonian Review, 7/1 (1977), 30–41. (60) Griswold, ‘Martolos’, 613; Anhegger, ‘Martoloslar Hakkında’, 288, 315. (61) Cengiz Orhonlu, ‘Kharādj’, EI2, 7: 1053–5. (62) See Halil İnalcık, ‘Osmanlılar’da Raiyyet Rüsûmu’, Belleten, 23 (1959), 603–5. (63) Halil İnalcık, Fatih Devri Üzerinde Tetkikler ve Vesikalar (Ankara: TTK, 1954), 155, 179; Orhonlu, Osmanlı˙İmperatorluğunda, 87; Anhegger, ‘Martoloslar Hakkında’, 293. (64) BOA, D.BKL 32187; D.BKL 32195; MAD 2052; MAD 5158; Ottoman Garrisons on the Middle Danube; Claudia Römer, Osmanische Festungsbesatzungen in Ungarn zur Zeit Murāds III (Vienna: VÖAW, 1995), 31. (65) İnalcık, ‘Stefan Duşan’, 243. (66) Ibid. (67) Ottoman Garrisons on the Middle Danube. (68) BOA, MAD 4000; MAD 5820; MAD 7208; Mühimme Defteri 216. (69) BOA, MAD 4000; MAD 5820. (70) Anhegger, ‘Martoloslar Hakkında’, 287. (71) Ottoman Garrisons on the Middle Danube. (72) Ibid.

Page 13 of 14 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Military Service and Material Gain on the Ottoman–Hapsburg Frontier (73) BOA, MAD 5158. (74) BOA, MAD 7320. (75) BOA, D.BKL 32208. (76) BOA, MAD 5820; MAD 7208; MAD 2113; MAD 4457; MAD 2846; MAD 3645. (77) BOA, MAD 2113; MAD 7162; MAD 4457; MAD 7003; MAD 16606; MAD 2846; MAD 6824; MAD 6790; MAD 3645.

Access brought to you by:

Page 14 of 14 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Ottoman Suakin 1541–1865: Lost and Found MICHAEL MALLINSON LAURENCE SMITH COLIN BREEN WES FORSYTHE JACKE PHILLIPS

DOI:10.5871/bacad/9780197264423.003.0024

Abstract and Keywords The island town of Suakin (Ott. Sevvakin) was one of the major Red Sea ports and, for a short period, the capital of the Ottoman eyelet of Habes. It lies 60 kilometres south of present-day Port Sudan, and has recently been the subject of a Sudanese-British collaborative archaeological project focusing on three main areas of research: archaeological study of the development of the settlement, architectural study of the ruins, and the future protection of the place as a cultural site. This chapter summarises the aspects of the project reflecting Suakin's Ottoman history. The study identifies material confirming the activities that led to this prosperity, namely trade. The archaeological evidence recovered in the recent excavations does support the existence of a wide-ranging trade network into which Suakin was linked from the earlier Ottoman period, covering neighbouring areas but also extending to east and south-east Asia. Keywords:   Suakin, Red Sea ports, site history, archaeology, Ottoman period, port trade

THE ISLAND TOWN OF SUAKIN (Ott. Sevvakin) was one of the major Red Sea ports and, for a short period, the capital of the Ottoman eyalet of Habeş (see Figure 24.1).1 It lies 60 kilometres south of present-day Port Sudan, and has recently been the subject of a Sudanese-British collaborative archaeological project focusing on three main areas of research: archaeological study of the development of the settlement, architectural study of the ruins, and the future protection of the place as a cultural site.2 This chapter summarises the aspects of the project reflecting Suakin’s Ottoman history. The port long had significance as a focus for trade situated at the frontier of the Ottoman and Funj territories. During this period, many of its more famous buildings were constructed. Tragically, most have vanished under the rubble of Suakin’s collapse since the site was

Page 1 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found abandoned in 1922. Thanks to the large photographic record and the excellent studies of J.-P. Greenlaw3 and F. W. Hinkel4 some of these buildings remain well known and the memory of Suakin as a place of cultural importance has survived. (p.470)

(p.471) Site History and Archaeology The earliest history of Suakin so far known5 dates back to the Christian Kingdoms of Sudan when in the tenth to twelfth centuries Suakin formed a trading point with the Red Sea, although an earlier Roman port, called Evangelon Portus by Ptolemy (fl. ad 121–51), may have existed there.6 How permanent this or the Christian port were remains unknown. Salim alAswani (fl. 975–96) describes Suakin as linked by caravan to Shankir near Berber, and this route would seem to be the reason for its survival. The port was under the joint control of both the traders and the local Beja tribes for most of its early history, although writers such as Mas‘udi (c.896– 956) and Ibn Sa‘id (1208–86) refer to a tribe called al-Khasa, who came from near Figure 24.1. Map showing location of Suakin Asmara, as controlling the town under their and other sites mentioned in the text. king. These early links with the Nile Valley and Ethiopia give a good indication of the port’s significance. The early importance of the town was shared with ‘Aydhab to the north and Badi’ to the south, the later importance of the town being mostly attributable to the decline of ‘Aydhab in the earlier fifteenth century. Later in the century, trade was taking place with Indian and Venetian traders, and by the sixteenth century the port was reported as having well-built houses. Having been under Egyptian Mamluk control in the late fifteenth century, Suakin was surrendered to the Ottoman Turks in 1517 and from then on, into the nineteenth century, had an Ottoman garrison, with two small forts recorded as being built on the mainland in the seventeenth century. First used as a base for the fleet fighting the Portuguese, it became in 1540 the supply base for the conquest of Ethiopia and a centre of the eyalet of Habeş. With the failure of that scheme by the 1580s, it was returned to the eyalet of Egypt, being later transferred to the control of Jeddah while remaining the main port for the Funj Sultanate. In the mid-nineteenth century the town came under Turco-Egyptian rule, initially under Mehmed Ali Pasha and then more permanently under the Khedive Isma‘il in 1865, and the port again prospered with the opening of the Suez Canal in 1869.7 This period of prosperity was short-lived, though responsible for most of the larger public buildings. Suakin was fortified during the Mahdiyya by Kitchener until 1888, and never succumbed to Mahdist troops. The revival of trade with the west following this period was responsible, however, for the final destruction of Suakin. A new port was built 60 kilometres Page 2 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found north at Shaykh al-Barghuth and by 1922 Suakin was effectively abandoned. The majority of the houses surviving in 1951 when surveyed by Greenlaw had a character that reflected the rich history of Suakin’s past, and its rapid (p.472) deterioration has been a great tragedy not just for the material culture of Sudan but also for the local population, as many of the houses are still privately owned. Cartographers provide further evidence of the site’s history presumably reflecting accounts supplied by travellers during this period, in a series of published maps showing the boundaries of Ottoman Egypt and the Funj Sultanate. These indicate that Suakin was not recognised by European cartographers as part of the Ottoman Empire until 1740 when Lotter’s map of the Turkish Empire8 shows it in the Kingdom of Habelthi. Earlier maps of Ortelius (1590),9 Linschoten (1596)10 and Speed (1626)11 show Suakin as part of Nubia. The map of Lázaro Luís in the Portuguese Academy of Sciences, dated 1563,12 shows a crescent flag flying over ‘Suaquem’. Morden’s map of 168713 shows Suakin as part of Beg de Habeleth. The beginning of this separation is most clearly indicated on Chatelain’s map of 1711,14 which divides Nubia, designated as a Province of Barberosa (i.e. Berberistan) under Turkish control, from the Kingdom of Sinnar (i.e. the Funj) in the south. Suakin appears just inside the area controlled by Sinnar. Although Reynolds’ map of 177115 shows it again as part of Nubia, Dunn’s map of 177416 has Habeş as a separate area, and so clearly still under nominally separate control. Bonne’s 178717 map shows Suakin on the coast of Habeş but in a Nubia divided between Turkish Nubia and the Kingdom of Funj. In the 1812 map of Pinkerton18 and in Meyer’s map of 1845,19 Suakin is shown as firmly part of Nubia, with Nubia as part of Egypt. The border with Ethiopia was quite mobile and appears as late as 1860 as just south of Tokar. Cartographers’ maps have also assisted us in identifying certain early key buildings of Suakin. Under the direction of Napoleon III, Guillaume Lejean published the first accurate street map following his visit between February and August 1862. This map (p.473) shows a number of features of the island including a building called Bayt al-Mufti.20 Several contemporary maps from 1860, such as that of Petermann21 and the 1874 map in the ‘Dépot des Cartes et Plans de la Marine’ series, based on Captain Pullen’s 1858 survey,22 are sufficiently detailed to show fortifications along the north shore relating to the late Ottoman or Turco-Egyptian defences, now lost. These may have been dismantled in order to build defences around the Geyf, the mainland part of the town, in the 1860s–80s and so also remove defences that could have been used against Gordon and, later, Kitchener’s naval support. Written and cartographic descriptions of Suakin show it as in a constant state of flux during the Ottoman period, and this tension between control from the mainland and the coastal states is still characteristic today. The role the Beja23 have played in this has been central to the development of the area. Their control of the hinterland has demanded that whoever sought to control Suakin needed their help for supplies or the delivery of trade goods to the interior. The importance of their role as pastoralists is not so visible in the material culture as the Island Town’s, but so many of the Hadradib24 traders married into Beja families that they created a separate tribe called Arteiga, whose ‘umda still has charge of Suakin.25 This has resulted in a Suakini culture reflecting Hadradib tradition while developing its own unique character, which the Hadendowa Beja also see as part of their own. It also meant that Suakin’s rulers in this period shared a culture with a tribe whose influence extended across the Bayuda and Butana Deserts as far as the river between Dongola and Shendi.26

Page 3 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found The 2002–7 Project Scientific study of the remains at Suakin has only recently begun. The first records were made of the island by J.-P. Greenlaw with students from the Faculty of Art at the then Technical College of Khartoum. Over several seasons in the 1950s the standing remains were drawn and measured, including detailed studies of parts of the buildings, (p.474) carved stone doors, rawshan windows27 and decorative plasterwork. The results were published in 1976 as the The Coral Buildings of Suakin. Later F. W. Hinkel undertook studies of a number of buildings still visible in the 1970s and published these in The Archaeological Map of the Sudan, which corrected some of Greenlaw’s dimensional surveys. The 2002–7 project has aimed to improve the level of knowledge contained in these two volumes by undertaking for the first time excavations on Suakin Island. Our research seeks to examine the development of the town, including its trading links and, in particular, how the unique structure of the island was attained. Study of the early quaysides at the Muhafaza (Governor’s Residence), and adjacent to the Bayt al-Mufti, clearly shows the addition of successive quays around an earlier core to the island (see Figure 24.2). This suggested that the earlier material would be found towards the centre of the island. This idea was supported by the excavations in 2006 outside the walls of the Bayt al-Basha. These uncovered fifteenth to sixteenth-century Mamluk to early Ottoman material, possibly reflecting the initial period of conquest of the island. The Bayt al-Mufti itself and Bayt Khurshid Efendi were excavated since they were considered to be significant examples of major early Ottoman buildings.28 Initial studies of the surrounding areas have uncovered only the remains of nineteenth-century fortifications. The missing Ottoman and Funj forts, noted above, seem to be buried under modern constructions. The structure uncovered by Chittick in 1954 would seem to pre-date the current archaeologically known earliest settlement on Suakin Island, which is from the Islamic period; it was made of bricks similar only to those found at Adobana near ‘Aqiq, where column drums in a Classical style have also recently been noted.29 The early Islamic site further south at Badi’ and the northern port of ‘Aydhab were both abandoned before the Ottomans had control in this region. Suakin alone holds the key to the archaeology of the Ottoman period on the Sudanese coast, as its history encompasses the entire period. It provides an interesting balance to the site at Qasr Ibrim in the Nile Valley,30 which remained firmly under Ottoman control, whereas the fortunes of Suakin fluctuated between the control of the Hadradib and the Beja. The rich local culture this created still survives in the traditions of local people. The third objective of our project, as noted above, is to restore the island as a cultural centre to display and preserve the unique history of this area. (p.475)

Page 4 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found (p.476) Excavations of Main Ottoman Buildings The earliest descriptions of the island up to the sixteenth century suggest a few poor stone buildings surrounded by straw huts as the main features of its architecture. From this grew the complete Island Town as surveyed in 1903 by the Sudan Survey Department. The first recorded depiction of the island was that by Juan de Castro in 1541.31 This was after the Ottoman occupation of the island in 1517 and shows a substantial number of stone buildings, including an apparent mosque, crowded together on an irregularly shaped island. A number of other salient features are apparent, the qubba on Condenser Island, the cemetery to the south of the Main Channel, and the qubba on Graham Point at the entrance to the channel. A small fort is also marked on the land which became an island at certain seasons, north of the Main Channel. Most of these features could still be identified by Hinkel in the 1970s. This, in part, is surprising because the island supposedly had been destroyed by de Castro’s fleet on returning from Suez in revenge for the islanders having warned the Ottoman Commander Selim that the Portuguese were on the way, so permitting him to draw his fleet out of the water and prevent its destruction.

Figure 24.2. Plan showing location of main sites investigated and other features on Suakin Island from plans by H. Barnard, M. Mallinson (after Greenlaw, Coral Buildings of Suakin), C. Breen and W Forsythe.

Bayt al-Mufti As the prime material on the site dates from the Ottoman period, the priority was to study the archaeology to clarify the development of the island. The excavations in 2002 started out by trying to identify buildings evidently linked to Ottoman rule. The Bayt al-Mufti (‘House of the Mufti’, named after the Ottoman religious official in the town) was a clear starting point (Figure 24.2, Plan A). The extreme erosion of the site at this location made excavation possible without clearing much building debris, and the foundations of a coral-block administrative building were revealed.32 It was constructed over a number of earlier burnt layers, perhaps remains of wooden buildings destroyed by fire by the Portuguese or a layer used to damp-proof the buildings;33 radiocarbon dating may clarify this by indicating if the layer dates to the earlier sixteenth century. Among the finds was a button with Turkish star and crescent. Excavations revealed the remains of a nineteenth-century cannon from the Northern Battery marked on the maps, being reused as a bollard, and another smaller cannon (p.477) from the Southern Battery was found on the beach inside the Customs Area.34 They are both light pieces, but sufficient to intimidate any smaller vessels; against any man-of-war they could have put up Page 5 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found no defence. Ottoman control, at least on the island, was evidently maintained with a very light touch according to the currently available data. Bayt Khurshid Efendi Khurshid Efendi’s house (on plot No. 35)35 lies on the easternmost part of the island, immediately beside the National Bank of Egypt building. Greenlaw reckoned it to be the oldest building on the island after Bayt al-Basha, originally dating probably to the sixteenth century or so. It follows a mostly single-storied plan, with a large open courtyard behind the main house that is surrounded by numerous outlier rooms (Figure 24.2, Plan B). Greenlaw’s drawing of its two-storey high diwan 36 has ensured the house architectural fame, thanks in large measure to its lavishly decorated plasterwork (Figure 24.3). In 2002 only the north wall of the diwan still remained standing, partially and rather precariously to 7 metres’ height. Everything else had completely fallen to rubble and was infilled with sand to at least a metre’s depth, often more. Three seasons, so far, have been devoted to clearance of four main areas of the building (see Figure 24.2, Plan B). Work in Areas 2 and 337 cleared the diwan interior and its associated storeroom, Area 6 behind, to floor level, mostly by removing the coral blocks of its collapsed walls and exposing the stone-flagged floor. Many of the decorated blocks are identifiable in old photographs and can be repositioned on paper. Also recovered here were large quantities of wood, some being from the interior fittings, including the decorated door into the extension, and others being furniture fragments and plain planks, probably from interior shelving. The mastaba of the upper diwan was also exposed. Area 4 is a trench that extended from the house façade to the shoreline. The façade was cleared to house foundation level, revealing five successive layers above the foundations. Layers 1, 3–5 were of grey or brown sand, while Layer 2 was a black layer apparently of burnt material. A few finds consisting of fragments of iron and copper alloy objects, and fragments of china and glass, were recovered from the top two layers. The excavations also exposed a hard cobbled stone surface, in Layer 2 in the front courtyard, which could be traced all the way to the shoreline. Nearing the water, the cobbles suddenly become much larger stones and then a double row of large ashlar blocks, at a different alignment to both the house and the then waterfront, probably the (p.478) remains of an unknown earlier structure. This building has not yet been investigated further, and it is unclear if the cobbled surface actually belongs with Khurshid Efendi’s house or the earlier structure, if either. Work in Area 1 initially consisted solely of exposing whatever remained of the rawshan, the Suakini version of the woodwork oriel window, at the front of the house that had collapsed in situ. Clearance exposed the considerable depth of deposit of the collapsed elements and their fragments protected by accumulated sand, many of which (p.479) could be identified

Figure 24.3. View of the south wall, Khurshid Efendi’s diwan, c.1929. Photograph courtesy of the Greenlaw Archive.

from Greenlaw’s drawing and early photographs.38 Many elements were in near-perfect condition, but others had rotted very badly from exposure to the salt air and wind. We retrieved all surviving pieces of the rawshan, and exposed its wooden and stone base supports and flanking mastabas, all still in situ. The wood has been considered to be teakwood imported from Java, although more locally available wood-types are present.39 In keeping with other known

Page 6 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found examples, the rawshan was painted in red and green, and the finials in dark turquoise, some still surviving. These are the elements that survived to the final use and collapse of the house. Close study of various photographs of the house and its façade, taken at different dates from 1889 to the 1980s, show that alterations, repairs and replacements were a regular feature of the building’s use and recorded lifespan. We can only speculate on how long any of the other elements have survived, and what was there before them. Very little, if anything, is as old as the original building. The last area upon which we have concentrated so far is the main house south of the diwan, incorporating the entrance passage, guest room, stairwell, side entrance and other rooms. In general many resembled Greenlaw’s published plan, but the details of the room layout bear little resemblance to it, although much of what he has recorded elsewhere has been verified. Opposite the stairwell was found a hamam, a toilet with slightly depressed floor, smoothed and bevelled at all edges with a hardened stone conglomerate, an open cesspit covered by positioned ashlar blocks, and a raised stand for a large water pot in the back corner. The westernmost room to be exposed in 2004 boasts two wall niches. A single post still at the northern end inset of the east wall strongly suggests that this room had a door at the entrance, since its post continued below floor level and would have to have been a pivot for a single door. This entire area also underwent multiple recognisable alterations over the years. The hamam room has a wall-blockage on its east wall, next to the south side entrance, that must at one time have been an open doorway, and the bevelled floor edging was added after this wall was blocked. This suggests that the room was not originally a hamam. Nonetheless, the present hamam entrance apparently was not cut through the wall after conversion. The narrow wall delineating the front courtyard all the way to the shoreline, not present on Greenlaw’s plan, also has a small wall-block at what must originally have been access to the front door from the south. The detailed arrangement of this area is still to be elucidated in future seasons. Work on this building has only just begun. The future plan is to reconstruct fully40 and redecorate the house as a prime example of Suakin domestic architecture, and turn it into a museum of the island’s life and history. (p.480) Bayt al-Basha In the centre of the island terrestrial excavations concentrated both within and around the Bayt al-Basha (House 184, ‘the Pasha’s House’, Figure 24.2 Plan C, Figure 24.4). Greenlaw considered this building to be the earliest in the town on the basis of local lore and through surviving decoration on the diwan arch, which featured motifs common in sixteenth- and seventeenthcentury palaces in Cairo.41 The house is located in the central, high point of the island, 50 metres south of the Hanafi Mosque. It formerly consisted of a two-storey house with a diwan and dihliz fronting on to a narrow laneway to the north and additional yards and ancillary buildings to the rear. Excavations demonstrated through a radiocarbon date on charcoal that the house was built in the sixteenth century (390±40 bp). It is likely that this was one of the primary administrative buildings of the Ottoman elite following their arrival in the second decade of the sixteenth century. The town at this time adopted the typical layout of an Islamic urban centre with the central area featuring mosques and the buildings of the mercantile and administrative elite. Over the next 300 years the building retained its importance. During the nineteenthcentury colonial period some refurbishment was carried out when a number of additional

Page 7 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found buildings were constructed in the rear yard. These served as quarters for visiting British military personnel, reflecting an interesting continuity as a focus for administrative staff. (p.481) Excavations outside this house site, c.30 metres to the west, have identified fifteenth-century pre-masonry period structures. A sequence of stratigraphic layers clearly shows the presence of Figure 24.4. Plan of Bayt al-Basha (House 184), showing diwan, dihliz, yard and buildings of post and probable wattle ancillary buildings. From Greenlaw, Coral construction. The remains of a sixteenthBuildings of Suakin, reproduced courtesy of century wooden building (354±30 bp), the Greenlaw Archive. consisting of a domestic house with central hearth, were uncovered during the course of the excavation. Pottery from this structure comprised exclusively local domestic wares. Its radiocarbon date range shows that it was in use during the initial phases of the Bayt al-Basha’s occupation. This implies the co-existence of both an Ottoman administrative presence and an indigenous population in Suakin in the sixteenth century.

Suakin Port and its Trade The Suitability of the Harbour The attractions of Suakin as a maritime port are varied. The island itself would originally have consisted of a low-lying natural coralline outcrop which became the central focus of settlement. Suakin provides a sheltered anchorage in an otherwise largely inhospitable coastline. Its approaches are safe and relatively easily navigable for shallow draught wooden sailing vessels. Its bathymetry and bottom type allow for good holding ground, given that the internal natural basin acts as a depositional environment for sediment coming in from the surrounding lands. Vessels can then lie safely at anchor for extended periods of time while being easily observed from the military and economic strategic points on both the island and the mainland, including the forts (p.482) and the entrance to the channel, from where the island and harbour could be observed. Although no deepwater berths existed prior to the twentieth century, they were hardly required given the diversity of other mechanisms visiting vessels used to both load and offload goods and passengers. While the island itself has no freshwater source, a number of wells lie close by on the mainland. It was here that the caravans from the interior stopped before undertaking dealings with the town’s merchants and administrative bodies. These caravans followed centuries-old routes through the Red Sea Hills. Three main land routes existed: Suakin to Berber, Suakin south to Ethiopia, and a northern route to Egypt. Thus Suakin was one of the primary ports of this region of the Red Sea since at least the thirteenth century and probably earlier, into the later decades of the first millennium AD. A number of historical sources attest its maritime importance. In 1330 the Maghribi traveller Ibn Battuta visited the town: We emerged at a roadstead called Ra’s Dawa’ir between ‘Aydhab and Sawakin [Suakin].… After two days’ travelling we reached the island of Sawakin. It is a large island lying about six miles off the coast and has neither water nor cereal crops nor trees. Water is brought to it in boats, and it has large reservoirs for collecting rainwater. The flesh of ostriches, gazelles and wild asses is to be had in it, and it has many goats together with milk and butter, which is exported to Mecca. Their cereal is jurjur, a kind of coarse grained millet, which is also exported to Mecca…We took ship at Sawakin for Yemen. No sailing is done Page 8 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found on this sea at night because of the number of rocks in it. At nightfall they land and embark again at sunrise. The captain of the ship stands constantly at the prow to warn the steersman of rocks.42 In 1541 the Portuguese mariner de Castro described the town as follows: The situation of the city is in this manner: in the midst of a circular nook stands a flat island, almost perfectly round, and level with the water, about a mile in compass. In this space there is not a foot of ground, but what is taken up with houses; so that all the island is a city, and all the city an island. This is Swaken…The road for ships lies round about the city to the distance of a great crossbow-shot; having everywhere six or seven fathom [of] water, so that ships may cast anchor at pleasure, in a mud bottom. This road is encompassed with a great shoal, and that by others, which render it almost inaccessible by sea.43 He later recorded that ‘the ships come up close to the shore quite around the city and may be laden by laying a plank from them to the merchant’s warehouses to the doors of which the Galleys are fastened with their beaks stretching over the streets which serve as bridges.’44 Sixteenth-century cartographic sources support this description and clearly show that Europeantype vessels berthed stem-first up to the island. This unusual arrangement (p.483) was demonstrated to be archaeologically correct following excavation at the northern edge of the island in the alleyway adjacent to the Muhafaza, a building probably significantly rebuilt by Mumtaz Pasha, the first Egyptian governor of the town, after 1866.45 A stone-built quay predating the Muhafaza building and structurally adjoined to an earlier stone building was located. It consisted of a masonry wall with steps leading down to below the water’s edge. The only in situ dating evidence found associated with it was a pre-eighteenth-century domestic potsherd. Much of the remaining strata were heavily disturbed by late nineteenth- and twentieth-century reclamation work, which created a regular stone-lined waterfront around the circumference of the island. This was used as both quayage and as a military ‘walkabout’ from the 1890s through to the town’s demise in the 1920s.46 It is tempting then to interpret this feature as the quay for an individual merchant house, given that the original entrance to this house was positioned to lead directly onto the area behind the waterfront wall. Given the cartographic evidence, further examples of such structures are likely to be replicated around the island. The Economy of the Port The harbour throughout the Ottoman period was the primary port of this Red Sea region, with merchants coming from as far afield as India.47 It was heavily involved in facilitating the movement of people performing the hajj. Suakin quickly established itself as the principal embarkation port for African pilgrims travelling to Jeddah and on to Mecca. In 1540 de Castro considered the level of commerce at Suakin comparable to Lisbon, saying that trade was carried on with both peninsulas of the Indies, within the Arabian Gulf, to Jeddah, to Cairo and Alexandria, and ‘with Ethiopia and the land of the Abeshins, from whence it hath vast quantities of gold and ivory’.48 Accounts of European visitors to the region can be used to provide a summary of the main areas with which Suakin traded and the major items traded here for most of the Ottoman period. Data for the period prior to the opening of the Suez Canal is presented in Table 24.1.

Page 9 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found The most commonly mentioned trade items during the entire period from the 1540s to 1860s are gold (from beyond Sinnar) and ivory; several records describe the former as gold dust. Pearls and tortoise shell (obtained locally) are mentioned by both Poncet at the beginning of the eighteenth century and Bruce towards its end. Although this (p.484) may be because Bruce (who never visited Suakin himself) based his account on that of Poncet, if both accounts can be taken at face value they indicate that these items of trade were important throughout the greater part of the eighteenth century. Further organic products exported included gum, coffee from Ethiopia, senna and ostrich feathers from Darfur and Kordofan, hides from Kassala, cotton, sesame oil, and cattle from local tribes. Table 24.1. Summary of trade contacts and trade items mentioned in European accounts of Suakin between the sixteenth and mid-nineteenth century.  

Places mentioned with which trade was carried on

Goods or products mentioned

de Castro, c. 1541a

Arabia

Gold

Egypt

Ivory

Ethiopia India Malaka Pegu Ovington, 1690sb

India

Gold Ivory Coffee

Poncet, 1700

c

‘the East’

Pearls Tortoises (for tortoise shell?)

Bruce, 1768–73d

China

Gold

India

Ivory Gum Arabic Cassia Myrrh Frankincense Pearls Tortoise shell Rhinoceros horn

Valentia, 1805e

Gold Ivory Slaves

Seetzen, 1808f

Gold Lime

Page 10 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found

 

Places mentioned with which trade was carried on

Goods or products mentioned Meerschaum Salt Pearls

Burckhardt, 1814g

Arabia

Gold

(Jeddah)

Dhurra Leather items Ostrich feathers Slaves Tobacco

Mansfield Parkyns, 1851h

Ivory Ostrich feathers

Baker, 1861i

Egypt

Ivory Beeswax Gum (Arabic?) Hides Senna Cotton

Sources: The works of Hinkel, Bloss, Foster and Kennedy Cooke, see below. (a) B. Kennedy Cooke, ‘The Red Sea coast in 1540’, Sudan Notes and Records, 16 (1933), 152– 3. (b) W. Foster (ed.), The Red Sea and Adjacent Countries at the Close of the Seventeenth Century, as described by Joseph Pitts, William Daniel and Charles Jacques Poncet (London: Hakluyt Society, 1949), 176. (c) Ibid., 107, 154. (d) J. F. E. Bloss, ‘The story of Suakin, Part I–II’, Sudan Notes and Records, 19 (1936), 292–3. (e) Ibid., 295–6. (f) Hinkel, Archaeological Map of the Sudan, 218. (g) Bloss, ‘The story of Suakin I-II’, 296–8. (h) Ibid., 298–99. (i) Ibid., 299.

Page 11 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found (p.485) Following the opening of the Suez Canal in 1869, Suakin’s importance in the context of trade with the Mediterranean rose. A snapshot of trade vessel activity in 1872 shows that eighteen vessels with a total tonnage of 4,331 departed Suakin (Table 24.2). The port was also involved in the transportation of slaves throughout its history. In 1874 Wylde reports that Suakin merchants were forwarding agents for mercantile houses of Jeddah and their only real trade was slaves, while a year later Junker states that the town’s ‘best traffic is in slaves.’49 Table 24.2. Tonnage of trade from Suakin in the second half of the nineteenth century. Export item

Details

Weight/form

Gum

Sent to Suez

3,460 kantars

Gum

Sent to Jeddah

13,670 kantars

Ivory

Shipped to Jeddah

30,042 kilograms

Sesame

Jeddah

4,514 ardebbs

Source: Statistique de l’Égypte, J. F. E. Bloss, ‘The story of Suakin, Part III’, Sudan Notes and Records, 20 (1937), 249. (p.486) The Extent of Trading Links of the Port During the early to mid part of the Ottoman period, particularly between the sixteenth and eighteenth centuries, trade appears to have been concentrated in four main areas (see Table 24.1): 1. Neighbouring countries then within the Ottoman Empire itself: the Arabian Peninsula through the port of Jeddah, and Egypt through Cairo and Alexandria. 2. Ethiopia. 3. South Asia. The places mentioned are generally identified with locations in the Indian subcontinent. 4. East and south-east Asia. The countries or areas specifically noted in the accounts include China and Malacca. These constitute the most extended Suakini trade links indicated by European descriptions. The main items or commodities shown in Table 24.1 have not yet been identified archaeologically. Although conditions for organic preservation on Suakin are not ideal, such material can survive, as is indicated by the presence of wooden rawshans, doors and window grilles at Khurshid Efendi’s house. However, macrobotanical remains so far identified do not include material from the organic products noted historically in these descriptions of the port. The most useful archaeological evidence in support of widespread and extensive trading links indicated by secondary European historical sources is pottery, which is not specifically mentioned in these sources. Archaeological remains significant for Ottoman-period trade have been recovered particularly in the Bayt al-Mufti and Bayt al-Basha excavations. Evidence indicating trade with Egypt is present in sherds (PS40) with a white interior glaze covered with irregular pale green and dark green stripes (Figure 24.5). Parallels have been tentatively identified among ceramics from excavations at Old Cairo, particularly in the Dayr Abu Sayfayn area. At Old Cairo, this ware is considered to be of later Ottoman date, most probably ascribable to the seventeenth to eighteenth centuries.50

Page 12 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found Several sherds may indicate trade with Asia, although the quantities of such ceramics recovered to date are not sufficient to confirm continuous trade over a substantial period. Two examples (Figure 24.6, far left, upper and lower) of blue-on-white, having decorations comprising ‘commas’ upon a number of irregularly spaced pale blue horizontal lines (PS20) and a vegetal motif (PS27), are possibly of Central Asian origin.51 Probable confirmation of the historically known trade with China is present in two sherds. One, from the Bayt al-Basha (PS01) is a small, thin-walled bowl (Figure 24.6, middle, upper), and exhibits quite characteristic underglaze enamel decoration, including a floral or vegetal motif in red and a metallic gold-coloured ‘rosette’ decoration. Although an identical decoration has not been found, the general style can be paralleled among the ‘Swatow’ wares, for example on a sixteenth- or seventeenth-century covered box,52 and the use of vegetal motifs is seen in the ‘stylised aquatic plants’ on a seventeenth-century dish.53 Such wares were widely exported from China to east and south-east Asia, Egypt (Fustat) and Europe,54 so the presence of a specimen at Suakin would be reasonable. The second sherd (PS06) is probably from a closed vessel, with blue-on-white decoration of a scene of a building by a river, with a background mountain (Figure 24.6, middle, lower). (p.487) The sherd’s fabric indicates an East Asian provenance. Although the decoration is not exclusively Chinese, the presence of the previous piece indicates that this sherd is also likely to be of Chinese origin.55 (p.488)

Figure 24.5. Sherds of green and white glazed ware, PS40. Photograph by G. Owen.

Figure 24.6. Sherds PS20, PS27, PS01, PS06 and PS09. Photograph by G. Owen.

Figure 24.7. Sherd PS03, part of open bowl. Photograph by G. Owen.

Page 13 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found (p.489) Five pieces provisionally indicate trade with south-east Asia. Two (PS09, PS03) comprise the most nearly complete vessels found (Figure 24.6, far right, and Figure 24.7). Both are open bowls, with Figure 24.8. Sherds PS24, PS25 and PS23. Photograph by G. Owen. black decoration on an off-white background, including stylised ‘tendril’ and ‘flower-head’ motifs, within panels defined by curvilinear lines. Two further sherds (PS24, PS25), blue-on-white and dark grey-on-white (Figure 24.8, left and centre), also have types of ‘tendril’ designs, with small round or oval ‘buds’, in dark paint. The last (PS23) is an undecorated base sherd with a moderately high footring, in light greyish-green celadon (Figure 24.8, right), probably of Vietnamese or Thai production, rather than Chinese.56 Although no identical design parallels to these four sherds have been noted, generally similar painted motifs and style are present on ceramics also from Vietnam and Thailand. The ‘tendril’ motif on PS25 is similar to some vegetal designs on Thai ‘Swankhalok Ware’.57 Some similarities in the representational style of the vegetal and floral motifs on PS03 and PS09 are present on bowls of the ‘Sukhothai Ware’ also originating in Thailand, dated fourteenth to fifteenth century,58 and on some examples of Swankhalok Ware, dating from the fourteenth to the first half of the sixteenth century.59 (p.490) However, parallels are not exact, and motifs with a comparable degree of similarity are present on slightly later Vietnamese wares. The motifs on the Suakin are similar in style of execution and general design, though different in specific form, to vegetal motifs on bowls and dishes of Vietnamese production dated from the late fifteenth to the early sixteenth century, recovered from the Hoi An shipwreck. The floral decoration on PS03 is similar in style to floral motifs on dishes and bowls,60 while the tendril motifs on the upper and the lower body of PS09 are similar to part of a motif on the lid of a covered box, and a motif on the side of another box, respectively.61 The fragmentary tendril motif on PS25 is reasonably well paralleled in the treatment of the ‘scrolling foliage’ design on small jars from the same collection.62 Links with Asia are supported by some identified wood types. Samples analysed to date from Khurshid Efendi’s house are either Acacia or Ficus,63 locally available at Suakin. However, identifications of samples from some window grilles carried out at the Technical University of Dresden in 1976 showed that wood from the genus Shorea (Dipterocarpaceae) had been used.64 This genus is found only in Asia (Sri Lanka to south China, Moluccas to Lesser Sunda Islands), being the most important timber genus in tropical Asia,65 and further demonstrates Suakin’s link with that region.66

Conclusion The project has investigated both historical and archaeological evidence for Ottoman Suakin. Evidence has been gained of the reasons for the initial establishment of the port and its significance during the Ottoman period that may be attributed partly to the earlier decline of other entrepôts, but also to its dynamic situation between the Ottoman Empire and the Funj Sultanate. Three of the major Ottoman buildings have been investigated, revealing both the early development of the Ottoman administration (Bayt al-Mufti, Bayt al-Basha) and the later period of the port’s prosperity (Bayt Khurshid Efendi), thus covering the entire Ottoman period at the site.

Page 14 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found (p.491) The study has identified material confirming the activities that led to this prosperity, namely trade. Although the majority of items stated in European historical accounts cited above to be traded at Suakin cannot yet be confirmed archaeologically, further work is planned to identify remains indicating the presence of the organic commodities mentioned. The archaeological evidence recovered in the recent excavations does support the existence of a wide-ranging trade network into which Suakin was linked from the earlier Ottoman period, covering neighbouring areas but also extending to east and south-east Asia. The project has also highlighted the significance of Suakin to Sudan as a historical and cultural centre. It reflects the prominent role of traders from Sudan in an international world during the medieval period, and with its trade in gold, gum arabic and incense its historical links to the trade described in this region in the classical and preclassical world, with its associated cultural memories of the land of Punt and the famed mines of Nubia and the ‘Mountains of the Moon’.67 Suakini links with the mainland populations would ensure also that, far from being an isolated kingdom, the Funj rulers of Sinnar had active links both with the Hadradib mainland and also with centres such as Mecca, Cairo and Jerusalem through both pilgrimage and trade. Suakin reveals fifteenth- to nineteenth-century Sudan as an African kingdom comparable to Ethiopia in its ancient lineage and cultural breadth. It contains in its religious buildings some of the few remaining monuments of this period. University of Khartoum excavations of the earlier phases of the Shafi’i mosque have shown that it has an earlier structure beneath it, possibly predating the Ottoman arrival. Apparently a large central mosque is depicted by de Castro a few decades after the supposed Ottoman conquest, but given the Muslim population it is likely that such a building would have existed well before this period. Suakin reflects a significant part of the early arrival of Islamic culture in Sudan, having a central position in the region through trade and its role as the point of access for Islam to Central Africa. The Sudanese are rightly proud of this role, and these studies have helped develop a clearer picture of its significance. It is hoped that they will also provide the basis for future cultural displays in the new museum and a focus for continuing research in the Red Sea area. Note. We are most grateful to Mr Hassan Hussein Idris, Director General of the National Corporation of Antiquities and Museums (NCAM) of Sudan, for his continued support of the Suakin Project and for the European section of the project; to the Director of Fieldwork, Dr Salah Mohammed Ahmed; and to the Head of Conservation, Mr Hayder Mukhtar. Special thanks are due to NCAM Architect Khalid Babikir Awad al-Karim, and to Inspector Balsam Abd al-Hamid. We gratefully acknowledge the funding from the D. M. McDonald Grants and Awards Fund of the McDonald Institute for Archaeological Research, University of Cambridge, the British Academy, (p.492) and the University of Ulster, and the support of the British Institute in Eastern Africa. We are further grateful for support provided by the academic institutions. We acknowledge the support and interest of HE the Minister of Culture, Youth and Sport, HE the Governor, HE the Minister of Culture, and the Muhafiz of the Red Sea State, Sudan, and the support and accommodation provided by Mr Mohammed Nour al-Hedaab of Suakin. The European section of the project is grateful for the collaboration of Sudanese colleagues Dr Intisar elZein, Professor Ali Osman, Suaad Osman Babikir, Nahid Abdel Latif and Ahmed Hussein Abd al-Rahman and Abdelrahman Ibrahim Said. Thanks are due to Ms J. Rippengal of the Department of Archaeology, University of Cambridge and Ms J. Boreham, then of the same department, for advice and loan of equipment. Thanks are also due to J. MacGinnis for

Page 15 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found initial clearance and excavation at Bayt Khurshid Efendi and to G. Breen for excavations near the Bayt al-Basha. Notes: Proceedings of the British Academy 156, 469–492. © The British Academy 2009. 470 (1) For further details on the eyalet of Habeş and references to Suakin in Ottoman archival sources, see Cengiz Orhonlu, Osmanlı İ mparatorluğu’nun Güney Siyaseti: Habeş Eyaleti (Ankara: TTK, 1996). (2) The Suakin Project has been carrying out research and restoration work at Suakin since 2002 under the National Corporation of Antiquities and Museums (NCAM) of Sudan. The project team has been made up of members from the Universities of Cambridge, Khartoum, UCLA, London (SOAS) and Ulster, as well as the staff members of NCAM under the leadership of Hassan Hussein Idris, its Director. (3) Jean-Pierre Greenlaw, The Coral Buildings of Suakin (Stocksfield: Oriel Press, 1976). Later republished as The Coral Buildings of Suakin: Islamic Architecture, Planning, Design and Domestic Arrangements in a Red Sea Port (London: Kegan Paul International, 1995). (4) Friedrich W. Hinkel, The Archaeological Map of the Sudan. Fasc. 6: The Area of the Red Sea Coast and Northern Ethiopian Frontier (Berlin: Akademie Verlag, 1992), 214–76. (5) Hinkel, Archaeological Map of the Sudan, 215–16. (6) In 1954 Neville Chittick cleared and recorded a cistern on nearby Condenser Island, proposing this as evidence for a Roman occupation of the town: H. Neville Chittick, ‘A cistern at Suakin and some remarks on burnt bricks’, Azania, 16 (1982), 181–3. (7) See Chapter 11 in this volume by John Alexander for the general context of Ottoman policies in the region. (8) Conrad Tobias Lotter, Magni Turcarum Dominatoris Imperium per Europam, Asiam, et Africam (Augsburg, 1740). (9) Abraham Ortelius, Africae Propriae Tabvla, In qua, Punica regna uides: Tyrios, et Agenoris vrbem (Antwerp, 1590). (10) Jan Huygen Van Linschoten, Deliniantur in hac tabula, Orae maritimae Abexiae, freti Mecani: al. Maris Rubri: Arabiae Freti Mecani: al. Maris Rubri: Arabiae, Ormi, Persiae, Supra Sindam usque (Amsterdam, 1596). (11) John Speed, The Turkish Empire (London, 1626). (12) Luís de Albuquerque and M. C. Henriques dos Santos, Atlas de Lázaro Luís 1563. Códice da Academia das Ciências de Lisboa (Lisbon: Imprensa de Coimbra e Simão Guimarães, 1990). (13) Robert Morden, A New Map of Arabia (London, 1687). (14) Henri Chatelain, Carte Particulière de L’Égypte, de La Nubie et de L’Abyssinie, Dressée sur les Mémoires les plus Nouveaux & les Observations les plus exactes (Amsterdam, 1711).

Page 16 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found (15) R. Reynolds, An Accurate Map of Africa from the latest Improvements and Regulated by Astronomical Observations from A new Universal Collection (London: T. Cooke, 1771). (16) Samuel Dunn, ‘A Map of Abyssinia and Nubia’, in New Atlas or Mundane System of Geography (London: Robert Sayer, 1774). (17) Rigobert Bonne, Égypte, Nubie et Abissinie (Paris, 1787). (18) John Pinkerton, Abyssinia, Nubia Etc (London, 1812). (19) Joseph Meyer, Neueste Karte von Arabien (Hildburghausen, 1845). (20) Guillaume Lejean, Voyage aux deux Nils. Nubie, Kordofan, Soudan oriental, exécuté de 1860 à 1864 par ordre de l’empereur Napoléon III (Paris: L. Hachette, 1865), 4: 192 ff. (21) A. Petermann, ‘Das rothe Meer und die wichtigsten Hafen seiner Westhalfte’, in Mittheilungen aus Justus Perthes’ geographischer Anstalt über wichtige neue Erforschungen auf dem Gesammtgebiete der Geographie (Gotha: Justus Perthes, 1860), 6, taf. 15. (22) Dépot des Cartes et Plans de la Marine, Mer Rouge: Port de Sawakin ou Soua Kin, Levé en 1858 par le Cap. W. J. S. Pullen R.N. (Paris, Lemercier et Cie., 1874). (23) The Beja are the nomadic tribes that live along the Red Sea coast; their three principal tribes are the Rasheeda, the Hadendowa and the Beni Amer. Their language is similar to the Tigrinian language of northern Ethiopia. (24) The name derives from the tradition of many inhabitants of Suakin being descendants of people originating in the Hadramaut (Hinkel, Archaeological Map of the Sudan, 218). (25) The ‘umda is the head of the community; he is leader in spirit, but the position is not a government post. (26) Gudrun Dahl and Anders Hjort-Af-Ornas, ‘Precolonial Beja: a periphery at the crossroads’, Nordic Journal of African Studies, 15 (2006), 473–98. (27) Rawshan is a projecting window or balcony closed by decorative wooden grilles (Hinkel, Archaeological Map of the Sudan, 73). (28) See following sections. (29) Examined by project members and earlier visitors as recorded by Hinkel, Archaeological Map of the Sudan, 312–14, 317. (30) J. A. Alexander, ‘The Turks on the Middle Nile’, Archéologie du Nil Moyen, 7 (1996), 15–35 and idem, ‘The archaeology and history of the Ottoman frontier in the Middle Nile Valley 910– 1233 AH / 1504–1820 AD’, Adumatu, 1 (2000), 47–61. (31) Juan de Castro, ‘The voyage of Don Stefano de Gama, from Goa to Suez in 1540, with intent to burn the Turkish galleys in that port. Written by Don Juan de Castro, then a Captain in the Fleet: afterwards Governor and Vice-Roy of India’, in A New General Collection of Voyages and Travels: Consisting of the Most Esteemed Relations, which have been hitherto published in any Language: Comprehending every Thing remarkable in its Kind, in Europe, Asia, Africa, and Page 17 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found America, trans. and ed. T. Astley (London, 1745–7), 1: 107–30, also published in B. Kennedy Cooke, ‘The Red Sea coast in 1540’, Sudan Notes and Records, 16 (1933), 151–9. (32) Excavations of H. Koefoed and H. Barnard, with assistance from N. Fayers-Kerr. (33) Suggested since charcoal absorbs moisture. (34) The cannons were found close to the sites of the two batteries as known from the maps, and a photograph in the case of the Northern Battery, and hence are most likely to have come from each of these locations. (35) Greenlaw, Coral Buildings of Suakin, 25–9. (36) Ibid., 28. (37) Overseen by John MacGinnis. (38) Greenlaw, Coral Buildings of Suakin, 27 top, 29 Figure right, 107 Figure top, 110 Figure top left. (39) Ibid., 103, but also see below. (40) The reconstruction is ongoing, supervised by NCAM Architect, Mr Khalid Babikir Awad alKarim. (41) Greenlaw, Coral Buildings of Suakin, 22. (42) Ibn Battuta, Travels in Asia and Africa 1325–1354, trans. and ed. H. A. R. Gibb (London: Broadway House, 1929), 104–7. (43) B. Kennedy Cooke, ‘The Red Sea coast in 1540’, Sudan Notes and Records, 16 (1933), 153. (44) J. F. E. Bloss, ‘The story of Suakin, Part I–II’, Sudan Notes and Records, 19 (1936), 290. (45) Greenlaw, Coral Buildings of Suakin, 14. (46) D. Roden, The Twentieth-Century Decline of Suakin (Khartoum: University of Khartoum, 1970). (47) Salih Özbaran, ‘A Turkish report on the Red Sea and Portuguese in the Indian Ocean (1525)’, in idem, The Ottoman Response to European Expansion: Studies on Ottoman-Portuguese Relations in the Indian Ocean and Ottoman Administration in the Arab Lands during the Sixteenth Century (Istanbul: Isis, 1994), 104, 108. (48) Kennedy Cooke, ‘The Red Sea coast in 1540’, 153. (49) J. F. E. Bloss, ‘The story of Suakin, Part III’, Sudan Notes and Records, 20 (1937), 251. (50) A. Gascoigne, pers. comm., 2005. (51) J. Lin, pers. comm., 2006.

Page 18 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Suakin 1541–1865: Lost and Found (52) H. Honda and N. Shimazu, Vietnamese and Chinese Ceramics Used in the Japanese Tea Ceremony (Singapore/Oxford: Oxford University Press, 1993), 140, Plates 127, 128. (53) Jorge Welsh, Zhangzhou Export Ceramics: The So-called Swatow Wares (London: Jorge Welsh Books, 2006), 133, 134, 136. (54) Ibid., 38–9. (55) J. Lin, pers. comm., 2006. (56) J. Lin, pers. comm., 2006. (57) B. Refuge, Swankalok, de export-ceramiek van Siam (Lochem: De Tijdstroom, 1976), Plate 108. (58) J. Guy, J. Shaw, L. A. Cort and N. Tingley, Thai Ceramics: The James and Elaine Connell Collection (Kuala Lumpur/Oxford: Oxford University Press, 1993), Plates 124, 126. (59) P. C. Howitz, Ceramics from the Sea: Evidence from the Koh Kradad Shipwreck, Excavated in 1979 (Bangkok: Faculty of Archaeology, Silpakorn University, 1979), Figures 3–5, 18c. (60) Butterfields, Treasures from the Hoi An Hoard: Important Vietnamese Ceramics from a Late 15th/Early 16th Century Cargo (San Francisco: Butterfields, 2000), 2: Nr. 1253, 1487. (61) Ibid., 2: Nr. 2107, 1959. (62) Ibid., 1: Nr. 453, 454. (63) A. Clapham, pers. comm., 2004. (64) Hinkel, Archaeological Map of the Sudan, 221. (65) D. J. Mabberley, The Plant-Book. A Portable Dictionary of the Vascular Plants (Cambridge: Cambridge University Press, 1997), 660–1. (66) The author of this section (L. S.) is most grateful to the following for advice on the identification of the pottery included in this paper: for Egyptian sherds to Dr A. Gascoigne, for identification of east Asian versus European porcelains to Mrs M. Ward, for Chinese and southeast Asian wares to Dr J. Lin. Thanks are due to Dr Alan Clapham for advice on the wood identifications and links with Asia, to Mr L. Tiffany and Ms K. Heawood for assistance with references, and to Mr G. Owen for sherd photography. (67) R. Fattovich, ‘The problem of Punt in the light of the recent field work in the Eastern Sudan’, in Akten des vierten internationalen Ägyptologen Kongresses, München 1985, ed. Sylvia Schoske, vol. 4 (Hamburg: Helmut Buske Verlag, 1991), 257–72, and F. Hinkel, ‘Note on Ptolemais Theron site NE-37-F/21-D-4’, Archaelogical Map of the Sudan, 313.

Access brought to you by:

Page 19 of 19 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760 COLIN HEYWOOD

DOI:10.5871/bacad/9780197264423.003.0025

Abstract and Keywords The North African maritime frontier in the period under review in this chapter presents a paradox. On the one hand, there was a situation in which, to borrow J. S. Bromley's luminous phrase, ‘two societies, two conceptions of justice, collaborated and collided’. Bromley was referring to the contemporary Caribbean world of the late seventeenth-century boucaniers, but, equally, in the western Mediterranean, and on the land frontier of the North African littoral, there was a common maritime culture which shared many traits across the religious divide. In the context of the Ottoman frontier, the discussion suggests that a British archive-based archaeography of the Ottoman North African maritime frontier for the entire period from 1660 to the end of the eighteenth century and even into the early nineteenth is a subject which both has something to contribute and deserves to be taken further. Keywords:   Ottoman frontier, North Africa, J. S. Bromley, boucaniers, maritime culture, archaeography

I FERNAND BRAUDEL, THE GRAND MAÎTRE OF MEDITERRANEAN HISTORY, has remarked that ‘to draw a boundary round anything is to define, analyse and reconstruct it’.1 Clearly, to define the complex and diverse frontiers of the Ottoman Empire, not only as they existed at the time of its greatest extent between the mid-sixteenth and the late-seventeenth century, but throughout the state’s history, presents a considerable challenge to both historian and archaeologist.2 But with what sort of a frontier, or frontiers, have we to deal? Daniel Power, drawing on the work of the classical historian C. R. Whittaker, has observed of the Roman Empire that ‘the Romans’ universalist creed divided the world into lands already conquered and lands to be conquered, rather than into the empire and its neighbours.’ He accurately describes this stance as ‘an outlook which was essentially expansionist and ideological rather than defensive and territorial’.3 To bring us forward from antiquity into the post-classical world, Power also observes that what he terms the ‘non-territorial, frontierless concept of a universal Christian empire’ persisted in the Byzantine Empire and was revived in the West by the

Page 1 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760 Carolingians. All, Romans, Byzantines, Franks, ‘sought to achieve hegemony and (p.494) clientage rather than to seal off their territory against their neighbours’.4 Do the earlier empires’ views of their confinia also hold good for the Ottomans? Prima facie, this would appear to be the case; but, more specifically, can we find utility in Power’s dictum and Braudel’s concept in the case of the Ottoman maritime frontier in the western Mediterranean between, let us say, the mid-seventeenth and the mid-eighteenth century? Here we find ourselves in a post-Braudelian world in both senses of the term, but one still not that distant in time from the beginnings of the late sixteenth-century ‘Northern Invasion’ of the Middle Sea.5 Certainly, the classical Islamic view of the world, which divided the oikoumene into Dar al-Islam or the lands already conquered for Islam, and Dar al-Harb, the ‘Abode of War’, or the lands which were yet to be conquered, may be seen as a phenomenon as much imperial and Mediterranean as it was specifically Islamic. The Ottoman state, with its fluid borderlands and complex clientages in the Dar al-‘Ahd, the ‘lands of truce’, border regions stretching from Transylvania and Wallachia via the Crimea to Kurdistan and the Persian Gulf, and with its essentially opportunistic character, may be fitted into an imperial and Mediterranean pattern older than Islam by a millennium.6 As Wittek remarked seventy years ago, the ‘potentiel militaire’ of the Ottoman state was always larger than its own circumference, in spite of the rapid growth of the latter.7 Exactly the same could be said of Rome in its heyday. The Ottoman frontier, of course, possesses its own history. The constantly evolving sequence of the first Ottoman frontiers, an ‘expansion frontier’ with a high level of internal dynamic, spreading out from fourteenth-century Bithynia and Thrace, have been invoked in Wittek’s ‘gazi thesis’ as an, if not the, determinant of the whole subsequent course of Ottoman history and the precursors and models of what developed after 1453 as the borderlands of what Wittek perceptively termed ‘an empire, with its own determined and determining features, recognising no limits to the further extension of its frontiers’.8 But in the case of the serhad-i mansura, the Ottomans’ ‘ever-victorious frontier’, we may add to Wittek’s phrase the words ‘at least until the end of (p.495) the seventeenth century’, when in the negotiations which brought to an end a long and unsuccessful war, the Ottoman central regime unwillingly accepted a clearly differentiated boundary with its former opponents, a specific provision for joint boundary demarcation commissions, a timetable for its implementation, and a declaration of adherence to the concept of territorial integrity9 And yet, despite his having had much to say about the early Ottoman frontier with Byzantium, Wittek’s formulations differ in this regard from two classic statements of the frontier als Problem, the Turner thesis and the Pirenne thesis. Turner seems to have been concerned with conceptualising the American frontier as a historical phenomenon, a ‘zone of settlement and struggle with nature, rather than as a political division’,10 while for Pirenne the frontier between Christendom and Islam is seen simply as the divider of two wholly different and mutually exclusive civilisations.11 In the case of the Ottoman frontier in North Africa (Figure 25.1), unlike the situation in Hungary or Iraq, there was no major hostile entity comparable to the Hapsburgs or the Safavids in the hinterland of the North African land frontier as an obstacle to further expansion, although the obstacles were political as well as ecological and geographical.12 The true frontier there was the maritime one, therefore, formed by Ottoman corsairs in the wake of the fall of Granada at the end of the fifteenth century and, as a notable article by Andrew Hess long ago pointed out, stiffened by the Moriscos expelled from Spain in the course of the succeeding hundred years.13

Page 2 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760 II The present chapter is not the fruit of any funded workshop or team research project, but of what in this post-Annales era tends to be written off as histoire artisanale. It builds on some earlier work on the North African maritime frontier, but I wish to pay particular attention to the manifestations of human realia in specific contact zones of the ‘forgotten frontier’ from Ceuta and Tetouan to Tripoli, with particular emphasis on the port city and corsair stronghold of Algiers.14 (p.496) (p.497) Caroline Finkel and Victor Ostapchuk have defined the Ottoman north Black Sea frontier as ‘a defensive line par excellence’ for most of its existence, and have contrasted it with the Ottoman frontier against the Hapsburgs in Hungary, where the Hapsburgs were thrown on the defensive for a century and a half from the 1520s.15 But the North African frontier? In the historical sense it had always been there, long pre-dating the Ottomans, who only took it over after some nine centuries, at least if we accept the Pirenne thesis of the fundamental division of the shores of the Figure 25.1. North Africa, showing the main places mentioned in the text. Mediterranean in the mid-seventh century, not in terms of economic history and Mediterranean trade, which has always moved freely between one side and the other, but in terms of cultures, civilisations and mentalités.16 Historiographically, too, the North African frontier seems always to have been with us, at least since Pirenne’s time, and particularly in the last fifty years, since the monu-mentalisation of its mid-sixteenth-century apotheosis in the great work of Fernand Braudel, who profited from his early days as a schoolteacher in Algiers, where he heard Pirenne speak on his ideas about the closure of the Mediterranean after the Arab invasions, and where he was able to perceive the Mediterranean ‘from the opposite shore, upside down’.17 Like the situation behind the Ottoman Black Sea frontier over against the Pontic steppe, in North Africa the hinterland was an at times hostile and always largely tribal or quasi-tribal world, over which an ascendancy could be maintained by clientage, or the building of outposts, or by often fruitless military expeditions into the (p.498) interior.18 In this sense, Ottoman and French colonial policies (if we may call them that) seem to have been remarkably similar. Equally, Ottoman North Africa possessed a double frontier, one on and in the sea against the north, and one in the hinterland against the south. The land frontier was not entirely irrelevant to the maritime frontier, in that the transSaharan caravan routes and their principal commodity, the slave trade, as in the hinterland of the Pontic steppe, made their way from the respective deep interiors (Poland and Ukraine / subSaharan Africa) to and from the port cities and transhipment points on the coast.19 The North African maritime frontier, then, was defensive, from the early sixteenth century onwards, as a bulwark against Spanish attempts at conquest or reconquest or, later, from the seventeenth century onwards, against the attacks of the maritime powers, England, Holland and Page 3 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760 France. It was also offensive, as providing the fortified land bases, with ideologically committed regimes, possessing defended harbours and the necessary reservoir of supplies both human and material for the active carrying on of the Algerine and Tripolitanian and, to a lesser extent, the Tunisian corso against Christian shipping and Christian coastal populations both within and outside the Straits of Gibraltar, as well as for forward operations on the land frontier.20 It was thus a frontier society with its own idiosyncratic institutions.21 (p.499) Trade (especially the slave trade) across and on the frontier is of particular interest in the context of Ottoman North Africa in the period under review, in that two separate and discrete aspects of slavery in the early modern period come together: what one might call the traditional trade in blacks from sub-Saharan Africa, destined for Ottoman markets on the north side of the Mediterranean, and the trade in what for a better term one must call white slaves, either captured at sea or taken in small-scale razzias directed for the most part against the coasts of Spain and Italy. This is a subject which has been much studied recently by Mediterraneanists as well as by Ottoman historians working within the post-Braudelian world of seventeenth- and eighteenth-century Mediterranean studies.22 Tripoli and Morocco appear to have had a preponderating role over Algiers in the African Mediterranean slave trade: from Tripoli there were age-old routes leading southwards through the Fezzan into the Bilad alSudan; equally, from Morocco, access to sub-Saharan Africa via the Atlantic littoral was relatively easy, while access from Algiers to the deep south was more problematical. The slave trade through Tripoli was not entirely one-way in the seventeenth century: witness the case of an unspecified ‘pasha of Tripoli’ who sent fifty young male Europeans to Bornu in a human commodity exchange which also included, inter alia, 200 horses and some muskets from Tripoli and one hundred young men and one hundred girls from Bornu. It is also reported that the succeeding maî of Bornu, in view of the satisfactory performance of these Europeans—in what capacity is not stated—sent to Tripoli to ask for, and to receive, some more, in exchange for a further 200 black slaves and some eunuchs.23 On the other hand, little attention has been paid by archaeologists to the marine fortifications dating from our period, apart from the two-volume study of the coastal fortifications of Tunisia published twenty years ago by Neji Djelloul.24 To mention one such maritime architectural complex which deserves to be better known: the walled and fortified port and chantier of Ghar al-Milh, north of Tunis—better known in the seventeenth century as Porto Farina—which was attacked from the sea in 1655 in a notable operation masterminded by Cromwell’s General-atSea, Robert Blake.25 Here (p.500) at Ghar al-Milh the entire külliye comprising the original fortified settlement, with its walls, its surviving covered cross-street and its arsenal and harbour, and with its mid-seventeenth-century post-Blake Ottoman inscriptions celebrating the fort’s rebuilding by the Ottoman grand vizier Ahmed Köprülü—inscriptions which are now deteriorating into illegibility at the hands of benign neglect and the salt sea air—is in desperate need of further investigation.26 And there are many other such sites across the whole region.

III Archaeology, like history, has become a flexible discipline. We already have (to mention the themes of two recently published monographs) archaeologies of polite manners (middle-class colonial Massachusetts) and of social space (coffee plantations in the Blue Mountains of Jamaica), not to speak of the well-known appropriation of the word if not the discipline of archaeology by Michel Foucault already a generation ago.27 Accordingly, maritime archaeology, and not marine archaeology, is the theme of this chapter. Marine archaeology, in which I hasten to disclaim any competence, may be defined as a search for people, through the discovery of

Page 4 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760 sunken cities and harbours and lost ships and their cargoes and artefacts (and of course also, for our period, through documents). Maritime archaeology (coastal archaeography might be a better term) can encompass coastal defences, as at Akkerman, and—of course—documents. And, to re-emphasise the point, Ottoman North Africa did have a land frontier: Algiers, Tunis, Tripoli, not to mention the many smaller harbours, active in the first part of the sixteenth century but (apart from Oran and Bizerta—and also Ghar al-Milh) little noticed subsequently, were fortified places, safeguarded against attack by sea (and also by land), with defensive works which were modified, extended or on occasion repaired during this period. We know about them from a copious iconography and through contemporary reports, such as the ‘Estat des Cannons, aux Chateaux et autres Fortifications à Alger’, drawn up by a French spy in 1736, listing the locations and sizes of no fewer than 474 cannon strategically disposed at 17 separate batteries around and outside the walls of the city. The list is now to be found in the East Yorkshire County Archives at Beverley, lurking unnoticed in the Chevalier de Caylus papers—which were (p.501) themselves taken from a French warship captured in the western Mediterranean not long after 1736 by Admiral Henry Medley, who had East Yorkshire roots and family connections.28 Other provincial archives and collections can also contain unexpectedly valuable material. What can we say about the possible sources for maritime archaeography, and the extent to which they can help fill archaeological gaps? For example, we have a fair amount of British archival documentation at our disposal concerning the ships of the late seventeenth- to early eighteenth-century Algerian corsair fleet, for which the Algerine records before the mideighteenth century appear to be lost. We know the names—the Marigold, the Lyon, the Seaven Starres, the Musketto, the Three Pigeons—by which the ships were referred to in western sources; we can list and statisticise their equipage and armaments, from the forty-eight guns and thirty pedreroes of the admiral’s flagship, the Goulden Marygold, down to the typical six guns and fourteen pedreroes carried by a small sitia. We are acquainted with the names and often the ethnic origin—‘Turk’, Tripoline, ‘Dantziger’, ‘Dutchman’, Genoese—of their commanders, the rüesa.29 We have a vast amount of anecdotal and circumstantial information, too great to describe here, of their corsairing activities and encounters on the high seas. But, as far as I am aware, not a single wreck site connected with the Algerine—or the Tunisian or Tripolitanian—war fleets has so far been identified, let alone excavated. On the other hand, the English/British archival sources for this period are extensive and allow the peopling of the North African maritime frontier—which extends to include Lisbon, the Balearics, Livorno, Sicily and Malta, not to speak of Marseilles and the smaller ports of Provence —in a way that may not be reproducible for some other sectors of the Ottoman frontier. I make no excuse for insisting on the equal validity for research purposes of what tends nowadays to be dismissed by some historians, though thankfully not all, as mere consular records, written by dead white males.30 It is not the records, but what can be done with them that is important, and what may have been done with them in the past that perhaps forms part of the problem. Maritime archaeology, then, as defined for the purposes of this paper, can illuminate the microhistory of the highly permeable North African frontier. It entails a search for lost people, but through documents, rather than through artefacts, via a search for those individuals who spent their lives in the Mediterranean carrying trade, not just in the caravane maritime—not by any means a French monopoly—but in other small-scale (p.502) maritime activities across and along the sea frontier, well ‘below the radar’ of the top-edge maritime enterprises of the time,

Page 5 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760 such as the relatively well-charted and documented activities of the English or Dutch Levant companies.31 Below that level, there was a mass of smaller craft, the successors of the ships of Braudel’s ‘poor wretches, who left few traces of their passage’, the harbingers of the ‘Northern Invasion’, who brought the first Atlantic vessels into the Mediterranean more than two centuries earlier.32 In our period the activities of such small craft on occasion generated written records, either when things went wrong, as they often did, and they were brought into contact or collision with the British consular authorities, or with the state authorities in the ports where they touched. These vessels, and their voyages along or in and out of sectors of the southern littoral of the Mediterranean proper, anywhere from Tangier or Ceuta to Alexandretta and Cyprus, and also within what I propose to call the ‘Mediterranean of the Atlantic’, that is, the African and Iberian littorals and coastal waters from Santa Cruz (Agadir) or thereabouts northwards to Cadiz and Lisbon, do not figure overmuch in the standard accounts of British or Dutch maritime trade with the Mediterranean, by Alfred Wood, Ralph Davis, Jonathan Israel and others.33 More light has been shed on the French component of the Mediterranean caravane maritime for this period by the work of M. Daniel Panzac, but the French (as his study shows), had by the eighteenth century a preponderance in but not a monopoly of the trade.34 On the other hand, although a very large part of the voluminous English consular correspondence from North Africa for this period has survived, the consular registers for the Mediterranean stations, which would have allowed detailed work to be done on shipping movements and marine assurance, seem—with some exceptions in the case of Aleppo—to have perished in their totality.35 (p.503) We need, therefore, to make use of a variation on the theme of historical archaeology, in order to uncover the traces of the British caravane maritime of the later seventeenth and eighteenth centuries, lying in single documents or small dossiers, buried in the deep strata of often apparently unpromising or irrelevant archival deposits.36

IV This approach, which might be termed collective microhistory, or namierisation (the ‘vigorous substitution of accurate detail for the generalisations which had contented older historians’37), can be applied to an otherwise intractable subject. Take, for example, the obscure late seventeenth-century figure of Lionel Crofts, to whom I recently devoted a small preliminary study38 In the early 1670s he is to be found at Lisbon, attracting the suspicions of the English factory there as someone untrustworthy and too close to the Algiers regime.39 Late in July 1683 we find him making a living in Algiers, and taking the opportunity of a dip in the Algiers shipping market caused by the French admiral Duquesne’s bombardment of the town, to purchase from two Muslim merchants from Tunis, for 900 reales in silver, a pink—thus a small, cheaply built, lightly manned cargo vessel, Dutch in origin, name unknown, but with a figure of a man carved in the stern. This he crewed at his own expense with ransomed English slaves from the bagnio, and dispatched to load corn at Oran. The pink, scarcely out at sea, was seized by Duquesne on a specious and legalistic pretext (the very spirit of the corso!) and sent to Toulon. Crofts, well out of pocket as a result, unsuccessfully sought redress through a petition to Sir Richard Graham, Viscount Preston, ambassador from James II to Louis XIV. As a result one volume of the Preston Papers, now (p.504) in the British Library, contains, among numbers of petitions of Westmorland officeseekers, Crofts’ petition concerning his ship, an illuminating small dossier of documents in Arabic, Spanish and English, which allow a detailed reconstruction of the affair.40 The late J. S. Bromley was well aware of these possibilities: the dubious dealings of the armateurs and fitters out in post-1704 Gibraltar, or post-1708 Minorca / Port Mahon, taking

Page 6 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760 advantage—not an unfair phrase—of the novel opportunity of operating under the Union flag.41 There is also the case of the wine drinkers of Tripoli. Captain Fearne, master of the Unity pink, arrived into Tripoli, that bastion of Muslim rectitude, on 18 June 1684, ‘with a lading of the most excellent Frontignan wine from Ligorne’. Who was drinking it, or the cargo of wine brought in from Cephalonia by a Greek vessel on 8 January the same year, or the ‘about 60 Butts’—that is to say, about seven and a half thousand gallons—of wine brought in by the Unity in the previous November?42 Or what was the prior and subsequent history of a ketch, the Oxinton, and its crew, ‘belonging to Sandwich, loaden with Pilchards, Conger [eels] and Suffolk Cheese’, bound for Alicante, which had been taken near Cape St Vincent, having no papers, and no Mediterranean pass, but a bill of health from the Port of Fowey, and was brought into Algiers in March 1716?43 The crew had fled—whither?—under the impression that the Algerine which took their vessel was from Salé (there was of course a treaty between Britain and Algiers, to be renewed later that year by Captain Coningsby Norbury, RN, but there was no treaty with the corsair republic of Salé, down on the Atlantic coast of Morocco). (p.505) More complex, and meriting a separate study, is the case of Daniel Israel, a shipmaster of London, whose adventures on both sides of the maritime frontier between the summers of 1716 and 1717 encompass the ports of Gibraltar, Tetouan, Ceuta, Algiers and Minorca. According to Congreve, the governor of Gibraltar, writing on 30 July of that year to Admiral Baker, Israel had loaded his ship—another ketch, named the Sarah—at Tetouan, bound for Algiers with Muslim passengers and their baggage. The Sarah had duly set sail, but sprang a leak and with difficulty managed to make it into Gibraltar. There the ship’s bottom, found to be ‘rotten all through’, was repaired and refitted; the ship, once more under way, was then ‘taken in the night’ by what Congreve terms ‘the Ceuta Row boats’—also termed ‘the galleots of Ceuta’— that is, Spanish privateers operating out of the Spanish enclave of Ceuta. Israel had Row boats’—also termed ‘the galleots of Ceuta’—that is, Spanish privateers operating out of the Spanish enclave of Ceuta. Israel had Muslim passengers on board, who would have been fair prize for the Ceutans. De Ribadeo, the Spanish governor of Ceuta, sent the ship back. Daniel Israel may have been, as it would appear, a small-time crook and ready street brawler, but his letter to Congreve, written from Ceuta on 2 August 1716, and sent on by Thomson in Algiers in November, contains a brilliant description of the fighting on board at the time of the ship’s seizure, together with probably well-founded allegations of Spanish complicity based on inside knowledge acquired from their spies in Gibraltar.44 There was a flurry of official correspondence between Congreve and de Ribadeo, and a growing file of depositions listing the names and possessions of ‘severall Algereenes’ taken by the Ceutans. Israel seems to have returned to Gibraltar, without securing a resolution. By May 1717 Thomson, in Algiers, was in correspondence with Methuen and Addison about the affair; in the following month Israel himself was back in Algiers, where he was accused by a bölükbaşı, that is, a high-ranking officer of the Algiers Janissary militia, of striking him in the face and calling him (in good lingua franca) a ‘sansa fida’, ‘with a great deal of other abusive language’. Israel was ordered by the authorities to be bastinadoed and deported; he appears to have finally departed from Algiers on 7 August, ‘master of a tartane which he had bought of the Dey’, and carrying away with him five absconding slaves. One escapee was missed by his master after two hours: the ‘Admirall’—that is, the senior corsair reis—sent a row frigate after the fleeing Israel, which caught up with him on the high seas. Israel, one of his sailors and four of the escaped slaves took to the tartane’s ship’s boat and made good their escape; the tartane, with one recaptured slave and one English sailor, was brought in. Attempts (p.506) were made to get Thomas Thomson to pay for all this; in the end he was given six months to get instructions from England, where, Israel had claimed, he owned three houses, all tenanted, in St Martin’s Court in the Strand. Some months later, in Page 7 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760 March 1718, Israel was seized at Majorca, and put in prison by Consul Peck to wait on the king’s pleasure: it was alleged that he was then intending to cruise against Algiers.45 The affair of Daniel Israel is typical, if somewhat more complicated than most, of cross-frontier incidents along the porous and permeable North African maritime frontier in the early eighteenth century. From this same year, 1717, derives a petition from Edward Holden and Charles Hudson, merchants residing at Algiers, detailing the case of the sloop Diligence, freighted from Algiers to Tetouan and back, with thirtynine ‘Moorish’ merchants and a registered cargo valued at $11,000 on board, but taken on the high seas by two Spanish row frigates and brought into Malaga, where the passengers were all imprisoned. This affair, too, rumbled on for a couple of years before it was resolved.46 If the English government was deeply interested in the inherently mutable politics of Algiers, with its frequent changes of regime and mood, there was also interest from the Algerine side in what was going on in an equally mutable Britain. The eventual success or otherwise of William of Orange’s coup d’état furnishes a case in point. In April 1690, when William and his dethroned predecessor were both in Ireland, Lawrence Wise, who was acting as consul after the death of Erlisman, wrote to Sunderland that one of the Dey’s favourites had told him that the French commissary in Algiers had presented the Dey with a letter from James II, ‘wherein it was earnestly requested that a Warr might be made with us, which he refused’.47 Wise commented, ‘I find this is the reason which made the Government enquire of our Druggerman soe strictly concerning the affaires of Ireland and the long stay [sc. non-arrival at Algiers] of our ships.’48 (p.507) Algerian interest in the affairs of William in Ireland were reciprocated by the king and his closest advisers, at least to the extent of the king’s writing from Dublin to Algiers on 6 July 1690 with the news of his victory at the Boyne less than a week previously. By the time the letter reached Algiers, on 8 March 1691 via Tunis, William’s letter to the ‘Dovelattly’ was old news, but Cole received the polite answer that ‘they were always glad to see the English here’.49 On the other hand, it is clear that the maritime frontier was a permeable one at least for those Muslims from North Africa and elsewhere who wished to travel or ship goods by sea, and for those European entrepreneurs such as Daniel Israel, often in a small way of business, who took their opportunities where they could on either side of the frontier. Not that to travel on English ships was entirely risk-free. In November 1689 Erlisman, English consul in Algiers, was writing to London about the unfinished business of the Swan, Capt. Hesketh Hobman, with Algerine passengers on board, taken in the Aegean island of Milo by a French privateer ‘about a year’ and brought into Toulon, where the Algerines were still being held.50

V In essence, the North African maritime frontier in the period under review presents us with a not untypical paradox. On the one hand we have a situation in which, to borrow J. S. Bromley’s luminous phrase, ‘two societies, two conceptions of justice, collaborated and collided…’.51 Bromley was referring to the contemporary Caribbean world of the late seventeenth-century boucaniers, but, equally, in the western Mediterranean, and on the land frontier of the North African littoral, there was a common maritime culture which shared many traits across the religious divide. The preceding remarks have only touched on a richly documentable subject. Whole aspects— relations on the high seas; trade, commerce and currency; slavery, captivity and redemption— have had to be omitted for lack of space. Yet the picture drawn here is only a part of the whole: Algiers, like Tunis, like Tripoli, also had a land frontier, equally tense and complex, with its Page 8 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760 hinterland. My one-time geographer colleague at SOAS, Keith McLachlan, looking some years ago not at Algiers but at Tripoli, observed that ‘[a] full and satisfying explanation of the city’s relations with its northern seaborne interests and with its southern landward territories is not so easily lighted upon.’ Algiers in the early modern period is better documented than Tripoli, but I feel that (p.508) McLachlan’s further observation on Tripoli, that ‘faute de mieux’, it ‘can be accepted as a city of aliens with an overwhelming preoccupation in maritime affairs which had precious little relevance to the interests of the Arab and Berber communities of the interior’, may need to be modified for Algiers in this period. Recent stimulating work by Algerian historians, now beginning to appear after decades of colonial and post-colonial marginalisation, gives hopes of a better understanding of these complex problems.52 In the context of the Ottoman frontier, I hope to have demonstrated that a British archive-based archaeography of the Ottoman North African maritime frontier in the years around 1700—and indeed for the entire period from 1660 to the end of the eighteenth century and even into the early nineteenth—is a subject which both has something to contribute and deserves to be taken further.53 This is the more so in that at exactly this time, when the central Ottoman state was accepting (albeit with a bad grace) the novel post-Westphalia European idea of demarcated and defined land frontiers, in North Africa pre-Karlowitz attitudes of jihad fi sabil Allah were to endure officially at sea, albeit in an increasingly attenuated form, at least until the fateful year 1830.54 Notes: Proceedings of the British Academy 156, 493–508. © The British Academy 2009. (1) Fernand Braudel, The Mediterranean and the Mediterranean World in the Age of Philip II, trans. Siân Reynolds (London: Fontana, 1975), 18. (2) For a discussion of some of the varying Ottoman responses to administering its frontiers see Gábor Ágoston, ‘A flexible empire: authority and its limits on the Ottoman frontiers’, in Kemal H. Karpat with Robert Zens (eds.), Ottoman Borderlands: Issues, Personalities and Political Changes (Madison: University of Wisconsin Press, 2003), 15–29. (3) Daniel Power, ‘Introduction. A. Frontiers: terms, concepts and the historians of medieval and early modern europe’, in Daniel Power and Naomi Standen (eds.), Frontiers in Question: Eurasian Borderlands, 700–1700 (Basingstoke: Macmillan, 1999), 1–12, at p. 4, citing C. R. Whittaker, Frontiers of the Roman Empire: A Social and Economic Study (Baltimore and London: Johns Hopkins University Press, 1994), 1–97. (4) Power, ‘Introduction’, 5, citing the references at Whittaker, Frontiers, 28, n. 14. (5) Braudel, The Mediterranean and the Mediterranean World, 615–41. Cf. for an alternative (one might almost say, allohistorical) interpretation of the phenomenon Molly Greene, ‘Beyond the northern invasion: the Mediterranean in the seventeenth century’, Past & Present, 174 (2003), 42–71. (6) Under the Ottomans grants of ‘ahd-names to neighbouring tributary princes were almost invariably the first stage in the process of absorption and conquest: Halil İnalcık, ‘Dār al- ‘ Ahd’, EI2, 2: 116–17, with further references; cf. also Inalcık, ‘Ottoman methods of conquest’, Studia Islamica, 2 (1953), 103–29. ‘Ahd-names were in principle truces but, in the case of the so-called Capitulations, they were truces that were often very long-lasting: cf. Halil İnalcık, ‘Imtiyāzāt, ii. The Ottoman Empire’, EI2, 3: 1179–89.

Page 9 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760 (7) Paul Wittek, The Rise of the Ottoman Empire (London: Royal Asiatic Society, 1938), 45–6. (8) Wittek, Rise of the Ottoman Empire, 51. On Wittek’s thesis and the frontier world to which it was applied, see Colin Heywood, ‘The frontier in Ottoman history: old ideas and new myths’, in Power and Standen (eds.), Frontiers in Question, 228–50, reprinted in Colin Heywood, Writing Ottoman History: Documents and Interpretations (Aldershot: Variorum, 2002). (9) Rifaat A. Abou-el-Haj, ‘The formal closure of the Ottoman frontier in Europe: 1699–1703’, Journal of the American Oriental Society, 89 (1969), 467–75. I develop these themes more fully in my ‘Frontier in Ottoman history’, especially at pp. 240–2. (10) For some parallels between the Ottomans’ Balkan expansion frontier and patterns of settlement in the Ottoman Balkans and the Spanish Americas, when viewed as respective ‘New Worlds’, see Heywood, ‘The frontier in Ottoman history’, 243–4. (11) Cf. Frederick Jackson Turner, ‘The significance of the frontier in American history’, Annual Report of the American Historical Association for the Year 1893 (Washington: Government Printing Office, 1894, 197–227, repr. Ann Arbor: University Microfilms, 1966); Henri Pirenne, Mohammed and Charlemagne [1937], trans. Bernard Miall (New York: Barnes and Noble, 1992). (12) Cf. Braudel, The Mediterranean and the Mediterranean World, 1189–95. (13) Andrew C. Hess, ‘The Moriscos: an Ottoman fifth column in sixteenth-century Spain’, The American Historical Review, 74 (1968): 1–25. (14) Andrew C. Hess, The Forgotten Frontier: A History of the Sixteenth-Century Ibero-African Frontier (Chicago and London: University of Chicago Press, 1978) does not go beyond the late sixteenth century. Cf. for the later period Colin Heywood, ‘A “forgotten frontier”? Algiers and the maritime frontier from the French bombardment (1682) to the Algiers earthquake (1716)’, Revue d’Histoire Maghrébine, 114 (2004), 35–50; idem, ‘An English merchant and consulgeneral in Algiers, c.1676–1712: Robert Cole and his circle’, in Abdeljelil Temimi and MohamedSalah Omri (eds.), The Movement of People and Ideas between Britain and the Maghreb (Actes du IIème Congrès du dialogue Britano-Maghrébin tenu à l’Université d’Exeter, 14–17 Sept. 2002) (Zaghouan: FTERSI, 2003), 49–66; idem, ‘Ideology and the profit motive in the Algerine corso: the strange case of the Isabella of Kirkcaldy, 1709–1714’, in Carmel Vassallo and Michela D’Angelo (eds.), Anglo-Saxons in the Mediterranean: Commerce, Politics and Ideas (XVII–XX Centuries) (Msida: Malta University Press, 2007), 17–42. (15) Caroline Finkel and Victor Ostapchuk, ‘Outpost of empire: an appraisal of Ottoman building registers as sources for the archaeology and construction history of the Black Sea fortress of Özi’, Muqarnas, 22 (2005), 154. Cf. Ágoston, ‘A flexible empire’, 23, 24–9. (16) Pirenne, Mohammed and Charlemagne, 162, 163, 164: ‘…after the [Muslim] conquest of Spain, and above all of Africa, the Western Mediterranean became a Musulman lake’ and ‘[t]he ancient Roman sea had become the frontier between Islam and Christianity’ According to Pirenne this was ‘the most essential event of European history that had occurred since the Punic Wars…It was the beginning of the Middle Ages…’. Cf., for the latest correctives to Pirenne, Cécile Morrisson, Richard Hodges and David Whitehouse, Mahomet, Charlemagne et les origines de l’Europe (Paris: P. Lethielleux, c. 1996), which suggests a continuation into the seventh century of active, though diminishing, Mediterranean East-West exchange.

Page 10 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760 (17) Fernand Braudel, ‘Personal testimony’, The Journal of Modern History, 44 (1972), 452, 450. (18) As indeed it always had been. We may extend to include the Ottomans and the Algerine corsair republic in C. R. Whittaker’s eminently sensible remark that ‘[n]either Carthaginian nor Roman nor even Arab colonisers of the Maghreb possessed the military capacity and the technological resources to alter the basic relations of the population to the land and the means of production’: ‘Land and labour in North Africa’, Klio, 60 (1978), 331 (reprinted in Land, City and Trade in the Roman Empire (Aldershot: Ashgate, 1993), §I). (19) Cf. Braudel, The Mediterranean and the Mediterranean World, 181 ff. and map at p. 183. On the slave trade through Tripoli see the following articles, which tend to concentrate on the later eighteenth and nineteenth centuries, but which have relevance for the earlier period: François Renault, ‘La traite des esclaves noirs en Libye au XVIIIe siècle’, Journal of African History, 23 (1982), 163–81; Daniel Panzac, ‘Le commerce maritime de Tripoli de Barbarie dans la seconde moitié du XVIIIe siècle’, Revue d’Histoire Maghrébine, 69–70 (1993), 154, 156–7; John Wright, ‘Enforced migration: the black slave trade across the Mediterranean’, The Maghreb Review, 31 (2006), 62–70. Cf. also Humphrey J. Fisher, Slavery in the History of Muslim Black Africa (London: Hurst and Company, 2001); and John Hunwick and Eve Troutt Powell, The African Diaspora in the Mediterranean Lands of Islam (Princeton: Marcus Wiener, 2002), especially Hunwick’s remarks on p. xiii. On the Tatars’ slaving routes (kos) from the Crimea into the Ukraine, Poland and Muscovy see L. J. D. Collins, ‘The military organisation and tactics of the Crimean Tatars during the sixteenth and seventeenth centuries’, in V. J. Parry and M. E. Yapp (eds.), War, Technology and Society in the Middle East (London: Oxford University Press, 1975), 257–76. (20) There is an extensive literature on the North African corso and its Christian counterpart across the maritime frontier of the western Mediterranean and extending eastwards as far as Cyprus and the ports of the Syrian littoral. The ouvrage de base is a famous essay, co-authored by Michel Fontenay and the late Alberto Tenenti, ‘Course et piraterie méditerranéennes de la fin du Moyen Âge aux débuts du XIXe siècle’, in M. Mollat (ed.), Course et piraterie (Paris: IRHT, CNRS, 1975), 78–131. The essay has now been reissued, with an extensive updated bibliography, by M. Fontenay, in Revue d’Histoire Maritime, 6 (2006), 173–228. (21) Andrew C. Hess, ‘The forgotten frontier: the Ottoman North African provinces during the eighteenth century’, in Thomas Naff and Roger Owen (eds.), Studies in Eighteenth Century Islamic History (Carbondale and Edwardsville: Southern Illinois University Press, 1977), 74–87, 371–3. Cf. Tal Shuval, ‘The Ottoman Algerian elite and its ideology’, International Journal of Middle East Studies, 32 (2000), 323–44; Heywood, A “forgotten frontier”?’; idem, ‘An English merchant and consul-general’; idem, ‘Ideology and the profit motive’. The important new study by Lemnour Merouche, Recherches sur l’Algérie à l’époque ottomane, II: La Course: mythes et réalité (Paris: Ed. Bouchene, 2007) reached me only after this chapter was completed. (22) For a useful overview of the relevant literature see Gelina Harlaftis and Carmel Vassallo (eds.), New Directions in Mediterranean History (St John’s, Newfoundland: International Maritime Economic History Association, 2004), especially 1–83, and Fontenay and Tenenti, ‘Course et piraterie’, ‘Bibliographie complémentaire’, 217–25. (23) Fisher, Slavery in Muslim Black Africa, 255.

Page 11 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760 (24) Neji Djelloul, Les Fortifications côtières ottomanes de la Régence de Tunis (XVIe-XIXe siècles) (Zaghouan: FTERSI, 1995), 177–92, 729–40 (pl. cii–cxiii). Ghar al-Milh was founded (1638–40) as a port and fortified place by the dey Usta Murad and colonised by Morisco immigrants; cf. J. D. Latham, ‘Contribution à l’étude de l’émigration andalouse et à sa place dans l’histoire de la Tunisie’, in M. de Epalza and R. Petit (eds.), Études sur les Moriscos andalous en Tunisie (Madrid and Tunis, 1973), 29. (25) Blake’s fifteen ships took out two shore batteries and destroyed nine ‘Algerine’ galleys, which it had been intended would form part of the Ottoman fleet engaged in the siege of Candia. The episode marks the emergence of England as a major player in North African Mediterranean politics (treaties with Algiers 1655 and with Salé 1656); cf. M. L. Baumber, ‘Robert Blake’, Oxford Dictionary of National Biography (Oxford, 2004, online edition, s.v.). (26) Djelloul, Fortifications côtières, 177–92, cf. Pls. cii–cxii (pp. 729–40). (27) Lorinda B. R. Goodwin, An Archaeology of Manners: The Polite World of the Merchant Elite of Colonial Massachusetts (New York: Kluwer, 1999); James A. Delle, An Archaeology of Social Space: Analysing Coffee Plantations in Jamaica’s Blue Mountains (New York: Kluwer, 1998); Michel Foucault, L’Archéologie du Savoir (Paris: Gallimard, 1969). (28) East Riding County Archives, Beverley, Chevalier de Caylus Papers (DDGR 40), 40/29, ‘Estat des Cannons aux Chateaux et autres Fortifications à Alger’, 1736. (29) Colin Heywood, ‘What’s in a name? Some Algerine fleet lists (1686–1714) from British libraries and archives’, The Maghreb Review, 31 (2006), 103–28. (30) For a notably positive and pertinent view of the utility of North African consular records as a source for economic history see Rifaat Abou-El-Haj, ‘An agenda for research in history: the history of Libya between the sixteenth and nineteenth centuries’, International Journal of Middle East Studies, 15 (1983), 314–15. (31) A. C. Wood, A History of the Levant Company (Oxford, 1935; repr. London: Cass, 1964); Jonathan I. Israel, ‘The phases of the Dutch Straatvaart (1590–1713): a chapter in the economic history of the Mediterranean’, Tijdschrift voor Geschiedenis, 99 (1986), 1–30. (32) Braudel, The Mediterranean and the Mediterranean World, 606–7. (33) On the English Southern European and Mediterranean trades see Ralph Davis, The Rise of the English Shipping Industry in the Seventeenth and Eighteenth Centuries (Newton Abbot: David and Charles, 1962), 228–56, and, inter alia, the important study by H. E. S. Fisher, ‘Lisbon, its English merchant community and the Mediterranean in the eighteenth century’, in P. L. Cottrell and D. H. Aldcroft (eds.), Shipping, Trade and Commerce: Essays in memory of Ralph Davis (Leicester: Leicester University Press, 1981), 23–44. (34) Daniel Panzac, La Caravane maritime: marins européens et marchands ottomans en Méditerranée (Paris: CNRS, 2004); cf. Greene, ‘Beyond the northern invasion’, 48 ff, and, as correctives, George B. Leon, ‘The Greek merchant marine (1453–1850)’, in Stelios A. Papadopoulos (ed.), The Greek Merchant Marine (1453–1850) ([Athens]: National Bank of Greece, 1972), 13–20, 25–32, 41–4, 469–74, 483, and Michela D’Angelo, ‘In the “English” Mediterranean (1511–1815)’, Journal of Mediterranean Studies, 12 (2002), 271–85.

Page 12 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760 (35) For example, the records of the British consulate at Lisbon, which had been sent for security to England on the outbreak of the Peninsular War, were lost when the ship conveying them went down in the Bay of Biscay: Sir Richard Lodge, ‘The English Factory at Lisbon’, Transactions of the Royal Historical Society, 4th ser., 16 (1933), 242. The Factory and consular records of Livorno were said to have been transferred to Malta in 1807 ahead of the French, but had disappeared already when searched for later in the nineteenth century: J. A. P. Lefroy, ‘The British Factory at Leghorn: some Huguenot associations’, Proceedings of the Huguenot Society of London, 22 (1971), 82. (36) Cf., from the outermost limits of the Mediterranean ‘near’ Atlantic, TNA, S[tate] P[apers] 110/91–2, two fragmentary merchants’ letter-books for the years 1697 and 1700–1 from Santa Cruz (Agadir, Morocco, not as miscatalogued, Santa Cruz del Mar) in ‘South Barbary’, with rich details of English trade in local commodities from the interior—‘Beeswax Cake, Copper, Sweet and Bitter Almo[nds], Cordivans dressed and undrest & cowhides’—and its connections with Lisbon. I am currently preparing a study based on these documents. (37) The Observer, 28 Aug. 1960, cited in OED, s.v. ‘namierisation’. (38) Colin Heywood, ‘Anglo-Maghrebi shipbroking in North Africa in the late seventeenth century: an Arabic document from Algiers (1094/1683)’, in Toni Cortis and Timothy Gambin (eds.), De Triremibus: Festschrift in Honour of Joseph Muscat (Malta, 2005), 619–34. (NB: The article as published supplies the facsimile and translations, but omits the text of the kadı of Algiers’ hujja (hüccet) for Crofts.) (39) Crofts was arrested in Lisbon on 25 July/4 August 1670 on the orders of the consul Francis Parry, ‘to prevent him giving information to the pirates [sic] of Algiers’. Parry had earlier written to Lord Arlington concerning Croft’s intended voyage to Algiers; later, in September 1670, the Privy Council approved the arrest, ordered him to be sent home as a prisoner, and requested an investigation to be made into Crofts’ supporters in the English Factory at Lisbon: TNA, SP 89/10, ff. 264, 273. It is worth noting that Portugal’s relations—largely unstudied—with the North African states were invariably poor and her shipping constantly subject to depredation: Fisher, ‘Lisbon and the Mediterranean’, 27. (40) B[ritish] L[ibrary] MSS Add. 63776, fol. 114; 63773, fol. 227. (41) Cf. Gonçal López Nadal, ‘Corsairing as a commercial system: the edges of legitimate trade’, in C. R. Pennell (ed.), Bandits at Sea: A Pirates Reader (New York: New York University Press, 2001), 125–36; idem, ‘La course majorquine en Méditerranée durant la seconde moitié du XVIIe siècle: 1652–1698’, in Économies Méditerranéennes. Équilibres et Intercommunications, XIIIe– XIXe siècles. Actes du IIe Colloque International d’Histoire, Athènes, 18–25 septembre 1983 (Athens: Centre de Recherches Néohélleniques, 1986), 1: 474–90. Algiers and its coastal hinterland, including Oran (retaken from Spain in 1708) were important sources of corn and other foodstuffs for allied armies operating in Spain during the War of the Spanish Succession, cf. J. S. Bromley, A letter-book of Robert Cole, British consul-general at Algiers, 1694–1712’, in idem, Corsairs and Navies 1660–1760 (London and Ronceverte: Hambledon Press, 1985), 38–9. From the Algerine side ships of Gibraltar and Minorca were still officially regarded as enemy vessels: cf. Heywood, ‘An English merchant and consul-general’, 53–4.

Page 13 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760 (42) C. R. Pennell, Piracy and Diplomacy in Seventeenth-Century North Africa: The Journal of Thomas Baker, English Consul in Tripoli, 1677–1685 (Rutherford: Fairleigh Dickinson University Press, 1987), 170, 165, 164. Cf. the observations (1682) of the Marseilles Levant merchant M. Lefèvre, in Anna Pouradier Duteil-Loizidou, ‘Le commerce des Français, Anglais, Hollandais et Vénetiens au Levant à la fin du 17e siècle’, Epeteris tou Kentron Epistemonikon Ereinon, Lefkosia, 13–16 (1984–87), 385: ‘On y [Tripoli] envoye des barques entièrement chargées de vin et aussy des piastres.’ The same thought as to the Tripolitans’ uncanonical consumption of wine a century or so later appears to have occurred to Daniel Panzac, ‘Commerce maritime de Tripoli’, 161: ‘du vin en proportion élevée…, en provenance de Marseille, Candie et plus encore de Malte’, forming at a rough computation 10 per cent of imports into Tripoli for the years 1776– 83. (43) TNA, SP 71/5, 357, Thomson to Stanhope, Algiers, 5 March 1715/16. (44) TNA, SP 71/5, 405, Daniel Israel to Congreve, Ceuta, 2 Aug. 1716 (copy); cf. 407, Congreve to de Ribadeo, governor of Ceuta (copy), 27 July 1716. Despite his name, Daniel Israel appears not to have been Jewish, but may possibly have been from a convert family. There seems to have been a seafaring family of that name living in Stepney and Southwark in the early eighteenth century: Charles Israel, ‘marriner’, will made 4 October 1691; proved 19 September 1712, and John Baptista Israel, ‘of St Thomas’s Hospital Southwarke, mariner’, serving/served on board HMS Tartar, d. 1715, one or both of these possibly related to Daniel (TNA, PROB 11/524, 11/644). Neither will is particularly informative, but it may be noted that both men’s executrixes were women with Christian names. (45) TNA, SP 71/5, 473, Thomson to Addison, 8 March 1717/18. Daniel Israel may have led a charmed life. Twenty years later, in June 1739, if it is he, we find him master of the Advice, ‘a square sterned ship with 12 carriage and 4 swivel guns’, making an agreement with the redoubtable Josiah Burchett, secretary of the Admiralty, to carry naval stores to Gibraltar, together with ‘Merchant goods’ for the Turkey Company, and seeking to proceed without convoy; later the same month we find Israel, now master of the Hadgee—a wonderful name for the period, also loaded with naval stores for Gibraltar—negotiating with Alexander Bennett, the ship’s owner, to buy her for £50 down on condition that he fit her out, at which time he was to pay the whole purchase money of £140. Something fishy was going on, according to the naval authorities at Deptford, but what it was remained unclear (TNA, ADM 354/109/154, 162–3, 174). (46) The papers regarding the Diligence sloop are scattered throughout TNA, SP 71/5. Cf. SP 71/5, p. 449, Thomas Thomson, Algiers, 14 Sept. 1717, on Spanish privateering against British vessels during and after the Spanish Succession wars, describing it as ‘a very high Affront to the Brittish Bandiera’. (47) On the other hand, presumably at James’s initiative, the amirautés in the French Mediterranean ports were not involved, as their ponantine counterparts were, in registering the powers given to him in 1692, and again in 1694, to commission privateers on French soil: J. S. Bromley, ‘The Jacobite privateers in the Nine Years War’, in idem, Corsairs and Navies 1660– 1760, 150. If evidence for Jacobite privateers operating against English vessels in the Mediterranean exists it would at least be a sign of a joined-up Mediterranean strategy on James’s part.

Page 14 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

A Frontier without Archaeology? The Ottoman Maritime Frontier in the Western Mediterranean, 1660–1760 (48) TNA, SP 71/3, ff. 144b-145a, Lawrence Wise to Sunderland, Algiers, 16 Mar. 1689/90: ‘the French are very dilligent in scandalizing of us, with many lying reports’, and fol. 150r, same to same, 11 April 1690. (49) TNA, SP 71/3, fol. 200, Cole to Nottingham, Algiers, 16 Mar. 1690/1. (50) TNA, SP 71/3, fol. 133b, Erlisman to [Shrewsbury], Algiers, 30 Nov. 1689. (51) J. S. Bromley, ‘Outlaws at sea, 1660–1720: liberty, equality and fraternity among the Caribbean freebooters’, in idem, Corsairs and Navies 1660–1760, 1. (52) Cf. Farid Khiari, Vivre et mourir en Alger. L’Algérie ottomane aux XVIe-XVIIe siècles: un destin confisqué (Paris: L’Harmattan, 2002), and Lemnouar Merouche, Recherches sur l’Algérie à l’époque ottomane, I: Monnaies, prix et revenus; II: La Course: mythe et réalité (Paris: Ed. Bouchene, 2002, 2007). (53) Overall, the Ottoman and Regencies’ view of the empire’s frontier zone in North Africa seems to have been remarkably close to that of the Romans: ‘to generate tax revenue and to provide for the security of its agents (sol diers and administrators)’: David Cherry, Frontier and Society in Roman North Africa (Oxford: Oxford University Press, 1998), 74. (54) Khiari, Vivre et mourir, 104–9; cf. Daniel Panzac, Barbary Corsairs: The End of a Legend 1800–1820 (Leiden: Brill, 2005).

Access brought to you by:

Page 15 of 15 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century PAUL LANE DOUGLAS JOHNSON

DOI:10.5871/bacad/9780197264423.003.0026

Abstract and Keywords This chapter presents a synopsis of the historical evidence concerning the expansion of slavery and the trade in ivory during the Turco-Egyptian era in the Sudan between 1820 and 1881, and a description of the results of recent and very preliminary archaeological investigations at three sites associated with this trade around the town of Rumbek in Lakes State, South Sudan. The chapter begins with a brief review of the establishment of Ottoman rule in Egypt, before moving on to consider the broader geopolitical forces that gave rise to the decision by the Egyptian Khedive, Mehmed Ali, to invade Sinnar, Kordofan and adjacent areas of northern Sudan. It then discusses the economic consequences and legal changes following the establishment of TurcoEgyptian rule that helped create the conditions for the expansion of slaving expeditions into southern Sudan and the establishment of a series of fortified camps or zaribas in these areas. Keywords:   Ottoman Egypt, slavery, Turco-Egyptian era, South Sudan, Mehmed Al, zaribas, ivory trade

IN THIS CHAPTER WE OFFER A SYNOPSIS OF THE HISTORICAL EVIDENCE concerning the expansion of slavery and the trade in ivory during the Turco-Egyptian (or Turkiyya) era in the Sudan between 1820 and 1881, and a description of the results of recent and very preliminary archaeological investigations at three sites associated with this trade around the town of Rumbek in Lakes State, South Sudan (Figure 26.1). The paper begins with a brief review of the establishment of Ottoman rule in Egypt, before moving on to consider the broader geopolitical forces that gave rise to the decision by the Egyptian Khedive, Mehmed Ali, to invade Sinnar, Kordofan and adjacent areas of northern Sudan. We then discuss some of the economic consequences and legal changes following the establishment of Turco-Egyptian rule that helped create the conditions for the expansion of slaving expeditions into southern Sudan via the Bahr al-Ghazal and the White Nile (or Bahr al-Jabal), and the establishment of a series of fortified camps or zaribas in these areas. Following this, the history of the site of Pendit and other zaribas around Rumbek and the available archaeological information from these sites is considered in

Page 1 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century more detail. The chapter concludes with suggestions for further archaeological research on this topic.

Ottoman Egypt Egypt was brought under the remit of Ottoman rule in early 1517, following the defeat of the Mamluk sultanate army at Raydaniyya on the outskirts of Cairo in late January by Sultan Selim I’s forces, and the subsequent capture and execution of the Mamluk sultan Tumanbey at the end of March.1 By 1520 all revolts against the Ottomans had been suppressed, and the remainder of the Mamluk elite were either banished or chose to swear allegiance to the Ottoman sultan. Egypt now changed from being an independent empire to a dependent province under the authority of a viceroy appointed by the Ottoman sultan. The viceroy held the rank of pasha, was the supreme commander of the army, and was personally responsible for protecting the empire’s interests. Each viceroy was appointed initially for only a year, although it was common for the appointment to be renewed for another two or three years. In administering the province, he was assisted by a council of state, or diwan, which met four times a week. (p.510) (p.511) Ottoman policy towards their provinces was generally one of restrained intervention in matters that were not of direct concern to the empire. As the empire’s largest province with huge economic and strategic value, Egypt seems to have been treated as a special case, since the timar system2 applied elsewhere did not operate in the province.3 Instead, a regime of tax collectors, known as amins, was used initially. These were ‘appointed from Istanbul and delivered the revenues they collected directly to the governor’s treasury’.4 During the course of the seventeenth century, this system was gradually replaced by a system of tax farming, under which local officials, military commanders and notables (beys) competed with one another for the right to collect taxes over a particular area or enterprise. This encouraged personal factionalism, especially between beys and regimental officers. Nevertheless, these new political structures coupled with the demands for Egyptian produce in the Ottoman heartland also facilitated the growth of a mercantile

Figure 26.1. Map showing the location of Rumbek, South Sudan and other places mentioned in the text.

economy, with many merchants forming strategic alliances with the more powerful households,5 while at the same time members of the military also developed trades as artisans, shopkeepers and merchants.6

Page 2 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century Throughout the period in question there was a permanent Ottoman garrison based in Egypt, comprising seven corps or ocaks totalling about 10,000 men made up of Janissaries, an azeb infantry corps and a number of smaller cavalry and infantry units, that included riflemen (tüfenkçiyan) and ‘volunteers’ (gönüllüyan). Given the size of the province, the Ottoman garrison was relatively small and its main role was largely defensive, with responsibilities for such tasks as guarding the capital, ports and the hajj caravan, although on occasion small Egyptian contingents joined Ottoman (p.512) expeditionary forces on other frontiers.7 Despite being defeated by Selim, the Mamluk military oligarchy was retained as part of the garrison,8 and over the next few centuries, especially following a series of revolts during the latter decades of the sixteenth century and the early seventeenth century, a mamluk elite, dominated by Caucasian (and especially Georgian) households, steadily re-established itself politically, economically and militarily. In the eighteenth century, the Kazdağlı household (which was rooted in the Janissary corps) and its associates became pre-eminent, partly as a consequence of the Janissaries’ control over the coffee trade from Yemen, the Alexandria customs and the agricultural tax farms. From c .1730 they were the de facto rulers of Ottoman Egypt, especially under Ali Bey Bulut Kapan (1760–73) and subsequently Mehmed Bey Abu’l-Dhahab (Ali Bey’s former mamluk), until the latter’s death in 1775, when there was a revival of factional disputes.9 Ottoman authority was briefly re-established, but following the French invasion and occupation of Egypt (1798–1801), a political vacuum emerged which was exploited by the Macedonian Mehmed Ali following the Anglo-Ottoman invasion to re-capture Egypt in 1801. Having arrived as second-in-command of the Albanian contingent of başıbozuks sent to assist the Ottoman invasion, following a mutiny prompted by the failure of the newly installed governor Hüsrev Pasha, Mehmed Ali became their commander in early 1803. In 1805 he was nominated vali or governor by the Sultan, but quickly came to wield considerable power and to act independently of Istanbul, contriving to remove any threats to his rule by ordering the massacre of the remaining mamluk elite following a banquet at the Cairo Citadel in 1811.10

Slavery in Ottoman Egypt in the Nineteenth Century At the time of the Ottoman conquest, the Mamluk Sultanate extended across much of what is modern-day Palestine, Israel, Syria, the Hijaz, the eastern Libyan Desert, the Egyptian Mediterranean seaboard and Nile delta, and the Nile Valley to about the Third Cataract (Lower Nubia, or Berberistan as it became known to the Ottomans).11 (p.513) The Red Sea coast of what is now the Republic of Sudan, including the important town of Suakin, which had been annexed by the Mamluks in the early fifteenth century, was also brought under Ottoman control, as were the ports of Massawa and Zayla following an attempt in 1557 during Süleyman I’s sultanate to capture Abyssinia.12 By contrast, the Sudanic kingdoms immediately to the south, notably the Keira sultanate of Darfur, the Funj sultanate of Sinnar, and the Taqali kingdom of southern Kordofan, were left undisturbed and trade between Egypt and these Sudanic states continued more or less as before. While raw materials and produce, such as coffee, were among some of the commodities traded, slaves were an equally significant component. Slavery and slave trading in Egypt and the Sudan have a long history among various Christian, Muslim and pagan communities, extending back at least to Pharaonic times.13 By the seventeenth century, the use of slaves as agricultural labourers, house servants or concubines was common across much of the Arab world and Africa, including Egypt, where slaves were valued according to a hierarchy based on a combination of colour and sex. In the nineteenth century, Circassian women were at the top of this hierarchy and tended to be found mainly in the wealthiest households and most prestigious harems. Other African women and also men were lower in the hierarchy. The Mamluks themselves were also of slave origin and following their Page 3 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century defeat by Selim, mamluks continued to be bought and sold in Ottoman Egypt by the new elite. {Abbas I (1848–54), who succeeded Mehmed Ali as khedive, is said to have owned 500 mamluks as part of his slave retinue,14 for example, and the practice of obtaining new slaves from Central Asia and training them for military and political service continued during the era of Ottoman rule. Other white male slaves originated as prisoners of war captured during various military campaigns, and white female slaves of Circassian, Greek or Georgian origin were highly prized as concubines and as domestics. Ethiopians ranked next in the slave hierarchy, with Ethiopian females, who were greatly prized for their beauty, being more highly valued than Ethiopian males.15 Agricultural slavery was also practised, and male slave labour was used on large cashcrop estates, especially during periods of increased economic production such as the cotton boom of the 1860s, when there was a large influx of slaves from the Caucasus, where agricultural slavery was common.16 By far the greatest number of slaves in nineteenth-century Egypt, however, (p.514) originated from other parts of Africa. Like slaves from other geographical areas, they were engaged in a range of different tasks. Some African males worked as agricultural labourers, some were eunuchs (as also in Ottoman Turkey),17 while others served as attendants to wealthier families and as assistants to artisans, shopkeepers and merchants. The majority of domestic slaves and menial household slaves were African females.18 Military slavery, that is the use of slaves as soldiers,19 was also a common practice in the later Muslim states, and from the nineteenth century the greater proportion of Ottoman Egypt’s slave soldiers were drawn from areas south of the Sahara.20

Slave Trading in the Sudan Up to the nineteenth century, much of the trade in slaves in the Sudan was conducted at a regional level and was especially geared towards meeting a variety of local domestic, agricultural and military needs. However, there was also a flourishing export trade especially from Sinnar where, in 1700, Theodoro Krump witnessed a vibrant slave market, with caravans arriving regularly from Dongola, Nubia, Darfur and Bornu bringing slaves, and departing to the Fezzan, Cairo and further away including India.21 Slaves were also central to the economy of Darfur sultanate under the Keira dynasty,22 from whence they were exported north to Egypt along the overland route to Asyut via Baris and Kharja oasis (the so-called Forty Days’ Road, or darb al-arba‘in), and eastwards via Sinnar and across the Red Sea through the port at Suakin. Initially, most slaves came from groups living in the lower and middle Nile Valley and Kordofan, especially the Nuba Mountains. The area along the Ethiopian border also became an important source of slaves, especially for Sinnar.23 While there is evidence to indicate that Dinka and some other southern peoples were captured following raids by the Baggara and ended up in Darfur, and some appear to have been recruited as mercenaries (p.515) into the Sinnar army by the late eighteenth century,24 the regional strength of the border peoples, notably Shilluk25 and Dinka, prevented major incursions into southern Sudan until the mid-nineteenth century. This situation changed following the Turco-Egyptian invasion of northern Sudan in 1820–1, when there was a rapid intensification in the trade, and state-sponsored organised raids, or razzias, by Turco-Egyptian forces soon became an annual event with the first recorded razzia (on the Shilluk Kingdom) occurring in 1826.26 The following year, Dinka living to the east of the White Nile were targeted, and around 500 taken captive.27 Slaving parties also began to penetrate further and further south, especially following the successful penetration of the marshy swamps of the Sudd in 1839–40 by Selim Kapudan, a Turkish naval officer whose expedition was sent by Mehmed Ali. Figures concerning the number of boats leaving Khartoum each year for the south provide an indication of this increase in trading and slaving. In 1852 twelve boats are recorded as sailing to Gondokoro, the trading station established on the east bank of the White Nile close Page 4 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century to what is now the modern town and regional capital of Juba. By 1856 this figure had risen to over forty, by 1859 some eighty boats left Khartoum for the southern Sudan, and by 1863 the number had reached 120 per annum.28 There was a variety of factors behind this intensification. A primary reason was Mehmed Ali’s decision to invade the Sudan, which was motivated by two core objectives29—the need to obtain slaves for his newly established army, the Nizam al-Jadid, and the hope to gain control over various mineral resources, especially the gold deposits in the Ethiopian border area.30 These motivations arose, in turn, from his imperialist ambitions and the increasing number of wars in which the Turco-Egyptian forces were engaged, especially in Syria and Arabia. During the early nineteenth century, there was also a revival of European interest in the Red Sea area, which Mehmed Ali sought to contain by establishing Egyptian control over the ports of Suakin and Massawa in 1843, by entering into a treaty with the Sultan of Lahj in southern Arabia (p.516) in 1822 so as to provide the garrison for Aden, and by launching expeditions to conquer southern Yemen and the Bedouin of ‘Asir in Arabia.31 All of these developments helped fuel Mehmed Ali’s need for more slaves and more mineral wealth. Furthermore, although the Egyptian army that invaded Sinnar initially consisted mainly of Turks, Albanians, Kurds, Circassians and Greeks, they were not well suited to the climate and quickly became insubordinate. In response, recruitment soon began to focus, therefore, on the local populace. As early as 1821, a training camp was set up in Aswan, where captured Sudanese were vaccinated, instructed in Islam and trained by military officers, most of whom were Europeans.32 Initially, these slave soldiers came mainly from tribes in old slave catchment areas of the Nuba Mountains and the Ethiopian border lands, but by the 1840s raiding had expanded into the White Nile and Sobat basins. Some individuals also appear to have joined voluntarily. With the opening of a route through the Sudd on the White Nile, traders steadily moved further south, mounting expeditions to the west of the river. Also in the mid-1850s, traders became aware of the Bahr al-Ghazal channel flowing into Lake No from the west. The river was navigable as far as Meshra er-Rek, where a river port was established, which in March 1863, as described by the Dutch explorers Mrs and Miss Tinne,33 consisted of ‘a small pond or lagoon, in which were crowded together, in the utmost confusion, twenty-five vessels of various descriptions’.34 By the 1860s merchants had extended their activities throughout the many tributaries to the south that feed into the Bahr al-Ghazal, with more and more captives being drawn from the southern groups, such as the Dinka, Nuer, Bongo and Jur. Figures concerning the size and composition of the Egyptian army in the Sudan give a further insight into the intensification of slaving during this period. In 1832, for instance, the army consisted of just one infantry unit of under 3,000 soldiers and a number of irregular cavalry and infantry. By the following year, the Governor of Sinnar (p.517) had 5,000 regular infantry and a few thousand irregular cavalry; while by early 1838 the Egyptian army comprised five regular infantry regiments and several irregular cavalry and foot-soldier units, numbering in total around 20,000 men. Forty years on, by the beginning of 1877, the regular and irregular army in the Sudan is estimated to have numbered between 30,000 and 35,000 troops in total.35 By then there were at least 3,000 black slave soldiers serving in the Egyptian army in Egypt, and other regiments had been sent abroad, including to Mexico to assist the French in the 1860s and to Turkey to aid the Ottomans in the war against Russia in 1877.36 Aside from the demand for slave soldiers, other factors also helped drive the intensification of slave trading in the Sudan following the Turco-Egyptian conquest. These included changes in

Page 5 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century local taxation practices in the north and its impact on land tenure. Prior to Turco-Egyptian invasion, land tenure among the northern riverine Sudanese was predominantly communal in nature. Although private ownership was permissible, individual landholding was only recognised where land was being used regularly for cultivation or some other productive activity. If unused for a number of years, it reverted to communal ownership.37 The system was thus based on the principle that cultivators had a right to a share in the fruits of the land, but not the land itself. This applied equally to land used for rain-fed agriculture and to land where cultivators had invested in elements of landesque capital, such as waterwheels and irrigation ditches.38 Tithes, payable in produce from the land or otherwise in kind, were levied by the government. The amounts varied according to whether a particular piece of land was simply rain-fed, irrigated, or in the zone inundated annually during the Nile floods, and in bumper years surtax was also levied. The government was expected to provide a range of services, including protection against commodity speculation and relief during drought years, in return.39 Following the imposition of Turco-Egyptian rule, the system of taxation was changed, in no small part because of the failure on the part of the new colonial power to locate the kind of vast mineral resources that had partially prompted the invasion. Taxes were now payable either in cash, or some other commodity needed by the Ottoman state, including ivory and slaves. Since not all northern Sudanese had the same access to these resources, the changes to the taxation system stimulated changes in the growth of private ownership and the pattern of land tenure (notably increased fragmentation of rights and ownership). This in turn resulted in the marketplace becoming the main mechanism whereby rights in land were transferred, (p.518) supplanting older systems based on kinship and rights of usufruct. As more land came to be held by the merchant class, so a demand for labour to work the land was created, which was met by a rapid upsurge in the supply of slaves.40

The Zariba System In the first few decades following the establishment of Turco-Egyptian rule, slave raiding remained a monopoly of the state. In 1853, however, Mehmed Said (1854–63), who had replaced {Abbas I as khedive, gave in to European pressure and abolished government raids, and private merchants from Egypt, Khartoum, northern Sudan, other parts of the Ottoman Empire and Europe rapidly assumed control of the trade,41 resulting in the establishment of a series of networks of zaribas, across large parts of southern Sudan. Zaribas (pl. zara‘ib) were fortified enclosures established by slave and ivory traders for use as temporary holding camps, as entrepôts from which to conduct slave raids and ivory collection expeditions, as garrison posts, and as more permanent settlements. Alphonse de Malzac, a French diplomat turned hunter and trader, was the first to break out of the restrictions the Egyptian government had imposed on ivory traders working along the Bahr al-Jabal. From a base at the river port of Shambe, he created a network of strategic alliances with the Cic and Agar Dinka in 1854–6, founding his first zariba at ‘Ronga’ among the Agar (or ‘Rol’) Dinka, near the modern town of Rumbek. De Malzac exploited tribal feuds and inter-tribal wars to penetrate the interior and expand his trading network. Using musket-armed Arab Sudanese retainers as the backbone of his military force, he allied with specific groups of Dinka against their local enemies. Cattle taken in these raids were used to feed his own troops, reward his allies, and trade for ivory. Captives taken in these raids were given to his retainers as slaves. Thus the connection between ivory-trading and slave-raiding was forged.42 De Malzac also established the first set of rules for the internal administration of zaribas. These were adopted by later commercial companies for running their zaribas and became known in Arabic as the

Page 6 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century qanun Malzac (Malzac’s laws). The Austrian botanistexplorer Dr Georg Schweinfurth provides a synopsis of these in his account of his travels in the region between 1868 and 1871: First of all, the territories immediately dependent [around a major zariba] were distinctly designated. Then it provided that the approaches to a meshera (river port) should only be used (p.519) by those who could establish a claim to it. Nearly every Seriba [zariba] has its separate avenues, upon which it levies a toll, and an avenue without tolls is not a legitimate highway.…Each separate company had its own route and its own train of captains, who purchased ivory and procured a market. No new-comers were allowed to intrude themselves into an established market, or to infringe upon its trade. Fresh marts could only be established by pressing farther onwards into the interior.43 De Malzac did not live long enough to enjoy his commercial success, however, dying in Khartoum in 1860 after a long illness. His commercial interests in the south were bought up by Franz Binder, a Transylvanian merchant.44 The zariba system, which has since been associated with such famous northern Sudanese Arab slave-merchants as Zubayr Pasha, and other Khartoumers such as Andrea Dobono from Malta, and the Turk Küçük Ali,45 was thus initiated by European traders, assisted by local Dinka allies. Zaribas varied in size, according to whether they were temporary way stations along a caravan route, or small outposts, or long-term permanent settlements. The earliest examples were probably simple thorn-enclosed squares containing no more than about twenty huts, but by the late 1860s the average-sized zariba had about 250 armed men, plus their women, children and slaves.46 Most (if not all) zaribas, even the smallest such as the one at Ayod on the Duk Ridge in current Jonglei State, were fortified by a trench with earthen embankments. (In fact, abandoned zariba sites can be located on the Sudan Survey 1:250,000 maps wherever the place name Khandaq, Arabic for trench or ditch, is found.) According to Schweinfurth, the palisade at one zariba he witnessed being built in 1871 was constructed from hundreds of tree trunks ‘erected side by side and close together in a deep trench; the trench was afterwards filled in with earth’, and the whole measured ‘a hundred feet square’.47 The major zaribas developed into large, stratified settlements, provisioned in part from land cultivated within a four to five mile radius, and partly through regular cattle raids on neighbouring communities—of whom the Dinka and Jur were some of the most heavily affected. ‘Each “great trader” had a number of stations, each under an agent, who in turn presided over the soldiers, their private slaves, itinerant djellaba, or merchants, who supplied the zeriba with a variety of goods, and the allied “chiefs” of the indigenous villages in the district claimed by the great trader.’48 Some zaribas contained many houses, or huts; others may have been less heavily built on. Describing his approach for the first time towards the chief zariba under (p.520) Ghattas’s49 authority, Schweinfurth observed that from a distance ‘there was hardly anything to distinguish it from any of the villages of the Dinka’, but on getting closer he could make out that ‘the area in the centre of the huts was surrounded by a lofty square palisade’ with a narrow gateway50 The actual zariba Schweinfurth estimated to be ‘about 200 paces square [and]…crammed with huts’.51 The combined population of the zariba, including those residing outside the palisade, he estimated to be at least 1,000.52 Berlioux’s 1872 description is briefer, but informs us that typically zaribas, were square ‘intrenched camps’ measuring ‘eighty-four paces, surrounded with palisades, in the interior of which there are a certain number of conical huts, used often as habitations, or for merchandise and provisions’.53 Later descriptions of the zariba at Rumbek

Page 7 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century (see below) by the Italian soldier of fortune and ex-Garibaldini, Romolo Gessi, who visited in 1878, and that provided by Emin Pasha,54 who was based at Rumbek in 1880 or 1881 while employed by Turco-Egyptian authorities as Governor of Equatorial Province, also indicate that huts were packed closely together. Writing about the chief zariba of Küçük Ali, near the Jur River, on the other hand, Schweinfurth recounts that the ‘store-houses and the governor’s dwelling stood alone on an open space within the palisade; around the exterior, at a considerable distance, were ranged the huts assigned to the soldiers and other dependants’.55 He goes on to describe these as ‘innovations’, introduced by the governor of this particular zariba as measures to prevent disease and the outbreak of fires. Since zaribas, varied in size and importance, it is difficult to give precise figures concerning their population and inhabitants. The overall impression from the historical sources, however, is that the bigger ones rapidly became quite cosmopolitan. In 1862, for instance, the zariba at Adael/Ayak, believed to have been near the modern village of Karic south of Rumbek (see below), is said to have contained 112 ‘men from Khartoum’ and ‘six elephant hunters’,56 while two decades later Emin Pasha stated that the non-southern Sudanese population alone totalled 151, comprising seven ‘pedlars’, fifty-seven tax-payers (ushurie), thirty irregular soldiers and fifty-seven ‘persons without (p.521) employment’. Emin Pasha reckoned that these alone owned between them 1,500 slaves, while there were also present seventy-three basingers (or bazinqir—armed slaves/retainers) and their slaves, ‘a large number of Nyam-Nyam, Monbuttu, Bongo, Mittu etc.’, all of whom also had slaves.57 Precise figures concerning the numbers of slaves exported are even harder to obtain. Some estimates for the late 1830s put the number of slaves from Sudan as a whole at around 2,500,58 which is probably far too low, given estimates of the numbers exported from Darfur alone in an individual year during the 1780s and 1790s of between 4,000 and 5,500.59 Collins gives a figure of 30,000 slaves being procured during 1822 and 1823 for the Egyptian army,60 and of an estimated 15,000 slaves being sent ‘down the Nile and another 2,000 overland in 1868’.61 Southworth, who visited in 1871–2 also mentions the difficulty of obtaining exact figures, although he estimated that around 25,000 slaves were extracted annually from the Blue and White Niles, and a further 15,000 were obtained from among the Galla, and from Darfur and Kordofan. He cites General Kirkham, at the time commander-in-chief of Emperor Johannes’ army, claiming that some 90,000 slaves were also obtained on an annual basis from Ethiopia, making a total of 130,000 per annum.62 For the same period, Millingen estimates that there were around 30,000 African slaves in Istanbul and its suburbs, averaging one slave for every two houses; not all of these, however, would have been derived from the Sudan.63 Even after the formal abolition of the trade, Emin Pasha released 567 slaves from Rumbek in just one day,64 indicating just how widespread slavery, in its many guises, was. This is also supported by data provided by Schweinfurth who, in 1870, estimated that the total immigrant population in the zaribas of the twelve great Khartoum trading companies in the Bahr al-Ghazal amounted to about 10,000 men:65 soldiers, merchants, clerks, itinerant traders, holy men, and other adventurers. In addition to this the camps (p.522) included a total of 5,000 to 11,000 slave soldiers, 40,000 to 60,000 personal slaves, and about 190,000 subject population settled around the camps and along the caravan routes, who were neither slave nor fully free, and were employed as porters, labourers and agricultural workers.66 The proportion of slave to free residents in the zaribas was thus between 4:1 and 6:1, while the proportion of combined servile and subject population to free immigrants was more like 26:1.67 This division between free, slave and subject populations was probably represented spatially in the organisation of the

Page 8 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century zaribas, whereby merchants, soldiers, and probably their personal slaves resided within the entrenchments, while the subject population lived around it.68

Later History of Rumbek By the late 1860s and 1870s the Turco-Egyptian state, trying to establish its authority over the area, had taken over these camps and Rumbek became a government station and the base for a mudir or governor, who was supposed to administer the district which was called Rhol, after the local river (Rol, now Na‘am). In 1878 Romolo Gessi used it as his base to defeat the revolt of the itinerant traders, or jallaba, in Bahr al-Ghazal, led by Zubayr Pasha’s son, Sulayman Zubayr. Gessi was not enamoured of Rumbek, however, writing later that his life there became insurmountable and that: One cannot imagine a more monotonous village; the huts crowded and the air pestilential. All the neighbourhood was full of filth. The upper floor of the hut is used by the master and his family, though the last in a separate tukul or apartment; the lower floor serves as a kitchen and sleeping place for the slaves. I am so disgusted that, if fate ever brings me to these parts again, I will do my best to avoid Rumbek.69 The town was also visited in 1880–1 by another Italian, Gaetano Casati, who accompanied Emin Pasha, and described it as ‘a very populous village,70 shut into a narrow space, with huts erected upon piles or short posts. The part below the habitation, which is not enclosed, is used for the purpose of household services.’71 It is quite possible that the houses Gessi and Casati saw resembled a style of Dinka and Jur housing that is still in use today. We should note, however, that their descriptions (p.523) are at odds with the depiction of Dinka houses provided by Schweinfurth. Emin Pasha’s description of Rumbek is even more comprehensive. He noted that the town had taken its name from a local chief, Bek72 (rum being the Dinka word for forested/wooded area)—which is in line with local oral traditions—and described it as a major slave-trading centre. At the present time the station has a very uncomfortable look, owing to the former attempt to construct a moat round the station, and the numerous trenches in the yellow clayey soil, from which clay has been taken to plaster the walls of the huts and to form the floors of platforms. Besides this, wells were constructed by means of a row of deep holes…In the midst of this chaos of trenches, mounds and pools the station buildings form a kind of island. They consist of a motley group of huts built upon platforms, the irregular arrangement of which defies all description. Paths about a foot broad, and covered by every kind of filth, lead through this confusion; and as the spaces underneath the platforms are occupied by crowds of slaves, who live after their own manner, the stench and dirt are prodigiously increased…. Water is very scarce; it is obtained from deep wells, and has a dirty colour and a nauseously earthy taste. The whole situation of the station so far from all water is a great mistake, and its formation into a central station can only be explained by its convenience for the slave trade.73 At the beginning of the Mahdiyya (1881–98), the Egyptian garrison at Rumbek was slaughtered in July 1883 by a combined force of Dinka and Nuer. The official version of this defeat (reported by Emin Pasha) was that the garrison was caught outside the zariba when on an unauthorised raid against the Agar Dinka.74 Emin’s Italian lieutenant, Casati, claimed that the Dinka attack on Rumbek was ‘preceded by a great sorcerer with the cry “Ti [sic] Sebil Allah” (By God’s Mercy), the deadly invocation of the Arab, with which Mahdism had filled the air…’.75 In fact the assault on Rumbek had nothing to do with the inspiration of Mahdism. The Dinka and Nuer versions of Page 9 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century the raid are much more interesting and revolve around the personality of the famous (or notorious) Agar Dinka leader Wol (or Awol) Athiang (Wuol Athiang as he is known to the Nuer). Wol had dealings with the Egyptian garrison, as his sister was the wife (or possibly concubine and/or slave) of the Egyptian commander there. According to the Nuer, it was the Egyptians who made Wol a chief of the Agar. He therefore had inside knowledge of the zariba’s strengths, weaknesses and activities. As the successes of the Mahdi grew throughout 1882 and 1883 Wol contacted the Dinka’s erstwhile enemies, the Western Nuer (including his immediate neighbours the Nyuong, but also attracting contingents from as far north as the Jagei, around Leer), who were led by one of their own religious figures, the prophet of the Nuer divinity Teny Prior to launching the assault on the zariba, however, Wol gave a separate set of instructions to the Agar (p.524) warriors: they were to wear a strip of white cloth around their arm or body to distinguish them from their Nuer allies. Then, he said, ‘As you spear the Turuk to your right, spear the Nuer to your left.’ The assault was successful, the Egyptian garrison was wiped out, but either during or after the battle (the Nuer say after) the Agar turned on the Nuer, trying to rid themselves of two enemies at once. The Nuer version of events emphasises Wol’s treachery, but they claim that the Dinka attack on them took place after the garrison fell. The prophet of Teny had been killed in the assault, and the Nuer were returning home with their share of the garrison’s captured cattle when Wol and the Agar attacked them.76 Rumbek was reoccupied later in 1883,77 but following a combination of incursions by Mahdist forces into Bahr al-Ghazal and revolts at various garrisons, Emin Pasha withdrew the garrison stationed at Rumbek (and also that at Ayak) so as to reinforce that at Ramadi.78 No foreign troops, whether Egyptian, northern Sudanese or European, were stationed at Rumbek until 1897, when a small French post was temporarily established as part of the expedition from Gabon led by Jean-Baptiste Marchand sent to establish a French foothold in the Upper Nile, and which was later to give rise to the so-called ‘Fashoda Incident’,79 which resulted in the French forces capitulating to a British force led by Lord Kitchener and the end of French aspirations to establish colonial rule over the Upper Nile. When the Anglo-Egyptian forces arrived in Bahr alGhazal at the beginning of the twentieth century they reoccupied the Rumbek area and established the current town at some distance from the old zariba site. The new town of Rumbek was, like its predecessor, relatively cosmopolitan, with a population of Greeks, Egyptians, and people from all over Sudan, as well as a few Britons and a contingent of Sudanese troops. By the end of the Condominium, there were a British District Commissioner and two Assistant District Commissioners based in the town, as well as a Sudanese ma‘mur (administrator), and the town had a population of around 3,100.80 Wol Athiang continued an ambivalent relationship with both (p.525) the new government and his Nuer neighbours, alternately collaborating and fighting with them both. The government temporarily exiled him to Wau in 1921 and allowed him to return to the vicinity of Rumbek in 1922. He died in 1948.

Recent Archaeological Survey As part of a broader strategy to encourage the revival of archaeological and other academic research in South Sudan,81 a preliminary archaeological survey was conducted around Rumbek town in Lakes Province, Western Equatoria over the first two weeks of February 2007.82 Rumbek was selected for two related reasons. Firstly, as already discussed, one of the zaribas to be built was established near Rumbek by Alphonse de Malzac. Secondly, the remains of a large ditched enclosure with surface scatters of nineteenth-century material survives on the edge of Rumbek town, which although now known simply as Pendit, or Old Town, is known from local oral traditions to have been associated with the nineteenth-century slave and ivory trade. This site has often been presumed to be the remains of de Malzac’s base. The condition of the site also Page 10 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century makes it suitable for interpretation to a general public, and thus it was reasoned that conducting research on the site might have the desired effects of raising local awareness of the importance and value of historical and archaeological research more generally, as well as producing tangible evidence concerning the nature of zaribas and the slave and ivory trades with which they were associated. To this end, a survey aimed at locating sites associated with the mid to late nineteenth-century slave and ivory trade around Rumbek, and collecting oral information about these sites and local Dinka settlement history was carried out. The survey was constrained by limited access to a suitable field vehicle, and it was only possible to explore areas within a 25 kilometre radius of Rumbek. The published documentary sources provided the initial starting point for enquiries, with each informant being asked a series of standard questions concerning their knowledge of the site of Pendit and its history, Dinka traditions concerning the foundation of Rumbek, slavery, the trade in ivory, and exchange practices. Informants were also asked if they were aware of any other places with banks or ditches, and if they were, to name these and provide information about their location. These locations were then visited and further informants drawn from the resident population of the new area were questioned about the possible presence of a site. This strategy resulted in the location of two more zaribas—Meen Atol and Lol Nhom (Figure 26.2). Two other localities were mentioned as possibly having similar enclosures—Langdit, near Gulnam to the north-east of Rumbek, and Wulu, to the south. Enquiries around Langdit resulted in the team being taken to a Dinka shrine with historical associations, but no one who was interviewed knew of anything resembling an enclosure similar to those found at Pendit, Meen Atol or Lol Nhom. The interviews at Wulu were more promising, and several sites were mentioned, but as this was the last day of the pilot season there was insufficient time to visit any of these. (p.526) (p.527) The recording strategy at each of the three sites visited was similar, and entailed mapping the outline of the main ditch or ditches and other features where present—such as banks, collapsed buildings and house floors, using a high-resolution handheld Trimble GeoExt GPS. Digital and slide photographs were taken at each site, and walkovers were conducted in an effort to locate surface finds. Local residents were also interviewed about the site and its history. Details of the three sites are given below.

Page 11 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century Pendit

Figure 26.2. General location (A), and plans The site now known as Pendit, which some of Pendit (B), Meen Atol (C) and Lol Nhom (D) zaribas. translate as ‘old town’ (but pandit means ‘big village’, in Dinka), is located at N 6° 49‘ 11" / E 29° 40’ 58", on the northern edge of Rumbek. It consists of a sub-circular causewayed enclosure, approximately 150 metres in diameter and bounded by a ditch approximately 13–14 metres wide and between 2 and 3 metres deep (Figures 26.2b, 26.3). There is no obvious bank associated with this ditch either around the outer ditch edge or its inner side. However, the centre of the enclosure is notably higher, by perhaps 50–60 centimetres, than the ground outside the ditch. There are a total of three causeways between 2 and 3 metres wide, situated on the southern, eastern and western sides of the monument. The ditched enclosure covers a total area of around 28,895 m2, although the internal surface area measures only around 16,550 m2. In addition to the main enclosure ditches there are at least two and possibly three shallow, banked ditches radiating out from it and, on the eastern side adjacent to one of the causewayed entrances, there is a sub-circular hollow around 2–2.5 metres deep and approximately 66 metres in diameter. In the rainy season this now serves as a water hole for cattle, but this was possibly not its original function. Local informants interviewed on site in April 2005, for instance, described it as a ‘slave pit’ where slaves were held until such time as the zariba commander was ready to send them north for embarkation at one of the river ports, such as that at Meshra er-Rek on the Bahr al-Ghazal, or Meshra Aboo-kooka on the Bahr al-Jabal—the latter being, according to Schweinfurth, the ‘proper place of embarkation’ for zaribas, on the River Rol (now known as the Na‘am).83 (p.528) There are two other external features of note, both located on the north-western side of the enclosure, opposite the causewayed entrance on this side. These features are in proximity to one another and are, respectively, a large mahogany tree, and about 20 metres away to the north-west, a low mound of grass- and weed-covered brick rubble about 1.5 metres high, beneath which the corners of a building measuring, minimally, 8.20 X 5.30 metres can just be discerned. The bricks used here are in the Egyptian style, being flatter and wider than European-style bricks. According to local Figure 26.3. Main enclosure ditch on the north side of Pendit zariba, Rumbek, looking informants interviewed at the site in early east. Photograph by Paul Lane, February February 2007, the brick building was the 2007. main office used by the Arab who controlled the site, and the nearby mahogany tree is named Ater Madat. Ater is the Dinka word for mahogany. Madat is the name of a Dinka man who was one of those captured and brought to Pendit.84 Traces of at least six wattle-and-daub structures are visible in the interior of the enclosure, one of which, near the west side, appears to have been rectangular in (p.529) plan. The shape of

Page 12 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century the other structures could not be discerned from the surviving surface remains, however. There is also a low-density background scatter of artefacts across much of the interior, comprising mostly local pottery sherds, some of which were decorated with a variety of cord-roulette motifs (at least five different types were noted, and also one possible example of the use of a carved roulette). In addition, several denser artefact concentrations were observed, especially around the causeway entrances on both sides of the enclosure ditch. In addition to local pottery, these often included fragmentary animal bones (notably cattle), glass beads and the occasional fragment of imported glazed-ware ceramics or bottle glass. Other surface finds included part of an iron knife, a stone musket ball and a possible musket flint.

Meen Atol Meen Atol, located about 5.4 kilometres south-south-west (bearing 201°) of Rumbek at N 6° 46‘ 24" / E 29° 39‘ 49", is another zariba with local oral traditions concerning its use as an ‘Arab’ camp where Dinka were brought to be ‘slaughtered’, in a manner similar to those recounted for Pendit.85 The site is markedly different from Pendit, however. Instead of being subcircular in plan, Meen Atol is an almost square enclosure formed by a low murram (gravelly, ironrich subsoil) bank, ranging between about 0.30 metres to 1.7 metres high (Figures 26.2c, 26.4). Where well preserved, this bank is approximately 4 metres wide, and there are several large doum palms growing out of its centre, suggesting that the site was abandoned some time ago. Two entrances were also noted, both on the east side. The more southerly of these is approximately 2 metres wide, while the northern one measured about 4 metres. The bank is surrounded by an outer ditch, which although currently rather hard to discern, is perhaps 3 to 4 metres wide and in places about 1 metre deep. When visited, the site was densely vegetated by a mix of doum palms, long grass and scrubby woodland, and it is evident that cultivation has occurred both within the enclosure and over the line of the bank and ditch, especially on the southern and south-western sides, where these features are now particularly indistinct. The vegetation cover restricted surface visibility and the only material noted was restricted to a lowdensity scatter of cord-rouletted local pottery in an area in the north-eastern part of the site that had been cultivated recently. No glass beads, imported ceramics or glass fragments were found, and neither were any traces of former structures seen, although this does not rule out the possibility that such material may occur on the site. Because of the vegetation cover, only the line of the bank was mapped by GPS. This suggests that the internal area of the enclosure is around 34,043 m2 and the enclosure sides are approximately 173 metres by 168 metres. (p.530)

Page 13 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century Lol Nhom The site at Lol Nhom (also known as Kalamei by Jur speakers) is similar in form and appearance to Meen Atol, although it is considerably larger and the bank and ditch are better preserved—and were possibly more substantial originally—compared with those at Meen Atol. The site is located at N 6° 28‘ 48" / E 29° 47‘ 16", about 35.3 kilometres south-south-east of Meen Atol (bearing 137°), and about 39.3 kilometres south-south-east (163°) from Pendit. It comprises a 1.6 to 2 metre high inner bank, approximately 4 metres wide at its base, Figure 26.4. Enclosure bank on the east side surrounded by an approximately 6 metre of Meen Atol zariba, from the enclosure interior. Photograph by Paul Lane, February wide ditch roughly 2 metres deep, although 2007. some sections are shallower and others are closer to 3 metres in depth (Figures 26.2d, 26.5). The enclosure has a slightly rhomboidal plan, the west side at 280 metres being about 65 metres longer than the eastern side. The total area enclosed by the bank is around 53,421 m2. There are three entrances, one in each of the northern, western and eastern banks, the widest being around 4 metres. As at Meen Atol, large mature trees, in this case mostly mahogany, were observed in several places growing out of the centre of the ditch and bank, suggesting that the enclosure was established some time ago and that once abandoned it was not intensively re-utilised. (p.531) At the time of the survey, much of the ground vegetation had been recently burned off, thereby improving surface visibility, although the ash deposits were not always conducive to observing surface artefacts. As at Meen Atol, there were very few artefacts on the surface. The few pieces of pottery that were seen resembled material noted at Pendit and Meen Atol (and also local contemporary ceramics). A single piece of bottle glass of possible nineteenth-century date was also noted. The interior of the site is relatively featureless, apart from a few low mounds that could be the remains of Figure 26.5. Southern corner of the main (p.532) collapsed mud-and-wattle houses, enclosure ditch, Lol Nhom zariba, looking although no obvious lumps of house daub north-west. Photograph by Paul Lane, February 2007. were seen. Since the remains of a couple of recently abandoned houses, and also a well, survive within the interior, it is possible that even if these mounds represent collapsed houses they may transpire to be of recent origin. One rather larger mound was also noted just outside the enclosure ditch, near the north-west corner. Page 14 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century This was said by local informants to be where the ‘Arabs’ lived,86 but there was little in the way of surface evidence except a few sherds of local pottery to distinguish it from the mounds seen in the interior. An alternative explanation, if the mound can be shown to be of human origin, is that it represents the remains of a rubbish dump, as might be inferred from Berlioux’s comment that ‘the dungheaps around the Séribes send forth filthy odours.’87 Finally, apart from its impressive size, another feature of Lol Nhom which distinguishes it from both Meen Atol and Pendit is its proximity to water, as the site lies within 150 metres of a perennial river, known locally as the Bahr Khawaja (Khawaja: White Person/European), and an area of reedy grassland that becomes inundated during the rainy season. Whether this influenced the choice of site for the enclosure is not yet known, however.

Matching the Historical Sources with Archaeological Evidence As already indicated, there are various primary documentary sources concerning different zaribas, in the vicinity of Rumbek, that include short descriptions of the conditions and inhabitants of the settlements at Rumbek and Adael/Ayak towards the latter part of the TurcoEgyptian era. There is also one map that shows the distribution of numerous zaribas, across much of what is now Bahr al-Ghazal and Western Equatoria Provinces compiled by Schweinfurth during his travels between 1868 and 1871 (Figure 26.6). Schweinfurth was hosted over this period by the Khartoum trader Ghattas whose network of zaribas extended from the Bahr alGhazal/Bahr al-Arab boundary to the south-western side of the Nile watershed, which also marked the approximate edge of the Azande kingdom (see Figure 26.6). Although Schweinfurth was based for much of his time at Ghattas’s principal zariba (marked simply as Seriba Ghattas), which was probably located to the north of the modern town of Tonj, Schweinfurth visited approximately forty zaribas and his map shows ten zaribas as being either part of, or formerly belonging to, Ghattas’s network. These included some ‘minor’ zaribas on the Rol,88 including de Malzac’s, Ronga and Adael. Adael is also marked on Petherick’s map of his journey through the area in 1855 on his way to Gondokoro. According to Schweinfurth, Adael was plundered and burned down by Agar Dinka in April 1870, and was not restored. However, as discussed above, Adael (if it was the same location) was clearly thriving again in 1880–1, when visited by Emin Pasha. De Malzac’s zariba is also shown on this map as being abandoned, although marked nearby is another zariba named Ronga (Figure 26.7a).89 ‘Ronga’ is the Dinka word for shea butter tree, and there is an area labelled thus on the 1976 revised version of the 1940 Sudan Survey 1:250,000 map of the area (Figure 26.7b). There is also an area around Rumbek known today as ‘Ronga’, although this is situated slightly to the north-west of the area so labelled on the Sudan Survey map. The modern location of Ronga was visited during the survey and found to have many shea butter trees. Investigations failed to locate any enclosure, however, and local informants were also adamant that they knew of none in the immediate vicinity90 (p.533)

Page 15 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century (p.534) (p.535) On the basis of these results, it would seem that the area in the vicinity of Rumbek known today as ‘Ronga’ does not correspond with general the location of the zariba named ‘Ronga’. This is not particularly surprising, since place names are by no means constant, especially where populations are relatively mobile. Schweinfurth encountered similar difficulties and observed that because of the high level of population mobility, settlements were ‘rarely…permanent beyond a period of at most ten years’ making it ‘very difficult to fix on the maps names and localities’.91 The problem of linking archaeological sites to historically documented names becomes even more acute when, as in this case, the place name is derived from the local vegetation. Specifically, there are many places where shea butter trees grow in abundance, and each would seem as likely to be named ‘Ronga’ as any other. Nevertheless, the documentary sources are not entirely consistent. Specifically, whereas Schweinfurth’s map implies that the zariba known as Ronga was distinct from the zariba associated with de Malzac, other sources (notably Franz Binder’s account of his arrival to take possession of de Malzac’s zaribas following the latter’s death), seem to imply that de Malzac’s first zariba in 1854–6 was at a place called Ronga which later became known as Rumbek.92 Matters are further complicated by the fact that the surviving zariba in the modern town of Rumbek is known only as Pendit. Additionally, when comparing the position of de Malzac’s, Ronga and Adael as shown on Schweinfurth’s map, with the locations of Pendit, Meen Atol and Lol Nhom as located in the 2007 survey, it is not clear whether the latter two sites correspond with Ronga (p.536) and Adael (see Figures 26.7a and b). To start with, Lol Nhom is some distance from the modern settlement of Karic. Adael,

Figure 26.6. Digitised version of Schweinfurth’s 1872 general map, showing the location of different zaribas, with those ‘owned’ by Ghattas highlighted. The location of Meshra er-Rek is also shown, and the shaded section reproduced as Figure 26.7.

Figure 26.7. Enlarged, digitised section of Schweinfurth’s 1872 map showing location of de Malzac’s, Ronga and Adael zaribas (a), left, compared with location of Pendit, Meen Atol and Lol Nhom zaribas as logged in 2007 (b), right.

Page 16 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century on the other hand, is said to have been situated close to Karic93 (although interviews here in 2007 failed to produce anyone who knew of a ditched enclosure nearby). Similarly, Meen Atol is on the wrong side of the Na‘am (Rol) river compared with Ronga. This said, since Schweinfurth appears not to have visited any of these zaribas, it is quite possible that his map is inaccurate— or, rather, more inaccurate than his understanding of the relative geography of those areas he did visit, since even in these areas he managed to get the position and orientation of a number of topographical features, including rivers, quite wrong. Moreover, an archival source cited by Gray94 implies that Ronga was about two miles south of the modern town of Rumbek.95 This could imply that the enclosure now known as Meen Atol may well have been once known as Ronga, but if so then on Schweinfurth’s map the position of Ronga relative to Ancient seriba Malzac’ (i.e. Rumbek) is wrong. At this stage, all that can be said with certainty is that the area round Rumbek contains three ditched enclosures, two of which conform in terms of their plan with the standard layout of a zariba (although both are larger than the figures that are quoted in the historical sources), while the site of Pendit is significantly smaller in area and of a quite different plan to that normally ascribed to a zariba (although this may be the result of changes to the original defences as part of the settlement biography of the site). That this site was the focus of more intensive trading activity is suggested by the higher proportion of surface finds of a non-local nature, although this would need to be confirmed by excavation at all three sites. Finally, the absence of any material evidence for the practice of slavery or the presence of slaves should also be noted— although given the almost universal low archaeological visibility of slaves, perhaps we should not make too much of this.

Conclusion The Turco-Egyptian invasion of northern Sudan in 1820–1 and the subsequent expansion of trading and military power across Equatoria and Bahr al-Ghazal introduced a number of changes that had profound consequences for the subsequent history of an area that lay beyond the southern frontier of the Ottoman Empire. As Richard Gray observed over forty years ago, just how significant these changes were is a matter of interpretation and perspective.96 Gray’s comments notwithstanding, as discussed in this paper, one outcome of the Turco-Egyptian invasion was the establishment of a network (p.537) of fortified trading bases across a vast area of the lands occupied by Dinka, Jur, Nuer, Bongo and other southern Sudanese peoples. However, instead of developing the entrepreneurial activities of local communities, the traders instead exploited ‘tribal jealousies and divisions…and raids for cattle and slaves became the economic basis of these settlements, the cattle being used to purchase ivory and the slaves to pay for the soldiers’. In due course, this created a context in which violence was established ‘as a normal relationship’ between southern Sudanese and the outside world.97 As Gray also observed, the majority of records available for scrutiny concern the activities of foreigners, who included individuals of Arab, Egyptian, northern Sudanese, European and Ottoman origin. As demonstrated in this chapter, the archaeological traces of the places that acted as the focus of the activities of these outsiders still survive. It is also highly probable that many more zaribas and related centres, such as the remains of the river ports, could be located through more intensive archaeological survey, especially if conducted in conjunction with the available documentary and cartographic sources, and it is hoped that such surveys will be conducted in the near future. Excavation at some of these sites may also generate information about the range of economic activities and material culture traditions that were coeval with the utilisation of the different zaribas, as well as potentially providing greater detail concerning the layout and construction of these enclosures and the range of trade goods introduced into the region. Page 17 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century Ultimately, however, if the consequences of the Turkiyya for the southern Sudan are to be more fully understood, such surveys and excavations will need to be extended to the wider area, in tandem with the collection of clan and lineage histories, so as to locate and investigate those places lived in, attacked and abandoned by the local inhabitants of the region as they were drawn inexorably into contacts with the outside world that were often brutal and violent, but which also offered new opportunities that could be manipulated for political and economic advantage. Note. The archaeological research on which this chapter is based was funded by the British Institute in Eastern Africa, and conducted at the invitation of the Government of South Sudan. Particular thanks are due to Mr John Luk Jok, former Minister of Culture and the Governor of Lakes State, for facilitating research clearance. Thanks are also due to Mr Deng Mabor, Ms Harjyot Hayer and Ms Sophie Butler for assistance in the field; Dr Justin Willis, BIEA Director and his staff in Nairobi for help with logistics; John Ryle and Dr Stephanie Wynne-Jones for assistance with references; Philip Owiti for preparing the main location map; and Geoff Arnott for preparing Figures 26.2, 26.6 and 26.7. Notes: Proceedings of the British Academy 156, 509–537. © The British Academy 2009. (1) Caroline Finkel, Osman’s Dream: The Story of the Ottoman Empire, 1300–1923 (London: John Murray, 2005), 110. (2) On the operation of the timar system, see Metin Hepper, ‘Center and periphery in the Ottoman Empire: with special reference to the nineteenth century’, International Political Science Review, 1 (1980), 81–105. (3) This was despite the existence in the Mamluk sultanate of a somewhat similar system of cavalry-supporting assignments of usufruct, or iqta‘s: see Jane Hathaway, ‘The military household in Ottoman Egypt’, International Journal of Middle East Studies, 27 (1995), 39–52. (4) Ibid., 39. (5) Ibid., 44–7. See also Gabriel Piterberg, ‘The formation of an Ottoman Egyptian elite in the eighteenth century’, International Journal of Middle East Studies, 22 (1990), 275–89. (6) André Raymond, ‘Soldiers in trade: the case of Ottoman Cairo’, British Journal of Middle Eastern Studies, 18 (1991), 16–37. (7) Michael Winter, ‘Ottoman Egypt, 1525–1609’, in M. W. Daly (ed.), The Cambridge History of Egypt, II: Modern Egypt, from 1517 to the End of the Twentieth Century (Cambridge: Cambridge University Press, 1998), 7–17. (8) For a discussion of some of the reasons why this happened, see David Ayalon, ‘The end of the Mamlūk Sultanate: (Why did the Ottomans spare the Mamlūks of Egypt and wipe out the Syrian Mamlūks?)’, Studia Islamica, 65 (1987), 125–48. See also F. R. C. Bagley, ‘Egypt and the eastern Arab countries in the first three cen turies of the Ottoman period’, in H. J. Kissling (ed.), The Muslim World A Historical Survey, III: The Last Great Muslim Empires, trans. F. R. C. Bagley (Leiden: Brill, 1969), 50–96.

Page 18 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century (9) Daniel Crecelius, ‘Egypt in the eighteenth century’, in Daly (ed.), The Cambridge History of Egypt, II, 73–6. (10) For more detailed historical background see Afaf Lufti Al-Sayyid Marsot, Egypt in the Reign of Muhammad Ali (Cambridge: Cambridge University Press, 1984); Khaled Fahmy, ‘The era of Muhammad ‘Ali Pasha, 1805–1848’, in Daly (ed.), The Cambridge History of Egypt, II, 139–79. (11) P. M. Holt, ‘Egypt and the Nile Valley’, in J. E. Flint (ed.), The Cambridge History of Africa, V: c.1790 to c.1870 (Cambridge: Cambridge University Press, 1975), 13–50. (12) Bagley, ‘Egypt and the eastern Arab countries’, 58; cf. Chapter 11 by John Alexander in this volume. (13) Peter F. M. McLoughlin, ‘Economic development and the heritage of slavery in the Sudan Republic’, Africa, 32 (1962), 355–91; Rex S. O’Fahey and Jay L. Spaulding, Kingdoms of the Sudan (London: Methuen, 1974); Robert O. Collins, ‘The Nilotic slave trade: past and present’, Slavery and Abolition, 13 (1992) (Special issue: The Human Commodity: Perspectives on the Trans-Saharan Slave Trade), 140–61; John Alexander, ‘Islam, archaeology and slavery in Africa’, World Archaeology, 33 (2001), 44–60. (14) Gabriel Baer, ‘Slavery in nineteenth century Egypt’, Journal of African History, 8 (1967), 417. (15) Ibid., 419. (16) Ehud R. Toledano, Slavery and Abolition in the Ottoman Middle East (Seattle: University of Washington Press, 1998), 81. (17) Ehud R. Toledano, ‘The imperial eunuchs of Istanbul: from Africa to the heart of Islam’, Middle Eastern Studies, 20 (1984), 379–90. (18) Toledano, Slavery and Abolition, 56–7. (19) Daniel Pipes defines a military slave as ‘a person of slave origins who is acquired in a systematic way, trained for military service, and spends most of his life as a professional soldier’: Daniel Pipes, Slave Soldiers and Islam: The Genesis of a Military System (New Haven: Yale University Press, 1980), 5. (20) Douglas H. Johnson, ‘Sudanese military slavery from the 18th to the 20th century’, in Léonie J. Archer (ed.), Slavery and Other Forms of Unfree Labour (London: Routledge, 1988), 142–56. (21) Gabriel Warburg, ‘European travellers and administrators in the Sudan before and after the Mahdiyya’, Middle Eastern Studies, 41 (2005), 59. (22) Rex S. O’Fahey, ‘Slavery and the slave trade in Dar Fur’, Journal of African History, 14 (1973), 29–43. (23) M. Abir, ‘The origins of the Ethiopian-Egyptian border problem in the nineteenth century’, Journal of African History, 8 (1967), 443–61. (24) Andrew N. M. Mawson, ‘The Triumph of Life: Political Dispute and Religious Ceremonial among the Agar Dinka of the Southern Sudan’ (unpublished Ph.D. thesis, University of Cambridge, 1989), 70. Page 19 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century (25) Patricia Mercer, ‘Shilluk trade and politics from the mid-seventeenth century to 1861’, Journal of African History, 12 (1971), 407–26. (26) Robert O. Collins, ‘Slavery in the Sudan in history’, Slavery and Abolition, 20 (1999), 69–95. (27) Richard L. Hill, Egypt in the Sudan, 1820–1881 (London: Oxford University Press, 1959), 63. (28) Richard Gray, A History of the Southern Sudan, 1839–1889 (Oxford: Oxford University Press, 1961), 76–8. (29) Egyptian historians have until recently emphasised other motives than those discussed here, including a desire to wipe out the residual community of Mamluks who had fled to Dongola in 1811 after Mehmed Ali’s rise to power, and also the latter’s wish to restore order and civilisation to the region, following the collapse of the Funj kingdom. For a review of these arguments, see Gabriel R. Warburg, ‘The Turco-Egyptian Sudan: a recent historical controversy’, Die Welt des Islam, n.s. 31 (1991), 193–215. (30) Hill, Egypt in the Sudan, 62–4. See also Alice Moore-Harell, ‘Economic and political aspects of the slave trade in Ethiopia and the Sudan in the second half of the nineteenth century’, International Journal of African Historical Studies, 32 (1999), 410–13. (31) Abir, ‘Ethiopian-Egyptian border problem’, 444–6; see also Janet J. Ewald, ‘Crossers of the sea: slaves, freedmen and other migrants in the northwestern Indian Ocean, c.1750–1914’, The American Historical Review, 105 (2000), 69–91. (32) The first training camp was established in October 1820 at Esna for young Mamluks destined to become officers in the new army. The camp at Aswan established the following year appears to have been intended from the out set for training black Africans from further south: see Hill, Egypt in the Sudan, 24–8. See also Alice Moore-Harell, ‘The Turco-Egyptian army in Sudan on the eve of the Mahdiyya, 1877–80’, International Journal of Middle East Studies, 31 (1999), 19–37. (33) Baroness Harriet Van Steengracht-Capellan, second wife of the Dutch merchant Philip F. Tinne, and their daughter Alexandrine Petronella Francina Tinne; they were also accompanied by Alexandrine’s aunt, Theodor von Heuglin and Dr Hermann Steudner: see Penelope Gladstone, Travels of Alexine (London: John Murray, 1970). Reports on their expedition were published by Theodor von Heuglin, Die Tinnésche Expedition im westlichen Nilgebiet 1863–1864 (Gotha: J. Perthes, 1865), and Reise in das Gebiet des Weissen Nils (Leipzig: C. F. Winter, 1869). (34) Mr Tinne, ‘“A communication from Mr Tinne relative to the Dutch ladies’ expedition from Khartum up the River Bahr-el-Ghazal”, commencing February 26th at a point on the White Nile’, Proceedings of the Royal Geographical Society, 8 (1863–4), 12–18. (35) Moore-Harell, ‘Turco-Egyptian army in Sudan’, 23. (36) See Douglas H. Johnson, ‘The structure of a legacy: military slavery in northeast Africa’, Ethnohistory, 36 (1989), 72–88. (37) Hassan A. A. Ahmed, ‘The Turkish taxation system and its impact on agriculture in the Sudan’, Middle Eastern Studies, 16 (1980), 105–14.

Page 20 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century (38) See especially Jay L. Spaulding, ‘Slavery, land tenure and social class in the northern Turkish Sudan’, International Journal of African Historical Studies, 15 (1982), 4–7. (39) Ibid., 3. (40) Spaulding, ‘Slavery, land tenure and social class’, 5–13. (41) P. M. Holt, A Modern History of the Sudan: From the Funj Sultanate to the Present Day (London: Weidenfeld & Nicolson, 1961), 64. (42) Gray, History of the Southern Sudan, 47–9. (43) Georg Schweinfurth, The Heart of Africa: Three Years’ Travels and Adventures in the Unexplored Regions of Central Africa from 1868 to 1871, trans. Ellen E. Frewer (New York: Harper & Brothers, 1874), 1: 226. (44) Richard Hill, A Biographical Dictionary of the Sudan (London: Frank Cass, 1967), 229. (45) Collins, ‘Slavery in the Sudan’, 77. (46) Douglas H. Johnson, ‘Recruitment and entrapment in private slave armies: the structure of the zarä’ib in the Southern Sudan’, Slavery and Abolition, 13 (1992) (Special issue, The Human Commodity: Perspectives on the Trans-Saharan Slave Trade), 170. (47) Schweinfurth, Heart of Africa, 2: 195. (48) Karen Sacks, ‘Causality and chance on the Upper Nile’, American Ethnologist, 6 (1979), 443. (49) Ghattas was an Egyptian Copt and one of the great Khartoum traders. (50) Schweinfurth, Heart of Africa, 1: 173. (51) Ibid., 178. (52) Ibid., 174. (53) Etienne F. Berlioux, The Slave Trade in Africa in 1872, Principally Carried on for the Supply of Turkey, Egypt, Persia and Zanzibar (London: E. Marsh, 1872), 22–3. (54) Emin Pasha was the name taken by the German traveller and naturalist Dr Eduard Schnitzer, who joined General Gordon’s administration of Egypt’s equatorial provinces in 1876, having previously assisted the governor of Ottoman northern Albania. By the time he joined Gordon, he was already known as Emin Efendi, and was later promoted to the rank of pasha around 1884, by which time he was Governor of Equatorial Province: see Georg Schweitzer, Emin Pasha: His Life and Work (London: A. Constable & Co., 1898). (55) Schweinfurth, Heart of Africa, 1: 189. (56) John Petherick and Katherine Petherick, Travels in Central Africa, and Explorations of the Western Nile Tributaries (London: Tinsley Brothers 1869), 1: 224.

Page 21 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century (57) E. Schnitzer (Emin Pasha), Emin Pasha in Central Africa. Being a Collection of His Letters and Journals, ed. G. Schweinfurth, F. Ratzel, R. W. Felkin and G. Hartlaub (London: George Philip and Son, 1888), 336. (58) Baer, ‘Slavery in nineteenth century Egypt’, 420. (59) Ralph A. Austen, ‘The Mediterranean Islamic slave trade out of Africa: a tentative census’, Slavery and Abolition, 13 (1992) (Special issue, The Human Commodity: Perspectives on the Trans-Saharan Slave Trade), 214–48. (60) Collins, ‘Nilotic slave trade’, 141. (61) Collins, ‘Slavery in the Sudan’, 77. (62) Alvan S. Southworth, ‘The Soudan and the Valley of the White Nile’, Journal of the American Geographical Society of New York, 5 (1874), 99. Using this total of 130,000 slaves and an estimated value of $60 per head, Southworth calculated the annual value of the trade to be $7.8 million. It is not clear if he is referring to US or Maria Theresa dollars, however. (63) Frederick Millingen, ‘On the Negro slaves in Turkey’, Journal of the Royal Anthropological Institute, 8 (1870–1), lxxxvi. (64) ‘In one day I set free and sent home 165 Monbuttu, of whom 41 were taken from the hut of the chief of this district, a certain Mula Effendi, a Dongolaui, [and] 400 Agar, Kich and Atwot slaves; these numbers are proof sufficient of the state of matters’: Schnitzer, Emin Pasha, 336. (65) Southworth, ‘The Soudan’, 99, gives the following estimates for some of the individual Khartoum traders: Agate 5,000 plus; Cushik Ali (Küçük Ali) 4,000; Gatase (Ghattas) 4,000; Bizzelli 800. (66) Schweinfurth, Heart of Africa, 2: 427. (67) Johnson, ‘Recruitment and entrapment’, 167. (68) For a possible analogue of the spatial organisation of the territory around a major zariba, see Dennis D. Cordell, Dar Al-Kuti and the Last Years of the Trans-Saharan Slave Trade (Wisconsin: University of Wisconsin Press, 1985), 84, Map 5. (69) Romolo Gessi, Seven Years in the Soudan (London: Sampson Low, Marston & Co., 1892), 213. (70) Estimated at around 2,500: see Paul Santi and Richard Hill (trans. and ed.), The Europeans in the Sudan, c.1834–1878 (Oxford: Clarendon Press, 1980), 124. (71) Gaetano Casati, Ten Years in Equatoria and the Return with Emin Pasha (London: Frederick Warne & Co., 1891), 1: 67. (72) Or ‘Bekei’, as he is referred to by Dinka today. (73) Schnitzer, Emin Pasha, 334–6.

Page 22 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century (74) Franz Stuhlmann (ed.), Die Tagebücher von Dr. Emin Pascha (Hamburg: G. Westermann, 1919), 2: 491. (75) Casati, Ten Years in Equatoria, 76. (76) Douglas H. Johnson, Nuer Prophets: A History of Prophecy from the Upper Nile in the Nineteenth and Twentieth Centuries (Oxford: Clarendon Press, 1994), 249. Mawson suggests that the attack may also have been motivated by the murder by a group of traders of Kejang, the eldest son of Macot, the Agar Dinka spear-master at the time, ‘Triumph of Life’, 75. (77) Robert O. Collins, The Southern Sudan 1883–1898: A Struggle for Control (New Haven: Yale University Press, 1962), 44; Gray, History of the Southern Sudan, 158. (78) Schnitzer, Emin Pasha, 468. (79) Fashoda was established as a landing and anti-slavery station in Shilluk territory on the White Nile by the Egyptian army in 1855, and was visited by Schweinfurth in 1869 and Junker in 1876, by which time it was a significant trading station and jumping-off point for travellers going south into equatorial Africa, where several Greek merchants had established themselves. However, when Marchand arrived, the fort was deserted and in ruins. See Narrative of the Expedition to the Bahr el Ghazal and Subsequent Occupation by El Karim W. Boulnois Bey (Dakhlia 112/13/84), cited by Mawson, ‘Triumph of Life’, 76; Marc Michel, La Mission Marchand 1895–1899 (Paris: Editions Mouton, 1972); and David L. Lewis, The Race to Fashoda: European Colonialism and African Resistance in the Scramble for Africa (New York: Weidenfeld & Nicolson, 1987). (80) Mawson, ‘Triumph of Life’, 77–85. (81) Paul J. Lane (ed.), Protecting and Preserving the Cultural Heritage of South Sudan: Defining Priorities for Action. Unpublished report on a two-day workshop held in Nairobi, Kenya 12th-13th April 2005 (Nairobi: British Institute in Eastern Africa, 2005), 77 pages. On file at the BIEA Library, Nairobi. (82) This research was led by Paul Lane. (83) Schweinfurth, Heart of Africa, 1: 224. (84) Interviews Pendit 1–2/2/2007. (85) Interviews Ronga 12/2/2007, Meen Atol 13/2/2007. These modern accounts of Dinka being slaughtered need to be understood against the context of the recently ended civil war in the Sudan, and probably should not be taken too literally. (86) Interview Lol Nhom 15/2/2007. (87) Berlioux, Slave Trade in Africa, 23. (88) Schweinfurth, Heart of Africa, 1: 224. (89) Schweinfurth, Heart of Africa, 2: endpiece.

Page 23 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

The Archaeology and History of Slavery in South Sudan in the Nineteenth Century (90) The same informants, however, were able to tell us of the existence of the ditched enclosure at Meen Atol situated about 4 kilometres away from the area known today as Ronga and 5.4 kilometres from the site of Pendit. (91) Schweinfurth, Heart of Africa, 1: 194. (92) See Gray, History of the Southern Sudan, 48: ‘…on de Malzac’s death in April 1860, his stations were bought and visited by Franz Binder…When after a week Binder reached the settlement at “Ronga” (Rumbek) he was greeted by many chieftains…A few days later at “Ronga” four chiefs of the Cic, Agar and “Agjel” Dinka tribes came…’. (93) Mawson, ‘Triumph of Life’, 71–2. (94) Gray, History of the Southern Sudan, 47, citing a letter from von Heughlin to Petermann, July 1862. (95) See also Mawson, ‘Triumph of Life’, 70. (96) Richard Gray, ‘The Southern Sudan’, Journal of Contemporary History, 6 (1971), 109–10. (97) Ibid., 111.

Access brought to you by:

Page 24 of 24 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Afterword

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

Afterword CAROLINE FINKEL

DOI:10.5871/bacad/9780197264423.003.0027

Abstract and Keywords This chapter comments on the Ottoman frontier, historical archaeology, Ottoman archaeology, and suggests future developments in these studies. The history of the frontiers of the Ottoman world played out in significantly different ways at each point along their great distance. Geographical and climatic circumstance and human conditions conspired to produce uniqueness. Meanwhile, the fortuitous degree of overlap between archaeological and historical data at Anavarin encourages people to search for a better understanding of the matters dealt with here. Each of the projects documented is this volume is tied to a specific geographic location. This simple fact opens up opportunities for virtual representation of historical and archaeological findings using GIS (Geographical Information Systems) software. GIS provides a means of digitally storing and analysing large amounts of data relating to defined locations. Keywords:   Ottoman frontier, historical archaeology, Ottoman archaeology, GIS, Ottoman Empire, Anavarin

The Ottoman Frontier WHENEVER YOU PICK UP A GENERAL BOOK ON THE OTTOMAN EMPIRE which has only a single map, you can be sure that it will be the familiar image showing the sultan’s domains at their greatest territorial extent. In the absence of precise calculation of the empire’s area, there is vagueness as to whether this was achieved as a consequence of the sixteenth-century conquests of Süleyman I, or in the later seventeenth century following the conquest of Crete in 1669 and Podolia in 1672, but it is the former image that is inevitably preferred. The utilitarian function of enabling the maximum amount of geographical information to be shown in a single spread is undeniably compelling but, as we are reminded in this volume, maps are representations of an imagined reality. Depiction of the Suleymanic conquests invites the viewer to be impressed that a single power could command such far-flung domains, while blurb writers inevitably provide complementary, stirring lines to reinforce the notion that what is most salient about the empire was its maximum size—and that the history of the time when this was achieved is the history that counts.

Page 1 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Afterword The ubiquitous map of the empire at its greatest extent is a relic of the days when two paradigms framed Ottoman history. Both concern frontiers, and both remain relevant to the purposes of this volume. One was Paul Wittek’s so-called ‘gazi thesis’ (discussed in the introductory chapter), whereby the Ottoman state was seen as a frontier principality impelled by the imperative of ‘holy war’; the increasing area of Ottoman territory was taken as ‘proof of its ideological efficacy as a creed that gave Ottoman warriors and bureaucrats their staying power as well as sustaining the momentum that pushed back the boundaries of the realm. The other paradigm was the ‘decline thesis’, whereby all Ottoman history after Süleyman’s time, or thereabouts, was a tale of unravelling from the high tide of invincibility that was peculiarly ascribed to him. Paradoxically, then—since both the ‘greatest extent’ map and the ‘decline thesis’ are products of the same mindset—the map of the Suleymanic conquests allows us to suspend disbelief: because the empire’s frontiers shifted but little in the century after his death, and its area possibly expanded, the glow of the so-called ‘golden age’ could persist. (p.540) As much as the extent of the Ottoman Empire laid out in the standard map, the starkness of the line delimiting the sultan’s territories is intended to impress. Within the space bounded by this periphery he is imagined to rule with confidence and to effect, while in the uncharted beyond there appears to be a dramatic falling-away of his, or the empire’s, influence. This ‘all or nothing’ image of Ottoman frontiers not only suppresses any hint of the varying degrees of integration of lands and peoples into the empire at a formal level, but its very linearity disguises the facts on the ground, masking the reality that any frontier, however defined, was a place of infinite complication both for the governments that sought to assert their power there, or acknowledged it as the limit of their reach, and for the people whose lives were shaped by its very existence. Frontier studies in general has been a lively field of enquiry for some time, and how best to conceptualise the frontier phenomenon so as to take account of its historical and geographical particularities has been well aired. Insofar as Ottoman studies is concerned, the most salient volume for present purposes is the collection of papers edited by Kemal H. Karpat and Robert W. Zens that appeared in book form as Ottoman Borderlands: Issues, Personalities and Political Changes.1 Theoretical issues are lightly alluded to in Karpat’s framing introduction: he writes that the existence of a ‘borderland’ implies the co-existence of a ‘coreland’, and discerns four main categories of Ottoman borderland, depending on varying degrees of linkage with the central power.2 The papers in Ottoman Borderlands confound easy assumptions about the linearity—whence perhaps Karpat’s preference for ‘borderland’ rather than ‘frontier’—and uniformity of Ottoman frontiers, and the papers in the present volume similarly point up their complexity. Apart from demonstrating in detail how the line on the map that bounds Ottoman territory misrepresents the interaction between societies and the mixing of cultures that characterises what was in reality a zone of variable morphology rather than a narrow boundary strip, they emphasise the ebb and flow of the formal and informal limits of the empire. By doing this both in terms of geography and in the reach and penetration of the governing power, they suggest that the rubric ‘Ottoman world’ well describes the space across which the ripples of the state centred at Istanbul ordered people’s lives in greater or lesser degree. The convenient line on the map may be suggestive, but its reductionism deprives it of interpretive merit. *** It goes without saying that the history of the frontiers of the Ottoman world is played out in significantly different ways at each point along their great distance. Geographical and climatic

Page 2 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Afterword circumstances alone would conspire to produce uniqueness, but adding to the mix the even greater unpredictability of the human condition makes for a startling (p.541) range of possibilities. The human aspect might include political, commercial, societal, economic, cultural, ethnic or religious factors—all of them intertwined, and none amenable to mapping as neatly matched layers. The study of the frontiers of the Ottoman world is a subject that, par excellence, integrates history and archaeology, and geography—people and place. Ottoman historians, and those who rely on their output, for too long believed the myth propagated by the dynasty and its advisers that sultans were not only just, but all-mighty—that they had only to command and it was done. The evidence of our eyes told another story, but we were willingly deceived. Obvious measures of resistance, such as need for repetition of edicts, indicating that relatively minor matters in the provinces were difficult of resolution, or the many and violent rebellions that punctuate the Ottoman centuries, were somehow glossed over and explained away. Had the sultan and those who exercised power in his name truly been able to exert their authority to the furthermost extremities of their domains according to a format devised in the council chambers of Topkapı Palace, we would expect to find that—allowing for the vagaries of geography and climate which even the sultans had no claim to control—their imprint should be more or less similar on every frontier. Instead, as the chapters in this volume show, we are confronted with a historical legacy of remarkable diversity. The wildest card in determining the limits of the possible when a region newly became Ottoman was its pre-Ottoman history. No conquering power could erase this, but was bound to acknowledge what had existed hitherto, and seek to transform it with a veneer of administrative and cultural ‘Ottomanness’. As time went by this veneer effaced living memory of a pre-Ottoman regime, and was itself pragmatically adopted as the norm. Insofar as the history of each region that became Ottoman was different, the process by which each region became part of the empire varied. The Ottoman ruling class understood the limitations on their power—and how these limitations were aggravated by distance—and that their imprint was resisted by the inertia of new subjects who had become accustomed to living their lives according to familiar certainties. The new rulers did not seek to impose what Karpat refers to as ‘an ideological blueprint’3 on newly conquered territory, but beyond the minimum requirement that the sultan henceforth be recognised as ultimate sovereign, and that the territory in question was now considered part of his domains, were forced to adapt their notional template to take account of prevailing local conditions and give a semblance of continuity despite the palpable rupture with the past. Discussion of the fringes of the Ottoman world reminds us of the extent to which the history of the sultan’s subjects was shared with numerous and diverse populations on three continents. The histories of the modern states whose precursors lay on the outer rim of the empire, as well as of the empire’s some thirty successor states, are (p.542) inescapably entwined. Widespread refusal to acknowledge Ottoman history as an integral layer of local and national history today is part and parcel of attempts to forge a past that will provide a sound basis for cohesion in the present. It is a denial that militates against self-knowledge. Academic study of the frontier, by contrast, allows recovery of the history of millions of people who may barely be aware that the lives of their ancestors were affected by the demands imposed by and the opportunities presented by life influenced by the dynasty that ruled from far-off Istanbul. The empire was an unavoidable fact, and what is commonly regarded as a detested and alien hegemon shaped the

Page 3 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Afterword past and affected the present every bit as much as did the indigenous cultures that were subsumed within its orbit. The ‘gazi thesis’ reminds us that Ottoman territory expanded from its meagre beginnings as a small north-west Anatolian principality with few apparent prospects. All parts of Ottoman territory were at some time frontiers of the expanding state—for longer or shorter periods—and were so again as, by lurches, it contracted. The ‘greatest extent’ map presents us with a fait accompli, and conceals this ever-unfolding process. Study of the frontiers of the Ottoman world is limited only by the extent to which those who inhabited or inhabit space that was once part of or contiguous to the empire can be considered to have or have had a common past with those we describe loosely as the Ottomans.

History and Archaeology—and Historical Archaeology Historians of all persuasions are apt to forget that the human activity they study takes place in space as well as in time. Indication of this is the dearth of maps in the books we write—although this may partly be accounted for by the parsimony of publishers. One of the main purposes of the London workshop was to bring together historians and archaeologists, to initiate a conversation that the division between the two disciplines typically precludes. Such a deliberate face-to-face encounter was perhaps a first in Ottoman studies and for those of us who identify ourselves as historians of the Ottoman world the results were salutary—we became aware of the work of the many colleagues who concern themselves with investigating similar topics to ourselves, but who adopt markedly different approaches. Even the empire seemed to have expanded, as we listened to the presentations of archaeologists working in north-east Africa and Arabia and, closer to home, and less forgivable an ignorance, of others engaged in projects in Greece and former Yugoslavia. And that was not the only matter for reflection: historians tend to spend their time in libraries and archives but archaeology cannot solely be conducted indoors, without ever visiting the site. Unlike historians, archaeologists are entirely familiar with the locations where the events they seek to understand took place, and they get their hands dirty through physical rather than virtual contact with the artefacts of the past. An archaeological approach of necessity (p.543) imposes a locational dimension on the study of frontiers, and one which historians, narrowly defined, neglect at their peril. The workshop demonstrated the importance of the geographic dimension of history to our research. The contributors to this volume are archaeologists and historians in almost equal number, and include others who might identify themselves as geographers or anthropologists. Different disciplines cluster in certain sections. In the section on fortifications, half of the eight chapters combine history and archaeology, while this is the case with only one of the six on administration. History and archaeology contribute in equal measure to the sections on society and on economy. The predominance of one discipline over another in individual chapters reflects the divide between specialisations that has been usual, or the interaction between disciplines that complement each other unequally in piecing together the story of the Ottoman frontier—at the present stage of our research it appears that some aspects, such as administration, are more likely to be revealed by a paper trail rather than by an assemblage of artefacts, but this preponderance of the written may shift as our interpretive skills develop. That any chapter combines disciplines presupposes the existence of a project bringing together the relevant specialists—unlike history, archaeology is never a solitary activity—and assumes an existing body of knowledge in one discipline that lays the basis for exchange with and enhancement by another. It might seem that it is easier for archaeologists than historians to contextualise their work by recourse to secondary sources on the historical aspects of the place Page 4 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Afterword where they work, than it is for historians to integrate archaeological findings—which may be highly technical, or tardily published—into their narrative. Archaeologists are more used to teamwork than are historians in general and Ottoman historians in particular, and the latter have much to learn in this regard. However, given their very different training, archaeologists have sometimes been insufficiently aware of the complexity of historical process. As this volume shows, much basic study of the Ottoman frontiers remains to be done in building up a reliable historical backdrop against which to interpret archaeological finds. But does the mere fact of partnership between archaeologists and historians qualify as historical archaeology? Much ink has been spilled in the attempt to find a universally applicable definition of this field, and in 1997 the editor of the then newly launched International Journal of Historical Archaeology (IJHA) eschewed defining the term because it had such varied meanings around the world.4 He instanced the USA, where, he writes, historical archaeology denotes archaeology of the period after Columbus, while elsewhere it signifies ‘archaeology that focuses on a literate period’ (sic), or ‘archaeology supplemented with historical writings’. In essence, ‘no sense of the term’ would be proscribed in contributions to the journal. Although he went on to set up an opposition between prehistoric archaeology and historical archaeology which suggests (p.544) an inching towards a tighter definition, we might be advised to embrace the ‘terminological chaos’5 that other writers see as attending the notion of historical archaeology. We should view the confusion as an opportunity that allows us to conduct research without worrying too much about the rigidity of its conceptual underpinnings. The question of the relationship between archaeology and history has been raised in this volume in an original manner. This concerns the question whether we should indeed be content—in situations where any nugget of data is a chance survival, and is to be prized and interrogated for the advance in our understanding it may yield—that specialists in these two disciplines work side by side, adding to the sum of knowledge by complementing each other’s work and adding discrete pieces to the jigsaw. Or should we be more ambitious, and strive for overlap between our disciplines, so that we gain a more complex picture of any aspect of our sites? Can what we do only then be considered ‘historical archaeology’? A limiting factor in achieving the denser picture that overlap provides is the serendipity of survival, and perhaps, here too, we cannot afford to agonise about attaining the ideal but simply be thankful when the findings of archaeology and history separately lend themselves to building up a picture of how life was lived on Ottoman frontiers.

Ottoman Archaeology Juxtaposing the words ‘Ottoman’ and ‘archaeology’ may cause no less confusion than juxtaposing ‘historical’ and ‘archaeology’. If the sweepingly named A Historical Archaeology of the Ottoman Empire: Breaking New Ground fails to deliver as much as it promises, its introductory essay is a heartfelt polemic in favour of archaeology of the Ottoman world.6 The preferred definition of historical archaeology of Uzi Baram and Lynda Carroll, editors of A Historical Archaeology, is somewhat imprecise: they opt for ‘historical archaeology as it is practiced in North America’, meaning (the IJHA definition notwithstanding) ‘the archaeological study of the recent past’, where ‘the explorations and conquests of Western Europeans starting in the 1400s’ sets the terminus for the ‘recent past’. Ottoman archaeology, on the other hand, they define simply as ‘an archaeology of the Ottoman period’, and see the discipline as having the potential to contribute to a number of debates. Ottoman archaeology, they write, ‘can (p. 545) be envisioned as borrowing and expanding on: global historical archaeology; the archaeology of Islam; Middle Eastern Studies; ethnoarchaeology; critical analyses of the

Page 5 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Afterword present’, and they go on to expound on these possibilities.7 Whether or not we choose to follow Baram and Carroll’s conceptual deliberations, a loose definition of Ottoman archaeology, such as they propose, may likewise serve us best. Baram’s fieldwork in Ottoman Palestine gives him an intimate perspective on the history of that region, and a first-hand awareness of how the disparagement of traces of the Ottoman past is worked out in a very specific geographic and ideological context. Modern Israel is only one of a number of instances of states whose ‘present-day ethnic identities and geopolitical borders have radiated from the distant past’—lending themselves to characterisation as what Baram and Carroll call ‘primordial’8—but is among the most salient for its disregard of the Ottoman layers in its archaeology. Rewriting people’s cherished view of their history meets with violent reaction: the relative newness of Ottoman archaeology as a discipline that dares to speak its name—even in Turkey, where Kemalist rhetoric aimed to discredit the Ottoman past—is testimony to this. In Ukraine, for instance, it is only now that the ring of impressive fortresses that once stood sentinel on the northern Black Sea littoral is being acknowledged as having spent the longest phase of their existence under Ottoman control—whether or not the Ottomans were the first to fortify the site in question—and Ottoman archaeology can at last proceed, albeit falteringly The extramuros Ottoman graveyard at Akkerman, for instance, is said to have been stripped away apparently in order to reveal the more acceptable past exemplified by the remains of the sixthcentury bc Greek colony of Tyras. As Baram and Carroll tellingly opine, ‘perusing an archaeology of a glorious, ancient past [is] a more desirable endeavor than developing the archaeology of a despised one.’9 If, as one reviewer of A Historical Archaeology of the Ottoman Empire wrote, ‘the aims of the editors do not principally concern physical remains from the time or territories of the Ottoman Empire, their recovery or preservation…’, this is not the case with two exemplary collective works on the frontiers of the Ottoman world, Archaeology of the Ottoman Period in Hungary,10 and A Historical and Economic Geography of Ottoman Greece.11 These two works offer models for collaborative work between archaeologists and historians studying varied aspects of the frontiers of the Ottoman world. (p.546) Hungarian scholars have perhaps made the greatest progress in their efforts to understand the cultural mixing characteristic of the once-Ottoman lands they inhabit and, unlike the present-day dwellers in most other like territories have, perhaps uniquely, laid aside their animus. This enables them to explore with great insight the historical archaeology of the period when their forebears dwelt in an Ottoman world. In Archaeology of the Ottoman Period we find, inter alia, essays grouped in sections entitled: Military Structures—Fortresses and Castles; Ecclesiastical and Civilian Architecture in the Towns; Abandoned Towns; Cemeteries; Material Culture. As in the present volume the contributors rely upon archaeological and architectural investigation as well as documentary and archival records in a variety of languages. It is gratifying that we can read the results of their efforts when those with no knowledge of Hungarian are often barred from access to the instructive writings of our Hungarian colleagues. A very different style is evinced in A Historical and Economic Geography of Ottoman Greece. Instead of investigating a multiplicity of sites on an Ottoman frontier over a period of some one hundred and fifty years as the Hungarian volume does, this study revolves around a single, albeit voluminous, Ottoman document that encapsulates the situation on a very limited stretch of the frontier at a particular moment in time. This is the first cadastral survey of the kaza of Anavarin (Navarino) following the expulsion of the Venetians, who had won the Peloponnese

Page 6 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Afterword from the Ottomans in 1685 but lost it in 1715. Although, as the authors write, the study ‘represents the fruits of a partnership between an Ottomanist [Fariba Zarinebaf]…, and two archaeologists [John Bennet and Jack L. Davis] who are both engaged in regional studies in Greece’,12 archaeology takes a back seat at this stage. Decipherment and translation of the cadastral survey reveals a map of the settlement and land use of the Anavarin district in 1715 which can realistically be expected to have been valid for longer than that year alone. It yields a wealth of toponyms, most of which have been assigned to a modern location, detailed information on the immovable property, produce and revenue of the named individuals of the area, and more. Properties are described in relation to adjacent properties, which enables a map of their relative locations to be drawn. The Anavarin project grew out of a greater, ongoing endeavour known as the Pylos Regional Archaeological Project (PRAP), which aims to collect archaeological evidence of all periods by means of intensive surface survey, and interpret it with the benefit of written records. The decision to thoroughly research the Ottoman cadastral survey gave an unanticipated direction to part of the PRAP and focused attention on the potential that such documents have for illuminating the past. It has been a productive detour because, as the authors write, ‘our close attention to the topography and toponymy…has permitted us to set our conclusions on a firm foundation…Such exhaustive… analysis has laid the groundwork (p.547) for future archaeological fieldwork, not only by identifying locations for excavation, but also by providing a cartography and geography of Ottoman Anavarin, with which the material culture may be integrated. It has also suggested how archaeological evidence can be employed to improve our understanding of the text itself…’.13 Even allowing that the intentions of every project, and the opportunities available to each, are different, and the aims of the PRAP and its Anavarin component may not be those of other researchers, it is abundantly clear that archaeologists of the frontiers of the Ottoman world need historians as much as historians need to work symbiotically with archaeologists. The present volume is a timely demonstration of this, as will be the future, fuller publication of the varied projects discussed herein. Moreover, the fortuitous degree of overlap between archaeological and historical data at Anavarin must encourage us to search for as nuanced an understanding of the matters of our concern as we can achieve.

A Way Forward The rich variety of subject matter and approaches in the presentations at the London workshop and between these covers suggests an equally varied set of possibilities for future research. As great a challenge to both archaeologists and historians as pursuing the trajectories laid out here is devising a framework that utilises the expertise and resources of all—most of? many of?—the micro-studies represented, an undertaking that somehow encompasses the complex histories of the frontiers of the Ottoman world. The problem is one of combining a number of smaller investigations into a larger project that has purpose. An achievable first step is that existing projects ensure that they bring together the full range of sources available to them, combining the archaeological with the historical as thoroughly as possible—some are already well along this road. Of all the directions we might take, one of the most valuable would be to consolidate the results of our research in an accessible format that will both provide a sound basis for future investigation of Ottoman frontiers and also, because bringing together disparate research reveals previously unobserved connections, inspire new thinking. As noted, each of the projects documented here is tied to a specific geographic location—whether a point or an area. This simple fact opens up opportunities for virtual representation of historical and archaeological Page 7 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Afterword findings using GIS (Geographical Information Systems) software. GIS provides a means of digitally storing and analysing large amounts of data relating to defined locations. Series of data sets (p.548) relating to a single location or vicinity at different times can be combined to represent change at that location over time. Or data sets relating to different locations can be juxtaposed to represent the varying conditions/parameters pertaining in these locations at the same time. Most ambitious of all, multiple data sets from a variety of places at a variety of times can be combined to give a rich and dynamic rendering of historical processes over a wide and diverse geography. Process and change can thus be explored spatially and temporally14 The China Historical GIS project is a possible model,15 but there are a surprising number of such projects under way, although most are not so well funded.16 In recent years the value of GIS for historical and archaeological analysis has become widely recognised, with university departments, conferences and websites dedicating resources to exploring the possibilities offered by this exciting tool. In the case of the frontiers of the Ottoman world, we might create an interactive map with our sites marked on it, and linked data sets, once compiled, to be accessed by means of a mouse click on any individual site. Thus we might, for instance, compare the number of Ottoman smoking pipe bowls, or Iznik pottery shards, found in widely spaced excavations in order to stimulate new insight into the reasons behind their distribution. We might tabulate the origins of captives taken by the corsairs of the western Mediterranean as against those brought by Tatar raiders to ports on the northern Black Sea, or the religious affiliation of plaintiffs in the kadı courts of Vidin by comparison with those of another frontier garrison city. However, given the disparate nature of the research discussed here, its multifarious aspects and range of possible interpretations, reducing our data to meaningful tabular format may be considered premature. Yet, despite and with awareness of the limitations and pitfalls of GIS, other historians and (p.549) archaeologists are embracing its possibilities and there is no reason why those studying the Ottoman world should be left on the sidelines. A realistic use of GIS at the present time is the possibility of linking to the Ottoman frontiers map an archive for each site. This might include photographs of the site or of artefacts found there, maps, narrative texts and other documents—all of which could be scanned into the GIS software program and accessed with a click. We all have documents and photographs and maps of the sites where we work that might be digitised to create a readily accessible archive on the Ottoman frontier. Each project could supply various sorts of information that once uploaded would provide a range of material that would allow ourselves and future researchers to ask questions which may not even have occurred to us as we work in our isolation. These could be linked to their location on the map as a first step in what has been characterised as ‘the steady and meticulous development of geographically referenced data resources that can be utilized for a variety of purposes over a period of years’.17 Without embarking upon the complicated, expensive and time-consuming task of creating numeric data sets, GIS technology can provide an aid to visualising the features of sites otherwise visited only by those researching there and those perusing coffee table books and the random possibilities for virtual travel afforded by the web. A basic text for historians and archaeologists who use GIS is Past Time, Past Place. This is a collection of papers demonstrating the use of the tool to elucidate a variety of historical and archaeological themes, most of which relate to American history. A further collection by the same editor, Placing History: How GIS is Changing Historical Scholarship, gives a sense of how

Page 8 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Afterword rapidly GIS techniques from the most basic to the most sophisticated are spreading in the fields of research that engage us.18 GIS allows familiar phenomena to be analysed in creative ways, and encourages the researcher to frame new questions as to why things were as they were. It is a tool that helps us challenge long-cherished interpretations—that, at least, is the result in projects where it has been utilised, and there is no reason to think that matters will be any different in regard to the history and archaeology of the frontiers of the Ottoman world. GIS allows a more comprehensive approach to research material, and can much better portray complexity, because it removes the need to exclude essential information to fit a printed format, which is by its very nature finite. The infinitely flexible nature of GIS obviates the need for ‘final’ conclusions to be drawn within a limited time period—new data can be brought to bear as they become available. GIS space and time is limited only by the availability of data. If potential funders enjoin researchers to think in new ways, and to synergise, and to make research relevant to modern concerns, and, in particular, to make results and (p.550) insights available to a wide audience, GIS offers a way of doing this that is becoming familiar in projects in histories and archaeologies other than Ottoman. Within Ottoman studies our present understanding of the frontiers of the Ottoman world is still too simplistic, and we urgently need a more sophisticated way of representing the wealth of research between these and similar covers. The papers here can contribute to the creation of a visual resource that permits of critical interpretation: they allow us to hope that we can consign the standard maps of the frontiers of the Ottoman world to the past, and create for ourselves and others an array of new maps that more nearly reflect how things really were. Notes: Proceedings of the British Academy 156, 539–550. © The British Academy 2009. (1) Kemal H. Karpat with Robert W. Zens (eds.), Ottoman Borderlands: Issues, Personalities and Political Changes (Madison: University of Wisconsin Press, 2003). (2) Kemal Karpat, ‘Comments on contributions and the borderlands’, in ibid., 1–2. (3) Ibid., 4. (4) Charles E. Orser, ‘Introductory Statement’, International Journal of Historical Archaeology, 1 (1997), 1. (5) Jószef Laszlovszky and Judith Rasson, ‘Post-medieval or historical archaeology: terminology and discourses in the archaeology of the Ottoman period’, in Ibolya Gerelyes and Gyöngi Kovács (eds.), Archaeology of the Ottoman Period in Hungary (Budapest: Hungarian National Museum, 2003), 377. (6) Uzi Baram and Lynda Carroll (eds.), A Historical Archaeology of the Ottoman Empire: Breaking New Ground (New York: Kluwer Academic/Plenum Publishers, 2000), 1–36. We are well aware that, as with meanness about maps, publishers have their own ideas about titles and will distort an author’s draft title that accurately describes a book’s contents in favour of an allembracing one intended to attract the widest possible readership—we give the editors of this important volume the benefit of the doubt!

Page 9 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Afterword (7) Ibid., 15–26. (8) Ibid., 8. (9) Ibid., 6. (10) Gerelyes and Kovács (eds.), Archaeology of the Ottoman Period in Hungary. (11) Fariba Zarinebaf, John Bennet and Jack L. Davis (eds.), A Historical and Economic Geography of Ottoman Greece: The Southwestern Morea in the 18th Century (Athens: American School of Classical Studies at Athens, 2005). (12) Zarinebaf, Bennet and Davis (eds.), A Historical and Economic Geography of Ottoman Greece, 1. (13) Ibid., 212. See also the companion volume to A Historical and Economic Geography, which came to my attention too late to be included in my discussion: Siriol Davies and Jack L. Davis (eds.), Between Venice and Istanbul: Colonial Landscapes in Early Modern Greece (Athens: American School of Classical Studies at Athens, 2007). (14) There is a growing literature on historical (and archaeological) GIS, and an increasing number of historical GIS projects accessible on the web. An invaluable beginner’s manual by the leading British specialist is Ian N. Gregory, A Place in History: A Guide to using GIS in Historical Research (2nd edition), www.ccsr.ac.uk/methods/publications/ ig-gis.pdf. Gregory’s Historical Geographical Information Systems Research Network (www.hgis.org.uk) is a con spectus of information on historical GIS, including an extensive bibliography. Within Ottoman studies, panels on GIS in Ottoman history were part of the XIth International Congress of the Social and Economic History of Turkey held at Bilkent University, Ankara on 17–22 June 2008 and also at the Middle Eastern Studies Association meeting held at Washington DC on 22–25 November 2008. A first published work is the collection edited by Okabe Atsuyuki, Islamic Area Studies with Geographical Information Systems (London: RoutledgeCurzon, 2004), which contains several papers of interest to Ottomanists, among them, for example: Egawa Hikari, ‘The use of GIS to locate abandoned villages listed in the Temettuat registers of Balikesir district, Anatolia’, 122–38; Yamashita Kimiyo, ‘The water supplies and public fountains of Ottoman Istanbul’, 162– 83. Georgios C. Liakopolous, A Study of the Early Ottoman Peloponnese in the Light of an Annotated editio princeps of the TT10–1/14662 Ottoman Taxation Cadastre (ca. 1460– 1463)’ (Unpublished Ph.D. thesis, University of London, 2008) concerns the geographic, economic and demographic aspects of the Peloponnese in the first years of the Ottoman conquest, and contains fifty-three digital maps using GIS. (15) www.fas.harvard.edu/~chgis/ (16) See Historical Geography, 33 (2005), ed. A. K. Knowles, for reports on national historical GIS projects in, inter alia, the USA, the UK, Ireland, Belgium, the Netherlands, Germany, Russia, China, Korea. Note also the overar ching Electronic Cultural Atlas Initiative (ECAI) discussed in the same place (www.ecai.org/). (17) Gregory, A Place in History, 16.

Page 10 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Afterword (18) A. K. Knowles (ed.), Past Time, Past Place: GIS for History (Redlands, CA: ESRI Press, 2002); idem (ed.), Placing History: How GIS is Changing Historical Scholarship (Redlands, CA: ESRI Press, 2008).

Access brought to you by:

Page 11 of 11 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Glossary

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

(p.551) Glossary Note: Many terms were used with different meanings in different periods. Only definitions relevant to this volume are included below. ağa lord; military commander akçe a silver coin (asper) akıncı irregular cavalry amir al-hajj official responsible for the hajj ardebb a unit of weight asker soldier; army azeb a type of troops, stationed in frontier garrisons başıbozuk irregular soldier bayt (Ar.) house bey lord; commander, governor beylerbeyi governor of a province; commander beylerbeyilik a province beylik principality bina emini official responsible for construction works cebelü (or cebeli) auxiliary soldier paid by timar or vakıf çiftlik an agricultural holding cizye (Ar. jizya) poll tax on non-Muslims defter register defterdar chief financial official; financial scribe derbendat başbuğu commander of local security forces (Balkans) Page 1 of 4 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Glossary derbendci guard of mountain passes devşirme the recruitment of Janissaries dey rulers of the North African coast, subject to the Porte dihliz space between house and street divan (Ar. diwan) (a) the Ottoman Imperial council; (b) public reception room; (c) collection of poetry dizdar garrison commander emin (Ar. amin) tax collection agent eyalet a province faki a holy man (from classical Ar. faqih and faqir; Sudan) farisan cavalry ferman decree (p.552) gaza (Ar. ghaza’) holy war, jihad gazi (Ar. ghazi) holy warrrior gönüllü (pl. gönüllüyan) volunteer soldier guruş a large silver coin hadd (pl. hudud) frontier hamam bath, bathhouse, toilet han (Ar. khan) a caravanserai harem women’s quarters; private domain hass (Ar. khass) private property of the sultan haydud (Slavic hajduk) bandit hüccet (Ar. hujja) kadı ’s certificate iltizam tax farm ispence poll tax on non-Muslims jallaba (Ar.) a slave trader kadı (Ar. qadi) religious judge kadıasker military judge kadı sicil records of the kadı ’s court kale castle, fort kantar a unit of weight kapudan pasą commander-in-chief of the Ottoman navy

Page 2 of 4 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Glossary kaşif governor of a district (Egypt and Sudan) kaymakam governor of a kaza kaza subdistrict of a sancak keşf appraisal (of a building before repair) keşf defteri appraisal record kila, kile a unit of weight külliye complex of buidings (e.g. markets, library, medreses) around a mosque liwan outer porch mamluk a slave soldier, after which the Mamluk dynasty takes its name martolos a Christian salaried Ottoman soldier (Balkans) mastaba bench medrese (Ar. madrasa) a school, especially religious miri lands belonging to the sultan or their revenue mirimiran provincial governor, equivalent to beylerbeyi mujtahid a Shi‘ite jurisprudent mukataa tax-farm müselleman auxiliary forces that gave logistical support rather than being engaged in fighting mustahfiz fortress guard nahiye a subdivision of a kaza naib deputy of a kadı naqib al-ashraf (Ar.) chief of the descendants of Ali b. Abi Talib, the leading figure in the Shi‘ite commmunity (Iraq) nazır overseer nezaret administration, ministry (p.553) ocak an inherited timar; a unit of recruitment in Ottoman military administration (e.g. the Janissaries); in general, a word used for Turkish soldiers in the Maghrib and Egypt ocaklık a hereditary family estate orta a company of Janissaries palanka a small fort, especially a wooden one paşalık a territory governed by a pasha, esp. a sancak qubba dome, epecially a domed tomb (Sudan)

Page 3 of 4 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Glossary rawshan (Ar., colloquial roshan) a projecting window or balcony covered by decorative wooden grilles (Sudan) reaya civilian subjects salyane taxes sent annually to Istanbul sancak a provincial district; a military command sancakbeyi the governor of a sancak; a military commander saqiyya water-wheel sekban irregular troops serasker military commander serhad frontier sicil register sipahi cavalry tahrir taxation survey of a province tapu title deed tapu tahrir defteri a tax register, tahrir defteri timar an apanage granted in return for military service trace italienne star-shaped fort design, originating from fifteenth- century Italy tüfenkçi (pl. tüfenkçiyan) rifleman uc frontier vakıf (Ar. waqf; pl. evkaf, awqaf) endowment vali governor vilayet a province voynuk local recruits to the Ottoman forces, with a high degree of autonomy (Balkans) yerlü local Janissary troops zaim holder of a zeamet zakat alms zariba a fortified enclosure used as a holding camp by slave traders zaviye (Ar. zawiyya) a dervish lodge; a shelter for travellers zeamet a large timar zimmi (Ar. dhimmi) a protected member of a religious group, a Christian or Jew

Access brought to you by: Page 4 of 4 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Sultans

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

(p.554) (p.555) Ottoman Sultans Osman I

?—c.1324

Orhan

c.1324–62

Murad I

1362–89

Bayezid I (Yıldırım, ‘The Thunderbolt’)

1389–1402

(Civil War

1402–13)

Mehmed I

1413–21

Murad II

1421–44, 1446–51

Mehmed II (Fatih, ‘The Conqueror’)

1444–6, 1451–81

Bayezid II

1481–1512

Selim I (Yavuz, ‘The Grim’)

1512–20

Süleyman I (Kanuni, ‘The Magnificent’)

1520–66

Selim II

1566–74

Murad III

1574–95

Mehmed III

1595–1603

Ahmed I

1603–17

Mustafa I

1617–18, 1622–3

Osman II

1618–22

Murad IV

1623–40

Ibrahim

1640–8

Mehmed IV

1648–87

Süleyman II

1687–91

Ahmed II

1691–5

Mustafa II

1695–1703

Ahmed III

1703–30

Mahmud I

1730–54

Page 1 of 2 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Ottoman Sultans Osman III

1754–7

Mustafa III

1757–74

Abdülhamid I

1774–89

Selim III

1789–1807

Mustafa IV

1807–8

Mahmud II

1808–39

Abdülmecid

1839–61

Abdülaziz

1861–76

Murad V

1876

Abdülhamid II

1876–1909

Mehmed V Reşad

1909–18

Mehmed VI Vahdeddin

1918–22

Access brought to you by:

Page 2 of 2 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Select Bibliography

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

(p.556) (p.557) Select Bibliography General Ágoston, G., ‘A flexible empire: authority and its limits on the Ottoman frontiers’, in Kemal H. Karpat with Robert W. Zens (eds.), Ottoman Borderlands: Issues, Personalities, and Political Changes (Madison: University of Wisconsin Press, 2003), 15–29 Ágoston, G., Guns for the Sultan: Military Power and the Weapons Industry in the Ottoman Empire (Cambridge: Cambridge University Press, 2005) Aksan, V., Ottoman Wars 1700–1870: An Empire Besieged (Harlow: Longman, 2007) Baram, U. and Carroll, L. (eds.), A Historical Archaeology of the Ottoman Empire: Breaking New Ground (New York: Kluwer Academic/Plenum Publishers, 2000) Brauer, R., ‘Boundaries and frontiers in medieval Muslim geography’, Transactions of the American Philosophical Society, n.s. 85/6 (1995), 1–73 Dankoff, R., An Ottoman Mentality: The World of Evliya Çelebi (Leiden: Brill, 2004) Ebel, K., ‘Representations of the frontier in Ottoman town views of the sixteenth century’, Imago Mundi, 60 (2008), 1–22 Faroqhi, S., The Ottoman Empire and the World Around It (London: I. B. Tauris, 2004) Finkel, C., Osman’s Dream: The Story of the Ottoman Empire, 1300–1923 (London: John Murray, 2005) Hepper, M., ‘Center and periphery in the Ottoman Empire: with special reference to the nineteenth century’, International Political Science Review, 1 (1980), 81–105 Heywood, C., ‘The frontier in Ottoman history: old ideas and new myths’, in Daniel Power and Naomi Standen (eds.), Frontiers in Question: Eurasian Borderlands, 700–1700 (Basingstoke: Macmillan, 1999), 228–50 İnalcık, H., ‘Ottoman methods of conquest’, Studia Islamica, 2 (1953), 103–29

Page 1 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Select Bibliography İnalcık, H. with Quataert, D. (eds.), An Economic and Social History of the Ottoman Empire, 1300–1914 (Cambridge: Cambridge University Press, 1994) Karpat, K. H. with Zens, R. W., Ottoman Borderlands: Issues, Personalities and Political Changes (Madison: University of Wisconsin Press, 2003); also published as International Journal of Turkish Studies, 9 (2003) Lowry, H., The Nature of the Early Ottoman State (Albany, NY: State University of New York, 2003) Murphey, R., Ottoman Warfare, 1500–1700 (New Brunswick: Rutgers University Press, 1999) Toledano, E., ‘The emergence of Ottoman-local elites 1700–1900: a framework for research’, in Moshe Maoz and Ilan Pappe (eds.), Middle Eastern Politics and Ideas: A History from Within (London: I. B. Tauris, 1998), 145–62 (p.558) Anatolian Frontiers Allen, W. E. D. and Muratoff, P., Caucasian Battlefields: A History of the Wars on the TurcoCaucasian Border, 1828–1921 (Nashville: Battery Press, 1999) Bruinessen, Martin van, Agha, Shaikh and State: On the Social and Political Organization of Kurdistan (Utrecht, 1978) Dündar (Aydın), Erzurum Beylerbeyliği ve Teşkilatι. Kuruluş ve Genişleme Devri (1535–1566) (Ankara: TTK, 1998) Har-El, S., Struggle for Domination in the Middle East: The Ottoman-Mamluk War 1485–1491 (Leiden: Brill, 1995) Kılıç, O., XVI. ve XVII. Yüzyıllarda Van (1548–1648) (Van: Belediye Başkanlığı Kültür ve Sosyal İşler Müdürlüğü, 1997) Kılıç, O., XVI Yüzyılda Adilcevaz ve Ahlat (1534–1605) (Ankara: Tamga Yayıncılık, 1999) Kırzıoglu, M. F., Osmanlılar’ın Kafkas-Elleri’ni Fethi, 1451–1590 (Ankara: TTK, 1993) Kuneralp, S., ‘The Ottoman Drang Nach Osten: the Turco-Persian border problem in Azerbeican, 1905–1912’, in Sinan Kuneralp (ed.), Studies on Ottoman Diplomatic History IV (Istanbul: Isis, 1990), 71–6 Kütükoglu, B., Osmanlı-İran Siyasi Münasebetleri, 1578–1590 (Istanbul: İÜ Edebiyat Fakültesi Matbaası, 1962) Lindner, R. P., Nomads and Ottomans in Medieval Anatolia (Bloomington: Indiana University, 1983) Murphey, R., ‘Resumption of the Ottoman-Safavid border conflict, 1603–1638: effects of border destabilization on the evolution of state-tribe relations’, in Stefan Leder and Bernard Streck (eds.), Shifts and Drifts in Nomad-Sedentary Relations (Wiesbaden: Ludwig Reichert Verlag, 2005), 307–23

Page 2 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Select Bibliography Öz, M., ‘Ottoman provincial administration in eastern and southeastern Anatolia: the case of Bidlis in the sixteenth century’, in Kemal H. Karpat with Robert W. Zens (eds.), Ottoman Borderlands: Issues, Personalities, and Political Changes (Madison: University of Wisconsin Press, 2003), 145–56 Shaw, S. J., ‘Iranian relations with the Ottoman Empire in the eighteenth and nineteenth centuries’, in Peter Avery et al. (eds.), The Cambridge History of Iran, VII: From Nadir Shah to the Islamic Republic (Cambridge: Cambridge University Press, 1991), 297–313 Sinclair, T., ‘The Ottoman arrangements for the tribal principalities of the Lake Van region of the sixteenth century’, in Kemal H. Karpat with Robert W. Zens (eds.), Ottoman Borderlands: Issues, Personalities, and Political Changes (Madison: University of Wisconsin Press, 2003), 119–43 Sinclair, T., ‘The city of Adilcevaz in the late Middle Ages and the early Ottoman period’, in C. Imber and K. Kiyotaki (eds.), Frontiers of Ottoman Studies, II: State, Province, and the West (London: I. B. Tauris, 2005), 109–25 Woods, J., The Aqquyunlu: Clan, Confederation, Empire (Salt Lake City: University of Utah Press, 1999) Zeyneloğlu, C., Şirvanşahlar Yurdu (Istanbul: Cumhuriyet Kitaphanesi, 1931) Arabia, Iraq, the Gulf and Yemen Abdullah, Thabit A. J., Merchants, Mamluks, and Murder: The Political Economy of Trade in Eighteenth-Century Basra (Albany, NY: State University of New York, 2001) (p.559) Anscombe, F., The Ottoman Gulf: The Creation of Kuwait, Saudi Arabia, and Qatar (New York: Columbia University Press, 1997) Bacqué-Grammont, J.-L. and Kroel, A., Mamlouks, Ottomans et Portugais en Mer Rouge: L’affaire de Djedda en 1517 (Cairo: IFAO, 1988) Bakhit, M. A., The Ottoman Province of Damascus in the Sixteenth Century (Beirut: Librairie du Liban, 1982) Barbir, K., Ottoman Rule in Damascus 1708–1758 (Princeton: Princeton University Press, 1980) Barr, J., Setting the Desert on Fire: T. E. Lawrence and Britain’s Secret War in Arabia, 1916–1918 (London: Bloomsbury, 2006) Blumi, I., ‘The Ottoman Empire and Yemeni politics in the sancaq of Ta‘izz, 1911–1918’, in J. Hanssen, T. Philipp and S. Weber (eds.), The Empire in the City: Arab Provincial Capitals in the Late Ottoman Empire (Beirut: Beiruter Texte und Studien, 2002), 349–67 Bostan, İ., ‘Basra Körfezinin Güney Kesimi ve Osmanlılar, 1876–1908’, Osmanlı Araştırmaları, 9 (1989), 311–22 Buzpınar, Ş. T., ‘Abdülhamid II and Sayyid Fadl Pasha of Hadramawt’, Osmanlı Araştırmaları, 13 (1993), 227–39 Buzpınar, Ş. T., ‘The Hijaz, Abdulhamid II and Amir Hussein’s secret dealings with the British’, Middle Eastern Studies, 31 (1995), 99–123 Page 3 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Select Bibliography Buzpınar, Ş. T., ‘Vying for power and influence in the Hijaz: Ottoman rule, the last emirate of Abdulmuttalib and the British (1880–1882)’, The Muslim World, 95 (2005), 1–22 Çetinsaya, G., ‘The Caliph and mujtahids: Ottoman policy towards the Shi[i community of Iraq in the late nineteenth century’, Middle Eastern Studies, 41 (2005), 561–74 Çetinsaya, G., ‘The Caliph and the shaykhs: Abdülhamid II’s policy towards the Qadiriyya of Mosul’, in Itzchak Weismann and Fruma Zachs (eds.), Ottoman Reform and Muslim Regeneration: Studies in Honor of Butrus Abu-Manneh (London: I. B. Tauris, 2005), 97–107 Çetinsaya, G., Ottoman Administration of Iraq, 1890–1908 (London: Routledge, 2006) Farah, C. E., The Sultan’s Yemen: Nineteenth-Century Challenges to Ottoman Rule (London: I. B. Tauris, 2002) Faroqhi, S., Pilgrims and Sultans: The Hajj under the Ottomans 1517–1683 (London: I. B. Tauris, 1994) Gündüz, A., Osmanlı İdaresinde Musul, 1523–1639 (Elazığ: Fırat Üniversitesi, 2003) Halaçoğlu, Yusuf, ‘Midhat Paşa’nın Necid ve Havalisi ile İlgili Bir Kaç Lâyihası’, Tarih Enstitüsü Dergisi, 3 (1972), 149–76 Hathaway, J. with Barbir, K., The Arab Lands under Ottoman Rule, 1516–1800 (London: Longman, 2008) Hourani, A., ‘The Fertile Crescent in the eighteenth century’, Studia Islamica, 8 (1957), 91–118 Hughes, M., ‘What did the Arab Revolt contribute to the Palestine campaign? An assessment’, The Journal of the T. E. Lawrence Society, 15 (2006), 75–87 Jafar Pasha Al-Askari, A Soldier’s Story: From Ottoman Rule to Independent Iraq (London: Arabian Publishing, 2003) Jomier, J., Le Maḥmal et la caravane égyptienne des pèlerins de la Mecque (XIIIe–XXe siècles) (Cairo: IFAO, 1953) Khoury, D., State and Provincial Society in the Ottoman Empire: Mosul, 1540–1834 (Cambridge: Cambridge University Press, 1997) Kurşun, Z., Necid ve Ahsa’da Osmanlı Hakimiyeti: Vehhabî Hareketi ve Suud Devleti’nin Ortaya Çıkışı (Ankara: TTK, 1998) Kurşun, Z., The Ottomans in Qatar: A History of Anglo-Ottoman Conflicts in the Persian Gulf (Istanbul: Isis, 2002) (p.560) Longrigg, S. H., Four Centuries of Modern Iraq (Oxford: Clarendon Press, 1925) Mandaville, J. E., ‘The Ottoman province of al-Hasa in the sixteenth and seventeenth centuries’, Journal of the American Oriental Society, 90 (1970), 486–513 Marufoğlu, S., Osmanlı Döneminde Kuzey Irak, 1831–1914 (Istanbul: Eren, 1998)

Page 4 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Select Bibliography Matthee, R., ‘The Safavid-Ottoman frontier: Iraq-i Arab as seen by the Safavids’, in Kemal H. Karpat with Robert W. Zens (eds.), Ottoman Borderlands: Issues, Personalities and Political Changes (Madison: University of Wisconsin Press, 2003), 157–73 Murphey, R., ‘The construction of a fortress at Mosul in 1631: a case study of an important facet of Ottoman military expenditure’, in Osman Okyar and Halil İnalcık (eds.), Social and Economic History of Turkey (1071–1920) (Ankara: Hacettepe Üniversitesi, 1980), 163–77 Nicholson, J., The Hejaz Railway (London: Stacey International, 2005) Nieuwenhuis, T., Politics and Society in Early Modern Iraq: Mamluk Pashas, Tribal Shaykhs and Local Rule between 1802 and 1831 (The Hague: Martinius Nijhoff, 1982) Özbaran, S., ‘The Ottoman Turks and the Portuguese in the Persian Gulf, 1534–1581’, in R. B. Serjeant and B. L. Bidwell (eds.), Arabian Studies IV (London: Hurst and Company, 1978), 125– 32 Özbaran, S., ‘Ottoman naval policy in the south’, in Metin Kunt and Christine Woodhead (eds.), Süleyman the Magnificent and his Age (London: Longman, 1995), 58–64 Petersen, A. D., ‘The fortification of the pilgrimage route during the first three centuries of Ottoman rule (1516–1757)’, in K. ‘Amr, F. Zayadine and M. Zaghoul (eds.), Studies in the Archaeology of Jordan V (Amman: Department of Antiquities, 1995), 299–305 Petersen, A., ‘Qal‘at Ras al-‘Ayn: a sixteenth century Ottoman fortress’, Levant, 30 (1998), 97– 112 Rogan, E., Frontiers of the State in the Late Ottoman Empire: Transjordan, 1850–1921 (Cambridge: Cambridge University Press, 1999) Shields, S., Mosul Before Iraq: Like Bees Making Five-Sided Cells (Albany, NY: State University of New York, 2000) Smith, C. K., ‘Kawkaban, the key to Sinan Pasha’s campaign in the Yemen (March 1569-March 1571)’, Proceedings of the Seminar for Arabian Studies, 32 (2002), 287–94 Wavell, A. P., The Palestine Campaigns (3rd edn., London: Constable, 1931) Central and Eastern Europe Abou-el-Haj, R. A., ‘The formal closure of the Ottoman frontier in Europe: 1699–1703’, Journal of the American Oriental Society, 89 (1969), 467–73 Adanır, F. and Faroqhi, S. (eds.), The Ottomans and the Balkans: A Discussion of Historiography (Leiden: Brill, 2002) Ágoston, G., ‘Habsburgs and Ottomans: defense, military change and shifts in power’, Turkish Studies Association Bulletin, 22 (1998), 126–41 Ágoston, G., ‘Information, ideology, and limits of imperial policy: Ottoman grand strategy in the context of Ottoman-Habsburg rivalry’, in Virginia H. Aksan and Daniel Goffman (eds.), The Early

Page 5 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Select Bibliography Modern Ottomans: Remapping the Empire (Cambridge: Cambridge University Press, 2007), 75– 103 Aksan, V., Ottomans and Europeans: Contacts and Conflicts (Istanbul: Isis, 2004) Anastasopoulos, A. and Kolovos, E. (eds.), Ottoman Rule and the Balkans, 1760–1850: Conflict, Transformation, Adaptation (Rethymno: University of Crete, 2007) Anscombe, F. (ed.), The Ottoman Balkans, 1750–1830 (Princeton: Markus Wiener, 2006) (p.561) Baer, M. D., Honored by the Glory of Islam: Conversion and Conquest in Ottoman Europe (New York: Oxford University Press, 2008) Bayerle, G., Ottoman Diplomacy in Hungary: Letters from the Pashas of Buda 1590–1593 (Bloomington: Indiana University, 1972) Bayerle, G., The Hungarian Letters of Ali Pasha of Buda, 1604–1616 (Budapest: Akadémiai Kiadó, 1991) Berindei, M. and Veinstein, G., ‘Les possessions ottomanes entre Bas-Danube et Bas-Dniepr: réglements fiscaux et fiscalité de la province de Bender-Aqkerman, 1570’, Cahiers du Monde russe et soviétique, 22 (1981), 251–328 Beydilli, Kemal, ‘Karadeniz’in Kapalılığı Karşısında Avrupa Küçük Devletleri ve “Miri Ticaret” Teşebbüsü’, Belleten, 55/214 (1991), 687–755 Bostan, İdris, ‘Rusya’nın Karadeniz’de Ticarete Başlaması ve Osmanlı İmparatorluğu, 1700– 1787’, in idem (ed.), Beylikten İmparatorluğa Osmanlı Denizciliği (Istanbul: Kitap Yayınevi, 2006), 285–324 Collins, L. J. D., ‘The military organisation and tactics of the Crimean Tatars during the sixteenth and seventeenth centuries’, in V. J. Parry and M. E. Yapp (eds.), War, Technology and Society in the Middle East (London: Oxford University Press, 1975), 257–76 Constantiniu, F., ‘Tradition and innovation in the eighteenth-century military structures of the Rumanian lands’, in Gunther Rothenberg, Bela Király and Peter Sugar (eds.), War and Society in East Central Europe, II: East Central European Society and War in the Prerevolutionary Eighteenth Century (New York: Brooklyn College, 1982), 387–99 Dávid, G., Studies in the Demographic and Administrative History of Ottoman Hungary (Istanbul: Isis, 1997) Dávid, G. and Fodor, P. (eds.), Hungarian-Ottoman Military and Diplomatic Relations in the Age of Süleyman the Magnificent (Budapest: ELTE, 1994) Dávid G. and Fodor, P. (eds.), Ottomans, Hungarians, and Habsburgs in Central Europe: The Military Confines in the Era of Ottoman Conquest (Leiden: Brill, 2000) Finkel, C., The Administration of Warfare: The Ottoman Military Campaigns in Hungary, 1593– 1606 (Vienna: VWGÖ, 1988)

Page 6 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Select Bibliography Finkel, C. and Ostapchuk, V., ‘Outpost of empire: an appraisal of Ottoman building registers as sources for the archaeology and construction history of the Black Sea fortress of Özi’, Muqarnas, 22 (2005), 150–88 Fodor, P., In Quest of the Golden Apple: Imperial Ideology, Politics, and Military Administration in the Ottoman Empire (Istanbul: Isis, 2000) Gerelyes, I. and Kovács, G. (eds.), Archaeology of the Ottoman Period in Hungary (Budapest: Hungarian National Museum, 2003) Gerő, G., ‘Türkische Keramik in Ungarn. Einheimische und importierte Waren’, in Géza Fehér (ed.), Fifth International Congress of Turkish Art (Budapest: Akadémiai, 1978), 347–61 Gradeva, R., Rumeli under the Ottomans, 15th-18th centuries: Institutions and Communities (Istanbul: Isis, 2004) Hickok, M., Ottoman Military Administration in Eighteenth-Century Bosnia (Leiden: Brill, 1997) Holl, I., Fundkomplexe des 15.-17. Jahrhunderts aus dem Burgpalast von Buda (Budapest: Archäologisches Institut der Ungarischen Akademie der Wissenschaften, 2005) İnalcık, H., ‘The origins of the Ottoman-Russian rivalry and the Don-Volga Canal’, Annales de l’Université d’Ankara, 1 (1947), 47–110 İnalcık, H., ‘Stefan Duşan’dan Osmanlı İmparatorluğuna: XV Asırda Rumeli’de Hıristyan Sipahiler ve Menşeleri’, in Mélanges Fuad Köprülü: 60 Doğum Yılı Münasebetiyle Fuad Köprülü Armağanı (Istanbul: Ankara Üniversitesi Dil ve Tarih-Coğrafya Fakültesi, 1953), 207–48 (p.562) Khodarkovsky, M., Russia’s Steppe Frontier: The Making of a Colonial Empire, 1500– 1800 (Bloomington: Indiana University Press, 2002) Kovács, G., ‘Iznik pottery in Hungarian archaeological research’, in Ibolya Gerelyes (ed.), Turkish Flowers: Studies on Ottoman Art in Hungary (Budapest: Hungarian National Museum, 2005), 69–86 Kurat, A. N., ‘The Turkish expedition to Astrakhan in 1569 and the problem of the Don-Volga Canal’, Slavonic and East European Review, 40 (1961), 7–23 Maxim, M., L’Empire ottoman au nord du Danube et l’autonomie des Principautés Roumaines au XVIe siècle: Études et documents (Istanbul: Isis, 1999) Minkov, A., Conversion to Islam in the Balkans: Kisve Bahası Petitions and Ottoman Social Life, 1670–1730 (Leiden: Brill, 2004) Mutafchieva, V., L’Anarchie dans les Balkans à la fin du XVIIIe siècle (Istanbul: Isis, 2005) Ostapchuk, V., ‘The human landscape of the Ottoman Black Sea in the face of the Cossack naval raids’, Oriente Moderno, n.s. 20 (2001) (Special Issue, The Ottomans and the Sea, ed. Kate Fleet), 23–95 Panaite, V., The Ottoman Law of War and Peace: The Ottoman Empire and Tribute Payers (New York: Columbia University Press, 2000) Page 7 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Select Bibliography Parvev, Ivan, Habsburgs and Ottomans between Vienna and Belgrade (1683–1739) (New York: Columbia University Press, 1995) Radušev, E., ‘Ottoman border periphery (serhad) in the Nikopol Vilayet, first half of the sixteenth century’, Études balkaniques, 31 (1995), 140–60 Reinkowski, M., ‘Double struggle, no income: Ottoman borderlands in northern Albania’, in Kemal H. Karpat with Robert W. Zens (eds.), Ottoman Borderlands: Issues, Personalities and Political Changes (Madison: University of Wisconsin Press, 2003), 239–53 Roksandić, D. and Štefanec, N. (eds.), Constructing Border Societies on the Triplex Confinium (Budapest: Central European University, 2000) Römer, C., Osmanische Festungsbesatzungen in Ungarn zur Zeit Murāds III (Vienna: VÖAW, 1995) Rothenberg, G., The Austrian Military Border in Croatia, 1522–1747 (Urbana, Illinois: University of Illinois Press, 1960) Rothenberg, G., The Military Border in Croatia 1740–1881: A Study of an Imperial Institution (Chicago: University of Chicago Press, 1966) Şahin, İlhan, ‘XVI. yüzyılda Akkerman’ın demografik ve sosyal durumu’, Güneydoğu Avrupa Araştırmaları Dergisi, 12 (1982–98), 319–23 Stein, M., Guarding the Frontier: Ottoman Border Forts and Garrisons in Europe (London: I. B. Tauris, 2007) Suceska, A., ‘The eighteenth-century Austro-Ottoman wars’ economic impact on the population of Bosnia’, in Bela Király and Peter Sugar (eds.), War and Society in East Central Europe (New York: Brooklyn College Studies in Society in Change, 1982), 2: 339–50 Sugar, P., ‘The Ottoman “professional prisoner” on the western borders of the Empire in the sixteenth and seventeenth centuries’, Études balkaniques, 7 (1971), 82–91 Vasić, M., ‘Die Martolosen im osmanische Reich’, Zeitschrift für Balkanologie, 2 (1964), 172–89 Vasić, M., ‘The martoloses in Macedonia’, Macedonian Review, 7/1 (1977), 30–41 Veinstein, G., ‘Les “çiftlik” de colonisation dans les steppes du nord de la Mer noire au XVIe siècle’, Istanbul Universitesi Iktisat Fakultesi Mecmuası, 41 (1982–3), 177–210 Zens, R., ‘Pasvanoglu Osman Paşa and the PaŞalık of Belgrade, 1791–1807’, International Journal of Turkish Studies, 8 (2002), 89–104 (p.563) The Eastern Mediterranean Frontier and Greece Anderson, R. C., Naval Wars in the Levant, 1559–1853 (Liverpool: Liverpool University Press, 1952) Andrews, K., Castles of the Morea (Princeton: Princeton University Press, 2006)

Page 8 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Select Bibliography Bracewell, W., The Uskoks of Senj: Piracy, Banditry and Holy War in the Sixteenth-Century Adriatic (Ithaca and London: Cornell University Press, 1992) Daskalakis, A., I Mani kai i Othomaniki Autokratia, 1453–1821 (Athens, 1928) Fleming, K., The Muslim Bonaparte: Diplomacy and Orientalism in Ali Pasha’s Greece (Princeton: Princeton University Press, 1999) Greene, M., ‘The Ottomans in the Mediterranean’, in Virginia H. Aksan and Daniel Goffman (eds.), The Early Modern Ottomans: Remapping the Empire (Cambridge: Cambridge University Press, 2007), 104–16 Guilmartin, J., Gunpowder and Galleys: Changing Technology and Mediterranean Warfare at Sea in the Sixteenth Century (Cambridge: Cambridge University Press, 1980) Hess, A., ‘The Battle of Lepanto and its place in Mediterranean history’, Past & Present, 57 (1972), 53–73 İnalcık, H., ‘Lepanto in the Ottoman documents’, in Gino Benzoni (ed.), Il Mediterraneo nella seconda metà del ’500 alla luce di Lepanto (Florence: Leo S. Olschki, 1974), 185–92 Karal, E., ‘Yunan Adalarının Fransızlar Tarafından İşgali ve Osmanlı-Rus Münasebatı, 1797– 1798’, Tarih Semileri Dergisi, 1 (1937), 120–4 Plomer, W., The Diamond of Jannina: Ali Pasha, 1741–1822 (London: Cape, 1970; 1st edn. 1936) Skiotis, D., ‘From bandit to pasha: first steps in the rise to power of Ali of Tepelen, 1750–1784’, International Journal of Middle East Studies, 2 (1971), 219–44 Skiotis, D., ‘The Greek revolution: Ali Pasha’s last gamble’, in Nikoforos Diamandouros, John Anton, John Petropoulos and Peter Topping (eds.), Hellenism and the First Greek War of Liberation (1821–1830): Continuity and Change (Thessaloniki: Institute for Balkan Studies, 1976), 97–109 Tenenti, A., Naufrages, corsaires et assurances maritimes à Venise, 1592–1609 (Paris, 1959) Thys-Şenocak, L., Ottoman Women Builders: The Architectural Patronage of Hadice Turhan Sultan (Aldershot: Ashgate Press, 2007) Uzunçarşılıoğlu, İ. H., ‘Arşiv Vesikalarına Göre Yedi Ada Cümhuriyeti’, Belleten, 1/34 (1937), 627–40 Wagstaff, J. M., ‘Settlements in the south-central Pelopónnisos c.1618’, in F. W. Carter (ed.), An Historical Geography of the Balkans (London: Academic Press, 1977), 197–238 Wagstaff, J. M., ‘John Philip Morier’s account of the Mani, 1804’, in Y. Saïtis (ed.), Máni: Témoignes sur L’Espace et la Société (Athens: Institut de Recherches Néohelléniques, 1996), 273–87 Zarinebaf, F., Bennet, J. and Davis, J. L. (eds.), A Historical and Economic Geography of Ottoman Greece: The Southwestern Morea in the Eighteenth Century (Athens: American School of Classical Studies at Athens, 2005) Page 9 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Select Bibliography The Western Mediterranean Frontier and North Africa Austen, R., ‘The Mediterranean Islamic slave trade out of Africa: a tentative census’, Slavery and Abolition, 13 (1992) (Special issue, The Human Commodity: Perspectives on the Trans-Saharan Slave Trade), 214–48 Braudel, F., The Mediterranean and the Mediterranean World in the Age of Philip II, trans. Siân Reynolds (London: Fontana, 1975) (p.564) D’Angelo, M., ‘In the “English” Mediterranean (1511–1815)’, Journal of Mediterranean Studies, 12 (2002), 271–85 Djelloul, N., Les Fortifications côtières ottomanes de la Régence de Tunis (XVIe–XIXe siècles) (Zaghouan: FTERSI, 1995) el-Moudden, A., ‘The Sharif and the Padishah: some remarks on Moroccan-Ottoman relations in the sixteenth century’, in Selim Deringil and Selim Kuneralp (eds.), Studies on Ottoman Diplomatic History, V: The Ottomans and Africa (Istanbul: Isis, 1990), 27–34 Fontenay, M. and Tenenti, A., ‘Course et piraterie méditerranéennes de la fin du Moyen Âge aux débuts du XIXe siècle’, in M. Mollat (ed.), Course et piraterie (Paris: IRHT, CNRS, 1975), 78–131; reissued, with an extensive updated bibliography by M. Fontenay, in Revue d’Histoire Maritime, 6 (2006), 173–228 Hess, A. C., ‘The Moriscos: an Ottoman fifth column in sixteenth-century Spain’, The American Historical Review, 74 (1968), 1–25 Hess, A. C., ‘The forgotten frontier: the Ottoman North African provinces during the eighteenth century’, in Thomas Naff and Roger Owen (eds.), Studies in Eighteenth Century Islamic History (Carbondale and Edwardsville: Southern Illinois University Press, 1977), 74–87, 371–3 Hess, A. C., The Forgotten Frontier: A History of the Sixteenth-Century Ibero-African Frontier (Chicago and London: University of Chicago Press, 1978) Heywood, C., ‘An English merchant and consul-general in Algiers, c.1676–1712: Robert Cole and his circle’, in Abdeljelil Temimi and Mohamed-Salah Omri (eds.), The Movement of People and Ideas between Britain and the Maghreb (Actes du IIème Congrès du dialogue Britano-Maghrébin tenu à l’Université d’Exeter, 14–17 Sept. 2002) (Zaghouan: FTERSI, 2003), 49–66 Heywood, C., ‘A “forgotten frontier”? Algiers and the maritime frontier from the French bombardment (1682) to the Algiers earthquake (1716)’, Revue d’Histoire Maghrébine, 114 (2004), 35–50 Hunwick, J. and Troutt Powell, E., The African Diaspora in the Mediterranean Lands of Islam (Princeton: Marcus Wiener, 2002) Khiari, F., Vivre et mourir en Alger: L’Algérie ottomane aux XVIe-XVIIe siècles: un destin confisqué (Paris: L’Harmattan, 2002) Merouche, L., Recherches sur l’Algérie à l’époque ottomane, I: Monnaies, prix et revenus; II: La Course: mythes et réalité (Paris: Ed. Bouchene, 2002, 2007)

Page 10 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Select Bibliography Panzac, D., ‘Le commerce maritime de Tripoli de Barbarie dans la seconde moitié du XVIIIe siècle’, Revue d’Histoire Maghrébine, 69–70 (1993), 141–67 Panzac, D., La Caravane maritime: Marins européens et marchands ottomans en Méditerranée (Paris: CNRS, 2004) Panzac, D., Barbary Corsairs: The End of a Legend 1800–1820 (Leiden: Brill, 2005) Pennell, C. R., Piracy and Diplomacy in Seventeenth-Century North Africa: The Journal of Thomas Baker, English Consul in Tripoli, 1677–1685 (Rutherford: Fairleigh Dickinson University Press, 1987) Renault, F., ‘La traite des esclaves noirs en Libye au XVIIIe siècle’, Journal of African History, 23 (1982), 163–81 Shuval, T., ‘The Ottoman Algerian elite and its ideology’, International Journal of Middle East Studies, 32 (2000), 323–44 Wright, J., ‘Enforced migration: the black slave trade across the Mediterranean’, The Maghreb Review, 31 (2006), 62–70 (p.565) Sub-Saharan Africa and the Sudan Abir, M., Ethiopia and the Red Sea (London: Cass, 1980) Ahmed, H., ‘The Turkish taxation system and its impact on agriculture in the Sudan’, Middle Eastern Studies, 16 (1980), 105–14 Alexander, J. A., ‘The Turks on the Middle Nile’, Archéologie du Nil Moyen, 7 (1996), 15–35 Alexander, J. A., ‘Qalat Sai: the most southerly Ottoman fortress in Africa’, Sudan and Nubia, 1 (1997), 16–20 Alexander, J. A., ‘The archaeology and history of the Ottoman frontier in the Middle Nile Valley 910–1233 AH / 1504–1820 AD’, Adumatu, 1 (2000), 47–61 Alexander, J. A., ‘The real Turkia’, Cahiers de Récherches de l’Institut de Papyrologie et d’Égyptologie de Lille: Mélanges offerts à Francis Geus, 26 (2006), 17–28 Baer, G., ‘Slavery in nineteenth century Egypt’, Journal of African History, 8 (1967), 417–41 Bloss, J. F. E., ‘The Story of Suakin, Part I–II’, Sudan Notes and Records, 19 (1936), 271–300 Bloss, J. F. E., ‘The Story of Suakin, Part III’, Sudan Notes and Records, 20 (1937), 247–80 el Zein, I. S., ‘The Ottomans and the Mahas in the Third Cataract region’, Azania, 39 (2004), 50– 7 Greenlaw, Jean-Pierre, The Coral Buildings of Suakin (Stocksfield: Oriel Press, 1976). Re-published as The Coral Buildings of Suakin: Islamic Architecture, Planning, Design and Domestic Arrangements in a Red Sea Port (London: Kegan Paul International, 1995)

Page 11 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Select Bibliography Hill, R., Egypt in the Sudan, 1820–1881 (London: Oxford University Press, 1959) Holt, P. M., ‘Sultan Selim I and the Sudan’, Journal of African History, 8 (1963), 19–23 Johnson, D., ‘Sudanese military slavery from the 18th to the 20th century’, in Léonie J. Archer (ed.), Slavery and Other Forms of Unfree Labour (London: Routledge, 1988), 142–56 Kennedy Cooke, B., ‘The Red Sea coast in 1540’, Sudan Notes and Records, 16 (1933), 151–9 Martin, B. G., ‘Maî Idrîs of Bornu and the Ottoman Turks, 1576–78’, International Journal of Middle East Studies, 3 (1972), 470–90 McGregor, A., ‘The Circassian qubba-s of ‘Abbas Avenue, Khartoum: governors and soldiers in the nineteenth century Sudan’, Nordic Journal of African Studies, 10 (2001), 28–40 Ménage, V., ‘The Ottomans and Nubia in the sixteenth century’, Annales Islamologiques, 24 (1988), 137–51 Millingen, F., ‘On the negro slaves in Turkey’, Journal of the Royal Anthropological Institute, 8 (1870–1), lxxxvi–xcvi Moore-Harell, A., ‘The Turco-Egyptian army in Sudan on the eve of the Mahdiyya, 1877–80’, International Journal of Middle East Studies, 31 (1999), 19–37 Moore-Harell, A., ‘Economic and political aspects of the slave trade in Ethiopia and the Sudan in the second half of the nineteenth century’, International Journal of African Historical Studies, 32 (1999), 407–21 O’Fahey, R. and Spaulding, J., Kingdoms of the Sudan (London: Methuen, 1974) Orhonlu, C., Osmanlı İmparatorluğu’nun Güney Siyaseti: Habeş Eyaleti (Ankara: TTK, 1996) Özbaran, S., ‘A Turkish report on the Red Sea and the Portuguese in the Indian Ocean (1525)’, in idem, The Ottoman Response to European Expansion: Studies on Ottoman-Portuguese Relations in the Indian Ocean and Ottoman Administration in the Arab Lands during the Sixteenth Century (Istanbul: Isis, 1994), 99–109 Petti Suma, M. T., ‘Il viaggio in Sudan di Evliya Čelebi’, Annali dell’ Istituto Orientale di Napoli, 14 (1964), 433–52 Rossi, E., ‘A Turkish map of the Nile River, about 1685’, Imago Mundi, 6 (1949), 73–5 Spaulding, J., ‘Slavery, land tenure and social class in the northern Turkish Sudan’, International Journal of African Historical Studies, 15 (1982), 1–20 (p.566) Toledano, E., ‘The imperial eunuchs of Istanbul: from Africa to the heart of Islam’, Middle Eastern Studies, 20 (1984), 379–90 Toledano, E., Slavery and Abolition in the Ottoman Middle East (Seattle: University of Washington Press, 1998)

Page 12 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Select Bibliography Udal, John O., The Nile in Darkness: Conquest and Exploration, 1504–1862 (Norwich: Michael Russell, 1998) Uğur, H., Osmanlı Afrikası’nda bir Sultanlık: Zengibar (Istanbul: Küre Yayınları, 2005) Yasamee, F. A. K, ‘The Ottoman Empire, the Sudan and the Red Sea coast, 1883–1889’, in Selim Deringil and Sinan Kuneralp (eds.), Studies on Ottoman Diplomatic History, V: The Ottomans and Africa (Istanbul: Isis, 1990), 87–102

Access brought to you by:

Page 13 of 13 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index

The Frontiers of the Ottoman World A.C.S. Peacock

Print publication date: 2009 Print ISBN-13: 9780197264423 Published to British Academy Scholarship Online: January 2012 DOI: 10.5871/bacad/9780197264423.001.0001

(p.567) Index Page numbers in italics denote illustrations. The letter n following a page number indicates the reference will be found in a note. Abaza Mehmed Pasha, 24 ‘Abbas I (Khedive), 513 ‘Abbas (Shah), 367 ‘Abbas Hilmi, 233 ‘Abd al-‘Aziz b. Sa‘ud, 250 ‘Abd al-Wahhab, 85–6 ‘Abdali, 290, 295, 297 ‘Abdallab tribal confederation, 228, 374 Abdić, Fikret, 409 Abdülhamid I, 176, 186 Abdülhamid II Hamidiye regiments, 249 in Iraq, 281–2, 284–5, 286, 287 photograph album, 200 title, 15 Abdülkadir Bey, 262, 264 Abdurrahman Pasha, 191 Abeshins, 483 Abu Bakr Mirza, 359 Abu Dhabi, 248 Abu ’l-Fida’, 97 Abu ’l-Lisan, 434, 440 Abu Numayy, 82 Abyssinia, 7, 513 Aceh, 8 Achaea, 113, 118, 120 Acre, 268 Acu-su Boğazı, 175n11 Adael, 520, 532–6 Adakale, 64 Page 1 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Adal, 227, 234 Addison, J., 505 Aden British control, 290–1, 292, 293–304 conquest 1538, 61, 82, 84 garrison, 516 Adilcevaz, 217, 219–20, 223 administration Adriatic frontier see Adriatic frontier, Napoleonic Wars Berberistan, 374–5 centre-periphery relations 1787–1915, 235–6, 250–1 Albania 1787–1820, 236–45, 237 eastern Arabia 1870–1915, 245–50, 246 frontier regions, 354 historical overview, 20–3 Iraq see Iraq, administration Karamanid principality, 323–4 north-east Africa see north-east Africa, frontier policies Van region, 16th century, 211–24, 212 Vidin, 347–9 Yemen see Yemen, administration Adobana, 474 Adony (Korkmaz), 400 Adriatic frontier Napoleonic Wars, 253–6 diplomatic problems, 256–61 discussion, 269–70 local interests versus imperial priorities, 262–9 Uskoks, 403 Afghans, 248 Africa see north Africa; north-east Africa Afyonkarahisar see Karahisar Agar Dinka, 518, 523–4, 535 Agate, 521n65 ‘Agayl, 108, 110 Agria (Erlau; Eger; Eğri), fortress geography and environment, 71, 73–4 map, 39–40 pottery, 392n29, 395, 396–7, 398, 399, 401 agriculture Iraq, 275, 277 north-east Africa, 230 slave labour, 513, 514 Vidin, 349–50 Agro-Commerce, 414 Ahıska (Akhaltsikhe), 361 Ahlat, 219, 220, 222, 223 Ahmad Ağa Karamanlu, 309n8 Ahmad Gran, 227 Page 2 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Ahmad Muhammad, 233 Ahmad Pasha Abu Adhan, tomb of, 371, 372 Ahmadi, 302 Ahmed I, 182 Ahmed Bey, 326 Ahmed Cevad Pasha, 340 Ahmed Cezzar, 242n22 Ahmed Izzet Pasha, 299n37 Ahmed Köprülü, 500 Ahmed Pasha, 227–8 Ahmed Raşid, 292 Ajrud, 97 Akhaltsikhe see Ahıska akhunds, 280, 281 akıncı troops, 342 (p.568) Akkerman (Bilhorod-Dnistrovs’kyi) fortress conquest, 139 description: 15th–17th centuries, 143–53, 144, 145, 146, 147; 18th-19th centuries 53–63, 159 discussion, 169–70 frontier life, evidence for, 163–9, 166, 168 frontier setting, 141–2 location, 140, 175 modifications, 19, 29 project, 140, 142–3, 170, 195 graveyard, 545 kaza, 139 sancak, 139 Yanık Palanka, 182 Akkoyunlu tribal confederation, 5, 316, 317, 318 Aksaray, 312, 314, 315 Akşehir administration, 312, 323 hinterland, 311 Karamanid control, 315, 317, 318 Via Sebaste, 322, 325, 326 Al Sabah family, 236, 246, 247 Al Thani family, 236, 248 al-Afghani, Jamal al-Din, 282 al-Barudi, 18 al-Damurdashi, 103 al-Diriya, 85 al-Dirr, 227, 374, 375 al-Diwan, 375 al-Hasa (Lahsa) administration, 245, 248, 249–50, 273 Ottoman control, 8, 10, 82, 83, 84, 94 Wahhabi raids, 86 Page 3 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index al-Jafr, 440 al-Jazari, 101 al-Khandaq, 371 al-Khasa, 471 al-Nasir Muhammad, 97 al-Qatif, 84 al-Sa‘id, 226 al-‘Ula, 88, 93 Alacahisar see Kruševac Alaeddin Bey, 315n17, 325 ‘Alawi, 290, 297, 299 Albak, 211–13, 217n21 Albania administration 1787–1820, 236–45, 237 Albanian migration, 18 frontier, 10, 16, 319 soldiers from, 244, 375, 516 workers from, 162 Albrizzi, G., plans by, 128, 133 Aleppo administration, 348n87 Arab Revolt, 437 consular records, 502 Ionian consulate, 268 trade, 215, 223 Zangi of, 13 Alexander I (Tsar), 255, 269 Alexander II (Khan), 359, 360 Alexandria, 268, 483, 486, 512 Alexandria Troas, 191 Algiers Braudel at, 497 corsairs, 26, 268, 355, 497, 498, 501, 503–6 Dey, 268, 505, 506 maritime frontier, 497–8, 500–1, 503–8 resources, 355, 504n41 slave trade, 26, 499 ‘Ali b. Mani‘, 295, 297, 300 ‘Ali b. Muqbil b. ‘Abd al-Hadi, 296 Ali Bey (of Egypt), 85, 104 Ali Bey (Turgutoğlu), 315, 327 ‘Ali Bey al-Dumyati, 103 Ali Bey Bulut Kapan, 512 Ali Bey Mihaloğlu, 342n52 ‘Ali Bey Zayn al-Faqar, 103 ‘Ali Dinar, 371 Âli Pasha (Mehmet Emin Âli Pasha), 178, 178n21, 182 Ali Pasha (Tepedelenli) see Tepedelenli Ali Pasha Alicante, 504 Page 4 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Altınekin see Zıvarık Alwa, 226 ‘Amara, 277 Amasya administration, 348n87 treaty of, 8, 215, 220, 223, 360, 361 Amid see Diyarbakır Amiens, treaty of, 261 Amin b. Qasim Husayn, 294n19 ‘Amiri, 290, 297, 299 Amla Diffi, 380 ‘Ammara, 304n58 Anagastum, 179 Anatolia administration, 20–1 canals, 16 conquest, 5, 12, 13 economy, 25 revolts, 25 see also Karamanid Principality; Safavids, frontier with; Van region Anavarin see Navarino Ancona, 64, 266 Andirion (Roumeli), 117 Andrews, K., 121, 129, 133 Androuvitsa (Ardovişta), 126 Anglo-Egyptian Condominium, 233 Ankara, 5, 7, 154 Antalya, 315, 315n20, 318 Anticythera (Cerigotto), 117 ‘Aqaba (Ayla) Arab Revolt, 437, 444 (p.569) caravan route, 440 Castle pre Ottoman period, 95–9 16th-17th centuries, 99–103, 100, 101, 110–12 18th-19th centuries, 103–8, 105, 106, 110–12 20th century, 108–12, 109, 111 location, 95, 96 plan, 99 ‘Aqabat Ayla see Naqb al-‘Aqaba ‘Aqrabi, 297 Arab Mehmed Pasha, 179 Arab Revolt 1916–18, archaeology of, 431–2 discussion, 449–51 Great Arab Revolt Project, 434–5 Ma‘an: first battle of, 447–9, 448; trench fortress at, 440–4, 442 Wadi Rutm railway station, 444–7, 445 guerrilla warfare, Lawrence’s conception of, 437–40 Page 5 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index historical background, 435–7 modern conflict, archaeology of, 432–4 Arabia administration of Eastern Arabia 1870–1915, 236, 245–50, 246 conquest of Period One, 1516–1700, 81–4, 83 Period Two, 1700–1800, 84–6 frontiers administration, 21 historical overview, 7, 8, 9, 10–11 society, 24 migrant holy men, 230 Syrian hajj route development, 86–7; 16th century, 87–90, 88, 89; 18th century, 90–1, 91, 92, 93 discussion, 93–4 trade, 484, 485, 486 see also al-Hasa; Arab Revolt; Hijaz; Yemen Arabian Gulf, 7–8, 16, 82, 274, 283, 284–7 Arad, 70, 71 Arapovac, 425 Aras, river, 361 archaeology, Ottoman, 3–4, 27, 544–7 future research, 547–50 maritime, 500–3 role of in frontier studies, 23, 542–4 Rumbek area, 525–32, 526, 528, 530, 531, 537 see also Akkerman, fortress; Arab Revolt, archaeology of; Hungary, garrisons and local population; Krajina Project; Suakin, archaeological project; Sudan, archaeology of middle Nile valley Archar, 334, 335, 337, 348n64 architects (mimar), 161, 195, 456–7; see also Mimar Sinan architectural models, 182–3, 186 architectural plans (resm), 202–3 architecture, Turko-Egyptian, 371, 372 archival materials, 3, 27 fortress works, 150–4, 155, 160–3, 170 Seddülbahir, 191–3, 194–6, 207–8 archival, 202–7, 204, 206 representational, 190, 196–202, 197, 198, 199, 200, 201, 202 see also building registers Ardovişta see Androuvitsa Arduan Diffi, 380 Arduan Island, 379–80 Arduan-Sageig, 380 Areopolis (Cibova; Tzimona), 127 Armenia, 10, 19, 141, 282, 285 Page 6 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Armenians, 216, 436 arms trade, 277, 286, 300, 301 army administration Albania, 1787–1820, 244–5 Arabia, 1870–1915, 247, 248, 249 Christian troops Anatolia, 13 Balkans, 185, 187 Hapsburg frontier, 464–8 Hungary, 25, 385, 400 Vidin/Danube, 338, 343–4, 345, 350–1 economic opportunities, Hapsburg frontier, 455–68 garrisons Adriatic frontier, 265–6 ‘Aqaba, 104–7, 108, 110 Arabia, 87 Danube, 335, 337–46, 349, 350–1 Egypt, 511–12 Hapsburg frontier, 403, 461 Hungary, 68, 71; see also Hungary, garrisons and local population, archaeology of Jordan, 441, 444, 450, 451n36 Karamanid lands, 329n91 north-east Africa, 17, 228, 230–1 Safavid frontier, 356–70 Sudan, 374–5, 376, 378, 379, 382, 516–17, 523–4 Van region, 217, 218, 221, 223 historical overview, 9–10, 13, 17, 19, 22, 25 Janissary corps see Janissaries modernisation, 435, 444 Nizam-ı Cedid, 241 slave soldiers, 513, 514, 515–17, 522 Arsenije III, 346 Arteiga, 473 Artindiffi, 380 Asafi, 366 Aşık Paşazade, 327 ‘Asir, 516 Asmara, 471 Astrakhan, 8, 60 Aşurviç, 177 Aswan Banu ‘Umar, 226 frontier, 226, 227, 231, 374 (p.570) garrison, 374 slave training camp, 516 trade, 230 Turkish education, 376 Page 7 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Asyut, 514 ‘Atabat, 280, 282 Augsburg, 35 Avnik, 219 ‘Awlaqi, 297 Awol Athiang see Wol Athiang Ayak, 520, 524, 532 ‘Aydhab, 471, 474 Aydin, 462 Ayla see ‘Aqaba Ayni Ali, 339 Ayo Varvara, 124, 126 Ayod, 519 Azak (Azov), 60, 140 Azande kingdom, 532 azeb troops, 461–3, 464, 467–8, 511 Azerbaijan, 213, 214, 278, 279, 356, 365–9 Aziziye Palanka, 180 Aznam, 98, 101 Azov see Azak Baban family, 273, 274 Babócsa (Bobofça), 392 Bács (Baç), 386 Badi‘, 471, 474 Baf, 203n39 Bağçesaray, 358n2 Baggara, 514 Baghdad administration, 272, 273–4, 275, 277, 285, 286 depiction of, 53–5, 54, 64 governor, 245 Shi‘i, 279 Baghdad railway, 10, 275, 284, 287 Bahr al-Ghazal, 516, 521–2, 524, 527, 532, 536 Bahr al-Jabal, 509, 518, 527 Bahrain, 82, 247, 248, 285, 287 Baker, Admiral, 505 Baker, V., 485 Balkans Balkan Wars, 1912–13, 279 civil buildings, 183–5 frontier formation and maintenance, 354–5 historical overview, 5, 9, 10, 15, 17–18 Ottomanisation, 183 palanka forts, 171–2, 172 background, late 18th century, 175–7, 177 Black Sea frontier, 182–3 Danube frontier, 181–2 Page 8 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index discussion, 185–7 geography, 173–5, 174 Hapsburg frontier, 178–81 Baluchis, 248 Balzac, H., 238 Bane see Soko Banja Baneh Süleyman Bey, 368n31 Bani Khalid tribe, 248–9 Bani Lam Bedouin, 88 Bani Sakhr, 93 Bani ‘Umar, 226 Banu ‘Umran, 104 Bar see Stari Bar Barakat II, 82 Baram, Uzi, 544–5 Barcs (Barç), 395, 396, 398, 400–1 Bargiri (Muradiye), 215 Baris, 514 Bartlett, W. H., 107, 108 Başkale, plain of, 213 Başkent, battle of, 5, 316 Basra administration, 273, 275, 277–8 fleet, 61, 273, 284, 286 governor, 249 Shi‘i, 279 trade, 84, 94, 273 Bátaszék (Batasek), 395, 396, 398, 400–1 Batn al-Ghul, 445, 447 Battlefield Archaeology, 432–3, 447–8 Bayazid, 279 Baybars, 95 Bayburd clan, 324n70 Bayda, 302 Bayezid I, 5, 193, 314, 315 Bayezid II campaigns, 7, 139, 143n14, 191n3, 308–9 maps, 64 mosque, 145 naval alliance, 81 Vidin defences under, 335 Bayhan, 302 Baysi, 299 Bazi‘, 248–9 Beaufort see Leftro Beaufort, Capt Francis, 307 Beçe see Becse Beçkerek see Becskerek Becse (Beçe), 70, 71 Page 9 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Becskerek (Beçkerek), 70, 71, 76 Bedouin Arab Revolt, 24, 437, 441 threat posed by, 24 ‘Asir, 516 eastern Arabia, 250 hajj routes, 88, 93, 95, 98–9, 102–3, 110 Hijaz, 84 troops, 107, 247, 248 Beg de Habeleth, 472 (p.571) Béga, River, 70 Behram Dimişki, 76 Beja tribes, 231, 471, 473, 474 Bek, 523 Békés (Bekeş), 395, 396, 399, 400n52, 401 Bela Krajina, 423 Bela Palanka, 171n1 Belgradcık (Belgrad; Belogradchik) administration, 347, 348n84 cavalry, 342 civil buildings, 184 fortress, 171n1, 173, 181, 184, 333, 334–5 Belgrade agrarian regime, 349 fortress, 58, 178 frontier, 178, 331, 332 Janissary rebellion, 241, 242, 244, 349 maps, 64, 178 road guide, 65 Serbian raids, 178 shipbuilding, 58 treaty of (1739), 332 Belogradchik see Belgradcık Bender, 139, 140, 143, 157, 175 Bengi Ali, 315n17 Bennett, Alexander, 506n45 Bent, J. Theodore, 327n85 Berber, 371, 471, 482 Berberistan map, 472 occupation background, 374–5 discussion, 382–3 historical and archaeological evidence, 375–82, 377 Berkovitsa, 335, 347 Berlin, treaty of, 279, 283, 284, 301 Berlioux, E. F., 520, 532 Bespirim see Veszprém Bessarabia see Bucak Page 10 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index beylerbeyi, role of, 20 beylerbeyilik provinces, 20 beyliks, 5, 12, 13 Beypazarı,315n17 Beyşehir administration, 312, 323, 324 hinterland, 311, 324n66 Karamanid control, 317, 318 Via Sebaste, 320, 322 Bihać(Bihke) administration, 173, 179 Bosnian War, 409 ethnography, 413, 414, 415, 418 fortress, 173, 177, 179–80, 406 Kapitanova Kula, 418n19 mosque, 412 public buildings, 184 Bihaćka Krajina see Krajina Project Bihke see Bihać Bilad al-Sudan, 499 Bilhorod-Dnistrovs’kyi see Akkerman bina emini, 161 Binder, Franz, 519, 535 Bingöl Dağ, 211, 213, 214 Bithynia, 12 Bitlis administration, 214, 215, 220, 221, 222–3 Evliya Çelebi on, 123 fortress, 220–1 garrison, 223 principality, 213, 214, 221, 222–3, 224 Bitlis pass, 213, 215, 220 Bitylos, 130 Bizerta, 500 Bizzelli, 521n65 Black Sea canal plan, 61–2 frontier description, 137–40, 138, 141–2, 169–70 historical overview, 5, 9 late military buildings, 182–3 see also Akkerman trade, 268 Blake, Robert, 499 Bobofça see Babócsa Boeotia, pottery, 165, 167 Boğazhisar see Kumkale Böğürdelen see Šabac Bon, A., 121, 133 Page 11 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Bongo, 516, 521, 537 Bonne, Rigobert, 472 Bornu, 8, 26, 499, 514 Borozdin, N., 265 Bosna, 173, 178 Bosnia administration, 179–80 civil buildings, 184 ethnography, 413, 414, 429 house types, 418 pottery, 421–3, 422, 424–5, 426, 427, 429 frontier, 240, 355, 403–4 soldiers from, 374, 375 see also Krajina Project Bosnian War (1992–5), 409, 410, 415, 421 Boz Ulus, 219 Bracewell, Wendy, 403 Bragadino, Marcantonio, 44–6 Braila (İbrail), 175, 182 Braudel, Fernand, 493, 494, 497 Brčko, 427 Brekovica, 412 Brezova Kosa, pottery, 422, 424, 425 Brezova Polje, 426 bridges, 58, 60, 179, 180, 184 Bringa, Tone, 413 Bromley, J. S., 507 Bruce, J., 483–5 Brza Palanka, 333, 334, 335n14 (p.572) Bucak (Bessarabia), 139, 141, 142 Bucovina, 175 Buda (Budin; Budun) administration, 58, 63, 71 fortress defences, 68–70 depiction of, 51–3, 52 garrison, 385, 466, 467 Ottoman capture, 58, 67 pottery, 386, 389, 392, 395, 396, 398, 399, 401 gunpowder works, 75 map commissioned, 66 shipbuilding, 58 siege (1684), 71 Budin/Budun see Buda building registers, 153–4, 155, 162, 195 buildings, civil, 183–5, 186–7; see also houses Buják (Buyak), 68 Bulgari, Count Pietro, 264 Page 12 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Bulgaria Christian risings, 345, 346n77 frontier, 5, 10, 173, 333 Second Bulgarian Empire, 331 voynuks, 185 Bunsuz, 321 Burchett, Josiah, 506n45 Burckhardt, John Lewis, on ‘Aqaba, 104, 107 Berberistan, 374, 375, 382, 383 Qasr Ibrim, 17 Suakin, 484 Bursa, 12, 14, 141, 315 Butrint (Butrinto), 236, 243, 256 Buyak see Buják Büyük Süleyman Pasha, 85 Bužim (Büzin), 410n12 Byron, Lord, 238 Byzantine empire, 5, 12, 14, 16 Caballa see Gevele Cafer Pasha, 353, 360, 366, 367, 368–9, 370 Cairo mamluk massacre, 512 military academy, 18 Ottoman conquest, 81 plague, 104 pottery, 486 slave market, 26, 514 trade/caravans, 98–9, 101, 103, 104, 483, 486, 491, 514 Calichiopoulo-Manzaro, Giorgio, 264 Çam see Tsamides Campo Formio, treaty of, 253 Çanad see Csanád Çanakkale, 196 canals, 16, 60–2, 216; see also Suez Canal Candia see Irakleion Canik, 244 Cankerman see Özi Cankurtalan see Korkmaz cannon warfare, 322–3, 328–9 Cantemir, Dimitri, 23 Cape Matapan (Tainaron), 114, 118, 131 Capodistria, Count Antonio, 262n45 Capodistria, Ioannis, 262n45 Cappadocia, 311 caravans ‘Aqaba, 97–104 Arabia, 84, 94, 250, 290 Page 13 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Jordan, 440, 446 north-east Africa, 226, 471, 482, 514, 522; see also zaribas trans-Sahara, 26, 498 caravanserais (hans) Akkerman, 145 ‘Aqaba, 97–101 Arduan Island, 379 Bitlis, 220, 221 Damascus, 87 Ljubuski, 179 Rahva, 220 Syrian hajj route, 87 Van, 216 Zıvarık Han, 326 Carroll, Lynda, 544–5 Casati, Gaetano, 522–3 Caspian Sea, 61, 137 castles, Bihaćka Krajina, 406–12, 407, 408, 411 Castro, Juan de, 476, 482, 483, 484, 491 Caucasia frontier First World War, 436 historical overview, 8, 17, 18, 19 maintaining, 358–60, 369 mamluks, 17, 512 Shi‘i challenge, 278 slavery, 18, 139, 513 see also Circassia Cazin, 413, 414, 418 Celali rebellions, 25, 463 Cem, 7 Cephalonia, 253, 261, 263, 504 Cerigo see Cythera Cerigotto see Anticythera Çerna Reka, 343n62, 347 Çeşme, 207 Çetin see Cetingrad Çetince see Cetinje Cetingrad (Çetin), 177, 410, 414 Cetinje (Çetince), 173 Ceuta, 497, 505 Cevdet, Ahmet, 123 Cevdet Pasha, 178–9, 180, 182, 183–4, 185, 186, 187 Chabot, General, 255 Chatelain, Henri, 472 Chiarello, Giovanni Battista, map by, 37–9, 38, 42, 46 (p.573) China pottery from Page 14 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Akkerman, 165–7, 166 Hungary, 388, 392, 398n39 Suakin, 486–8, 487 trade, 484, 486–8 Chittick, N., 474 Christians in administration, 21, 22 civil buildings erected for, 184–5, 187 ethnography in Bihaćka Krajina, 414–15 independence movements/rebellions, 10, 241, 242 in military Anatolia, 13 Balkans, 185, 187 Hapsburg frontier, 464–8 Hungary, 25, 385, 400 Vidin/Danube, 338, 343–4, 345, 350–1 missionaries, 286 Ottoman policy, 7, 13–14, 242–3 The Chronicle of the Morea, 113, 118–21 churches, 14, 183, 184, 187 Bihaćka Krajina, 412 Ioannina, 242 Vidin, 346 Cibova see Areopolis Cic, 518 Ciğirdelen see Párkány Çıldır Eyaleti, 23 Cilicia, 314 Cilicia Trachea (Rough Cilicia), 311, 312, 314, 318, 321, 327 Circassia mamluks, 17 slaves, 26, 513 soldiers/officials from, 18, 371, 375, 516 Ciriaco of Ancona, 130 civil officials, 235 Clausewitz, Carl von, 439 Clissa (Klis), conquest of, 48–50, 49 coffee, 290, 291, 294, 484, 512, 513 coinage, 21, 167 Committee of Union and Progress, 10, 24, 279, 436 Compana, 129 Congreve, W., 505 Constantine Copronymous, 183n56 Constantinople, fall of, 5, 14; see also Istanbul Constantinople Protocol, 279 Corfu campaign, 1715, 242n21 fortresses, 117 Page 15 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Napoleonic Wars, 253, 255–6, 269 diplomatic problems, 261 local interests versus imperial priorities, 262, 263, 264–6, 267 Corinth, Gulf of see Lepanto Coron see Koroni Coronelli, Vincenzo map by, 44–6, 45, 46, 47 plan by, 128, 129, 133 corsairs Adriatic, 64, 253, 266, 268–9 north Africa, 26, 355 maritime frontier formed by, 495, 497, 498, 503–4; collective microhistory, 503–6; fleet, 501–2 see also Piri Reis Cossacks, 139–40, 141, 142, 150–1, 156, 167 Crete, 9, 18, 114, 118, 191, 348n87 Crikvenica, 427 Crimea administration, 21, 137–9 annexation by Russia, 153 coinage, 167 frontier, 5, 9 Khanate, 15, 21, 137–9, 140, 141, 142, 358 pottery production, 164 trade, 141 Crimean War, 247–8, 278 Crişul Repede see Sebes-Körös Croatia, frontier, 173, 175, 176–7, 177, 355; see also Krajina Project Crofts, Lionel, 503–4 Csallóköz, 74 Csanád (Çanad), 70, 71 Csepel Island, 68 Çukur-i Saad, 361 Cyprus, 40, 44, 183, 203n29, 312, 314 Cyprus War (1570–3), 117 Cythera (Cerigo), 117, 253 Dab‘a, 90 Dali‘, Amir of, 298–9, 300, 302–4 Dalmatia, 173, 179, 256 Damascus, 81, 85, 87, 286, 437 Danube fleet organisation, 58–60, 336–7 sailors, 343n62, 343n64, 344, 348, 465 seat of admiral, 333, 337 shipbuilding, 337 frontier Page 16 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index administration, 22, 347–50 geography and environment, 58–60, 59, 66–78, 69 human resources, 337–46 palanka forts, 173, 175, 181–2, 186, 335 Vidin, role of, 331, 333–7, 334, 350–1 trade, 140, 141 Dar al-‘Ahd, 494 Dar al-Hamra’, 90 Dar al-Harb, 12, 494 Dar al-Islam, 12, 225, 233, 494 Dar‘a, 109 Dardanelles, 189, 190, 191–3, 207n42; see also Kumkale; Seddülbahir (p.574) Darfur administration, 232, 233, 234, 371 trade, 485, 513, 514, 521 Darija, 300, 300n41 Debar (Debre-i Bala), 180 Debrecen, 78 deforestation, 72–6, 79 Delatte, A., 131 Delgo, 378 Deliorman see Dobrudja Demir Kapu see Iron Gates Demirhisar see Sidirokastro Demiševci, pottery, 422 Derbend, 358, 358n2, 359 derbendci troops, 343, 466 Develi Karahisar, 317 Dhat al-Hajj, 88 Diffi Dunni, 380 Dinaric Alps, 176 Dinka alliances, 518 cattle raids, 519, 537 Rumbek attacked by, 523–4 settlements/houses, 520, 522–3, 525 shrine, 527 slaves, 514–15, 516, 528, 529, 537 Diregel see Drégely Diwaniya, 277 Diyala, 277 Diyar-i Yunan, 312 Diyarbakır (Amid) Akkoyunlu tribal confederation, 5 Evliya Çelebi on, 123 province, 214, 215, 273, 274, 348n87, 368 trade, 215, 217 Djibouti, 291 Page 17 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Dnieper, river, 139, 140, 141, 150, 156 Dniester, river, 139, 140, 141, 145, 157, 182 Dniprovske-2, 164 Doboj, 427 Dobono, Andrea, 519 Dobrudja (Deliorman), 173, 244 Doha, 248n40 Dolenja Vas, 427 Don, river, 60–1, 62, 140, 150 Don John of Austria, 117 Dongola, 18, 228, 376, 377, 514, 515n29 Doria, Andrea, 116 Drégely (Diregel), 68 Drežnik Grad (Dreznik), 177 Drina, river, 173, 179 Drury, W. B., 109–10 Druze, 9 Dubica, 177 Dubrovnik (Ragusa) couriers, 33n3 map, 64 republic, 173, 256, 257, 258, 263, 268 Dulkadirid Türkmen, 317 Dumas, A., 238 Dunaföldvár (Fedvar), 400 Dunn, Samuel, 472 Duquesne, Admiral, 503 Dürrühand Hatun, 327 Düş Hasan Pasha, 180n36 economy Hapsburg frontier, 455–68 historical overview, 25–6 see also trade Ecsed, 76 Eger (Eğri) see Agria Eğriboz see Euobea Eğridir, 315n17 Egypt administration, 234, 509–12 1517–55, 226–7 1555–89, 227–8 1589–1820, 228–31 1820–1914, 231–3 canal, 61 economy, 25 French occupation, 231, 255–6, 512 frontier garrison, 511–12 historical overview, 7, 9, 10, 11, 16, 17, 18 Page 18 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index trade coffee, 512, 513 slave trade, 512–14, 517, 518 Suakin, 482, 484, 485, 486 see also hajj, Egyptian route el-Ghoury see Qansuh al-Ghawri Eleşkirt, 211, 219 Elliot, Sir Henry, 295, 296 Emin Pasha, 520–1, 522–3, 524, 535 Emirşah Bey, 326, 327 environment, frontier zones, 19, 58–66, 72–8 Epirus, 117, 236, 240, 241 Equatoria, 232, 233, 536 Erçek valley, 213 Erçin see Ercsi Erciş 215, 217–18, 223 Ercsi (Erçin), 400 Érd-Hamzsabég (Hamzabey), fortress defences, 68 pottery, 395, 397, 398, 399, 400, 401 Ereğli, 312 Eretnids, 12, 325 Eritrea, 234, 291 Erlau see Agria Erlisman, Consul, 506, 507 Ermenek, 318, 329n91 Érsekújvár see Nové Zámky Erzurum garrison, 24, 356–8, 362–4 province, 214, 217, 218, 219 (p.575) timars, 348n87 Estergon see Esztergom Eszék (Osijek), 76, 183 Esztergom (Estergon), 58, 67, 68, 386, 467 Ete, 390n18 Ethiopia frontier, 16, 234, 472 1517–55, 227, 471 1555–89, 227–8 1589–1820, 230 1820–1914, 232, 233 trade slave trade, 513, 514, 516, 521 Suakin, 471, 482, 483, 484, 485, 486 ethnicity, Hungary, 399–401 ethnography, Bihaćka Krajina, 406, 413–15 discussion, 429–30 funerary stelae, 418, 419 pottery making, 421–9, 422, 428 Page 19 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index rural houses, 415–18, 416, 417 watermills, 419–21, 420 Euboea (Eğriboz; Negroponte), 318n31 Euphrates, river, 58, 64, 283 Evangelon Portus, 471 Evliya Çelebi accompanies Hüseyin Pasha, 94 on Akkerman, 143n14, 145, 148, 149, 149n25, 182n48 on Berberistan, 373–4, 375, 376, 381–2, 383 Arduan Island, 379 Hannek, 380–1 Jabal Sese, 378 Kada Musa, 379 Sai, 376 Simit Island, 380 Tinare, 378 on Bihke, 180n36 on Dardanelles forts, 192 on Erciş 217n23 on Fethülislam, 181 on fortresses, 53n39, 171n1, 171n3 on Mani, 114, 122–9, 134–5 itinerary, 125 on raiding, 458–9 on Vidin, 339, 342 expansion, ideology of, 11–16 eyalet, 20 Fadl family, 290, 297 ‘Abdallah, 297–8 Fadl b. Muhsin, 293–4, 295, 296, 297–8 Fakhr al-Din Ma‘an, 9 fakis, 230, 233 Falkenhayn, General von, 444 Famagusta, 44 Faroqhi, Suraiya, 124 Fashoda Incident, 524 Fassu‘a (Zahr al-‘Aqaba), 90, 92, 93 Faw, 286–7 Faysal, Amir, 439 Fazıl Ahmed Köprülüzade, 124, 134 Fearne, Captain, 504 Fedvar see Dunaföldvár Fehérvár (İstolni Belgrad), 70 Fejér, 399–400 Fer, Nicholas de, map by, 190, 196–7 Ferdinand I, 51, 65–6 Ferhad Bey, 355 Ferid Uğur, M., 320 Fethülislam see Kladovo Page 20 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Fez, 7 Fezzan, 499, 514 Fifteen Years War, 399 filurci troops, 343, 344 Filurdin see Florentin First World War ‘Aqaba, 110 Arabia, 11; see also Arab Revolt, archaeology of Gallipoli, 202, 436 jihad, 13 north-east Africa, 233, 234 fleet see navy Florentin (Filurdin), 333, 334, 335 Floret of Hainhaut, 120 Förster, E. H., plan by, 158, 159 fortification, Van region, 16th century, 216–21, 223 fortresses, 19–20 Balkans, Bihaćka Krajina, 406–12, 407, 408, 411 Black Sea, 140, 156–7, 545; see also Akkerman Corfu, 264–6 Danube, geography and environment, 58–60, 59, 67–78; see also Vidin Dardanelles see Kumkale; Seddülbahir Egyptian hajj route see ‘Aqaba Hapsburg frontier building and maintaining, 456–8 geography and environment, 57–79, 59, 69 Hungary garrisons and local population, 385–401 geography and environment, 57–79, 59, 69 Iraq, 286–7 Jordan see Ma‘an Karamanid lands, 319–23, 328, 328–9 Mani, 113–35, 116, 125, 128, 131, 134 mapping of, 31–2, 55 by Europeans, 32–3, 50, 66; Agria, 39–40; Clissa, 48–50, 49; Győr, 35–7, 36; Hatvan, 33–5, 34; Kelefa, 128; Nayhaysel, 37–9, 38; Nicosia, 40–4, 41, 43; Santa Maura, 44–6, 45, 46, 47 by Ottomans, 50–5, 52, 54 north Africa, 499–500 north-east Africa, 226–7, 228, 230–1, 376–83, 377; Page 21 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index see also Suakin Syrian hajj route, 87, 93–4 16th century, 87–90, 88, 89 18th century, 90–1, 91, 92, 93 (p.576) technology, advances in, 156–60 Van region, 216, 217–18, 219–21, 223 see also palanka forts Foscolo, Leonardo, 48 France alliances with, 14, 207n42 Arab revolt, 435–6, 437, 441 Cem held in, 7 foreign policy Arabian Gulf/Red Sea, 16, 291, 301 Egypt and Sudan, 231, 255–6, 512, 524 Ionian Islands, 117, 261 north African frontier, 498, 500, 502 Mexican war, 517 Napoleonic Wars see Adriatic frontier, Napoleonic Wars frankincense, 484, 491 Frere, Sir Bartle, 295, 296 Frontier Thesis, 455 frontiers environment and geography, Ottoman understanding of, 58–66 formation and maintenance, 354–8 discussion, 369–70 Erzurum, 362–4 Kars, 360–2 Shirvan Province, 358–60 Tabriz, 364–9 geography landscape and climate, 72–8 rivers, terrain and fortifications, 66–71 historical overview, 4–11, 6 administration, 20–3 economy, 25–6 fortifications, 19–20 frontiers, frontiersmen and the centre, 16–18 ideology of expansion, 11–16 society, 23–5 nature of, 31, 493–5 study and understanding of, 539–50 Fuat Pasha, 178, 180 funerary stelae, 418, 419, 429 Funj Sultanate frontier with archaeology of, 373–4, 375–83, 377 maps, 472 Ottoman policies 1517–1914, 226, 227, 228, 230, 231–2, 234 Page 22 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index study of, 27 trade, 26, 471, 491, 513, 514–15, 516 Gaïtsa (Gaçiça), 126 Galip Efendi, 261n35 Galla (Omoro) tribes, 227, 521 Gallipoli, 202, 436, 462 Garbelia, 126 Gazi Mihal, 342n52 gazi thesis, 12–14, 494–5, 539, 542 Gedik Ahmed Pasha, 321 Gelviye, 347 Genghis Khan, 15, 137–9 Genoa, 5, 116, 142 geography, Ottoman understanding of, 58–66 George of Trebizond, 16 Georgia allies, 358 frontier, 5, 17, 213 governors, 23 mamluks, 17, 273, 512 slaves, 513 Geraki, 119 Germany engineers, 179, 284, 435 First World War, 435, 436n15, 444 maps, 33–7, 34, 36 mercenaries, 78 see also Prussia Germiyanids, 12, 13, 25 Gessi, Romolo, 520, 522–3 Gevele (Caballa), 309, 311, 317, 319–23, 328–9 Ghalib b. Muhammad, 292 Ghar al-Milh (Porto Farina), 499–500 Ghattas, 520, 521n65, 532 ghuzz, 374, 375 Gibraltar, 504, 505, 506n45 Girays, 15 GIS, 547–50 Giurgiu (Yergöği), 333, 337, 344 Giymir see Kirmir Glavica, 426 Göçü, 324 Goethe, J. W., 238 gold trade, 225, 230, 483, 484, 491, 515 Goli Vrh, 426 Golubac (Güvercinlik), 58 Gondokoro, 232, 515, 535 gönüllü troops, 460–1, 511 Gordon, General Charles, 232, 473 Page 23 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Gornji Košinj, 426 Gornji Lađevac, 425 Gornji Šiljkovača, 414, 415, 416, 419, 420, 421 Goslett, Capt. Raymond, photograph by, 111 Grabovac, house plan, 417 Gračanica, 427 Gradiška (Gradişka), 177, 456, 457 Graham, Sir Richard, Viscount Preston, 503–4 Gramacor (Grand Major), 179 Grand Magne (Maina; Máïni; Mégali Máïni; Palaiá Máïni), identification, 113 discussion, 132, 133 evidence, 118, 119–20, 121–2, 124, 130, 132 see also Kelefa Granville, Earl, 295 Gray, Richard, 536–7 (p.577) Great Arab Revolt Project see Arab Revolt, archaeology of Great Britain, foreign policy Arabia, 16, 247, 248, 250 Arab Revolt, 435–6, 437, 450–1 Ionian Islands, 237, 243, 255, 256, 261, 262, 263, 269 Iraq, 271, 279, 280–1, 282–7 north African frontier, 498, 502–7, 508 Sudan, 524 Yemen, 289–91, 292, 293–304 see also India, foreign policy Great Hungarian Plain, 72–3 Greece administration 1787–1820, 236–45, 237 frontiers, 5, 10, 114–18 Ottoman history, study of, 546–7 revolt, 239, 241–2 slaves, 513 see also Adriatic frontier, Napoleonic Wars; Ionian Islands; Mani; Rumeli Greenlaw, J.-P., 469, 471, 473–4, 477, 479, 480 Gregory, Patriarch, 264 Grelot, W., 196 Grenzer troops, 403 Grimani, Francesco, 129, 132–3 Gritsena, 119 Groote, Alexander de, 37n12 Gülhane Decree, 10 Güllü, 127 gum arabic, 484, 485, 491 gunpowder works Hungary, 75–6, 79 Estergon, 68 Van region, 223 Ahlat, 220 Erciş, 217–18 Page 24 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Van, 216 Güvercinlik see Golubac Guwayra, 434 Gvozd (Vrginmost), pottery, 423, 427, 428, 428, 429 Győr (Iavarinum; Yanık) defences, 76 garrison, 467 Ottomans capture, 71 siege map, 35–7, 36, 46 timber supply, 74 Gyula, 76 Habelthi, kingdom of, 472 Habeş, 17, 22, 227–31, 469, 471, 472 Hacı İsmail, 179n35 Haçova (Mezőkeresztes), battle of, 64 Hadım Süleyman Pasha, 381 Hadiyya, 90 Hadradib, 473, 474, 491 Hafir-i Kabir, 379–80, 381 Hafir-i Sagir, 381 Hafiz Efendi, 179 hajj African route, 226, 227, 483 Dardanelles route, 191 Egyptian route, ‘Aqaba, 95–110, 96 Syrian route, 86–7, 440, 446 development: Phase One (16th century), 87–90, 88, 89; Phase Two (18th century), 90–1, 91, 92, 93 discussion, 93–4 threats to Bedouin, 88, 93, 95, 98–9, 102–3, 110 Portugal, 61 Hakkari principality, 213, 214, 217n21, 222, 223 Halil Bey, 63 Halil Bey (Turgutoğlu commander), 326 Halil Hamid Pasha, 176 hamams Akkerman, 143, 147, 148–9, 152, 170 Balkans, 183 Seddülbahir, 149n28, 191, 205 Suakin, 479 Hamidi, 322–3 Hamilton, William J., 320 Hamza, 324 Hamza Bey, 65–6 Hamzabey see Érd-Hamzsabég Hanekiyye, 53 Hannek, 228, 374, 376, 380–1, 382 hans see caravanserais Page 25 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Hapsburgs, frontier with defining and mapping European, 33–40, 34, 36, 38 Ottoman, 51–3, 51 geography and environment, 57 discussion, 78–9 landscape and climate, 72–8 Ottoman understanding of, 58–66, 59 river systems, 66–71, 69 historical overview, 2, 7, 9, 13, 17–18 military service and material gain, 455–68 palanka forts, 173, 175–7, 177, 178–81, 186, 457–8, 467 taxation, 21 Vidin, 331–2, 336 see also Krajina Project haracgüzar provinces, 21 Haram, 58 Harameyn, 181 Harar, 16, 227, 232 Hart, Basil Liddell, 438 Hasa, 248n40 Hasan Agha, 104 Hasan Bey, 381 Hashemite army, 437, 441, 450–1 (p.578) hass lands Hapsburg frontier, 456 Van region, 222 Vidin region, 338, 339, 340, 347, 348 Hatvan, 33–5, 34, 68 Hawshabi, 290, 295, 297, 299, 300, 302 Haydud troops, 403 Herzegovina (Hersek) civil buildings, 184, 187 Montenegrin attacks, 178 palanka, 180 pottery, 421 sancak, 173 tower houses, 418 see also Krajina Project Hetherington, P., 120 Heywood, Colin, 13 Hijaz administration, 21, 249 Arab Revolt, 437 conquest and control, 7, 82, 83, 84–6, 93–4 Hijaz railway, 10, 94 Great Arab Revolt Project, 434–5, 434, 437, 449, 451n36 Ma‘an, 440–1, 448 Wadi Rutm station, 444–7, 445 Page 26 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Hilla, 64, 277 Himara, 243 Hınıs, 213, 214–15, 217, 218, 219, 223 Hinkel, F. W., 469, 474, 476 Hırşova (Hiršovo), 175 historical archaeology, 543–4 Hobman, Captain Hesketh, 507 Hofkriegsrat, 176 Holden, Edward, 506 Holıvnik see Turnu Măgurele Hóllokő (Holloka), 68 Holy League, war with, 8, 77, 331, 333, 336, 345 Hóman, Bálint, 72 Horta, 324 Hoşap, 211, 214 houses Bihaćka Krajina, 415–18, 416, 417 Rumbek, 522–3, 528–9, 532 Suakin, 477–81, 478, 480 hüccet, 152, 154, 161 Hudayda, 294, 300 Hudson, Charles, 506 Hufuf, 82 Hugo, Victor, 238 hükümet provinces, 21 Humada, 304n58 Humar, 299 Humaydi, 302 Hummayma, 86 Hungarian Diet 1563, 74 Hungarian Diet 1600, 74 Hungarian language, 65 Hungary administration, 21, 25, 183, 355 frontier defences, 331, 333 environment, 57; discussion, 78–9; geography, Ottoman understanding of, 58–66, 59; landscape and climate, 72–8; river systems, terrain and fortifications, 66–71, 69 garrisons, 466, 467 historical overview, 7, 9, 14 garrisons and local population, archaeology of background, 385–6 pottery: analysis, 395–7; classification, 388–94, 388, 389, 391, 393, 394; discussion, 398–401; new ceramic types, 386–8 Page 27 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index kingdom of, 67 Ottoman history, study of, 545–6 see also Hapsburgs, frontier with Hunkar İskelesi treaty, 207n42 Hurmuz, straits of, 82 Husayn Kamil, 233 Hüseyin Ali Pasha, 94, 124, 126, 127, 134 Hüseyin Bey, 265n65 Hüseyin Efendi, Yüzbaşı, 180, 184 Hüseyin Kazım Kadri, 172n4 Hüsrev Pasha, 220, 221, 512 Iavarinum see Győr Ibb, 294 Ibn Battuta, 482 Ibn Iyas, 97–8 Ibn Janbalan, 374, 382 İbn Kemal, 322 Ibn Rashid, 285, 286 Ibn Sa‘id, 471 Ibn Sa‘ud, 248n40, 250, 285 Ibrahim (sultan), 191 İbrahim II (Karamanoğlu), 308, 315–16, 317, 318, 321 İbrail see Braila Ibrim, 227, 228, 230, 231, 374, 378; see also Qasr Ibrim İçel, 308, 312, 325n76, 327n85, 329n91 iconography, maps, 33, 35, 39, 40, 48 Idrisi, 11 Ilgın, 312, 323, 326 İman Kulu, 218–19 imperial diploma, 258–60, 263, 268 India foreign policy eastern Arabia, 247 Iraq, 271, 273, 280–1, 283–4 Yemen, 290, 293–304 pilgrims, 61, 280, 283 slave trade, 514 specialist units (Arab Revolt), 437 trade, 82, 231, 471, 483, 484, 486, 514 Indian Ocean, 8, 61, 83, 226 Indonesia, 8 (p.579) Inebahtı, siege of, 51n37 İnsuyu, 323, 324 Ioannina massacre, 345n71 Tepedelenli Ali Pasha, 236, 239, 241, 242, 243–4, 245 Ionian Islands banner, 260–1, 260 centre-periphery relations, 236–7, 243 Page 28 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Napoleonic Wars background, 253–6 diplomatic problems, 256–61 discussion, 269–70 local interests versus imperial priorities, 262–9 Venetian control, 114, 117 Irakleion (Candia), 114, 117, 118, 124, 191 Iran frontier 19th century, 271, 273, 278–82 historical overview, 2, 7, 8, 9, 13, 14 pottery, 392, 398n39 trade, 223 see also Safavids, frontier with Iranian Constitutional Revolution, 280 Iraq administration, 19th-century background, 271–4, 272 British penetration, 282–7 conflicts in provincial administration, 275–6 Iranian (Shi‘i) challenge, 278–82 reform, dilemmas of, 275 tribal unrest, 276–8 frontier, historical overview, 2, 7, 8 Wahhabi raids, 85, 86 Iron Gates (Demir Kapu), 173 irrigation, 230, 275, 368, 517 İsa of İstrumca, 324 Isakce (İsakçı), 175, 348 Isauria, 307, 308, 309, 311, 312 Isfahan, 214 İsfirlik see Svrljig İshak Bey, 316–18, 321, 323 Iskandar Beg, 360 İşkodra see Shkodër Islam Islamisation, 14, 15 Bihaćka Krajina, 413–14, 415 in military, 341, 342, 466–7 north-east Africa, 233 Mehmet Ali’s attitude to, 241–3 see also gazi thesis; hajj; medreses; mosques; Shi‘i; Sufism İsmail (town) see Izmail Isma‘il (khedive), 232, 233, 471 İsmail Pasha (building overseer), 182 İspozi see Spuž Israel, Daniel, 505–6, 507 Istanbul Aya Sofya (Haghia Sophia), 14, 181 Page 29 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index food supply, 141 gunpowder shipped to, 216, 218, 223 Kağıthane, 183, 186 Patrona Rebellion, 25 royal architects, 195 slavery, 26, 521 Tophane-i Amire, 181 İstineli see Stijena İstolni Belgrad see Fehérvár; Székesfehérvár İstrumca (Strumitza), 324 István Bocskai, 339 Italo-Turkish War, 109–10 Italy in Arabian Gulf/Red Sea, 16, 291, 301 Cem held in, 7 map-makers, 33, 37, 42 mercenaries, 78 see also Genoa;Venice Ithaca, 253 Ivan IV (Tsar), 8, 60 ivory trade Arduan Island, 379 Suakin, 483, 484–5 Sudan, 509, 517, 518–19, 525, 537 Izačić, Ivan, 412 Izmail (İsmail), 175, 182, 186 Izmir, 268 İzmit, 61, 62 Iznik, pottery, 164, 165, 166, 387, 392, 398n39, 548 İzvornik see Zvornik Jabal al-Batra, 440 Jabal Samna, 441 Jabal Sese, 378–9 Jabal Shammer confederation, 285 Jagei, 523 Jalilis, 273 jallaba, 522 James II, 506 Janissaries at Algiers, 505 Belgrade, 241, 242 Berberistan, 374–5 Egypt, 511, 512 Erciş, 217 Erzurum, 363–4 Hungary, 398 Nikopol, 342 Qal‘at Sai, 228 Page 30 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Qasr Ibrim, 17, 230 Shirvan, 359 Vidin, 335, 340–1, 344, 345, 349, 350, 351 economic opportunities, Hapsburg frontier, 458, 460–1, 462, 463–4, 467 historical overview, 9, 17, 25 Jaqeli family, 23 (p.580) Java, 11 Javurdzija, Fehim, 422 Jazirat Fara‘un, 95 Jazirat Tombus, 381 Jeddah administration, 228 governor, 94 trade, 26, 230–1, 471, 483, 484, 485, 486 Jerković, Stjepan (Stipo), 424 Jerusalem, 491 Jews, 14, 242, 345, 346n75, 351 Jireček, Constantin Josef, 171n3 Joachim of Brandenburg, 78 Johannes, Emperor, 521 Johi, 426 John Zapolya, King, 387 Jontić family potter, 428 Jordan see Arab Revolt, archaeology of Joseph II, 176 Josipdol, 427 Juba, 515 Juba, river, 232 Julfa, 215 Junker, W., 485 Jur, 516, 519, 522, 537 Kaçanik, 53n39 Kada Musa, 379, 381 kadıs Akkerman, 152, 161, 163 Berberistan, 376 Hapsburg frontier, 462, 463 north-east Africa, 233 Van region, 221 Vidin, 22, 332, 337, 346, 350 Kajtezi, 418 Kakhetia, kingdom of, 358, 360 Kalamata, 124, 126, 132 Kalamei see Lol Nhom Kale-i Sultaniye, 192 Kaliya see Karea Kalmeta, A., 424 Kaluđerovac, pottery, 421, 422, 426 Kanije see Kanizsa Page 31 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Kanizsa (Kanije), 51n37, 66, 71, 76, 458–9, 467 Kaposvár, 76 Kara (‘Black’) Ahmed Pasha, 70 Kara Mahmud Pasha, Buşatlı, 238n6, 240, 243 Karabagh, 356 Karahisar (Afyonkarahisar; Karahisar-ı Sahip), 318, 319 Karaman (city) see Larende Karaman (dynasty founder), 321 Karaman island, 84 Karamanid Principality beylik, 5, 12, 16–17 conquest of, 307–11 discussion, 328–9 geography, 310, 311–14, 313 Gevele and Konya, 319–23 hinterland, 323–8 relations and succession struggle 1464–8, 314–16 succession, Ottoman interference, 316–19 history of, 25 Karasi, 13 Karataş, 327 Karayazı Düzü, 219 Karbala, 64, 86, 277, 280 Kardamyli see Mila Karea (Kaliya), 127 Karioupolis (Kıryopoli), 126 Karlovac, 406, 426 Karlowitz, treaty of, 13, 15, 175, 335, 341, 347 Karpat, Kemal H. and Zens, Robert W., 540 Kars, 356, 358, 360–2 kaşifs, Sudan, 374, 375, 378, 380, 382 Kasım Bey, 60 Kašpar, 426, 428 Kasr-ı Şirin, 53, 370 Kassala, 485 Kastah-ı Fireng, 132 Kastamonu, 154 Kastania (Kastaniya), 126 Kastellesua, 185 Katsafados, Takis, 120 Kauffer, François, plan by, 158, 159, 160 Kavo Grosso, 132 Kaya Pasha Limanı see Porto Kaio kaymakams, 247, 376 Kayseri, 12, 312, 315n17, 317, 318 Kazan, 8 Kazdağlı household, 512 Kazimayn, 280 Kefe, 141, 151, 169 Page 32 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Keira sultanate, 513, 514 Kejang, 524n76 Kelapish tribe, 381 Kelefa, fortress discussion, 132–5, 134 documentary evidence, 118–23, 127 other evidence, 129–32 strategic location, 113, 118 Venetian plans, 128 Keli, 127 Kerch see Kerş Keria, 127 Kerrarish nomads, 378 Kerş (Kerch), 140 keşf processes, 151–2, 154, 161, 162 Khalil Bey, 103 Khalyl, 104 Kharaba, 302 Kharja, 514 Khartoum administration, 234 slave trade, 515, 518, 519, 520, 521, 532 (p.581) Technical College, 473 tombs, 371, 372 Khayir Bey al-‘Ala’i (Khayirbak Mi‘mar), 97, 98 Khoy, 215, 216 Khurshid Efendi, 477–9 Khushqadam, al-Zahir, 317 Kiev, 64 kılavuzan, 65 Kılburun (Kinburn), 140, 156 Kilia (Kili), 58, 139, 140, 168, 175, 182 Kılıç Arslan II, 13 Kilis, 355 Kilitbahir, 192, 196 kilns, 421, 423, 424, 427, 428 Kinburn see Kılburun Kirkham, General, 521 Kirkuk (Şehrizor), 273 Kirmir (Giymir), 324 Kırşehir, 312, 315n17 Kırya Nero see Kryoneri Kıryopoli see Karioupolis Kisterna, 119, 120 Kitchener, Lord, 471, 473, 524 Kivella, 126 Kızılbaş, 7 Kladovo (Fethülislam) administration, 347, 348 Page 33 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index fortress, 173, 178, 181–2, 333, 334, 335 garrison, 343n62 Kladuša, 407, 408, 413, 414 Klanac, 426 Klis see Clissa Knights of St John, 7, 114, 117 Kocaeli, 62 Koçi Bey Risalesi, 172 Koçkaya see Zengicek Kokka kingdom, 381 Kolaşin (Kolašin), 173, 180 Komárom, 74 Kommer, 381 Kompolje, 423, 427 Konjic (Koniça), 184 Kontkaşa see Zengicek Konya administration, 312, 323 conquest of, 309–11, 319, 321–2, 323, 327–9 hinterland, 311, 328 inscriptions, 13 Karamanid control, 311, 314n15, 314–15, 317, 318, 326, 327 mausoleum, 326 mint, 315n17, 318n34 Via Sebaste, 322, 325 Koppan see Törökkoppány Kör Hüseyin Bey, 376, 378, 379, 380, 382 Kordofan, 232, 485, 509, 513, 514, 521 Korianiç, 184 Korita, 180 Korkmaz see Adony Korkmaz (Cankurtalan), 68 Körmend, 460 Koroni (Coron; Corone; Koron) fortress, 116, 122, 126, 129, 130, 132 gazi, 127 Köse Hamza Bey, 318 Kosovo, 18 Kotor corridor, 213, 216–17, 279 Kougeas, Z., 120 Koutifari (Kutufari), 126 Krajina, military administrator, 179 Krajina Project background, 403–6, 404, 405 castles and fortifications, 406–12, 407, 408 discussion, 429–30 ethnography, 413–15 funerary stelae, 418, 419 pottery making, 421–9, 422, 428 Page 34 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index rural houses, 415–18, 416, 417 watermills, 419–21, 420 Krajna (Krayna), 348, 348n84 Kralje, 414 Kritovoulos, 193 Krivina, 347, 348n84 Krump, Theodoro, 514 Kruševac (Alacahisar), 58 Kryoneri (Kırya Nero), 126 Küçük Ali, 519, 520, 521n65 Küçük Kaynarca, treaty of, 9 kula (kule), 406, 415–18 Kulubnarti, 230 Kumkale (Boğazhisar), fortress historical background, 189, 190, 191–3 historical evidence, representational, 190, 196, 199, 200, 200, 207 oral history, 196n22 Kupa, river, 173 Kürd Mehmed Kapudan, 265n65 Kurdish principalities, 211, 213–15, 221, 223–4 Kurdish tribes, 273, 277, 278–9, 368 Kurdish troops, 244, 248, 249, 516 Kütahya, 12 pottery from, 165, 166 Kutlofçe see Montana Kutufari see Koutifari Kuwait, 10, 245, 246–8, 285, 287 Laborde, de, Léon, illustrations by, 104–7, 105, 106, 108 Lahj sultanate, 290, 293–4, 295–8, 299, 302, 515–16 village leaders, 293n16 Lahsa see al-Hasa Lakmat, 302 Lala Mustafa Pasha, 203n29, 361 Land Law 1858, 249, 276–7 Langdit, 527 (p.582) language, 23, 232, 239, 250 Lankada, 126 Larende (Karaman), 312–14, 315n17, 319, 326, 327 Larissa (Yenişehir), 241 Lawrence, T. E., 110, 431–2, 437–40, 449–50 Laz, 244 Leake, Capt. William Martin, 117, 121, 245n34 Lebanon, 9 Lefcochilo, Antonio Tomaso, 266n72 Leftini, 126 Leftro (Beaufort), 119, 121 Lejean, Guillaume, 472–3 Lemberg see Lviv Page 35 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Leontari (Londar), 124 Leopold II, 176 Lepanto (Navpaktos), 8, 33, 62, 117, 337 Leptini, 126 Lesin, plain of, 126 Leucada, 253 Levant companies, 502 Levkas, 44, 117, 243 Libya, 11 Liješevo, 425, 425n29 Lika, 426, 427 Lim, river, 173, 179 Linant, M., 108 Linardovich, 179 Linschoten, Jan Huygen Van, 472 Liplje, 426 Lippa (Lipova), 70, 71 Lisbon, 502n35, 503 Lithuania see Poland-Lithuania Little Nogay Horde, 139 Livorno, 502n35, 504 Ljubuški (Liyubuşka), 179 Lockwood, William G., 413 Lőcse, 78 Lokman ibn Seyyid Hüseyin, 50 Lol Nhom (Kalamei), 525–7, 526, 530–2, 531, 535–6 Lom see Polomiye Londar see Leontari Long War (1593–1606), 33, 71, 77, 78, 344 looting, 458–60, 466 Lotter, Conrad Tobias, 472 Lučka, stelae, 418, 419 Luís, Lázaro, 472 Lviv (Lemberg; Lwów), 141 Lycaonia, 311 Lynch company, 283, 286 Ma‘an archaeology first battle of, 447–9, 448, 451 trench fortress, 440–4, 442, 449 fort built, 88, 89 railway station, 435, 441, 448 Macedonia, 180 Mackay, Pierre, 124 Mackintosh, Major, photograph by, 111 Mada’in Salih, 90 Madat, 528 Madinat al-Kubra, 379 Mafraq see Tall Far‘un Page 36 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Maghribis, 107, 244 mahalle, 183 Mahas archaeology background, 371–6, 373 discussion, 382–3 sites described, 376–82, 377 mek, 376, 377, 378, 379, 381 sancak, 372, 373, 376, 382 Mahdiyya, 232–3, 380, 471, 523 Mahmud II, 13, 181, 241, 242, 273 Mahmud Pasha, 327 Mahmud Raif Efendi, 262, 265, 266 Mahmudi principality, 213–14, 222 Maina (Máïni; Mani; Mégali Máïni; Palaiá Máïni) see Grand Magne Mair, Alexander, map by, 35, 36, 46 Majorca, 506 Mala Kladuša, 420 Malacca (Malaka), 484, 486, 490 Malaga, 506 Malazgirt, 214, 217, 218–19 Malta consular records, 502n35 expedition against, 226 French occupation, 255 Knights Hospitaller, 114, 117 siege plan, 51n37, 64 slave merchant, 519 Malvar see Márévár Malzac, Alphonse de, 518–19, 525, 532, 535 mamluks historical overview, 7, 9, 10, 14–15, 17 in Anatolia, 308, 315, 317, 319 ‘Aqaba, 98, 103 Arabia, 81–2 Baghdad, 273 Berberistan, 374, 375 Egypt, 95–8, 226, 230, 509–11, 512–13 Sudan, 515n29 The Mani, 114–29, 125; see also Kelefa Manuel II Palaeologos, 120 maps, 1–2, 61, 63–6, 254; see also fortresses, mapping of Maqraq, 381–2 Marchand, Jean-Baptiste, 524 Márévár (Malvar), 392 Marj Dabiq, battle of, 81 Page 37 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Marjanovac, 414, 419, 420 Maros (Mures), river, 70 Marsigli, Luigi Ferdinando, 77, 459, 462, 464 palanka drawing, 172 (p.583) martolosan troops, 343, 403, 465–7 Massawa, 226, 230, 232, 375, 513, 515 Mas‘udi, 471 Matrakçı Nasuh, maps by, 53, 54, 55, 64–5 Matthias, King, 458 Mavga, 329n91 Mecca Amir of, 436 Ottoman protection of, 82, 93–4, 226 trade, 482, 491 Wahhabi missionaries, 86 see also hajj Mecseknádasd (Nadasd), 392 Međimurje see Muraköz Medina, 11, 82, 226, 436, 437, 451n36 Medley, Admiral Henry, 501 medreses Ahlat, 220 Akkerman, 145 Balkans, 183 Damascus, 87 Erciş, 218n32 Iraq, 281 Malazgirt, 218n35 Seddülbahir, 191, 205 Van,216 Meen Atol, 525–7, 526, 529–30, 530, 535–6 Mégali Máïni see Grand Magne Mehmed I, 315 Mehmed II conquests, 5, 14, 16 Black Sea, 137, 316 Crete, 191 Greece, 130, 318n31 Karaman, 311, 314, 315, 316, 318–19, 322, 329 Dardanelles fortresses, 192, 193 maps, 64 son of, 7 titles, 13 Mehmed IV, 15, 189, 191 Mehmed, kaşif of Ibrim, 378 Mehmed Ali (Muhammad ‘Ali) ‘Aqaba under, 104, 107, 112 centre-periphery relations, 237–8, 245 historical overview, 10, 11, 18 Page 38 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Iraq threatened by, 274 in north-east Africa, 231–2, 371, 509, 512, 515–16 Suakin under, 471, 515 Mehmed Bey II (Karamanoğlu), 314–15 Mehmed Bey Abu’l-Dhahab, 512 Mehmed Edib, 87–8 Mehmed Giray II (Khan), 358 Mehmed Mirza, 315n17 Mehmed Pasha, 180, 181 Mehmed Said, 518 Mehmet Agha, 175n11 Mehmet Emin Âli Pasha see Âli Pasha Mekreke, 382 Melek Ahmed Pasha, 143n14 Melingoi, 118, 119 Mennan, 321, 329n91 Meropi (Pırastoz), 126 Meshra Aboo-kooka, 527 Meshra er-Rek, 516, 527 Messina, 118, 130–1 metalwork, Akkerman, 165, 167 Methoni (Modon), 116, 117, 127, 129 Methuen, Lord, 505 Mexico, 517 Meyer, Joseph, 472 Mezapo Bay, 120, 132 Mezistre see Mistra Mezőkeresztes see Haçova Micaniko, General, 259 Michael the Brave, 335, 350 Michael Palaeologos, 119 Midhat Pasha, 94, 245–9, 251, 274, 281, 285 migration, 17, 183, 184, 187, 455–6 Mila (Kardamyli; Milia; Skardamoula), 126 Milani, 426 Millingen, F., 521 Milo, 507 mimar see architects Mimar Sinan, 62, 87, 195 Minorca, 504, 505 Mişik, 185 missionaries, Christian, 286 Mistra (Mezistre), 118, 119, 124 Mittu, 521 Modon see Methoni Mohács (Mohaç), 58 Moks, 221 Moldavia building materials and workers, 162, 170 Page 39 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index coinage, 167 prince of, 23 Principality administration, 21 Akkerman under, 142, 148, 156 frontier, 137, 139, 142 slave raids, 141 Molina Family Atlas, 48, 49 Monbuttu, 521, 521n64 Monemvasia, 119 Mongol Empire, 11, 12, 15, 272, 325 Montana (Kutlofçe), 334, 335, 348 Montenegro administration, 173, 184, 187 frontier, 16, 173, 175, 176, 179–80, 184, 240 raids by, 178, 180, 185 revolt 1875–6, 179 Moraca, river, 180 Morava, river, 65 morbus hungaricus, 78, 79 Morden, Robert, 472 Morea campaigns, 117, 129, 130, 318n31 fortresses, 129 (p.584) maps, 133 troops from, 124 under Tepedelenli Ali, 237, 241 see also Chronicle of the Morea Morea, Castle of (Rion), 117 Morocco, 7, 499, 504 Morosini, Francesco, 122 mosques Ahlat, 220 Akkerman, 145 ‘Aqaba, 98 Arduan Island, 379 Balkans, 183 Bihaćka Krajina, 412 Bihać, 412 Pećigrad, 408–9, 408, 410, 412 Stijena, 411–12 Damascus, 87 Erciş, 218 Ilgın, 326 Ioannina, 242 İşkodra, 185 Istanbul, 14 Kelefa, 133 Nile valley, 230, 371 Page 40 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Sai, 376 Sarayönü, 326 Seddülbahir, 191, 196–7, 197, 198, 206 Suakin, 476, 491 Van,216 Mostar, 184 Mosul, 273–4, 278 Mountains of the Moon, 491 Mubarak al-Sabah, 247 Mudawwara, 90 müfettis, 161 Muhammad ‘Ali see Mehmed Ali Muhammad b. Sa‘ud, 85 mühimme defteris, 3, 8 Muhsin b. ‘Ali b. Mani‘, 300 Muhtar Pasha, 244n30 Muhtar Pasha, Ahmed, 292, 293, 294n19, 295 mukataas, 348 Mukha, 291, 294, 300, 301n44 Mula Effendi, 521n64 mültezim, 20 Mumtaz Pasha, 483 Münif Pasha, 180 Muntafiq tribe, 249, 277 Muqbil, Muhammad Nasir, 299–300 Murad II, 315n20 Murad III, 8, 102 Murad IV, 14, 15 Muradiye see Bargiri Muraköz (Međimurje), 76 Mures, river see Maros Murteza Agha, 179 Muş, 211, 213, 215, 219, 222 Musa Pasha Hamdi, tomb of, 371, 372 Muscat, 285 Muscovy, 8, 60, 139, 140, 142 müselleman troops, 343 Musil, Aloïs, 108 Muşinzade Abdullah Pasha, 175, 178, 186 Mustafa, 308–9 Mustafa Ağa, al-Hajj, 265, 266 Mustafa Ali, 355 Mustafa Pasha, 46, 51n37 Mustafa Reşid Efendi, 263, 265, 266 Mustawfı, 314n15 Musul, 381 Mutahhar, 84 mûtemed, 456 Muzayrib, 87 Page 41 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Mytilene, 348n87 Na‘am (Rol), river, 522, 527, 532 Nadasd see Mecseknádasd naibs, 161 Naima, 192 Najaf, 64, 280 Najd administration, 273, 284, 285, 286, 287 trade, 84, 85, 250 Wahhabis, 85–6, 250 Nakhchevan, 215 Nakhl, 97, 103 Nakhlatayn, 90 Napoleon, 236, 253, 267 Napoleon III, 472 Napoleonic Wars see Adriatic frontier, Napoleonic Wars Naqb al-‘Aqaba (‘Aqabat Ayla), 95 Narante, 379 Narnante-i Sagira, 379 Nasir, 249 Nasir al-Din, 282 Navarino (Anavarin) Bay of, 117 kaza, 546–7 see also New Navarino Navpaktos see Lepanto navy Adriatic, 256, 262, 265, 265n65, 267 Aegean, 207 Danube organisation, 58–60, 336–7 sailors, 343n62, 343n64, 344, 348, 465 seat of admiral, 333, 337 shipbuilding, 337 Don, 60 Gallipolli, 462 Lepanto, defeat at, 8, 62 Red Sea/Arabian Gulf/Indian Ocean, 61, 226, 284, 286, 471, 476 (p.585) see also corsairs Nayhaysel (Neuhäusel; Uyvar), siege of, 37–9, 38, 46 Naznarinte, 379 Negroponte see Euboea Nelson, Admiral Lord, 266 Neretva, river, 184 Netherlands, 498, 502 Neuhäusel see Nayhaysel Neutra (Nitra), river, 37 New Navarino see Pylos Nicopolis, battle of, 331 Page 42 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Nicosia, maps of, 40–4, 41, 43 Niğbolu see Nikopol Niğde, 312, 315 Nikopol (Niğbolu) building workers from, 162n56 fortress, 58, 333, 335 garrison, 342, 343n65, 344 Nikšić (Nikşik), 173, 176, 179, 185 Niš, 5, 9, 65, 178, 351 nišani see funerary stelae Nitra see Neutra Nizam-ı Cedid, 9, 241 Nogays, 139, 141 Nográd see Novigrad Nomitsa (Nomçin), 126 Norbury, Capt Coningsbury, 504 north Africa frontiers, historical overview, 7, 8, 9, 11, 26 maritime frontier archaeology, 500–3 background and history, 493–500, 496 collective microhistory, 503–7 discussion, 507–8 formation and maintenance, 355 north-east Africa, frontier policies background and summary, 225, 234 Period One (1517–55), 226–7 Period Two (1555–89), 227–8 Period Three (1589–1820), 228–31, 229 Period Four (1820–1914), 231–3 see also Egypt; Sudan Nové Zámky (Érsekújvár; Uyvar), 71, 76, 455–6, 461 Novi Pazar, 180 Novigrad (Nográd), 67, 68 Nubia, 232, 472, 491, 514, 516; see also Berberistan; Sudan Nuer, 516, 523–4, 525, 537 Numan Bey, Esseyid, 182 Nureddin, 325 Nyam-Nyam, 521 Nyuong, 523 Ochakiv see Özi Oitylo see Porto Vitiloz Olay Bey, 460 Omoro see Galla Onogost, 179 Oradea see Várad oral history, 196n22, 525–7, 528, 529, 532, 535 Oran, 500, 503, 504n41 Page 43 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Orhan, 14, 193, 465 Orşova, 58, 173, 333 Ortelius, Abraham, 66, 472 Osijek see Eszék Osman b. Ertuğrul, 5 Osman Gazi, 465 Osman Pasvanoğlu Pasha centre-periphery relations, 240–1, 242, 243 at Vidin, 332, 336, 337, 341, 349 Osmanikki, 380, 383 Oştarije, 426 Oteš, 426 Otlukbeli, battle of, 316 Otoka, 419 Otranto, Strait of, 118 Ottendorf, Heinrich, 171n3, 400 Ottomanisation, 23–4, 183, 186, 541 Oudh Bequest, 280–1, 284 Ovington, J., 484 Özdemir Pasha, 17, 227, 374 Özdemiroğlu Osman Pasha, 17, 366 Özi (Cankerman; Ochakiv) administration, 139 fortress, 140, 156, 157, 175, 195 pottery, 163–4, 165, 167 trade routes, 141 Öziçe see Užice Ozora conquest, 70 pottery, 394, 395, 397, 399, 400, 401 Padalonya (Pano; Pende Alonia), 126 Pajić, Husejn, 425 Paks (Pakşa), 400 Palaiá Máïni see Grand Magne palanka forts Balkans, 171–2, 172 background, late 18th century, 175–7, 177 Bihaćka Krajina, 406 Black Sea frontier, 182–3 Danube frontier, 181–2, 335 discussion, 185–7 geography, 173–5, 174 Hapsburg frontier, 178–81, 457–8, 467 Hungary, pottery, 390, 392, 395, 396–7, 398, 399, 400 Palestine, 85, 90, 107, 231, 437, 545 Pálffy, Miklós, 76 Palota, 71 Pannonhalma see Szentmárton Pano see Padaolonya Page 44 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Pápa, 71 Papazoğlu, 346n77 (p.586) Parga, 236, 243, 245, 256, 261 Paris, treaty of, 247–8, 283 parkan, 68 Párkány (Ciğirdelen), 68 Parkyns, Mansfield, 484 Parry, Francis, 503n39 Pasha, 348 Pasin, 219 Passarowitz, treaty of, 336, 341, 344 Passava (Pasova), fortress Chronicle of the Morea, 118, 119, 120, 121 Evliya Çelebi visits, 127 plan, 128 Spanish attack, 114 Patrona Rebellion, 25 Paul I (Tsar), 255, 259, 261, 262, 269 Pavia, battle of, 37n12 Paxos, 253 Pećigrad (Pekleri) castle, recording, 406–11, 407, 408, 412 ethnography, 414, 418, 419 pottery, 429 Peck, Consul, 506 Pécs (Peç), 70, 346, 390n20 Pegu, 484 Pekleri see Pećigrad Pende Alonia see Padaolonya Pentele, 400 Persia see Iran Persian Gulf see Arabian Gulf Pest (Peşte), 58, 467 Peter the Great, 62, 346n77 Petermann, A., 473 Petherick, J., map by, 532–5 Petrovaradin, 58 Philip II, 63 Pigadia (Piğa), 126 Pılaça see Platsa pilgrims, Shi‘i, 94, 280, 283; see also hajj Pinargenti, Simon, map by, 40–2, 41, 50 Pinkerton, John, 472 Pir Ahmed, 311, 316–17, 318–19, 323, 327 Pir Hüseyin, 326, 327 piracy, 301; see also corsairs Pırastoz see Meropi; Prasteio Page 45 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Pirenne, Henri, 495, 497 Piri Reis, 61, 64, 132 Pirot, 423, 427, 428 Pitts, Joseph, 102 Platsa (Pılaça), 126 Pleternica, 427 Podzvizd, 412 Pojega, 58, 173 Pokrajna Zapadna Bosnia movement, 414 Poland (Podolia), 141, 167 Poland-Lithuania, 140, 142 Polomiye (Lom), 334, 335, 347, 348, 348n64, 350n95 Poncet, C. J., 483–5 Popović, C, 425 Porto Kaio (Kaya Pasha Limanı), 114, 118, 119, 127, 131 Porto Quaglio, 129 Porto Vitiloz (Vitylo/Oitylo) fortress, 122, 127, 130, 131–2, 134–5 harbour, 114, 118, 120, 127, 130–2 houses, 126–7 Portugal foreign policy Arabian Gulf/Indian Ocean, 7–8, 16, 61, 82, 83, 226–7 Red Sea, 22, 226–7, 471, 476 spies, 22 Third Coalition War, 255 see also Lisbon postal service, 283, 286 Potoćari, 413 pottery, 23, 548 Akkerman, 163–7, 166, 168–9 Hungary analysis, 395–7 discussion: ethnicity of supply hinterlands, 398–9; garrison composition, 398; as reflection of ethnic changes, 399–401 new types, 386–8; classification, 388–94, 388, 389, 391, 393, 394 Pećigrad, 411 Rumbek, 529, 531 Suakin, 486–90, 487, 489 pottery manufacture Akkerman, 148, 163n60, 164 Bihaćka Krajina, 421–9, 422, 428 Hungary, 386 Power, Daniel, 493–4 Pozsony, 78 Prasteio (Pırastoz), 126 Preveza, 236, 243, 256 Page 46 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Pribach, Sebold von, 66 Prijedor, 427 Prizren, 180, 181 provinces administration, 20–3 surveys, 63 Prussia, 176 Prut, 64 Ptolemy, 471 Pullen, Captain W. J. S., 473 Punt, land of, 491 Pylos (New Navarino), 117, 130 Pylos Regional Archaeological Project, 546–7 Qaitbay, 319 Qalandariyya, 90 Qal‘at al-Hasa, 90 Qal‘at Balqa, 90 Qal‘at Sai archaeology, 372, 375, 376–7, 377, 383 (p.587) garrison, 228, 230, 374, 382 Qansuh al-Ghawri, 97, 98, 104 Qasim, 248 Qasr Ibrim fortress, 227, 228, 371, 377 garrison, 17, 230, 374, 375, 376 Qa‘taba, 295, 296, 299 Qatar, 10, 245, 248, 250, 284, 285, 287 Qatrana, 88, 88 Qazvin, 214 qubbas, 371, 372, 476 Querini, Marco, 114 Raab, river, 35 Rác, 399–400 Radetina Kula, 406 Ragusa see Dubrovnik Rahva, 220 railways, 435, 436; see also Baghdad railway; Hijaz railway Ramadi, 524 Rastoki, 423, 427, 428 Rawanduz, 274 Raydaniyya, 509 razzias, 515 reaya Hapsburg frontier, economic opportunities, 456, 461–8 Vidin, 341–4, 345, 346, 348 Red Ruthenia, 141, 142 Red Sea, 61, 226–7, 228, 234, 291; see also ‘Aqaba; Suakin Page 47 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Revan (Yerevan), 19 Reynolds, R., 472 Rhine-Maas canal, 62 Rhodes, 7 Rhol, 522 Ribadeo, de, 505 ribat, 220 Ribnica, 423, 427 Rifa‘at Abou-el-Hajj, 13 Rijeka, 427 Rindomo Gorge, 126 Rion see Morea, Castle of Ripač, 415 roads ‘Aqaba, 97, 98 Balkans, 184, 186 Syrian hajj route, 90, 93–4 Roberts, David, 108 Robinson, Edward, 107 Rol, river see Na‘am Romania, 173, 185 Ronga, 518, 532, 535–6 Rosaccio, Giuseppe, map by, 42–4, 43, 50 Rough Cilicia see Cilicia Trachea Roumeli see Andirion Rudolph II, 35 Rukn al-Din Kılıç Arslan IV, 321 Rumbek archaeological survey, 525–7, 526 Lol Nhom, 530–2, 531 matching with historical sources, 532–6, 533, 534 Meen Atol, 529–30, 530 Pendit, 527–9, 528 zariba historical evidence, 520, 521 later history, 522–5 Rumeli (Rumelia) administration, 139, 347 building resources, 162, 170 coastal route, 141, 157 Cossacks raid, 151 defence, 186, 324, 462 governors, 70, 237 Rüppell, Eduard, 107 Ruse (Rusçuk) battle, 264 building workers, 162n56 dockyard, 337 fortress, 58, 333, 335, 337, 342n56 Page 48 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Russia canal, 16 First World War, 435 foreign policy Adriatic, Napoleonic Wars, 253–5, 256–9 Balkans, 173–5, 186, 240, 244, 336 Black Sea, 150–1, 156, 158, 167, 182, 186 Crimea, 153, 278 Greece, 114 historical overview, 9, 10 Iraq, 278, 279 navy, 207 spy, 158 wars with, 153 1768–74, 114, 175 1789–92, 158 1806, 259 1877–8, 278, 279, 284, 517 see also Muscovy Rüstem Bey, 326, 327 Saabindiffi, 380 Šabac (Böğürdelen), 178, 181 Safavids, frontier with 1535, 64–5 1578–1605 background, 353–4 formation and maintenance, 355–8, 357, 369–70; Erzurum, 362–4; Kars, 360–2; Shirvan Province, 358–60; Tabriz, 364–9 campaigns, 17th century, 82–3 historical overview, 2, 7, 8, 14–15, 16, 17 mapping, 53–5, 54 (p.588) spies, 309, 327 Van region, 213, 214, 215, 217, 219–20, 221 Sahib Giray, 369, 370 Sahra, 348n84 Sai see Qal‘at Sai St Gotthard, battle of, 461 St Petersburg, 262, 263 Sakarya-İzmit canal, 61–2 Saladin, 95 Salé, 504 Salih, 378 Salim al-Aswani, 471 Salonica, 268 salt trade, 185 salyaneli provinces, 20, 21, 273 Page 49 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Samarra, 280 San‘a’, 11, 84, 94, 292 Sanamayn, 87 sancak system administration, 20 Danube, 70, 71 Hapsburg frontier, 63 Karamanid lands, 312 north-east Africa, 226, 227, 228, 230–1, 372, 374 Van region, 214–15, 215–16, 217, 218, 221–2 Vidin, 347–8 sancakbeyis, 20, 161 Sanski Most, 425 Santa Maura, 44–6, 45, 46, 47, 117 Sanuto, Matteo, 44 Sarajevo, 178, 426, 427 Sarayköyü, 320 Sarayönü, 326 sarıca units see sekban units Sava, river, 67, 175, 176 Sayyid ‘Ali b. Ahmad, 293 Sayyid Talib Pasha, 249 Scanderbeg, 319 Schneider, Brigadier-General, 294–5, 296–7 schools, Christian, 184, 185; see also medreses Schwarzenberg, Adolph, 35 Schweinfurth, Dr Georg, 518–20, 521, 523, 527, 535, 536 map by, 532, 533, 534 Sebes-Körös, river (Crişul Repede), 71 Seçen see Szécsény Second World War, 409 Seddülbahir, fortress hamams, 149n28 repairs, 153n41 restoration project background, 189–93, 190 discussion, 207–8 historical evidence, 194–6; archival, 202–7, 204, 206; representational, 190, 196–202, 197, 198, 199, 200, 201, 202 oral history, 196n22 Seetzen, U. J., 484 Seged; Segedin see Szeged Segesd (Şegeş), 392 Şehrizor see Kirkuk Şehzade Korkud, 81 Şehzade Mustafa, 328 Page 50 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index sekban (sarıca) units, 463–4 Selaniki, 359, 363, 367 Seleucia, 312 Selim I, campaigns, 7, 15, 17 Arabia, 87 Black Sea, 145n17, 149 Egypt, 226, 509 Selim II, 8, 60, 61, 90, 145n17 Selim III alliance with France, 207n42 army reformed by, 9, 241 frontiers under, 182, 183, 186, 244 tribute received by, 258n20 Selim (Commander), 476 Selim Bey, Miralay, 180 Selim Kapudan, 232, 515 Seljuks, 12, 13, 321 Semendire see Smederevo Semeonis, Symon, 129 Senj, 403 Serbia frontier, 5, 10, 173, 331, 336 Orthodox Christians, 345–6 pottery, 390, 421, 423, 427–8 Tsar, 180 uprisings, 178, 242, 332, 349 Şeremet Mehmed Bey, 265, 265n65 serhad-i mansura, 16 serhadli troops, 342 Seriba Ghattas, 532 Seven United Islands see Ionian Islands Severin, 58, 333 Seydi (Sidi) Ali Reis, 61 Seydisehir, 312, 318n32, 326 Şeyhi, 25 Seyid Ali Efendi, 261n35 Seyyid Harun Veli, 326 Sha‘iri, 302 Shambe, 518 Shams al-Din, 222 Shankir, 471 Sharaf, 222–3 sharia law, 15, 249 Sharif Husayn, 110, 436 Shatt al-‘Arab, 286 Shaw, Brent D., 309 Shaykh al-Barghuth, 471 Shayqiyya confederation, 377 Shemakha, 359 Page 51 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Shiglindiffi, 380 Shi‘i, 7, 14–15 in Arabia, 84, 85, 86 in Iraq, 271, 279–82, 283 in Yemen, 292 Shilluk, 515, 524n79 (p.589) shipbuilding, 58, 337 Shirvan, 61, 356, 358–60 Shkodër (İskodra), 173, 179, 185, 238n6, 240 Sidirokastro (Demirhisar), 126 Sigetvar, 77 Şikari, 25, 325 Siklós (Şikloş), 68–70 Silahdar Fındıklı Mehmed Ağa, 124 Silifke, 312, 317 Silistra (Silistre) fortress, 58, 175, 333, 335, 342, 344 sancak, 139, 182 Šiljkovača, 421 silk trade, 141, 215 Sille (Su-diremi; Zillieh), 319, 320 Simit Island, 380, 381 Simontornya (Şimontorna), 70, 399 Sinac, 426 Sinan Pasha, 37, 53n39, 84, 94 Sindi, 380 Sinnar, 232, 514–15, 516–17; see also Funj Sultanate Sinop, 141, 151 Sinzendorf, Joachim von, 66n22 sipahis, 20, 68 Karaman, 323, 324 Van, 214, 224 Vidin, 337–40, 341 Sirem (Srem; Srjem), 70, 71 Sırrı Bey, 326 Sistova see Svištov Sivrihisar, 315n17 Skardamoula see Mila Skokovi, 407, 424, 425 slave trade, 18, 26 Arabia (eastern), 249 Black Sea region, 139, 141, 498 Egypt, 512–14 Greece, 127 Hungary, 458–9 north Africa, 498, 499, 503 north-east Africa, 225, 230, 233, 484, 485 Suakin, 484, 485 Page 52 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Sudan background, 509–14, 510 description, 514–18 discussion, 536–7 Rumbek: archaeological survey, 525–32, 526, 528, 530, 531; later history, 522–5; matching with historical sources, 532–6, 533, 534 zaribas: Rumbek, later history, 522–5; system of, 518–22 Slavonia, 175, 426, 427 Slavonski Brod, 406 Slovenia, 423, 427, 429 Slunj, 410 Smederevo (Semendire), 58, 173, 178, 181 society, historical overview, 23–5 Sofia, 351 Söğüt, 12 Soko Banja (Bane), 333, 335n14, 340n46, 342, 343n62, 347, 348 Sokol, 178, 182 Sokullu Mehmed Pasha, campaigns under, 59, 60–1, 70 Solnok see Szolnok Somalia, 234 Somogy, 399–400 Sopron, 78 Southworth, A. S., 521 sovereignty, concept of, 11–12 Spain, 355, 495, 498, 505, 506 Speed, John, 472 spice trade, 82, 141, 226 Spon, Dr Jacob, 122, 127 Spuž (İşpozi), 176, 179n31 Srem; Srjem see Sirem Stari Bar (Bar), 173, 179 Stefan Dušan, 180 Steiner, Erich, 109 Stephan the Great, 139 Stijena (İstineli), 406–7, 410, 411–12, 411, 418 Strumitza see İstrumca Su-diremi see Sille Suakin (Sevvakin) administration, 22, 226–7, 230, 232, 233, 234, 515 archaeological project description, 473–6, 475; Bayt al-Basha, 480–1, 480; Bayt al-Mufti, 476–7; Bayt Khurshid Efendi, 477–9, 478 discussion, 490–1 port: economy, 26, 483–5, 514; extent of trade, 486–90; Page 53 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index harbour, 481–3 site history and location, 469–73, 470 garrison, 230–1 houses, 377 Ottoman remains, 371 Subayhi, 297 Sudan administration, 232, 233, 234 archaeology of middle Nile valley background, 18, 371–5 discussion, 382–3 Funj-Ottoman frontier region, 375–6; Arduan Island, 379–80; Hannek, 380–1; Jabal Sese, 378–9; Kada Musa, 379; Qal‘at Sai, 376–7, 377; Simit Island, 380; Tinare, 378; unidentified sites, 381–2 Khartoum, 371, 372 Christian kingdoms, 226, 471 slave trade, 19th century background, 509–12, 510 description, 514–18 discussion, 536–7 in Egypt, 512–14 Rumbek: archaeological survey, 525–32, 526, 528, 530, 531; later history, 522–5; matching with historical sources, 532–6, 533, 534 zariba system, 518–22 see also Suakin (p.590) Sudd, 515, 516 Suez, 226, 485 Suez Canal, 60, 61, 292, 436, 471, 485 Sufism, 233, 239, 277 Sugar, Peter, 459–60 Sukkot, 227, 228 Sulayman Janbulad, 374 Sulayman Zubayr, 522 Sulaymaniya, 273, 274 Süleyman I campaigns, 7–8, 14 Arabia, 61 Black Sea, 139 Dardanelles, 191n3, 193 depiction of, 50–1, 53–5, 54 Eastern, 64, 360, 365, 370 Hungarian, 58, 59, 64, 337 Page 54 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index north-east Africa, 227, 513 canal projects, 60, 62 hajj route under, 87–8, 99–101 kanunname, 462 Süleyman Ağa, 107 Süleyman Pasha, 61 Suli, 239, 243 Suliotes, 259, 267 Sultan Hatun, 316n21, 326, 327 Sultaniye, 64 sürgün, 17 Svištov (Sistova), treaty of, 176 Svrljig (İsfirlik), 333, 335, 342, 347, 348 Syria Arab Revolt, 437 frontier, 7, 9, 231, 274 hajj route, development, 86–7 16th century, 87–90, 88, 89 18th century, 90–1, 91, 92, 93 discussion, 93–4 Mosul, ties with, 273 Syrmia, 175 Szécsény (Seçen), 68 Szeged (Segedin; Seged), 70 Székesfehérvár (İstolni Belgrad), 65 Szekfű, Gyula, 72 Szekszárd-Újpalánk (Yenipalanka), 390n18, 392 Szent Tamáshegy (Tepedelen), 68 Szentmárton (Pannonhalma), 71 Szigetvár, 33, 64, 76 Szolnok (Solnok), 70, 71, 76, 390n20 Taber, Robert, 450 Tabriz, 8, 64, 215, 316, 353, 364–9 Tabuk, 88 Tahir pass, 219 Tahirids, 83 Tahmasp, 217n25, 220 tahrir surveys, 3, 51n37 Akkerman, 145 Balkans, 183 Cyprus, 183 Hungary, 63, 71, 183 Karamanid lands, 312 Mani, 127 Van region, 221, 223, 224 Ta‘izz, 84, 294, 295, 296, 297, 299 Takkeli Mountain, 320 Tall Far‘un (Mafraq), 87 Talleyrand, Charles, 261 Page 55 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Taman, 140 Tamási (Tomaşin), 70 Tamerlane see Timur Tanzimat reforms, 9–10 Arabia, 245, 250–1 Iraq, 271, 274, 275, 276 Taqali kingdom, 513 Tara, river, 180 Tarih-i Kale-i Seddülbahir, 191–2 tariqas, 277 Tata, 71, 76 Tatars, 5, 60, 137–40, 142, 167, 358 Taurus mountains, 211–13, 307, 309, 312 taxation administration, 3, 20–1, 25 Akkerman, 161, 162 Arabia, 247, 248, 249, 250 Balkans, 181, 184 Berberistan, 374–5 Egypt, 511 Hapsburg frontier, 465–6 Ionian refugees, 267 Iraq, 271, 273, 276, 277 Sudan, 517 Vidin, 349 Yemen, 293, 296, 303 Tehran, 283 telegraphy, 274, 283 Teleki, Pál, 72 Temes (Timiş), river, 70, 76 Temesvár (Temeşvar; Timişoara) administration, 347, 348 Banat, 175 fortress, 68, 70, 71, 76, 331 gunpowder works, 75 Teny, prophet of, 523, 524 Tepedelen see Szent Tamáshegy Tepedelenli Ali Pasha Adriatic frontier, 255–6, 255n2, 259–60, 261, 266–7, 270 relations with Istanbul, 9, 236–45 Tercuman Mahmud, 66 Tetouan, 497, 505, 506 Thailand, pottery from, 488, 489 Theodora, Princess, 14 Therand, 180 Thessalonike, 5 Third Coalition War, 255, 261, 269 Thomas, Lowell, 431 (p.591) Thomson, Thomas, 505, 506 Page 56 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Thrace, 5, 189 Tiflis, 359, 364 Tighinia, 139 Tigris, river, 58, 64, 274, 283, 286 Tıhala see Trikala Tihany (Tihon), 71 Tilsit, treaty of, 255, 261, 269 timar system Hapsburg frontier, 456, 460, 461 Iraq, 273 Karaman, 323–4, 325, 327, 328 Van, 214, 222 Vidin, 332, 337–41, 344, 347, 348, 350 timarlı provinces, 20, 21, 25 Timiş, river see Temes Timişoara see Temesvár Timok, 340n43, 347, 348, 348n84 Timur (Tamerlane), 5, 314–15 Tinare, 378, 381 Tinne, Mrs and Miss, 516 Tinos, 114 Tisza, river, 67, 68, 70 tobacco pipes, 23, 548 Akkerman, 163, 167, 168, 168 Pećigrad, 411, 429 Tobacco Protest, 280 Todorovo, 410, 412 Tokaj, 76 Tokat, 348n87, 364 Tolna, 399–400 Tomara, M., 263, 264, 266 Tomaşin see Tamási Tomić, Persida, 421 Törökkoppány (Koppan), 392, 399 Toulon, 266, 503, 507 Tournefort, J. P. de, engraving, 197, 198 Trabzon (Trebizond), 5, 216, 218, 244, 316 trade Arabia, 84, 85, 249, 250; see also Yemen Bihaćka Krajina, 411, 424–5, 426, 427, 429 Black Sea, 141, 145, 148, 162, 165, 168–9, 268 Egypt, 512, 513 Ionian Republic, 268 Iraq, 273, 274, 283, 286 north Africa, 26, 498, 499, 502, 503–4 north-east Africa, 225, 226, 230–1, 233, 376; see also Suakin Suakin, 471, 473, 482–90, 491, 514 Page 57 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Van region, 215, 216–17, 220, 221, 223 Yemen, 231, 290, 294, 298, 299–300, 512 see also arms trade; caravans; gold trade; ivory trade; salt trade; silk trade; slave trade; spice trade; wine trade Transylvania frontier, 71, 175, 335 Long War, 78 merchant, 519 prince of, 50–1 revolt, 339 Traquair, R., 121, 133 Trebinje (Trebin), 173, 176 Trebizond see Trabzon Tremenhere, Major-General, 294 tribal systems Iraq, 276–8 Van,213–14 Yemen, 290–1, 291, 297, 299, 303–4 tribute (harac/cizye), 21, 232, 256–8, 359 Trikala (Tıhala), 241 Tripoli, 26, 498, 499, 500, 504, 507–8 Troad, 189 Trucial Coast, 285 Tržac, 410, 412, 418, 419 Tsamides (Çam), 259 Tulcea (Tulça), 175 Tumanbey, 226, 509 Tunis, 268, 498, 499–500, 503, 507 Turgut, 325–6 Turgut tribesmen, 308–9, 323–7, 328 Turgutoğlu Hüseyin, 324 Turhan Sultan (Hadice Turhan Sultan), 189, 191–2, 193 Turkiyya, 18, 232, 371, 509, 537 Turner, F. J., 455, 456, 495 Turnu Măgurele (Holıvnik), 333, 344 Turnu Severin, 173 Tursun Bey, 319 typhus, 78, 79 Tyras, 142, 545 Tzimona see Areopolis uc, 12, 16, 333 Uganda, 234 Ukhaydhir, 87–8 Ukraine Cossacks, 150–1 frontier, 137, 142, 175, 545 pottery, 165, 167 slave trade, 139 ulema, 11, 230, 233, 249, 280, 281, 282 Page 58 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index ‘Umar Bey Qatamish, 103 ‘Umar Salim, 301n44 Una, river, 173, 175, 176, 180, 184 ‘Unayza, 90 Upper Kolaşin (Kolaşin-i-Bala), 180 Ushakov, Admiral, 256, 262, 263, 264, 266 Uskoks, 403 Üstad Osman, 50n35 Üveys Pasha, 51n37, 66 Uyvar see Nayhaysel; Nové Zámky Užice (Öziçe), 178, 182 Uzun Hasan, 316, 317, 318 (p.592) Vác (Vaç), 58, 68 vakıf foundations, 145, 183, 273, 326, 327, 347 Valentia, Viscount, 484 Van (city), 64, 214, 215, 216, 223 Van region 16th-century administration and fortification administration, 221–4 area defined, 211–13, 212 capture of, 214–16 fortification, 216–21 tribal principalities, 213–14 province of, 215, 221–3, 368 Várad (Varad; Oradea), 71 Varna, 58 Varsak tribesmen, 308–9 Vasieur (Le Vaseur; Vaseur; Vassor), plan by, 128, 129, 133 Vatheia (Vatya), 127 Vatican map, 379, 380, 381, 383 Vatya see Vatheia Vauban, Marquis de, 156, 158, 186, 336 Vazne, 279 Vefa, 203 Vejnac, 419, 420 Veleşniçe, 347 Velika Kladuša, 409, 410, 414, 415 Venice conquered by Napoleon, 236, 253 foreign policy, 7, 12, 15 Albania, 319 Balkans, 173, 185–6 Dardanelles, 191, 192 Karamanids, 318 Mani, 113, 114–17, 121–2, 129, 130, 132–3 fortresses mapped by, 37, 40–50, 41, 43, 45, 46, 47, 49, 128, 254 map of lagoon fortifications, 64 trade, 471 Veselići, 426 Page 59 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Veszprém (Bespirim), 71, 74n41 Via Aegnatia, 186 Via Militaria, 171n3, 186 Via Sebaste, 313, 322n55, 325 Vidin administration, 22, 181, 347–9 centre-periphery relations, 240, 243 frontier, historical overview, 4 garrison, 264, 337–46 Ottoman rule, 15th–18th centuries, 331–2 administration and agrarian regime, 347–50 discussion, 350–1 fortifications, 333–7 human resources, 337–46 plan to transport Muslims to, 178 public buildings, 184 Vienna campaigns 1529, 7 1593–1606, 71 1683, 9, 82, 459, 464 map, 64 pottery, 394n30 spy, 66 Vietnam, pottery from, 488–90, 489 vilayet province, 312 Villehardouin, Guillaume de, 113, 118–19 Virdan, 379 Visegrád (Visşegrad), 58, 68, 395, 397, 399, 401 Višnjica, 425 Vitiloz see Porto Vitiloz Vittolo, 130, 131 Vitylo see Porto Vitiloz viziers, role of, 161 Vlach, 343n65, 399 Vojna Krajina, 403, 404–5, 425; see also Krajina Project Volga, river, 60–1, 62 Volgograd (Tsaritsyn; Stalingrad), 60 Volhynia, 141 Vonitsa, 236, 243, 256 Vostan, 213 voynuk troops, 343, 344, 403, 465 Vranina, 185 Vrginmost see Gvozd Vrkašić, 414, 424, 425, 426, 428–9 Vroom, Joanita, 165, 167 Wadi Halfa, 227, 228, 231 Wadi Rutm, 435, 444–7, 445, 449, 451 Page 60 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Wadi Safiyya, 304n58 Wahab, Colonel, 300, 302–3, 304n58 Wahhabi forces, 84, 85–6, 94, 250 Wahid Bey, 441 Wajd, 437 Wallachia administration, 21 building workers and materials, 162, 181 frontier, 173, 176, 331, 335, 336, 344, 350 Washshasha, 103 watermills, 419–21, 420 Wau, 525 Wavell, Lt.-Col., 451n36 Wheler, George, 122, 127 Whittaker, C. R., 493 William of Orange, 506–7 wine trade, 26, 504 Wingate, General, 109–10 Wise, Lawrence, 506 Wittek, Paul, 13, 494–5, 539 Wol Athiang (Awol Athiang; Wuol Athiang), 523–5 wood, analysis of, 490 Woods, John E., 317 Wulu, 527 Wuol Athiang see Wol Athiang Wylde, A. B., 485 Wyngaerde, Anton van den, 63 (p.593) Yafi‘i, 290, 297 Yahşi Bey, 342n52 Yahya Pasha, 103 Yanık see Győr Yanık Hisar, 182n48 Yanık Palanka, 182 Yemen administration 1872–1914 background and context, 289–96, 291 British response, 296–8 discussion, 303–4 imperial patronage, 298–301 territorialising, 301–3 frontier conquest and reconquest, 82, 83–4, 83, 94, 226, 231 historical overview, 7, 8, 10–11, 16, 17 Ottomanisation, 23 trade, 231, 290, 294, 298, 299–300, 512 Yeni Imam, 53 Yenikale, 140 Yenipalanka see Szekszárd-Újpalánk Yenişehir see Larissa Page 61 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020

Index Yerevan see Revan Yergöği see Giurgiu Yıva, 325n76 Young Turk revolution, 435–6, 436n15 Yusuf Agha, 185 Yusuf Şah Bey, 326 Zab valley, 213 Zabid, 83 Zagoriye, 347 Zagorje, 427 Zagreb, 426 Zagyva, river, 68, 70 Zahir al-‘Umar, 85 Zahr al-‘Aqaba see Fassu‘a Zakany, river, 76 Zangi of Aleppo, 13 Zante, 253, 259, 261, 263 Zanzibar, 11, 232 Zapolya, John, 51 Zapolya, John Sigismund, 51, 52 Zaporozhia, 139–40, 141, 150–1 zaribas Rumbek archaeological survey, 525–32, 526, 528, 530, 531 later history, 522–5 matching with historical sources, 532–6 system, 509, 518–22, 537 Zarnáta, 121, 122, 126, 127, 129, 134 zaviyes, 216, 220 zawiyyas, 230 Zayla, 227, 230, 232, 513 Žegar, 414 Zemun (Zemin), 58 Zengicek (Kontkaşa; Koçkaya), 323, 324, 326 Zillieh see Sille Zıvarık (Altınekin), 326 Zrínyi (Zrinski), Miklós (Nikola), 76 Zrinyis, 180n36 Zubayr Pasha, 26, 519, 522 Župan, Grga, 426 Zürcher, Erik, 24 Zvornik (İzvornik), 58, 427 (p.594) (p.595) (p.596) (p.597) (p.598) (p.599) (p.600)

Access brought to you by:

Page 62 of 62 PRINTED FROM BRITISH ACADEMY SCHOLARSHIP ONLINE (www.britishacademy.universitypressscholarship.com). (c) Copyright British Academy, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in BASO for personal use. Subscriber: OUP-Reference Gratis Access; date: 07 October 2020