From Albania to Arrān: The East Caucasus between the Ancient and Islamic Worlds (ca. 330 BCE–1000 CE) 9781463239893

This volume is the first publication in English to discuss the nature and identity of the polity in the East Caucasus re

186 99 26MB

English Pages 405 [399] Year 2020

Report DMCA / Copyright

DOWNLOAD PDF FILE

Recommend Papers

From Albania to Arrān: The East Caucasus between the Ancient and Islamic Worlds (ca. 330 BCE–1000 CE)
 9781463239893

  • 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

From Albania to Arrān

Gorgias Studies in Classical and Late Antiquity

25 Series Editorial Board Johannes Niehoff-Panagiotidis Rebekka Nieten Adrian Pirtea

Irene Schneider Manolis Ulbricht

Advisory Editorial Board Stefan Esders Thomas Figueira Christian Freigang David Hernandez de la Fuente Markham J. Geller Susan Ashbrook Harvey

Nicola Denzey Lewis AnneMarie Luijendijk Roberta Mazza Arietta Papaconstantinou Meron-Martin Piotrkowski Shabo Talay

This series contains monographs and edited volumes on the Greco-Roman world and its transition into Late Antiquity, encompassing political and social structures, knowledge and educational ideals, art, architecture and literature.

From Albania to Arrān

The East Caucasus between the Ancient and Islamic Worlds (ca. 330 BCE–1000 CE)

Edited by

Robert *Hoyland

gp 2020

Gorgias Press LLC, 954 River Road, Piscataway, NJ, 08854, USA www.gorgiaspress.com Copyright © 2020 by Gorgias Press LLC All rights reserved under International and Pan-American Copyright Conventions. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise without the prior written permission of Gorgias Press LLC.

‫ܐ‬

1

2020

ISBN 978-1-4632-3988-6

Library of Congress Cataloging-in-Publication Data A Cataloging-in-Publication Record is available from the Library of Congress. Printed in the United States of America

TABLE OF CONTENTS Table of Contents ....................................................................................................... v Acknowledgments..................................................................................................... vii Abbreviations.............................................................................................................. ix Transliteration & Conventions ............................................................................... xv Introduction ................................................................................................................. 1 I. HISTORIES & PERSPECTIVES ............................................................... 7 1. Albania in Greek and Latin Texts: The Use and Utility of ‘Views from the West’..................................................................................................................... 9 Lara Fabian 2. The Chronology of the Arsacid Albanians ....................................................... 29 M.S. Gadjiev 3. %XOGćQDO-5ćQ: The Many Definitions of Caucasian Albania in the Early Abbasid Period ................................................................................................. 37 Alison Vacca 4. “The Wall of Rock and Lead”: Abbasid Reflections on Sasanian Caucasian 3ROLF\LQ$UUćQ.................................................................................................. 85 Ryan Lynch 5. The Mission of BishRS,VUD\ĔOLQWKH&RQWH[WRIWKH+LVWRULFDO*HRJUDSK\ of Caucasian Albania ..................................................................................... 101 M.S. Gadjiev 6. $UUćQLQWKHth-10th Centuries: 6DKOLEQ6XQEćʜ (d. ca DQG0DU]XEćQ b. MuʘDPPDGE0XVćILU G ............................................................... 121 V. Minorsky 7. The Caucasus between Byzantium and the Caliphate (9th–10th c.) ............. 149 Constantin Zuckerman 8. Georgian Images of Caucasian Albania .......................................................... 191 Stephen H. Rapp Jr. II. MATERIAL CULTURE ....................................................................... 217 9. On the Cities of Ancient Caucasian Albania .................................................. 219 I.A. Babaev and J.A. Khalilov v

vi

FROM ALBANIA TO ARRAN

10. Jerusalem in the Caucasus: Church Building and Relic Traditions in Late Antiquity .......................................................................................................... 233 Annegret Plontke-Lüning 11. The Medieval City of Bardha‘a ....................................................................... 253 A.B. Nuriev and Paul Wordsworth 12. Caucasian Albania and its First Capital City: Qabala .................................. 277 Jeyhun Eminli III. LANGUAGE & LITERATURE ........................................................... 305 13. Peoples and Languages of Caucasian Albania: On the Language Continuum as an Alternative for Koine: Written Language and Vernacular Language ..................................................................................... 307 A.K. Alikberov 14. The Deciphering of the Script of the Caucasian Albanians ...................... 335 Zaza Aleksidzé, with Jean-Pierre Mahé 15. &DXFDVLDQ$OEDQLDDQGLWV+LVWRULDQ .............................................................. 351 James Howard-Johnston Indices ....................................................................................................................... 371

ACKNOWLEDGMENTS This volume has had a long gestation period. It began in summer 2013 when I went to Azerbaijan to investigate the possibility of excavating there. I became interested in Barda (Arm. Partaw, Arab. Bardha‘a), since it suited my expertise in that it flourished through the late antique – early Islamic period and went from being a capital of a Christian kingdom to the base of a Muslim garrison. While I was there, I was invited to meet Professor Nargiz Pashayeva, rector of the Baku branch of Moscow State University, as someone interested in this project. We quickly realized that we shared a strong belief in making scholarship more accessible and more interconnected, and we talked about the idea of a research center that would be able to promote these aims. Under Professor Pashayeva’s patronage, the Oxford Nizami Ganjavi Program was established, which had as its mandate the provision of postgraduate scholarships in the field of Caucasian Studies, the management of an archaeological mission at Barda (by the program’s postdoctoral fellow, Paul Wordsworth) and the identification and translation of some key research produced in Russian and Azeri concerning the pre-modern history of the Caucasus. In order to complement the excavations being carried out at Barda, I thought it would be a good idea to direct some of our translation activity to past publications about that city and the polity of which it was the capital, and the result is the chapters below by Gadjiev, Babaev/Khalilov, Nuriev and Alikberov. To make it into a full-length volume, I included contributions from other scholars, leading to the remaining chapters, which are either original works (Fabian, 9DFFD/\QFK5DSS(PLQOL:RUGVZRUWKpVDGGHQGXPWR1XULHYDQG+RZDUG-Johnston) or translations/reprints of classic articles (Minorsky, Zuckerman, Plontke-Lüning and Aleksidze/Mahé). The final outcome is the first volume in English on Caucasian Albania and an exploration of its culture by a diverse array of experts from many different language and cultural backgrounds, fulfilling the Oxford Nizami Ganjavi Program’s aim of promoting cross-cultural academic collaboration. It will be published just as this program is transformed by a further large grant overseen by Professor Pashayeva into the permanently endowed ‘‘Oxford Nizami Ganjavi Centre IRUWKH6WXG\RIWKH+LVWRU\/DQJXDJHVDQG&XOWXUHVRI$]HUEDLMDQWKH&DXFDVXVDQG Central Asia,’’ which will facilitate research into these fields for a long time to come. The translations were all done by Maroussia Bednarkiewicz, the first official Oxford Nizami Ganjavi Program translator and Oxford doctoral student, and myself, except for chapter 2 below, which was done by Jordan Lian (NYU) and myself. I am very grateful to Lara Fabian and Alison Vacca, who are much more knowledgeable about Caucasian archaeology and history than I and who have been very generous vii

viii

FROM ALBANIA TO ARRAN

with their expertise, for responding to my frequent questions and requests without demur. The translations, except for chapter 2 below, were paid for by the Oxford Nizami Ganjavi Program, and so I am indirectly indebted to the three people who in different ways brought this program into being and also its successor, the Oxford 1L]DPL*DQMDYL&HQWHU7KH\DUH3URIHVVRU(GPXQG+HU]LJZKRWRRNRYHUPDQDJH ment of this venture after I left Oxford for the Institute for Study of the Ancient World (New York University) and who has always been very supportive; Leyla Najafzada, who began as my doctoral student and, by virtue of her ability to translate between English, Azeri and Russian as well as her passionate belief in the worth of the Oxford Nizami Ganjavi Program/Center, kept the whole project on track; and, most importantly, Professor Nargiz Pashayeva, who not only oversaw the funding of the project but facilitated all aspects of its activities (excavations, translations, workshops and so on). Without her neither this project as a whole nor this volume in particular would have seen the light of day.

ABBREVIATIONS Arab. Arabic AN Akadamiia Nauk / Academy of Sciences Arm. Armenian b. ibn/‘‘son of’’ BGA Bibliotheca Geographicorum Arabicorum BSOAS Bulletin of the School of Oriental and African Studies CRAIBL Compes rendus de l’Académie des Inscriptions et Belles-Lettres Ed. edited EI The Encyclopedia of Islam, 3 editions (Leiden, 1913–36, 1954–2005, 2007– ). EIr Encyclopedia Iranica (New York, 1982–; free online at http://www.iranicaonline.org/) Georg. Georgian Gr. Greek Lat. Latin Pers. Persian REArm Revue des Etudes Arméniennes Tr. translated Turk. Turkish

ABBREVIATIONS FOR COMMONLY USED PRIMARY SOURCES Caucasian (Armenian and Georgian) Agat‘angelos, History of the Armenians, ed. and tr. Robert Thomson (Albany NY, 1976). Brosset = M.F. Brosset, Collection d’historiens arméniens I (St Petersburg, 1874). 'DV[XUDQFoL 0RYVĔV'DV[XUDQFoL 3DWPXWoLZQ$øXDQLFoDx[DUKL, ed. Varag Aѻak‘elyan (Yerevan, 1983); tr. C.J.F. Dowsett as History of the Caucasian Albanians (London, 1961); tr. S.B. Smbatyan as ǖǿȀǼǾǶȍ ǿȀǾǮǻȉ ǎǹȁǮǻǸ (Yerevan, 1984). Note that Dowsett’s translation omits the heading of chapter 3.17 so that the numbering of subsequent chapters lags one behind that of Arak‘elyan. ix

x

FROM ALBANIA TO ARRAN

(øLxĔ = (øLxĔ Vasn Vardanay ew Hayocұ paterazmin, ed. E. Ter-Minasean (Yerevan, 1957), tr. Robert W. Thomson as History of Vardan and the Armenian War (Cambridge, MA, 1982). Kұartұlis cұ[RYUHED (= The Life of K‘art‘li/Georgia), ed. 64DX[ă‫ދ‬ixvili, vol. 1 (Tbilisi, 1955); tr. Robert W. Thomson as Rewriting Caucasian History: The Medieval ArmeQLDQ$GDSWDWLRQRIWKH*HRUJLDQ&KURQLFOHVWKH2ULJLQDO*HRUJLDQ7H[WVDQGWKH$UPHQLDQ Adaptation (Oxford, 1996); tr. M-F. Brosset as Histoire de la Géorgie I (St Petersburg, 1849). Koriwn = Koriwn, Varkұ Maxtocұi, ed. 0 $EHøHDQ /RQJ@= $x[DUKDFoR\Fo0RYVĔVL;RUHQDFowR\>/RQJ5HFHQVLRQ@, ed. A. 6RXNU\ 9HQLFH 7KHZRUNZDVLQLWLDOO\LQFRUUHFWO\DWWULEXWHGWR0RYVĔV Xorenac‘i and so it has sometimes been referred to as Ps-Xorenac‘i’s Geography. —— $x[DUKDFoR\Fo>6KRUW@ = Anania Širakac‘u Matenagrut‘yunʠHG$xRW$UVHQL$EUD hamyan (Yerevan, 1944). —— $x[DUKDFoR\Fo>8QLTXH@ = $x[DUKDFoR\Fo 0RYVĔVL;RUHQDFowR\>8QLTXH5HFHQVLRQ@ (Venice, 1865). —— Geography = 7KH*HRJUDSK\RI$QDQLDVRIkLUDN$x[DUKDFұoycұ, the Long and the Short RecensionsWU5REHUW+HZVHQ :LHVEDGHQ  Step‘DQRV$VRøLN = Step‘DQRV$VRøLNRI7DUňQHistoire universelle, part 1 (= Books 1– 2), tr. E. Dulaurier (Paris, 1883); part 2 (= Book 3), tr. F. Macler (Paris, 1917). Step‘anos Orbelean = Step‘anos Orbelean, Histoire de la Siounie, tr. M. Brosset (St Petersburg, 1864) in vol. 1 with commentary in vol. 2. T‘ovma Arcruni = T‘ovma Arcruni, Patmut‘iwn tann Arcruneac‘, ed. K. Patkanean (St Petersburg, 1887); HG9UH{9DUGDQ\DQ (Yerevan, 1985); tr. Robert Thomson as History of the House of the Artsrunik‘ (Detroit, 1985); tr. Brosset, 1–266. ;RUHQDFoL 0RYVĔV;RUHQDFoLPatmut‘iwn Hayoc‘, ed. M. Abeøean and S. Yarut‘iwnean (Tbilisi, 1913); tr. Robert Thomson as History of the Armenians (Cambridge MA,  WU+(PLQDVIstoriia Armenii Moiseiia Khorenskago (Moscow, 1858). Greco-Roman 1%/RHELQWKLVOLVWVWDQGVIRUWKHVHULHVSXEOLVKHGE\+DUYDUG8QLYHUVLW\3UHVV

ABBREVIATIONS & SOURCES

xi

Aelian = Aelian, De natura animalium/ On the Characteristics of Animals, ed./tr. A.F. Scholfield (Loeb, 1958–1959). App. Mithr. = Appian, Mithridatic Wars, vol. 2 of Appian’s Roman HistoryHGWU+ White (Loeb, 1912). Aristotle Meter. = Aristotle, MeteorologicaHGWU+3'/HH /RHE  Arr. Anab. = Arrian, $QDEDVLVRI$OH[DQGHU, ed./tr. P.A. Brunt (Loeb, 1976–83). RG = Augustus, Res Gestae, ed./tr. A.E. Cooley (Cambridge, 2009). Dio = Cassius Dio, Roman History, ed./tr. E. Cary (Loeb, 1914–27). Eutr. = Eutropius, BreviariumWU+:%LUG /LYHUSRRO  Fest. Brev. = Festus, Breviarium, ed. M.-P. Arnaud-Lindet as Abrégé des hauts faits du peuple romain (Paris, 1994). Flor. = Florus, Epitome of Roman History, ed./tr. E.S. Forster (Loeb, 1929). Front. Str. = Frontinus, Strategemata, ed. M.B. McElwain, tr. C.E. Bennett (Loeb, 1925). +HFDWDHXV)*U+LVW) +HFDWDHXV'LH)UDJPHQWHGHUJULHFKLVFKHQ+LVWRULNHU Frg. 1 F 291, ed. F. Pownall in Brill’s New Jacoby (Leiden, 2013). +GW +HURGRWXVThe Histories, tr. R. Waterfield (Oxford, 1998). +RUCarm. +RUDFHCarmina, ed./tr. N. Rudd (Loeb, 2004). Joseph. A.J. = Josephus, $QWLTXLWDWHV-XGDLFDHHGWU/+)HOGPDQet al. (Loeb, 1930– 65). Joseph. B.J. = Josephus, Bellum JudaicumHGWU+6-7KDFNHUD\ /RHE–28). Justin = Justin, Epitome of Trogus, ed. O. Seel (Stuttgart, 1985). Livy Per. = Livy, Periochae, ed. P. Jal as Abrégés des livres de l’histoire romaine de Tite-Live, (Paris, 1984). Pliny N.H. = Pliny the Elder, Naturalis historia/Natural HistoryHGWU+5DFNKDPet al. (Loeb, 1938–63). Plut. Luc. = Plutarch, Vitae parallelae, Lucullus, ed./tr. B. Perrin, vol. II (Loeb, 1914). Plut. Pomp. = Plutarch Vitae parallelae, Pompeius, ed./tr. B. Perrin, vol. V (Loeb, 1917). Pomponius Mela = Pomponius Mela, De situ orbis, tr. F.E. Romer as Pomponius Mela’s Description of the World (Ann Arbor, 1998). Ptolemy = Ptolemy, Geographia, ed. A. Stückelberger and G. Grasshoff (Basel, 2006). Sen. Med. = Seneca, Medea, ed./tr. J.G. Fitch, vol. I (Loeb, 2018). Sen. Thy. = Seneca, Thyestes, ed./tr. J.G. Fitch, vol. II (Loeb, 2018). Serv. Aen. = Servius, Commentary on the Aeneid of VergilHG*7KLORDQG++DJHQ vol. I (Leipzig, 1881). 6+$Had. = 6FULSWRUHV+LVWRULDH$XJXVWDH, Hadrian, ed./tr. D. Magie, vol. 1 (Loeb, 1921). Suet. Ner. = Suetonius, Nero, ed./tr. J.C. Rolfe (Loeb, 1914), vol. 2.

xii

FROM ALBANIA TO ARRAN

Strabo = Strabo, Geographica, ed. S. Radt (Göttingen, 2002– WU+-RQHV /RHE 1930). Tac. Ann. = Tacitus, Annales/Annals, tr. C. Damon (London and New York, 2012) Tac. His. = Tacitus, Historiae, ed. R. Ash and tr. K. Wellesley (London, 2009). Vell. Pat. = Velleius Paterculus, Historiae, ed./tr. F.W. Shipley (Loeb, 1924). Muslim (Arabic and Persian) %DOćGKXUĪ = $ʘPDG E &DXFDVLDQ$OEDQLDDQG'DJHVWDQ+LVWRULFR-Geographical and Politico-Administrative Aspects], Albania Caucasica 2015, 28–42; G.S. Svazyan, “Snova o iuzhnoi granitse Albanii (I v. do n.e. – ser. V v. n.e.)” [Once Again on the Southern Border of Albania (1st c. BCE– Mid-5th c. CE)], Albania Caucasica 2015, 48–56. Another example of a historical geographical approach to Albania comes from the work of S.N. Murav’ev: “Ptolemeeva karta Kavkazskoi Albanii i uroven’ Kaspiia” [The Ptolemian Map of Caucasian Albania and the Level of the Caspian], Vestnik drevnei istorii 163, no. 1 (1983), 117–47; id., “Problem Araksa-Tanaisa-Iaksarta i uroven’ Kaspiia v 6th-3rd vv. do n.e.” [The Problem of the Aras-Tanis-Jaxartes and the Level of the Caspian in the 6th-3rd c. BCE], in Mathesis. Trudy seminara po istorii antichonoi nauki i filosofii (Moscow, 1991), 115–5++HZVHQq0XUDYLHYpV7KHVLVRQWKH+LVWRULFDO*HRJUDSK\RI Southeastern Caucasia,” The Annual of the Society for the Study of Caucasia 2 (1990), 51–+RZ ever, the modelling of Caspian Sea levels that underpin these arguments should be regarded with deep caution – see A.V. Kislov et al., “Current Status and Palaeostages of the Caspian Sea as a Potential Evaluation Tool for Climate Model Simulations,” Quaternary International 345 (2014), 48–55. 26 For an example of polemical discourse, compare the following two works: Mamedova, Politicheskaia istoriia >3ROLWLFDO +LVWRU\@ –31; A.A. Akopyan et al., “K izucheniiu istorii Kavkazskoi Albanii: po povodu knigi F. Mamedovoi ‘Politicheskaia istoriia i istoicheskaia geRJUDILLD.DYND]VNRL$OEDQLLpr>&RQFHUQLQJWKH6WXG\RIWKH+LVWRU\RI&DXFDVLDQ$OEDQLDRQ WKH%RRNRI)0DPHGRYDo7KH3ROLWLFDO+LVWRU\DQG+LVWRULFDO*HRJUDSKy of Caucasian Albania’”], in P. Muradyan, ed., K osveshcheniiu problem istorii i kul‘tury Kavkazskoi Albanii i vostochnykh provintsii Armenii (Yerevan, 1991), 321–52. 27 Chiefly Strabo (11.2–5 and 14) and Ptolemy (5.8, 5.10–13), along with descriptions in Pliny the Elder (N.H. 6.28–31; 6.38–40) and briefer mentions in Dio (35.54–55; 37.3–5), Plutarch (Pomp. 34–36), Tacitus (Ann. 6.33), and Pomponius Mela (3.34). See Bais, Albania caucasica, 149–84 for a full overview. 28 The debates about ancient territorial extent have become firmly embedded in modern discussions of national boundaries; for context see N. Dudwick, “The case of the Caucasian Albanians: Ethnohistory and ethnic politics,” &DKLHUVGXPRQGHUXVVHHWVRYLÆWLTXH, 1990, 377–83. There is now a sizeable body of literature on the entanglements of nationalism and archaeology, both in the Caucasus and beyond, see for example P.L. Kohl and C. Fawcett, eds., Nationalism, Politics and the Practice of Archaeology (Cambridge, 1995); M.S. Gadjiev et al., “Mythologizing the Remote Past for Political Purposes in the North Caucasus,” in B. Grant and L. Yalcin-+HFNPDQQ HGV Caucasus Paradigms: Anthropologies, Histories and the Making of a World Area (Berlin, 2007), 119–42. (1st

ALBANIA IN GREEK AND LATIN TEXTS

17

history, and present a challenging and internally inconsistent body of evidence. 29 Although the sources likely reflect changing conditions in a land of fluid borders, the strident debates have left little room for ambiguity or uncertainty. Our classical sources on Albania are not simply a compendium of neutral data points that have happened to survive the vicissitudes of history and are ready to be strung into a representative whole. They are, instead, the products of the western gaze of their individual authors, shaped by understandings of concepts like ‘territory’ and ‘ethnicity,’ and presented in the context of precise historical HYHQWV$V+HZVHQULJKWO\ pointed out, the modern questions about ancient borders “will not be settled by scholars rummaging about in the fragments of data which have come down to us on the ethno-history of southeastern Caucasia two millennia ago.” 30

HISTORY AND MYTH IN THE EASTERN CAUCASUS Historical Framework Most classical references to Albania come in the course of descriptions of a relatively limited number of Roman historical events. Albania appears most often in accounts of Pompey’s march through the Caucasus in 66–65 BCE, a planned (but unrealized) Roman campaign aimed at the Caucasus ca. 68 CE, and a sequence of military encounters among local dynasts, mobile pastoralists, and imperial powers in the first two centuries CE. Many of these episodes occurred during the long-running military and political conflict between Rome and the Arsacid dynasty of Parthia, and specifically their struggles for power in Armenia. Descriptions of Albania do not occur because of inherent Roman interest in the Caucasus or Albania itself, but rather because of the way that local histories interacted with Roman military activities. There is very little information to contextualize the relationship between Albania and Rome or how it was understood by local political authorities and residents, who

29

E.g. at one point, Pliny (N.H. 6.29) writes that the Albanians dwell in planitiem omnem a Cyro (“WKHHQWLUHSODLQIURPWKH.XUD>QRUWKZDUG@”), which is a straightforward association use of the river as a border. Elsewhere, he writes (N.H. 6.39) that flumina per Albaniam decurrunt in mare &DVXVHW$OEDQXVGHLQ&DPE\VHVLQ&DXFDVLVRUWXVPRQWLEXVPR[&\UXVLQ&RUD[LFLV(“The rivers running down to the sea through Albania are the Casus and the Albanus, then the Cambyses, which rises in the Caucasus Mountains, and then the Kur, ULVLQJLQWKH&RUD[DFL”). By calling the Kura one of the rivers that runs per Albaniam, Pliny seems to be implying that it is not a border. Bais resolves this inconsistency by arguing that, “a river dividing two territories can be taken to belong to either of them,” (“The Southern Border,” 343), which is an interesting argument, but is not definitive. 30 5++HZVHQq(WKQR-+LVWRU\DQGWKH$UPHQLDQ,QIOXHQFHXSRQWKH&DXFDVLDQ$OED nians,” in Classical Armenian Culture: Influences and Creativity, ed. T.J. Samuelian (Philadelphia PA, 1982), 35.

18

LARA FABIAN

often seem as much allies of the Arsacid Empire as the Roman Empire. 31 In many cases, Albania is mentioned by only one or two authors in conjunction with a specific historical episode, leading to questions about why some authors chose to include details about the distant and unfamiliar land, while others omitted the space. An understanding of the place of the Caucasus within the mytho-geographic landscape of antiquity provides some context for these authorial choices. General Mytho-Geographic Tropes As mentioned in the introduction, there are several central tropes embedded within Greek and Roman descriptions of the Caucasus: that the Caucasus mountains were the site of Prometheus’ punishment; that the region marked the farthest reaches of Jason’s journey with the Argonauts; and that the mountains were the home of the Amazons. 32 Although the first two of these myths were anchored in the Pontic basin, the stories developed tendrils reaching far to the east. Thus we hear in one account that the Albanians were originally from Mount Alban on the Italian peninsula, having IROORZHG+HUFXOHVHDVWDQGWKDWWKLVFRQQHFWLRQHQDEOHG-DVRQWRPDNHDWUHDW\ZLWK them (Justin 42.3.4).33 Greek legendary accounts also connect the founding of Armenia to companions of Jason. 34 We have good reason to suspect, therefore, that these mythological narratives had broad currency in Mediterranean perceptions of the Caucasus. As David Braund has argued, the general mythological landscape highlights the Caucasus’ “curiosity” – its oddity and deviation from the social norms. 35 This is a place where gold flows in the rivers, and where the mountains are the domain of the profoundly unnatural women warriors. At the same time, the myths also frame the Caucasus as a space of adventurism and exploration. Following the pattern of Jason and the Argonauts, anyone who ventured there existed within a continuum of mythological campaigns of heroes and gods. 36 Finally, by the time of the Roman Empire, mere mention of the Caucasus served as shorthand for authors looking to describe a EOHDNDQGKDUVKODQGVFDSH +RUCarm. 1.22.7; Sen. Med. 43; Sen. Thy. 1048; Serv. Aen. 4.367). 31

The archaeological and textual sources for Iberia are more numerous. For a recent overview see A.E. Furtwängler et al., eds., ,EHULDDQG5RPH7KHH[FDYDWLRQVRIWKHSDODFHDW'HGRSOLV Gora and the Roman influence in the Caucasian Kingdom of Iberia (Langenweissbach, 2008). 32 For an overview of the vast classical source material referencing these tropes see Braund, “Caucasian Frontier.” 33 See L.E. Patterson, “Pompey’s Albanian Connection at Justin XLII, 3,4,” Latomus 61.2 (2002), 312–25. On kinship myth in the Caucasus more broadly see Braund, “Caucasian Frontier,” 40–41. 34 G. Traina, “Traditions on Armenia in Submerged Greek Literature: Preliminary Considerations,” in G. Colesanti and L. Lulli, eds., Submerged Literature in Ancient Greek Culture. Vol. 2: Case Studies, (Berlin, 2016), 11–112. 35 Braund, “Caucasian Frontier,” 41. 36 Ibid., 39

ALBANIA IN GREEK AND LATIN TEXTS

19

The region’s physical reality – its mountainous topography – seems to play an outsized role in all of these presentations. In classical geographical thought more generally, highlands served a tropic function as “an alter-ego of urban civilization.” 37 Lowland and highland peoples were understood to occupy fundamentally different places within society; the former as standard bearers of civilization, and the latter as war-loving bandits threatening civilization. 38 Considering the evolution of the Caucasus in Greek geographic thought, it is noteworthy that the earliest extant descriptions of it are nothing more than discussions of its mountainous character. Early Historical Accounts on the Eastern Caucasus General reports about the region appear from the sixth century BCE onward in the works of the Ionian geographers as well as Greek poets. 39 7KHDFFRXQWVRIERWK+H rodotus (1.203) and Aristotle (Meter. 1.12 350a 26–36) capture the massive nature of the Caucasus range, situating them as the largest mountains of antiquity and placing them next to the Caspian Sea. 40 The understandings of both mountains and sea, however, are imprecise. In the case of the Caspian, for instance, the sources reflect a deep and ongoing uncertainty about whether the body of water was an inland sea or a gulf of the great ocean thought to surround the inhabited world. 41 Thus, precise and specific information about Caucasia is missing in the earlier &ODVVLFDODQG+HOOHQLVWLFDFFRXQWVHYHQRQWKHOHYHORIFRQFUHWHXQGHUVWDQGLQJVRI geography and topography. 42 Rare accounts of local populations and socio-political structures provide even fewer details about the people inhabiting the landscape. 37

%0HL¼QHUq9RUVWHOOXQJHQGHU*ULHFKHQYRQGHQ%HUJHQrLQ(2OVKDXVHQDQG+ Sonnabend, eds., Gebirgsland als Lebensraum (Amsterdam, 1996), 369. 38 %'6KDZq%DQGLW+LJKODQGVDQG/RZODQG3HDFH7KH0RXQWDLQVRI,VDXULD-Cilicia,” Journal of the Economic and Social History of the Orient 33 (1990), 199–233. 39 Also, +HFDWDHXV BNJ 1 F 291). See Meißner “A Belated Nation: Sources on Ancient Iberia and Iberian Kingship,” AMIT 32 (2000), 177–78, for a discussion of the early textual sources, with a focus on Iberia. 40 6HH$+HUUPDQQq ¸ŧÁ¸ÊÇË (3),” RE (Pauly-Wissowa), XI.1 (Stuttgart, 1921) for references. 41 On ancient debates about the Caspian VHH+-*HKUNHq7KHo5HYROXWLRQpRI$OH[DQ der the Great: Old and New in the World’s View,” in S. Bianchetti et al., eds., Brill’s Companion to Ancient Geography. The Inhabited World in Greek and Roman Tradition (Leiden, 2016), 91–93; J.O. Thompson, History of Ancient Geography (New York, 1965), 79–86. The confusion of ancient authors on this point echoed long past antiquity, into the work of medieval Islamic geographers; see W. Raczka, “A Sea or a Lake? The Caspian’s Long Odyssey,” Central Asian Survey 19.2 (2000), 189–221. 42 See the development of understandings of the Caucasus mountains and the Aras river, which in both cases only crystalized in the Roman period: A. Dan, “Les Leukosyriens: quelques notes d’ethnographie sinopéenne,” Ancient Civilizations from Scythia to Siberia 16.1 (2010), 73–102; G. Traina, “La découverte de l’Araxe,” in A. Dan and S. Lebreton, eds., Étude des IOHXYHVGp$VLHPLQHXUHGDQVOp$QWLTXLWÆTome II (Arras, 2018), 235–44.

20

LARA FABIAN

+HURGRWXVQRWHVVHYHUDOJURXSVZKRPLJKWEHSODFHGQHDUWKHFHQWUDORUHDVWHUQ&DX casus: one group called the Saspeirians (1.104) are said to dwell between the Medes and the Colchians. The Caspian coast is described as the home of a multitude of peoples (1.203), with one group called the Caspians (3.92), perhaps connected to the eastern Caucasus, although their geography is confused.43 Albanians do not appear explicitly in &ODVVLFDO*UHHNRU+HOOHQLVWLFDFFRXQWV 44 Albania on the World Stage Direct mention of the Albanians and Albania appear only in texts composed in the very late first century BCE, under the Roman Empire. This likely reflects the growth of interest in this region during this period, which resulted in a larger number of textual accounts. But the lack of earlier sources is due also to the idiosyncrasies of textual transmission, producing lacunae that do not reflect reality. Indeed, these ‘later’ accounts regularly reference historical events from earlier periods, and provide sources for their information. 45 ,WLVFOHDUWKDWE\WKHODWH+HOOHQLVWLFSHULRG LIQRWHDUOLHU WKH eastern Caucasus had come into the orbit of authors and authorities in the Mediterranean world. Furthermore, the evidence suggests that a group known as the

43

Meißner, “A Belated Nation,” 178 n. 8, for a discussion of these two groups and the confusions surrounding their identification and localization. For a discussion of other groups LQWKLVUHJLRQLQWKH+HOOHQLVWLFSHULRGVHH56\PHq7KH&DGXVLLLQ+LVWRU\DQGLQ)LFWLRQr The Journal of Hellenic Studies 108 (1988), 137–50. 44 Later sources, however, describe interactions between the Albanians and the Caspians. Strabo seems to describe the transformation of the Caspian territory from being independent (Eratosthenes apud Strabo 11.8.8), to being a part of Albania (Theophanes apud Strabo 11.4.5). These accounts have led some scholars to propose that the Albanians began as a subgroup of the Caspians, going so far as to call the early population the ‘‘Albano-Caspians’’ (Trever, Ocherki [Studies], 52). 45 Quellenforschung has been a preoccupation for scholars discussing Strabo’s description of the South Caucasus. See A.I. Boltunova, “Opisanie Iberii v ‘Geografii’ Strabona” [Description of Iberia in Strabo’s Geography], Vestnik drevnei istorii no. 4 (1947), 142–60; Trever, Ocherki [Studies], 1–11; K.G. Aliev, “K voprosu ob istochnikakh Strabona v opisanii drevnei Kavkazskoi Albanii” [On the Question of the Sources of Strabo in his Description of Caucasian Albania], Doklady AN AzSSR 16.4 (1960), 419–22; Babaev, Goroda [Cities], 52; M.S. Gadjiev, “Strabon o karavannoi torgovle aorsov” [Strabo about the Caravan Trade of Aorses], in Dorogo Strabona kak chast’ Velikogo Shelkovo puti (Samarkand-Tashkent, 2009), 25–30. Most draw on K.J. Neumann, Strabons Quellen im elften Buche. (Leipzig, 1881); K.J. Neumann, “Strabons Landeskunde von Kaukasien: eine Quellenuntersuchung,” Jahrbücher für classische Philologie, Supplementband 13 (1883), 322–354; W. Fabricus, Theophanes von Mitylene und Quintus Dellius als Quellen der Geographie des Strabon (SWUDXVVEXUJ (+RQLJPDQQq6WUDERQ  rRE (PaulyWissowa), IVA.1 (Stuttgart, 1931); and M. Dubois, ([DPHQGHODJÆRJUDSKLHGH6WUDERQ(WXGHGHOD methode et des sources (Paris, 1891).

ALBANIA IN GREEK AND LATIN TEXTS

21

Albanians was active in some way in this space at the time, though the socio-political nature of the Albanians in this period is not clear from the sources.46 This uncertainty stems from the fact that early evidence about the Albanians is exceedingly scarce. The first appearance of the Albanians on the world stage dates to the late fourth century BCE, preserved most directly in Arrian’s second-century CE work Anabasis    +HUH we read that the Albanians participated in the fighting force of the Median satrap Atropates on the side of the Persians during the Battle of Gaugamela (331 BCE). Following this episode, there is a brief mention in Pliny of an anonymous Albanian king who gifted a dog to Alexander the Great (N.H. 8.149). 47 Then, there is a period of some two centuries during which the Albanians are not mentioned by any extant sources. They reappear in the historical record during the early first century BCE, in Plutarch’s account of Lucullus’ time in Armenia (Luc. 26), appearing more frequently thereafter. As noted above, most of the references to Albania relate to moments when local histories intersected with Roman military activity, but they also seem to occur in contexts where a sense of Argonautic adventurism is appropriate. Following the discussion of Lucullus, the next sustained accounts of interaction describe the march of Pompey through the Caucasus in 65 BCE. This episode is the subject of more classical mentions of Albania than any other historical moment. 48 Although there is good reason to believe in its historicity, the frequent appearance of the Albanians in these particular accounts also reflects a more general sense within the historiographic tradition that compared Pompey’s actions to Jason’s. The nine ancient accounts of the episode generally depict the Iberians and the Albanians working in concert to repel the foreign Roman army. Despite the conceptual pairing of the Albanians and Iberians in these accounts,49 several later reports provide elusive glimpses into local strife within the territory in the form of Albanian46

Debates about the date of the emergence and consolidation of Albania as a formal political unit have been widespread in historical and archaeological literature. For an opinion WKDWSODFHVWKLVGHYHORSPHQWLQWKHHDUOLHU+HOOHQLVWLFSHULRGVHH,$%DEDHYq.YRSURVXR vozniknovenii gosudarstva Albanii (Kavkazskoi)” [On the Question of the Rise of the State of Caucasian Albania], Izvestiia AN AzSSR: SIFP no. 4 (1976), 40–51. For an opinion that the SURFHVVKDSSHQHGODWHULQWKH+HOOHQLVWLFSHULRGVHH7UHYHUOcherki [Studies], 144. See also Bais, Albania caucasica, 69–72. 47 On this passage and questions surrounding it, Traina, “Due note,” 318–19. 48 Livy Per. 101.4; Justin 42.3.4; Vell. Pat. 2.40.1; Front. Str. 2.3.14; Plut. Pomp. 34–36.2; Flor. 1.40.21; App. Mithr. 103; Dio 36.54, 37.1–6; Eutr. 6.14. This is perhaps the most discussed episode in Roman-Albanian relations among contemporary scholars as well. See particularly the explorations in Dreher, “Pompeius,” and M.-L. Chaumont, “Expédition de Pompée le Grand en Arménie et au Caucase: 66–65 av. JC,” Quaderni Catanesi di Studi Classici e Medievali 6.11 (1984), 17–94. 49 With the two described by Plutarch (Pomp. 34.1) as the “most important” populations in the region (ÄšºÀÊ̸»ò¸ĤÌľÅ ëÊÌÀÅ ì¿Å¾ ¹¸ÅÇĖÁ¸Ėa¹¾É¼Ë). On the conceptual pairing of Albania and Iberia see Traina, “Roman Representations,” 43–44.

22

LARA FABIAN

Iberian violence, which is attested in the first century CE (Tacit. Ann. 12.45.1; Dio 49.24.1–2 DQGDOVRLQWKHVHFRQG 6+$Had. 17.10–12; Dio 69.15.1–2). 50 Fluctuating patterns of allegiance are also a central feature of descriptions of the relationships between the Albanians and their neighbors to the north, a steppe-connected population often described broadly as ‘Sarmatian,’ or with increasing specificity as ‘Alan,’ over the course of the first and second centuries CE. 51 The accounts concerning these northern neighbors in the first and second century describe raids of mobile pastoralists targeting Armenia, Parthia and Roman Anatolia, as well as Albania itself, occurring in 35 CE (Tacit. Ann. 6.33–35; Joseph. A.J. 18.96–97), ca. 72 CE (Joseph. B.J. 7.244– 251), and 135 CE (Dio 69.15.1–3). 52 Furthermore, both the Albanians and the Alans are also frequently mentioned by modern scholars in connection with military activity planned by Nero in the Caucasus in 68 CE (Pliny N.H. 6.40; Tacit. His. 1.6.2; Suet. Nero 12.2; Dio 63.8.1) – another moment when the Caucasus becomes a territory for adventurism. 53

50

On the question of these Albanian-Iberian conflicts see Juntunen, “Pharasmanes and the Iazyges”; A.S. Schieber, The Flavian Eastern Policy (PhD; University of New York at Buffalo, 1975), 85; 96ff; E.L. Wheeler, Flavius Arrianus: A Political and Military Biography (PhD; Duke University, 1977), 111; 122–23; 242–44. 51 On sources about the Alans see A. Alemany, Sources on the Alans: A Critical Compilation (Leiden, 2000). The similarity of the terms alanoi and albanoi have created considerable confusion. See D. Braund, “Map 85: Oudon-Rha,” in R.J.A. Talbert, ed., Barrington Atlas of the Greek and Roman World (Princeton, 2000), 1213, for an argument that scholars are over-confident in our ability to untangle the textual uncertainties concerning these two groups, and see n. 20 for an example of such confusion. 52 Discussions of these raids are scattered, see Bais, Albania caucasica, 82, 88, 93–96, for summaries as well as bibliography. See also A.B. Bosworth, “Vespasian’s Reorganization of the North-East Frontier,” Antichthon 10 (1976), 63–78; Bosworth, “Arrian and the Alani”; F. Carrata Thomes, Gli Alani nella politica orientale di Antonino Pio 7XULQ 'ċEURZDq3RP pey to Domitian”; Giardina, “Roma e il Caucaso,” 131ff; Ia. Kharmatta, “Iz istorii alano-parILDQVNLNKRWQRVKHQLLr>)URPWKH+LVWRU\RI$ODQR-Parthian Relations], $FWD$QWLTXD$FDGHPLDH Scientiarum Hungaricae 13 (1965), 127–47; S.M. Perevalov, “Alanskii nabeg 136 g. n.e. v strany Zakavkaz’ia: problemnye voprosy” [The Alan Raid in 136 CE into the Countries of the Transcaucasus: Problematic Issues], in Antichnaia tsivilizatsiia i varvary (Moscow, 2006), 318; Trever, Ocherki [Studies],130–31; A.A. Tuallagov, Alany Pridar’ial’ia i zakavkazskie pokhody I–II vv. [The Alan Predarial and Transcaspian Campaigns of the 1st and 2nd Centuries] (Vladikavkaz, 2014); Wheeler, “Flavius Arrianus,” 100–101, 124–30, 227ff. 53 The connection of the Alans to this planned campaign is not supported in the sources. +RZHYHUVHHMommsen, Römische Geschichte, V, 394 n. 1, for an emendation to Tacitus’ account, changing albanoi to alanoi and thus shifting the target of the campaign to the North Caucasus group. Although this emendation is not widely accepted today, the perception that the steppe people were somehow involved in Nero’s plans (even if not as the main target) has remained. For recent discussions of this episode that provide further bibliography see Gadjiev, “O Kaspiiskom pokhode”; Kolendo, “Le projet d’expédition.”

ALBANIA IN GREEK AND LATIN TEXTS

23

Finally, despite the passing nature of Roman interest in the space, there are occasional details that allow for more detailed consideration of Albania’s internal political affairs. For instance, the Roman sources on the campaign of Pompey provide the earliest reference to a specific Albanian king, Oroises, who mounted opposition against the Roman forces (Dio 36.54). 54 Evidence from the time of Augustus, meanwhile, suggests that (from a Roman perspective at least), the Albanians and Iberians had sought amicitia with Rome, and were therefore participants in a formal political relationship with the Roman Empire (RG 31). 55 The discovery of a Latin military inscription near the petroglyph site of Qobustan in 1948 provided important evidence in discussions of Albanian-Roman interaction in the period of Domitian. 56 It bears witness to the presence of Roman soldiers along the Caspian coast, although the reasons for their presence remain a cause for debate. 57 Scattered accounts, meanwhile, attest to continuing diplomatic relationships between the Albanians, Iberians and Romans through the second century CE, particularly in the form of official visits and gift-JLYLQJ 6+$Hadr. 21.13). 58

SOCIO-POLITICAL FLEXIBILITY IN THE CASE OF THE ALBANIANS There are a handful of geographic and ethnographic descriptions of the eastern Caucasus that paint a more complicated picture of local socio-political configurations. 59 The most extensive of the ethnographic accounts, that of Strabo’s Geography (11.4), was composed in the very late first century BCE or early first century CE, incorporating earlier sources. Strabo had more reason than most Roman authors to proffer 54

Another king, Zober, is attested in Dio (49.24.1). nostrum amicitiam appitiverunt per legatos… $OEDQRUXPTXHUH[HW+LEHURURXP There are also several mentions of kings being “given” to the Albanians by the Romans (Eutr. 8.3; Fest Brev. 20), but these late accounts are very thin evidence for such direct political involvement. 56 I.M. Dzhafarzade, “Drevnelatinskaia nadpis’ y podoshvy gory Beiukdash” [An Ancient Latin Inscription from the Foot of the Beiukdash Mountain], Doklady AN AzSSR 4.7, (1948), 304–11; E.A. Pakhomov, “Rimskaia nadpis’ 1 v. n.e. i legion XII Fulminata” [A Roman Inscription of the 1st century CE and the 12th Legion Fulminata], Izvestiia AN AzSSR no. 1 (1949), 79–88. See also discussions in Bosworth, “Vespasian’s Reorganization,” 75; D. Braund, “Notes from the Black Sea and Caucasus: Arrian, Phlegon and Flavian Inscriptions,” Ancient Civilizations from Scythia to Siberia 9.3 (2003), 189–90; Grosso, “Aspetti della politica,” 117ff. 57 See Braund, “Notes from the Black Sea and Caucasus,” 190, for a brief consideration of the inscription. 58 The evidence for diplomatic visits is clearest in the case of Iberia, as epigraphic evidence from Ostia confirms a visit of the Iberian king Pharasmanes during the reign of Antoninus Pius. FRUFLWDWLRQVDQGDGLVFXVVLRQVHH'%UDXQGq+DGULDQDQG3KDUDVPDQHVrKlio 73 (1991), 208–19. 59 The central account is that of Strabo, discussed below. Other reports of regional geography and social organization come from Ptolemy (5.8–12) and Pliny (N.H. 6.28–31, 6.38– 40), while Aelian’s De natura animalium (17.32–33) provides a description of some of the wildlife of the Caspians. 55

24

LARA FABIAN

VRXQGLQIRUPDWLRQDERXWWKH&DXFDVXV+HZDVDQDWLYHRI$PDVHLDLQ$QDWROLDZKLOH his uncle had been a Pontic governor in Colchis under Mithridates (11.2.18). Discussions of Strabo’s text have formed the backbone of many of the analyses of Albania, with specific attention paid to highlighting, as well as speculating on the causes of, supposed factual errors in his report, that could be compared with archaeological data. 60 Equally important, however, is reading these sources with an eye to Strabo’s presentation of space and ethnicity. 61 Strabo’s treatment of the Albanians comes after his discussion of the Iberians (11.3), and immediately before that of the Amazons (11.5). Strabo comments that a curious mixture of myth and reality is at play in descriptions of the Amazons, saying “…our accounts of other peoples keep a distinction between the mythical and the historical elements…. But as regards the Amazons, the same stories are told now as in early times, though they are marvelous and beyond belief” (11.5.3). In juxtaposing the Albanians with the Amazons, then, Strabo makes neighbors of myth and history. Indeed, Strabo’s entire presentation of the Albanians plays with this theme, blurring the conceptual line between a straightforward ‘civilized’ group and one belonging to the world of the ‘barbarians’ beyond. 62 On the one hand, allusions to the mythical Golden Age are prominent in Strabo’s account of the Albanians. They are inefficient cultivators living a cyclopean life, having no need for the sea and inhabiting a region so fertile that it produces a great bounty in spite of them (11.4.3). They allow fruit to wither without being harvested and they use only unworked pieces of wood as tools. Their cultural inadequacy is supported by the fact that they do not participate in economic activity (11.4.4), and that they are “closer to nomadic stock” than their neighbors the Iberians (11.4.1), who in contrast have proper markers of settlement sophistication (11.3.1). On the other hand, despite these characteristics of paradigmatic primitiveness, Strabo asserts that the Albanians were fundamentally different from their nomadic neighbors. If supplied with “men to lead them,” the Albanians (along with the Iberians and Armenians) made excellent subjects of the Romans, whereas the nomads were impossible to control (6.4.2). It is tempting to connect this comment with Strabo’s observation that the Albanians had recently been united under a single king, having 60

I.G. Aliev, “K interpretatsii paragrafov 1.3.4 i 5, 4 glavy 11 knigi ‘Geografii’ Strabona” [On the Interpretation of Paragraphs 1, 2, 3, 4 and 5 of Section 4 of the Eleventh Book of Strabo’s “Geography”], Vestnik drevnei istorii 133, no. 3 (1975), 150–65; Mamedova, Politicheskaia istoriia >3ROLWLFDO+LVWRU\@–20; Trever, Ocherki [Studies], 147. 61 For Strabo on related areas see E. Ilyushechkina, “Strabo’s Description of the North and Roman Geo-3ROLWLFDO,GHDVrDQG*7UDLQDq6WUDERDQGWKH+LVWRU\RI$UPHQLDrERWK in D. Dueck, ed., The Routledge Companion to Strabo (London, 2017), 60–69 and 93–101. On Strabo’s geographic presentation more generally see D. Dueck, “The Geographical Narrative of Strabo of Amasia,” in K.A. Raaflaub and R.J.A. Talbert, eds., Geography and Ethnography: Perceptions of the World in Pre-Modern Societies (Malden MA, 2009), 236–51. 62 On Strabo’s representation of the interaction between these poles of social organization see Dueck, “The Geographical Narrative,” 243.

ALBANIA IN GREEK AND LATIN TEXTS

25

before been ruled by separate kings for each of the 26 local language groups (11.4.6). Although this passage is usually discussed in the context of dating the consolidation of the Albanian federation, 63 the socio-cultural implications bear further thought. Although we have no idea who these 26 tribes were, Strabo and the textual traGLWLRQJRLQJDOOWKHZD\EDFNWRWKHGHVFULSWLRQLQ+HURGRWXVVXJJHVWVDGHQVHQHWZRUN of multiple populations in the eastern Caucasus. The picture is striking, furthermore, in that it presents a variety of populations with ties to the steppe coexisting alongside the Albanians from the middle of the first century BCE onward. 64 To name just a few of these steppe-connected neighbors, we find the Udini, said by Pliny to be a Scythian people living just south of Albania (N.H. 6.38),65 as well as the Sakacene (Strabo 11.8.4, 11.14.4; Arr. Anab. 3.8.4, Ptolemy 5.12), who according to Strabo trace their roots to the Saka of Central Asia, appearing in the Caucasus in the course of Saka raids from Bactria (11.8.4). There are also the Aorsi, another steppe group (Strabo 11.5.8; Pliny N.H. 6.39), 66 said by Strabo to run Caspian coastal trade routes. And finally, the land just north of Albania was the domain of the Alans,67 said to have descended from either the Massagetae (Dio 69.15.1; Amm. Marc. 23.5.16), the Scythians (Joseph. B.J. 7.244; Arr. Ect. 26) or the Sarmatians (Marcianus 2.39). We should suspect that these groups, defined by the etic sources in strict categorical terms, were likely both more contingent and more overlapping than the 63

Traina, “Due note,” 318–19. On classical perceptions of nomads see B.D. Shaw, “Eaters of Flesh, Drinkers of Milk: The Ancient Mediterranean Ideology of the Pastoral Nomad,” Ancient Society 13–14 (1982), 5– 31. 65 For an overview of sources on the Udini and debates, see Bais, Albania caucasica, 54– 55. The Udini/ Udi/ Uti linguistic evidence from recent decades suggests that there are close relationships between the modern East Caucasian language Udi, still spoken in several pockets of Azerbaijan, and the late antique language of Caucasian Albanian known from inscriptions as well as later palimpsests. On the linguistic issues see W. Schulze, “From Caucasian Albanian to Udi,” Iran and the Caucasus 19.2 (2015), 149–77. Debates remain, however, about the Udini/ 8GL8WLRIDQFLHQWVRXUFHV&RPSDUH+HZVHQq(WKQR-+LVWRU\rWR-$*UHSSLQq7KH Language of the Caucasian Albanians,” Folia Slavica 5 (1982), 163. 66 For more on the Aorsi see M.J. Olbrycht, “Die Aorser, die Oberen Aorser und die Siraker bei Strabon. Zur Geschichte und Eigenart der Völker im nordostpontischen und nordkaukasischen Raum im 2.-1. Jh. v. Chr.,” Klio-Beiträge zur Alten Geschichte 83.2 (2001), 425–50. 67 Literature about the Alans from both textual and archaeological perspectives is vast, though mostly focused on later periods of history. For an accessible introduction to some of the questions surrounding this group as well as further bibliography, see M.P. Abramova, “Sarmatians in the North Caucasus,” in J. Davis-Kimball et al., eds., Nomads of the Eurasian Steppes in the Early Iron Age (1995), 165–84. For a recent consideration of Alans in the South Caucasus on the basis of archaeological material, see A. Sagona et al., “Alans in the Southern Caucasus?,” in F.J. Vervaet et al., eds., (XUDVLDQ(PSLUHVLQ$QWLTXLW\DQGWKH(DUO\0LGGOH$JHV &RQWDFWDQG([FKDQJHEHWZHHQWKH*UDHFR-Roman World, Inner Asia and China (Cambridge, 2017), 205–50. 64

26

LARA FABIAN

sources attest.68 The Albanians, furthermore, are not described as outsiders to these underlying local networks. Instead, their ties to the mobile pastoralists are couched in genetic terms, with Strabo reporting that both the Iberians and the Albanians have kinship ties to their nomadic neighbors (11.3.3 and 11.4.5), and cooperate with them in military encounters against outsiders for this reason (11.4.5). This approach to the eastern Caucasus shifts our perspective. We see the area not just as the space of a fixed polity, but instead as a dynamic zone of overlapping populations and identities. It was the site of cohabitation between groups with explicit steppe connections as well as those without, present in the territory for various periods of time. These populations were living side-by-side and interacting both broadly and deeply. The flexibility implied by this system was likely an important factor in enabling diversity of social and political interactions that played out across the Caucasus in antiquity, with networks stretching into Roman and Arsacid space, as well as to the steppe systems beyond.

CENTRAL ISSUES AND ENDURING COMPLEXITIES The Caucasus was a physically distant place for classical authors, although it was familiar from a deep current of mytho-geographic thought that positioned the region as a wild periphery for the performance of heroism. The classical sources discussed in this chapter each approach the question of Albania from a different perspective: there is no one single canonical ‘Roman’ view of this space, despite the shared tropes and understandings. We cannot forget, furthermore, that this evidence represents just a sliver of what must have been a more robust dialogue about the eastern Caucasus among its neighbors, with perspectives from the Iranian plateau and the Eurasian Steppe (to say nothing of local understandings) now lost. Even considered holistically and contextually, the narrow range of classical witness that has been transmitted is poorly suited to address the positivist questions we have tended to ask, about states and borders. The first modern local historian of the area, Bakikhanov, warned us in 1841 that we should not expect to reach comprehensive understanding about the history of this complex space. “This land especially, because of the coming and going of many different peoples, was the place of disturbances, chaos and upheaval, and many books and documents and buildings and artifacts have perished.” And yet, he wrote “if you do not comprehend all of it, it does not mean that you are abandoning all of it.” 69 The ‘views from the west’ in antiquity are a particular piece of the puzzle of ancient 68

For example, speaking about the western reaches of the Caucasus near the city of Dioscurias, Strabo notes that, of 70 tribes present in the area, “the greater part of them are Sarmatian, but all of them are Caucasian” (11.2.16). The acceptance of hyper-schematic understandings of identity in ‘Sarmatian’ studies are more acute because of facile acceptance among archaeologists of textual categorization schemes; see A. Dan, “The Sarmatians: Some 7KRXJKWVRQWKH+istoriographic Invention of a West Iranian Migration,” in F. Wiedemann et al., eds., Vom Wandern der Völker: Migrationserzählungen in den Altertumswissenschaften (Berlin, 2017), 97–134; V. Mordvintseva, “The Sarmatians: The Creation of Archaeological Evidence,” 2[IRUG-RXUQDORI$UFKDHRORJ\ 32.2 (2013), 203–19. 69 Bakikhanov, The Heavenly Rose-GardenWU:0)ORRUDQG+-DYDGL

ALBANIA IN GREEK AND LATIN TEXTS

27

Albania’s history. Moving beyond the tempting desire to string them together into a cohesive narrative, we reach a more comprehensive understanding of Albania if we instead appreciate them for their ambiguity and fragmentation.

2. THE CHRONOLOGY OF THE ARSACID ALBANIANS * M.S. GADJIEV The question regarding the time of the accession of the new Arsacid dynasty to the throne of Caucasian Albania is disputed. Numerous scholars 1 have repeatedly paid attention to this particular problem and have, in general, arrived at two opposing points of view. According to the first view (K.V. Trever, F.J. Mamedova and others), the offshoot of the branch of the Parthian dynasty of the Arsacids, represented by 9Dă‘agan I, began to rule in Albania during the first century CE. This conflicts, however, with the fact that the third king (Uѻnayr) of the new dynasty participated in the Battle of Dzirav (Jirov), also referred to as the Battle of Bagrevand (or Baghrawand), in 371 and so ruled in the fourth century. Such a huge chronological gap is, of course, LPSRVVLEOHDQGZHKDYHQRUHDVRQWRWKLQNWKDWWKHKLVWRULDQ0RYVĔV'DV[XUDQFoL who notes only one king LQWKHSHULRGEHWZHHQ9Dă‘agan I and Uѻna\UQDPHO\9Dă‘Ĕ I, omitted anyone from his list of Albanian Arsacid rulers (1.15). The assignment of the rise to power of the Arsacids in Albania to the first century is based not on the historical facts cited in the sources, but on the fact of the establishment of their power at this time in Armenia and Iberia, on the basis of which conclusions were then drawn (through the principle of analogy) about their accession in Albania. Yet, we have no grounds on which to doubt the chronological sequence and continuity of the rule of the kings of Caucasian Albania from the Arsacid dynasty presented by Dasxuranc‘i. Moreover, the dates provided by him for certain events associated with one or other of the Albanian kings categorically prevents us from accepting the theory of the accession of the Arsacids in Albania during the first century.

7UDQVODWHGIURPqǣǾǼǻǼǹǼDZǶȍǎǾȆǮǸǶDzǼǰǎǹǯǮǻǶǶr Albania Caucasica 1 (2015), 68– 75. 1

For example: M.I. Brosse, S.T. Eremyan, K.V. Trever, K.L. Tumanov, N.A. Akinyan, B.A. Ulubabyan, F. J. Mamedova, R.B. Geyushev, T.M. Mamedov, I.G. Semenov, and others. For an overview of these opinions, see Mamedova, ǝǼǹǶȀǶȅdzǿǸǮȍǶǿȀǼǾǶȍ [Political +LVWRU\], 168–71; I.G. Semenov qǜ ǰǾdzǺdzǻǶ ǰǼȄǮǾdzǻǶȍ ǰ ǘǮǰǸǮǵǿǸǼǷ ǎǹǯǮǻǶǶ ǎǾȆǮǸǶDzǿǸǼǷ DzǶǻǮǿȀǶǶr>The Accession of the Arsacid Dynasty in Caucasian Albania], Albania Caucasica 1 (2015), 57–67.

29

30

M.S. GADJIEV

According to the second point of view (M.I. Brosse, S.T. Eremyan and others), the Arsacids assumed power in Albania at the beginning of the fourth century. The first king was allegeGO\ 6DQDWUXN ZKR LV LGHQWLILHG ZLWK 6DQĔVDQ 2 the king of the Maskut.3 After the death of the Armenian king Trdat III in roughly 330, SanaWUXN6DQĔVDQDFWLYHO\SDUWLFLSDWHGLQWKHVWUXJJOHIRUWKHWKURQHRI$UPHQLDDJDLQVW WKHOHJLWLPDWHKHLU;RVURY,,,QDGGLWLRQWKLV6DQDWUXN6DQĔVDQPHQWLRQHGLQFRQ nection with the developments of the 330s, would, in the eyes of those who propose this reconstruction, have marked the beginning of Maskut rule in Albania as well as of the Arsacid dynasty of Albania. This opinion, advocated by some scholars, is only VXSSRUWHG E\ WKH VWDWHPHQW RI 3oDZVWRV %X]DQG   WKDW q6DQĔVDQ NLQJ RI the Maskut” crossed Albania and invaded Armenia in order to attack the Arsacid Xosrov, ZKRLVGHVFULEHGDVWKHqNLQVPDQrDQGqEURWKHUrRI6DQĔVDQ7KLVLPSOLHVWKDWWKH latter too heralded from the Arsacid family, though this suggestion is not supported by the archaeological evidence. In determining this matter, the archeological materials, which attest to the presence in northeast Albania, to the south of the Darband pass, of Alan Iranian-speaking Maskuts and which find correspondence in the narrative sources, carry primary significance and contradict the theory of the Maskut ancestry of the Albanian Arsacid dynasty. The specific archeological evidence is thousands of kurgans (burial mounds) dating to the mid-fourth – mid-fifth century and concentrated in the coastal parts of southern Dagestan, between the Rubas, Gyulgary-chay, and Samur rivers (the kurgan burials Palasa-sirt, Sugyut, Donguz-nour, Kukhmazkunt). In their funerary rituals, they exhibit two cultural components: on the one hand, they maintain the most intimate and direct connection to the burial monuments of the Alan culture in the central regions of the northern Caucasus (catacombs type I – with T-shaped arrangement of entrance and burial chambers). On the other hand, they are related to funerary traditions of Middle Sarmatian culture (catacombs type IV – with N-shaped arrangements of the entrance and burial chambers, podboy [with underground chamber] and pit tombs), in particular, the kurgan necropolises of northern Dagestan (Lvov tombs, with the exception of Lvov VI, in which catacombs of type I prevail). 4 A

2

7KLVFRQIXVLRQJRHVEDFNWR0RYVĔV;RUHQDFoLERWKKHDQG3oDZVWRV%X]DQGQDUUDWH the killing of St Grigoris and a challenge to Xosrov son of Trdat, but whereas P‘awstos attribXWHVERWKDFWLRQVWR6DQĔVDQq$UVDFLGNLQJRIWKH0DVNXWr 6–7), Xorenac‘i assigns them to one Sanatruk, also an Arsacid (3.3 and 9). Whether two different people have become muddled RUZKHWKHUWKHWZRGLIIHUHQWQDPHVDSSO\WRWKHVDPHSHUVRQLVGLIILFXOWWRVD\>5*+@ 3 I give the most commonly used form of this name, which is a slight simplification of $UP0DVNoXWoLQ$UDELFLWLVUHQGHUHGDV0DVTDʜ7KHQDPHUHIHUVWRDSHRSOHOLYLQJLQWKH area of modern northeast Azerbaijan/southeast Dagestan, and sometimes, especially in Arabic, it is a geographical rather than an ethnic deVLJQDWLRQ>5*+@ 4 M.S. Gadjiev and V.lu. 0DODVKHYqǜǯȋȀǻǼǸȁǹȊȀȁǾǻǼǷǽǾǶǻǮDzǹdzǴǻǼǿȀǶǸȁǾDZǮǻǻȉȃ ǺǼDZǶǹȊǻǶǸǼǰIV– V ǰǰǬǴǻǼDZǼ ǒǮDZdzǿȀǮǻǮ” [On the Ethnocultural Affiliation of Mound Tombs from IV-V centuries in Southern Dagestan], in Ǔ.ǖ. ǘǾȁǽǻǼǰ Ƕ ǾǮǵǰǶȀǶdz ǮǾȃdzǼǹǼDZǶǶ

T+(C+5212/2*2QWKH6WXG\RIWKH/DZLQ&DXFDVLDQ Albania], ǝǾǼǯǹdzǺȉ ǶǿȀǼǾǶǶ, ȂǶǹǼǹǼDZǶǶ, ǸȁǹȊȀȁǾȉ 16.1 (Magnitogorsk, 2006), 249. It is known that the Albanian church headed by Catholicos Mikael (705–42) cursed the family of the Albanian prince (Lx[DQ) Varazo, who married his cousin, while the Iberian catholicos Talilu, who blessed this marriage, betrayed anathema (Dasxuranc‘i, 3.13). 13 PerikhǮnyan, ǜǯȇdzǿȀǰǼ Ƕ ǽǾǮǰǼ [Society and Law], 65.

T+(C+5212/2*WKHGLVWULFWRI@ZKLFKIORZVWKH5LYHU6ĔERƲ southwards by way of Lesser Armenia. 6 ɔɊါ Ɋʎʄɸʗʇ Ɋʉɹɸʍʂɸ, ɸʌʔʂʍʛʍ Ɋʉʏʙɸʍʛ, ʌɼʃʂʘ ɧʗɸʘ, ʌɼʗʂ ɦɸʗʋɸʖʂʏʌ ɸʓ ɘɸʙʆɸʔʏʕ, ʋʂʍʐɼʙ ʘəɸʌʏʘ ʔɸʇʋɸʍɸʙ, ɸʓ ɘʏʙʗ ɺɼʖʏʕါ ʀɾʑɾʖ ɼʙ ɸʔʖʂ ʘɘʏʙʗ ɽɸʋɼʍɸʌʍ ʔɸʇʋɸʍʔ ʇɸʍɼɸʃ ɾ ʂ əɸʌʏʘʟ ɋɸʌʘ ʋɼʛ ɸʔɸʔʘʏʙʛ ɽɹʏʙʍ ɸʎʄɸʗʇʍ Ɋʉʏʙɸʍʂʘ ʏʗ ɿʍɻ ʋɾʒʔ ɾ ʋɼʅʂ ɺɼʖʏʌʔ ɘʏʙʗɸʌ ɼʙ ɘʏʕʆɸʔ ʃɼʗʂʍʍʟ ɟɸʄ ɸʓ ɧʗʜʛ, Ɏʄʍʂ ɺɸʙɸʓ ɸʓ Ɋʉʏʙɸʍ ɺɼʖʏʕʍ, ɼʙ ɭɸʋɹɾʊɸʍ ɸʓ ɘʏʙʗɸʙါ ɼʙ ɿʔʖ ʇɸʗɸʙʏʌ ʍʏʗɸ ɹɼʗɻʍ ɧɸʗɸɽʋɸʍɸʙɸʓ, ʇɸʍɻɼʗʈ ɘʏʙɻʗɸʀ ɺɼʉɸʛɸʉɸʛɸʙ, ɼʙ ʕɸʌʗʛʍ ɸʍɸʑɸʖ ʋʂʍʐɼʙ ʘɘʏʙʗ ɺɼʖါ ʌʏʗʏʌ ʌɼʃʂʘ ʆʏʉʋɸʍɾ Ɍɿʙɺɸʙ ʛɸʉɸʛ, ɸʓ Ɋʉʏʙɸʍ ɺɼʖʏʕ, ɼʙ ɋʂʄ ɺɸʙɸʓ ɸʓ ɘɸʙʆɸʔɸʙ, ɼʙ ʌɼʃʂʘ

4

“Ananias of Shirak,” in EIr. I would like to thank Tim Greenwood for allowing me access to this entry before its publication. 5 Širakac‘i, Geography, 16–27. 6 Širakac‘i, Geography, 57 and 59.

B8/'Ć1$/-RĆ1

41

ʍʏʗɸ, ɠɸʛɾ, ɼʙ ɍɾɺɸʓʏʙ ɺɼʖါ ɸʓ ʏʗʏʕ ʇʏʋɸʍʏʙʍ ɺɸʙɸʓɵ ɸʓ ɦɸʍʂ ɺɼʖʏʕʟ Ɋʌʔ ɸʋɼʍɸʌʍ ʇɸʔʂʍ ʂ ɘʏʕʆɸʔɸʌ, ɼʙ ʄɸʓʍɼɸʃ ʂ ʌɊʉʏʙɸʍ ɺɼʖɵ ɸʍʆɸʍʂʍ ʂ ɘʏʙʗ ɺɼʖါ ʂʔʆ ʂ ʌɼʃʂʘ ʆʏʙʔɾ ɭɸʕɸʉɸʆ ʛɸʉɸʛ Ɋʉʏʙɸʍʂʘ ʋɼʗʈ ʂ ɘɸʙʆɸʔါ ɿʍɻ ʏʗʏʌ ʋɾʒ ɺɼʖʍ ɦɾɹʔʒ [sic] ɻɾʑ ʂ ʇɸʗɸʙʏʌ ɿʔʖ ʚʏʛʗ əɸʌʏʘʟ 7

Širakac‘i’s perspective here is invaluable, as he indubitably receives his information from earlier Greek sources (hence the need to translate ¹¸Åţ¸ DV$øXDQNoIRUKLV Armenian-speaking audience), which he is expressly unable to reconcile with the politiFDOUHDOLWLHVRIKLVRZQWLPH+LVXQGHUO\LQJVRXUFHXQGHUVWDQGV$OEDQLDWRUHIHU exclusively to the territory north of the River Kura, which he clarifies as “the original lands of Albania” (ɹʏʙʍ ɸʎʄɸʗʇʍ Ɋʉʏʙɸʍʂʘ) for his audience who presumably would expect an extended Albania already by the seventh century. Some of the Albanian districts in Širakac‘i’s time appeared in his underlying Greek text as Armenian. In the long recension, Širakac‘i understands Greater Armenia as a sprawling province that reaches as far as the Caspian to the east. Greater Armenia comprised fifteen lands, including the “tenth, Arjax [sic] which lies beyond it [Siwnik‘]; eleventh, [the region of] the city of P‘aytakaran, [extending] to the shore of the Caspian west of the [River] Araxes; twelfth, the land of the Utians, bordering Albania and the River Kura” (ʖɸʔʍɼʗʏʗɻɵ Ɋʗʈɸʄ ʏʗ ʌɼʗʂ ʍʏʗɸ ʆɸʌါ ʋɼʖɸʔɸʍɼʗʏʗɻ ɸʎʄɸʗʇɵ ɬɸʌʖɸʆɸʗɸʍ ʛɸʉɸʛʏʕ ʏʗ ɸʓ ɼɽɼʗɹʍ ɘɸʔɹʂʘ, ʂ ʋʏʙʖʔ Ɏʗɸʔʛɸʌါ ɼʗʆʏʖɸʔɸʍɼʗʏʗɻ ɸʎʄɸʗʇɵ ɡʙʖɾɸʘʏʘ, ʏʗ ɸʓ Ɋʉʏʙɸʍʂʙʛ ɼʙ ɘʏʙʗ ɺɼʖʏʕ). 8 In other words, Armenia stretched even past the Araxes, reaching territories that by the seventh century belonged to Georgia, Albania and Azerbaijan. 9 Širakac‘i is explicit that his Greater Armenia was not relevant to the political claims of his own day. So, for example, after his description of Arc‘ax, 10 he adds “all these [districts] the Albanians have taken from Armenia” (ɽɸʌʔ ɸʋɼʍɸʌʍ Ɋʉʏʙɸʍʛ ʏʙʍʂʍ ʇɸʍɼɸʃ ʂ əɸʌʏʘ). 11 N. Garsoïan sees Širakac‘i’s extended description of Armenia as less historical reality than “the new concept now achieved that Armenia was one single, indivisible entity,” an Armenianness capable of ignoring the vicissitudes of fractured political claims. 12 As Garsoïan herself notes, Širakac‘i acknowledges explicitly in more than one instance that Utik‘ and Arc‘ax belong to the Albanians. But she argues that he tags them as “originally Armenian” because he embraces an extensive definition of

7

Širakac‘i, $x[DUKDFoR\Fo>/RQJ@, 28–29. Širakac‘i, Geography, 59; $x[DUKDFoR\Fo>/RQJ@, 29. 9 In this chapter “Azerbaijan” refers to the late antique/early Islamic province (Pers. ĆGŠUEćGDJćQ, Arab. ĆGKDUED\MćQ), which roughly corresponded to modern northwest Iran and sometimes the southern edges of modern Azerbaijan. 10 Arc‘ax (or Artsakh) does not, to my knowledge, appear as such in Arabic sources, but UDWKHUVLPSO\DV.KćVKĪQ (Arm. ;DăoĔQ); note Mis‘DUE0XKDOKLO$EŠ'XODI Travels in Iran FLUFD$' HGDQGWU9ODGLPLU0LQRUVN\ &DLUR  (QJ  $UDELF KDV.KćMĪQ 11 Širakac‘i, Geography, 65; $x[DUKDFoR\Fo>/RQJ@, 33. 12 Nina Garsoïan, Interregnum: Introduction to a Study on the Formation of Armenian Identity (ca 600– (Leuven, 2012), 120. 8

42

ALISON VACCA

what it means to be Armenian. This is a very tempting reading of this geography. The province must have meant different things to different people, reflecting the multiple claims on the legacies of the past and the communal value of any given place. In the case of Širakac‘i, though, the question of Armenianness is too nebulous. Rather than seeing his extended Greater Armenia as a sort of ethnic consciousness, we might instead read this merely as the geographer floundering in the attempt to adapt an early WH[WWRFRQWHPSRUDU\JHRJUDSK\+LV*UHDWHU$UPHQLDLVQRWDOWRJHWKHUGLIIHUHQWIURP Armenian territory prior to 387. Širakac‘i likely files Arc‘ax, Utik‘ and P‘aytakaran under Armenia simply because he inherits the structure of an earlier text, written before the territorial expansion of Albania. This is not necessarily a question of “Greek” v. “Armenian” geographical traditions, let alone a burgeoning sense of Armenianness, but rather the problem of writing seventh century geography based on a fourth century model. The short recension of the $x[DUKDFoR\Fodiffers substantially from the long recension because it updates the text to expand Albania into the territories south of the River Kura. This version of the text is quite similar to the earliest concept of Albania as found in Arabic geographies, which list many of the same districts. It reads: The twenty-VL[WK>ODQGLQ$VLD@$OEDQLDZKLFKLV$øXDQNoLVHDVWRI,EHULDERUGHU ing Sarmatia along the Caucasus as far as the Caspian Sea [extending] to the frontier RI$UPHQLDE\WKH5LYHU.XUDf,WVGLVWULFWVDUH([QL%H[.oDPEĔăDQ+RøPDø kDNoĔ*HJDZX2VWDQ+DEDQG 0DU]SDQDQ 13 .oDøDFoGDxW,ED]NDQDQGRWKHUGLV WULFWV WDNHQ IURP $UPHQLD kDNDxĔQ *DUGPDQ .oXVWLSoDѻQHV .RøWo $øXĔ 7oXăNoDWDNѺtѻostak, ѺRWSDHDN*UHDWHU.XDQNo*UHDWHU,UDQNo3LDQNo+DUăODQNo Pacank‘, Muxank‘, Vakunik‘, LesseU+DEDQG6LVDNDQѺotastak, Berjor, and Asѻot, [extending] as far as the juncture of the Araxes and the Kura. 14 ɭʔɸʍ ɼʙ ʕɼʘɼʗʏʗɻɵ Ɋʉɹɸʍʂɸ ɾ, ɸʌʔ ʂʍʛʍ Ɋʉʏʙɸʍʛ, ʌɼʃʂʘ ʆɸʃʏʕ ɧʗɸʘ ɸʓ ɼʗʂ ɦɸʗʋɸʖʂʏʌ ɸʓ ɘɸʙʆɸʔɸʙ ʋʂʍʐʞ ʘɘɸʔɹʂʘ ʅʏʕʍ ʞ ʘəɸʌʏʘ ʔɸʇʋɸʍʍ ɸʓ ɘʏʙʗ ɺɼʖʏʕʟါါါɎʙ ɺɸʙɸʓʛ ɼʍ ɸʌʔʏʛʂʆါ Ɏʄʍʂ, ɋɼʄ, ɭɸʋɹɼʊɸʍ, əʏʉʋɸʉ, ɠɸʛɾ, Ɍɼɺɸʙʏʙ, ɡʔʖɸʍ, əɸɹɸʍɻ, ɔʋɸʗʅ, ɣɸʍɸʍ, ɭɸʉɸʘ ɻɸʎʖʍ, ɔɹɸɽʆɸʍါ ʞ ɸʌʃ ɺɸʙɸʓʛ, ɽʏʗ ʂ əɸʌʏʘ ʇɸʍɼɸʆ ɾါ ɠɸʆɸʎɾʍ, Ɍɸʗɻʋɸʍ, ɭʏʙʔʖʂʚɸʓʍɼʔ, ɘʏʉʀ, Ɋɽʏʙɾ [for which, read Ɋʉʏʙɾ], ɒʏʙʐʛɸʖɸʆ, ɥʖʓʏʔʖɸʆ, ɥʏʖʑɸɼɸʆ, ɝɼʅ ɘʏʙɸʍʛ, ɝɼʅ ɔʗɸʍʛ, ɣʂɸʍʛ, əɸʗʊʃɸʍʛ, ɣɸʅɸʍʛ, ɝʏʄɸʍʛ, ɧɸʆʏʙʍʂʛ, ɬʏʛʗ əɸɹɸʍɻ, ɦʂʔɸʆɸʍ, ɥʏʖɸʔʖɸʆ, ɋɼʗʈʏʗ, Ɋʔʓʏʖ ʋʂʍʐʞ ʂ ʄɸʓʍʏʙʋʍ Ɏʗɸʔʄɸʌ ʂ ɘʏʙʗ ɺɼʖ͎ 15

13

Reconstructed from Imarc Panan; this should likely be paired with Ostan as Ostan-i marzpan, i.e., the seat of the Sasanian governor in Albania; Širakac‘i, Geography, 145 n. 78. 14 Širakac‘i, Geography, 57A and 59A. 15 Širakac‘i, $x[DUKDFoR\Fo>6KRUW@, 348; $x[DUKDFoR\Fo>8QLTXH@, II 606.

B8/'Ć1$/-RĆ1

43

Map 2AB. The Caucasus in the view of Anania Širakac‘i: A = Albania, B = Armenia

This definition of Albania includes not only the lands north of the Kura, but also all of the districts of Arc‘ax and Utik‘ south of the Kura, which were the heartlands of Abbasid-era Albania.

44

ALISON VACCA

The short recension also follows the long by listing Arc‘ax as a district in Armenia, reduplicating some of these same regions: Arc‘ax borders Siwnik‘. It contains twelve districts which are in the possession of the Albanians: the other Xaband, Vakunik‘, Berjor, Greater Irank‘, Greater Kuank‘, +DUƲlank‘, Muxank‘, Piank‘, Packank‘, Sisakank‘, Kuakk‘, K‘ustip‘aѻQHVDQG.RøWo where bdellium is found. 16 [ɨɸʔɼʗʏʗɻɵ] Ɋʗʘɸʄ ʌɼʗʂ ʆɸʌ ɦʂʙʍɼɸʘʟ Ɏʙ ɺɸʙɸʓʛ ɼʍ ʍʏʗɸ ɼʗʆʏʖɸʔɸʍ, ɽʏʗ Ɋʉʏʙɸʍʛ ʏʙʍʂʍါ ʋɼʙʔ ɖɸɹɸʍɻ, ɧɸʆʏʙʍʂʛ, ɋɼʓʈʏʗ, ɝɼʅʂʗɸʍʛ, ɝɼʅʆʏʙɸʍʛ, əɸʗʒʃɸʍʛ, ɝʏʙʄɸʍʛ, ɣʂɸʍʛ, ɣɸʅʆɸʍʛ, ɦʂʔɸʆɸʍʛ, ɭʏʙɸʆʛ, ɭʏʙʔʖʂʚɸʓʍɾʔ, ɘʏʉʀ, ʌʏʗʏʙʋ ʃʂʍʂ ʛɸʗɸʄʏʙʍʆʟ 17

This description of Arc‘ax is nearly verbatim the same in the long and short recensions. The difference, though, is that in the short recension the copyist lists these districts twice so that they appear in both Albania and in Armenia, whereas the copyist of the long recension left Arc‘ax only as a region in Armenia, adding the comment that it was controlled by the Albanians. The copyist of the short recension instead not only replicated his earlier text, but also rearranged it, maintaining the original while repeating the description of Arc‘ax in the “correct” place, that is, Albania. Both P‘aytakaran and Utik‘ are missing from the long recension (supplied in brackets in Soukry’s edition) but appear in the unique Venice manuscript, the former as part of Azerbaijan and the latter as Albania: P‘aytakaran is east of Utik‘ along the Araxes. It contains twelve districts which toGD\EHORQJWR$WURSDWHQH+UDNoRW-3HUR{9DUGDQDNHUW(DZWonp‘orakean-Bagink‘ [for which, read: Ewt‘np‘orakean-Baginn], K‘oekean, ѺRYWLEDøD .oDøDQѻot, BoѻRVSLăDQ +DQL $WoOL %DJDZDQ 6SDQGDUDQ-SHUR{ 2UP]GSHUR{ DQG $OD wan…Uti [sic] is west of the Araxes between Arc‘ax and the River Kura. It has seven districts, which are in the possession of the Albanians: Aranѻovt, Tѻi, ѺotSWDN$øXĔ7XăoNDWDN*DUGPDQkDNDxĔQDQG8WL3URSHULQZKLFK>LVORFDWHG@WKH city of Partaw. 18 ɬɸʌʖɸʆɸʗɸʍ ʌɼʃʂʘ ʆɸʌ ɡʙʖʂʏʌ ɸʓ Ɏʗɸʔʄɸʙါ ɡʙʍʂ ɺɸʙɸʓʔ ɼʗʆʏʖɸʔɸʍ [ɽʏʗ ɸʌʁʋ Ɋʖʗʑɸʖɸʆɸʍ ʏʙʍʂʟ] 19 əʗɸʛʏʖʑɼʗʏʁ, ɧɸʗɻɸʍɸʆɼʗʖ, Ɏʙʀʍʚʏʗɸʆɼɸʍ ɋɸɺʂʍʍ, 20 ɭʏɼʆɼɸʍ, ɥʏʕʖʂɹɸʉɸ, ɭɸʉɸʍʓʏʖ, ɋʏʓʏʔʑʂʊɸʍ, əɸʍʂ, Ɋʀʃʂ, ɋɸɺɸʙɸʍ, 16

Širakac‘i, Geography, 65A. Širakac‘i, $x[DUKDFoR\Fo>6KRUW@, 350. 18 Širakac‘i, Geography, 65A. 19 This passage does not appear in Abrahamyan’s edition of the short recension, but it is in the unique Venice manuscript; Širakac‘i, $x[DUKDFoo\Fo>8QLTXH@, II 609. 20 ༴འརཪཱཱུུ here follows Abrahamyan’s edition of Širakac‘i, $x[DUKDFoR\Fo >6KRUW@, 350, though +HZVHQWransliterates this following Širakac‘i, $x[DUKDFoo\Fo>8QLTXH@, II 609. 17

B8/'Ć1$/-RĆ1

45

ɦʑɸʍɻɸʗɸʍʑɼʗʏʁ, ɡʗʋɽɻɼʑɼʗʏʁ, Ɋʃɸʙɸʍʟါါါɡʙʖʂ ɸʓ ʋʖʂʘ ʆɸʌ Ɏʗɸʔʄɸʌ ɿʍɻ ʋɾʒ Ɋʗʘɸʄɸʌ ʞ ɘʏʙʗ ɺɼʖʏʌါ ʏʙʍʂ ɺɸʙɸʓʔ ɽʏʗ Ɋʉʏʙɸʍʛ ʏʙʍʂʍɵ ɼʙʀʍါ Ɋʗɸʍʓʏʖ, ɨʓʂ, ɥʏʖʑɸʘɼɸʍ, Ɋʉʏʙɾ, ɨʏʙʐʛɸʖɸʆ, Ɍɸʗɻʋɸʍ, ɠʂʆɸʎɾʍ, ɡʙʖʂ ɸʓɸʍʈʍɸʆɵ ʌʏʗʏʙʋ ɣɸʗʖɸʙ ʛɸʉɸʛʟ 21

Modern scholars do not uniformly embrace this definition of P‘aytakaran, as some FROODSVH3oD\WDNDUDQDQG%DøDVDNDQLQRUGHUWRSODFHWKHSURYLQFHQRUWKRIthe river Araxes. 22 There are two main reasons to be cautious about this. First, an inconclusive passage in an eighth-century Armenian history has the Khazars cross the Araxes into these same districts.23 Second, a ninth-century Arabic geography mentions some of these places, locating them south of the Araxes as part of Azerbaijan instead of Albania: Vardanakert is :DUWKćQ (ɰȆțȲɼ), immediately on the Araxes; 24 Bagawan is %ćMDUZćQ (ɰȄɼȳȡȆȉ); 25 and Spandaran-SHUR{ LV 6ćGUćVE (ȑȷȄȲȮȆȷ), between Barzand and Zaharkash. 26 In short, a more conservative reading of the Armenian and Arabic sources would leave P‘aytakaran in Azerbaijan. This is not a wholly comfortable conclusion, though, as it cannot make sense of the association of the Armenian toponym P‘aytakaran with WKH$UDELF%D\ODTćQZKLFKLVDVVXUHGO\QRUWKRIWKH$UD[HV0DUTXDUWXVHVWKLVDV evidence that Azerbaijan had once stretched as far as %D\ODTćQ and argues that

21

Širakac‘i, $x[DUKDFoo\Fo>/RQJ@, 33; $x[DUKDFoo\Fo>8QLTXH@, II 609–10. While this passage about P‘aytakaran and Uti also appear in the short recension, it lists the districts only and omits the explanation that P‘aytakaran belonged to Azerbaijan. Širakac‘i, $x[DUKDFoR\Fo>6KRUW@, 350. 22 Several historians understand P‘aytakaran to be an earlier term that is absorbed into the territory of the Sasanian-HUDSURYLQFH%DøDVDNDQ ɋɸʉɸʔɸʆɸʍ), which also appears in Arabic as ɰȆ‫ֺܥݨ‬ȉ and in Persian as ɰȆΪȷֺȉ. 7KH WHUP %DøDVDNDQ/%DOćVDMćQ survives into the $EEDVLGSHULRGDVZHOODVZHVHHEHORZLQ÷HZRQGDQG%DOćGKXUĪ 23 ÷HZRQGURQWKH.KD]DUUDLGVq7KH\PDUDXGHGLQWKHODQGRI3oD\WDNDUDQDQG FURVVHGWKH$UD[HVULYHULQWRWKHODQGRIWKH3HUVLDQV7KH\GHVWUR\HG$UWDZĔWWKHFDSLWDO Ganjak WKH GLVWULFW WKDW ZDV FDOOHG $WoxLEDJZDQ D FHUWDLQ 6SDWDU 3oHUDZ] DQG 2UPL]G P‘erawz” (ɸʔʑɸʖɸʆɾʂʍ ʌɸʎʄɸʗʇʍ ʚɸʌʖɸʆɸʗɸʍ. ɸʍʘɸʍɾʂʍ ɿʍɻ ɺɼʖʍ ɼʗɸʔʄ ʌɼʗʆʂʗʍ ʑɸʗʔʂʘ. ɸʙɼʗɾʂʍ ɽɸʗʖɸʙɾʖ ʞ ɽɺɸʍʈɸʆʍ ʎɸʇɸʔʖɸʍ. ʞ ɽɺɸʙɸʓʍ ʏʗ ɸʀʎʂɹɸɺʏʙɸʍʍ ʆʏʐʂ. ʞ ɽʔʑɸʖɸʗ ʏʋʍÚ ʚɼʗɸʙɽ. ʞ ɽʏʗʋʂɽɻ ʚɼʗɸʙɽ). These districts render Širakac‘i’s At‘li, Bagawan, Spandaran-SHUR{ DQG 2U P]GSHUR{$OOUHIHUHQFHVWR÷HZRQGLQWKLVFKDSWHUDUHWUDQVFULEHGIURP0DWHQDGDUDQ the oldest extant manuscript of the text, ed. and tr. Sergio La Porta and Alison Vacca (in preparation). 24 :DUWKćQLVFRQVLVWHQWO\UHSUHVHQWHGDVWKHERUGHUEHWZHHQ$OEDQLDDQG$]HUEDLMDQ,Q VRPHVRXUFHVLWDSSHDUVDV$OEDQLD VHHEHORZ +RZHYHULWDOVRDSSHDUVDV$]HUEDLMDQIRO lowing this example here. Cf. Ibn KhurUDGćGKELKq7KHURDGIURP%DU]DQGWRWKHGHVHUW RI%DOćVDMćQDQG:DUWKćQZKLFKDUHWKHODVWVHWWOHPHQWVRI$]HUEDLMDQLVWZHOYHIDUVDNKV.” 25 q%ćMDUYćQr EIr LQ ZKLFK %RVZRUWK DVVRFLDWHV %ćMDUYćQ ZLWK 0XTDGGLVĪpV PDGĪQDW 0ŠqćQVRXWKRIWKH$UD[HV 26 ,EQ.KXUUDGćGKELK

46

ALISON VACCA

P‘aytakaran fell to the Albanians only at a later date and, thereafter, settled by Armenians. 27 Marquart’s suggestion does little to situate the district on any particular side of the Araxes, nor does it help us understand the categorization of P‘aytakaran. First, it could be understood as part of Azerbaijan, since it was “originally” part of the province and, specifically, fell to the south of the Araxes. Second, it might instead appear as part of Albania based on the identity of its rulers, i.e., whether being ruled by the Albanians was enough to make a district Albanian. Finally, it might, as Širakac‘i organizes it, be part of Armenia, informed either on legacies of long-past claims or on the identity of its inhabitants, i.e., P‘aytakaran might in fact be Armenia given the settlement of Armenians there. Again, the threads of various definitions of Albania weave across literary traditions and at times even present difficulties to untangle ZLWKLQDVLQJOHVRXUFH+RZHYHUZHPLJKWSDXVHWRTXHVWLRQZKHWKHUWKLVLVVLPSO\D matter of figuring out the definition of the province from one period to the next, or rather of envisioning echoes of earlier definitions coexisting simultaneously, depending on whom you ask, their familiarity with geographical texts and their position vis½-vis political claims in the North. Despite (or, more aptly, because of) the difficulties in reading Širakac‘i against both Arabic and later Armenian sources, the $x[DUKDFoR\Fo serves as a useful introduction to the definition of Albania in the early Islamic period. In part, this usefulness stems merely from its composition during a significant moment of change in the region, during or immediately after the appearance of caliphal troops in the North. More importantly, though, he demonstrates a few main themes to keep in mind about how the transmission of geographical texts informs our understanding of Albania. First, Širakac‘i is already toggling between different definitions of Albania. It is not a question of simply shifting definitions, as if Albania was one thing and became something else; instead, multiple definitions of Albania coexist in the seventh century. As much as he works to harmonize them, the fissures remain. These versions are not defined by their origin in Greek v. Armenian texts, as Širakac‘i is well aware of both traditions, in addition to whatever Middle Persian texts he had at his disposal, now lost to us. Second, the difference between these definitions of Albania are informed at least in part by the political claims made to the land. Širakac‘i – or perhaps more correctly, those who transmit Širakac‘i’s $x[Drhac‘oyc‘ – does not always agree on the territoriality of political claims. On the one hand, Utik‘ is part of Greater Armenia. On the other hand, it is ruled by Albanians and should appear as part of Albania. The political realities of the seventh century are not enough to force him to abandon the structure of a fourth-century text, in part because a subtle differentiation remains between the definition of a province and the control over a province. 6HEĔRVpVP atmut‘iwn (“History”) 6HEĔRVpVPatmut‘iwn, written in the early Umayyad period, is an invaluable resource for Sasanian and conquest-era Armenia, but it tells us next to nothing about the definition Marquart, ēUćQxDKU, 111. See also 118: “Die Landschaft ¸ÊÈÀ¸Åû, das spätere P‘aitakaran, hatte früher eine Provinz von Atropatene gebildet und war dann diesem von Artaxias entrissen worden, zur Zeit des Theophanes aber gehörte sie zu Albanien.” 27

B8/'Ć1$/-RĆ1

47

RI$OEDQLD8QOLNHkLUDNDFoL6HEĔRVUHFRJQL]HVODQGVRXWKRIWKH.XUDDV$OEDQLDDQG bears no trace of the discussion about “original” Armenian lands. During the reign of Maurice (582–602), two Armenian noblemen fled to the Caucasus via Albania, as 6HEĔRVH[SODLQVq3DVVLQJWKURXJKWKHYLOODJHFDOOHG6DZGNoWKH\UHDFKHGWKHODQGRI WKH$øXDQNoDQGPDGHIRUWKH+XQV$IWHUFURVVLQJWKHULYHU.XUDWKH\FDPSHGRQ its bank” (ʚɸʄɿʔʖɸʆɸʍ ɺʍɸʘʂʍ ɿʍɻ ʛɸʉɸʛɸɺʂʙʉʍ ʆʏʐɼʘɼɸʃ ɦɸʙɻɸʌ, ʇɸʔɸʍɾʂʍ ʌɸʎʄɸʗʇʍ Ɋʉʏʙɸʍʂʘ, ʂ əʏʍʔ ɻɾʋʔ ɸʗɸʗɼɸʃ, ʞ ɸʍʘɼɸʃ ɿʍɻ ɺɼʖʏʕʍ ʏʗ ʆʏʐʂ ɘʏʙʗɵ ɹɸʍɸʆɾʂʍ ɸʓ ɺɼʖɼɽɼʗɹʍʟ). 28 In other words, they reached Albania after they passed Sawdk‘ (directly east of Lake Sevan) and only later, RQ WKHLU ZD\ WR WKH 1RUWK &DXFDVXV GLG WKH\ FURVV WKH .XUD 6HEĔRV XQGHUVWRRG Albania to refer to the land south of the Kura. )XUWKHU6HEĔRVGHVFribes the political claims combining Armenia and Albania in the Sufyanid period. 0XoćZL\D UHSRUWHGO\EHVWRZHGWKHHQWLUHUHJLRQRQ7oĔRGRURV ѺxWXQLq+HJDYHKLPWKHUDQNRISULQFHRI$UPHQLD,EHULD$øXDQNoDQG6LZQLNoDV far as the Caucasus mountains and the Pass of Chor” (Ɏʙ ʂʎʄɸʍʏʙʀʂʙʍ ɼʖ ʍʋɸ ɽəɸʌʔ ʞ ɽɧʂʗʔ ʞ ɽɊʉʏʙɸʍʔ ʞ ɽɦʂʙʍʂʔ, ʋʂʍʐʞ ʘɘɸʑʆʏʇ ʞ ʘɣɸʇɸʆʍ ɜʏʗɸʌ).29 This reflects the administrative norms of the Umayyad period, collapsing all of the territory from the Araxes to the Caucasus into a single unit to be ruled together. We will return to this administrative grouping below, as it continues into the Abbasid period (albeit with Arab, Iranian or Turkish governors in the place of ArmeQLDQRU$OEDQLDQRQHV )RURXUSXUSRVHVLWLVLPSRUWDQWWRQRWHWKDW6HEĔRVOLVWV Armenia, Georgia, Siwnik‘ and Albania as separate provinces. Though he does not explicitly refer to the Caspian territories, he does explain that the Sufyanid North UHDFKHGĂoRøD\RQWKH&DVSLDQ 30 ÷HZRQGpVP atmabanut‘iwn (“History”) ÷HZRQGDQ$UPHQLDQFOHULF vardapet) who lived in the early Abbasid North, wrote his Patmabanut‘iwn in the late 780s. 31 LLNHWKHGHVFULSWLRQVLQ6HEĔRVpVPatmut‘iwn÷H wond’s Albania is a reflection of the contemporary political claims of the Caliphate. +RZHYHU XQOLNH 6HEĔRV ÷HZRQG DOVR UHYHDOV WKH ODVWLQJ LQIOXHQFH RI *UHHN JHR graphical paradigms. In 762, some two and a hDOIGHFDGHVEHIRUH÷HZRQGFRPSRVHG his history, the Khazars ravaged caliphal territory: 28

6HEĔRV HG$EJDU\DQ  WU7KRPVRQ  Ibid., 169 (ed. Abgaryan), 143 (tr. Thomson). 30 ĂoRø (Chol) [genitive form  Ă‘olay: short for “pass oI ĂoRøpp@ RU ĂoRU &KRU FI Širakac‘i, Geography, 57 and 122 n. 106) is how Armenian sources render the name of this site of/by Darband+HZVHQQRWHV ibid., 123 n. 106) that there probably underlies it “the native name of a village at which the Persians built Darband”; for arguments against the identification of Chol/ ĂoRU with Darband see ch. 5 below and Vacca, Non-Muslim Provinces, 96, n. 77. 31 7KHWUDGLWLRQDOGDWHIRU÷HZRQGLVDVSUHIHUUHGKHUHEXWWKLVKDVUHFHQWO\EHHQ contested. See 7LPRWK\*UHHQZRRGq$5HDVVHVVPHQWRIWKH+LVWRU\RI÷HZRQGrLe Muséon 125 (2012); Jean-Pierre Mahé, “Le problème de l’authenticité et de la valeur de la chronique GH÷HZRQGrLQL’Arménie et Byzance: histoire et culture (Paris, 1996). 29

48

ALISON VACCA They spread their raids north of the strong river called Kura and took many disWULFWV+ĔƲar, K‘aøD2VWDQLPDU]SDQ+DEDQG*HøDZXkDNoĔ%L[;HQL.DPEĔăDQ and Xozmaz. These districts are in the land of Albania. They also took the desirable SODLQRI%DøDVDNDQf7KH\DOVRWRRNVHYHQGLVWULFWVIURPWKHODQGRIWKH*HRUJLDQ SULQFLSDOLW\kXăoNo.oZHøGDSoRUĂoHOGWo-XNoĔWo9LOLVFL[Ĕ7oLDQĔWoDQG(UN ʏʗʏʘ ʔʚʓɼɸʃ ɽɸʔʑɸʖɸʆʔ ʂʙʗɼɸʍʘ ɿʔʖ ʇʂʙʔʂʙʔʏʌ ɺɼʖʏʌʍ ʇɽɸʙʗɸɺʏʙʍʂ ʏʗ ʆʏʐʂ ʆʏʙʗ. ɸʓʍʏʙʂʍ ɹɸɽʏʙʋ ɺɸʙɸʓʔ. ɽʇɾʒɸʗÚ ɽʛɸʉɸ. ɽʏʔʖɸʍʂ ʋɸʗɽʑɸʍɼɸʍ. ɽʇɸɹɸʍɻ. ɽɺɼʉɸʙʏʙ. ɽʎɸʛɾ. ɽɹʂʄ. ɽʄɼʍʂ. ɽʆɸʋɹɼʄʊɸʍ. ɽʄʏɽʋɸɽʟ ɸʌʔ ɺɸʙɸʓʛ ɸʎʄɸʗʇʂʍ ɸʉʏʙɸʍʂʘ: ɸʓʂʍ ʞ ɽʘɸʍɺɸʃʂ ɻɸʎʖʍ ɹɸʉɸʔɸʆɸʍ…ɸʓʂʍ ʞ ʌɸʎʄɸʗʇɾʍ ʂʎʄɸʍʏʙʀɼɸʍʍ ʕʗɸʘ ɺɸʙɸʓʔ. ɼɸʙʀʍ. ɽʎʏʙʐʛ. ɽʛʏʙɼʉɻɸʚʏʗ. ɽʐɼʃɻʀ. ɽʅʏʙʛɼʀ. ɽʕʂʃʂʔʅʂʄɾ. ɽʀʂɸʍɼʀ. ɽɼʗʆ: 32

Although the spelling of nearly every toponym on this list differs from the $x[DU hac‘oyc‘, ÷HZRQGpVGHILQLWLRQRI$OEDQLDLVTXLWHLQOLQHZLWKkLUDNDFoLpVLHLWLVHQWLUHO\ north of the Kura and thereby omits the main centers of Abbasid-era Albania such as 3DUWDZ7KLVFRXOGEHDUHVXOWRI÷HZRQGpVHDUOLHUVRXUFHVWKDWH[FOXGHWKHODQGVVRXWK of the Kura from Albania or it might simply be a reflection of the reach of the .KD]DUV÷HZRQGLVDIWHUDOOQRWZULWLQJWRGHVFULEH$OEDQLDEXWWRFKURQLFOHWhe .KD]DUUDLGV7KHVROHDGGLWLRQWRWKHGLVWULFWVRI$OEDQLDLV+ĔƲar, to the east of these other districts and absent in the $x[DUKDFoR\Fo ÷HZRQGpV UHIHUHQFHWR%DøDVDNDQDOVR brings us farther east, but the addition of “also” (ew) here suggests that he differentiDWHVEHWZHHQ%DøDVDNDQDQGWKHGLVWULFWVRI$OEDQLD0RVWRI÷HZRQGpV*HRUJLDQGLV tricts also appear in Širakac‘i, some under Georgia and others in Gugark‘, “originally” $UPHQLDQWHUULWRULHVWKDWWKH*HRUJLDQVFRQWUROOHG:KHQ÷HZRQGGLIIHUHQWLates between Albania and Georgia, he contrasts with Abbasid-era Arabic sources of the %DONKĪVFKRROZKLFKFROODSVHHDVWHUQ*HRUJLDLQWR$OEDQLD VHHEHORZ  ÷HZRQGOLNHkLUDNDFoLKDGWRKDYHDFFHVVWRJHRJUDSKLFDOWH[WVLQ*UHHNXQOLNH Širakac‘i, thouJK÷HZRQGpVDFFHVVDSSHDUVWRKDYHEHHQVHFRQGKDQG:HVHHWKLVIRU H[DPSOH LQ ÷HZRQGpV UHIHUHQFH WR q9Ĕѻi, who are Georgians” (ʂ ʕɾʓʂɸʌ ʏʗ ɼʍ ʕʂʗʛ). This echoes Širakac‘i, who rightly understood 9Ĕѻia as a toponym instead of ethnonym: “Veѻia, that is Virk‘ [Georgia]” (ɧɼʓʂɸ ɾ, ɸʌʔʂʍʛʍ ɧʂʗʛ), where ɧɼʓʂɸ is the Armenian transliteration of the Greek `¹¾Éţ¸. 33 7KHIDFWWKDW÷HZRQG misunderstood the term suggests that he did not personally know Greek. The simiODULWLHVEHWZHHQkLUDNDFoLDQG÷HZRQGPLJKWVXJJest that he is working with the $x[ arhac‘oyc‘ at hand. 6KRUW@, 348. I would like to thank Sergio LaPorta for pointing this out. 33

B8/'Ć1$/-RĆ1

49

Rutaks, Zidѻoy, Tasuk, Gaznak, Yormi, Surenapat, and still other neighboring districts” (ʞ ɼʃɼɸʃ ɸʔʑɸʖɸʆ ʔʚʓɾʗ ɽʆʏʉʋɸʋɹʛ ɸʖʗʑɸʖɸʆɸʍ ɸʎʄɸʗʇʂʍ ʂ ɽɸʗɸʙɸʍɻ ɺɸʙɸʓ. ʂ ʗʏʙʖɸʆʔ. ʞ ʂ ɽʂɻʓʏʌ. ʂ ʖɸʔʏʙʆ. ʂ ɺɸɽʍɸʆ. ʂ ʌʏʗʋʂ. ʂ ʔʏʙʗɼʍɸʑɸʖɵ ʞ ʌɸʌʃ ʞʔ ʋɼʗʈɸʆɸ ɺɸʙɸʓʔʍ). 34 Only Zarawand appears in the $x[DUKDFoR\Fo, as part of Parskahayk‘, which Širakac‘i understood to be part of Greater Armenia. There are a couple of different possible readings RIWKLVSRWHQWLDOO\WKDW÷HZRQGKDGDFFHVVWRDQXQGHUO\LQJWH[WRUWKDWKHLQKHULWHG a different – albeit Greek – source; or perhaps his addition of Azerbaijan is his own, merely reflecting an eighth-century Armenian understanding of Abbasid Azerbaijan. ÷HZRQG DOVR UHIHUV WR DQ $OEDQLD WKDW ZRXOG KDYH EHHQ TXLWH XQKHDUG RI LQ Širakac‘i’s time, including two further references worth noting here. First, during the UHLJQRI0DKGĪ 5–87), the governor o8WKPćQ “gathered the troops of the Armenian nobles (QD[DUDUNo) in the land of the Albanians at the Caspian gates in the city called Darband” (ɺʏʙʋɸʗɾʗ ɸʙʀʋɸʍʍ ɽɽɸʙʗʔ ʍɸʄɸʗɸʗɸʘʍ ʇɸʌʏʘ ʌɼʗʆʂʗʍ ɸʉʏʙɸʍʂʘ. ʂ ɻʗʏʙʍʔ ʆɸʔɹʂʘ. ʂ ɻɸʗʑɸʍɻʍ ʆʏʐɼʘɼɸʃ ʛɸʉɸʛ).35 ,QWKLVSDVVDJH÷HZRQGXQGHUVWDQGV'DUEDQGWREHDFLW\LQ$OEDQLDVRPHWKLQJWKDW is remarkably common in Arabic geographies but absent in Širakac‘i and other earlier WH[WV $OWKRXJK ZH VDZ DERYH WKDW 6HEĔRV XVHV WKH QHDUE\ ĂoRøD\WR GHlineate the H[WHQWRIFDOLSKDOWHUULWRU\KHGRHVQRWFODLPĂoRøD\DVDQ$OEDQLDQFLW\÷HZRQGpV explicit inclusion of Darband “in the land of the Albanians” (ʌɼʗʆʂʗʍ ɸʉʏʙɸʍʂʘ) is characteristic of the Abbasid-HUDJHRJUDSKLHVSDUWLFXODUO\WKH%DONKĪVFKRRl. Next, ÷HZRQG FODLPV WKDW +ćUŠQ DO-5DVKĪG 6–809) “gave his brother Atrpatakan and Armenia, along with Georgia and Albania” (ʖɸʌʗ ɼʉɹɸʙʗ ʂʙʗʏʙʋ ɽɸʖʗʑɸʖɸʆɸʍ` ʞ ɽʇɸʌʔ. ʇɸʍɻɼʗʈ ʕʗɸʙʛ ʞ ɸʉʏʙɸʍʂʙʛ).36 +HUH DJDLQ ÷HZRQGpVGHVFULSWLRQHFKRHVLQWKH$EEDVLG-era Arabic geographies, again particularly those associated with the %DONKĪ school, in that he combines Azerbaijan, Armenia, Georgia and Albania as a single province. These comments, therefore, reveal the polyvocality of our sources, as we saw above with the examination of Širakac‘i’s $x[DUKDFoR\Fo+HHFKRHVkLUDNDFoLLQVRPH details, but not in others, and clearly had access – though probably secondhand – to the Greek tradition. And yet he embeds these legacies of the past into an Abbasid context, all the while demonstrating that other definitions existed reflecting the political norms of his time. General Remarks on the early Armenian sources The Definitions of Albania kLUDNDFoLVHWVXSWKHGLVFXVVLRQRIWKHGHILQLWLRQRI$OEDQLDTXLWHZHOO+HLGHQWLIies Albania as a land (Dx[DUK QRUWKRIWKH5LYHU.XUDZLWKLWVFDSLWDODW.DSDøD\+HDOVR admits that both Utik‘ and Arc‘ax were “originally Armenian” but were in his own time ruled by the Albanians, leading to the reduplication of those two districts in the

34

÷HZRQGY–100r. Ibid., 123r. 36 Ibid., 127v. 35

50

ALISON VACCA

short version of the $x[DUKDFoR\FoVRWKDWWKH\IDOOLQERWK$UPHQLDDQG$OEDQLD÷H wond inherits the northward interest in the definition of Albania, failing to include Utik‘ and Arc‘ax, though he updates his description of Albania to include Darband. This detail is certainly not coming from either Širakac‘i or the early Greek sources, EXWUDWKHUUHIOHFWVDVSHFLILFDOO\$EEDVLGFRQFHSWXDOL]DWLRQRI$OEDQLD6HEĔRVpVWH[W similarly reveals a contemporary understanding of Albania, but he shows no familiarLW\ZLWK*UHHNWUDGLWLRQVRUWKHFRQFHSWRIqRULJLQDOr$OEDQLDQODQGV6HEĔRVFOHDUO\ understood Utik‘, i.e., land south of the Kura, as Albania. Given the underlying sources of each text, as well as differences in genre and the dramatic political changes during the period between Širakac‘i and ÷HZRQGLWLVQRWVXUSULVLQJWKDWWKHVHWH[WV do not project a single definition of Albania. Entangled Traditions The interrelation between Armenian and Greek concepts of Albania is undeniable here, given the Grecisms in both Širakac‘i’s $x[DUKDFoR\Fo DQG ÷HZRQGpV Patmabanut‘iwn$VVXFKVHSDUDWLQJ*UHHN$OEDQLDIURPWKH$UPHQLDQ$øXDQNoGRHVQRWUH flect the interrelated nature of the Near Eastern imagination. Texts passed back and forth readily. That said, part of this process of textual exchange is the rewriting, correcting, updating, and even mistranslating of texts as they passed from one language to another. This set of sources demonstrates that transmission was multipronged, meaning that there were multiple voices engaged in passing on geographical material. We see this in the differences between the different recensions of Širakac‘i’s text as ZHOODVWKHGLVWDQFHEHWZHHQ÷HZRQGDQGKLVLQIRUPDWLRQFXOOHGIURP*UHHNVRXUFHV There are likely a number of layers underpinning these texts, such that additions, for H[DPSOH WKH GLVWULFWV RI $]HUEDLMDQ LQ ÷HZRQGpV KLVWRU\ WKDW DUH PLVVLQJ IURP Širakac‘i, refuse to allow the construction of a clear linear stemma as information passed from one text to the next. This paints a very organic view of Near Eastern transmission, where instead of a single translator rendering a cohesive text verbatim from one language to the next, we might instead see a far more entangled path of information from one language to the next, due to flexible attributions of authorship and the constant revision of content. Province and Power Širakac‘i’s text seems to set up the argument that provinces are eternal constructs that continue to exist regardless of political claims to the contrary. Otherwise, he would just place Utik‘ and Arc‘ax in Albania, Gugark‘ in Iberia and P‘aytakaran in Azerbaijan. Yet there are alternative ways to explain this by rethinking how geographers understood the relationship between the definition of a province and claims to rule a province. In other words, was Albania “the land of the Albanians,” suggesting that it was home to Albanian people, or was it “the land ruled by the Albanians,” suggesting a potentially more diverse population? Very similar to the Arabic texts, these Armenian sources exhibit significant fluidity between the various terms applied to geographical units. Based on the earlier layer of histories about the Arsacids, Garsoïan draws a distinction between the Armenian word Dx[DUK, literally: world, i.e., land in the political sense of a realm, and erkir,

B8/'Ć1$/-RĆ1

51

literally: earth, i.e., land in the geographical sense of province. 37 kLUDNDFoLDQG6HEĔRV both refer to the Dx[DUKRI$OEDQLDZKLOH÷HZRQGFODLPVWKDW'DUEDQGLVSDUWRIWKH erkir of Albania. If we make the unsupported leap to maintain such Arsacid-era definitions for these terms, perhaps this implies that Darband was geographically combined with Albania, but not controlled by Albanians. We cannot read too much into this, though. If the differentiation between province and power were a given in this period, we would have a hard time understanding the differences between the two recensions of the $x[DUKDFoR\Fo. When the transmitter of the short recension moved Utik‘ and Arc‘ax to Albania, he is updating the Dx[DUKof Albania based on contemporary political claims. The long recension of Širakac‘i’s $x[DUKDFoR\Foby refusing to recognize Utik‘ and Arc‘ax as Albania, indicates that he did not see political rule to adequately inform geographical constructs. We therefore encounter two approaches to geography, the first conservative with an eye to rationalizing earlier constructs despite their inability to reflect contemporary norms and the other revisionist, where province aligns with power.

ARABIC GEOGRAPHICAL TEXTS: THE IRAQI SCHOOL OF THE 9–10TH CENTURIES

The earliest geographers writing in Arabic form a loose group known as the “Iraqi school” for ease of exposition rather than as a reflection of any shared genre or the inheritance of common teachers. They were participants in the flowering of the sciences under Abbasid supervision, meaning that they inherited much of their conceptual approach and vocabulary from the Greek Ptolemaic tradition.38 As such, a more correct designation may follow Kratchkovsky to label these as the LTOĪPL\\Dgeography, WKDWLVJHRJUDSK\WKDWQDUUDWHGWKHFOLPHV+RZHYHUWKHGHVLJQDWLRQRIWKHVHJHRJ raphers as Iraqi does not intend to supply precision, as if to imply that these geographers only wrote in Iraq, but rather serves as a reminder of the significance of the Abbasid state in cultivating such works. Many of these geographers were agents of the court, working in the early Abbasid GĪZćQ, who had access to the archives or bureaucratic workings of the state. As they pertain to the study of Albania, the works of this group demonstrate several common points of interest. For one, these geographers understand Albania to be a district of Armenia instead of a province in its own right. Second, the geographers rarely defined the province explicitly, but instead list the name Albania along with other toponyms such as Bardha‘a%D\ODTćQDQG4DEDOD $UP.DpDøDk). This makes it difficult to pinpoint precisely what “Albania” means, particularly whether these cities fall within Albania or rather should be seen as something else. Third, Albania is

37

Garsoïan, Interregnum, 105 makes this differentiation, separating Dx[DUK +D\RFo (“the country of Armenia”), erkir Hayoc‘ (“the land of Armenia”), and erkir Haykazean lezui (“the land of the Armenian tongue or speech”). 38 2QWKHWHUPop,UDTLVFKRROppVHH60DTEXO$KPDGoo'MXJKUćIL\ćppLY7KH&ODVVLFDO Period EI2. On the thorny relationship between Ptolemy and the early Arabic geographical WUDGLWLRQVHH-RKQ+RSNLQVq*HRJUDSKLFDODQG1DYLJDWLRQDO/LWHUDWXUHrLQ0-/