207 80 98MB
English Pages 368 [377] Year 2017
i
Power, Patronage, and Memory in Early Islam
ii
iii
Power, Patronage, and Memory in Early Islam Perspectives on Umayyad Elites Edited by
ALAIN GEORGE AND ANDREW MARSHAM
1
iv
3 Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and certain other countries. Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America. © Oxford University Press 2018 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 Oxford University Press, or as expressly permitted by law, by license, or under terms agreed with the appropriate reproduction rights organization. Inquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above. You must not circulate this work in any other form and you must impose this same condition on any acquirer. CIP data is on file at the Library of Congress ISBN 978–0–19–049893–1 9 8 7 6 5 4 3 2 1 Printed by Edwards Brothers Malloy, United States of America
v
Contents
Acknowledgments
vii
List of Contributors
ix
Introduction: Umayyad Elites and the Foundation of the Islamic Empire—alain george and Andrew Marsham
xiii
PART I: Caliphal Authority in Text and Image
1. “God’s Caliph” Revisited: Umayyad Political Thought in Its Late Antique Context—A ndrew Marsham
3
2. Paradise or Empire? On a Paradox of Umayyad Art—A lain George
39
3. A Qurʾanic Script from Umayyad Times: Around the Codex of Fustat—F rançois Déroche
69
PART II: Patronage and Power: The Evidence of the “Desert Castles”
4. Hishām’s Balancing Act: The Case of Qaṣr al-Ḥayr al-Gharbī— Robert Hillenbrand
83
5. Khanāṣira and Andarīn (Northern Syria) in the Umayyad Period and a New Arabic Tax Document—Robert G. Hoyland
133
vi
vi
Contents
6. Two Possible Caliphal Representations from Qaṣr al-Ḥayr al-Sharqī and Their Implication for the History of the Site—Denis Genequand
147
7. Umayyad Palace Iconography: On the Practical Aspects of Artistic Creation—Nadia Ali and Mattia Guidetti
175
PART III: Historiography and Historical Memory
8. Christians in Umayyad Iraq: Decentralization and Expansion (600–750)—Philip Wood
255
9. The Future of the Past: Historical Writing in Early Islamic Syria and Umayyad Memory—A ntoine Borrut
275
10. Caliphs and Conquerors: Images of the Marwanids and Their Agents in Narratives of the Conquest of Iberia—N icola Clarke
301
11. The Umayyads in Contemporary Arab TV Drama: ʿUmar ibn ʿAbd al-ʿAzīz and al-Ḥajjāj ibn Yūsuf in Musalsalāt—Jakob Skovgaard-P etersen
321
Index
335
vi
Acknowledgments
is based on papers given at a conference convened by the editors on June 24–25, 2011, at the University of Edinburgh, The Umayyads: History, Art and Culture in the First Century of Islam. We thank the Centre for the Advanced Study of the Arab World (CASAW) and Edinburgh College of Art (ECA) at the University of Edinburgh for their generous support, which made the event possible; Sophie Lowry, for her invaluable administrative skill; and our graduate student helpers, Emily Goetsch, Hannah Hagemann, Maud Hurley, and Majied Robinson. We are very grateful to Sarah Pirovitz of Oxford University Press, and Rajakumari Ganessin of Newgen for their support, assistance and patience. The Moray Endowment Fund supported the production of color plates, and Jaakko Hämeen-Anttila, Yagnaseni Datta and Simon Loynes provided valuable help and editorial assistance at various stages in the book’s production. Our speakers, colleagues, and audience came from near and far to discuss a variety of perspectives on and evidence for the Umayyad period. We are very grateful to all of them for their contributions. THE PRESENT VOLUME
vi
ix
Contributors
Nadia Ali is a postdoctoral researcher with the Empires of Faith project, a collaboration between the British Museum and Oxford University (Wolfson College). Before joining this project she completed a PhD in Islamic art history at the University of Aix en Provence, France, and has trained in Arabic and Islamic studies. Her research interests focus on the emergence of early Islamic art in the context of Late Antiquity. Antoine Borrut is Associate Professor of History at the University of Maryland, College Park. He is the author of, among other works, Entre mémoire et pouvoir: l’espace syrien sous les derniers Omeyyades et les premiers Abbassides (v. 72-193/692-809) (Leiden: Brill, 2011). Nicola Clarke is Lecturer in the History of the Islamic World at Newcastle University. She is the author of The Muslim Conquest of Iberia: Medieval Arabic Narratives (Routledge, 2012), which won the 2014 La Corónica International Book Award. François Déroche holds the chair “Histoire du Coran: Texte and Transmission” at the Collège de France. He is the author of numerous books and articles about early Qurʾanic manuscripts, including Qur’ans of the Umayyads: A First Overview (2014). He was previously Directeur d’Etudes at the Ecole Pratique des Hautes Etudes (EPHE, Paris). Denis Genequand is specialized in Islamic archaeology, with a strong focus on the transition from Late Antiquity to Early Islam. He has conducted extensive fieldwork in Syria and Jordan, and also in Yemen, Central Asia, and West Africa. He was the director of the Syrian-Swiss projects in Qaṣr al-Ḥayr al-Sharqī (2002–2011) and Palmyra (2008–2011). He has published widely on Umayyad Syria and is the author of Les établissements des élites omeyyades en Palmyrène et au Proche-Orient (2012). He holds a
x
x
Contributors
lectureship in Islamic archaeology at the University of Geneva and is also in charge of Roman archaeology at Geneva’s archaeological office. Alain George is Ieoh Ming Pei Professor of Islamic Art and Architecture at the University of Oxford. He has previously taught at the University of Edinburgh. His main areas of research are early Qur’ans, Islamic calligraphy, the Islamic arts of the book, and Umayyad to early Abbasid art and architecture. He was awarded a Philip Leverhulme Prize for his research in 2010. Mattia Guidetti is Universitätassistant in Islamic art at the Institut für Kunstgeschichte at the University of Vienna. He is the author of In the Shadow of the Church: the Building of Mosques in Early Medieval Syria (Leiden: Brill, 2016). He researches and publishes on the role of Late Antique Christian art in the making of Islamic art in the Syrian area. Robert Hillenbrand is a Professorial Fellow at the University of St Andrews and a Honorary Professorial Fellow at the University of Edinburgh. Between 1971 and 2007, he taught the history of Islamic art and architecture at the University of Edinburgh. His interests include Umayyad iconography and architecture; Islamic architecture, with a focus on the Iranian world, Greater Syria, and North Africa; and Persian miniature painting. Robert G. Hoyland is Professor of Late Antique and Early Islamic Middle Eastern History at New York University’s Institute for the Study of the Ancient World. He has written numerous books and articles exploring different aspects of the region’s history in the period 300–900. Andrew Marsham is Reader in Classical Arabic Studies at the University of Cambridge and formerly Head of Islamic and Middle Eastern Studies at the University of Edinburgh. He is the author of Rituals of Islamic Monarchy: Accession and Succession in the First Muslim Empire (Edinburgh: Edinburgh University Press, 2009) and a number of articles and chapters on Islamic political thought and practice. Jakob Skovgaard-Petersen is Professor of Islamic and Arabic Studies at the University of Copenhagen. His field of research is contemporary Islam, with a focus on the establishment of a modern Muslim public sphere, and the role of the Muslim ulama in modern Arab states. His recent publications include Global Mufti: The Phenomenon of Yusuf al-Qaradawi, coedited with Bettina Gräf (London, New York: Hurst/Columbia University
xi
Contributors
xi
Press, 2009); Islam på TV i den arabiske verden [Islam on TV in the Arab World] (Copenhagen: Vandkunsten, 2013); Arab Media Moguls, coedited with Donatella della Ratta and Naomi Sakr (London: I. B. Tauris, 2015). Philip Wood completed his DPhil at Oxford with Averil Cameron. He has taught at Oxford, Cambridge, and SOAS and now holds an Associate Professorship at Aga Khan University, Institute for the Study of Muslim Civilisations in London.
xi
xi
Introduction: Umayyad Elites and the Foundation of the Islamic Empire Alain George and Andrew Marsham
the wealthy merchant ʿUthmān ibn ʿAffān (d. 656) became the leader of the new Arabian community that his father-in-law, the Prophet Muḥammad (d. 632), had founded just over two decades earlier. For most of the following century, his relatives, starting with Muʿāwiya (r. 661–680), retained leadership of what we now know as the caliphate— the first Muslim empire. Collectively, they came to be known as the Banū Umayya—the Umayyads—after their shared ancestor, ʿUthmān’s great- grandfather, Umayya ibn ʿAbd Shams. Although the beginning of the Umayyad caliphate is usually placed in 661, with Muʿāwiya’s accession to power, ʿUthmān’s dependence upon his Umayyad relatives could equally suggest 644 as its starting point.1 From this period until their downfall in 750, the Umayyads presided over the evolution of the Muslim polity from a federation of Arabian tribes that had challenged the hegemony of the Roman (or “Byzantine”) and Iranian empires into an established Arabian and Islamic empire that dominated a vast swath of western Eurasia. It is under their rule, and partly through their actions, that a civilization of Islam began to coalesce and flourish, building up foundations that last to this day. The Umayyads lead us back to a period that is not easily documented yet is of long-lasting importance. The present collection of essays explores ways in which power and identity, both sacral and temporal, were articulated and projected by this first great dynasty of Islam. The Umayyad Empire was the largest empire in terms of land area that the world had thus far ever seen.2 The core lands of the caliphate had been captured by Arabian armies in the 630s, 640s, and 650s: Roman Syria and IN 644 AD,
xvi
xiv
Alain George and Andrew Marsham
Egypt, Sasanian Iraq, and parts of Sasanian Iran. Under the Umayyads after ʿUthmān, two further major waves of conquest, in the 670s and early 700s, saw extensive settlement in Sistan and Khurasan, at the eastern end of the Iranian plateau, and in Roman North Africa and Visigothic Spain in the south and west Mediterranean. At the same time, recurrent land campaigns penetrated the South Caucasus and the Anatolian highlands, while naval campaigns contested control of the Mediterranean Sea. Sind, at the northwestern corner of the Indian subcontinent, also fell under Muslim rule. By the apogee of the empire, in the 720s and 730s, commanders who had paid allegiance to the Umayyads were raiding Merovingian France in the West and contesting power in Central Asia with the Turgesh allies of the Chinese Tang Empire. The success of the Umayyad dynasty was founded upon the political skill of its most successful rulers, such as Muʿāwiya and ʿAbd al-Malik, and their alliances with the Arabic-speaking tribes of post-Roman Syria. The loyalty of the Syrians provided crucial support during two major civil wars (656–661, 683–692) and then helped the Umayyads suppress the rebellions against their rule that broke out in Iraq in each subsequent generation. Eventually, in the 740s, unrest in Iraq and Iran coincided with internal conflict within the Umayyad elite, leading to the events now known as the “Abbasid Revolution.” In 750, the leading members of the Umayyad clan were massacred by the revolutionary armies and Umayyad rule was thereafter confined to the far-flung Iberian Peninsula, with the Abbasid family—distant cousins—taking over the leadership of the central lands of the empire. These Abbasid successors of the Umayyads were by far the longer-lasting dynasty. The Abbasids (named after the Prophet Muḥammad’s uncle, al-ʿAbbās) ruled much of the empire for the next two hundred years, down to 945, and retained the title of caliph—and so ultimate authority over what had by then become the Sunni Islamic world—for a further three hundred years, until 1258. As a result, the Umayyad period risks being viewed as something of a temporary aberration—a period of interest for the astonishingly rapid conquests and the establishment of the Muslim empire, but of little lasting significance compared to the subsequent half millennium of the Abbasid era.3 A further problem is posed by the written record: all the extant literary evidence for early Islam was composed in its extant form only during the Abbasid period, and hence many historians have tended to be skeptical of the prospects for writing detailed Umayyad
xv
Introduction
xv
history.4 Others have accepted later representations of the Umayyads as impious and ungodly—not truly “Islamic” monarchs—a portrayal that continues to hold sway in much of the Islamic world today.5 For some time, the perspectives offered by art, architecture, and numismatics have provided a counterpoint to these widespread views, by showing that Umayyad experiments and syntheses laid the cornerstones for later developments in their respective fields.6 The perception of these evolutions has gained in complexity over the years, thanks to such factors as the more refined study of texts; the uncovering of new archaeological evidence; and the gradual identification of Umayyad Qurʾans.7 Epigraphy, notably in modern Saudi Arabia, has also yielded a growing body of inscriptions from the period: short texts set in stone, often formulaic, yet of undoubted historical importance.8 This body of material evidence, ranging from small-scale coins to monumental mosques, represents the most direct remains of the period—and it is extensive. It has brought widely held notions into question and fueled ongoing debates about issues of meaning and sociocultural context. In a parallel line of investigation, textual research is progressively showing that the Umayyad period was foundational on several other levels, and that its documentation is not as hopeless as it once seemed. Both strands, the textual and the material, are increasingly being combined as facets of the same history. Challenges to deeply ingrained assumptions are emerging, and approaches revealing continuities and discontinuities with the Late Antique and Abbasid eras are gradually opening up new perspectives. Historians working primarily with texts increasingly recognize the possibilities for more nuanced approaches to Abbasid and later sources, taking into account the processes of their transmission in order to not only understand the later contexts that shaped them but also discern materials that are likely to be much earlier than the date at which they were finally set down. At the same time, both the potential of earlier non- Arabic material and the complexities of its relationship to the later Arabic texts have been recognized.9 One important consequence of these new approaches is that many of the developments in Islam that were formerly attributed to the Abbasid period—for example, the emergence of Islamic theology and law, the rise of science in the Islamic world, and the development of key institutions of Muslim government—are now dated by many scholars to the Umayyad period.10 As a result, the Umayyad era is coming vividly to life,
xvi
xvi
Alain George and Andrew Marsham
not only as the crucial moment of imperial conquest that established the borders of what would become the Arab world, but also as a turning point in world history, in which the tribes of Arabia joined the literate world of Mediterranean and West Asian monotheism with their own highly successful iteration of this tradition. Umayyad elites are being reevaluated— and to some extent rehabilitated—as actors within this new tradition in its formative phase, not simply as the “secular” and impious aristocrats that can easily be derived from the later Arabic tradition. The chapters in the present volume reflect many of these current developments. Two principal themes run through them. The first of these is the material culture of the Umayyad era and its relationship (or not) to articulations of legitimate authority. The second is the historiography of the Umayyad period and the implications of the nature of the texts for our understanding of this period. The book is divided into three parts in which these two strands intermingle more or less closely. Part I, “Caliphal Authority in Text and Image,” addresses articulations of caliphal legitimacy and power as they are found both in texts and in iconography. Part II, “Patronage and Power: The Evidence of the ‘Desert Castles,’ ” looks specifically at the Umayyad patronage of residences, often surrounded by agricultural estates, in the Syrian steppe and its fertile peripheries. Part III, “Historiography and Historical Memory” focuses on the sources and their uses in Abbasid Iraq, Umayyad Spain, and among Christian communities under Umayyad rule. The volume ends with a reflection about the distant echoes of this era in the modern media culture of Egypt and Syria. What was the relationship between earthly and sacral authority in Umayyad times, and upon whom was ideological authority vested? These are key issues when considering any of the Late Antique polities of western Eurasia, and they lie at the heart of part I. The essays in this section suggest that the Umayyads were patrons of carefully composed articulations of sacral monarchy that took a distinctive and self-confident position within a wider discourse of images and texts, directly challenging the stance of Roman emperors and early Christian civilization with whom they were in competition. Andrew Marsham investigates the ideological underpinnings of the Umayyads’ title “God’s caliph,” suggesting, through a close reexamination of the evidence, that it should be understood in the wider context of Late Antique monotheist political culture and, specifically, as an epithet in a competitive and theologically charged dialogue with Roman imperial titulature. Alain George considers the mosaic
xvi
Introduction
xvii
decorations at the Dome of the Rock in Jerusalem and the Great Mosque of Damascus and explores how their iconography simultaneously projected claims to earthly rule and religious authority, the latter reflected in paradisiacal imagery. François Déroche shows that patronage of monumental Qur’an manuscripts was also an act of imperial monarchy—promoting a distinctive Umayyad style of Qurʾanic script and its decoration. Part II includes four chapters on Umayyad-era settlements in Greater Syria. Robert Hillenbrand argues that the palatial architecture of Qaṣr al-Ḥayr al-Gharbī represented a deliberate synthesis, under the aegis of Hishām ibn ʿAbd al-Malik (r. 724–743), of Roman and Sasanian forms and decoration in order to lay claim to inheritance of both empires. Robert Hoyland reflects on the archaeology of Khanāṣira and Andarīn in Syria in the context of late Roman and Umayyad rule; he presents an edition and translation of a new ink inscription—a request for a tax payment drafted on behalf of an Umayyad amīr, probably a son of the caliph Hishām. Denis Genequand presents evidence from recent excavations he led at Qaṣr al- Ḥayr al-Sharqī, showing that the settlement was much more extensive than is usually recognized. Some stucco images recently found there, he argues, may be atypical representations of the caliph, or at least leading members of the ruling elite. Nadia Ali and Mattia Guidetti question the dominant paradigm for interpreting the “desert castles.” Taking Quṣayr ʿAmra and Khirbat al-Mafjar as case studies, they argue that the habitual repertoires of craftsmen played a determining role in shaping their much- discussed decoration—thereby challenging the idea that buildings can be read as simple “texts” authored by their patrons. Part III presents four chapters on sources and historical memory. Philip Wood looks at textual evidence about Iraqi Christianity, mainly from Syriac sources, to suggest that the Umayyad period was a unique moment between the Sasanian and Abbasid periods when the link between the imperial state and the Christian leadership in Iraq was briefly interrupted. Antoine Borrut makes the case that elements of Umayyad historical writing long obliterated or obscured in the Arabic tradition can be recovered from Syriac sources. Nicola Clarke offers a different perspective on the historiographic tradition, considering how the western limits of the great Umayyad-era conquests were remembered by Arabic writers in Abbasid times and then in Islamic Spain—the one region of the Islamic world to remain in Umayyad hands after 750 (for another 259 years, down to 1009). Finally, Jakob Skovgaard-Petersen looks ahead a further one thousand
xvi
xviii
Alain George and Andrew Marsham
years, to the present-day representation of Umayyad rule in Ramadan TV serials from Egypt and Syria, reminding us of the potent (and often stereotyped) images that still hold wide currency in the Arab world today.
Notes 1. Hodgson 1974, 212. 2. Taagepera 1978. 3. In her recent book on the Abbasid Empire and its society and culture, Amira Bennison notes this emphasis on the Abbasid era as the period of “classical” Islam, while acknowledging deeper roots in earlier times: “I shall focus on the Abbasid era, which is generally considered the classical age of Islamic civilization and the formative century which preceded the Abbasid Revolution in 750, during which the seeds of many of the achievements of the ‘Abbasid age were actually planted” (Bennison 2009, 4). 4. See, for example, Crone 1980, 7–8: “The religious tradition of Islam is thus a monument to the destruction rather than the preservation of the past. It is in the Sīra of the Prophet that this destruction is most thorough, but it affects the entire account of the religious evolution of Islam until the second half of the Umayyad period; and inasmuch as politics were endowed with religious meaning, it affects political history no less.” Likewise, Hillenbrand (2005, 325, 344): “In spite of the vast mass of scholarly writings and the plethora of theories in recent times about the phenomenon of the rise of Islam and the early Muslim empire up to 750 ad, the structure of detailed ‘facts’ underpinning this phenomenon remains historiographically unsure… . It is difficult to give a balanced view of the Umayyad period. There are few extant contemporary sources and the dynasty’s achievements have been distorted by the partisan accounts of the chroniclers of the ‘Abbasids, their rivals who ousted them in the revolution of 750. Shiʿite groups roundly condemned the Umayyads as political usurpers rather than true theocratic rulers, alleging that they were mere ‘kings’ who had introduced the principle of hereditary succession into the umma.” 5. See Skovgaard-Petersen (this volume), and the summary by Judd (2014, 2–9). This view of the Umayyads as “secular” was criticized by Watt (1971, 572): “From a modern historical standpoint the alleged absence of a theological justification for the Umayyad claim to the caliphate, cannot be regarded as a factor in the dynasty’s downfall.” However, the notion that Umayyad rule was “secular” persists—for example in the current edition of the Encyclopaedia Britannica, which refers to “traditional Muslim disapproval of the secular nature of the Umayyad state”; Encyclopaedia Britannica Online, s.v. “Umayyad Dynasty”; www.britannica. com; (accessed 6 July 2017).
xi
Introduction
xix
6. This line of work owes much to Oleg Grabar’s influential book The Formation of Islamic Art, first published in 1973. 7. See, respectively, for instance, Elad 1995; Genequand 2012; Déroche 2014. 8. E.g., Imbert 2011. 9. Among a number of important recent publications, see, for example, Borrut and Cobb 2010; Borrut 2011; Borrut and Donner 2016. 10. On law, see Hallaq 2005. On science, see Saliba 2007, esp. 49–72. On government, see the papers collected in Donner 2012.
Bibliography Bennison, A. 2009. The Great Caliphs: The Golden Age of the Abbasid Empire. London: I. B. Tauris. Borrut, A. 2011. Entre mémoire et pouvoir: l’espace syrien sous les derniers Omeyyades et les premiers Abbassides (v. 72-193/692-809). Leiden: Brill. Borrut, A. and P. Cobb, eds. 2010. Umayyad Legacies: Medieval Memories from Syria to Spain. Leiden: Brill. Crone, P. 1980. Slaves on Horses: The Evolution of the Islamic Polity. Cambridge: Cambridge University Press. Déroche, F. 2014. Qur’ans of the Umayyads: A Preliminary Overview. Leiden: Brill. Donner, F., ed. 2012. The Articulation of Early Islamic State Structures. London: Ashgate/Variorum. Elad, A. 1995. Medieval Jerusalem and Islamic Worship: Holy Places, Ceremonies, Pilgrimage. Leiden: Brill. Genequand, D. 2012. Les établissements des élites omeyyades en Palmyrène et au ProcheOrient. Beirut: Institut Français du Proche-Orient. Grabar, O. 1973. The Formation of Islamic Art. Yale: Yale University Press. Hallaq, W. 2005. The Origins and Development of Islamic Law. Cambridge: Cambridge University Press. Hillenbrand, C. 2005. “Muhammad and the Rise of Islam.” In The New Cambridge Medieval History: Volume 1, c.500–c.700, edited by P. Fouracre and R. McKitterick, 317–345. Cambridge: Cambridge University Press. Hodgson, M. 1974. The Venture of Islam: Conscience and History in World Civilisation. Volume 1. The Classical Age of Islam. Chicago: Chicago University Press. Imbert, F. 2011. “L’Islam des pierres: l’expression de la foi dans les graffiti arabe des premiers siècles.” Revue des Mondes Musulmans et de la Mediterranée 129: 57–77. Judd, S. 2014. Religious Scholars and the Umayyads: Piety-Minded Supporters of the Marwanid Caliphate. London: Routledge. Saliba, G. 2007. Islamic Science and the Making of the European Renaissance. Cambridge, MA: MIT Press.
x
xx
Alain George and Andrew Marsham
Taagepera, R. 1978. “Size and Duration of Empires: Systematics of Size.” Social Science Research 7: 108–127. Watt, W. M. 1971. “God’s Caliph: Qurʾānic Interpretations and Umayyad Claims.” In Iran and Islam: A Volume in Memory of Vladimir Minorsky, edited by C. E. Bosworth, 565–572. Edinburgh: Edinburgh University Press.
1
Plate 1 Late Antique Syrian mosaic of Adam enthroned. Hama (Syria), Hama Museum.
2
Plate 2 Crown of Recceswinth, treasury of Guarrazar, seventh century. Rim height 10 cm, diameter 20.6 cm. Madrid, Museo Arqueológico Nacional, 653.72 (after Ripoll López 1993, cat. 12a).
3
Plate 3 Qaṣr al-Ḥayr al-Gharbī: lioness bringing down its prey, stucco sculpture, ca. 727. Damascus, National Museum.
Plate 4 Qaṣr al-Ḥayr al-Gharbī: walking man with head turned back, stucco sculpture, ca. 727. Damascus, National Museum.
4
4
Plate 5 Qaṣr al-Ḥayr al-Gharbī: two linked figures seated and reclining, stucco sculpture, ca. 727. Damascus, National Museum.
Plate 6 Qaṣr al-Ḥayr al-Gharbī: painting of an ugly man, wall painting, ca. 727. Damascus, National Museum.
5
Plate 7 Arabic text on marble from Andarīn.
6
Plate 8 Arabic text on marble from Andarīn.
Plate 9 Aerial view of the northern wing of Building E.
7
Plate 10 Panel H: a falconry scene.
Plate 11 Al- Ḥumayma, carved ivory panel with a soldier (courtesy of Rebecca Foote).
Plate 12 Al- Ḥumayma, carved ivory panel with a frontally seated personage holding a staff or a sword (courtesy of the Metropolitan Museum of Art, New York City; post conservation supported by the Jordanian Department of Antiquities in preparation for the Museum’s 2012 exhibition Byzantium and Islam: Age of Transition [7th –9th Century]).
8
Plate 13 Drawing of the calendar wall painting from the ceiling of the central span, Quṣayr ʿAmra, Jordan (C. Vibert-Guigue and G. Bisheh, Les peintures de Qusayr ‘Amra [Beirut: Institut Français d’Archéologie du Proche-Orient, 2007], pl. 48b).
1
PART I
Caliphal Authority in Text and Image
2
3
1
“God’s Caliph” Revisited Umayyad Political Thought in Its Late Antique Context Andrew Marsham*
the caliph al-Walīd II (r. 743–744) is said to have circulated a letter to the eastern provinces of the Muslim empire. In the letter, al- Walīd II set out his plans for the succession to the caliphate. He framed this testament with a now well-known statement about the status of the Muslim monarch as “God’s caliph” (khalīfat Allāh—that is, “God’s deputy” or “God’s viceregent”). The caliphs are presented as the new intermediaries between God and Man, replacing the prophets, whose era had ended with the death of Muḥammad 111 years earlier: IN 743 AD,
Then God deputed His caliphs (khulafāʾahu) over the path of His prophethood—that is, when He took back His Prophet and sealed His revelation with him—for the implementation of His decree
I would like to thank my fellow participants in the 2011 Edinburgh conference, The Umayyads: History, Art and Culture in the First Century of Islam, as well as the convenors and participants at the 2012 Late Antique and Byzantine Seminar in Oxford, the 2012 Cambridge Late Antique Network Seminar, the 2012 Middle Eastern Studies Seminar at Manchester, and the Concepts and Their Transmission in the Medieval Euromediterranean conference at the German Historical Institute in Paris for their helpful comments and criticisms of earlier versions of this chapter. I would also like to thank Jere Bacharach, James Montgomery and Luke Treadwell for helpful comments and discussions. All remaining errors are of course my own. I acknowledge the support of the AHRC during 2012 for my project The First Muslim Empire: Re-Framing the Umayyads, during which much of the work for this chapter was carried out [grant number AH/1026731/1]. *
4
4
A ndrew M arsham
(ḥukm), the establishment of his normative practice (sunna) and restrictive limits (ḥudūd) … strengthening the strands of His covenant (ḥabl) … God (blessed and exalted is He) says, “And if God had not kept back the people, some by means of others, surely the earth would have been corrupted; but God is bounteous to mankind.” (Q. 2:252) So, God’s Caliphs followed one another… . No-one can dispute their right without God casting him down, and nobody can separate from their polity (jamāʿa) without God destroying him, nor can anyone hold their government in contempt or query the decree of God (qaḍāʾ Allāh) concerning them without God placing him in their power and giving them mastery over him, thus making an example and a warning to others. This is how God has acted towards anyone who has departed from the obedience to which He has ordered [people] to cling, adhere and devote themselves, and through which it is that the heaven and earth came to be supported. God (blessed and exalted is He) has said, “When your Lord said to the angels, ‘I am placing a deputy (khalīfa) on earth,’ they said, ‘Are You placing in it someone who will act corruptly and shed blood while we are celebrating Your praise and sanctifying You?’ He said, ‘I know what you know not.’ ” (Q 2:28) Patricia Crone and Martin Hinds translated the letter and discussed it in some detail in their 1986 book, God’s Caliph.1 They accepted the unanimous opinion of previous scholars that it was an authentic copy of a real document issued by the caliphal dīwān al-rasāʾil (“office of public letters,” or “chancery”). Although there are some reasons for caution, circumstantial evidence—discussed further below—does tend to suggest that the letter may indeed be at least an approximation of an original.2 That said, whether or not this specific letter is authentic is less significant when the wider corpus of public letters that purport to be from the later Umayyad period are read together. Then, certain features recur that are unique to this corpus and do indeed appear to reflect distinctively Umayyad-era political culture.3 Among these features are the use of the title khalīfat Allāh, “God’s caliph,” in the complete absence of what might (from a later “classical” perspective) be the expected formula—khalīfat rasūl Allāh (“successor to God’s Messenger”). In Iḥsān ʿAbbās’s edition of letters by two of the last heads of the Umayyad dīwān al-rasāʾil, which includes
5
“God’s Caliph” Revisited
5
this letter, the title khalīfat Allāh occurs on about eight further occasions, whereas the later title khalīfat rasūl Allāh does not occur at all.4 The same pattern is found in the poetry, particularly from the Marwanid Umayyad period (684–750), but also some which purports to date from the very first decades of the caliphate.5 Furthermore, some of the Umayyads’ enemies also acknowledged that this was the title they used of themselves.6 In most classical Islamic political thought, it is held that this Umayyad title, khalīfat Allāh, was an aggrandizing innovation, and that the original title, khalīfat rasūl Allāh, in use under the very first caliphs, had been distorted. The “correct” title was then restored to use during the Abbasid era (750–1258), although rulers continued to use khalīfat Allāh as well in certain contexts.7 In 1986, Crone and Hinds sought to rethink this classical position. They argued that it was simpler and more elegant to suggest that the title khalīfat Allāh in fact predated khalīfat rasūl Allāh.8 Furthermore (and departing from the earlier work of Montgomery Watt on this point), they argued that the origins of khalīfat Allāh as a title for the Muslims’ leader lay at the very beginning of Islam, with the succession to Muḥammad in 632.9 It reflected a “common conception of the caliphal office” in the first centuries of Islam, which was of “the Shīʿite type,” uniting authority over both political matters and religious law; only later, with the development of classical Sunni Islam during and after the ninth century, did the division between caliph and scholars emerge, where the latter held authority over the religious law.10 They also thought it most unlikely that the title “God’s caliph” derived from any Roman or Iranian influence, but rather that it was an autochthonous Arabian (or just possibly Samaritan) formulation, which reflected the distinctive fusion of religious and political leadership in Islam.11 Of their four conclusions, none has been uncontroversial, but the first and second have best stood the tests of the last thirty years: the idea that khalīfat Allāh in fact predated the longer formula that included the Prophet is now quite widely (but by no means universally) accepted.12 It is also quite widely accepted that the more plenipotentiary understanding of caliphal authority over the law predated, or at least coexisted with, other sources of authority within the wider Muslim community.13 The idea of the title’s unique Arabian origin has fared less well. Indeed, in her 2004 Medieval Muslim Political Thought, Crone herself reassessed the evidence for the origins of the title, concluding that the title khalīfat Allāh in the sense of “God’s caliph,” or “God’s deputy,” rather than being
6
6
A ndrew M arsham
distinctive to Islam, “seems to be rooted in Hellenistic conceptions,” as they had been taken up in the Christian Eastern Roman Empire: “The [Hellenistic] king had been fashioned on the model of his Maker; he was the image of God, the archetype of the true king, as the pagans said and the Christians started agreeing when Constantine converted.”14 When the Hellenistic conception of monarchy was adopted by Muslims is not clear, but “these ideas had … contributed to the Muslim concept of the imamate, perhaps from the start, certainly from the Umayyad period onwards.”15 The early Islamic title khalīfat Allāh is now presented as related directly to late Roman conceptions of monarchy and as originating in the middle decades of the seventh century.16 However, the evidence for these very earliest usages of “God’s caliph” derives from the much later literary tradition. As a consequence, a number of scholars have recently arrived at different conclusions. In 2000, Ella Landau-Tasseron downplayed any as it were proto-or quasi-Shīʿite notions of the caliphate belonging to a bloodline or having a sacral status at the outset.17 Instead she maintained that the earliest sense of the title “caliph” (khalīfa) was simply, pace Crone and Hinds, “successor”—that is, to the Prophet in the leadership of the Muslims—in a context where leadership was elective.18 Khalīfat Allāh, she suggested, might have been an Umayyad-era innovation, intended to transcend divisions in the nascent Muslim empire by asserting the divine right of the Umayyads, elevating them above the elected leaders of earlier times.19 In 2005, Avraham Hakim took a very different approach to the problem through a study of the attribution, in the ninth-and tenth-century Hadith literature, of the titles khalīfat Allāh and khalīfat rasūl Allāh to the second caliph, ʿUmar ibn al-Khaṭṭāb (r. 634–644). Hakim was able to show that both titles were in circulation by the early eighth century at the very latest. Although this says nothing about what the Umayyad caliphs called themselves, it does tend to suggest a more complex picture than that proposed by Crone and Hinds, with “deputy” and “successor” existing in parallel in some circles from quite early on.20 Others have looked to the documentary evidence. In 2006, Robert Hoyland picked up on what he described as Landau-Tasseron’s “evolutionary” perspective on the caliphate and noted the coincidence of the first documentary attestation of the title on coins dating from 694–695 (75 ah): “Was he [ʿAbd al-Malik] the first to use the title khalīfat Allāh as the numismatic record implies?”21 In 2010, Fred Donner developed this line of reasoning, in an argument about the doctrinal position of the
7
“God’s Caliph” Revisited
7
seventh-century Muslims, whom he prefers to term “Believers,” on the basis that muʾminūn (“faithful” or “believers”) appears to have been the main label used by the early generations of Muḥammad’s community. He suggests that the caliphate of ʿAbd al-Malik (r. 685–705) was a turning point in the development of a distinctively “Islamic” identity (as opposed to a more generally monotheist one) and that the “adoption, for a brief period, of the title khalifat allah … on a few transitional coins” was part of ʿAbd al-Malik’s “determined program of emphasizing the status of the Qurʾan and its legitimating value, among the Believers.” Donner also notes that the title echoes the Qurʾan’s description of King David as khalīfat Allāh fi’l- arḍ (“God’s deputy on earth,” Q. 38:26), and that its use in 694–695 followed ʿAbd al-Malik’s completion, in 691–692, of the Dome of the Rock in David’s city, Jerusalem.22 In his 2011 Empires of Faith, Peter Sarris accepts both Donner’s model of the early polity and his conclusions about the title khalīfat Allāh, but he combines these with the observations that the title may have been used “in direct rivalry to the ideological claims made by emperors in Constantinople.”23 Hence, whereas for Crone and Hinds (1986) the title khalīfat Allāh went back to the 630s, and for Crone (2004) it was at least a consistent element of Umayyad ideology, other commentators have focused on the point that the only documentary (as opposed to literary) attestations of khalīfat Allāh before the early ninth century occur on a very rare series of coins struck between ca. 694 and ca. 697. For Hoyland, Donner, and Sarris, this contributes to the sense that the title was an exception and probably an innovation, and it is the basis of their suggestion that the reign of ʿAbd al-Malik (r. 685–705, uncontested 692–705) was the moment that this title was first used. For Landau-Tasseron, of course, khalīfat Allāh was not only innovation but also a departure from the use of khalīfa in the sense of “successor”—in line with the classical tradition and as opposed to Crone and Hinds’s critique of that same tradition. Meanwhile, Hakim has identified a discourse going back to the early eighth century where the two titles existed in parallel, in some circles at least. What all scholars apart from Crone and Hinds share is skepticism concerning the centrality of the title in the “common conception of the caliphate” and all (except Hakim) have a sense that the title khalīfat Allāh is rather somehow exceptional—even for some unique to the mid-690s— and, respectively, related directly either to the war with ʿAbd al-Malik’s rival for the caliphate, Ibn al-Zubayr (r. 683–692) or to competition with the Roman Emperor Justinian II (r. 685–695, 705–711).24
8
8
A ndrew M arsham
Certainly, khalīfat Allāh was not the caliph’s most common title. As Crone and Hinds observed: It was not of course the title commonly used for purposes of address and reference to individual Umayyad caliphs. For such purposes amīr al-muʾminīn, “commander of the faithful,” was adopted, and this title is far more densely attested in the sources than khalīfa. Indeed, the more quotidian amīr al-muʾminīn appears very often as a pairing, ʿabd Allāh … amīr al-muʾminīn, “God’s servant (or slave) … commander of the faithful.”25 Typically, this would bracket the name of the caliph, as in the earliest extant examples on coins, inscriptions, and papyrus protocols (the earliest of which date from 661–662), ʿabd Allāh Muʿāwiya amīr al-muʾminīn (“God’s Servant Muʿāwiya, Commander of the Faithful”).26 In the edited letters of Sālim and ʿAbd al-Ḥamīd, which mostly date from the 730s and 740s, khalīfat Allāh appears about ten times and amīr al-muʾminīn on more than 40 occasions.27 The latter then remained the standard caliphal title throughout the history of Islam.28 However, absence of evidence for the Umayyad period is not evidence of absence and, as we have seen, both the poetry and prose—as preserved in the later classical tradition—record “God’s caliph” quite frequently, albeit far less often than “commander of the faithful.” In what follows, an attempt is made to make sense of this evidence and to move forward from the diverse positions of the recent commentators. It is proposed that a new theoretical perspective on titulature is the best starting point for understanding the development of the title “God’s caliph.” Leaders’ titles are a form of political communication—they are articulations directed at specific audiences and to specific ends. From this follow two further points. First, the use of caliphal titles needs to be carefully contextualized: the title could have very different purposes and different receptions depending upon whether it was deployed, for example, in contests with the Roman Empire or with other members of the new Arabian monotheist elite. Second— and crucially— there are different types of title and epithet, and so it is a mistake to treat khalīfat Allāh in isolation from other caliphal titles, or as somehow equivalent in purpose to those other titles; different titles were appropriate for different contexts and purposes.29 This latter point is acknowledged by Crone and Hinds when they observe that khalīfat Allāh was not “commonly used for purposes of address
9
“God’s Caliph” Revisited
9
and reference to individual Umayyad caliphs.” It is further suggested here that there is a near exact parallel between the caliphs “protocollary” title (i.e., amīr al-muʾminīn) and his more aggrandizing titles, appropriate usually for court contexts or polemical purposes (e.g., khalīfat Allāh), on the one hand, and the Roman imperial equivalents (e.g., autokrator, or “emperor,” as opposed to eikōn theou, or “image of God”), on the other. This is not necessarily to suggest any simple process of “borrowing” or “influence.” Rather, it probably indicates a shared cultural milieu, which can indeed, after Crone (2004), be traced to Hellenistic kingship in its post-Constantinian, Near Eastern iteration. In early Islam, as elsewhere, royal titles (and caliphs are, in structural or “outsider” terms, in some sense kings) are interventions in discourse—arguments in a contest—and not statements of “fact.”30 By beginning with the evidence for this discourse in the Christian Roman Empire we can establish the basis for a comparison with the situation in early Islam. Then we can turn to the Islamic evidence, returning in particular to al-Walīd II, because the evidence for his use of the title “God’s caliph” is unusually rich. This allows us to see the extent to which the emperors and the caliphs were participating in a shared, or at least overlapping, discourse or discourses. From there, we can broaden our perspective on the Islamic evidence to take in the coins struck by ʿAbd al-Malik in order to identify some other ways in which “God’s caliph” may have been used in the conflict between the two empires, as well as within the new Muslim empire itself. Finally, in the conclusion, we can turn to consider the possible origins of the caliphal titles and the milieu in which they took shape and became normative. A single specific origin for the title “God’s caliph” cannot be identified if (as seems unlikely) it ever existed; what is clear is that a “common conception of the caliphate” was something that emerged with the formation of the Muslim empire, was never stable and was always contested. As a result, the meaning of the title itself was open to interpretation. In Islam as in Rome there was a tension between earthly authority and transcendent power—who, ultimately, had the right to “bind and loose,” was the crux of the matter. In Christian Rome, the monarch was in some senses both “priest” and “king”; as such, he could not simply inherit power but also needed to be acclaimed by God’s chosen people.31 The status of the caliph in Islam was similar—even where a bloodline was acknowledged, pledged allegiance (the reenactment of the covenant) made the caliph a sort of “sacral king.”32
01
10
A ndrew M arsham
Roman Imperial Titulature in Late Antiquity As Crone suggests, a crucial context for thinking about caliphal titles is Roman Christian political thought. Indeed, there is a near exact analogy here with the Roman emperor’s claim, in the Hellenistic tradition of sacral monarchy, to be the deputy or the image—in Greek, the hyparchos or eikōn, in Latin, the vicarius or imago—of God. In Trinitarian thought (at least as it was understood in court circles), the emperor was God’s image in the dual sense that he was both the image of Christ and also in that Adam (and so all mankind) was understood to be made in God’s image. (As Crone notes, in the strictly monotheist early Islamic context the ruler could not easily literally be God’s image, but he could be God’s deputy.)33 Just as the title khalīfa was almost never used in formal, protocollary contexts by the Umayyads, these claims to “resemblance” to, or “deputyship” of, God were rarely used in formal protocol among the Romans. Rather, the emperor had a number of formal titles, including “Augustus” and, of course, “Emperor” (Greek, autokrator; Latin imperator). Hence, the sixth-century Oxyrhynchus papyri include contracts dated by formulas such as “the reign of our godlike and pious master Flavius Justinian the eternal Augustus [and] Emperor” (basileias tou theiotatou kai eusebestatou hēmōn despotou flaouiou Ioustinianou tou aiōniou Augoustou [kai] autokratoros).34 By the early seventh century “king” (basileus), which had long been used informally and in diplomatic language, had also become a formal imperial title; Heraclius (r. 610–641) and his son referred to themselves in an edict of March 21, 629, as “Heraclius and Heraclius the New Constantine, Kings Faithful in Christ” (Herakleios kai Herakleios neos Kōnstantinos pistoi en Christō basileis).35 For claims that the emperor was “God’s deputy” or “God’s image” we must turn to panegyric, to the more aggrandizing imperial decrees, and to the iconography of imperial ceremonial. The starting point is the Tricennial Oration of Eusebius, Bishop of Caesarea (d. 339), performed in 335–336 at the thirtieth jubilee of the Emperor Constantine (r. 306–337). In this speech, Eusebius used the ancient Hellenistic idea of the sacral king to present the newly Christian emperor as God’s deputy on earth. In the passage where hyparchos appears, Eusebius is describing Constantine’s conversion of pagan temples into churches: For while the Common Saviour of the Universe punished invisible beings invisibly, he [Constantine], as the deputy of the
11
“God’s Caliph” Revisited
11
Supreme Sovereign (hoia megalou basileōs hyparchos) proceeded against those so vanquished, stripping their long and utterly dead corpses and distributing the spoils freely among the soldiers of the victor.36 In the three centuries after Constantine and before the caliphate, monotheist sacral monarchy was elaborated, becoming part of the ideology of Christian empire. In court circles and in panegyrics the emperor was not only God’s delegate on earth, ruling with God’s authority, but also God’s likeness or image. Hence, Emperor Justinian (r. 527–565) declares in the preface to his Digest that “governing our empire by the authority of God, which has been entrusted to us (deo auctore nostrum gubernantes imperium quod nobis a caelesti maiestate traditum est), we win victory in war, cultivate the arts of peace, and preserve the State.”37 Of Justinian’s successor, Justin II (r. 565–574), the panegyricist Corippus says, “Christ granted the lords of the earth to have power over all things. He is omnipotent; and he [the emperor] is the likeness (or image) of the omnipotent.”38 That the same ideas circulated far beyond Constantinopolitan court circles is suggested by the anonymous Armenian history attributed to Sebeos. This text, which probably dates to the latter part of the seventh century, refers to an Armenian nobleman who refused to rebel against Emperor Heraclius, justifying his loyalty on the basis of the emperor’s status as God’s deputy: “but he did not agree to the murder of the king [Heraclius] and his sons. Rather he said, ‘You call them vicars (i.e., ‘deputies,’ tełepah) of God; so it is not right to participate in that act.’ ”39 The Bible was the source of archetypes that could be deployed to emphasize the sacred status of the emperor as the “re-born” and “de- judaised” Israelite monarch of God’s chosen people.40 Hence, a homily on the 626 siege of Constantinople: Our emperor [Heraclius] is a new David in his piety and his clemency to his subjects. And the Lord will crown him with victories like David, and his son who reigns with him, making him wise and peaceable like Solomon, and bestowing on him and his father piety and orthodoxy.41 Similarly, Corippus alluded to the parallel between Adam’s dominion and the emperor in his panegyric for Justin II.42 These biblical archetypes were also invoked visually in the spaces where imperial power was enacted
21
12
A ndrew M arsham
publicly. For example, at Justinian’s foundation of St. Catherine’s in Sinai, the mosaic image of the transfigured Christ is surrounded by roundels depicting the patriarchs, with the crowned figure of David—perhaps also to be identified with Justinian himself—directly below Christ’s feet.43 The more aggrandizing imperial claims did not go uncontested. Agapetus (fl. ca. 527)44 acknowledges that the emperor is the image of God but advises that the emperor is also, after all, a mortal human—in an echo of Paul’s letters, God’s servant or slave—like everybody else: In his bodily essence, the emperor is the equal of every man, but in the power of his rank he is like God over all men… . For if he is honoured for his divine image, he is nevertheless bound to his earthly image through which he is taught equality with other men… . Treat your servants as you pray that your Master will treat you… . The emperor is the master of everyone, but together with everyone he is the slave of God. He will then most properly be called master, when he masters himself.45 Others were more vociferous. The career of Maximus the Confessor (d. 662) began at the court of Heraclius and ended with mutilation and exile under Heraclius’s son, Constans II (r. 641–668). Maximus had been critical of the monothelete theology promulgated by both emperors and of the aggrandizing claims of the emperors to have authority over religious dogma, usurping (in his view) a role proper to priests. The eyewitness (though highly tendentious) account of Maximus’s interrogation in 655 describes his response to the question “Is not the Christian emperor also a priest? [i.e., And so may he not involve himself in dogma?]” No, because he does not have access to the altar; and after the sanctification of the bread he does not raise it, saying: the holy things to the holy; he does not baptise, nor does he perform initiation with the myron [i.e., the chrism or oil of anointing]; he does not ordain, nor does he make bishops, priests and deacons; he does not consecrate churches by unction; he does not bear the symbols of priesthood, the omophorion [vestment] and the Gospel, as he bears those of kingship, the crown and the purple… . But what purpose is served by continuing? During the holy anaphora at the holy table, it is after the hierarchs and the deacons and the whole order of clergy that one remembers the emperors at the same time as the
13
“God’s Caliph” Revisited
13
laity, the deacon saying: and of the laymen who have died in the faith, Constantine, Constans etc. He also mentions living emperors after the clergy.46 For Maximus a clear distinction ought to be maintained between royal and sacerdotal power—a distinction that had been eroded by the claims of the emperors to decide matters of dogma, as kings in Christ and images of Him.
The Ruler as God’s Deputy in Islamic Late Antiquity: The Claims of al-Walı̄d II In sixth-and seventh-century Rome, the emperor could claim to be both king and priest—to be the image of Christ and God’s deputy—but these claims were confined to panegyric and court contexts, provoking both politic reactions, like those of Agapetus, and impolitic ones, like that of Maximus. Protocollary titles remained more reserved: “commander” (imperator or autokrator) and, at the height of imperial claims to religious authority, “king” (basileus); the emperor could also be reminded that he was, in the end, doulos theou—God’s servant. The parallels with the situation of the Umayyad rulers are clear: in protocollary contexts, an Umayyad ruler was “God’s servant … commander of the faithful.” In certain ceremonial and court environments, however, the caliph was associated much more closely with God. But the Muslim monarch could not claim to be the image of God. Instead he was “God’s deputy” (as the Roman emperor sometimes was, too). By the later Umayyad period, at the latest, the title “God’s caliph” could be understood as alluding to what might be seen as the Qurʾanic idea of all Mankind’s stewardship of the world (e.g., Q. 6:165, 57:7), and more specifically to the patriarchs Adam and David, as God’s delegates on earth (Q. 2:28 and 38:26).47 Garth Fowden has argued convincingly that we see this idea rendered iconographically under the patronage of the future al-Walīd II.48 In around 735, while he was still just heir-apparent to the caliphate (and about a decade before he sent the letter to the provinces about the succession, cited above), al-Walīd built the small palace complex in the desert east of Amman, in modern Jordan, now famous as Quṣayr ʿAmra.49 A small stone-built audience chamber, on a plan familiar from pre-Islamic Romano-Arabian sites, adjoins a bath complex. Nearby are a small mosque, irrigation works, and other buildings. The interior of the audience chamber and baths is
41
14
A ndrew M arsham
remarkable—and well known—for the well-preserved frescoes that cover it. Crucially, for the purposes of this discussion, there is an image of a ruler placed directly above the space for the enthroned heir, which draws heavily on Roman, and, specifically, Romano-Syrian, forms (fig. 1.1).50 To an observer familiar with Roman imperial iconography (and, indeed, Sasanian iconography), the visual referents would be numerous and resonant with all the symbolism of monotheist imperial monarchy: this was God’s deputy enthroned, as visible in imperial contexts from Constantinople to Taq-i Bustan. However, Fowden suggests that a more specific, and local, reading was also possible. He proposes that the image of the ruler at Quṣayr ʿAmra was a deliberate allusion to images prevalent in Late Antique churches in Greater Syria of Adam as God’s delegate, holding dominion over the earth. These include the mosaic of Adam enthroned, now in the Hama Museum, in modern Syria. Elements of this image, including the rounded cushions on each side of the throne, the pillars with capitols supporting an arch, the draped clothing of the seated figure, and the feet placed on a footstool, all bear a particularly close resemblance to the fresco at Quṣayr ʿAmra (plate 1).51 Fowden also observes that other Syrian images likewise depict the animals over which Adam had been given dominion. In the mosaic floor at Ḥawīrtah, the position of the birds flanking Adam’s shoulders resembles that of those flanking the ruler at Quṣayr ʿAmra. In the latter fresco, the ruler is seated above water, where fishes swim, in another more distant echo of the aquatic life depicted some distance from Adam at Ḥawīrtah (figs. 1.2 and 1.3).
Figure 1.1 Drawing of the extant image of the enthroned ruler at Quṣayr ʿAmra and artist’s reconstruction (after Vibert-Guigue and Bisheh 2007).
15
“God’s Caliph” Revisited
15
Figure 1.2 Floor mosaic depicting Adam and the animals, late fifth century ad. Ḥawīrtah (Syria), North Church (drawing after Laroche in Canivet and Canivet 1987).
For Fowden, the visual evidence for an allusion to local ideas about Adam’s dominion over the earth (and so also its Qurʾanic retelling) was compelling, but what clinched the case was its patron’s later letter, where the relevant Qurʾanic verse was invoked to provide an archetype of his own caliphal dominion: “When your Lord said to the angels, ‘I am placing a deputy (khalīfa) on earth,’ they said, ‘Are You placing in it someone who will act corruptly and shed blood while we are celebrating Your praise and sanctifying You?’ He said, ‘I know what you know not’ ” (Q. 2:28). (The full verse continues: “He taught Adam the names of all things; then He placed them before the angels, and said: ‘Tell me the names of these if ye are right’ ”). This Qurʾanic text, cited by Quṣayr ʿAmra’s patron, does seem to tie the iconography of that palace to the idea of Adam’s dominion over Creation, and so also to the Late Antique Christian version of the story as depicted in a number of Syrian churches. This is not to suggest that the palace fresco had a single meaning, but it does build a strong case that many Syrians would have read it as an allusion both to imperial power
61
16
A ndrew M arsham
Figure 1.3 Detail of Adam enthroned, late fifth century ad. Ḥawīrtah (Syria), North Church (after Canivet and Canivet 1987).
and to Adam, and that these ideas were in the minds of its patron and painter.52 Fowden’s case ends here and is quite convincing on its own terms. However, it is also possible to provide a wider Late Antique context for al-Walīd’s iconography and for its Qurʾanic referent, when it is noted that a number of Late Antique Jewish and Christian texts from the regions bordering the Arabian Peninsula are especially close to the Qurʾanic text. The relevant Biblical story about Adam is found in Genesis: Now out of the ground the Lord God had formed every beast of the field and every bird of the heavens and brought them to the man to see what he would call them. And whatever the man called every living creature, that was its name. The man gave names to all livestock and to the birds of the heavens and to every beast of the field. But for Adam there was not found a helper fit for him (Genesis 1:19–20).
17
“God’s Caliph” Revisited
17
Then God said, “Let Us make man in Our image, according to Our likeness; let them have dominion over the fish of the sea, over the birds of the air, and over the cattle, over all the earth and over every creeping thing that creeps on the earth.” So God created man in His own image; in the image of God He created him; male and female He created them. (Genesis 1:26–27; my emphasis) In the Late Antique Middle East, this story had been elaborated to emphasize the idea of Adam as the archetypal first ruler of the world—and the first in a series of successors to God’s covenant; most likely these elaborations derived ultimately from Hellenistic Jewish sources, such as Philo of Alexandria (d. ca. 50 AD).53 Hence, the Syriac Book of the Cave of Treasures—probably composed in the sixth century—describes itself as a “Book of the Succession of Generations,” setting out the transmission of God’s covenant through the patriarchs from Adam down to the time of Christ. A passage on the naming of the animals expands on Genesis 1.19–20: Adam was created in Jerusalem. There he was arrayed in the apparel of sovereignty… . And all the wild beasts, and all the cattle, and the feathered fowl were gathered together, and they passed before Adam and he assigned names to them; and they bowed their heads before him; and everything in nature worshipped him, and submitted themselves to him. And the angels and the hosts of heaven heard the Voice of God saying unto him, “Adam, behold; I have made thee king, and priest, and prophet, and lord, and head, and governor of everything which hath been made and created… .” And when the angels heard this speech they all bowed the knee and worshipped Him.54 Likewise, the Hebrew Midrash Rabbah—perhaps compiled around 500 ad—comments on Genesis by developing the theme of God’s being questioned about humanity’s destructive nature, as also in the Qurʾan: Then God said, “Let us make Man in our likeness, and let there be a creature not only the product of the earth, but also gifted with heavenly, spiritual elements, which will bestow on him reason, intellect, and understanding.” Truth then appeared, falling before God’s Throne, and in all humility exclaimed: “Deign, O God, to refrain
81
18
A ndrew M arsham
from calling into being a creature who is beset with the vice of lying, who will tread the earth under his feet.” Peace came forth to support this petition, “Wherefore, O Lord, shall this creature appear on earth, a creature so full of strife and contention, to disturb the peace and harmony of Thy Creation? He will carry the flame of quarrel and ill-will in his trail; he will bring about war and destruction in his eagerness for gain and conquest.”55 The same Late Antique milieu that produced these commentaries on Genesis, and on the idea of Adam’s dominion over the world, also produced debate on the question about Adam’s “likeness” to God. Here, a prominent strand of the Syriac tradition comes closest to the Qurʾanic stance; Ephrem (d. ca. 373 ad) argues in his commentary on Genesis that Adam’s “likeness” to God pertains only to his dominion and rule over Creation and is not to be interpreted literally (i.e., not that Adam actually looked like God): “And God said, ‘Let Us make man in Our image.’ ” According to what has been the rule until now, if it pleases God, He will make it known to us, Moses explains in which sense we are the image of God, saying: “That he may rule over the birds and beasts and fish of the sea and over all the creatures that move along the ground.” Therefore, Adam accepted power over the land, and over everything in it; he was in that the likeness of God, in whose power were the heavens and the earth.56 On the basis of these texts, it seems almost certain that the Qurʾanic use of khalīfa to refer not just to Adam and David, but also to other, unnamed “successive generations” (khulafāʾ) of human beings, is to be understood as relating to this wider discourse, in which Adam’s “dominion” and his role as the first in a series of holders of “God’s covenant,” and, in some (highly contested) sense also “God’s likeness,” is discussed.57 For the later exegetes of the Qurʾan—who tended to want to draw a clear line between prophetic and caliphal authority—there was no connection between the caliphate of Adam and David, and the caliphate of the Muslim rulers of their own day.58 Nor—perhaps less surprisingly given their aversion to Christian apocrypha—did they expound on the
91
“God’s Caliph” Revisited
19
question of God’s likeness to Man in connection with this passage.59 However, al-Walīd II made much of them and sought to participate in a long-established and live Late Antique discourse about monotheist monarchy and its patriarchal archetypes; these particular implications of the images—and also of texts—he sponsored would have resonated with his post-Roman Syrian audience.60 As noted above, this association between royal power and Adam’s dominion was also familiar in Constantinople.61
Caliphal Titulature as Discourse That is to say, the meaning of this claim to be God’s deputy would, of course, vary with its context. In al-Walīd II’s iconography and in his letter—as in other purportedly Umayyad-era texts—the caliphs are placed at the end of the succession of prophets, as guardians of God’s covenant with Mankind. In this, the Umayyad material echoes both the Syriac Book of the Cave of Treasures and some of the claims of the Roman emperors. However, in the Umayyad version, the covenant is passed on from Adam to Christ, then to Muḥammad, and thence to the Umayyad monarchs themselves. Thus, the Umayyad caliphs’ claims to be khulafāʾ Allāh were at once an analogue of the Roman emperors’ claims to be God’s hyparchoi (and, potentially, a critique of the emperors’ claims to be God’s eikōnes) and also an assertion of their own place as successors to God’s covenant in both generally monotheist and more specifically Qurʾanic terms. The audiences for these articulations were diverse, and reception of the title would vary accordingly: allies among the wider Umayyad clan would no doubt receive it approvingly, as might the loyal members of the (post-Roman and Christianized) Arabic-speaking tribes of Syria; other groups—notably the Arabian settlers in Iraq and the population of the Hijaz— might have been less receptive.62 All the audiences, however, would have perceived the Umayyads’ claims as a challenge to contrasting articulations of legitimacy, whether Roman or Arabian. In the case of the two best-attested assertions of “God’s caliphate,” these contexts and audiences can be partially reconstructed. The second of these is the material attributed to al-Walīd II, to which we will return with this question of audience in mind shortly. Before that, however, we should consider the first well-attested assertion of “God’s caliphate,” by al-Walīd II’s grandfather, ʿAbd al-Malik.
02
20
A ndrew M arsham
ʿAbd al-Malik’s reform of the coinage system of his empire, which departed from both the metrical and iconographic forms of the Roman gold (and Iranian silver) system that the Arabians had adopted after the conquests, would certainly have amounted to a very direct challenge to Roman imperial authority. Dinars (gold coins), most likely struck in Syria in the 690s, mark a clear departure in weight and style from previous copies of Roman solidi, replacing the image of the emperor with an image of the caliph, and, on the obverse, a “pole-on-steps,” which amounted to a denial of Roman Trinitarian monotheism (fig. 1.4).63
Figure 1.4 “Standing caliph” dinar, 76 ah/696 ad. London, British Museum, CM 1954, 1011.2.
Figure 1.5 “Standing caliph” dirham 75 ah/694-5 ad. Private collection (after Walker 1952, pl. IX.4).
21
“God’s Caliph” Revisited
21
While these so-called standing caliph dinars have no dates marked on them, dirhams (silver coins), derived from a Sasanian prototype, are marked with a year in the lunar Hijri calendar. An example marked 75 ah (694–695 ad) is the first of a very few dirhams that appear to have been closely modeled on the “standing caliph” dinars, but with the addition of the labels amīr al-muʾminīn and khalifat Allāh flanking the standing figure (sic: the latter title appears without the long ‘ī’ familiar from classical Arabic, and with a tāʾ instead of a tāʾ marbūṭa, as it also does in the Syrian copper coinage from this period; fig. 1.5).64 The same caliphal titles also appear on some of the so-called miḥrāb and ʿanaza dirhams from the same period, together with the phrase naṣr Allāh or naṣara Allāh (“God’s victory” or “May God assist,” perhaps likely the former; fig. 1.6).65 Two years later, in 77/697, a further reform saw the production of entirely aniconic coins, which placed the assertions of God’s unity and the denial of the Trinity at the center of the two fields (fig. 1.7). This was a design that broke with the long tradition of royal images on coins in the Middle East. It stuck, lasting in various iterations for over a millennium. All reference to the ruler had disappeared (though it was reinstated by various dynasties after the Umayyad period).66 These iconographic developments have tended to be characterized as the product of a single, specific cause. Often, this is a direct diplomatic
Figure 1.6 “Miḥrāb and ʿanaza” dirham, ca. 75–77 ah?/ca. 694-7 ad?. Private collection (after Morton & Eden Auction 85, 26 April 2017, lot 3).
2
22
A ndrew M arsham
Figure 1.7 “Post-reform” epigraphic dinar, ca. 77 ah/696-7 ce. London, British Museum, CM 1874, 0706.1.
exchange between the two empires, with ʿAbd al-Malik’s coins and those of Justinian II (r. 685–695 and 705–711) in dialogue—an interpretation with some basis in the literary sources.67 In this assessment, Justinian II’s new coinage, which juxtaposed the bust of Christ and an image of the monarch, and which in some cases were labeled “Christ’s servant” (or “slave,” servus Christi), might be interpreted as a response to the Arabians’ new coinage, or vice versa.68 However, some recent examinations of the development of the seventh-and early eighth-century Roman coins have emphasized internal developments as opposed to relations with the Arabians.69 That gold was used in payments to the Romans and was the main precious metal currency in former Roman territories might suggest that the anti-Trinitarian messages were directed primarily at Constantinople itself, as well as the armies in Syria, while the specific inscriptional assertion about “God’s caliph,” which occurs only on the dirhams, may have been directed primarily against internal rivals. After all, it is a little difficult to see the “standing caliph” dirhams as part of this direct dialogue between monarchs—one would rather expect the dinars that are said to have been the medium of diplomatic exchange. The obvious alternative, “internal” context for ʿAbd al-Malik’s coins is the civil war with Ibn al-Zubayr (r. 683– 692): ʿAbd al-Malik’s reform came in the immediate wake of a war over the leadership of the Arabian empire, in which his opponent had made much of his status as a Companions of Muḥammad and his close association with the pilgrimage site of Mecca, and had struck copies of Sasanian coins
23
“God’s Caliph” Revisited
23
marked bismillāh (“In the name of God”).70 Although there is some metrological evidence that, unusually, ʿAbd al-Malik’s silver coins were actually struck in Syria, the coins may have been intended for export to Iraq or Iran.71 What is more certain is that reading these coins as the product of a one-dimensional political context is wrongheaded.72 In the 690s, ʿAbd al-Malik oversaw the promulgation to the Arabian conquerors on a new scale of claims to monotheist, imperial authority.73 Competition with Constantinople in Syria, Asia Minor, the Transcaucasus, North Africa, and the Mediterranean was one powerful consideration. So, too, were his ongoing problems in asserting his claims to legitimate authority in the Hijaz, in northern Syria, and in Iraq in the wake of a decade of war in which he had overthrown the widely recognized caliphate of Ibn al- Zubayr. In the conflict with Rome, his claim to be khalīfat Allāh would be a rebuke—after Ephrem, Maximus, and the Qurʾan—to the notion of the emperor as literally Christ’s image, and so an assertion of the superior piety of the Arabians. In the conflict with the Zubayrids it provided a link, via its Qurʾanic resonances, to Adam and David, two of the patriarchs of Jerusalem, where he was building the Dome of the Rock; in a rivalry with (primarily Abrahamic) Mecca, this was powerful rhetorical material. That is, the khalīfat Allāh coin certainly formed part of a wider program of public enunciations of legitimate power, directed at diverse audiences within the caliphate and without. A panegyric poem by al-Akhṭal, most likely performed at ʿAbd al- Malik’s Syrian court in 691 during the closing stages of the conflict with Ibn al-Zubayr, points to both this “internal” context and to the wider, long- standing tropes of Near Eastern monarchy and Arabian poetry (at the same time, however, the Romans are present in the background of the riverine metaphor): He who wades into the deep of battle, auspicious his augury, the Caliph of God [khalīfat Allāh], through whom men pray for rain. When his soul whispers its intention to him it sends him resolutely forth, his courage and caution like two keen blades. In him the common weal resides [wa’l-mustamirr bihi amr al- jamīʿ], and after his assurance no peril can seduce him from his pledge. Not even the Euphrates when its tributaries pour seething into it and sweep the giant wallow-wort from its banks into the middle of its rushing stream,
42
24
A ndrew M arsham
And the summer winds churn it until its waves form agitated puddles on the prows of ships, Racing in vast and mighty torrent from the mountains of the Romans whose foothills divert its course, Is ever more generous than he to the supplicant or more dazzling to the beholder’s eye.74 The image of the ruler as sacred bringer of the rain is an Ancient Near Eastern motif, where the water stands for the material and spiritual blessings that are gained through loyalty to the king; the image of the ship on the Euphrates in flood also directly recalls a poem attributed to al-Nābigha al-Dhubyānī, a pre-Islamic poet addressing the King al-Nuʿmān of al-Ḥīra in Iraq (r. ca. 580–602).75 Hence, it may not be a coincidence that a ship sails on waters below the throne of the prince at Quṣayr ʿAmra—constructed about 45 years after al-Akhṭal performed his panegyric for ʿAbd al-Malik. However, if ʿAbd al-Malik’s use of “God’s caliph” can be connected to his projection of claims to monotheist imperial authority as he fought to establish his power in contest with both the Romans and Ibn al-Zubayr, al-Walīd II’s use of similar iconography and ideas nearly fifty years later should be linked to his struggle, first, to retain his place in the succession during the reign of his uncle Hishām (r. 724–743) and, second, to retain the caliphal succession for his own sons at the expense of other branches of ʿAbd al-Malik’s family in 743.76 The audience for the palace iconography at Quṣayr ʿAmra was probably quite limited, and in this respect, the allusions to biblical and Qurʾanic archetypes conform to the pattern that we see in the Roman context, where such assertions were for an elite, usually courtly, audience. Given that the palace is now known to have been constructed while he was defending his status as walī al-ʿahd (“heir apparent”—literally, “successor to, or possessor of, the covenant”), the primary purpose may have been to assert his royal credentials among his supporters.77 In contrast, the letter was intended for a much wider audience: it was to be read out publicly at the congregational mosques of Iraq and Khurasan, and perhaps elsewhere, too.78 The letter reads not only as an assertion of al-Walīd II’s claim to be “God’s caliph,” but also as a defense of caliphal prerogratives in the interpretation of the law: by the 740s, the idea that the wider community of Muslims had some claim to interpret the law had gained ground, and there is some evidence that al-Walīd II
25
“God’s Caliph” Revisited
25
sought to reassert a more plenipotentiary role for the caliph.79 As the letter has it: “God deputed His caliphs … for the implementation of His decree (ḥukm), the establishment of his normative practice (sunna) and restrictive limits (ḥudūd), and for the observance of His ordinances (farāʾiḍ) and His rights (ḥuqūq).” One is reminded of the lament by the former Umayyad secretary, Ibn al-Muqaffaʿ, composed only about ten years later, that, with regards to a legal matter, “traced from the forefathers (al-salaf) upon which there is not agreement—one group manages it one way, others manage it another way.” Ibn al-Muqaffaʿ’s famous proposed solution was that the caliph should look into “which of the two is most worthy of trust (taṣdīq) and which of the two most resembles justice (ʿadl)” in order to produce a single law code; this suggestion does not seem ever to have been taken up.80 That is, al-Walīd II’s letter was also an intervention in the ongoing and evolving contest over the relationship between the monarch and the law in early Muslim society—one that was already framed in terms of “book” and sunna, ḥudūd, farāʾiḍ, ḥuqūq, and ʿadl. The connection with the discussion of the succession that followed was clear: al-Walīd II, as monarch, was making a claim to have the capacity of “loosing and binding” (in later classical thought, al-ḥall wa’l-ʿaqd), against the rival claims of those who also might assert that they represented God’s covenant, and God’s law—in the Roman Empire, this group were priests, but in the Muslim empire, they could be understood as God’s community of the Faithful.81
Conclusions The years 694 to 697 and 743 were unusual moments in the public articulation of Umayyad legitimacy. Media that would reach large numbers of ordinary Muslims—coins and a public decree, respectively—deployed the epithet “God’s deputy,” where “commander of the faithful” was much more usual. However, in other contexts “God’s deputy” was quite widely used—it is found in poetry and prose and is referred to by the Umayyads’ rivals. The analogy with the Roman imperial titles is striking: just as the emperor was imperator, autokrator, or basileus in protocol and only eikōn theou or hyparchos theou in panegyric, so, too, the caliph. Both, of course, were “God’s slaves.” That said, there is a strong impression that the Islamic context was more hostile to public expressions of royal sacral power. The clearest evidence for this comes from the coinage, where the first modifications to the Roman and Sasanian coins entailed adding the
62
26
A ndrew M arsham
tag “In God’s Name” (and similar pious phrases) and where a flurried period of experimentation with “caliphal image” coinages in the 690s led quickly to a dramatic departure from centuries of Middle Eastern custom and the complete disappearance of any mention of the monarch from the precious coins. It is probably wrongheaded to explain this simply in terms of “pious opposition” to the early experiments—after all, as Luke Treadwell has recently shown, the dramatic break in iconography helped the practical aim of facilitating the unification of the Roman and Sasanian fiscal systems.82 Nonetheless, there certainly was also an ideological dimension to this change. Ultimately, it appears to stem from the idea of the amr Allāh—“God’s command”—which the caliphs claimed to represent on behalf of God’s community of the faithful.83 This idea of the sovereignty of God appears to be echoed in the Qurʾanic use of the term khalīfa. Whereas for many Christians, particularly in the Greek-speaking world, Adam was both the image and the deputy of God, in the Qurʾan, as in much of the Late Antique Judaic and Syriac tradition, he is merely God’s delegate and in no literal sense his “image.” In such a context, “God’s deputy” becomes a title at once humble and aggrandizing: true sovereignty is deferred to God, while the monarch rules on His behalf on earth. It is also worth observing that there is no necessary connection between the title “God’s caliph” and a particular understanding of the sources of God’s law: pace Crone and Hinds, God’s deputy might still be beholden to God’s law, as the Roman discourse reminds us. However, “God’s caliph” does appear to have gained an ideological charge during the early to mid-eighth century (not lost, it seems, on al-Walīd II) that has left its mark in the Hadith literature.84 This meant that when the Abbasids came to power, they tended to pair the title with one that accentuated the founding position of God’s Messenger—“successor to God’s messenger” (khalīfat rasūl Allāh).85 Again, the analogies with the Roman situation, and the question of “church” and “state,” are striking. However, the old problem played out rather differently in the Islamic context: as Crone and Hinds observed, “the caliphate did fuse religion and politics from the start,”86 or, to put it another way, the “state” was the “church” in very early Islam. Hence, the caliph was in some sense both a “king” (i.e., a mere secular potentiary) and also “priest” (i.e., a sacerdotal or even soteriological figure). He was made such by the pledge of allegiance, or bayʿa (a covenant at once earthly and salvific); the pledge made the ruler, but it also was
2 7
“God’s Caliph” Revisited
27
understood, by some at least, to be the basis of salvation.87 This recalls, of course, the covenants of ancient Israel, and the caliphs, like the Roman emperors, quickly came to think of themselves as new Davids and new Solomons and, in the case of al-Walīd II at least, as new Adams. However, whereas the Roman emperors, even if they sometimes disliked it, were raised to office by God’s priests, there were no priests in Islam, just the community of the faithful, their past monarchs, and the memory of God’s Prophet. These various sources of authority were the basis of competition over ideological power in early Islam; if al-Walīd II was trying to assert a more plenipotentiary understanding of God’s vicariate, he was probably already looking back to a bygone era (if it ever existed) from his own, very different, context, which is vividly described by his contemporary, Ibn al-Muqaffaʿ. One question that remains is at what point the Muslims’ leaders first began to use their various titles. With the protocollary titles we are on fairly secure ground. It is striking that in one of the very earliest attestations of these—an inscription at the bathhouse at Hammat Gader, on the banks of Lake Tiberias—the inscription itself is in Greek, but Muʿāwiya’s title is transliterated (as opposed to translated) as abd alla Maavia amēra al-moumenēn. That is, in 662–663, these titles were conceived by masons working in Greek, or their patrons, as Arabic titles—even as part of the name of the Arabian monarch.88 Thus, while it might be tempting to see them as translations from the Greek doulos theou, and some combination of autokrator and pistoi, they appear to have already been established in Arabic, and so any such “translation” would date to the earliest conquest period at the latest. Likewise, it is striking that in seventh-and early eighth-century Syriac texts, the word amīr is transliterated as the title for the Arabians’ commanders, indicating that the Arabic term for “commander” was also already in use in the conquest armies.89 These protocollary titles very likely echo the practice of the Hijazi community in its earliest decades. That khalīfa, in the sense of “deputy on earth,” appears in the Qurʾan, and is attested in pre-Islamic South Arabian inscriptions in the sense of “deputy” or “governor” (of the king), might perhaps suggest the same for the roots for this title, but, as we have seen, the evidence is much less secure.90 Landau-Tasseron and Donner look to the 690s for the origin of the title (and Hoyland cautiously suggests the same). ʿAbd al-Malik’s use of khalīfat Allāh on the coinage does stand out as unique, and it is very likely a response to very specific circumstances. It is best understood both
82
28
A ndrew M arsham
as a product of the pressing need to reform the coinage for ideological and fiscal reasons as well as in the wider context of the literary evidence. However, the literary evidence, albeit thin, also tends to suggest a sort of prehistory for the title before it briefly appeared on the coins in the mid- 690s.91 Indeed, the inscription on the coin would have been futile if it did not resonate with existing ideas. Furthermore, it is also very important to note that, by analogy with the Roman situation, such a proclamation on a coin would, by definition, be highly unusual, and so on its own it indicates very little about the prevalence of the concept and its articulation in other, less durable, media. Although the title “God’s caliph” may sometimes have been deployed as a direct riposte to Roman imperial claims, it does not seem necessary to see it as having been directly borrowed or adapted from Roman usage. Rather, it appears that Islam took shape in a context where the same conceptual materials were to hand: the intertextuality of the Qurʾan with various Late Antique religious materials has already been noted. That said, it does seem most likely, pace Crone and Hinds, that the adoption of the title “God’s caliph” as a monarchic and imperial formula was an initiative of the ruling Umayyad family and their allies: after all, the earliest attested poetic use of the title appears in a poem about ʿUthmān attributed to Ḥassān ibn Thābit, apparently dated to the mid-650s. Like al- Akhṭal, Ḥassān had been associated with the Christian, Arabic-speaking tribes of the Syrian desert before he turned his talents to supporting the Muslim elite (al-Akhṭal, in contrast, remained a Christian); the possible allusion to the poetry of the Naṣrid panegyrist al-Nābigha al-Dhubyānī and of al-Akhṭal in the painting Quṣayr ʿAmra would also match this hypothesis.92 In such a context the formula khalīfat rasūl Allāh might have developed in parallel, as part of a discourse within the conquering elite about the proper titles of the ruler. However, in the present state of the evidence, this hypothesis is impossible to prove categorically—after all the term khalīfa is in the Qurʾan and the idea that the ruler was in some sense the image or deputy of gods was a very ancient one in the Middle East.
Notes 1. Crone and Hinds 1986, 116–126. Quotation above at 120, with some modifications to the translation.
92
“God’s Caliph” Revisited
29
2. The letter is unique to al-Ṭabarī’s (d. 923) History. It probably reached him from Aḥmad ibn Zuhayr (d. 892), who was in turn relying on al-Madāʾinī (d. ca. 830– 50): al-Ṭabarī 1879–1901, 2:1755–1756; Ibn al-Nadīm 1971, 115; Sezgin 1967, 311, 320; Rotter 1974, 103–133; Marsham 2009, 147. For skepticism concerning its transmission, see Calder 1987, 375–378. 3. Al-Qāḍī 1992, 215–276; Al-Qāḍī 1993, 285–313; Marsham 2009, 145–155. 4. ʿAbbās 1988, 191, 193, 209, 216, 271, 273, 298 (in two places); see also, for related phrases, 195, 217, 298, 301, 318. Among the former group of letters, those that can be attributed to a particular caliph’s reign belong to those of al-Walīd II and Marwān II (r. 744–747). 5. Crone and Hinds 1986, 6–11, 30–32. The earliest attestation of the title in poetry appears to be in a poem attributed to Ḥassān ibn Thābit (d. before 674), where the third (and first Umayyad) caliph is described as khalīfat Allāh: Ḥassān ibn Thābit 1971, 1:96, no. 20, l.10; Crone and Hinds 1986, 6. Among the other early instances in the poetry are verses attributed to the Iraqi poet Miskīn al-Dārimī (d. ca. 708)—where the caliph Muʿāwiya is called amīr al-muʾminīn, but the Umayyad family in general, “the tribe (or sons) of God’s Caliphs” (banū khulafāʾ Allāh): al-Iṣfahānī 1868, 18:71; Enyclopaedia of Islam, 2nd ed., s.v. “Miskīn al- Dārimī” (by C. Pellat); Marsham 2009, 92–94. 6. A point observed by Crone and Hinds 1986, 12–13. 7. Crone and Hinds 1986, 12–21. 8. Crone and Hinds 1986, 1–3. 9. Crone and Hinds 1986, 42. Cf. Watt 1971, 565–572. 10. Crone and Hinds 1986, 116–126. See now also Judd 2014. 11. Crone and Hinds 1986, 111–115. 12. See, e.g., Berkey 2003, 78–80, 124–129; Hallaq 2005, 43–44; Rippin 2012, 67– 69. In contrast, Al-Azmeh accepts that it was very early, but nonetheless would have the sense of “succession” (just) precede it: Al-Azmeh 1997, 74. Hakim suggests that the Hadith literature indicates the parallel existence of both titles by the early eighth century at the latest—khalīfat Allāh and khalīfat rasūl Allāh—where khalīfat Allāh was not at first controversial: Hakim 2005, 207– 223. Likewise Dagron suggests that the two meanings probably coexisted from the outset: Dagron 2003, 50–53. However, he develops understanding of their implications in new ways through a very suggestive comparison with the Roman situation. 13. For arguments that partially accept and also modify Crone and Hinds’s views on this point, see Zaman 1997; Hallaq 2005, 43–78. 14. Crone 2004, 40. 15. Crone 2004, 195. 16. The Umayyad period is usually dated from ca. 661, but this could be pushed back to 644 (if ʿUthmān’s caliphate is accepted as the beginning of the Umayyad
03
30
A ndrew M arsham
period), and Crone’s current assessment leaves open the possibility of a still earlier date. 17. Cf. Crone and Hinds 1986, 2–3. 18. Landau-Tasseron 2000, 187. However, the “elective” basis of the caliphate is not in fact in tension with its sacral status. 19. Landau-Tasseron 2000, 213. 20. Hakim 2005; a point anticipated by Dagron on the basis of very different evidence Dagron: 2003, 50–53. 21. Hoyland 2006, 405, n.50. 22. Donner 2010, 99, 209–211. 23. Sarris 2011, 299–300. 24. This aspect of the question does not concern Hakim, of course. 25. Crone and Hinds 1986, 11. 26. They are collected in Hoyland 1997, 690–691. 27. ʿAbbās 1988, 191–192, 194, 196, 197–198, 201, 205–208, 213, 215, 216, 217, 234– 235, 237, 253, 267–268, 270, 272–276, 280, 297–299, 305–310, 316–319. 28. Encyclopaedia of Islam, 3rd ed., s.v. “Commander of the Faithful” (by A. Marsham). 29. Bates makes a similar point in 2003, 280–283. 30. For caliphs as “kings,” see Al-Azmeh 1997, passim; Marsham 2009, 1–2. 31. See the discussion in Dagron 2003, 36–53. 32. On the pledge of allegiance as “God’s covenant,” see Marsham 2009, esp. 74–75, 109–110, 169–178. 33. Crone 2004, 40. 34. Grenfell and Hunt 1898, 1:224, no. 140, dated to 550 ad. 35. Chrysos 1978, 31. 36. Eusebius 1902, 215; Drake 1976, 97. Cf. Dvornik 1966, 622. The translation here follows Drake. 37. Cited in Anastos 1978, 37. 38. Bell 2009, 31, citing Corippus 1976, 2:ll.427–428, (= p. 60 [tr. 102]). 39. Thomson, Howard-Johnston, and Greenwood 1999, 2:93, n. 576. 40. Dagron 2003, 48–50. 41. Cameron 1979, 21. 42. Corippus 1976, 2:ll.11–46 (= pp. 48–49 [tr. 94–95]). Cf. Eusebius 1902, 6:208; Drake 1976, 92. 43. Cormack 2000, 50–51. 44. On Agapetus, see Bell 2009, 8–9. 45. Bell 2009, 107 (§21), 120 (§68). Cf. Titus 1:1. 46. Translated in Dagron 2003, 170. 47. Various Qurʾanic and biblical patriarchs are invoked throughout Umayyad panegyric (not to mention in the names of some of the senior members of the family): Crone and Hinds 1986, 44, 54–57. See also the discussion in Crone 2004, 4–11, 40–41. However, this interpretation where the Qurʾanic caliphs are
31
“God’s Caliph” Revisited
31
archetypes of the current ruling caliphs is absent from the (later) exegesis: see Al-Qāḍī 1988, 392–411. It is also implicitly criticized by various later scholars: Crone and Hinds 1986, 21–22, n. 86; 100, n. 17; Hakim 2005. 48. Fowden 2004, 127–141. 49. This is the date proposed by Fowden and certainly inscriptional evidence does strongly suggest that the palace belongs to the period of al-Walīd ibn Yazīd’s heir apparency (wilāyat al-ʿahd): Fowden 2004, 142–163; Imbert 2007, 45–46. See now also Imbert 2016. 50. Fowden 2004; Vibert-Guigue and Bisheh 2007. 51. Fowden 2004, 115–141. 52. See George in this volume on polysemy in Umayyad architecture. 53. Philo 1854–1890, 3:75 (I): “For some persons say, and not without some reason and propriety, that this is the only way by which cities can be expected to advance in improvement, if either the kings cultivate philosophy, or if philosophers exercise the kingly power. But Moses will be seen not only to have displayed all these powers—I mean the genius of the philosopher and of the king—in an extraordinary degree at the same time, but three other powers likewise, one of which is conversant about legislation, the second about the way of discharging the duties of high priest, and the last about the prophetic office; and it is on these subjects that I have now been constrained to choose to enlarge; for I conceive that all these things have fitly been united in him, inasmuch as in accordance with the providential will of God he was both a king and a lawgiver, and a high priest and a prophet, and because in each office he displayed the most eminent wisdom and virtue. We must now show how it is that every thing is fitly united in him.” For what follows, see also Reynolds 2010, 39–54. 54. Bezold 1888, 14. The translation is from Wallis Budge 1927, 53. 55. “The Bereshith or Genesis Rabba,” in Horne 1917, 4:46. 56. Ephraem 1955, 2:17; Ephraem 2010, 94. Further references to similar ideas are found in Ephraem 1998, 12, n. 69. See also Ruzer 2001, 261–262. This interpretation is in distinction from the rabbinical tradition, which does not make this connection between Adam being made in God’s image and his dominion: Gottstein 1994, 185. 57. As noted by Melchert 2011, 120, where he observes, “It is apparently a commonplace among students of Christian apocrypha, although not (yet) students of the Qurʾan, that the Qurʾanic story presupposes this apocryphon,” citing Anderson 2001, 25, and others. See also Reynolds 2010, 39–54. 58. See above, n. 47. 59. Al-Qāḍī 1988. Anthropomorphism was, of course, of ongoing concern to Muslim theologians, but not in connection with this passage. This in spite of a Hadith that refers to Adam “being made in God’s image,” which is discussed by Melchert (2011, 119–120), where he notes that the exegetes do not make anything of this in commenting on the Qurʾan.
23
32
A ndrew M arsham
60. For the suggestion that al-Walīd II also did so in poetry, where he referred to Qurʾanic terminology in defense of his caliphate, see Judd 2008, 439–458. 61. See the discussion of Roman imperial titulature, above in this chapter. 62. Conflicts internal to the Umayyad dynasty may also have had an ideological dimension, as in the case of the conflict between al-Walīd II and Yazīd III: Judd 2008, 454–456. 63. On the “pole-on-steps”, cf. Jamil 1999; Heidemann 2010, 23–34. However, given the paucity of evidence and the multiple possible receptions of this image, it seems wise to conclude that the strongest implication was a clear denial of Roman imperial Christianity, in the context of a hardening of ideological boundaries and ongoing military conflict, especially in the light of the anti-Trinitarian inscriptions that eventually replaced the “pole”: Treadwell 2009, 368, 373. On what follows, see also Bacharach 2010, who emphasizes conflict with the Kharijites in Iraq. 64. Transliterated fairly literally, the slogans would read amīr almwmnīn khlft allh. While abbreviation for reasons of space is possible (but very unusual), it is striking that in Epigraphic South Arabian (ESA), the yāʾ of khalīfa is not marked (and the tāʾ marbūta does not exist) as on the Abraha stele from the mid-sixth century, which refers to one “Yazīd son of Kabshat governor (kh-l-f-t)”: Corpus inscriptionum semiticarum 1881, no. 541, ll. 10–11. On the Syrian copper coins, see Walker 1952, 110. 65. “God’s Victory” is a Qurʾanic phrase, e.g., Q 30:5, 48:3, 110:1. 66. For a recent summary and reinterpretation of the evidence for the reforms, see Treadwell 2009. 67. See, e.g., King 1985, 273–275; Treadwell 2009, 361, 365; Sarris 2011, 299–300. 68. Sarris 2011, 300. 69. Grierson 1999, 7–8. 70. Johns 2003, 426–427. 71. Treadwell 2005, 11; 2009, 370–372; cf. Bacharach 2010. 72. Indeed, Luke Treadwell recently suggested that the creation of a unified bimetallic currency system has been underestimated as a key causal factor in ʿAbd al-Malik’s reform. In this analysis, the “standing caliph” and then the “miḥrāb and ʿanaza” coins were a brief experiment, and the khalīfat Allāh inscription was “to avoid the danger of misidentification” (of the new standing figure as either the Roman emperor or the Sasanian king) and then retained almost by accident on the “miḥrāb and ʿanaza”: Treadwell 2005, 14; 2009, 361, 370. However, Treadwell also acknowledges ideological motivations as well as more practical ones: Treadwell 2009, 361, 365, 369. 73. While the extent of the change between the reigns of Muʿāwiya and ʿAbd al- Malik is debated, there is no doubt that the program of monumental architecture, coin reform, and other public articulations of authority by ʿAbd al-Malik and his son al-Walīd I (r. 705–715) was on an unprecedented scale: see, e.g., Johns 2003; Treadwell 2009, 369; cf. Bacharach 2010.
3
“God’s Caliph” Revisited
33
7 4. Stetkevych 1997, 103, 119–120; Marsham 2009, 102–106. 75. Stetkevych 1997, 107; Marsham 2009, 33–34, 103. 76. On these events, see Marsham 2009, 118–130, with references, and Judd 2008. 77. On the date of the palace, see above, n. 49. On the term walī al-ʿahd, see Marsham 2009, 113–119. 78. Marsham 2009, 155–158. 79. On the law, see Hallaq 2005. On al-Walīd II’s stance, see Judd 2008. 80. Pellat 1976, §34–§38. Jokisch has argued that Hārūn al-Rashīd did codify the law. There is little concrete evidence for this: Jokisch 2007. 81. For the Roman context, see Dagron 2003, passim and esp. 3, 295. 82. See above, n. 72. 83. A point noted by Hallaq 2005, 31. For one apparently early instance of the phrase, see Marsham 2012, 72. The term is Qurʾānic: see, e.g., Q 4:47, 9:48, 9:106. 84. See above, nn. 12 and 20 and cf. Alajmi 2011. 85. Crone and Hinds 1986, 80–96. 86. Crone and Hinds 1986, 115. 87. Crone 2004, 21–23; Marsham 2009, 97–100. 88. Green and Tsafrir 1982. Conversely, the title was translated into Pahlavi on some seventh-century coins: Hoyland 1997, 690. 89. As in the case of the “Maronite Chronicle,” which dates to before 727, and probably before 680: Brooks 1955, 71; Palmer and Brock 1993, 32. For the term in a South Arabian inscription, see Corpus inscriptionum semiticarum 1881, 4:363, ll. 3–4. For early Arabic attestations, see Marsham 2012, 83, with references. 90. For the Qurʾān, see above pp. 13 and 15 and nn. 47 and 48; for the South Arabian inscriptions, see above, n. 64. 91. See above, nn. 5 and 20. 92. See above, pp. 23–24, and n. 76.
Bibliography ʿAbbās, I. 1988. ʿAbd al-Ḥamīd b. Yaḥyā al-Kātib: wa-mā tabqā min rasāʾilihi wa-rasāʾil Sālim Abī al-ʿAlā. Amman: Dār al-Shurūq. Alajmi, A. 2011. “ʿUlama and Caliphs New Understanding of the God’s Caliph Term.” Journal of Islamic Law and Culture 13: 102–112. Anderson, G. A. 2001. The Genesis of Perfection: Adam and Eve in Jewish and Christian Imagination. Louisville, KY: Westminster John Knox Press. Anastos, M.A. 1978. “Byzantine Political Theory: Its Classical Precedents and Legal Embodiment.” In The “Past” in Medieval and Modern Greek Culture, edited by S. Vryonis, 13–53. Malibu, CA: Undena. Al-Azmeh, A. 1997. Muslim Kingship: Power and the Sacred in Muslim, Christian and Pagan Polities. London: I. B. Tauris.
43
34
A ndrew M arsham
Bacharach, J. 2010. “Signs of Sovereignty” The Shahāda, Qurʾānic Verses, and the Coinage of ʿAbd al-Malik.” Muqarnas 27: 1–30. Bates, M. 2003. “Khurāsānī Revolutionaries and al-Mahdī’s Title.” In Culture and Memory in Medieval Islam: Essays in Honour of Wilferd Madelung, edited by F. Daftary and J. W. Meri, 280–283. London: I. B. Tauris. Bell, P. N. 2009. Three Political Voices from the Age of Justinian: Agapetus, Advice to the Emperor, Dialogue on political science; Paul the Silentiary, Description of Hagia Sophia. Liverpool: Liverpool University Press. Berkey, J. P. 2003. The Formation of Islam: Religion and Society in the Near East. Cambridge: Cambridge University Press. Bezold, C. ed. 1888. Die Schatzhöhle: nach dem syrischen Texte der Handschriften zu Berlin, London und Rom nebst einer arabischen Version nach den Handschriften zu Rom, Paris und Oxford. Leipzig: J.C. Hinrichs. Brooks, E. W. ed. 1955. Chronica Minora (Textus), “Corpus Scriptorum Christianorum Orientalium 3, Scriptores Syrii 3.” Leuven. Calder, N. 1986. “Patricia Crone and Martin Hinds, God’s Caliph: Religious Authority in the First Centuries of Islam, Cambridge University Press, 1986.” Journal of Semitic Studies 22: 375–378. Cameron, A. 1979. “Images of Authority: Elites and Icons in Late Sixth-Century Byzantium.” Past and Present 84: 3–35. Canivet, P., and M. T. Canivet 1987. Ḥūarte: sanctuaire chrétien d’Apamène IVe–VIe s. 2 vols. Paris: Librairie Orientaliste Paul Geuthner. Chrysos, E. K. 1978. “The Title Βασιλευς in Early Byzantine International Relations.” Dumbarton Oaks Papers 32: 29–75. Corippus. 1976. In laudem Iustini Augusti minoris libri IV. Edited and translated by A. Cameron. London: Athlone. Cormack, R. 2000. Byzantine Art. Oxford: Oxford University Press. Corpus inscriptionum semiticarum ab Academia Inscriptionum et Litterarum Humaniorum conditum atque digestum: Pars IV inscriptiones Ḥimyariticas ae Sabæas continens. 1881. Paris: E Reipublicae Typographeo. Crone, P. 2004. Medieval Islamic Political Thought. Edinburgh: Edinburgh University Press. Crone, P., and M. Hinds. 1986. God’s Caliph: Religious Authority in the First Centuries of Islam. Cambridge: Cambridge University Press. Dagron, G. 2003. Emperor and Priest: The Imperial Office in Byzantium. Translated by J. Birrell. Cambridge: Cambridge University Press. Donner, F. M. 2010. Muhammad and the Believers: At the Origins of Islam. Cambridge, MA: Belknap Press of Harvard University Press. Drake, H. A. 1976. In Praise of Constantine: A Historical Study and New Translation of Eusebius’ Tricennial Orations. Berkeley: University of California Press. Dvornik, F. 1966. Early Christian and Byzantine Political Philosophy: Origins and Background. Washington, DC: Dumbarton Oaks Center for Byzantine Studies.
35
“God’s Caliph” Revisited
35
Ephraem. 1955. Sancti Ephraem Syri In Genesim et in Exodum commentarii. Edited and translated by R. Tonneau. Leuven: L. Durbecq. ———. 1998. The Armenian Commentary on Genesis attributed to Ephrem the Syrian. Edited by E. G. Mathews. Leuven: Peeters. ———. 2010. Selected Prose Works: Commentary on Genesis, Commentary on Exodus, Homily on Our Lord, Letter to Publius. Translated by E. G. Mathews and J. P. Amar. Washington: Catholic University of America Press. Eusebius. 1902. “Tricennatsrede an Constantin.” In Eusebius, Band 1, edited by I. A. Heikel, 193–259. Leipzig: J. C. Hinrichs. Fowden, G. 2004. Quṣayr ʿAmra: Art and the Umayyad Elite in Late Antique Syria. Berkeley: University of California Press. Gottstein, A. 1994. “The Body as Image of God in Rabbinic Literature.” Harvard Theological Review 87: 171–195. Grabar, A. 1957. L’iconoclasme byzantin: Dossier archéologique. Paris: Collège de France. Green, J. and Y. Tsafrir. 1982. “Greek Inscriptions from Ḥammat Gader: A Poem by the Empress Eudocia and Two Building Inscriptions.” Israel Exploration Journal 32: 77–96. Grenfell, B. P., and A. S. Hunt, eds. 1898. The Oxyrhynchus Papyri. London: Egypt Exploration Fund. Grierson, P. 1999. Byzantine Coinage. Washington, DC: Dumbarton Oaks. Hakim, A. 2005. “‘Umar b. al-Khaṭṭāb and the title khalīfat Allāh: a Textual Analysis.” Jerusalem Studies in Arabic and Islam 30: 207–223. Hallaq, W. B. 2005. The Origins and Evolution of Islamic Law. Cambridge: Cambridge University Press. Ḥassān ibn Thābit. 1971. Dīwān of Ḥassān ibn Thābit. Edited by W. ‘Arafat. London: Luzac. Heidemann, S. 2010. “The Standing Caliph Type: The Object on the Reverse.” In Coinage and History in the Seventh Century Near East: Proceedings of the 12th Seventh Century Syrian Numismatic Round Table Held at Gonville and Caius College, Cambridge on 4th and 5th April 2009, edited by A. Oddy, 23–34. London: Archetype. Horne, C. F., ed. 1917. The Sacred Books and Early Literature of the East, vol. 4. Medieval Hebrew. New York: Parke. http://www.sacred-texts.com/jud/mhl/ mhl05.htm (accessed July 16, 2013). Hoyland, R. 1997. Seeing Islam as Others Saw It: A Survey and Evaluation of Christian, Jewish and Zoroastrian Writings on Early Islam. Princeton, NJ: Darwin. ———. 2006. “New Documentary Texts and the Early Islamic State.” Bulletin of the School of Oriental and African Studies 69(3): 395–406. Ibn al-Nadīm. 1971. Kitāb al-Fihrist. Edited by R. Tajaddud. Tehran: Dānishgāh. Imbert, F. 2007. “Note Épigraphique et Paléographique.” In Les Peintures de Qusayr ‘Amra: Un bain omeyyade dans la bâdiya jordanienne, by C. Vibert-Guigue and G. Bisheh, 45–46. Amman: Institut Français d’archéologie du Proche-Orient.
63
36
A ndrew M arsham
______. 2016. “Le prince al-Walīd et son bain: itinéraires épigraphiques à Qusayr ʿAmra.” Bulletin d’Études Orientales 64: 321–363. al-Isfahānī. 1868. Kitāb al-Aghānī. 20 vols. Edited by Naṣr Hūrīnī. Būlāq: Dār al-Ṭibāʿah al-ʿĀmirah. Jamil, N. 1999. “Caliph and Quṭb: Poetry as a Source for Interpreting the Transformation of the Byzantine Cross on Steps on Umayyad Coinage.” In Bayt al-Maqdis: Jerusalem and Early Islam, edited by J. Johns, 11–57. Oxford: Oxford University Press. Johns, J. 2003. “Archaeology and the History of Early Islam: The First Seventy Years.” Journal of the Economic and Social History of the Orient 46(4): 411–436. Jokisch, B. 2007. Islamic Imperial Law: Harun- al- Rashid’s Codification Project. Berlin: Walter De Gruyter. Judd, S. 2008. “Reinterpreting al-Walīd b. Yazīd.” Journal of the American Oriental Society 128(3): 439–458. ———. 2014. Religious Scholars and the Umayyads: Piety-Minded Supporters of the Marwānid Caliphate. London: Routledge. King, G. R. D. 1985. “Islam, Iconoclasm, and the Declaration of Doctrine.” Bulletin of the School of Oriental and African Studies 48(2): 267–277. Landau- Tasseron, E. 2000. “From Tribal Society to Centralized Polity: An Interpretation of Events and Anecdotes in the Formative Period of Islam.” Jerusalem Studies in Arabic and Islam 24: 180–216. Marsham, A. 2009. Rituals of Islamic Monarchy: Accession and Succession in the First Muslim Empire. Edinburgh: Edinburgh University Press. ———. 2012. “The Pact (Amāna) between Muʿāwiya ibn Abī Sufyān and ʿAmr ibn al- ʿĀṣ (656 or 658 ce): “Documents” and the Islamic Historical Tradition.” Journal of Semitic Studies 57(1): 69–96. Melchert, C. 2011. “God Created Adam in His Image.” Journal of Qurʾanic Studies 13: 113–124. Miles, G. C. 1952. “Miḥrāb and ʿAnazah: A Study in Early Islamic Iconography.” In Archaeologica Orientalia in Memoriam Ernst Herzfeld, edited by G. C. Miles, 155– 171. New York: J. J. Augustin. Palmer, A. and S. Brock, eds. 1993. The Seventh Century in the West-Syrian Chronicles. Liverpool: Liverpool University Press. Pellat, C. 1976. Ibn al-Muqaffaʿ : mort vers 140/757, “Conseilleur” du Calife. Paris: G.-P. Maisonneuve et Larose. Philo. 1854–1890. “On the Life of Moses, Book II.” In The Works of Philo Judaeus: The Contemporary of Josephus, vol. 3, translated by C. D. Yonge, 74–88. London: G. Bell & Sons. Al-Qāḍī, W. 1988. “The Term “Khalīfa” in the Exegetical Literature.” Die Welt des Islams, New series 28.1: 392–411. ———. 1992. “Islamic State Letters: The Question of Authenticity.” In The Byzantine and Early Islamic Near East I: Problems in the Literary Source Material, edited by A. Cameron and L. I. Conrad, 215–276. Princeton, NJ: Darwin.
37
“God’s Caliph” Revisited
37
———. 1993. “The Impact of the Qurʾān on the Epistolography of ʿAbd al-Ḥamīd.” In Approaches to the Qurʾān, edited by G. Hawting and A. A. Shareef, 285–313. London: Routledge. Reynolds, G. 2010. The Qurʾān and Its Biblical Subtext. London and New York: Routledge. Rippin, A. 2012. Muslims: Their Religious Beliefs and Practices. 4th edition, Abingdon: Routledge. First published 1990. Rotter, G. 1974. “Zur Überlieferung einiger historischer Werke Madāʾinīs in Ṭabarīs Annalen.” Oriens 23: 103–133. Ruzer, S. 2001. “The Cave of Treasures on Swearing by Abel’s Blood and Expulsion from Paradise: Two Exegetical Motifs in Context.” Journal of Early Christian Studies 9: 251–271. Sarris, P. 2011. Empires of Faith: The Fall of Rome to the Rise of Islam, 500–700. Oxford: Oxford University Press. Sezgin, F. 1967. Geschichte des arabischen Schriftums, Band 1: Qur’anwissenschaften, Hadit Geschichte, Fiqh, Dogmatik, Mystik bis ca. 430 H. Leiden: Brill. Stetkevych, S.P. 1997. “Umayyad Panegyric and the Poetics of Islamic Hegemony: Al- Akhtal’s Khaffa Al-Qaṭīnu (‘Those That Dwelt with You Have Left in Haste’).” Journal of Arabic Literature 28(1–2): 89–119. al-Tabarī. 1879–1901. Taʾrīkh al-rusul wa’l-mulūk. 15 vols. Edited by D. Heinrich, E. Prym, J. Barth, M. J. de Goeje, S. Guyard, T. Nöldeke, and V. R. Rozen et al. Leiden: Brill. Thomson, R. W., J. Howard-Johnston, and T. Greenwood. 1999. The Armenian History Attributed to Sebeos. Liverpool: Liverpool University Press. Treadwell, L. 2005. “ ‘Mihrab and ʿAnaza’ or ‘Sacrum and Spear’: A Reconsideration of an Early Marwanid Silver Drachm.” Murqanas 22: 1–28. ———. 2009. “ʿAbd al-Malik’s Coinage Reforms: The Role of the Damascus Mint.” Revue numismatique 6(165): 357–381. Vibert-Guigue, C., and G. Bisheh. 2007. Les Peintures de Qusayr ‘Amra: Un bain omeyyade dans la bâdiya jordanienne. Amman: Institut Français d’Archéologie du Proche-Orient. Walker, J. 1952. “Some New Arab-Sasanian Coins.” Numismatic Chronicle 6(12): 106–110. Wallis Budge, E. A. 1927. The Book of the Cave of Treasures. London: R.T.S. Watt, M. 1971. “God’s Caliph: Qurʾānic Interpretations and Umayyad Claims.” In Iran and Islam: A Volume in Memory of Vladimir Minorsky, edited by C. E. Bosworth, 565–572. Edinburgh: Edinburgh University Press. Zaman, M. Q. 1997. Religion and Politics under the Early ʿAbbāsids: The Emergence of the Proto-Sunni Elite. Leiden: E. J. Brill.
83
93
2
Paradise or Empire? On a Paradox of Umayyad Art Alain George*
R E L I G I O U S A R C H I T E C T U R E was the earliest expression of Islamic art on a grand scale, and it has continued to exert its fascination upon generation after generation of visitors ever since. To the modern student, it is also a field riddled with elusive meanings and apparent paradoxes. For decades, divergent interpretations have been put forward about its two most emblematic monuments: the Dome of the Rock, in Jerusalem, and the Great Mosque of Damascus. Indeed, the more one reflects about them, the more the same conclusions seem to repeat themselves, with a single referent—the representation of a plant, a precious stone, a building—seemingly evoking both paradise and empire. This puzzling situation is, of course, partly due to the lack of human figures and identifying captions in their iconography, leaving the viewer to construe, after a gap of over a millennium, what may have been obvious to original audiences. What is more, any imperially sponsored sacral building or object is inherently bound to be a manifestation of both spirituality and power. Yet the reality observed in practice goes beyond the level of truisms, reaching a degree of articulation that suggests the possibility of a deliberate choice, a conscious tendency to elude one-dimensional readings and to conflate the spiritual and earthly planes. This possibility and its cultural context form the subject of the present chapter, with a focus U M AY YA D
The research for this chapter was undertaken with the support of the Leverhulme Trust, which is gratefully acknowledged. *
04
40
A lain G eorge
on the Dome of the Rock, the Great Mosque of Damascus, and Umayyad Qurʾan illuminations.
The Dome of the Rock The Dome of the Rock was built in 692 by the caliph ʿAbd al-Malik (r. 685–705), which makes it the earliest Islamic monument still standing in something close to its original state. Its wall mosaics are concentrated on the inner octagon and drum, covering almost all of their vertical surfaces above column level, as well as their soffits. With a total area of some 1,280 square meters, they are the largest repository of this art form to survive from before the twelfth century.1 Initially, similar mosaic ornament clad the exterior of the building, thereby proclaiming a message not only to those within its walls, but also to Jerusalem at large.2 The extant decoration has been the subject of numerous studies.3 It essentially consists of semi-abstract depictions of plants, fertile, lush and ever expanding, along with more naturalistic trees. These vegetal elements are intermingled with jewels, crowns, and precious stones. In most panels, both thematic strands are blended, with plants carrying jewels, crowns, pearls, and precious stones (fig. 2.1). The resonance of this decoration with the Qurʾanic imagery of paradise has been articulated by several modern writers.4 Yet as one delves into the iconography, unmistakable references to earthly rulership also emerge.5 As the latter have received less emphasis in recent scholarship, it is with them that I begin.
Crowns, Jewels, and Rulership In the spandrels of the inner octagon and most of all in the drum, several bejeweled vases with plants carry a pair of wings at their apex (fig. 2.1).6 These wings reflect a type of pre-Islamic crown widely attested in Sasanian coins, metalwork, and stuccoes from the fifth century onward and perpetuated in Islamic coinage from Iran and Iraq down to the very years in which the Dome of the Rock was built.7 The correspondence is precise: note, in particular, the geometrical patterning of the bejeweled base of each wing, the transversal line that separates this base from the raised feather tips, and the way the uppermost feather curves into a hook at its tip. Most pairs of wings frame pointed, bejeweled ovals that are themselves adorned with crowns, diadems, and pearls. The two pairs of wings in the inner octagon are
41
Paradise or Empire?
41
Figure 2.1 Mosaic decoration of the drum. Jerusalem, Dome of the Rock, 72 ah/692 ad.
themselves part of a range of other regalia, which makes it clear that they are crowns. In one of them, a pole carrying a moon crescent emerges from the wings, thereby making the analogy with Sasanian crowns complete (fig. 2.2). It has been suggested that these wings might represent angels.8 But upon close consideration, this seems unlikely. In its Iranian context, the winged crown signified the ruler’s khvarnah (an Old Persian term, New Persian farr), his good fortune of divine origin, or in the words of the Avestan Zāmyād yasht (fifth century bc and later), “the mighty, gleaming glory created by [Ahura] Mazda,” the “Wise Lord.”9 The wings conveyed the spiritual dimension of divinely ordained kingship, expressed in different forms that have in common their royal connotations: in some Sasanian stuccoes, silvers, seals, and textiles, the same symmetrical pair of wings thus serves as the base for another motif, such as a moon crescent (like those often seen in crowns; it appears, in fig. 2.3, with the Pahlavi word nishān, “[royal] emblem”), or the head of a ram (itself also associated with khvarnah), again with a floating ribbon.10
24
Figure 2.2 Mosaic decoration, western arcade of inner octagon. Jerusalem, Dome of the Rock, 72 ah/692 ad.
Figure 2.3 Stucco plaque from the area of Ctesiphon, Iraq, sixth or seventh century. Height 39.5 cm. Berlin, Museum für Islamische Kunst, KtO 1084.
43
Paradise or Empire?
43
This was thus the clear connotation of the motif in a Sasanian context, and the intention was apparently to retain it at the Dome of the Rock. This is, first of all, a matter of composition: the mosaic wings are shown as a small part of larger arrangements of jewels, crowns, and plants, too small and enclosed to act as primary signifiers for whole beings. For local viewers in Jerusalem, the motif would have been reminiscent of Arab-Sasanian coins still very recently minted in the eastern parts of the Islamic empire, where it capped the head and crown of the ruler, but fundamentally at odds with the Christian iconography of angels, in which the wings are more naturalistic, have a less compact shape, and are typically shown at rest, pointing downward, rather than upward, or extending horizontally if in flight.11 Such would also have the habitual repertoire of the mosaicists who worked on the Dome of the Rock (indeed, even in Sasanian art, comparable modes of depiction were adopted for celestial beings—mostly fravashis, protective spirits).12 In other words, the wings seem to represent the conscious import of a distinctively Sasanian royal motif into this Umayyad monument. This gesture of appropriation, which finds some echoes in sixth-century Constantinople, would have been all the more potent in early Islamic Jerusalem, at the hands of the polity that had recently brought about the final demise of the Sasanians.13 Its meaning, as will soon become apparent, probably combined a political and a votive dimension. Jewels and crowns adorn much of the remaining mosaic decoration at the Dome of the Rock with such abundance as to conjure an imagined treasury. These precious objects could carry spiritual connotations linked to the divine source of earthly rule and to the jewels of paradise, two themes already present in early Christian mosaics. Notwithstanding this possible layer of meaning, to which we shall return, at an immediate level, they evoke kingship, looking west as well as east. The round crowns inlaid with pearls and colored precious stones are reminiscent of the crowns of Justinian (r. 527–565) depicted in sixth-century mosaics at Ravenna, in the church of San Vitale, and possibly Sant’ Apollinare Nuovo, while the larger crowns with pearl hangings recall the tiara of Theodora, Justinian’s empress, again at San Vitale (fig. 2.4).14 Both scenes at San Vitale are themselves framed by mosaic depictions of gold columns inlaid with jewels and set within a bejeweled rectangle. These representations, in turn, reflect real jewelry of the period. Crowns stemming from the Byzantine tradition have survived at the treasury of the Lombard queen Theodelinda in Monza (ca. 600) and at the Visigothic treasury of Guarrazar, in Spain (seventh century). The contents
4
44
A lain G eorge
Figure 2.4 Mosaic panel showing Empress Theodora with attendants. Ravenna, Church of San Vitale, ca. 527–546 AD.
of the latter, in particular, are so close in quality and style to the finest extant Byzantine jewels that they may have been executed by Byzantine craftsmen brought to Toledo.15 The most lavish crown from Guarrazar, that of king Recceswinth (r. 653–672), is inlaid with rock crystal, pearls, sapphires, and other precious stones (plate 2 in the insert): its form and adornment directly resonate with the mosaic crowns at the Dome of the Rock (compare it, for example, with the crown in the lower register of the spandrel in figure 2.2).16 These objects were made not to be worn, but suspended above church altars as offerings from kings and dignitaries. This usage is attested not only at the Visigothic court but also, at least as early as the sixth century, at Hagia Sophia in Constantinople and the Holy Sepulchre in Jerusalem, a few steps away from the Dome of the Rock. In this context, they symbolized surrender of sovereignty to God, the ultimate source of power, a theme also emphasized in the mosaic inscription at the Dome of the Rock; in the same vein, Christ, the Virgin, and Christian saints were often represented with such regalia.17 As long ago argued by Oleg Grabar, the mosaic crowns at the Dome of the Rock might likewise have represented votive offerings signifying the
45
Figure 2.5 Relief sculpture, Tāq-i Bustān, Iran, fifth to seventh century.
64
46
A lain G eorge
submission of rulers defeated at the hands of Islam, on this most sacred spot associated with Creation and Judgement in Syrian Muslim circles.18 A tradition recorded by al-Wāsiṭī (Jerusalem, wr. before 410/1019) as the narration of a servant (khādim) appointed to the upkeep of the Dome of the Rock under the Umayyads adds a further element to this layer of meaning: “During the time of ʿAbd al-Malik, there was hanging on the chain above the Rock under the dome the Yatīma pearl, the horns of Abraham’s ram and the crown of Kisrā [the Sasanian emperor]. When the Banū Hāshim [the Abbasids] took over the caliphate, they sent them to the Kaʿba.”19 This anecdote, like the vast majority of those about the sacred Rock, is derived from an earlier compilation by al-Walīd ibn Ḥammād al-Ramlī (d. 912). It thus existed in written form by the ninth century, and much of this material appears to have entered into circulation in Jerusalem by the late Umayyad period.20 This is particularly likely for the present anecdote, first of all because of the chain of textual transmission (isnād), going from al-Ramlī through one ʿAbd al-Raḥmān ibn Muḥammad to the latter’s father and grandfather, with his great-great grandfather as the narrator. This makes the story akin to a family memory handed down over little more than a century. Second, the pattern of sending regalia and precious objects to sacred monuments is confirmed, with respect to Mecca, by several independent sources, including the near- contemporary Syriac chronicle of Theophilus of Edessa (ca. 695–775), as reconstructed from later citations. According to these sources, the crown of Kisrā, the horns of Abraham’s ram, and two ruby-encrusted crescents from Damascus were kept at Mecca before the days of ʿAbd al-Malik. Crowns, other precious objects, and spoils of war continued to be sent there in subsequent decades.21 Whether the crown and horns at the Dome of the Rock were different from those in Mecca or were taken from there to Jerusalem upon ʿAbd al-Malik’s defeat of Ibn al-Zubayr in 692 (and eventually brought back to Mecca by the Abbasids) cannot be determined.22 But the nature of the source and ubiquity of the practice make it entirely plausible that such objects did exist at the Dome of the Rock in the Umayyad period. Hanging from the dome, they would have found multiple echoes in the surrounding drum, where most of the mosaic crowns are concentrated. This contiguity of real and imagined objects, this juxtaposition of ornaments and colors across different media, is reminiscent of the Late Antique aesthetic mode labeled the “jeweled style” by Michael Roberts.23
4 7
Paradise or Empire?
47
In sum, the mosaic crowns and jewels at the Dome of the Rock carried incontrovertible connotations of earthly rulership, both Sasanian and Byzantine, expressed in concrete terms, with a votive dimension. Yet this is not the end of the story, for motifs in the same compositions do also resonate with the Muslim imagery of paradise.
Touching Paradise, Making the Rain Come The Qurʾan evokes paradise as a fertile garden underneath which rivers flow, with gushing springs, lush plants, lofty chambers, and precious stones. To cite but one of the many verses about its denizens: Those—theirs shall be gardens of Eden, underneath which rivers flow; therein they shall be adorned with bracelets of gold, and they shall be robed in green garments of silk and brocade, therein reclining upon couches—O, how excellent a reward! And O, how fair a resting place! (Q. 18:31)24 In the mosaics, most of the plants carrying crowns are fantastical scrolls of fruit-bearing vine rooted in precious vases. The theme of the vine stemming from a vase was ubiquitous in Christian art of the period, though at the Dome of the Rock, the schematized, symmetrical form of this opulent growth is also reminiscent of Sasanian art (fig. 2.5). In a Christian context, the motif would have naturally evoked the True Vine, as explicitly shown by a stone carving from Dayr al- Zaʿfarān, in northern Mesopotamia (sixth century), where a cross around which vines scroll is rooted in a vase. A similar motif, with the vase on one panel and the cross on the next, appears on an ivory pyxis probably produced in Greater Syria in the Umayyad period.25 At the Dome of the Rock and in the normative context of the Qurʾan, these images would have rather resonated with the supernatural fertility of paradise. The otherworldly character of the plants is suggested by their inlaid jewels and the frequent appearance of gold in their nervures, bringing light into their very fabric. The marble paneling below the mosaics, with its wave patterns, might have evoked water sliding down from this fantastical garden, both within and without the monument, a visual impression reinforced by their coolness to the touch. Such associations with marble are, at least, known from numerous Byzantine texts and from rarer Arabic buildings descriptions of the ninth century onward.26 In this
84
48
A lain G eorge
reading, the monument offers an almost tactile glimpse of paradise on earth, both through its specific motifs and through the overall aesthetic impact of the decoration.27 Some of the plants, however, remain free of jewels and gold and appear more earthly in nature. In the Qurʾan, God is also lauded as the giver of rain, water sent down from the sky (mentioned twenty-six times, alongside other references to rain).28 In a world where seasonal irrigation lay at the root of agriculture, the life-giving nature of rain, year after year, was felt more acutely than in modern societies. In the ideological construct of the early Islamic empire, the caliph played a pivotal role in obtaining this divine gift for the community, as intercessor and spiritual pole. In terms familiar from court poetry of this period, al-Akhṭal thus eulogized ʿAbd al- Malik as “the caliph of God through whom men pray for rain,” the executor of divine victory, justice, and the light and guidance of the erring.29
The Qurʾanic Art of Polysemy Ultimately, two visions emerge from the mosaics: the surrender of earthly power to God combined with God-given fertility (two pillars of caliphal ideology in early Islam), and the ethereal landscapes of paradise. Each is supported by specific elements, so that neither can be dismissed lightly, but each may also be extended individually to the whole decoration. One is thus tempted to ask whether aesthetic sensibilities had any role to play in this puzzling situation. Polysemy was a core value of Arabic poetry, the most developed art form in pre-Islamic Arabia, which retained unrivaled cultural standing in the Islamic era. It is, even more importantly, a fundamental aspect of the Qurʾan, where clear admonitions rhythmically alternate, often in the same verse, with words or clauses that have a wide openness of meaning.30 Far from being accidental, or from solely reflecting semantic difficulties, this was a self-proclaimed value of the Qurʾanic text: It is He who sent down upon thee the Book, wherein are verses clear (muḥkamāt) that are the Essence of the Book, and others ambiguous (mutashābihāt). As for those in whose hearts is swerving, they follow the ambiguous part, desiring dissension, and desiring its interpretation; and none knows its interpretation, save only God. And those firmly rooted in knowledge say, “We believe in it; all is from our Lord”; yet none remembers, but men possessed of minds. (Q. 3:7)31
94
Paradise or Empire?
49
The interpretation of this verse has itself been the subject of much discussion on the part of Qurʾan commentators (not unlike the toil of modern scholars seeking to understand the Dome of the Rock). But such ambiguity was perceived, in relation to the Qurʾan, as the source of inexhaustible depths of meaning.32
The Umayyad Qurʾan of Fustat An additional piece of evidence should now be considered: the “Umayyad codex of Fustat” (thus labeled by François Déroche after its city of discovery in the nineteenth century; previously also referred to as “Marcel 13”).33 This Qurʾan manuscript is closely related to the Dome of the Rock through its illuminations, and previous studies have shown that it was probably produced in Greater Syria around the reign of ʿAbd al-Malik.34 The heading of sura 38 (Ṣād) carries the image of a green column topped by a niche with a hanging lamp, and, above, a second, red column with a fluted fountain at its apex (fig. 2.6). The lamp conjures associations with divine light, potently expressed in the famous “light verse”: God is the Light of the heavens and the earth; the likeness of His Light is as a niche wherein is a lamp, the lamp in a glass, the glass as it were a glittering star, kindled from a Blessed Tree, an olive that is neither of the East nor of the West whose oil well nigh would shine, even if no fire touched it. Light upon light, God guides to His light whom He will. (Q. 24:35) Once again, the verse is fundamentally polysemic. Earth and heaven are not neatly set apart but melded in a succession of images that open onto a boundless spiritual horizon. The lamp in the illumination, likewise, could
Figure 2.6 Illumination band marking the beginning of sura 38. Total page dimensions 37 × 31 cm. Saint Petersburg, National Library of Russia, Marcel 13, f. 33r.
05
50
A lain G eorge
naturally signify divine light, yet, with its two strings, girdles, and glass casing, it is also a physical object. Thousands of lamps were deployed across the ḥaram al-sharīf, the esplanade of Temple Mount. They were reportedly lit all day long; some served to mark out holy spots, while scented lamps suspended on chains were concentrated inside the Dome of the Rock.35 The fountain, in turn, resonates with the Qurʾan’s gushing waters, fountains, and rivers of paradise.36 According to al-Ṭabarī (d. 310/923), the early Qurʾan commentators Mujāhid (Meccan, d. 104/722) and Hishām ibn al-Kalbī (Kufan, d. 146/763) saw the fount of paradise as pouring its waters downward from higher spheres.37 Closer to the probable geographical origin of the illumination, traditions from Jerusalem state that the four rivers of paradise flowed beneath the sacred Rock. This vision, being rooted in the Old Testament (Ezekiel 47:1–12), is likely to have emerged at an early stage in the Islamic era.38 On an earthly level, the image can also be associated with the gushing fountains or springs brought forth by God as sources of human sustenance (as for instance in Q. 36:34). The latter image also resonates with early Syrian traditions according to which all the sweet water of the world and its mountains flowed from beneath the Rock. At the same time, the particular type of this fountain, fluted, standing on a column, with its pistil-shaped spout occurs as a palatial attribute in the Theodora mosaic panel at San Vitale (sixth century, fig. 2.4), in the Rotunda mosaics at Thessaloniki (fifth century), and in a Syriac manuscript at the Bibliothèque Nationale de France (sixth century).39 It is thus, once again, a concrete object. The red and green columns, with their tiny white specks, also echo a distinctive type of column found at the Dome of the Rock.40 Numerous early Syrian traditions compiled in the ninth century by al-Ramlī construed the sacred Rock around which the monument is built as an earthly locus that touches paradise, the site of God’s Throne on earth, from which He ascended to heaven after Creation, and as the navel of the world.41 The columns in this Qurʾan could thus notionally have conveyed an additional symbolical layer of meaning, for instance as a celestial pole uniting heaven and earth; but the motif is so generic as to preclude any certainties. Taking the image as a whole, a paradisiacal reading may be put forward, but an interpretation of the illumination as exalting Umayyad architectural programs is equally possible, given the way the latter are outwardly referenced. The latter dimension is even more obvious in the remaining sura headings from the same Qurʾan, most of which directly mirror ornaments at the Dome of the Rock and in other Umayyad monuments.42
51
Paradise or Empire?
51
Moving one step further, the very dichotomy on which this distinction rests—between heaven and earth, the spiritual realm and earthly rule—is not entirely self-evident. An Arabic tradition about Temple Mount and the Rock reads: “The sons of Hārūn (Aaron), may God pray for him, used to come to the Rock and call it the Temple in Hebrew, and a fountain of olive oil would descend unto them, and it [the oil] would circle and fill the lamps without [human] contact.”43 The narration shares with the above illumination the image of a lamp and fountain, although again it is not possible to assert any direct relationship between them. It also dims the boundary between the earthly and heavenly realms in a way often encountered in this period. A tradition recorded by Ibn al-Murajjā (Jerusalem, mid-eleventh century) states that an area on its esplanade known as “the miḥrāb [prayer niche or space] of our Prophet Muḥammad” stood opposite the “lamp of God” (qindīl Allāh), also known as the “lamp of paradise” (qindīl al-janna).44 These names evoke sacred objects set in physical space, while opening onto a spiritual realm of undefined expression. The four rivers of paradise that were believed to flow beneath the Rock were named, in one tradition, as Jaxartes (Sayḥān), Oxus (Jayjān), the Nile, and the Euphrates: in other words, earthly and heavenly rivers were conceptually blended. This particular anecdote again stems from a biblical source (Genesis 2:10– 14), which implies the likelihood of an early date.45 Such traditions could be multiplied. The precise period at which they entered into circulation is subject to interpretation, but at least some of them are early, and they collectively reflect a worldview that cannot have emerged suddenly. It implies that in the physical space of Temple Mount, as often with early Islamic holy sites, both levels of reality, the physical and spiritual, were perceived as conjoined, rather than separate. The conceptual boundary was pliable, as it might likewise have been in monumental mosaics or manuscript illumination. These conflations make the words of Peter Brown resonate with an early Muslim context: Like their pagan rivals, Christians thought of themselves as enveloped, for good or ill, in a mundus, a visible universe shot through with etherial (and so, usually, though not invariably, invisible) benign and hostile powers. Their problem was not to envision a so-called Other World, a world “out there.” … What mattered was to bring into their own, perilous existence in this world touches of paradise—a region of delight hauntingly adjacent to themselves.46
25
52
A lain G eorge
A kindred perspective is, again, implicit in the Qurʾan, notably with the jinn, invisible creatures fashioned by God from “smokeless fire” (Q. 55:14) that inhabit the earth and, most of all, through the omnipresence of God himself. As far as man-made images are concerned, however, one is left, as in poetry, with an open field of possibilities: a field that may be mapped but intrinsically resists being narrowed down.
The Great Mosque of Damascus ʿAbd al-Malik’s son and successor, al-Walīd (r. 705–715), was the foremost patron of religious architecture in the Umayyad era, being responsible for the building, rebuilding, or redecoration of the mosques of Damascus, Medina Mecca, Fustat, Sanaa, and the Aqsa. The Great Mosque of Damascus initially had its whole courtyard and prayer hall covered with mosaic decoration above the level of its marble dadoes and columns. The largest extant sections lie on the western arcade and on the inner and outer façades of the axial nave. They show landscapes with tall trees adjoined by buildings of different shapes and sizes, set against skies of gold with water running in the foreground. The buildings are adorned with jewels, particularly pearls that hang from gates (fig. 2.7).
Figure 2.7 Detail of the western arcade seen from the courtyard. Great Mosque of Damascus, 706–715 ad.
53
Paradise or Empire?
53
A Vision of Paradise and Empire An evocation of the Qurʾanic imagery of paradise again suggests itself, as articulated by several modern writers.47 In this perspective, most elements seem to fall into place: the ethereal skies of light, the jewels, and the pearls, themselves also symbols of light in Late Antiquity.48 The tall buildings might evoke a further facet of the Qur’anic imagery of paradise: But those who fear their lord, for them await lofty chambers above which are built lofty chambers (ghuraf min fawqihā ghuraf mabniyyatun), underneath which rivers flow—God’s promise, God fails not the tryst (Q. 39:20). Blessed be He who, if He will, shall assign to thee better than that— gardens underneath which rivers flow, and He shall assign to thee palaces. (Q. 25:10)49 The odd proportions of the buildings, which tend to be markedly smaller than the trees and to be shown in logically incompatible perspectives and scales, could be interpreted as contributing to the same otherworldly effect.50 One of the earliest preserved historical writings on the imperial mosques of al-Walīd, the (lost) history of Medina by Ibn Zabāla (late eighth century), lends some support to this interpretation. It puts the following statement in the mouth of mosaicists who worked on al-Walīd’s rebuilding of the Prophet’s mosque at Medina, which had similar decorations to the ones at Damascus: “We made it according to the pictures of the trees of paradise and its palaces.”51 Interestingly, however, the anecdote also asserts that one of these Byzantines (al-rūm) tried to urinate on the Prophet’s tomb (but died of supernatural causes in the act). A craftsman also drew pigs on arches before being caught out and punished at the order of ʿUmar ibn ʿAbd al-ʿAzīz, who was in charge of the works. These alleged provocations could imply a tacit rebuke of al-Walīd’s employment of Christian workers, and they may arguably be a literary fabrication after the fact. However a polemical intent is not perceptible in the extract as a whole.52 Furthermore, whatever their region of origin (a debated issue), these mosaicists are likely to have been sent to work at Medina against their will, as a form of forced labor. The Aphrodito papyri, a body of administrative documents dating to the very same years in Egypt, shows that people often fled from such assignments.53 One could thus
45
54
A lain G eorge
equally envisage that some Christians on whom work in distant Arabia had been imposed sought to perform hostile actions. The credibility of Ibn Zabāla’s account is not undermined by these details, and if teams of Christian mosaicists were briefed to represent paradise without showing humans, animals, and angels, the extant mosaics would represent one way to do so with the pictorial tools at their disposal. But once again, this may not be the whole story, for the mosaics at Damascus show an array of different buildings in a way that does also convey a real landscape, from luxurious palaces and pavilions to smaller dwellings reminiscent of country houses.54 This possibility comes to life as one steps back to view the whole panel on the western arcade, and back again to imagine the entire courtyard adorned with similar decoration (fig. 2.8). A vision of earthly dominion would have been an almost instinctive response to this vast scenery, especially since floor maps showing cities and their buildings were to be seen in church mosaics of the period.55 This is the impression conveyed by virtually all medieval visitors who wrote about the mosque, starting with the well-known testimony by al-Muqaddasī (tenth century) that “there is hardly a tree or a notable town which has not been depicted on these walls.”56 Ibn Shākir al-Kutubī (fourteenth century) even cited an account according to which there was a representation of the Kaʿba above a miḥrāb of this mosque.57
Figure 2.8 Courtyard of the Great Mosque of Damascus, looking toward the west. Great Mosque of Damascus, 706–715 ad.
5
Paradise or Empire?
55
Furthermore, when al-Walīd’s court poets Jarīr, al-Farazdaq, and al- Nābigha al-Shaybānī wrote panegyrics to celebrate the construction of the mosque, none of them alluded to a paradisiacal dimension of its decoration.58 The most detailed description in these poems, that by al-Nābigha, extols the precious materials used in the mosque and the towering height of its dome, but rather than to link the lamps to divine light, for example, he proclaims that their radiance reaches “the Lebanon and the coast.” As he turns to the mosaic landscapes, the poet describes the way the courtyard encloses “the rivers and the countryside” (al-anhār wa’l-rīf): the second term probably implies that the image pertains to the earthly realm. Al-Nābigha’s only allusion to paradise and Judgment arises with respect not to these images, but to the inscription and its “detailed verses containing a promise and an admonition from our Lord.”59
The Lost Mosaic Inscription at Damascus This inscription, now lost, has been recorded in textual sources, and its contents provide a further hint at the perspective of Umayyad ruling elites on this building. It was made of gold mosaic against a dark blue ground inscription and originally occupied much of the qibla wall. Its text referred repeatedly to Judgment, as well as the fire of hell and the garden of paradise. For example, one could read toward the end: When heaven shall be stripped off When Hell shall be set blazing When Paradise shall be brought nigh Then shall a soul know what it has produced. (Q. 81:11–14)60 But God was invoked as the sustaining force of this world with just as much vigor. Both of these themes, eschatology and earthly abundance, appeared in conjunction in two long passages that lay at the core of the inscription and formed a large part of its text (Q. 79:1–46; 80:24–42). To cite but one of these: Let man consider his nourishment We poured out the rains abundantly, Then We split the earth in fissures and therein made the grains to grow
65
56
A lain G eorge
and vines, and reeds, and olives, and palms, and dense-tree’d gardens, and fruits, and pastures, an enjoyment for you and your flocks. And when the Blast shall sound upon the day when a man shall flee from his brother, his mother, his father, his consort, his sons, every man that day shall have business to suffice him. Some faces on that day shall shine laughing, joyous; some faces on that day shall be dusty o’erspread with darkness Those—they are the unbelievers, the libertines. (Q. 80:24–42) It is, in other words, as if passages bringing these two themes together had been expressly chosen. The evocation of grapes, olives, dates, enclosed gardens, and lofty trees directly echoes textual descriptions linking the mosaic decoration to earthly landscapes. Al-Walīd was indeed lauded by al-Nābigha as “the caliph of God through whom clouds of rain are sought” and by al-Farazdaq as “the shepherd of God on earth.”61 But this dimension also leads into the hereafter in these verses, just as in the mosaics. In a possible allusion to the Prophet or caliph as intercessor, the inscription also read: “His are all things in the heavens and on earth. Who is there to intercede in His presence except as He permits?” (Q. 2:255). In front of this wall, the caliph would stand and sit atop the minbar during the khuṭba, the public sermon—a few years earlier, the court poet Ibn Qays al- Ruqayyāt had evoked ʿAbd al-Malik as “the deputy of God on his minbar.”62 During ritual, he would lead prayer in the enclosed area of the maqṣūra, facing the miḥrāb, often protected by armed guards. Heated political arguments could arise.63 The mosque at Damascus was inherently related to the person of the caliph, its patron and the central presence in its social and ritual life, who was named at the end of the inscription and whose domed palace stood behind the qibla wall. The different messages emitted by the mosaics thus appear to coalesce around the early Islamic concept of the caliphate, in a way that is illuminated by the inscriptional program. The dimension of
57
Paradise or Empire?
57
paradise does not rule out that of empire, nor was it necessarily meant to: both appear to have been intimately related. Two traditions cited by al-Rabaʿī (eleventh century) assert that the Great Mosque of Damascus would remain a place of prayer for forty days after the end of the world, thereby linking this site to Judgment.64 In the days of al-Walīd, al-Nābigha also evoked, in relation to the mosque, the “navel of the world” (surrat al-arḍ), the meeting point between heaven and earth,65 an idea already encountered above in relation to the Dome of the Rock.
The Umayyad Codex of Sanaa Probably dating to the same years as the mosque of Damascus, and at any rate to the first decades of the eighth century, is a monumental Qurʾan discovered some four decades ago at the Great Mosque of Sanaa.66 This manuscript is famous for its three opening pages with architectural illuminations, a theme that seems to have originally existed in other Qurʾans. The first image (fig. 2.9) shows an abstract geometrical design, with a circle surrounded by a double square that forms a star. It is followed by the representation of a mosque, with its miḥrāb, minbar, corner towers, and mosque lamps (fig. 2.10). The depiction of a central nave, together with the type of decoration on the spandrels and columns, suggests an Umayyad imperial mosque. Corner towers were rare in mosques of this period and might thus reflect a specific reference to the mosque of Damascus or Medina. The trees behind the qibla, which are very large in relation to the building, are reminiscent of those in the Damascus mosaics. They bear fruit and are placed behind the qibla wall, in an evocation of paradise amalgamated with the direction of prayer. This places the mosque at the conceptual threshold between an actual building and a vision of paradise. The third image shows a more generic mosque, with its miḥrāb, lamps, and trees behind the qibla.67 Returning to the first illumination (fig. 2.9), the geometrical figure of a double square enclosing a circle is the exact same one that served as a basis for the ground plan of the Dome of the Rock. The trees are shown in groups of three: a large central tree flanked by two smaller ones. All of them stem out of the same trunk and thereby replicate a composition with three trees specific to the mosaics at the Dome of the Rock. The trees in the Sanaa Qurʾan are rooted in the circle, whereas they would have been more clearly visible had they been placed on the edge of the double square. This suggests the recurrence of a paradisiacal symbolism linked this time
85
58
A lain G eorge
Figure 2.9 First illumination of the Umayyad Codex of Sanaa. Page height ca. 41 cm, page width ca. 37.1 cm. Sanaa, Dār al- Makhṭūṭāt al-Yamaniyya, IN 20–33.1.
not to the qibla wall, but to the dome of heaven. On one level, these images thus represent a glorification of Umayyad buildings, which they depict schematically. On another level, their paradisiacal symbolism is reenacted with more clarity than in the extant mosaics. These dimensions are conflated into a single pictorial plane, like two images conjured by the same verse in the Qurʾan, but also possibly to show them as part of the same reality. In the original setting, this vision of paradise and empire was articulated on a grand scale, by integrating the mass of architectural structure with surface ornament, objects such as mosque lamps and pearls, and the divine Word in inscriptions and Qurʾan manuscripts. Textual sources suggest that Umayyad Qurʾans probably resembling the Sanaa manuscript were placed next to the miḥrāb of major Umayyad mosques, where they would be read on Thursday evening and Friday.68 Their pages of calligraphy
95
Paradise or Empire?
59
Figure 2.10 Second illumination of the Umayyad Codex of Sanaa. Page height ca. 41 cm, page width ca. 37.1 cm. Sanaa, Dār al- Makhṭūṭāt al-Yamaniyya, IN 20–33.1.
would have resonated with the inscriptions in gold against a dark green or blue ground, like the illumination echoed the mosaics, with their buildings and plants, as well as the actual lamps, columns, and objects. The trees and lush plants, both in manuscript and mosaic form, may well have represented, at one and the same time, paradisiacal evergreens and lands blooming through the intercession of the caliph. The caliph, through the attributes of his office, represented the connecting point between these two realms, as a celestial pole and a gateway to salvation.
Conclusion The Dome of the Rock and the Great Mosque of Damascus were built during crucial years that saw the Umayyad caliphate consolidate its foundations, expand territorially, and seek to firmly establish the standing of Islam as a religion. Among other tools, architecture and iconography
06
60
A lain G eorge
were called upon to impart the sense of a new emerging order on the public’s imagination. In both monuments, the themes of paradise and empire are so closely intertwined as to make attempts to disentangle them, or assert one at the expense of the other, seem to be in vain. The same images, the same perceived architectural structures, could emit such different messages partly because the underlying concepts were themselves in proximity: partly, also, because such openness of meaning was a deeply ingrained aesthetic mode notably embodied by the Qurʾan. The mosaics thus seem to have conveyed cognate messages about heaven and earth that reflected the early Islamic conception of the caliphate and opened onto a spiritually charged invisible realm. It is perhaps for this reason that both monuments were prone to associations with the navel of the earth, the meeting point of different temporal dimensions and planes of reality. It may be necessary, when studying the art and culture of this period, to leave behind our modern predilection for neatly defined categories: for what appears to us as paradox might have been, in the eyes of its makers, a consciously crafted polysemy based on an overarching sense of the divine.
Notes 1. Grabar 1996, 71. An extensive set of color images can be consulted in this publication; see also Nuseibeh and Grabar 1996. 2. On the original appearance of the building, see Allen 1999; Van Berchem 1969, 237. 3. For overviews of the main interpretations, see Grabar 2006, 109–119; Shani 1999, 158–165. See also, to cite but some of the most seminal studies, Grabar 1959; Grabar 1996, chap. 2; Rosen-Ayalon 1989; Rabbat 1989; Shani 1999; Necipoğlu 2008. 4. The most detailed discussion of this theme to date is Rosen-Ayalon 1989, 46–69. 5. These have been most thoroughly articulated in Grabar 1959. Grabar’s own understanding of the building later evolved to include additional layers of meaning. 6. E.g., Grabar 1996, figs. 29, 30, 32 (drum), 45 (right), 48 (left) (inner octagon). Maria Vittoria Fontana (2012, 97) has counted twenty-six of these occurrences in the drum. 7. See Erdmann 1951; Grabar 1967, Cat. 5, 16, 69, 82u, 82w, 82x; Harper 1978, Cat. 46; Kröger 1982, Taf. 93.5; Demange 2006, Cat. 33, 201–208 (several of the same objects appear in these different publications). For further examples of coins from the Islamic period, see Jamil 1999, fig. 14b; Treadwell 1999, figs. 2–3 and pp. 261– 269; Treadwell 2005, figs. 2–3, 7. Cf. also, for a broader discussion of this symbol and its Islamic adaptations, Fontana 2012.
61
Paradise or Empire?
61
8. Rosen-Ayalon 1989, 54–55. 9. Rose 2011, 28. Cf. also Gherardo Gnoli, art. “Farr(ah),” in Encyclopaedia Iranica, http://www.iranicaonline.org/articles/farrah (accessed November 28, 2012). 10. For a discussion of this association with khvarnah, see Soudavar 2003, 19–25; for examples of objects, Grabar 1967, Cat. 61, 70; Harper 1978, Cat. 39, 43; Kröger 1982, Taf. 14.3, 58.2–3, 76.3, 81.3, 83.5, 94.1; Demange 2006, Cat. 10; Evans and Ratliff 2012, Cat. 17. Wings of the same type, but composed as an overlapping pair, often adorn a Sasanian griffin (senmurv) or a winged horse associated with the heavens. 11. For some Christian examples, see Deliyannis 2010, figs. 45, 73, 83, 86, 87, 89, 93 and pl. IIIa, V, VIa, VIb; Cormack and Vasilaki 2008, Cat. 21–23; Evans and Ratliff 2012, fig. 19 (p. 53) and Cat. 26–27; Miziołek 1990, fig. 2a. 12. Grabar 1967, Cat. 15; Demange 2006, 40, fig. 2. 13. For Hagia Sophia at Constantinople, see Canepa 2010, 217–221. 14. See Deliyannis 2010, 172–174, 238–243. As noted by Deliyannis, the male portrait at Sant’ Apollinare Nuovo could represent either Justinian or the Ostrogothic king Theodoric. 15. Ripoll López 1993, 53–59; Ager 2010, 80; Conti 1983, 39–42 (not consulted). The Guarrazar treasury was first noted in relation to the Dome of the Rock by Grabar (1959, 49). See also, for links between Byzantine jewelry and its representations at San Vitale, Brown 1979. 16. Cf. also, for example, Grabar 1996, figs. 44 (right), 46 (right and left), 49 (far right), and so on. 17. Lethaby and Swainson 1894, 72–73; Grabar 1959, 48–50; Wilkinson 2002, 139; Necipoğlu 2008, 55. The sacral and paradisiacal dimensions of these representations are emphasized in Rosen-Ayalon 1989, 49, 52; Mekeel-Matteson 1999, 161–162. 18. Grabar 1959, 47–52; Grabar 2006, 113. 19. Rabbat 1993, 71; Wāsiṭī [2009], 339 (no. 120). Cf. also Elad 1995, 52; Kaplony 2002, 349 and n. 2. 20. Elad 1995, 6–22; Mourad 2008. 21. Grabar 1959, 50–51; Flood 2009, 28–29; Hoyland 2011, 136–137. 22. Cf. Rabbat 1993, 71–73. 23. Roberts 1989; cf. Bolman 2006, 18, 20, for early Christian Egypt. 24. See also L. Kinberg, art. “Paradise” (EQ); Schimmel 1976, 13–21. The translation used for all Qurʾanic quotations is by A. J. Arberry, unless otherwise stated. 25. See, respectively, Hawkins, Mundell, and Mango 1973, fig. 13; Evans and Ratliff 2012, Cat. 120A. 26. Barry 2007 (where the similar marble panels at the Great Mosque of Damascus are mentioned at p. 639); Milwright 2007, 214; Flood 2001, 67. 27. For a full development of this reading, see Rosen-Ayalon 1989, 46–69. 28. D. Waines, art. “Agriculture and Vegetation” and “Weather” (EQ).
26
62
A lain G eorge
29. Crone and Hinds 1986, 8. Cf. Marsham (this volume); Jamil 1999, 39–40; Necipoğlu 2008, 39. 30. H. Berg, art. “Polysemy in the Qurʾan” (EQ). This echoes the conclusions of Gülru Necipoğlu about the Dome of the Rock, which point to “the parallelism between the eternal heavenly kingdom and its earthly counterpart entrusted to the caliph” and the way the monument “blurred the temporal boundaries between the past, the present and the eschatological future” (Necipoğlu 2008, 56). 31. Also cited and discussed in Kermani 2015, 103. 32. See Kermani 2015, esp. 100–106. Cf. also the reflections put forward about this theme, along with many others, in relation to early Islamic architecture in Hamdouni Alami 2011, esp. chaps. 1, 3. 33. The name was coined in Déroche 2014, 76. Cf. Déroche (this volume) for further information and references. The label “Marcel 13,” proposed earlier by Déroche (2004, 242), reflects the shelfmark of one of its fragments held at the National Library of Russia (Saint Petersburg). The other extant fragments are Marcel 11 and Marcel 15 in the same collection and Arabe 330c at the Bibliothèque Nationale de France (Paris). 34. See Déroche 2004, esp. 242, 253–260; George 2010, 75–78. 35. Elad 1995, 46–47, 51, 56, 64; Kaplony 2002, 52–53, 245, 250–253, 325, 328. 36. Déroche has remarked (2004, 247–248) that the water at the top appears to be a later addition. 37. M. Radscheit, art. “Springs and Fountains” (EQ). 38. Mourad 2008, 93–94. Cf. Kaplony 2002, 242; Livne-Kafri 2006, 389–390. The tradition about the lamp of God also states that the angel Gabriel had suspended in the Jewish Temple the Greatest Lamp (al-qindīl al-akbar), the seven-armed Menorah, brought from heaven, which burned day and night “without smoke or impurity” (Kaplony 2002, 234). 39. Thessaloniki: Bakirtzis, Kourkoutidou-Nikolaidou, and Mavropoulou-Tsioumi 2012, 82–83, fig. 45; Paris, Bibliothèque Nationale de France, Syriaque 33, f. 8v: Leroy 1964, 1:200 (for the date); 2: pl. 41–1 (image). The latter has been mentioned in relation to the present illumination by Déroche (2014, 90). The same type of fountain occurs in another Umayyad Qurʾan: Istanbul, Türk ve Islam Eserleri Müzesi, ŞE 321, f. 57v; Déroche (this volume), fig. 2; Déroche 2014, fig. 24 (color image). 40. For some images, see Grabar 1996, figs. 25–27. 41. al-Wāsiṭī [2009], 287 (no. 87), 318 (no. 106), 321 (no. 108), 322 (no. 109), 325 (no. 112), 327 (no. 113), 329 (no. 114), and so on. All of the above traditions are based on al-Ramlī (see Mourad 2008 for further background). 42. Déroche 2004, 243–259. Cf. Déroche (this volume), fig. 1, for an image. 43. Elad 1995, 107. 44. Kaplony 2002, 234. 45. al-Wāsiṭī [2009], 321 (no. 108); Mourad 2008, 93. 46. Brown 1999, 31.
6 3
Paradise or Empire?
63
47. Finster 1970, 118–121; Brisch 1988; van Lohuizen-Mulder 1995, 207, 209; Flood 2001, s.v. “Paradise.” 48. Flood 2001, chap. 2. 49. For further references, see Brisch 1988, 16–18. 50. Brisch 1988, 14–16. 51. al-Samhūdī 2001, 2:269–270. Also cited in Sauvaget 1947, 81; Finster 1970, 120; Brisch 1988, 18; van Lohuizen-Mulder 1995, 193, 209. 52. As already noted by Gibb (1958, 228) with regard to related traditions. 53. To give but one example, see Bell 1911, 269–270 (no. 1333). 54. Ettinghausen 1977, 22–28; Rabbat 2003, 90–93. 55. Bowersock 2006, chaps. 1 and 3 (where Damascus is mentioned at pp. 80–81). 56. Van Berchem 1969, 233. 57. Rabbat 2003, 91; Van Berchem 1969, 238. Cf. also, for further accounts along the same lines, Rabbat 2003, 91–92; Van Berchem 1969, 235, 239. 58. These will be studied in full and translated in George forthcoming. 59. al-Nābigha 1996, 103; Rabbat 2003, 90; George forthcoming. 60. The full text of the inscriptions is given in Flood 2001, 247–254 (citing Yusuf Ali’s translation, as opposed to A. J. Arberry’s in the present essay). 61. Crone and Hinds 1986, 9. 62. Crone and Hinds 1986, 8. 63. Sauvaget 1947, 134–153. For an English translation, see Sauvaget 2002, 122– 142. Estelle Whelan (1986) has sought to link the concave miḥrāb, using iconographic and textual parallels, to the commemoration of the Prophet. But this hollow niche was also directly related, in mosque ritual, to the caliph, whether in person during prayer or at other times, when it may have served as a reminder of his presence—a possibility suggested by the work of Sauvaget (1947, 145–149), which emphasized its sociopolitical aspects. A reference to either the Prophet or the caliph—or both, since they are not mutually exclusive—may have been intended. 64. al-Rabaʿī 1950, 38 (Nos. 66, 67); cf. Ibn Kathīr 1966–1967, 9:154. 65. al-Nābigha 1996, 103; Rabbat 2003, 90. 66. The present discussion is based on George 2010, 79–86, to which readers are referred for further detail and references. See also, about this manuscript, Von Bothmer 1987; Déroche 2014, 111–116. 67. Reproduced in Von Bothmer 1987, Abb. 1; George 2010, fig. 53. 68. George 2010, 86.
Bibliography Ager, B. 2010. “Byzantine Influences on Visigothic Jewellery.” In Intelligible Beauty: Recent Research on Byzantine Jewellery, edited by Chris Entwistle and Noel Adams, 72–82. London: British Museum.
46
64
A lain G eorge
Allen, H. R. 1999. “Some Observations on the Original Appearance of the Dome of the Rock.” In Bayt al-Maqdis. Jerusalem and Early Islam, edited by Jeremy Johns, 197–215. Oxford: Oxford University Press. Bakirtzis, C., E. Kourkoutidou- Nikolaidou, and C. Mavropoulou- Tsioumi. 2012. Mosaics of Thessaloniki: 4th to 14th Century. Athens: Kapon. Barry, F. 2007. “Walking on Water: Cosmic Floors in Antiquity and the Middle Ages.” Art Bulletin 89 (4): 627–656. Bell, H. I. 1911. “Translations of the Greek Aphrodito Papyri in the British Museum.” Der Islam 2: 269–283. Bolman, E. 2006. “Late Antique Aesthetics, Chromophobia and the Red Monastery, Sohag, Egypt.” Eastern Christian Art 3: 1–24. Bowersock, G. W. 2006. Mosaics as History: The Near East from Late Antiquity to Early Islam. Cambridge, MA: The Belknap Press of Harvard University Press. Brisch, K. 1988. “Observations on the Iconography of the Mosaics in the Great Mosque at Damascus.” In Content and Context of Visual Arts in the Islamic World, edited by Priscilla Soucek, 13–20. University Park: Pennsylvania State University Press. Brown, K. 1979. “The Mosaics of San Vitale: Evidence for the Attribution of Some Early Byzantine Jewelry to Court Workshops.” Gesta 18 (1): 57–62. Brown, P. 1999. “Images as a Substitute for Writing.” In East and West: Modes of Communication, edited by E. Chrysos and I. Wood, 15–34. Leiden: Brill. Canepa, M. 2010. The Two Eyes of the Earth: Art and Ritual of Kingship between Rome and Sasanian Iran. Berkeley: University of California Press. Conti, R. 1983. Il tesoro: Guida alla conoscenza del tesoro del Duomo di Monza. 2nd ed. Monza: Museo del Duomo di Monza. Cormack, R., and M. Vasilaki, eds. 2008. Byzantium, 330–1453. London: Royal Academy of Arts. Creswell, K. A. C. 1969. Early Muslim Architecture. Volume I: Umayyads, a.d. 622–750. Oxford: Oxford University Press. Crone, P., and M. Hinds. 1986. God’s Caliph: Religious Authority in the First Centuries of Islam. Cambridge: Cambridge University Press. Deliyannis, D. M. 2010. Ravenna in Late Antiquity. Cambridge: Cambridge University Press. Demange, F., ed. 2006. Les Perses sassanides: Fastes d’un empire oublié. Paris/Suilly-la- Tour: Paris Musées/Findakly. Déroche, F. 2004. “Colonnes, vases et rinceaux sur quelques enluminures d’époque Omeyyade.” Comptes Rendus des Séances de l’Académie des Inscriptions et Belles Lettres: 227–264. ———. 2014. Qurʾans of the Umayyads: A First Overview. Leiden: Brill. Elad, A. 1995. Medieval Jerusalem and Islamic Worship: Holy Places, Ceremonies, Pilgrimage. Leiden: Brill.
6 5
Paradise or Empire?
65
Erdmann, K. 1951. “Die Entwicklung der Sāsānidischen Krone.” Ars Islamica 15/ 16: 87–123. Ettinghausen, R. 1977. Arab Painting. 2nd ed. Lausanne: Skira. Evans, H. C., and B. Ratliff, eds. 2012. Byzantium and Islam: Age of Transition. New York: Metropolitan Museum of Art. Finster, B. 1970–1971. “Die Mosaiken der Umayyadenmoschee von Damaskus.” Kunst des Orients 7: 83–141. Flood, F. B. 2001. The Great Mosque of Damascus: Studies in the Makings of an Umayyad Visual Culture. Leiden: Brill. ———. 2009. Objects of Translation: Material Culture and Medieval “Hindu-Muslim” Encounter. Princeton, NJ: Princeton University Press. Fontana, M. V. 2012. “The Meaning and the Iconographic Development of the Winged Sasanian Crown in Early Islamic Art.” Beiträge zur Islamischen Kunst und Archäologie 3: 95–112. George, A. 2010. The Rise of Islamic Calligraphy. London: Saqi. ———. Forthcoming. The Great Mosque of Damascus: Art, Faith and Empire in Early Islam [Working title]. Gibb, H. A. R. 1958. “Arab-Byzantine Relations under the Umayyad Caliphate.” Dumbarton Oaks Papers 12: 219–33. Grabar, O. 1959. “The Umayyad Dome of the Rock in Jerusalem.” Ars Orientalis 3: 33–62. ———. 1967. Sasanian Silver: Late Antique and Early Mediaeval Arts of Luxury from Iran; August-September 1967, the University of Michigan Museum of Art. Ann Arbor: University of Michigan Museum of Art. ———. 1987. The Formation of Islamic Art. 2nd ed. New Haven, CT: Yale University Press. ———. 1996. The Shape of the Holy: Early Islamic Jerusalem. Princeton, NJ: Princeton University Press. ———. 2006. The Dome of the Rock. Cambridge, MA: The Belknap Press of Harvard University Press. Hamdouni Alami, M. 2011. Art and Architecture in the Islamic Tradition: Aesthetics, Politics and Desire in Early Islam. London: I. B. Tauris. Harper, P. 1978. The Royal Hunter: Art of the Sasanian Empire. New York: Asia Society. Hawkins, E. J. W., M. Mundell, and C. Mango. 1973. “The Mosaics of the Monastery of Mār Samuel, Mār Simeon, and Mār Gabriel Near Kartmin, with a Note on the Greek Inscription.” Dumbarton Oaks Papers 27: 279–296. Hoyland, R. 2011. Theophilus of Edessa’s Chronicle and the Circulation of Historical Knowledge in Late Antiquity and Early Islam. Liverpool: Liverpool University Press. Ibn Kathīr. 1966–1967. al-Bidāya wa’l-nihāya. 14 vols. Beirut: Maktabat al-Maʿārif. Jamil, Nadia. 1999. “Caliph and Quṭb: Poetry as a Source for Interpreting the Transformation of the Byzantine Cross on Steps on Umayyad Coinage.” In
66
66
A lain G eorge
Bayt al- Maqdis: Jerusalem and Early Islam, edited by Jeremy Johns, 11– 58. Oxford: Oxford University Press. Kaplony, A. 2002. The Ḥaram of Jerusalem, 324–1099: Temple, Friday Mosque, Area of Spiritual Power. Stuttgart: Franz Steiner. Kermani, N. 2015. God Is Beautiful: The Aesthetic Experience of the Quran. Translated by T. Crawford. Cambridge: Polity Press. Kröger, J. 1982. Sasanidischer Stuckdekor: Ein Beitrag zum Reliefdekor aus Stuck in sasanidischer und frühislamischer Zeit. Mainz am Rhein: P. von Zabern. Leroy, J. Les manuscrits syriaques à peintures conservés dans les bibliothèques d’Europe et d’Orient. 2 vols. Paris: Geuthner, 1964. Lethaby, W. R., and H. Swainson. 1894. The Church of Sancta Sophia, Constantinople: A Study of Byzantine Building. London: Macmillan. Livne-Kafri, O. 2006. “Jerusalem in Early Islam: The Eschatological Aspect.” Arabica 53 (3): 382–403. Mekeel-Matteson, C. 1999. “The Meaning of the Dome of the Rock.” Islamic Quarterly 43: 149–185. Milwright, M. 2007. “‘Waves of the Sea’: Responses to Marble in Written Sources (Ninth-Fifteenth Century).” In The Iconography of Islamic Art: Studies in Honour of Robert Hillenbrand, edited by Bernard O’ Kane, 211–221. Edinburgh: Edinburgh University Press. Miziołek, J. 1990. “Transfiguratio Domini in the Apse at Mount Sinai and the Symbolism of Light.” Journal of the Warburg and Courtauld Institutes 53: 42–60. Mourad, S. A. 2008. “The Symbolism of Jerusalem in Early Islam.” In Jerusalem: Idea and Reality, edited by Tamar Meyer and Suleiman Ali Mourad, 86–102. Milton Park: Routledge. al-Nābigha (al-Shaybānī). 1996. Dīwān. Edited by ʿUmar Fārūq al-Ṭabbāʿ. Beirut: Dār al-Arqam. Necipoğlu, G. 2008. “The Dome of the Rock as Palimpsest: ʿAbd al-Malik’s Grand Narrative and Sultan Süleyman’s Glosses.” Muqarnas 25: 17–105. Nuseibeh, S., and O. Grabar. 1996. The Dome of the Rock. London: Thames and Hudson. al-Rabaʿī. 1950. Faḍāʾil al-shām wa-dimashq. Edited by Ṣalāḥ al-Dīn al-Munajjid. Damascus: Al-majmaʿ al-ʿilmī al-ʿarabī bi-dimashq. Rabbat, N. 1989. “The Meaning of the Umayyad Dome of the Rock.” Muqarnas 6: 12–21. ———. 1993. “The Dome of the Rock Revisited: Some Remarks on al-Wasiti’s Accounts.” Muqarnas 10: 66–75. ———. 2003. “The Dialogic Dimension of Umayyad Art.” RES: Anthropology and Aesthetics 43: 78–94. Ripoll López, G. 1993. “The Formation of Visigothic Spain.” In The Art of Medieval Spain, a.d. 500–1200, 41–69. New York: Metropolitan Museum of Art.
6 7
Paradise or Empire?
67
Roberts, M. 1989. The Jeweled Style: Poetry and Poetics in Late Antiquity. Ithaca, NY: Cornell University Press. Rose, J. 2011. Zoroastrianism: An Introduction. London: I. B. Tauris. Rosen-Ayalon, M. 1989. The Early Islamic Monuments of al-Haram al-Sharif: An Iconographic Study. Jerusalem: Institute of Archaeology, Hebrew University of Jerusalem. al-Samhūdī. 2001. Wafāʾ al- wafā bi- akhbār dār al- muṣṭafā. Edited by Qāsim al- Sāmarrāʾī. 5 vols. London: Al-Furqān Islamic Heritage Foundation. Sauvaget, J. 1947. La mosquée omeyyade de Médine: Etude sur les origines architecturales de la mosquée et la basilique. Paris: Vanoest. ———. 2002. “The Mosque and the Palace.” In Early Islamic Art and Architecture, edited by Jonathan Bloom, 109–148. Aldershot: Ashgate. Schimmel, A. 1976. “The Celestial Garden in Islam.” In The Islamic Garden, edited by Richard Ettinghausen, 11–39. Washington, DC: Dumbarton Oaks. Shani, R. 1999. “The Iconography of the Dome of the Rock.” Jerusalem Studies in Arabic and Islam 23: 158–207. Soudavar, A. 2003. The Aura of Kings: Legitimacy and Divine Sanction in Iranian Kingship. Costa Mesa, CA: Mazda. Treadwell, L. 1999. “The ‘Orans’ Drachms of Bishr Ibn Marwān and the Figural Coinage of the Early Marwanid Period.” In Bayt al-Maqdis: Jerusalem and Early Islam, edited by Jeremy Johns, 223–269. Oxford: Oxford University Press. ———. 2005. “‘Mihrab and ʿAnaza” or ‘Sacrum and Spear’? A Reconsideration of an Early Marwanid Silver Drachm.” Muqarnas 22: 1–28. Van Berchem, M. 1969. “The Mosaics of the Dome of the Rock in Jerusalem and of the Great Mosque in Damascus.” In Early Muslim Architecture, Volume I: Umayyads, a.d. 622–750, by K. A. C Creswell, 213– 372. 2nd edition. Oxford: Oxford University Press. Van Lohuizen- Mulder, M. 1995. “The Mosaics in the Great Mosque at Damascus: A Vision of Beauty.” Babesch 70: 193–213. Von Bothmer, H.-C. 1987. “Architekturbilder im Koran: Eine Prachthandschrift der Umayyadenzeit aus dem Yemen.” Pantheon 45: 4–20. al-Wāsiṭī. [2009]. Faḍāʾil al-bayt al-muqaddas. Edited by Abū al-Mundhir al-Juwaynī. Cairo: np. Whelan, E. 1986. “The Origins of the Miḥrāb Mujawwaf: A Reinterpretation.” International Journal of Middle East Studies 18 (2): 205–223. Wilkinson, J. 2002. Jerusalem Pilgrims before the Crusades. Warminster: Aris and Phillips.
86
96
3
A Qurʾanic Script from Umayyad Times Around the Codex of Fustat François Déroche
of the Qurʾanic fragments of the Great Mosque in Sanaa, the identification of Umayyad copies of the Qurʾan has made some headway.1 In this process, fragments like Marcel 13, a group of folios kept in the National Library of Russia in Saint Petersburg have played an important role in our attempts at reconstructing the history of the Arabic script used for the transcription of the Qurʾan by the end of the first or beginning of the second century of Islam.2 Although nine folios from this manuscript were kept in the Bibliothèque Nationale by the middle of the nineteenth century (Arabe 330c), they baffled the cataloguers of the Jean- Louis Asselin de Cherville collection. Michele Amari stated that they were written in a “script of the Hidjâz, verging towards the Kufic script”3 and suggested a date in the third/ninth century.4 The present writer preferred to leave them outside of the typology he proposed for the early Qurʾanic scripts.5 Most of what remains of this manuscript, sixty-four folios, is kept in the National Library of Russia.6 The seventy-three parchment folios in vertical format (37 × 31 cm) are written with twenty-five lines to the page. The average height of the line is about 11.5 mm. The surviving quires are homogeneous, with five bifolios. One of the most salient features of this copy are, of course, the sura headbands, which are crucial when assessing the date of this copy. I have suggested to call it the Codex of Fustat (Umayyad) (fig. 3.1).7 SINCE THE DISCOVERY
07
70
F ran ç ois D é roche
Figure 3.1 Folio from the Umayyad Codex of Fustat. Page dimensions: 37 × 31 cm. St. Petersburg, National Library of Russia, Marcel 13, f. 1r.
The Damascus and Fustat Codices A fragment kept in the Turkish and Islamic Arts Museum (TIEM) in Istanbul, ŞE 321, has also been important in this identification process.8 It was kept for a long time in the Umayyad mosque in Damascus, until its eventual transfer to Istanbul before the fall of the Ottoman Empire.9 Its seventy-eight parchment folios, which were once part of the original manuscript, were brought together under the provisional shelf number TIEM ŞE 321, but it shall be referred to as the Codex of Damascus (Umayyad) (fig. 3.2).10 This copy was written with eighteen to twenty-one lines to the page on folios in vertical format, their height being roughly 24 cm (228 to 242 mm) and their width slightly over 19.5 mm (186 to 197 mm). The folios were ruled with a dry point: a complete frame with two vertical lines indicating the margins and eighteen to twenty-one horizontal lines. The quires are unfortunately not complete but might be tentatively reconstructed: from the material at hand, we can conclude with reasonable certainty that the quires were usually made of ten bifolios, a comparatively
71
A Qurʾanic Script from Umayyad Times
71
Figure 3.2 Folio from the Umayyad Codex of Damascus. Page dimensions: ca. 24 × 19.5 cm. Istanbul, Türk ve Islam Eserleri Müzesi, ŞE 321, f. 57v.
rare structure of quire in the Islamic tradition.11 The manuscript contains twenty sura headbands, which are in this case again an important clue as to the date of copy. The sura headings of the Codex of Damascus do not contain information about the text (title of the sura, number of verses, for example); when some information appears in the case of sura 33, it has been written afterward in red outside the frame of the illumination. The same can be said of the Codex of Fustat: from the structure of the illuminations, it is clear that it was not meant to receive a title. At a later date, someone added below the old formula fātiḥat al-sūra … (“Beginning of the sura …”), followed by the title and the number of verses. In both manuscripts, the sura headings fall into two broad groups based on a formal feature: the first one covers those illuminations that have a roughly rectangular shape, their long sides being rectilinear; the headings of the other group are not inscribed within a frame and their shapes develop freely within the limits of an imaginary rectangle roughly equivalent to a line of text. As a rule in both manuscripts, the illumination between the suras is contained within a theoretical rectangle, the long sides of which are the horizontal of the upper line of script and that of the line left empty for the heading, whereas the short vertical sides are
27
72
F ran ç ois D é roche
drawn by the vertical limits of the justification. As a result, no component of the illumination significantly encroaches on the margins. Only in a few instances did the illuminators of the Fustat codex have a small tip of the ornament in the margin.12
Illumination In both the Fustat and the Damascus codices, the quality or even the presence of illumination is quite irregular. In the Codex of Fustat, there are twenty-one beginnings of a sura.13 Of those, only twelve are underscored by illumination. Their presence seems to be somehow connected with the quire structure of the manuscript, as if the illuminator had been working on loose quires, which would explain why three quires are almost completely decorated, two do not have any ornament at all, and one is partly illuminated. Moreover, two illuminations are actually preparatory drawings that never received their paint.14 In the Damascus codex, the situation is different: there are huge variations in the quality of the ornaments, ranging from crude to sophisticated, but the sequence cannot be related to any material feature of the manuscript, like the quires, nor to what could be a division of the work between, say, two illuminators. The palette is based on four components: gold, dark red, light blue, and green, varying in the Damascus codex from a darker hue to a lighter one. The unpainted parchment has been used instead of white—notably between colored stripes. The repertoire, which is being analyzed elsewhere, can be directly related to Umayyad works.15
Orthography The orthography of both manuscripts can be defined as scriptio defectiva, mainly characterized by the omission of notation for the sound /ā/.16 It is in keeping with what we find in copies in Hijazi script. On the other hand, changes appear and the scriptio plena is gaining ground. The case of qāla in the Damascus codex is especially tale-telling of this evolution: on f. 41a for instance, the copyist wrote on line 5 qālū (Q. 34:22), then q(ā)la (same verse) and on the next line q(ā)lū (same verse); on the verso, on line 5, we find qāla (Q. 34:32), then q(ā)la on 1. 7 (Q. 34:33), and again the scriptio plena on lines 14 and 15 (Q. 34:34 and 34:35). A survey of the Fustat codex shows that qāla is written most of the time (88%) in scriptio plena, but that conversely, qālu
73
A Qurʾanic Script from Umayyad Times
73
is written with medial alif in eighty percent of cases. Keeping to the defective orthography, the plural ʿibād written without alif belongs here (see, for instance, Q. 34:13, 38:83, 40:85, and so on), while the word ʿadhāb is illustrative of an almost complete transition to the scriptio plena. The orthography of both the Damascus and Fustat codices seems less defective than that of the Codex Parisino-petropolitanus, and the scriptio plena seems to have made substantial advances in the famous Qurʾan from Sanaa (Inv. 20–33.1) dated to the reign of al-Walīd, in comparison with the previous copies—according to what I could glean from the pictures published so far. The two copies may thus be considered as an intermediary stage in the effort toward a canonical text of the Qurʾan.
Script Turning to the script of both copies, it looks as a whole comparatively slightly heavier than Hijazi as written by Hand A of the Codex Parisino- petropolitanus;17 but Hand D of the latter manuscript or other Hijazi fragments like London, BL Or. 2165 are rather close to the Damascus codex in terms of relation between the height of the shafts, the width of the stroke, and its neatness. The letters are spread over the page by the extensive use of the elongation of the horizontal connections or by spacing regularly the letters that are not connected—be they part of the same word or not: we find here again the transposition of the principles of late antique scripta continua to the Arabic script.18
Letter Forms In both copies, the shaft of the alif is usually vertical, although in some instances one can find alifs that are almost Hijazi in appearance. The lower return of the letter is short and the upper end of the shaft is bevelled—as in the case of the shaft of other letters. The qāf—as an isolated or final letter—has a noticeably important down-falling tail that interferes somewhat with the line of text below. One of the most interesting letters is certainly the final or isolated kāf.19 Its very developed lower horizontal stroke is a noticeable feature of our manuscript: when the opportunity arises, it may be substantially lengthened. The upper horizontal stroke is strictly parallel to the lower one; at its extremity on the left, at a point located before the middle of the lower stroke, a vertical shaft surges from it.
47
74
F ran ç ois D é roche
Another typical letter form is that of the mīm at the end of a word. It looks like a semibreve set on the line; on the left side, there is usually no trace of a tail—as one would expect—but the shape is carefully rounded. The “eye” of the mīm is clearly visible. The letter cannot be confused with the tā’ marbūṭa, which keeps the short rectilinear stroke of the hā’ on its right side. In this case, it is clearly a peculiarity of this manuscript—and of the group of fragments that is related to it. It should be added that mīms with a horizontal tail do also appear in the Codex of Damascus, but this does not seem very frequent. The nūn retained some of the features of the Hijazi nūn as it appears, for instance, on the Codex Parisino-petropolitanus20 or BNF Arabe 6140a.21 But the use of another qalam (or at least of another way of cutting the tip of the reed) gives to the scribe the opportunity of stressing the contrast between the two halves of the crescent and the point where they are connected. The sickle-shaped letter opens more or less widely; the upper part is in some cases almost vertical, but it is more commonly oblique.
Line Layout In both copies, line-end fillers are commonly used, a feature that is not found in earlier copies. In the Codex of Fustat, we even find cases of the use of such strokes in the middle of a line, when the copyist extended too much toward the bottom the lower stroke of letters like ʿayn or qāf so that they disrupted the line and forced the copyist to distribute the text in order to avoid leaving any blank space.22 The margins are more generous in the Fustat codex than in the Damascus codex and also distinguish both copies from the earlier tradition illustrated by the Codex Parisino-petropolitanus, where margins are almost nonexistent.
Different Hands, New Standard The two copies are clearly in two different hands (actually a closer study of the script might shed light on a collective work in at least one manuscript), but the palaeographical examination suggests that they were working according to the same standards, in contradistinction to the earlier situation exemplified by the Codex Parisino-petropolitanus and related copies belonging to the Hijazi stage of handwritten transmission of the Qurʾanic text.23 The Codex Parisino-petropolitanus has been transcribed by five copyists whose hands exhibit strong idiosyncrasies, which suggest a
75
A Qurʾanic Script from Umayyad Times
75
great freedom in the process of copying the holy text. I suggested dating that manuscript to the third quarter of the first/seventh century.24 As noted by Amari, this script (which I call “O.I”) retains something of its Hijazi origins,25 to such an extent that it might be useful to distinguish between a version still influenced by the style (O.Ia) from the more elaborate script of the Fustat codex (O.Ib). As indicated previously, Hand D of the Codex Parisino-petropolitanus can be seen as a forerunner.26 It is still a little hesitant and some of the letter shapes, the hāʾ, for instance, are very different from those of the new Umayyad script. What is more interesting is the relationship between the script of the Qurʾanic manuscripts and some of the Umayyad inscriptions, notably ʿAbd al-Malik’s milestones: “[their] writing as a whole echoes manuscript calligraphy.”27 The isolated or final qāf is almost identical with that found on stone and the same can be said of the final kāf in the caliph’s name on three of the milestones;28 to these comparative examples can be added the shape of the kāf found on two of the inscriptions of the Dome of the Rock, that in mosaic and another one on a copper plate at the northern entrance of the building.29 Alain George has suggested identifying in all these inscriptions the use of a grid defining the size of the various letters and of their constitutive parts.30
A Change in Production More important, the change is not only a matter of aesthetics. It has to do with a sharp change in the production of copies that are what I would call (with caution) official copies, as the size of the Codex Parisino-petropolitanus and the Fustat codex, both large quarto manuscripts, imply a possible public use. London, BL Or. 2165 on the one hand, as well as both the Damascus and Fustat codices on the other hand, witness a completely new feature in the—young—history of the Arabic script: the deliberate iteration of a style of writing. The two hands cooperating in the transcription of Or. 2165 or the two (or more) copyists of the other two copies of the Qurʾan were able to transcribe the text in such a way that the difference between the hands was not immediately detectable. In other words, they belonged to a world where scribes had a professional approach to their trade, learning a specific style and using it for specific tasks. We may go a step further: we have before our eyes the beginnings of a new concept, that of Qurʾanic script, something that is perhaps not completely new in the area but is assuredly new in the Arabic manuscript tradition. A style now becomes specific to
67
76
F ran ç ois D é roche
a certain use. These elements point into the same direction: at the end of the first/seventh century, under the reign of the caliph ‘Abd al-Malik, a fateful change occurred in the chancery of the empire.31 Arabic, both language and script, became the official medium of the administration. From some sources, we know that the caliphs had specific scripts devised for their own correspondence.32 The relationship between the script of ʿAbd al-Malik’s milestones and that of contemporaneous muṣḥafs suggests that the latter were involved in this transformation, which concurred with reforms involving the Qurʾanic text. A group of fragments in the O.I script provides us with an idea of the nature of the change.33 The evidence is found in the collection of old Qurʾanic codices, which was formerly kept, like the Damascus codex, in the Umayyad mosque in Damascus before it was taken to Istanbul (some fragments have also been scattered among other collections), but also in the Marcel collection in Saint Petersburg. In the largest copies, the twenty-five lines of text to the page tally with the Codex of Fustat. However, Marcel 12, in spite of its similarities with the latter, has only twenty-two lines to the page. A few copies illustrate what might be another tendency, with only sixteen lines to the page, which could point to multivolume sets. In spite of this diversity, the module for those scripts remains fairly constant. The majority comprised between 10 and 12.7 mm. They retain the main characteristics that have been described (scriptio continua, margins, consistent division into verses), but a fair proportion exhibits a new feature: short vowels marked with red dots, as in the Damascus and Fustat codices. The date of their introduction remains open to debate. They could, however, be contemporary with the diffusion of O.I. More surprisingly, the script is also found on fragments from manuscripts in oblong format and the varying quality of its execution from one copy to another betrays a diffusion that was not restricted to the ruling elite or to imperial commissions. The sometimes conservative orthography may point to a rather early date.
Conclusion There is no doubt that the ruling circles of the Umayyad state became concerned with the Qurʾan as a written text. According to Omar Hamdan, an effort was made under al-Ḥajjāj’s supervision between 84 and 85 ah/ 703 and 705 ad.34 It aimed at providing the Muslim community with an improved muṣḥaf: the orthography was reformed; the diacriticals added
7
A Qurʾanic Script from Umayyad Times
77
to a rasm, which did not have any; the components of the Qurʾan were enumerated; and in the end copies were sent in various large cities of the empire. Some of our manuscripts exhibit such features. In addition, the political overtone of the move is quite obvious and the implication of the ruling circles may have resulted in the commissioning of copies that were to reflect the might of the dynasty and its religious commitments. The Codex of Fustat may have been one of those copies as we know that al-Ḥajjāj had sent one to Fustat.35 As we have seen, other copies of the Qurʾan can be directly related to the same group of script and give more strength to the hypothesis of some form of control over the transcription. There is now a common reference (with the same shapes and some common habits) as if some sort of teaching/training had been set up for copyists/calligraphers and possibly an institution where the transcription was performed or at least controlled. The fragments that survive were not all produced on the same scale as the Fustat or even Damascus codices, as far as the illumination is concerned, but their script exhibits consistent features. The Fustat and Damascus codices are, of course, important for the information they provide about their date and the chronological implications they entail. They also show a dramatic change in the conception of the muṣḥaf and the emergence of new concerns. First, these books reflect through their beauty the importance and the perfection of the text. Among the reasons behind this evolution is a general tendency, which has been described in the following terms by Barry Flood about architectural realizations of the Umayyads: The desire to rival the best efforts of the Christians and the need to convince by appearances were adequately addressed by the construction of a monumental ensemble which was not only worthy of an imperial capital, but strongly redolent of that most familiar by sight or reputation to the Syrian subjects of the Umayyads.36 In the field of manuscript production, the new muṣḥaf, eventually produced under official patronage, was challenging by its appearance Christian Bibles. Next to this apologetic concern, which addressed both inner and outer needs, the sharp contrast between the earlier muṣḥafs in Hijazi style with their highly individual appearance and these several copies that were all transcribed in O.Ia suggests an increased control over the
87
78
F ran ç ois D é roche
text and its transmission. On the other hand, these Qurʾanic manuscripts, transcribed in the same script, stress the unity of the community and the uniformity of its sacred text. Choices involving the materiality of the book, like the regularized script, the calamus adapted to it, the margins, and in some cases the illuminations, helped bring about these aesthetically and ideologically motivated changes.
Notes 1. This chapter largely reproduces material published in Déroche 2014, chap. 3, courtesy of Brill Publishers. 2. Déroche 2004, 227–264; also George 2010, 75–78, figs. 50–51. 3. The descriptions prepared by M. Amari were included by W. McGuckin de Slane (1883–1895, 91). 4. De Slane 1883–1895, 92. 5. Déroche 1983, 144–145, no. 268. 6. Vasilyeva 1996, 20. 7. See Déroche 2014, chap. 3. 8. Déroche 2002, 629–634. 9. Demirkol and Kutluay 2010, 139–140. Also see Déroche forthcoming. 10. Déroche 2002, 629. 11. Compare with Déroche et al. 2006, 72–76. 12. A similar situation can be observed in the manuscript Sanaa, Dar al-Makhtutat, Inv. 20–33.1 (see Von Bothmer 1987, figs. 14 or 18, for instance). 13. Suras 16 to 23, 26, 27, 31 to 41. 14. Headings for sura 35 (f. 29b) and 37 (34a). 15. See Déroche forthcoming. 16. See Déroche 2009, 60–63. 17. Déroche 2009, 31–34, pl. 1–2. 18. See Diem 1983, 386–387 (parag. 242). 19. See Déroche 1999, 87–94 and pl. 15–16 [In memoriam M. Meinecke]. 20. See Déroche 2009, pl. 2, 4, 6. 21. See Déroche 1983, 61, no. 6. 22. See f. 21b, line 15, for instance. 23. See Déroche 2009, 127–130, pl. 18–28. 24. Déroche 2009, 31–43, pl. 1–10. 25. In principle, the next letter currently available in the typology proposed by the present writer (Déroche 1992, 34–47, 132–137; a development of Déroche 1989, 35–47; see also George 2010, 148–161) is “G.” However, since other groups (Hijazi and NS) do not comply with the sequence, I decided for mnemonic reasons to choose for the new group the letter “O” (for “omeyyade”).
97
A Qurʾanic Script from Umayyad Times
79
26. Déroche 2009, 39–41, pl. 7–8. 27. George 2010, 69. For a comparison between the script of Saint Petersburg, NLR Marcel 13 and ‘Abd al-Malik’s inscriptions, see Déroche 2006, 231–234, fig. 2. 28. Van Berchem 1922, 17–29; Déroche 2006, 231–234, fig. 2. 29. Déroche 2006, 232–234. 30. George 2010, 56–74. 31. al-Jahshiyārī 1938, 37; 1958, 85–86. 32. al-Jahshiyārī 1938, 47; 1958, 94. 33. A first descriptive list of these fragments is available in Déroche 2014. 34. Hamdan 2006, in particular, p. 141. 35. For this episode, see Ibn Duqmāq 1893, 72–74; al-Maqrīzī 2001, 30–31. M. Tillier suggested that the Codex Parisino-petropolitanus may have been Asma’s muṣḥaf (2009, 112–113). Actually, I stated that it seemed unlikely that the manuscript was transcribed “dans un contexte officiel” (Déroche 2009, 153), which excludes Tillier’s interpretation and in any case before the “Maṣāḥif project” of al-Ḥajjāj. 36. Flood 2001, 226.
Bibliography Demirkol, A. S., and S. Kutluay. 2010. “Türk ve Islam Eserleri Müzesi Kur’an-ı kerim koleksiyonu hakkında.” In 1400: Yılında Kur’an-ı kerim, edited by Müjde Unustası, 139–140. Istanbul: Antik A.Ş. Kültür Yayınları. Déroche, F. 1983. Catalogue des manuscrits arabes, Partie 2 Manuscrits musulmans, T. 1:1 Les manuscrits du Coran: Aux origines de la calligraphie coranique. Paris: Bibliothèque Nationale. ———. 1992. The Abbasid Tradition: Qurʼans of the 8th to the 10th Centuries ad. London: Oxford University Press. ———. 1999. “Un critère de datation des écritures coraniques anciennes: Le kâf final ou isolé.” Damaszener Mitteilungen 11: 87–94. ———. 2002. “New Evidence about Umayyad Book Hands.” In Essays in Honour of Salâh al-Dîn al-Munajjid, 629–634. London: Al-Furqan Islamic Heritage Foundation. ———. 2004. “Colonnes, Vases et Rinceaux: Sur quelques enluminures d’époque Omeyyade.” Comptes-Rendus des Séances de l’Académie des Inscriptions et Belles- Lettres 148(1): 227–264. ———. 2006. Islamic Codicology: An Introduction to the Study of Manuscripts in Arabic Script. Translated by D. Dusinberre and D. Radzinowicz. London: Al-Furqān Islamic Heritage Foundation. ———. 2014. Qur’ans of the Umayyads: A First Overview. Leiden: Brill. ———. Forthcoming. “La bibliothèque de la mosquée des Omeyyades: Les documents qui accompagnent les manuscrits.” In Ecrire l’histoire de Damas: Nouvelles données archéologiques et nouvelles sources sur une métropole arabe à l’époque médiévale, edited by J. M. Mouton.
08
80
F ran ç ois D é roche
De Slane, W. M. 1883–1895. Catalogue des manuscrits arabes. Paris: Imprimerie Nationale. Diem, W. 1983. “Untersuchungen zur frühen Geschichte der arabischen Orthographie IV: Die Schreibung der zusammenhängenden Rede, Zusammenfassung.” Orientalia NS 52(3): 357–404. Flood, B. 2001. The Great Mosque of Damascus: Studies in the Makings of an Umayyad Visual Culture. Leiden: Brill. George, A. 2010. The Rise of Islamic Calligraphy. London: Saqi. Hamdan, O. 2006. Studien zur Kanonisierung des Korantextes: Al-Ḥasan al-Baṣrīs Beiträge zur Geschichte des Korans. Wiesbaden: Harrassowitz. Ibn Duqmāq. 1893. Description de l’Egypte par Ibn Doukmak publiée d’après le manuscrit autographe conservé à la Bibliothèque khédiviale, part 1. Edited by K. Vollers. Cairo: Imprimerie Nationale. al-Jahshiyāri, Muḥammad b. ʿAbdūs. 1938. Kitab al-wuzarāʾ wa’l-kuttāb. Edited by M. al-Saqqā, I. al-Ibyārī, and ʿA. Shalabī. Cairo: Muṣṭafā al-Bābī al-Ḥalabī wa-Awlāduhu. ———. 1958. Das Buch der Wezire und Staatssekretäre von Ibn ʿAbdūs Al- Gahšiyārī: Anfänge und Umaiyadenzeit. Translated by J. Latz. Walldorf: Verlag für Orientkunde H. Vorndran. al-Maqrīzī. 2001. al-Mawāʿiẓ wa’l-iʿtibār bi-dhikr al-khiṭaṭ wa’l-āthār, vol. 4. Edited by A. Fuʾād Sayyid. London: Al-Furqan Islamic Heritage Foundation. Tillier, M. 2011. Review of La transmission écrite du Coran dans les débuts de l’Islam: Le codex Parisino-petropolitanus, by F. Déroche. Journal of Qurʾanic Studies 13(2): 109–115. Van Berchem, M. 1922. Matériaux pour un Corpus inscriptionum arabicarum, Deuxième partie: Syrie du Sud. t. 1. Jérusalem “ville.” Cairo: Institut Français d’Archéologie Orientale. Vasilyeva, O. 1996. “Oriental Manuscripts in the National Library of Russia.” Manuscripta Orientalia 2(2): 19–35. Von Bothmer, H. C. 1987. “Architekturbilder im Koran: Eine Prachthandschrift der Umayyadenzeit aus dem Yemen.” Pantheon 45: 4–20.
81
PART II
Patronage and Power: The Evidence of the “Desert Castles”
28
83
4
Hishām’s Balancing Act The Case of Qaṣr al-Ḥ ayr al-Gharbī Robert Hillenbrand*
two principal aims. The first is to draw attention to the borrowings from classical and Byzantine and, to a lesser extent, Sasanian art at Qaṣr al-Ḥayr al-Gharbī, assessing the changes to which some of them were subjected in the process of transference. The second aim is to determine whether these borrowings from such very disparate traditions were random or whether they were coordinated by some underlying purpose. A third and related question of equal interest, namely the spirit in which these themes, motifs, and ideas are used,1 will also make an occasional appearance in what follows. These questions illuminate the gradual strengthening of eastern traditions and, indeed, of the political impact of the eastern provinces of the Umayyad empire in its closing decades.2 The discussion will, in passing, draw attention to some unfamiliar material. The larger topic of the site of Qaṣr al-Ḥayr al-Gharbī and its wider context, whether military, palatial, commercial, or agricultural, is outside the remit of this essay, not least because it is these topics that have attracted the lion’s share of recent scholarship on the site, to the detriment of its decoration.3 T H I S E S S AY H A S
*
My warm thanks to Alain George for his helpful suggestions about this essay.
48
84
R obert H illenbrand
Borrowings from the Classical, Early Christian, and Byzantine Traditions The borrowings from the classical, early Christian, and Byzantine traditions are easily outlined. They infiltrate every part of the building, extending from the architectural type itself to the smallest details of its decoration, and including actual spolia, for example, in the entrance gateway (fig. 4.1). It is enough to list them in rather cursory fashion here, since the presence of strong classical and Byzantine elements in this building has never been in dispute.4 Even so, the way that this imprint finds expression, as distinct from its mere existence, will stand out in sharper relief if one or two examples are selected for brief analysis. The total of these borrowings may seem formidably large, but that is no reason to assume that Qaṣr al-Ḥayr al-Gharbī is predominantly a classicizing monument. A similarly long list of Sasanian motifs and echoes of Sasanian styles could also be drawn up, and their cumulative evidence could be used to reach a precisely contrary conclusion. To analyze them in the necessary depth, however, would require another study all to itself.5 The point to stress here is that the presence of both western (classical and Byzantine) as well as eastern (Sasanian and Central Asian) motifs6 announces a building in which
Figure 4.1 Qaṣr al-Ḥayr al-Gharbī: stone classical sculpture, date uncertain, reused in the gateway.
85
Hishām’s Balancing Act
85
a delicate balance is kept between the two principal artistic traditions of western Asia in Late Antique times.7 The implications of these remarks will be assessed in the second part of this essay. In the meantime, it will be convenient first to deal with the classical and Byzantine elements in the ground plan, elevation, architectural vocabulary, sculpture, and frescoes of this building.8 Before that is attempted, however, it might be as well to spell out what is implied by the notion of “the classical imprint” that can be detected at Qaṣr al-Ḥayr al-Gharbī. To begin with, “classical” may seem an unacceptably broad and unspecific term. Yet the range of pre-Islamic art forms, whether Hellenistic, Roman, early Christian, or Byzantine—all of them, incidentally, branches from the same tree—whose traces can be detected in this building is remarkably varied. It quite excludes the possibility that the Umayyads drew for inspiration only on the living, that is Byzantine, architectural tradition of Syria. There was no reason why they should have experienced the kind of inhibition that might have beset a contemporary Christian architect at the thought of employing motifs or techniques that had specifically pagan associations or that were many centuries out of date. Anachronisms held no terrors for them. Moreover, Hellenistic, Roman, Palmyrene, and early Christian monuments were all around them, and to a degree that is hard for us to imagine today. And whether the actual craftsmen employed had been trained in a local Christian Syrian idiom, or in a Sasanian Iraqi one, or indeed in one or another of the substyles current in the Near East at that time, the key fact was that they were working to the orders of an Umayyad patron. In that sense they were Umayyad craftsmen. Small wonder, then, that, if so directed, Umayyad craftsmen should draw impartially on the accumulated architectural heritage of the previous millennium in the Mediterranean world and that of western and central Asia. Parallels with, say, Roman architecture are therefore intrinsically more plausible than they would be in the case of buildings erected in Christian Syria a century earlier. What more likely than that one world empire should exert an impact on another? Even the notion of a “classical imprint” demands a brief gloss in this context.9 In the present discussion it can have two markedly different meanings. It can refer, first, to the presence of craftsmen who, whatever their land of origin, brought to the construction of Qaṣr al-Ḥayr al-Gharbī their distinctive local combination of material, technique, style, and image making. Alternatively, it can refer to the memory, or the actual presence,
68
86
R obert H illenbrand
of a foreign object or a foreign way of doing things, which duly had their effect on the craftsmen. With these introductory remarks in mind, one may proceed to an examination of the building. Qaṣr al-Ḥayr al-Gharbī, then, is an Umayyad princely residence in Syria, set in the solitude of the Palmyrene steppe roughly halfway between Damascus and Palmyra itself. Schlumberger, who excavated the site from 1936 to 1938, attributed the palace to the caliph Hishām (r. 724–743) and dated it to the same time as the nearby building, which he identified as a khān, namely 727.10 It is no part of the purpose of this essay to challenge these suggestions. To recapitulate familiar material, in form the building is a square of some seventy meters per side—a favorite Roman unit—whose regular bastions and single entrance recall the Romano-Byzantine forts that protected the Provincia Arabia. Inside, however, the flavor is domestic, not military; yet here again a Roman model is ready to hand—for the courtyard with living units opening off it irresistibly recalls the atrium of the typical Roman house with its enclosing portico.11 The hybrid nature of this arrangement—private house inside, fortified castle outside (fig. 4.2)— serves notice of the attitude the Arabs held toward the architecture they encountered. They tended to devise their own combinations of familiar types, and some of the results are so unlooked-for that they suggest a shotgun marriage.12
Figure 4.2 Qaṣr al-Ḥayr al-Gharbī: plaster model of the palace. Damascus, National Museum.
8 7
Hishām’s Balancing Act
87
The Gateway Equally hybrid is the gate itself, which brings to mind some gigantic wedding cake (fig. 4.3). Broadly speaking it conforms to Roman type in its basic form of two salient bastions flanking the entrance proper. But, as Qaṣr al-Ḥayr al-Sharqī demonstrates, Umayyad architects could achieve a far more authentically Roman grandeur in this idiom when they chose.13 At Qaṣr al-Ḥayr al-Gharbī the impact is transformed by the fussy plethora of ornament that this structural framework is called upon to bear. Abstract repetitive stucco decoration of a broadly organic nature invades the lower half of the gateway. With its even covering of the field, its sense of endless pattern, guillotined by the frame so as to suggest infinity, and its lack of background, it is typical of Sasanian architectural ornament from Iran (fig. 4.4). This is not nature, but nature methodized, with originally meaningful motifs like the pearl roundel or the rosette demoted to mere ornament. The combination of geometric structure and vegetal infill is also typically Sasanian.14 At the halfway mark, however, the nature of the decoration changes abruptly. Now it employs a simplified classical vocabulary of blind arcades, triangular pediments, scallop-shell tympana, and figures
Figure 4.3 Qaṣr al-Ḥayr al-Gharbī: entrance gateway, mainly mud brick and stucco, ca. 727. Damascus, National Museum.
8
88
R obert H illenbrand
Figure 4.4 Qaṣr al-Ḥayr al-Gharbī: stucco panels from lower part of gateway, ca. 727. Damascus, National Museum.
clad in classical drapery. This vocabulary is manipulated with easy assurance, but the discrepancy between its tiny scale and its location high up on the monument is plain to see. If this section had all along been intended to be seen only indistinctly, that would be explanation enough for the rather summary execution of its finer details, including drapery, faces, and capitals. This factor would not, however, account for such minor solecisms as the replacement of standard semicircular arches by triangular forms, or the aberrant capital types. These, unlike their counterparts inside the building—a telling detail, this, for it points to the presence of a different team of craftsmen—belong in no category known to the classical world, and they thus flout the extreme conservatism of this genre, where quite small changes might exert major impact.15 Interspersed with this decoration, which might justifiably claim kinship with the figural triumphal arches of Late Antiquity, such as the Arch of Constantine, are other forms that derive from dado or even floor decoration found in the villas of Byzantine Syria—plain adjoining lozenges with some central motif.16 Such simple designs depended for their success on the colored stone or marble of which they were made;17 reduced to plaster and transported 10 meters aloft, they understandably lose much of their appeal. One may feel, too, that the addition of a Palmyran-style bust at the center of such a lozenge devalues an image that in its original context, as a painstakingly carved funerary bust—often attempting a close likeness of the deceased—had both style and power.18 The verdict on the classicizing decoration of the upper half of this gateway must therefore fall some way short of unqualified approval. But while no Roman or Byzantine architect
98
Hishām’s Balancing Act
89
could of his own volition have executed the gateway as it stands, he would have had in his head if not in his repertory each of the models on which it depends—fort, villa, and triumphal arch. It is both the combination and the fleshing out of these three basic skeletons that is new.
The Interior of the Palace Sculpture What of the interior? Here interest centers on the frescoes, the tympana, and the figural sculptures festooning (according to the current restoration) the upper story of the portico enclosing the courtyard. The latter, being more numerous and varied, and less fully published, will receive more detailed treatment here.19 Of particular interest is the context chosen for these sculptures. The idea of placing figural sculpture high up in an architectural setting has an honorable ancestry in Mediterranean art, for it goes back at least as far as the Parthenon. By degrees this theme developed certain standard variations. One conventional rendering of it that enjoyed particular favor in the Hellenized communities of the Middle East and Central Asia was the disposition of figures in a continuous arcade.20 This version of the theme excluded any substantial degree of interaction between the figures, for they had very little room in which to move. At best, two figures might face each other; but the battle and processional themes of earlier centuries were scarcely a viable option within this formula. Thus the architecture to a certain extent dictated the subject matter, privileging some kind of interconnected or narrative iconography. It is to the credit of the sculptors at Qaṣr al-Ḥayr al-Gharbī that they rejected the rather crude simplifications of this theme, popular, for example, in the Hellenized cities of central Asia,21 and tried to infuse it with greater spatial freedom and human interest. They did so by resurrecting the theme of a procession. This ensures that the figures interact effectively enough, while the unbroken architectural framework—still an arcade, but one subdivided into panels in which two rounded terminations flank a triangular one—ensures a basic continuity between each figure or pair of figures and the next.22 It also meant that the potentially wearisome formality of successive frontal images—as at the Hellenistic site of Ai Khanum on the Oxus23—could be relieved, since the figures at Qaṣr al- Ḥayr al-Gharbī are depicted by turns in frontal, profile, or three-quarter- view poses. Their activities, too, are varied—though their lofty location on the first floor of the courtyard portico imposed a need for simplified
09
90
R obert H illenbrand
Figure 4.5 Qaṣr al-Ḥayr al-Gharbī: tiger pacing to the right, stucco panel, ca. 727. Damascus, National Museum.
Figure 4.6 Qaṣr al-Ḥayr al-Gharbī: tiger pacing to the left, stucco panel, ca. 727. Damascus, National Museum.
depiction. Hence their pronounced three-dimensional, volumetric character and the highlighting of limbs by carefully spaced, sinuous drapery folds. Animals include two pacing affronted tigers (figs. 4.5–4.6), a horse and rider,24 a lioness bringing down its prey (Plate 3 in color insert), several birds (fig. 4.7),25 a pair of affronted lions (fig. 4.8), and a dog. But the majority of the figural sculpture depicts people, most of them male.
91
Hishām’s Balancing Act
91
Figure 4.7 Qaṣr al-Ḥayr al-Gharbī: bird, stucco sculpture, ca. 727. Damascus, National Museum.
Figure 4.8 Qaṣr al-Ḥayr al-Gharbī: affronted lions, stucco sculpture, ca. 727. Damascus, National Museum.
An armored warrior stands at attention;26 servants bring food and drink (fig. 4.9); a grim-visaged bearded man in Roman dress sits enthroned; a man walks forwards but with his head turned back (Plate 4 in color insert); another is depicted with raised knee and extended leg in movement, as if mounting steps, while yet another standing figure, with pointed headgear, carries a bird; and a bearded man strokes his chin in the manner of
29
Figure 4.9 Qaṣr al-Ḥayr al-Gharbī: servants offer food and drink, stucco sculpture, ca. 727. Damascus, National Museum.
Figure 4.10 Qaṣr al-Ḥayr al-Gharbī: bearded man strokes his chin pensively, stucco sculpture, ca. 727. Damascus, National Museum.
9 3
Hishām’s Balancing Act
93
Figure 4.11 Qaṣr al-Ḥayr al-Gharbī: naked child, stucco sculpture, ca. 727. Damascus, National Museum.
a classical philosopher deep in thought (fig. 4.10).27 Several of these males are young: a naked boy (fig. 4.11),28 a young man throwing a ball (fig. 4.12), another youth in a challenging pose with head cocked and hands on hips (fig. 4.13), while a third shoulders a palm branch (fig. 4.14),29 and a fourth with a page-boy haircut and downcast eyes was perhaps left unfinished (fig. 4.15).30 But women also make an appearance; the program includes topless maidens offering beverages (fig. 4.16),31 half-length, fully clothed matrons depicted frontally, a bejeweled woman in frontal pose seated on a bench with legs akimbo (fig. 4.17), a half-naked woman holding aloft two birds in her right hand while her left grasps a cornucopia over her shoulder, and a woman carrying a basket of fruit.32 Three sculptures have a strong local and classical flavor: a couple comprising a seated and a reclining figure (Plate 5 in color insert), a bare-armed youth carrying an animal in his arms (fig. 4.18), and an image of a bare-breasted woman holding one of her plaits with one hand and a dove with the other (fig. 4.19).33 The first is of unmistakably Palmyrene inspiration, the
49
Figure 4.12 Qaṣr al-Ḥayr al-Gharbī: youth throwing a ball, stucco sculpture, ca. 727. Damascus, National Museum.
Figure 4.13 Qaṣr al-Ḥayr al-Gharbī: youth with upturned head, stucco sculpture, ca. 727. Damascus, National Museum.
9 5
Figure 4.14 Qaṣr al-Ḥayr al-Gharbī: youth with palm branch, stucco sculpture, ca. 727. Damascus, National Museum.
Figure 4.15 Qaṣr al-Ḥayr al-Gharbī: youth with page-boy haircut, stucco sculpture, ca. 727. Damascus, National Museum.
69
Figure 4.16 Qaṣr al-Ḥayr al-Gharbī: maiden offers beverage, stucco sculpture, ca. 727. Damascus, National Museum.
Figure 4.17 Qaṣr al-Ḥayr al-Gharbī: seated woman, stucco sculpture, ca. 727. Damascus, National Museum.
9 7
Figure 4.18 Qaṣr al-Ḥayr al-Gharbī: youth carrying an animal in his arms, stucco sculpture, ca. 727. Damascus, National Museum.
Figure 4.19 Qaṣr al-Ḥayr al-Gharbī: bare-breasted woman holding a bird, stucco sculpture, ca. 727. Damascus, National Museum.
89
98
R obert H illenbrand
second recalls a standard sacrificial scene, and here too a Palmyrene model has been suggested;34 while the third, perhaps an image of the goddess Atargatis, follows Nabatean models.35 Finally, it is worth noting a branch or small tree in full leaf (fig. 4.20), of which more anon. Two distinct styles have been identified in this sculpture.36 Talgam contrasts the sharp angles and strong oppositions of light and shade that characterize the first style with the soft and rounded impression of the second style. These contrasting techniques can be seen at work in the treatment of eyes in much the same way as Hamilton noted at Khirbat al- Mafjar:37 either as a double line around the eye, creating a sunken effect, or as a strongly protruding eyeball. Other resemblances to the sculpture at Khirbat al-Mafjar include the reduction of drapery folds into boldly diagonal stripes and tiers of half ovals, as well as a certain flattening and abstraction of the face. In both cases, too, the sculptures project a presence, a monumentality that belies their small size (which is almost always less than 70 centimeters).38 But much of this sequence of images has been lost, and that creates obvious problems of interpretation. While an ancient
Figure 4.20 Qaṣr al-Ḥayr al-Gharbī: tree or leafy branch on entrance façade, stucco sculpture, ca. 727. Damascus, National Museum.
99
Hishām’s Balancing Act
99
Near Eastern origin for the ideas behind such images can be proposed,39 the visual language of most of these figures is indisputably classical— and classical, moreover, in the proto-medieval mode of Late Antiquity. Yet there are also sculptures in which a Sasanian manner dominates, notably the much-discussed ruler image over the entrance (fig. 4.21).40 Its huge scale, its location at the dead center of the gateway, and its easy visibility all combine to give it pride of place. So the importance of the Iranian tradition at Qaṣr al-Ḥayr al-Gharbī makes itself felt at the very outset. Unfortunately, the problem is not confined to missing elements. The exact context of some of the crucial images, namely the tigers and the principal enthroned dignitary (fig. 4.22),41 whose thoroughly Western character immediately invites comparison with the thoroughly Eastern character of the royal image over the gateway, remains doubtful, an uncertainty compounded by differences in scale and in the framing devices employed in these figural sculptures. Thus their full significance is
Figure 4.21 Qaṣr al-Ḥayr al-Gharbī: ruler image, eastern style, over entrance, stucco sculpture, ca. 727. Damascus, National Museum.
001
100
R obert H illenbrand
Figure 4.22 Qaṣr al-Ḥayr al-Gharbī: ruler image, western style, stucco sculpture, ca. 727. Damascus, National Museum.
elusive. Nevertheless, the wide range of connected images here, and their subordination to an overall program, makes it tempting to see this ensemble as a forerunner of those later Islamic princely cycles which (just like this one) operated in a partially Christian social context, were located high up in the building they graced, and gave a detailed if symbolic account of court life under Islamic rule. Foremost among them, of course, are the Armenian Church of Aght’amar, dated 921,42 and the Cappella Palatina in Palermo, datable after 1143.43 In both these examples the Islamic subject matter gains added impact from its non-Islamic surroundings. Yet it should not be forgotten that the princely cycle closest in date to Qaṣr al-Ḥayr al-Gharbī is not Islamic at all but Byzantine. It is to be found at Kiev; but this remoteness is deceptive, for the images in question— which include circus scenes—have been shown to derive from a probably much fuller cycle in Constantinople.44 The portico sculptures at Qaṣr al- Ḥayr al-Gharbī—assuming that this was indeed their location within the building—may therefore have a purer Byzantine pedigree than has yet been suspected. And a further possibility might be canvased—that, just as at
0 11
Hishām’s Balancing Act
101
Quṣayr ʿAmra, and perhaps also at Khirbat al-Mafjar, these images could be interpreted as a calendrical program on the Roman model.45 This, as Hanfmann has shown,46 is replete with associations of cosmic order, eternal happiness, and the prosperity and fertility guaranteed by divinely ordained government—themes with a powerful propagandist charge. It would have been a typically Umayyad move to lay claim to these ideas on behalf of the new imperium. The relevance of such an interpretation to an agricultural manor is plain. Nor is there any need to be surprised if not all the iconographical details represent an exact fit, for it is wholly characteristic of Umayyad art to make unexpected adaptations to inherited themes.
Lunettes There are some 60 door-lunettes at Qaṣr al-Ḥayr al-Gharbī, so many that they cannot very well be ignored.47 They derive from the claustra of Late Antique buildings,48 models that were later taken up in Christian and Muslim Spain and in the churches of medieval Byzantium and Italy.49 Their purpose at Qaṣr al-Ḥayr al-Gharbī, set as tympana above doorways, was to let light and air into the cramped, dingy cells surrounding the courtyard, and prevent the entry of birds and swirling sand. Theirs is an art of silhouette, and photographs that emphasize the surface ornamentation of these grilles misrepresent them. Roman and Byzantine claustra were uniformly geometric, and simple checkerboard grilles were preferred. Apparently the possibility of developing such designs further was not seriously considered before the Islamic period, although classical and Byzantine floor mosaics provide abundant proof that extremely complex geometric designs were frequently used in that medium from the time of Augustus (d. 14 ad) onward.50 In early Islamic times craftsmen experienced no inhibitions in putting such designs to work in windows and tympana, and they lost no time in developing the ideas that they incorporated. Certainly the tympana of Qaṣr al-Ḥayr al-Gharbī leave Roman and Byzantine claustra far behind in the sophistication of their design, and for good measure they also draw on the highly abstract plant ornament of Sasanian Iran (fig. 4.23).51 Those tympana and windows that exploit geometrical motifs, however, are a relatively rare example of the classical source inspiring more complex work executed in the self-same idiom (fig. 4.24).52 The marble grilles of the windows on the west back wall of the Great Mosque of Damascus (built 706–715) showed the way here.53 The balustrades at Qaṣr al-Ḥayr al-Gharbī also have strongly Sasanian elements (fig. 4.25).
021
Figure 4.23 Qaṣr al-Ḥayr al-Gharbī: lunette of eastern type, stucco sculpture, ca. 727. Damascus, National Museum.
Figure 4.24 Qaṣr al-Ḥayr al-Gharbī: lunette of western type, stucco sculpture, ca. 727. Damascus, National Museum.
013
Hishām’s Balancing Act
103
Figure 4.25 Qaṣr al-Ḥayr al-Gharbī: balustrade, stucco sculpture, ca. 727. Damascus, National Museum.
The Frescoes The frescoes are far more varied and numerous than published accounts suggest. They include a relatively well-published female head (fig. 4.26)54 and one face of surpassing ugliness—Beauty and the Beast. The latter forms part of a grotesque scene of low life (Plate 6 in color insert),55 for which Roman parallels, especially in scenes from the theater,56 abound, and for which analogies from the floor of the Great Palace at Constantinople,57 from Khirbat al-Mafjar,58 and even from central Asian Buddhist temples59 may be cited; architecture-scapes, a familiar category in Roman60 and early Byzantine61 wall decoration; and exotic creatures such as the peacock, which bring to mind the mosaics of Antioch62 and other eastern Mediterranean sites.63 In others, to be mentioned briefly later, the subject matter is geometric design.64 Pride of place, however, must go to a pair of gigantic frescoes placed—most unexpectedly—on the floor.65 It is a context that would naturally—especially in Late Antique Syria66—call for mosaic work. When this substitution of an expensive material by a cheap one is considered alongside the use of stucco (not stone) on the external façade, and of plaster (not marble) for the dadoes of the inner chambers (fig. 4.27), a pattern of cost cutting emerges. It is entirely consonant
041
Figure 4.26 Qaṣr al-Ḥayr al-Gharbī: image of a comely woman, wall painting, ca. 727. Damascus, National Museum.
Figure 4.27 Qaṣr al-Ḥayr al-Gharbī: inner chamber with painted plaster dado and lunettes, wall painting, ca. 727. Damascus, National Museum.
015
Hishām’s Balancing Act
105
Figure 4.28 Qaṣr al-Ḥayr al-Gharbī: southern floor fresco of musician and hunter, ca. 727. Damascus, National Museum.
with the lurid tales of Hishām’s miserliness that characterize the literary sources dealing with his reign.67 One of these frescoes depicts musicians and a prince hunting gazelle (fig. 4.28), the whole conceived in convincingly Sasanian fashion but for the youthfulness of the hunter. A horizontal frame separates these two scenes, while the lowest third, largely destroyed, depicted a retainer leading a tagged animal into a game preserve (fig. 4.29). While al-ḥayr can have that meaning, among others, and the palace bears that name, the archaeological evidence for a network of canals points compellingly to an enclosure intended for agriculture, not hunting.68 The tripartite division of the fresco is reflected in some Sasanian rock reliefs.69 In technical terms
061
106
R obert H illenbrand
Figure 4.29 Qaṣr al-Ḥayr al-Gharbī: modern painted record of southern floor fresco (after Schlumberger 1946–1948).
the treatment is at once linear and graphic, with no painted background. One detail after another is unmistakably Persian, from hairstyles to costume, from headgear to fluttering ribbons. The ornamentation of the horse, down to the plaited tail and painted hooves, also follows Sasanian precedent, as does the flying gallop.70 All this is in the spirit of Sasanian imperial silverware, a spirit faithfully recaptured in the later Islamic cycle of princely life and leisure. While this essay is not concerned with the minutiae of Sasanian modes in the art of Qaṣr al-Ḥayr al-Gharbī, it is very relevant to the main argument outlined below that the Sasanian element is so strong here.
017
Hishām’s Balancing Act
107
Figure 4.30 Qaṣr al-Ḥayr al-Gharbī: northern floor fresco with tritons and Gaea, ca. 727. Damascus, National Museum.
The other floor fresco (fig. 4.30) is much less readily characterized, and this difference is in itself worthy of comment. For while the “Sasanian” fresco adopts a predominantly narrative mode, the “Byzantine” one can be read only in a symbolic sense.71 It makes no pretense of storytelling. The central image, a woman in a medallion, is identified by her attributes—namely the snake around her neck and her lapful of produce—as the goddess Gaea, Gaia, or Ge, the earth (fig. 4.31). As such the image is not technically Byzantine at all—though Coptic versions exist72—but a throwback to Roman art.73 What can such an image possibly be doing in an Islamic context? Perhaps its immediate meaning was that the entire world—earth, sea, and air74—was now under the authority of the Muslim caliph. But that is by no means the only possible interpretation. It may be that since Christianity and not paganism was the major living religion of Syria at that time, the image of a personified Greco-Roman goddess had already been shorn of its power to awe and therefore (in this context) to shock.75 If so, this was no longer an icon exerting spiritual authority.
081
108
R obert H illenbrand
Figure 4.31 Qaṣr al-Ḥayr al-Gharbī: northern floor fresco, detail of Gaea, ca. 727. Damascus, National Museum.
Its obvious association with Rome would also have established a political rather than religious connection between the Islamic empire and its famous predecessor, and thus have proclaimed—as did the neighboring fresco of Sasanian inspiration—the right of the new Islamic polity to belong to that most exclusive club of all, the Family of Kings.76 Or one might suggest that its interest could have been largely antiquarian. And it was removed still further from the real world by its attendant— and obviously mythical—sea centaurs. These, by the way, seem to be an Islamic invention, a felicitous conjunction of triton and sea serpent, which develops an idea found at Dura Europos on the Euphrates in the third century ad (fig. 4.32).77 Nor do these suggestions exhaust the range of possible meanings of this enigmatic image, which may very well have been intended to speak with multiple voices. For though Gaea is divested of her religious significance here, she retains in full measure her connotations of power and fertility. The latter association might have been peculiarly gratifying to Hishām, a man generally criticized for his obsession with digging canals and making the desert bloom.78 It would have been an association especially à propos for Qaṣr al-Ḥayr al-Gharbī, for the Muslims built the largest pre-modern dam in the Near East a mere 16.5
091
Hishām’s Balancing Act
109
Figure 4.32 Roof tile with sea creature, synagogue, Dura Europos, ca. 245 ad. Damascus, National Museum.
kilometers away,79 thereby enabling the royal residence to function also as the center of an agricultural estate. Similar references to agricultural fertility have been invoked to explain the riddling mosaic of knife and fruit at Khirbat al-Mafjar,80 another image perhaps deliberately intended to be susceptible to multiple interpretations. The lower or eastern part of the fresco depicts birds and game, thus perhaps taking up the hunting theme of the other floor fresco (fig. 4.33). Given that the two frescoes were intended to be seen as a pair, it seems permissible to suggest that their meanings are linked or even complementary. If that is indeed so, the “classical” fresco might be intended as a reference to the ruler as lord of the manor—a manor that is not limited to the agricultural estate of Qaṣr al-Ḥayr al-Gharbī but extends to most of the known world from France to the Far East and thus comprises the earth itself. That earth, at once local and universal, offers him her fruits. The sea and the air also submit to him. The Sasanian fresco could be seen as the
0 1
110
R obert H illenbrand
Figure 4.33 Qaṣr al-Ḥayr al-Gharbī: modern painted record of northern floor fresco (after Schlumberger 1946–1948).
private counterpart to this public statement and would thus refer to the off-duty diversions of the ruler. Whatever the likelihood of the explanation just given, there can be no doubt that here, in the image of Gaea,81 Umayyad art skirted perilously close to the sin of unbelief (kufr). No matter how neutralized that image had become by the passage of time and by virtue of its physical context, it still represented a religion that in Muslim eyes was idolatrous—and that, moreover, within a very few years of the iconoclastic edicts of both Yazīd II (r. 720–724) and Leo the Isaurian (r. 717–741), in a country too where pagan beliefs and rites still clung tenaciously to life in some places.82 Of course, one could argue that the intention was to humiliate the image of Gaea; hence its location on the floor, where it could be trodden underfoot. Perhaps this would have been Hishām’s explanation if he had been challenged by any of the theologians at his court.83 Yet the popularity of divine images in Roman floor mosaics reveals the inadequacy of such an excuse,84 as does the use of Christian images in similar contexts.85 Be that
111
Hishām’s Balancing Act
111
as it may, nothing of this kind was attempted again in Umayyad art, perhaps because it was regarded as a dangerous experiment.
The Rationale for East versus West at Qaṣr al-Ḥ ayr al-Gharbı̄ The evidence presented so far is necessarily abbreviated, since space is limited, but it indicates clearly enough that the craftsmen responsible for Qaṣr al-Ḥayr al-Gharbī trawled very widely in search of ideas that could be put to use in its architecture and decoration. The obvious question to present itself now is whether they (or their overseers) were guided by any underlying aim in their selection of motifs. It is certainly tempting to suggest that, judging by the farrago of decoration at the site, their borrowings were random and innocent of any grand design.86 The well-known Umayyad practice of conscripting labor from the various provinces of the empire—the so-called leiturgia—could be invoked at this stage as an explanation for the otherwise baffling juxtaposition of motifs from unrelated cultures.87 Other scholars have indeed tried to account for the apparent indifference to propriety and context in the decoration of Mshatta and Khirbat al-Mafjar by means of just such arguments. Yet the more closely Umayyad architecture is studied, the less satisfactory do such arguments appear. Lurking behind them, after all, is the unspoken implication that the Arabs did not really know what they were doing, that they were nouveaux riches of the most egregious kind, content if their buildings made a splash but too ignorant to take on the architects and craftsmen of Byzantine and Sasanian Persia at their own game. Yet close analysis of precisely these maligned palaces of Mshatta88 and Khirbat al-Mafjar,89 both datable (perhaps significantly in this context) to the closing years of the Umayyad dynasty, has revealed an acute sensitivity to nuance on the part of those who built and decorated them. Nor are these isolated instances; those who raised the Dome of the Rock, the very first Muslim building to survive, had nothing to learn about the role of architecture as propaganda. It is permissible to assume, therefore, that the particular combination of motifs at Qaṣr al-Ḥayr al-Gharbī had some underlying purpose. To establish what that purpose was, some understanding of the political scene at the time is imperative. As noted above, the building is customarily regarded as contemporary with the nearby khān whose inscription mentions the caliph Hishām and gives the date Rajab 109 ah/November 727 ad.90 It was therefore built a scant ten years after the third abortive Muslim
2 1
112
R obert H illenbrand
attempt on Constantinople. That was to be the last serious Muslim attack on the city for some seven centuries; but no one could have foreseen that in the time of Hishām. For the time being the conflict was prosecuted with greater vigor on ideological grounds. The adoption of an official iconoclastic policy by the Muslim authorities in 721—the celebrated Edict of Yazīd II—was the signal for the widespread destruction of Christian images in Muslim territory.91 That this event predated by only five years the similarly iconoclastic edict of Leo the Isaurian is one of the ironies of history, but in the interim the Christians of Syria suffered none the less acutely for that. It must have seemed an equally bitter irony to them that the very caliphs associated with this iconoclastic policy in the religious sphere should embrace secular images so fervently in their palatial architecture. In comparison with this ideological struggle, actual warfare with Byzantium slackened. Along the Cilician marches and further east the military situation stabilized into a predictable annual sequence of raid and counter-raid.92 Yet Qaṣr al-Ḥayr al-Gharbī proves that the Muslims were by no means fossilized in a stance of enmity toward the Christians, whether as neighbors or as subjects. The option of persuasion remained open. That it should be especially associated with the cautious, long-headed caliph Hishām is entirely appropriate. For if it be asked how this puritanical ruler, this celebrated miser, this wrecker of mandolins93 and assaulter of poets,94 whose shamefully exiguous wardrobe excited a buzz of comment among contemporaries,95 could be responsible for as sybaritic a palace as this, the answer is surely at hand. What Hishām the private citizen would perhaps have abhorred, Hishām the caliph it seems undertook as a duty of his office, in much the same way as his predecessor Muʿāwiya I (r. 661–680) did.96 As at the Dome of the Rock, architecture is being used to make a political statement. This time, however, the message is not so nakedly polemical.97 It suggests reconciliation rather than harping on victory. The architectural vocabulary of Sasanian Iran and of the classical and Byzantine world is integrated into that of Islam. So it is that the building proclaims not conflict but co-existence under the dominion of Islam. All this may sound a trifle pat, and to forestall any misunderstandings it might be as well to emphasize first that the theory enunciated here is only a theory—others could readily be proposed—and second that it is not intended to imply that Hishām himself was an art historian born before his time, or indeed a precocious master of the propaganda machine.
13
Hishām’s Balancing Act
113
So what was his role? The entire question of patronage and its manifold implications for how buildings were conceived and executed in Umayyad times is contentious in the extreme. Nor can it be otherwise, given that not a single one of the Umayyad “palaces”—to use a conventional term—bears the name of a patron, architect, or craftsman in the form of a monumental inscription of the type so familiar in later centuries.98 So far, history has not even yielded up the names of those who designed the Dome of the Rock99 and the Great Mosque of Damascus. Scholarship therefore faces a critical, indeed inescapable dearth of information. In such a situation, normal art-historical practice dictates that the object itself—whether painting, sculpture, building, or whatever it may be—should be read intensively for relevant clues. In the case of Qaṣr al-Ḥayr al-Gharbī this procedure seems doubly justifiable in view of the confident wielding of symbolic visual language in such earlier Umayyad buildings as the Dome of the Rock. Modern scholarship has generally100 come to accept the propaganda intent of that building, and indeed fresh evidence on that score is still coming to light. Admittedly its location, its purpose, and its context— both political and religious—made the Dome of the Rock unique, but this monument nevertheless allows an a priori assumption that in its wake significant Umayyad public buildings might very well incorporate a purposefully directed iconography.101 The lengths to which such an iconography might be developed are illustrated at Khirbat al-Mafjar. Modern ignorance of the part played by its putative patron102 must not be allowed to obscure the fact that someone had the responsibility of choosing what decoration should be used as well as how and where it should be distributed. It all cost hard cash, after all. The lavish decoration of Khirbat al-Mafjar was formerly thought to reflect little more than an aimless desire for display.103 While Ettinghausen’s later interpretation of the bath hall104 might be contested at one point or another, the general thrust of his argument—that the architecture and its decoration were a vehicle for exalting the ruler, his legitimacy, and his absolute power—has carried general conviction. This evidence from a near-contemporary palace is compelling. It suggests that there would be nothing strange in the presence of a meaningful iconography in the decoration at Qaṣr al-Ḥayr al-Gharbī. Who, then, was responsible for the iconographic program? Various possibilities could be canvased. Perhaps the most straightforward one— though admittedly not the most likely105—is that the caliph himself laid down fairly explicit guidelines for a building that should incorporate both eastern and western, Sasanian and Byzantine, ideas. It is also not
4 1
114
R obert H illenbrand
inconceivable that the role of Hishām himself might have been limited to giving the crucial cue to the official charged with erecting the palace. In such a situation it would have been enough had the caliph ordered, for example, not just a palace but a palace worthy of his status as ruler of east and west. The architect, or perhaps the superintendent of works—distinctions between such offices are perforce nebulous in view of the dearth of information—might have been inspired by this hint to draw up a program characterized by a willed dichotomy between the two. In this scenario the role of the caliph as an active patron would be reduced to authorizing a program drawn up by someone else. In other words, even if Hishām himself had been utterly incapable of conceiving the details of the program, there is enough reason to believe (given the evidence of both earlier and later Umayyad architecture) that he would have had no difficulty in finding somebody who could do so. The crucial detail is that he must have approved of the result. Alternatively, the occasion for this contrast of styles could have been accidental and thus somewhat more mundane: for example the arrival—perhaps by some vagary of the corvée system—of two teams of artisans, one working within an eastern tradition, the other in a western one. The speedy completion of the building would presumably have been most readily achieved by setting both teams different and well- defined areas to decorate.106 Nor should one exclude the possibility that the choice of teams of craftsmen who represented two very different traditions was itself deliberate, and indeed that it followed naturally from the decision to embody both eastern and western ideas in this palace. These are no doubt only some of the theories that could be proposed to account for the layout and execution of Qaṣr al-Ḥayr al-Gharbī. Irrespective of which one of them ultimately finds favor, however, all share one characteristic: they assume the deliberate segregation of the teams of workmen involved. This seems to have been an essential part of the creative process, which resulted in the iconographic program under discussion. It may be urged that the assumptions built into these various theories are simply too many and too bold. Yet in the absence of any clear textual or epigraphic evidence, the problem of reading this program and unraveling its genesis poses a stark dilemma: hypothesis or silence. The former option seems preferable because at least it provides a basis for further discussion. This essay has confined itself to analyzing the sculpture at Qaṣr al- Ḥayr al-Gharbī from the particular viewpoint of the interplay between eastern and western elements in this program.107 Yet this is by no means
15
Hishām’s Balancing Act
115
the only way to look at it, and there is far more to be said about it than has yet appeared in print. For example, these sculptures demonstrate time and again that Umayyad art in general has one foot in Late Antiquity and another in early medieval times. The sculpture of a heavily laden branch from the façade of this palace (fig. 4.20), though only a fragment, can serve as a benchmark in that transition, which can be traced all over the Mediterranean world in various media and local styles between the fifth and eighth centuries.108 At the heart of this process is the move from naturalism to abstraction. Hence, for all the technical virtuosity and mastery of surface detail exhibited here, notably in the confident use of the drill, so that the densely packed foliage floats above a dark ground, the branch cannot be identified with a given species, and indeed it bears grapes and fruit or nuts simultaneously. Yet the sinuous rhythm and mastery of overlap suggest close observation. The material is stucco, perhaps molded and then hand-carved,109 not the stone usual in Syria—although some small-scale Roman stucco sculpture has survived in Palmyra.110 The use of stucco on a large scale suggests familiarity with late Sasanian art in Iraq and Iran. Trees of transcendental size and foliage occur frequently in Umayyad art (in Jerusalem, Damascus, ʿAnjar, and Khirbat al-Mafjar).111 They sometimes have otherworldly associations, but they also evoke the dramatic fertility of an oasis in a dry land. While the exact context of this piece, as that of so much of the sculpture from this site, is lost, it is sufficiently clear that this small masterpiece could have operated at several levels of meaning.
The Political Aspects of the Decorative Program So far the emphasis of this essay has been directed toward the classical and Byzantine elements in the architecture and decoration of Qaṣr al-Ḥayr al-Gharbī. The monument is not, however, aimed simply at Byzantium. For there is a larger political situation to consider. Murmurs of the insurrection that finally took shape as the Abbasid Revolution were already to be heard in the eastern provinces of the Islamic empire,112 and it seems that—with extraordinary though ultimately vain prescience?—Hishām (or his Master of Works) angled the images of Qaṣr al-Ḥayr al-Gharbī in that direction too. It would of course be a gross exaggeration, and indeed inherently absurd, to suggest that Hishām was trying, through the medium of his palatial architecture in Syria, to stem the current of the incipient Abbasid Revolution in far-off Khurasan, or that disaffected Hijazi
6 1
116
R obert H illenbrand
Arabs living in Merv could be placated by the use of Sasanian images in such buildings. Rather does this seem to be a case of an iconography that, while intentionally up-to-date and relevant in its own time, so accurately captures the Zeitgeist that it acquires an extra dimension of almost prophetic significance not originally intended by its creator. It is also worth remembering at this stage that, from the vantage point of Hishām himself, “east” and “west” did not necessarily represent irreconcilable extremes. He had his de facto capital at Rusafa, in eastern Syria, and the whole eastern part of the empire was ruled by his governors— Khālid al-Qaṣrī, and then Yūsuf ibn ʿUmar—from Ḥīra or Baṣra in lower Mesopotamia, while the erstwhile Sasanian capital of Ctesiphon, in central Iraq, was still closer to Syrian territory. East and West, therefore, were separated by little more than the width of the Euphrates. Such proximity would, of course, have facilitated the use of eastern modes at Qaṣr al-Ḥayr al-Gharbī while not detracting one whit from the intention of the overall design as an expression of cultural unity, of the symbiosis of east and west. Nor did this proximity impair the kind of sensitivity to what is eastern and what is western that is fundamental to the success of the whole design. One sees at work, therefore, an even-handed policy that evokes echoes of both east and west, while securely anchoring both traditions within a new Islamic context. In that sense there may be some wishful thinking at work, for the Byzantine heritage is placed on a par with that of Sasanian Iran and we are therefore invited to regard Islam as the successor of Byzantium just as much as of Persia. Is this a case, then, of sympathetic magic? In that case, the underlying assumption is that the living rival empire will share the fate of its dead counterpart. The ramifications of this theory extend well beyond the history of architecture, decoration, style, or even iconography and enter the troubled arena of Umayyad political history. As long ago as 1958, Sir Hamilton Gibb had argued, admittedly on somewhat limited evidence, that the reign of Hishām had witnessed an increasing concern on the part of the Umayyad government with the eastern provinces of the empire.113 The notion of a shift of emphasis from west to east in the reign of Hishām has since won general acceptance. The conclusions reached here, though provisional, sound a warning note against implicit faith in Gibb’s assessment. It is only right to attempt a detailed justification of this skepticism. The evidence to be cited will be taken from the visual arts and may therefore shed a novel light on problems usually debated on the basis of literary sources alone. And the visual arts, unlike the literary sources, are contemporary
17
Hishām’s Balancing Act
117
documents. And carefully read, they can shed light on knotty problems of the period. Beyond question, the pattern of evolution discernible in the first century of Islamic architecture does indeed reveal an inexorable and notably consistent movement from a western to an eastern orientation. The termini of this process—the Dome of the Rock and al-Manṣūr’s Baghdad— exemplify its extremes. One of them is overpoweringly Byzantine, the other correspondingly Persian.114 To discover how this came about, one may imagine a graph whose limits are marked by these two monuments. The various subsequent Umayyad buildings disposed along it would reveal a striking chronological consistency in their relative placing. It is not the intention here to dispute the absolute dating of Qaṣr al-Ḥayr al- Gharbī, but merely to draw attention to its central position on that imaginary graph. It should therefore not be lumped with the later and much more strongly orientalizing palaces of Khirbat al-Mafjar and Mshatta.115 It is entirely plausible—and also, as it happens, entirely consistent with the evolution outlined above—that the Umayyad preoccupation with eastern affairs did indeed gather momentum in the later years of Hishām’s reign. The value of Qaṣr al-Ḥayr al-Gharbī on the wider canvas of Umayyad history, however, is its evidence that in the early years of Hishām’s caliphate neither east nor west had priority—indeed, that there was a deliberate intention to keep a just balance between them. Such an interpretation accords gratifyingly well with the historical accounts of Umayyad military activity in Spain and France as well as central Asia in the 720s.116 Mention of Hishām’s even-handed policy117 prompts a closer look at how the impact of the decoration in this building is calculated. It is quickly apparent that a deliberate harmony and balance is maintained throughout. To prove this contention in the requisite detail would take another study in itself. Suffice it to say that this balance finds its clearest expression in the floor frescoes, where the great divide between east and west is asserted with iron logic not only in iconography but also in abstract decoration, visual conventions, and even brushwork technique.118 The dominance of Islam, too, may be inferred from this stark juxtaposition, for these frescoes are placed on the floor. Perhaps in Muslim eyes at least, then, they were subject to the ultimate indignity of being trodden underfoot.119 Yet such a humiliation would have been more apparent than real, for the frescoes are set in the wells of staircases that in all probability led up to the royal audience chamber, and would thus have been accessible to the eye rather than the foot.120 Nevertheless, their joint message is
8 1
118
R obert H illenbrand
clear enough: the subservient powers of Byzantium and Persia guard the approaches to the caliphal throne. Some echo of that exclusive club of the kings of the earth may be detected here.
Concluding Remarks Once noticed, this theme of east pitted against west, which is indeed a marked characteristic within the narrower field of Late Antique art,121 can be recognized almost everywhere in the palace. It creates a continuous dualism where neither triumphs. It announces itself from the outset in the contrasting upper and lower sections of the entrance gateway; it continues through the contrasting ruler sculptures in stucco; and it resurfaces in the contrasting floor frescoes, juxtaposed in perfect balance. It lurks even in the window grilles—classical and geometric versus Sasanian and organic122—or the wall frescoes, where the Sasanian rosette123 and sīmurgh124 yield to coarse plaster imitations of a sophisticated technique familiar in Byzantine churches,125 whereby quartered marble dadoes are so laid that the grain of the stone runs from one slab to the next.126 A Sasanian imperial image over the main gate seems to have been flanked by spandrels filled with mosaic, a Byzantine technique,127 while the Western classical imperial image (fig. 4.21), whose exact original emplacement is uncertain,128 was executed in the Persian medium of stucco. The parallelism could scarcely be more refined. Some of the women depicted in the sculpture bear the attributes of classical goddesses such as Atargatis,129 or perpetuate the distinctive classicism of Palmyra, complete with geometrically precise damp-fold drapery;130 others are recognizably associated with the Bacchic females found on Sasanian silverware, especially ewers, exhibiting the same perfunctory drapery style and frankly pneumatic charms.131 To a modern art historian, the cumulative effect of all these details is convincing evidence of the intentions of the person who designed this program; but, of course, contemporaries, deprived of the advantages of photography and a professionally trained eye, could be expected to register only a fraction of these subtleties. Such subtleties were nevertheless there. Thus they could have had some effect at least, even if the persons affected could not have analyzed their reactions with the requisite precision. And in this antithetical context, the layout of the gate itself must not be forgotten. As analyzed earlier, its lower half is decorated in a pervasively Sasanian idiom, while the architectural vocabulary of the upper half is
9 1
Hishām’s Balancing Act
119
basically classical. Nor is this an isolated refinement. Entrance to the palace is through a doorway taken intact from some Late Antique building132— there is no need to belabor here the psychological impact of using such spolia133—over which broods a Sasanian kingly image (fig. 4.21). The road to Islamic power, therefore, is in a very literal sense via Byzantium and Persia, a theme to be taken up again, as noted above, in the floor frescoes. Moreover, while it is possible to interpret the gateway as some gigantic crown, and thus a symbol of royal power, the strands that mingle to make that power effective are also tellingly present—for while the top and bottom of the gateway assert notions of military fortification, the central section conveys opulence and wealth. At its dead center is placed the double window from which the ruler, standing directly above his stucco Sasanian alter ego and surrounded by decoration of unmistakably classical type, would have been visible (perhaps alongside a royal prince) to his people gathered below. This suggests a ceremonial epiphany of a kind already familiar for many centuries in the Graeco-Roman world.134 The gateway of Qaṣr al-Ḥayr al-Gharbī, so delicately poised between competing cultures, is a political as well as an art-historical metaphor. It is an image that—to move the clock forward into recent times—recalls a compelling photograph of the desperately neutral King Dāʾūd of Afghanistan lighting an American cigarette with a Russian match. Neutrality, alas, rarely serves for long. The nimble balancing act of the caliph Hishām was a gallant attempt to hold together an empire bursting at the seams with centrifugal forces. But its success was precarious. The tide of history was already flowing against such an attempt. A generation earlier, the great Umayyad religious buildings—the Dome of the Rock and the Great Mosque of Damascus—owed most of their splendor to classical and Byzantine traditions. Under Hishām, as this essay has tried to show, eastern and western traditions enjoyed a brief period of equilibrium, a condition illuminatingly illustrated at Qaṣr al-Ḥayr al-Gharbī. A generation later, the building of Baghdad symbolized the end of all such accommodations. Thereafter, Islamic architects and craftsmen in the central Islamic lands turned east, not west, for inspiration.
Notes 1. This is a topic in itself; it was the subject of the nineteenth annual Sir Steven Runciman lecture, which I gave at King’s College, London, in February 2012. 2. Marsham 2009, 142.
201
120
R obert H illenbrand
3. Most of this work has been done by Denis Genequand. See especially Genequand 2006a, 63–83; Enciclopedia Archeologica, s.v. “Qasr al-Hayr al-Gharbi” (by D. Genequand); and Genequand 2004, 3–44. Qaṣr al-Ḥayr al-Gharbī also figures repeatedly in his most recent major work of synthesis: Genequand 2013, esp. 166 and 168. 4. See, for example, Fowden 2004a, 282–304. 5. Meanwhile, see Talgam 2004, 1:63–66. 6. Talgam 2004, 1:69–72. 7. For detailed remarks on this topic, see Hillenbrand 1973, 169–176; and, in the particular context of ruler images, Fowden 2004b, 121–123. 8. For a general survey of this element in the particular context of Qaṣr al-Ḥayr al- Gharbī, see Talgam 2004, 1:94–97. 9. This term seems preferable in the present context to the much-abused word “influence”; Michael Baxandall (1985, 58–62) has drawn attention to the cavalier and lazy misuses of this word in art-historical literature. 10. Schlumberger 1939, 332–336. 11. For a general account of this type of building in what was to be Islamic territory, see McKay 1975, 210–237. For a color illustration of a surviving example—the House of the Vettii at Pompeii—see Wheeler 1964, fig. 106. 12. For a general discussion of this process, see Genequand 2006b, 3–25, esp. 25. 13. Grabar et al. 1978, 1:21–22, 29 and 2:9, 82, 96. 14. For a detailed analysis of this gate, see Hillenbrand 1973, 169–171, 176. 15. Kautzsch 1936, 2–3, 234–235, and elsewhere. The history of the acanthus capital is a striking illustration of the point (Kautzsch 1936, 115–152, 213–215). For a discussion of new developments in capital design in the immediately pre- Islamic period, see Maguire 1987, 353, 361. 16. Levi 1947, II: pl. XXXVIa, CIIa-e and CIIIa, d and f. 17. As, for example, at Sta. Sabina in Rome; see, for a color illustration, Grabar 1966a, 11, fig. 7, or at Ostia (Brenk 1977, fig. 40 a, b) (in color). 18. Cf. a Palmyrene version of this theme: Colledge 1976, pl. 97. For the significance of the motif, see Winkes 1969, especially pl. 61–96 for a representative selection. 19. These sculptures, like the smaller frescoes, await full publication by a member of staff of the Damascus museum, though some have already been briefly published in various catalogues and elsewhere. I am most grateful to Dr. Afif al-Bahnassi, the Director of the Damascus Museum, and to his colleagues Dr. Kassem Toueir and Mr. Muhammad al-Kholi, for having facilitated my work on the Umayyad material under their care, and for having permitted me to photograph it. I also acknowledge with thanks a grant from the British Academy, which made my research in Syria possible. The discussion that follows is based on my own notes and photographs.
2 11
Hishām’s Balancing Act
121
20. As in a Sogdian ossuary from Central Asia; for an illustration, see Ghirshman 1962, 323, fig. 434. For an example nearer home, see the Umayyad bronze and iron brazier from al-Fudayn, Mafraq, Jordan; see Bienkowski 1991, 99 (color plate); Ballian 2012, 212–213, with further references. See also Naghawy 2015. 21. Such as Panjikent; see Belenizki 1980, fig. on p. 210; though pl. 68 on pp. 152– 153 shows that more sophisticated variants of this theme were also known at the same time and place. 22. Schlumberger 1986, pl. 68 d and e. 23. Ghirshman 1962, 468, fig. 347. 24. For an analysis of these three images, see Hillenbrand 1973, 173–174, illustrated on pl. 45a-b and 46. 25. For these images in color, see Mouliérac 1993, 416–417, 413. 26. Talgam 2004, 2: fig.79. 27. Schlumberger 1986, pl. 68d–e, 67a, 70e, 70f, 67e, 70b and 70d, respectively. 28. For a color image, see El-Kholi 1990, 39. 29. For these images in color, see Mouliérac 1993, 414–415. 30. For a color plate, see El-Kholi 1995, 78. 31. Hillenbrand 1973, 171–172 and pl. 44. 32. Schlumberger 1986, pl. 64e and 67b, c and d, respectively. 33. Schlumberger 1939, 1:pl. XLVI/2, pl. XLVI/I1 and fig. 21. 34. Schlumberger 1939, 350. 35. Glueck 1940, 184–194, and figs. 116–117, 119–120, and 122. 36. Talgam 2004, 1:20–21. 37. Hamilton 1969, 62–63 and pl. XVII. 38. Talgam 2004, 1:23. 39. Many of these themes are already well developed at Persepolis (Roaf 1983, figs. 53, 61, 111, 114, 116 and 119 and pl. XIXa. XXXIVa–b and XLVIIa). 40. For analyses of this key image, see Schlumberger 1939, 1:328–329 and pl. XLV/ 3; Brisch 1973, 182–183; and Talgam 2004, 1:22. 41. Hillenbrand 1973, 163–164 and fig. 38. Schlumberger (1986, pl. 81d and e) suggests that two life-size sculptures, one of a warrior and the other depicting a woman in sensitively executed damp-fold drapery, originally flanked this image. 42. Der Nersessian 1965; Otto-Dorn 1961, 1–69; Jones 2007, 54–65, with the important suggestion that this frieze depicts the lifestyle of the Armenian king Gagik rather than that of the Abbasid caliph. 43. Brenk 2010; Dittelbach 2011; especially Anzelmo 2005; and Johns 2015, 59–89. 44. Grabar 1936, 63–70, figs. 1–8; for a discussion, see 70–74. 45. I am most grateful to Nadia Ali for this suggestion: see also Ali and Guidetti (chap. 7 below); Ali 2008; and Ali 2012, 17–18. 46. Hanfmann 1952.
221
122
R obert H illenbrand
47. El-Kholi 1990, 26. See also ʿAbd al-Ḥaqq 1951, 5–57, with a French summary on 129–133. The most detailed visual presentation of these claustra to date is in Schlumberger 1986, pl. 74–81a and c. But the discussion of them (Schlumberger 1986, 15–16) is skeletal. 48. Franz 1956, 468–472, with an attempt to devise categories for the lunette surround: either a simple rounded arch with an ornamental extrados or an elaborately architectonic rectangular framework carried on engaged columns and crowned with a dwarf arcade. He notes the distinctively Sasanian character of the second type (Franz 1956, 465). 49. Franz 1956, 472–482. See also Brisch 1966, 42 and pl. 47a, 53a and b, 58a, 60a and b, 71b and 72b. 50. Levi 1947, II: passim; Allen 1988, 1–17. 51. Herzfeld 1920, 117, 119, and pl. XXXVIII– XLI, LIV, LVII and LIX– LX; cf. Hillenbrand 1973, 175–176, and fig. 47a, b, and c. For the impetus to invention triggered by the move from stone to stucco, see Hamilton 1947, 55–56; 1953, 53–55. 52. Franz (1956, 469) noted the balance between Sasanian and Syro-Byzantine traditions in these lunettes. 53. Creswell 1969, 1/1: 202–4, figs. 92, 118, and 127 and fig. 59a–d. 54. See Grube 1968, 11–12; Jones and Michell 1976, 316; and Meinecke 1982, 272 and color plate on 232. 55. Meinecke 1982, 272 and color plate on p. 232, with the suggestion that the fresco depicts prisoners. 56. Bieber 1961, passim, but especially figs. 551–564 and 815–831. 57. Trilling 1989, fig. 35 (man kicked by mule). 58. See Grabar in Hamilton 1959, 310 and fig. XCVIIa. 59. Esin 1967, color fig. 7; fig. XXII, figs. 4 and 5; and fig. XXXI, fig. 3. See also Bussagli 1963, color plate on p. 103, for a particularly telling parallel from Bezeklik, conveniently dated to the eighth century. 60. Förtsch 1993, especially 188–192, citing examples from Rome (Casa di Livia and the villa under the Farnesina), Sabratha, Boscoreale, and, above all, Pompeii (especially Casa del Criptoportico, Casa di Sulpicius Rufus, Casa della Fontana piccola, and Casa di Lucrecius Fronto), all of which he illustrates (pl. 45–50). See also McKay 1975, pl. 60–63. 61. Crowfoot 1941, fig. XVa, b; Grabar 1966a, 79, pl. 82–83. 62. Levi 1947, II: pl. XXXVIIa, XLIIa, CLXXVIIIc-d, and CLXXXId. 63. Such as the sixth-century floor mosaic from a Christian basilica at Delphi, or another in the Heptapegon at al-Tabgha on the Sea of Galilee (Crowfoot 1941, fig. XII). 64. Schlumberger 1939, 1: pl. XXXVIII/3 and XXXIX/4. 65. Schlumberger 1946–1948, 86–102. 66. Among the thousands of square meters of floor decoration encountered at Antioch, not one floor fresco was found. 67. Encyclopaedia of Islam, 2nd ed., s.v. “Hishām” (by F. Gabrieli); Gabrieli 1935, 135, 137; Wellhausen 1963, 348–349; Blankinship 1994, 79, 227, 302.
213
Hishām’s Balancing Act
123
68. Strika 1968, 146–147. El-Kholi suggests that the game park and the agricultural estate adjoined each other (1990, 27). 69. E.g., Vanden Berghe 1983, pl. 23 and 40. 70. Schlumberger 1946– 1948, fig. B opposite p. 96; Ettinghausen 1962, plate on p. 37. 71. Schlumberger 1946– 1948, fig. A opposite p. 88; Grube 1967, 18, fig. 3; Ettinghausen 1962, plate on p. 35. 72. Grabar 1966a, color plate on 323. 73. Levi 1947, I:232, 346, 356, 365, and 577–579 and II: pl. LXXXIb, CLXIXa, LXXXIVd, and XCd, presents a whole range of Gaea images from Antioch (and elsewhere, e.g., the Gaea at Bait Jibrin, Levi 1947, I:579, fig. 213). For a detailed account of the classical iconography of Gaea—an account insufficiently exploited in earlier discussions of the floor fresco at Qaṣr al-Ḥayr al-Gharbī—see Levi 1947, I:263–269. 74. Note the birds in the lower part of the fresco. 75. Thus the floor mosaic in the villa of a wealthy Antiochene Christian of the sixth century could feature in a prominent position a roundel of the personified attribute Magnanimity—megalopsychia (Grabar 1966a, 107, fig. 109 [in color]). Nevertheless, it must be acknowledged that Megalopsychia is a virtue personified, not a goddess, and the image would therefore be unlikely to attract opprobrium from pious Christians. Gaea, as a goddess that was once widely venerated, would have been an altogether different proposition. At the very least, however, the Megalopsychia mosaic proves that Christian Syria had inherited from the Graeco-Roman world the custom of placing symbolic portraits within roundels and using them as the centerpiece of floor mosaics. The ancestry of the Gaea image at Qaṣr al-Ḥayr al-Gharbī is therefore sufficiently assured. 76. Grabar 1954, 185–187. For the fullest discussion of this theme to date, see Fowden 2004b, 197–226. 77. Compare the sea creatures depicted on roof tiles from Dura Europos (pl. 36): Chi and Heath 2011, 50–51 and pl. 2–12 and 2–17; and, more distantly, the tritons found on marine mosaics from Timgad (Dunbabin 1978, color plate G) and from Ostia (Strong 1976, pl. 119). 78. Wellhausen 1963, 332. Hishām was by no means alone in this predilection; his representative in Iraq, the governor Khālid al-Qasrī, spent some 15 years on such projects (I am grateful to Lawrence Conrad for this information). 79. Schlumberger 1939, 201–202 (he identifies it as Roman); Genequand 2006a, 66– 69, makes a convincing argument for an Umayyad date. 80. Ettinghausen 1972, 21–24, 35–36. 81. Gaea is, incidentally, not the only goddess depicted in the decoration of Qaṣr al- Ḥayr al-Gharbī. A female bust bearing the attribute of the goddess Atargatis, a dove, occurs on the main façade (Schlumberger 1939, 349, fig. 21; discussed by M. Meinecke and better illustrated in Kohlmeyer and Strommenger 1973, 268 and 269, respectively).
241
124
R obert H illenbrand
82. Even today they are not entirely extirpated; see Fedden 1965, 71–72; and Haas 1982, 333 and 337. 83. And in similar fashion he could have explained away the luxurious lifestyle evoked in the other floor fresco, which could also have attracted censure. 84. Dunbabin 1978, color pl. H and pl. 3, 13, 30, 68, 98, 143–145, 147–154, 159, 165, 175–178, and 181–183. 85. Note, for example, the Good Shepherd in the floor mosaic of the Basilica at Aquileia (Grabar, 1966b, 29, fig. 27 [in color]) or the bust of Christ at Hinton St. Mary, Dorset, now in the British Museum (Neal, 1981, 87–89 and color plate 61 following p. 104). 86. Grabar 1973, 161. 87. Herzfeld 1911, 60–62; Becker 1924, 207. 88. For a compendium of the earlier literature on Mshatta, see Creswell 1969, 1(2):604–606 and 623–638; Hillenbrand 1981, 63–86; Enderlein, Meinecke, and Blower 2010, 309–372. 89. Ettinghausen 1972, 17–65. 90. Combe, Sauvaget, and Wiet 1931, 23, no. 27. 91. Vasiliev 1956, 25–47. According to al-Maqrizi—though he is a very late source indeed—Hisham revoked this edict (Vasiliev 1956, 39). 92. Kennedy 2001, 98, 105–108. 93. al-Ṭabarī 1885–1889, 1733, line 11. 94. al-Iṣfahānī 1868, 4:125, lines 12–30; but cf. 4:121, lines 12–14, where the caliph responsible is al-Walīd ibn Yazīd. 95. Gabrieli 1935, 135. 96. Ibn Khaldūn 1967, 1:417. When he was governor of Syria he was reproached by the caliph ʿUmar for his unseemly display of Persian royal manners (kisrāwiyya); he excused himself by saying that he had no choice but to do so, since “I am in a border region facing the enemy.” 97. Grabar 1959, 53–55. 98. Coincidentally it is in the two Qaṣr al-Ḥayr sites—east and west—that partial exceptions to this rule may be discovered. The khān at Qaṣr al-Ḥayr al-Gharbī (though not the palace itself) bears a building inscription in the name of one Thābit ibn Abī Thābit (see note 90 above), while the inscription at Qaṣr al-Ḥayr al-Sharqī, removed by Rousseau in 1808 and subsequently lost, was sited in the mosque. It mentions that the city was constructed in 110/728 at the order of the caliph Hishām by one Sulaymān ibn ʿUbayd (Combe, Sauvaget, and Wiet 1931, 1:23–24, no. 28). 99. An article by Peters (1983, 119–138) does not entirely fulfill the promise of its title, in that it does not address the question of who actually designed the building. Peters does, however, make the interesting suggestion that Muʿāwiya I might have been the moving spirit behind the erection of the Dome of the Rock (1983, 133). Grabar treats this as a possibility (2006, 62). As for the identity
215
Hishām’s Balancing Act
125
of those who actually designed, constructed, and decorated the Dome of the Rock, this remains a mystery; the men entrusted with the project by ʿAbd al- Malik, Rajā ibn Ḥaywa al-Kindī and Yazīd ibn Sallām, supervised the project and administered the budget (Rabbat 1993, 68–69). 100. See, however, the article by Ali and Guidetti in this volume. 101. Cf. the review by Rogers (1973–1974, 155) of Grabar’s book. Sheila Blair has revealed hitherto unrecognized associations between the inscriptions and the decoration (Blair 1992, 70–85). 102. Grabar 1973, 177–178, argues for a koine of early medieval princely art that transcended the barriers of culture and religion and was concerned above all to express a luxurious lifestyle. 103. Hamilton 1969, 336. 104. Ettinghausen 1972, 63–65. 105. Cf. the disparaging remarks of Rogers on such a possibility (1973– 1974, 158–159). 106. As appears to have been done at Mshatta—see Kühnel 1965, 132–133 and 145– 146, who suggests the activity of at least four master carvers, each with numerous workmen at his disposal. The matter is dealt with in much greater detail in Herzfeld 1911, 135–143. Cf. too the remarks by Ettinghausen (1972, 63) à propos of Khirbat al-Mafjar: “the basic Byzantine and Sasanian elements co- exist here as ‘equal but separate’ entities; there seems thus to be no true intermingling of the two strains, only a skilful coordination.” 107. On this theme, see in particular Talgam 2004, 1:19–30, 45–46, 63–67, 94–97, and 2:figs. 62–79 and 110. Hers is the fullest attempt so far to analyze this body of sculpture. 108. For a still classic analysis of this process, see Kitzinger 1940. 109. Jones and Michell 1976, 298. 110. Colledge 1976, 104 and pl. 135; Kohlmeyer and Strommenger 1982, 206–209, nos. 186–191 (K. Parlasca); and for good color images of seven examples, Cluzan, Moulierac, and Bounni 19993, 308–309 (K. El-Assad). 111. For the Jerusalem examples, see Grabar and Nuseibeh 1996; for the examples in the Great Mosque of Damascus, Degeorge 2010; and for Khirbat al-Mafjar, Hamilton 1947, color frontispiece. Cf. too a capital from Anjar: Finster 2007, fig. 54. 112. Kennedy 2005, 1–10. 113. Gibb 1958, 232–233, reprinted in 1962, 59–60. 114. The point is made in telling detail in Creswell 1962, 2:18–23. 115. Northedge 1992, 100–104. 116. Blankinship 1994, 176–185, 196–197, 230–236. 117. For comments on the context of Hishām’s reign and his ideology, see Blankinship 1994, 4–9, 96–97. 118. Hillenbrand 1973, 172–173.
261
126
R obert H illenbrand
119. See, however, note 83 above. For the view that the location of an important image on the floor does imply humiliation, see Ettinghausen 1972, 24. 120. Similarly, in Christian buildings sacred images placed on the floor were so located that they were safe from being accidentally stepped on, while the Edict of 427 AD forbade the sign of Christ on the floor. See Kitzinger 2002, 250, 258; originally published in 1971, 640–641. 121. Förtsch 1993, 180. 122. Cf. Abb. 1 and 3 with Abb. 2 and 4 in Franz 1956, 466–467. 123. Schlumberger 1939, fig. XXXIX/4. 124. Schlumberger 1939, 352 and fig. 26 on 355. 125. E.g., at San Vitale, Ravenna (Lassus 1967, 56–57, fig. 41 [in color]). 126. Schlumberger 1939, pl. XXXVIII/3 and XXXIX/4. It seems unlikely, however, that the Muslims attributed to such decoration the complex programmatic meaning that it seems to have held for connoisseurs in Late Antique times (Onians 1980, 1–24). 127. Schlumberger 1939, 328. 128. Genequand (2013, 166) places it in the upper part of the eastern façade above the entrance, a location that sits ill with the presence of a double arcade; Talgam (2004, 2:53) more plausibly locates it (though without suggesting a compass point) on “the façade facing the courtyard.” Grabar (1987, fig. 80) identifies it as “from the courtyard.” For an analysis of this image, see Hillenbrand 1973, 163–164. 129. Schlumberger 1939, 349, fig. 21. Cf. Grabar 1987, pl. 84. 130. Grabar 1987, pl. XLVI/2 and XLVII/3. 131. Cf. Spuler and Sourdel-Thomine 1973, fig. 44; Ghirshman 1962, 215–217, figs. 256–258; Ettinghausen 1967–1968, pl. 1–7, 9–10, 13–15, and 17. 132. Schlumberger 1939, 225–226 and fig. XXXVII/1. 133. Yet even here the Umayyad craftsman left his mark, for the weathered base of one of the jambs has been repaired in stucco (Schlumberger 1939, 226 and pl. XXXIII/2). 134. Baldwin Smith 1978, 22, 182, and 200.
Bibliography ʿAbd al- Ḥaqq, S. 1951. “Iʿādat tashyīd janāḥ Qaṣr al- Ḥayr al- Gharbī fī matḥaf Dimashq.” Les Annales Archéologiques de Syrie 1: 5–57. Ali, N. 2008. “Processus de fabrication de l’iconographie omeyyade: L’exemple des décors palatiaux du Bilād ash-Shām.” Unpublished diss., Université Aix Marseille. ———. 2012. “Pour une étude globale de l’iconographie des omeyyades de Syrie et d’Espagne (VIII–XIe): Le cas des calendriers et des thèmes agricoles.” Anales de Historia del Arte 22: 9–26.
217
Hishām’s Balancing Act
127
Allen, T. 1988. Five Essays on Islamic Art. Sevastopol, CA: Solipsist. Anzelmo, F. M. 2005. “Classification of the Decorated Garments and Headdresses in the Ceiling Paintings of the Cappella Palatina.” In The Painted Ceilings of the Cappella Palatina (Islamic Art, Supplement I), edited by E. J. Grube and J. Johns, 499–506. Genoa: Bruschettini Foundation for Islamic and Asian Art. Assad, K. 1993. “La Syrie à l’époque hellénistique et romaine.” In Syrie: Mémoire et civilisation, edited by S. Cluzan, J. Mouliérac, and A. Bounni, 303– 313. Paris: Institut du Monde Arabe. Baldwin Smith, E. 1978. Architectural Symbolism of Imperial Rome and the Middle Ages. Repr. New York: Hacker Art Books. Ballian, A. 2012. “Al-Fudayn.” In Byzantium and Islam: Age of Transition 7th-9th Century, edited by H. C. Evans and B. Ratliff, 212–215. New York: Metropolitan Museum of Art. Baxandall, M. 1985. Patterns of Intention: On the Historical Explanation of Pictures. New Haven, CT: Yale University Press. Becker, C. H. 1924. Vom Werden und Wesen der islamischen Welt. Islamstudien 2. Leipzig: Quelle and Meyer. Belenizki, A. M. 1980. Mittelasien: Kunst der Sogden. Translated by L. Schirmer. Leipzig: Seemann. Bieber, M. 1961. The History of the Greek and Roman Theatre. Princeton, NJ: Princeton University Press. Bienkowski, P., ed. 1991. Treasures from an Ancient Land: The Art of Jordan. Stroud: A. Sutton. Blair, S. 1992. “What is the Date of the Dome of the Rock?” In Bayt al-Maqdis: ʿAbd al-Malik’s Jerusalem. Part One, edited by Julian Raby and Jeremy Johns, 59–87. Oxford: Oxford University Press. Blankinship, K.Y. 1994. The End of the Jihad State: The Reign of Hisham Ibn ‘Abd al- Malik and the Collapse of the Umayyads. Albany: State University of New York Press. Brenk, B. 1977. Spätantike und frühes Christentum. Berlin: Propyläen Verlag. ———, ed. 2010. La Cappella Palatina a Palermo. 4 vols. Modena: Franco Cosimo Panini Editore. Brisch, K. 1966. Die Fenstergitter und verwandte Ornamente der Hauptmoschee von Córdoba: Eine Untersuchung zur spanisch- islamischen Ornamentik. Berlin: De Gruyter. ———. 1973. “Omayyaden.” In Die Kunst des Islam, edited by B. Spuler and J. Sourdel-Thomine, 139–145, 171, 177–186. Berlin: Propyläen Verlag. Bussagli, M. 1963. Painting of Central Asia. Translated by L. Small. Geneva: Skira. Chi, J. Y., and S. Heath, eds. 2011. Edge of Empires: Pagans, Jews, and Christians at Roman Dura-Europos. New York: New York University Press. Cluzan, S., J. Mouliérac, and A. Bounni, eds. 1993. Syrie: Mémoire et civilisation. Paris: Institut du Monde Arabe.
281
128
R obert H illenbrand
Colledge, M. A. R. 1976. The Art of Palmyra. London: Thames and Hudson. Combe, E., J. Sauvaget, and G. Wiet. 1931. Répertoire chronologique d’épigraphie arabe I. Cairo: Imprimerie de l’Institut français d’Archéologie Orientale. Creswell, K. A. C. 1969. Early Muslim Architecture: Umayyads. a.d. 622–750. Oxford: Oxford University Press. Crowfoot, J. W. 1941. Early Churches in Palestine. London: Oxford University Press. Degeorge, G. 2010. La Grande Mosquée des Omeyyades, Damas. Paris: Imprimerie Nationale. Der Nersessian, S. 1965. Aght’amar: Church of the Holy Cross. Cambridge, MA: Harvard University Press. Dittelbach, T., ed. 2011. Die Cappella Palatina in Palermo: Geschichte, Kunst, Funktionen: Forschungsergebnisse der Restaurierung. Künzelsau: Swiridoff Verlag. Dunbabin, K. M. D. 1978. The Mosaics of Roman North Africa: Studies in Iconography and Patronage. Oxford: Oxford University Press. El-Kholi, M. 1990. Châteaux omayyades de Syrie: Collections du Musée National de Damas, 16 septembre 1990–17 mars 1991. Paris: Institut du Monde Arabe. ———. 1995. “Le jardin aux époques omeyyade et abbasside.” In Syrie: Un patrimoine inédit, edited by H. Chidiac, 76–83. Paris: Institut du Monde Arabe. Enderlein, V., M. Meinecke, and J. Blower. 2010. “Excavation— Investigation— Presentation: Problems of Representing Past Cultures in the Example of the Mshatta Façade.” Art in Translation 2(3): 309–372. Esin, E. 1967. Antecedents and Development of Buddhist and Manichaean Turkish Art in Eastern Turkestan and Kansu. Istanbul: Milli Egitim Basimevi. Ettinghausen, R. 1962. Arab Painting. Geneva: Skira. ———. 1967–1968. “A Persian Treasure.” Arts in Virginia 8(1-2): 28–41. ———. 1972. From Byzantium to Sasanian Iran and the Islamic World: Three Modes of Artistic Influence. Leiden: Brill. Fedden, R. 1965. Syria and Lebanon. London: John Murray. Finster, B. 2007. “Researches in ‘Anjar. II: Preliminary Report on the Ornaments of ‘Anjar.” Bulletin d’Archéologie et d’Architecture Libanaises (BAAL) 11: 143–165. Förtsch, R. 1993. “Die Architekturdarstellungen der Umaiyadenmoschee von Damaskus und die Rolle ihrer antiken Vorbilder.” Damaszener Mitteilungen 7: 177–212. Fowden, G. 2004a. “Late Antique Art in Syria and its Umayyad Evolutions.” Journal of Roman Archaeology 17: 282–304. ———. 2004b. Qusayr ‘Amra: Art and the Umayyad Elite in Late Antique Syria. Berkeley: University of California Press. Franz, H. G. 1956. “Die Stuckfenster im Qasr al-Hair al-Gharbi.” Wissenschaftliche Annalen 5: 465–483. Gabrieli, F. 1935. Il Califfato di Hisham: Studia di storia omeyyade. Alexandria: Société de publications égyptiennes.
291
Hishām’s Balancing Act
129
Genequand, D. 2004. “Châteaux omeyyades de Palmyrène.” Annales Islamologiques 38: 3–44. ———. 2006a. “Some Thoughts on Qasr al-Hayr al-Gharbi, Its Dam, Its Monastery and the Ghassanids.” Levant 38: 63–84. ———. 2006b. “Umayyad Castles: The Shift from Late Antique Military Architecture to Early Islamic Palatial Building.” In Muslim Military Architecture in Greater Syria: From the Coming of Islam to the Ottoman Period, edited by H. Kennedy, 3–25. Leiden: Brill. ———. 2012. Les établissements des élites omeyyades en Palmyrène et au Proche-Orient. Beirut: Institut Français du Proche- Orient, Bibliothèque archéologique et historique. Ghirshman, R. 1962. Iran: Parthians and Sassanians. Translated by S. Gilbert and J. Emmons. London: Thames and Hudson. Gibb, H. A. R. 1958. “Arab-Byzantine Relations under the Umayyad Caliphate.” Dumbarton Oaks Papers 12: 221–233. ———. 1962. “Arab-Byzantine Relations under the Umayyad Caliphate.” In Studies on the Civilization of Islam, edited by J. S. Shaw and W. R. Polk, 47–61. Princeton, NJ: Princeton University Press. Glueck, N. 1940. The Other Side of the Jordan. New Haven, CT: American Schools of Oriental Research. Grabar, A. 1936. L’empereur dans l’art byzantin: Recherches sur l’art officiel de l’empire d’Orient. Strasbourg: Publications de la Faculté des lettres de l’Université de Strasbourg. ———. 1966a. Byzantium: From the Death of Theodosius to the Rise of Islam. Translated by S. Gilbert and J. Emmons. London: Thames and Hudson. ———. 1966b. The Beginnings of Christian Art: 200–395. Translated by S. Gilbert and J. Emmons. London: Thames and Hudson. Grabar, O. 1954. “The Painting of the Six Kings at Qusayr ʿAmrah.” Ars Orientalis 1: 185–187. ———. 1959. “The Umayyad Dome of the Rock in Jerusalem.” Ars Orientalis 3: 33–62. ———. 1973. The Formation of Islamic Art. London: Yale University Press. ———. 1987. The Formation of Islamic Art. 2nd ed. London: Yale University Press. ———. 2006. The Dome of the Rock. Cambridge, MA: Belknap Press of Harvard University Press. Grabar, O., R. Holod, J. Knustad, and W. Trousdale, eds. 1978. City in the Desert: Qasr al-Hayr East. Cambridge MA: Harvard University Press. Grabar, O., and S. Nuseibeh. 1996. The Dome of the Rock. New York: Rizzoli. Grube, E. J. 1967. The World of Islam. London: Hamlyn. ———. 1968. The Classical Style in Islamic Painting: The Early School of Herat and Its Impact on Islamic Painting of the Later 15th, the 16th and 17th Centuries. Venice: Edizioni Oriens.
301
130
R obert H illenbrand
Haas, V. 1982. “Götter, Kulte und Mythen.” In Land des Baal: Syrien, Forum der Völker und Kulturen, edited by K. Kohlmeyer and E. Strommenger, 330–337. Mainz: P. von Zabern. Hamilton, R. W. 1947. “Plaster Balustrades from Khirbat al Mafjar.” Quarterly of the Department of Antiquities in Palestine 13: 1–58. ———. 1953. “Carved Plaster in Umayyad Architecture.” Iraq 15(1): 43–55. ———. 1959 (with a contribution by O. Grabar). Khirbat al Mafjar: An Arabian Mansion in the Jordan Valley. Oxford: Oxford University Press. ———. 1969. “Who Built Khirbat al-Mafjar?” Levant 1: 61–67. Hanfmann, G. M. A. 1952. The Season Sarcophagus in Dumbarton Oaks. 2 vols. Cambridge, MA: Harvard University Press. Herzfeld, E. 1911. “Die Genesis der islamischen Kunst und das Mshatta-Problem.” Der Islam II: 105–144. ———. 1920. Am Tor von Asien: Felsdenkmäler aus Irans Heldenzeit. Berlin: D. Reimer. Hillenbrand, R. 1973. “Omayyaden.” In Die Kunst des Islam, edited by B. Spuler and J. Sourdel-Thomine, 145–177. Berlin: Propyläen Verlag. ———. 1981. “Islamic Art at the Crossroads: East versus West at Mshatta.” In Essays in Islamic Art and Architecture: In Honor of Katharina Otto-Dorn, edited by A. Daneshvari, 63–86. Malibu, CA: Undena. Ibn Khaldūn. 1967. The Muqaddimah: An Introduction to History. 2nd ed. Translated by F. Rosenthal. London: Routledge and Kegan Paul. al-Iṣfahānī. 1868. Kitāb al-aghānī. 20 vols. Cairo: Bulaq. Johns, J. 2015. “Muslim Artists and Christian Models in the Painted Ceilings of the Cappella Palatina.” In Romanesque and the Mediterranean: Patterns of Exchange across the Latin, Greek and Islamic Worlds c. 1000–c. 1250, edited by R. Bacile and J. McNeill, 59–89. Leeds: Maney. Jones, D., and G. Michell, eds. 1976. The Arts of Islam. London: Arts Council of Great Britain. Jones, L. 2007. Between Islam and Byzantium: Aght’amar and the Visual Construction of Medieval Armenian Rulership. Aldershot: Ashgate. Kautzsch, R. 1936. Kapitellstudien: Beiträge zu einer Geschichte des spätantiken Kapitells im Osten vom vierten bis ins siebente Jahrhundert. Berlin: W. de Gruyter and Co. Kennedy, H. 2001. The Armies of the Caliphs: Military and Society in the Early Islamic State. London: Routledge. ———. 2005. When Baghdad Ruled the Muslim World: The Rise and Fall of Islam’s Greatest Dynasty. Cambridge, MA: Da Capo. Kitzinger, E. 1940. Early Medieval Art in the British Museum. London: British Museum. ———. 1970. “The Threshold of the Holy Shrine: Observations on Floor Mosaics at Antioch and Bethlehem.” In Kyriakon: Festschrift für Johannes Quasten, edited by P. Granfield and J. A. Jungmann, 639–647. Münster: Aschendorff.
131
Hishām’s Balancing Act
131
———. 2002. “The Threshold of the Holy Shrine: Observations on Floor Mosaics at Antioch and Bethlehem.” In E. Kitzinger, Studies in Late Antique, Byzantine and Medieval Western Art, I, 244–259. London: Pindar. Kohlmeyer, K., and E. Strommenger, eds. 1982. Land des Baal: Syrien, Forum der Völker und Kulturen. Mainz: P. von Zabern. Kühnel, E. 1965. “Some Notes on the Façade of Mshatta.” In Studies in Islamic Art and Architecture in Honour of Professor K. A. C. Creswell, 132–146. Cairo: American University in Cairo Press. Lassus, J. 1967. The Early Christian and Byzantine World. London: Hamlyn. Levi, D. 1947. Antioch Mosaic Pavements. 2 vols. Princeton, NJ: Princeton University Press. Maguire, E. D. 1987. “Range and Repertory in Capital Design.” Dumbarton Oaks Papers 41: 351–361. Marsham, A. 2009. Rituals of Islamic Monarchy: Accession and Succession in the First Muslim Empire. Edinburgh: Edinburgh University Press. McKay, A. G. 1975. Houses, Villas and Palaces in the Roman World. Ithaca, NY: Cornell University Press. Meinecke, M. 1982. “Islamische Zeit.” In Land des Baal: Syrien, Forum der Völker und Kulturen, edited by K. Kohlmeyer and E. Strommenger, 254–291. Mainz: P. von Zabern. Naghawy, A. 2015. “Brazier.” In “Discover Islamic Art.” Museum with No Frontiers. http://www.discoverislamicart.org/database_item.php?id=object;ISL;jo;Mus01; 6;en&cp (accessed 30 September 2015). Neal, D. S. 1981. Roman Mosaics in Britain: An Introduction to Their Schemes and a Catalogue of Paintings. London: A. Sutton. Northedge, A. 1992. Studies on Roman and Islamic ‘Amman: The Excavations of Mrs. C.-M. Bennett and Other Investigations. Oxford: Oxford University Press. Onians, J. B. 1980. “Abstraction and Imagination in Late Antiquity.” Art History 3(1): 1–24. Otto- Dorn. K. 1961. “Turkisch- islamisches Bildgut in den Figurenreliefs von Achthamar.” Anatolia 4: 1–69. Parlasca, K. 1982. “Römische Zeit.” In Land des Baal: Syrien, Forum der Völker und Kulturen, edited by K. Kohlmeyer and E. Strommenger, 186–226. Mainz: P. von Zabern. Peters, F. E. 1983. “Who Built the Dome of the Rock?” Graeco-Arabica: First International Congress on Greek and Arabic Studies 2: 119–138. Rabbat, N. 1993. “The Dome of the Rock Revisited: Some Remarks on al-Wasiti’s Accounts.” Muqarnas 10: 67–75. Roaf, M. 1983. “Sculptures and sculptors at Persepolis.” Iran 21: 1–164. Rogers, J. M. 1973–1974. Review of The Formation of Islamic Art, by O. Grabar. Kunst des Orients 9(1–2): 153–166.
321
132
R obert H illenbrand
Schlumberger, D. 1939. “Les fouilles de Qasr al-Hair al-Gharbi (1936-1938): Rapport préliminaire.” Syria 20(4): 324–373. ———. 1946–1948. “Deux fresques omeyyades.” Syria 25: 86–102. ———. 1986. Qasr el-Heir el-Gharbi. Paris: Librairie Orientaliste P. Geuthner. Spuler, B., and J. Sourdel-Thomine. 1973. Die Kunst des Islam. Berlin: Propyläen Verlag. Strika, V. 1968. “Sull’interpretazione di ḥayr nel periodo ommiade.” Annali dell’Istituto orientale di Napoli, N.S. 18(2): 139–149. Strong, D. 1976. Roman Art. Harmondsworth: Penguin. al-Ṭabarī. 1885–1889. Tārīkh al-rusul wa’l-mulūk. Secunda Series, III. Edited by I. Guidi, D. H. Müller, and M. J. Goeje. Leiden: Brill. Talgam, R. 2004. The Stylistic Origins of Umayyad Sculpture and Architectural Decoration. 2 vols. Wiesbaden: Harrassowitz. Trilling, J. 1989. “The Soul of the Empire: Style and Meaning in the Mosaic Pavement of the Byzantine Imperial Palace in Constantinople.” Dumbarton Oaks Papers 43: 27–72. Vanden Berghe, L. 1983. Reliefs rupestres de l’Iran ancien. Brussels: Musées Royaux d’Art et d’Histoire. Vasiliev, A. A. 1956. “The Iconoclastic Edict of the Caliph Yazid II, a.d. 721.” Dumbarton Oaks Papers 9–10: 25–47. Wellhausen, J. 1963. The Arab Kingdom and Its Fall. Translated by M. G. Weir. Beirut: Khayats. Wheeler, R. E. M. 1964. Roman Art and Architecture. London: Thames and Hudson. Winkes, R. 1969. Clipeata Imago: Studien zu einer römischen Bildnisform. Bonn: R. Habelt.
313
5
Khanāṣira and Andarı̄n (Northern Syria) in the Umayyad Period and a New Arabic Tax Document Robert G. Hoyland
forms a pyramid-shaped wedge that thrusts upward from the Arabian Peninsula, its apex nudging the Anatolian uplands in modern southeast Turkey. Its southern portion is stony desert, but to the north, where the climate is slightly cooler and wetter, it gives way to open steppe where, now as in the far-distant past, pastoralists roam freely with their flocks of sheep and goats. Around the edges of this steppe lie scattered a myriad of small villages, home to seminomadic herders and small- scale farmers, and a handful of larger settlements that served as market towns and administrative centers for these steppe communities. THE SYRIAN DESERT
Khanāṣira and Andarı̄n in the Late Roman Period Two well-known examples of such places are Khanāṣira and Andarīn (ancient Anasartha and Androna) in modern north central Syria (thirty- five miles south-southeast and forty-eight miles south of Aleppo, respectively). Both are very extensive sites, occupying approximately one and a half square kilometers and possessing substantial encompassing walls and numerous churches (twelve in the case of Andarīn). Both seem to have begun their lives as small roadside settlements (Andarīn is designated a mansio by the third-century traveler Antoninus) in the second to
341
134
R obert G . H oyland
third centuries ad, servicing the needs of travelers and merchants. Both then boomed in the fourth to sixth centuries, as is indicated by a good number of dated inscriptions commemorating new construction. Possibly this efflorescence was a result of increased military expenditure in the eastern frontier provinces in response to the successful Persian incursions into Roman territory in the mid-fourth century. We can see evidence of this expansion most clearly in Andarīn where there have been excavations by British, German, and Syrian teams in recent years. These have uncovered a very large four-towered building (ca. seventy-five square meters) with luxurious décor that belies its military designation (kastron, barracks, in an inscription dated 559 ad); two sets of public baths (the earlier one built about the same time as the kastron), both decked out with fine frescoes and much marble; and a couple of houses, one equipped with its own arched cistern and stables and a superb mosaic floor portraying various animals, including a griffin and a lion in the act of killing a gazelle.1 Khanāṣira was elevated to the status of a polis by Emperor Justinian (527–565), that is, it enjoyed the formal status of city and the handouts from the state that such a designation could bring, whereas Andarīn, despite its large size, was only a village (kōmē) in official terms (and described as such in a fifth–sixth century mosaic inscription). This is reflected in the fact that none of the names on the thirty-nine lintels bearing inscriptions in Andarīn are accompanied by a title whereas in the inscriptions of Anasartha we encounter a bishop and praetorian prefects.2 Both settlements, however, must have been heavily dependent for their income on agriculture (and many of the owners of their smart houses and sponsors of their buildings were probably wealthy landowners), since in the vicinity of both lie the remains of an impressive network of underground canals that tapped wadis and aquifers to feed reservoirs around the village (four in the case of Andarīn, the two major ones decorated with niches, colonettes, and animal carvings), that in turn watered the surrounding fields. And both were likely, given their position on the edges of the steppe, to have engaged in commerce with the pastoralists living all around them. Still now numerous bedouin clans spend the summer months in the vicinity of these two sites to escape the oppressive heat of the interior of the Syrian Desert, pasturing their flocks on the recently harvested fields, offering their labor to their settled agriculturalist neighbors, selling their own animal products, and buying what they need for the winter ahead. And it is plausible that this situation continues a very ancient tradition, with exchanges between these groups an important feature of the commercial life of Khanāṣira and Andarīn for centuries before the modern day.
315
Khanāṣira and Andarīn
135
The exchanges were apparently not limited to the realm of business, as we can see from the many distinctively Arab names, often with Greco-Latin ones, in the inscriptions (mostly in the official language of Greek) of Khanāṣira, Andarīn, and the surrounding area. For example, the martyrium of St. Sergius at Zebed, only a very short hop away, is commemorated by a bilingual Greek and Syriac inscription (dated 512 ad) and by a short memento in Arabic signed by, among others, Sergius, son of Amat-Manāf, and Ḥaniyā, son of Mar al-Qays (fig. 5.1). The parents of these two have Arab pagan names, and presumably either they or their sons converted to Christianity, perhaps after settling down in places like Anasartha and Androna and/or marrying one of its inhabitants. In all such exchanges, Christianity was important for facilitating intercourse between the various groups, and sources for the fourth to sixth centuries ad regale us with numerous stories about Christian missionary work among the Arab tribes of the Syrian steppe that won over innumerable souls to the faith. The chiefs of these newly converted tribes would often then involve themselves in local Christian politics and patronize Christian culture, further forging the link between the settlements and tribes of the steppe.3
Figure 5.1 A section of the Zebed inscription showing the name Ḥaniyā, son of Mar al-Qays, in Arabic script.
361
136
R obert G . H oyland
Khanāṣira and Andarı̄n in the Umayyad Period We know that both settlements remained occupied during the Umayyad period, though our evidence for each place is of a different kind. Perhaps because it had been a polis—and thus of some importance—Khanāṣira is mentioned frequently by Muslim historians. Apparently its Christian residents went out to meet the first governor of Syria, Abū ʿUbayda ibn al-Jarrāḥ, to make a peace agreement with him.4 A number of prominent Umayyads are reported to have visited it, and the caliph ʿUmar ibn ʿAbd al-ʿAzīz (r. 717–720) is said to have been appointed over it by ʿAbd al-Malik and to have used it as his residence during his time as caliph. It would seem to have had a mosque from a very early date, since ʿUmar is reported as having delivered the Friday sermon while he was there.5 By contrast, we hear almost nothing about Andarīn from Muslim writers,6 but excavations there reveal that it, too, received the attention of the Muslim authorities in the Umayyad period. For a long time only Greek inscriptions were found on the site (around seventy in total), but recently a Syriac text that accompanied a fresco depicting the annunciation scene was discovered in the kastron,7 and now a couple of Arabic inscriptions have come to light. One was found in the same kastron (fig. 5.2) and is a simple graffito on a limestone block asking for God to pray for Ghanam,
Figure 5.2 Andarīn, graffito on a limestone slab found in the kastron.
317
Khanāṣira and Andarīn
137
son of … al-Ramlī: الرميل. . … صىل اللّٰه عىل غنم بن.8 As regards its date, the slender extended alifs and lāms, the elongated low ṣād, the long line of the alif maqṣūra of ʿalā, and the ʿayn/ghayn with its curve leaning back and its initial line stretching forward, all suggest a date in the second/eighth century. Also, assuming that the reading of the last part of the name, al- Ramlī, is correct, then the text must postdate the foundation of Ramla in 715 ad.
A New Umayyad Tax Document This eighth-century date is corroborated by a second Arabic text from Andarīn, which was found during excavation of the later of the two bath complexes by a Syrian team headed by Radi al-ʿUqda. The earlier baths were built around 558 ce and remained in use until destroyed, perhaps by an earthquake, in the late seventh/early eighth century.9 Thereafter a new bath complex was built, adjacent to the old one and making use of its water source. The inscription is written in what looks like ink on a small marble tablet, probably a piece of the marble revetment from the earlier baths. This might seem strange at first, but the adjacent ruined baths would mean that a lot of marble revetment was just lying around, and it does make a very nice flat writing surface. It is then akin to potsherds and simply serves as rough paper, useful for writing the draft of a document before committing it to papyrus or parchment, which was very expensive.10 It is a common-enough phenomenon and we have numerous examples of doodles, receipts, and the like inscribed on marble, and even a letter to a caliph (fig. 5.3).11 The other advantage of using a marble tablet is that it is easy to wash off the text and use it again. Unfortunately for the modern scholar this makes it more difficult to decipher, since the latest text is obscured by traces of earlier texts. I will give first my suggested reading of this Andarīn bath text (Plates 7 and 8 in color insert) and then discuss its significance: الرحيم بسم اللّٰه الرحمن١ معوية من الليث بن الذيال عامل االمري٢ واهله بن امري المومنني عىل ارض قنرسين٣ االول تكفي مكوس من اقليم رعبان٤ غسان من كورة قورص فاعطها٥ …مطرف بن٦
381
138
R obert G . H oyland
5cm
Figure 5.3 Arabic text on marble of a letter to Caliph Hishām from Khirbat al- Mafjar (after Baramki 1939, p. 53 and pl. XXXIV, no. 2).
1. In the name of God the Merciful the Compassionate 2. From al-Layth ibn al-Dhiyāl, governor for the emir Muʿāwiya, 3. son of the commander of the faithful, over the province of Qinnasrin 4. and its people. You should pay in full the taxes of the district of Raʿbān al-Awwal 5. of the canton of Cyrrhus (Qūruṣ) and then give them to Ghassān 6. son of Muṭarrif12 … [the rest of the line and the following line are unclear] The text begins in the customary way for an item of correspondence: an invocation, followed by the identification of the sender: min (“from”) x, of such and such a position. Usually the addressee would come next: ilā (“to”) x; perhaps it is understood to be the relevant official of the village in question or maybe this is an oversight in the draft, which was corrected in
391
Khanāṣira and Andarīn
139
the final version. Next comes the request, namely to remit in full the taxes of Raʿbān al-Awwal; otherwise one could read this as a statement, namely that “the taxes of Raʿbān al-Awwal are sufficient,” that is, they have been paid in full. Raʿbān is a small place “between Aleppo and Samosata near the Euphrates,”13 which was evidently tributary to the canton of Cyrrhus (Qūruṣ), a famous city in Late Antiquity, lying some seventy kilometers north of Aleppo. The word for taxes here, mukūs, appears in both contemporary papyri and literary sources, though usually in the singular (maks).14 The administrative hierarchy is the same as that found in the Nessana papyri of southern Palestine: the smallest unit is the iqlīm, which is part of a larger unit called a kūra, which is itself part of the province, here designated arḍ,15 of which there were five in Greater Syria (Palestine, Jordan, Damascus, Homs, and Qinnasrin). This is also neatly demonstrated by a lead seal from the province of Homs, probably of the second/eighth century ad (fig. 5.4), which includes the same three administrative levels. The taxes, once collected in full, are then to be handed over to the designated recipient and so we get the command fa-ʿṭihā (“so give them to”) x. We see the same construction used in the tax documents of the Nessana collection of the 60s/680s.16 So who is the emir Muʿāwiya? The first three words of line three are very faint except for the letters al-mu of the third word. Given that some sort of title is expected here, amīr al-muʾminīn (“commander of the faithful”) immediately springs to mind, and it is certainly possible to read the second word as amīr. Since Muʿāwiya has already been designated as emir,
Figure 5.4 Lead seal illustrating administrative levels of iqlīm, kūra, and province (here Homs). 27 × 26 × 3 mm. London, British Museum, OR 5289.
401
140
R obert G . H oyland
he cannot himself be the commander of the faithful, and so the only solution is to read the first word as ibn (“son of”),17 which then yields the standard formula “emir x, son of the commander of the faithful.”18 Since the script is most likely of the second/eighth century ad,19 the reference must be to Muʿāwiya, son of the caliph Hishām (724–743). He is closely connected with the province of Qinnasrin, launching numerous raids from it into Byzantine territory and residing in Dayr Ḥanīnāʾ, near Rusafa, which lay within that province and served as a residence and base of the caliph Hishām and his son Sulaymān.20 An interesting question is how did these marble tablets end up in the baths in Andarīn? The latter is only about 80 miles west of Rusafa. It has excellent water sources and so would make a good stop if one were then traveling on to Hama (ca. 40 miles to the southwest) or Aleppo (ca. 50 miles to the north). The barracks had already been transformed in the late sixth century to early seventh century to an intramural monastery, with a church erected at its center. However, the destruction in the late seventh century to early eighth century,21 possibly attributable to an earthquake, would seems to have brought to an end this phase of the building’s life. Thereafter it apparently served an administrative function, perhaps dealing with fiscal matters on behalf of the Umayyad government in Damascus.22 Certainly there is material evidence to suggest that Andarīn was still flourishing in the late Umayyad and early Abbasid
Figure 5.5 a–b Umayyad copper coin found during excavation of a kiln by the southeast reservoir of Andarīn. Obv. marg.: muḥammad rasūl all[āh arsalahu bi’l-hudā…] li-yuẓhirahu ʿalā al-d[īn]; obv. field: lā ilāha illā allāh [waḥdahu] lā sha[rīka lahu]. Rev. marg.: bismillāh ḍur[iba]…sanat…; rev. field: allāh aḥad allāh al-ṣamad lam yalid walam yūlad. No mint, probably Harran; AH 116 = AD 734. Hama Museum Store (after Walker 1956, no. 954, p. 293).
4 11
Khanāṣira and Andarīn
141
Figure 5.5 c–d Abbasid copper coin found during excavation of the northern end of the southeast reservoir of Andarīn. Obv. marg.: bismillāh mā amara bihi ṣāliḥ ibn ʿalī; obv. field: lā ilāha illā allāh waḥdahu lā sharīka lahu. Rev. marg.: ḍuriba [hādhā al-fals] bi-khaznat ḥalab sanat sitt wa-arbaʿīn wa-miʾa; rev. Field.: muḥammad rasūl Allāh. Mint: Aleppo; ah 146 = ad 763; governor: Ṣāliḥ ibn ʿAlī. Hama Museum Store (after Shamma 1998, 91; Lowick 1996, 318–319).
periods, such as ceramic and glass finds of the seventh to eighth centuries in the east hall of one of the houses23 and two copper coins found in excavated contexts. One of the coins, from the late Umayyad period, was found in a kiln near the settlement’s southeast reservoir (figs. 5.5a–b), and the second, from the early Abbasid period, was uncovered at the bottom of the same reservoir, at its northern end (figs. 5.5c–d).24 It is possible, then, that al-Layth ibn al-Dhiyāl stayed here for a time with his scribe and dealt with some of his correspondence. It may even be that the new baths were being constructed while he was there and his draft correspondence, once finished with, was tossed into the early construction trenches. The tablets were certainly buried quite deeply, rather than lying amid general debris, and this is how they were preserved from the deleterious effects of the rain.
Conclusion Although these two texts are not much to look at, they constitute an interesting addition to the growing body of evidence attesting to the strong presence of the Umayyad family and their allies in northern Syria in the first half of the eighth century, a presence more solidly attested by the estates and irrigation projects that they commissioned along the River Euphrates from Balis to Raqqa.25 A contemporary Syriac chronicle, completed ca. 732 ad,
421
142
R obert G . H oyland
states explicitly that the caliph Hishām, upon his accession in 724, “ordered waters and canals to be dug in the direction of the desert by free and forced labour … from the Euphrates above Callinicum (Raqqa) and he established plantations and gardens.”26 And many of the Marwanid branch of the Umayyad family had residences in northern Syria and even based their rule there for a time: ʿUmar II at Anasartha, al-Walīd ibn Yazīd at Palmyra, Hishām at Rusafa and Qaṣr al-Ḥayr al-Sharqī and Marwān II at Harran.27
Notes 1. The information in this and the following paragraph, plus further discussion, can be found in Strube 2003, 2008; Mundell Mango 2002, 2008, and 2016. One of the houses is dated to 583–584 ad. 2. Prentice 1908, no. 318. 3. There are many studies of this process; see in particular Trimingham 1979 and Fisher 2011, chap. 2. 4. al-Balādhurī 1866, 149. 5. al-Maqdisī 1899–1919, 6:47; al-Suyūṭī 1967–1968, 2:93; Ibn Kathīr 2010, 9:381, 10:86; Abū Nuʿaym 1932–1938, 5:286. 6. Some Muslim writers connect the north Syrian village of Andarīn with a reference to “the wines of al-Andarīna” that occurs in the first line of the famous pre- Islamic Arabic poem (muʿallaqa) of ʿAmr ibn Kulthūm (e.g., Yāqūt 1866–1873, s.v., who notes that the settlement was in ruins in his day), but this was by no means the only interpretation of this verse. 7. Brock 2005. 8. C. P. Haase reads mḥf[l] ṣlybʾ qlm ʿly ʿy, which he translates as “Versammlungsort? Das Kreuz (Akusativ) des … au [ʿsa/Jesus?]” (Strube 2003, 114); to reach this reading he takes some extraneous marks above the first line to be part of the main inscription and makes no attempt to relate the text to the norms and conventions of Arabic epigraphy. Possibly one could read the first line as ṣallaba ʿalā = made the sign of the cross over, but this expression is not known in the extant graffiti whereas ṣallā ʿalā occurs very often. I prefer to assume that the writer simply forgot to put the tail on the alif maqṣūra rather than to opt for a more strained interpretation. For a survey of the phraseology of early Arabic graffiti, see Hoyland 1997, esp. 80 and n. 15 for the use of ṣallā ʿalā. 9. Mundell Mango 2002, 310. Something like an earthquake is probable because the kastron, which lies directly to the east of the baths, also suffered destruction at the same time (personal communication from Christine Strube).
413
Khanāṣira and Andarīn
143
10. Thus the seventh-century abbot of Iona Adomnan, in composing his work De locis sanctis (Adamnan 1983, 36), tells us that he first wrote it on (wax) tablets before committing it to parchment. 11. Baramki 1939, pl. 34 (“to ʿAbd Allāh Hishām, commander of the faithful”); another twenty Arabic texts “written on marble fragments in carbon ink” are published in Baramki 1953, 105– 117; for the Greek texts on marble from this site, see Schwabe 1945–1946, 20–30. For examples from elsewhere, see Schlumberger 1939, fig. 29 (Qaṣr al-Ḥayr al-Gharbī: “from Hishām”); Walmsley et al. 2008, 122–126 (a list of accounts from a shop in Jerash). 12. The reading of the name is not certain; in particular, the last letter of the first name and the first letter of the second name are unclear. A little more of this text is perhaps decipherable and I hope to return to it at some future date and also to work on a second fragment that was found with it and may be related to it (see Plate 7 in the color insert). 13. Yāqūt, Muʿjam, s.v. “Raʿbān.” The identification is not certain, because on the tablet it seems to be accompanied by the epithet al-awwal, “the first” (though the word is not very clear), which does not feature in Yāqūt, and the place where a bāʾ would be is overlaid by an alif from the line below. However, the location of Raʿbān would fit well. 14. For examples of its use in literary sources, see Lecker 2001, 33–34. For discussion of its exact meaning (customs duty/sales tax?), see Becker 1906, 51–56 (papyrus 2, line 24, mentions a ṣāḥib al-maks). 15. Although jund is the word that is used in ninth-and tenth-century literary texts to designate a province, it does not occur in any inscription or document of the seventh or eighth century. 16. Kraemer 1958, 175–197. 17. The curved stroke above the kāf at the beginning of line 4 might well be the tail of a nūn. 18. Thus the Qaṣr Burquʿ inscription from northeast Jordan (al-amīr al-walīd ibn amīr al-muʾminīn) of 91 ah (Gruendler 1993, 18–19) and a Jabal Says inscription from southeast Syria (al-amīr khālid ibn amīr al-muʾminīn) also of 91 ah (ʿUshsh 1964, 357, no. 21). 19. In an earlier article (Hoyland 2006, 446), having only a bad photograph to work from, I had linked the text with Muʿāwiya I (r. 661–680), but with the permission of Jamal Ramadan, director of the Antiquities of Hama, I was able to examine the tablet itself and I am now convinced that a second/eighth-century date is the most plausible. Alain George kindly looked at the text for me and corroborated this dating for the script. I am grateful to Jamal Ramadan and Radi al-ʿUqda for allowing me to publish this inscription. 20. al-Ṭabarī 1866–1901, 4:120–157, has him lead raids into Byzantine territory in the years 107, 109, and 110–117 ah; Yāqūt 1866-1873, s.v. “Ḥanīnāʾ”; Hoyland 2011, 252.
441
144
R obert G . H oyland
2 1. See note 6 above. 22. Personal communication from Christine Strube, and see Strube 2008, 57–58. 23. Strube 2008, 59. 24. I am very grateful to Marlia Mango, director of the Oxford archaeological excavations at Andarīn (in which I participated during 1998–2010), for allowing me to publish these two coins and to Luke Treadwell (Oxford) for help with identifying them. 25. See the articles by T. Leisten (“The Marwanid Mansion at Balis”), C-P. Haase (Madinat al Far/Hisn Maslama), T. Ulbert, and D. Sack (Rusafa) in Bartl and Moaz 2008. 26. Hoyland 2011, 223–224; cf. p. 259: “Marwan returned to Callinicum and ordered his troops to plunder the properties of Hishām, which were on the banks of the Euphrates.” 27. For Palmyra and Qaṣr al-Ḥayr al-Sharqī, see Genequand 2012. For Rusafa, see the previous note, and for Marwān II at Harran, see Hoyland 2011, 281–282.
Bibliography Adamnan. 1983. De locis sanctis, edited by D. Meehan. Dublin, 1983. al-Balādhurī. 1866. Futūḥ al-buldān, edited by M. J. de Goeje. Leiden: Brill. Baramki, D. C. 1939. “Excavations at Khirbet al-Mefjer.” Quarterly of the Department of Archaeology in Palestine 8: 51–53. ———. 1953. “Arab Culture and Architecture of the Umayyad Period.” Unpublished PhD diss., University of London. Bartl, K., and A. R. Moaz, eds. 2008. Residences, Castles, Settlements: Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham. Rahden, Westf.: Verlag Marie Leidorf. Becker, C. H. 1906. Papyri Schott-Reinhardt I. Heidelberg: C. Winter. Brock, S. 2005. “The Syriac Inscription of Androna/ al- Andarin.” In Christliche Wandmalereien in Syrien, edited by A. Schmidt and S. Westphalen, 199–202. Wiesbaden: Reichert. Fisher, G. 2011. Between Empires: Arabs, Romans and Sasanians in Late Antiquity. Oxford: Oxford University Press. Genequand, D. 2012. Les établissements des élites Omeyyades en Palmyrène et au Proche- Orient. Beirut: Institut Français du Proche-Orient. Gruendler, B. 1993. The Development of the Arabic Scripts: From the Nabatean Era to the First Islamic Century according to Dated Texts. Atlanta: Scholars Press. Hoyland, R. 1997. “The Content and Context of Early Arabic Inscriptions.” Jerusalem Studies in Arabic and Islam 21: 77–102. ———. 2006. “New Documentary Texts and the Early Islamic State.” Bulletin of the School of Oriental Studies 69: 395–416. ———. 2011. Theophilus of Edessa’s Chronicle and the Circulation of Historical Knowledge in Late Antiquity and Early Islam. Liverpool: Liverpool University Press.
415
Khanāṣira and Andarīn
145
Ibn Kathīr. 2010. al-Bidāya wa’l-nihāya, edited by Muhyī l-Dīn Mastū. 10 vols. Beirut: Dār Ibn Kathīr, 2010. al-Iṣfahānī. 1932–1938. Ḥilyat al-awliyāʾ. 10 vols. Cairo: Maktabat al-Khānjī. Kraemer, C. J. 1958. Excavations at Nessana, Volume 3: Non-Literary Papyri. Princeton, NJ: Princeton University Press. Lecker, M. 2001. “Were Customs Duties Levied at the Time of the Prophet Muhammad?” Al-Qantara 22: 19–43. Lowick, N. 1996. Early ʿAbbāsid Coinage: A Type Corpus 132–218 ah/ad 750–833, edited by E. Savage. London: British Museum. al-Maqdisī, Muṭahhar ibn Ṭāhir. 1899–1919. Le livre de la création et de l’histoire, publié et traduit d’après le manuscrit de Constantinople. Kitāb al-badʾ wa’l-taʾrīkh. Edited and translated by Clément Huart. 6 vols. Paris: Ernest Leroux. Mundell Mango, Marlia. 2002. “Excavations and Survey at Androna, Syria.” Dumbarton Oaks Papers 56: 307–315. ———. 2008. “Baths, Reservoirs and Water Use at Androna in Late Antiquity and the Early Islamic Period.” In Residences, Castles, Settlements: Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham, edited by K. Bartl and A. R. Moaz, 73–88. Rahden, Westf.: VML, Verlag Marie Leidorf. ———. 2016. “Byzantine Settlement Expansion in North Central Syria: The Case of Androna/Andarin.” In New Cities in Late Antiquity, edited by Efthymios Rizos and Alessandra Ricci, 93–122. Turnhout: Brepols. Prentice, W. K. 1908. Greek and Latin Inscriptions in Syria. Publications of the Princeton University Archaeological Expedition to Syria. New York: Century. Schlumberger, D. 1939. “Les fouilles de Qasr el-Heir el-Gharbi (1936–38).” Syria 20: 324–373. Schwabe, M. 1945–1946. “Khirbat Mafjar: Greek Inscribed Fragments.” Quarterly of the Department of Archaeology in Palestine 12: 20–30. Shamma, S. 1998. A Catalogue of Abbasid Copper Coins. London: Al-Rafid. Strube, C. 2003. “Androna/al-Andarin: Vorbericht über die Grabungskampagnen in den Jahren 1997–2001.” Archäologischer Anzeiger 1: 25–115. ———. 2008: “Al-Andarin/Androna: Site and Setting.” In Residences, Castles, Settlements: Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham, edited by K. Bartl and A. R. Moaz, 57–72. Rahden, Westf.: VML, Verlag Marie Leidorf. al-Suyūṭī. 1967–1968. Ḥusn al-muḥāḍara fī akhbār miṣr wa’l-qāhira, edited by M. Ibrāhīm. Cairo: ʿIsā al-Bābī al-Ḥalabī. al-Ṭabarī. 1866–1901. Taʾrīkh al-rusul wa’l-mulūk, edited by M. J. de Goeje. 15 vols. Leiden: Brill. Trimingham, J. S. 1979. Christianity among the Arabs in Pre-Islamic Times. London: Longman. al-ʿUshsh, M. 1964. “Kitābāt ʿarabiyya ghayr manshūra fī jabal usays.” Al-Abḥāth 17: 227–316.
461
146
R obert G . H oyland
Walker, J. 1956. A Catalogue of the Arab-Byzantine and Post-Reform Umayyad Coins. Vol. 2: A Catalogue of the Arab-Byzantine and Post-Reform Umaiyad Coins. London: British Museum. Walmsley, A., L. Blanke, K. Damgaard, A. Mellah, S. McPhillips, L. Roenje, I. Simpson, and F. Bessard. 2008. “A Mosque, Shops and Bath in Central Jarash.” Annual of the Department of Antiquities of Jordan 52: 109–137. Yāqūt al-Ḥamawī. 1866–1873. Muʿjam al-buldān, edited by F. Wüstenfeld. 6 vols. Leipzig.
417
6
Two Possible Caliphal Representations from Qaṣr al-Ḥ ayr al-Sharqı̄ and Their Implication for the History of the Site Denis Genequand
the Syrian- Swiss archaeological mission at Qaṣr al-Ḥayr al-Sharqī excavated an aristocratic residence situated in the northern part of the famous Umayyad site. Called Building E, this residence was occupied for a rather short time during the first half of the eighth century and was decorated by carved stuccoes, including geometric, vegetal, and figurative motifs. A preliminary plan of the building and some of the stucco decorations have already been published.1 The present chapter aims to provide further precisions on the plan of Building E, now better known after further field seasons, and to return to the possible interpretation of two of the stucco panels with figurative motifs.2 Qaṣr al-Ḥayr al-Sharqī is an Umayyad aristocratic settlement, one of the so-called desert castles. It lies in the steppe lands some 110 kilometers to the east-north-east of Palmyra. The site was investigated during the 1960s by an American project led by Oleg Grabar and more recently by the Syrian-Swiss archaeological mission.3 It is an enormous archaeological site comprising a caliphal palace; a large enclosure including a mosque, BETWEEN 2007 AND 2011,
48 1
148
D enis G enequand
dwelling units, and an industrial unit with oil presses; a bath; extended areas covered with mud-brick structures (mostly houses) to the north and east of the palace; and an impressive hydraulic and agricultural system with two irrigated enclosures and a number of other agriculture-related structures. An inscription found in the nineteenth century, and now lost, attributes its construction as a madīna to the caliph Hishām ibn ʿAbd al- Malik in 110 ah/728–729 ad.4
Building E Building E at Qaṣr al-Ḥayr al-Sharqī was excavated during five field seasons up until the summer of 2011, when fieldwork was interrupted due to the political situation in Syria.5 With the exception of two areas and large surfaces in the courtyards, it was almost completely excavated. The building is situated in a densely built-up area to the north of the palace (fig. 6.1). Extended excavations and smaller trenches showed that all the surrounding structures were mud-brick houses with rows of square rooms covered by earthen domes organized around rectangular courtyards. Most of these were settled in the eighth century and some of them continued to be occupied until the mid-ninth century. Even before excavations, Building E appeared to be a more impressive and more formally planned structure than those around it. It measures 44.50 by 65.50 meters and comprises two adjoining but independent parts, each organized around its own rectangular courtyard (fig. 6.2). The western part had a residential function, while the eastern part was some sort of storehouse. The building is entirely built in mud bricks (ca. 40 by 40 by 8–9 centimeters and half modules).
Western Part The western part comprises four wings of rooms and was accessible by a gateway in its southeastern corner. The western, southern, and eastern wings have only one row of rooms and were single-story structures, while the northern one has two rows and also had an upper story. The southern wing comprises five rooms and an L-shaped entrance hall (1) linking the main gateway and the central courtyard. The first room beside the entrance hall (2) has a rectangular plan with side entrances and benches along three sides; it is likely to have been a vestibule. The
491
Figure 6.1 Plan of the northern area of Qaṣr al-Ḥayr al-Sharqī (survey and drawing by Sophie Reynard, Denis Genequand, and Marion Berti).
501
150
D enis G enequand
Figure 6.2 Plan of Building E (drawing by Marion Berti).
other rooms are square in plan. Several of them seem to have had a service function: room 4 includes separation walls and was likely devoted to storage; room 6 had a low bench and a raised plastered basin with a water drain, which may have served either as a shower tub or to clean kitchen utensils. Room 5 is better arranged and the floor and walls are uniformly plastered. Half of these rooms open into the courtyard; the others have side entrances. It is not known whether the square rooms were covered by barrel vaults or by an earthen dome, like in the surrounding houses. The eastern wing comprises three rooms: two square ones and one with a rectangular plan. The door between rooms 8 and 22 was blocked at a very early stage, and room 8 also seems related to the gateway and entrance corridor. Rooms 22 and 23 form a single unit and the latter includes a raised and plastered basin with water drain. Part of the western wing still has not been excavated, but walls are visible on the surface and form one small rectangular room (21) and two regular square rooms (19 and 20). It also includes the southern part of an extremely long rectangular room belonging to the northern wing (15).
151
Two Possible Caliphal Representations
151
Being larger and including an upper floor, the northern wing is by far the most elaborate of the whole building (Plate 9 in color insert). With the exception of corridors 9 and 10, service room 11, and bathrooms 16 and 17, all the other rooms have a higher standing. They are well built and their floors and walls are systematically covered by plaster, sometimes forming thicker plinths on the lower part of the walls. They are all rectangular in plan and were covered by barrel vaults. Some of them (12 and 15) are subdivided by diaphragm arches supported by pillars. The walls of some of the rooms (12, 13, 15, 18) are also reinforced by vertical and horizontal wooden beams inserted into the mud-brick masonries. There are almost no features or other indications to help define the function of each room precisely, and we may only presume that they were primarily intended as living and reception quarters for the owner of the building. Room 11 is divided by an asymmetrical row of pillars and was perhaps not covered by a vault, but by a flat roof. It includes four cylindrical bread ovens (tannūr) in one corner, a low basin that may be a kneading trough, a raised platform, and a large raised basin; the poor state of preservation of the draining part of the latter makes it difficult to be sure whether it is a structure combining a shower tub and latrine, or only a basin devoted to kitchen-related activities. Small rectangular bathrooms 17 and 18 have exactly the same plan and contain raised platforms, accessible by a few steps, with a concave plastered surface followed by a drain and latrines in a recessed niche in the back wall. The drains led to a large pit dug in bedrock to the back of the building. The first bathroom opens from service room 11 and the second is only accessible from the central room (13) of the wing, giving it some more importance as a private bathroom. Two staircases, abutted to the northern and western façades of the courtyard, led to the upper floor. The observation of the collapsed walls and destruction levels suggests that the upper floor had roughly the same plan as the ground floor. The roof was organized as a terrace and was covered by a thick layer of plaster.
Eastern Part The eastern part has only three wings with a single row of rooms (fig. 6.3). The gateway is certainly in the unexcavated area in its southwestern corner, close to the gateway of the western part. Originally the northern wing had only three square rooms. Room 27 includes an elevated
521
152
D enis G enequand
Figure 6.3 Aerial view of the eastern part of Building E.
platform serving both as a shower tub and latrine, and remains of several bread ovens and fireplaces. Room 26 has no internal installations that would allow its function to be determined. Room 25 has some small storage compartments in the corners and its floor was covered by many similar broken jars, pointing to a storage function. The space between room 27 and the dividing wall originally belonged to the courtyard and was built up at a later stage. It became a rectangular vaulted room (28) subdivided in three parts by pilasters supporting diaphragm arches. An upper story existed over this room, as shown by collapsed remains and by a staircase abutted to its southern wall in the courtyard. The eastern wing comprises only one very long hall, subdivided into seven square rooms by pilasters supporting arches. Most of these rooms were accessible directly from the courtyard and some also have doors facing the outside of the building. The southern wing has more or less the same plan, with an elongated hall subdivided in at least three rooms, and it also included the gateway. Internal installations are rare and comprise low platforms, storage compartments in the corners or along the walls, or fireplaces. Only room 36 has two circular silos dug in the floor; these have a diameter of 1.20 meters and a depth of 1.10 meters and are completely plastered. It is not clear whether the halls or rooms were covered by barrel vaults or by domes.
513
Two Possible Caliphal Representations
153
Date of Building E There are no inscriptions or deposits belonging to the construction phase of the building. Nevertheless, the comparison of building techniques and materials suggests that the date of construction is quite close to those of the other structures built in Qaṣr al-Ḥayr al-Sharqī during the first decades of the eighth century. But for the time being, there is no way of knowing precisely whether it predates or postdates the construction of the madīna in 110/728–729. In any case, the building does not seem to have been occupied for very long. The date of abandonment is somewhat easier to define and this is first and foremost indicated by the pottery retrieved from several layers related to the last phase of occupation (filling of pits, trash deposits, abandoned pots in some rooms). The pottery assemblage includes a number of brittle ware types that are diagnostic for the Umayyad period (cooking pots with high conical necks); north Syrian painted amphorae attributed to the seventh and early eighth centuries; a number of jars and drinking cups with a red and mineral fabric quite different from the soft sandy fabric of the Abbasid period; and several imports from the southern Levant (painted and cooking wares).6 This pottery assemblage points to abandonment close to the mid-eighth century or during the second half of the same century. All the coins retrieved during the excavation seem to be Umayyad fulūs belonging to the series later than the monetary reform of ʿAbd al-Malik in the 70s/690s.
Interpretation of the Plan It is quite clear from the above description that the western part of Building E functioned as a residence and housed someone important; this will be even clearer after the presentation of architectural decoration. The eastern part had only limited housing possibilities and, with its elongated halls on two of the wings, is best interpreted as some sort of warehouse with storage facilities. The combination of these two elements in one single building is not often found in early Islamic architecture, although less extensive storerooms are sometimes assumed in some Umayyad palaces or aristocratic residences.7 Warehouses conceived as independent buildings are attested in a few instances within Umayyad aristocratic settlements, for example at Jabal Says or in Qaṣr al-Ḥayr al-Sharqī itself with the so-called southern castles (fig. 6.4).8 We shall come back later to a possible interpretation of this rather odd combination in Building E.
541
154
D enis G enequand
Figure 6.4 Plan of the so-called southern castles at Qaṣr al-Ḥayr al-Sharqī (survey and drawing by Sophie Reynard, Denis Genequand and Marion Berti).
Stucco Decoration from Building E Four main groups of stucco decorations were found in Building E: inside the gateway and entrance corridor of the western part; in the western courtyard directly in front of the façade of the northern wing (in front of
515
Two Possible Caliphal Representations
155
rooms 13 and 14); within the rooms of the northern wing of the western part (rooms 11 to 15 and 18); and in the northwestern corner of the eastern courtyard (in front of room 28ter).9 The first group comprises some rectangular or diamond-shaped medallions with interlaced designs and a number of engaged half-columns bordered by serrated motifs. These were likely the decorations of the gateway and the sidewalls of the entrance corridor. The second group is very homogeneous and comprises at least thirty arches or window frames found collapsed at the foot of the façade in the courtyard. The third and fourth groups, despite being found in different places, are quite similar in their composition together and include at least twenty- six rectangular panels with geometric, vegetal, or figurative motifs. It is likely that the panels found in the eastern courtyard were moved there after the abandonment of the building and that these two groups are in reality only one, whose original position was in the different rooms of the first floor of the western part. The two panels, which will be discussed in more details here, belong, respectively, to the third and fourth groups and were found in the destruction layers of room 11 (panel H) and on the floor of the eastern courtyard (panel Q).
Panel H: A Falconry Scene This panel measures 49.5 by 66 centimeters (4 centimeters thick). It has a rectangular frame and is partly openwork. It represents a man riding a horse and holding a bird (Plate 10 in color insert). They are turning to the left and framed by palm leaves: four small alternating leaves with two to four leaflets on a narrow vertical strip to the left; a much larger leaf with more leaflets develops to the right side and in the upper right corner. The horse and the bird are represented in profile, as well as the lower part of the rider, whose torso and head are in frontal position. An element that looks vegetal, a simplified leaf of sorts, sits between the legs of the horse.
The Horse The horse has some lacunae, all of them at the level of the legs, particularly the lower legs. Its head is turned down, but its eye, represented by a large beveled incision underlined by a second very fine incised line, is
561
156
D enis G enequand
horizontal as if the head were upright. On the top of it, there are two pointed ears and a lock of the mane rests on the forehead. The rest of the mane is not figured on the crest. A vertical line deeply incised down the muzzle characterizes the mouth, but there is no nostril. The limit between the head and the neck is indicated by a small incision between the extremity of the lower mandible and the ears. Apart from the saddlery, there are no details on the breast, the croup, or the rest of the horse’s body. It has a long vertical tail, almost reaching the ground. The position of the horse, with three stretched legs, is one of pause, but the bending movement of the anterior right leg indicates that it has just stopped or is ready to move on. The natural bending of the posterior legs is slightly exaggerated. Only the heel and pastern of the anterior left leg are preserved and overlap onto the frame of the panel; it was also the case for the posterior heels and pasterns. One or two incised and incurved lines underline the joints of the knees and fetlocks according to their bending. Criteria for identifying a small-sized horse rather than a donkey or a mule are very clearly indicated: small ears, flowing mane symbolized by the lock on the forehead, and a long horsehair tail. Several elements of the bridle and saddlery are represented in detail. A bit goes out of the horse’s mouth and is attached to reins held by the rider. The rest of the bridle consists of a crownpiece over the horse’s poll and a browband across its forehead. Surprisingly, these are not followed by a cheekpiece attached to the bit. Both the crownpiece and browband are decorated by oblique incised lines giving the idea of woven material unlike the reins, which are smooth and would normally consist of a single leather strap. The saddle has two different elements. The first is the saddle itself, on which the rider is sitting, and which is represented by a rounded cantle behind and by a rising pommel in front. The saddle rests on a trapezoidal piece of cloth, the saddle blanket, though one cannot completely exclude that this element is only hung on to the side of the saddle or even covers it. Its surface is covered by more or less perpendicular incisions forming a square to diamond-shaped pattern. A girth goes under the horse barrel and maintains the saddle in place. Two other straps, a breastcollar, and a crupper somehow reinforce it, both seemingly attached to the front and back of the saddle. The girth and the breastcollar have a woven design and it should be noted that there is no phalera on the latter; the crupper is a smooth strap with four ovoid pendants.
517
Two Possible Caliphal Representations
157
The Rider The rider holds the bird in his right hand and the horse’s reins in the left. His torso, arms, and head are proportionally too big by comparison to the lower part of his body and to the horse. His left leg is slightly bent in a normal riding position. His arms are also either slightly or completely bent and are joined with no breaks up to very rounded shoulders. The figure wears a one-piece ankle-length garment with long sleeves. Lines incised horizontally, vertically, or diagonally form an irregular pattern of lozenges (these are more rectangular on the leg and lower part of the trunk, and diamond-shaped on the torso and arms). Similar incised lines on the saddle blanket and on the garment of the figure on the next panel (panel Q) clearly show that it is related to the cloth’s texture or motifs. Similar motifs also appear on the trousers of several stucco figures in the round from Khirbat al-Mafjar10 and on different pieces of clothing drawn on earlier Roman and Parthian-Sasanian graffiti, which will be described below. The diagonal or slightly curved incised lines give some relief and movement to the garment and to the body wearing it, especially to the shoulders and elbows. At the end of the sleeves and at the neck, the garment stops without particular edging. It is likely that the garment represents some sort of dress or long tunic, and comparison with the garment of the next figure, with which it has many similarities, makes it unnecessary to interpret it specifically as a hunting suit. At the level of the man’s ankle, a slightly recessed strip without the check patterns probably indicates the presence of trousers below the dress. The rough treatment of the foot makes it impossible to determine whether the figure is wearing shoes or not, and, if so, which type. The rider’s hands are quite detailed. The right one is pointing up and shows three bent fingers and the index finger stretched out; the thumb is not visible. The left foot of the bird grips the whole hand and the right foot only the index finger. The left hand, under the figure’s chest, holds the reins behind the neck of the horse with four bent fingers and the thumb resting on top of them. The head of the rider is shown frontally. The nose is represented without detail, only by a vertical bridge in the continuation of the arches of the eyebrows. Lower and upper eyelids, clearly indicated in relief, frame the almond-shaped eyes. There are still some traces of black paint that accentuated the upper eyelids. The mouth is figured by a depression having the precise contours of lips and is enhanced by red paint. Between the
581
158
D enis G enequand
nose and the mouth, a series of short vertical incisions indicate a moustache, which is then continued horizontally by some longer horizontal incisions. Directly under the mouth, some more vertical incisions indicate a small goatee and a thicker beard is shown by oblique incisions covering the whole chin and large parts of the cheeks. The beard still has traces of black paint.11 The figure wears a pointed, almost conical headgear, which frames his face and covers his ears, the back of his head, and part of his nape. The whole surface of the headgear has received the same treatment as the garment described before: vertical and horizontal incised lines form a checkered pattern. A continuous strip including two rows of squares or lozenges characterize the edge framing the face. It is a hat or more likely some peculiar sort of soft helmet covered by the same kind of cloth as the rider’s garment.
The Bird The bird held by the rider is in a rather odd position, with the head facing down; this is probably largely due to the remaining space in the upper left side of the panel. The bird’s wings are partly spread; a series of long, deep, and curved incisions give the impressions of remiges. A network of curved incisions organized in a parallel and perpendicular manner covers its body. The tail is triangular and stops abruptly because of the border of the panel. As already described, the legs and feet are settled on the right hand of the rider; only two forward digits of the left foot are represented covering the whole hand, while the right foot forms a half-circle with two forward digits gripped on the index finger of the rider and a back digit extending toward the small palm in the upper corner of the panel. The head of the bird presents a bulging forehead with some sort of caruncle reaching the base of the beak, an almost round eye with an excrescence toward the back, and a straight beak with a rounded extremity. The main question regarding the representation of the bird concerns the determination of the species. At first sight, because of the general composition of the panel and its presence in an aristocratic residence, one might think of a bird of prey used in falconry—a falcon, a sparrowhawk, or, less frequently, an eagle—and thus interpret the panel as a hunting scene. Nevertheless, the bird presents some characteristics that make this interpretation questionable.
591
Two Possible Caliphal Representations
159
The head of the bird looks exactly like a pigeon’s head (neck, bulging forehead, caruncle, rounded eye, shape of the beak) and the short triangular tail is also reminiscent of that of a columbinae. Two more elements point in the same direction. First, the rider holds the bird without gloves (his fingers are visible), a somewhat difficult task with a bird of prey because of its claws; but perhaps the sculptor did not wish to enter into these details, like for the foot of the rider, where no shoe is discernible. Second, the downward position of the bird is like that of a carrier pigeon just before the throwing, and taking flight. Apart from the general composition of the scene, which remains a deciding factor, the main argument in favor of a bird of prey is the large feet. But the sculptor did not differentiate between the digits and the claws: the falcon, which kills with its beak, has rather short claws, while the eagle’s claws are very long because it uses them to kill its prey. The motifs on the body of the bird may be reminiscent of the markings on a sparrowhawk, but they may also simply symbolize the texture of the feathers. Comparison with other contemporary representations of birds—in stucco at Qaṣr al-Ḥayr al-Gharbī and Khirbat al-Mafjar, painted in Quṣayr ʿAmra—does not really help, as in most cases the species are not really defined or definable, or are definitely species other than birds of prey or pigeon.
Comparisons and Interpretation Despite some uncertainties regarding the species of the bird, the composition of the panel strongly suggests a hunting scene with a bird of prey, most likely a falcon, as it is both the most common species in the Near East and the most often used in falconry. The rider, despite wearing no gloves, is in a position typical for this kind of activity usually conducted on a horse and with an extended arm in order to hold the falcon away from the hunter’s face. The limited available space in the rectangular panel may have forced the sculptor to represent the bird of prey in the remaining space, hence an odd downward position close to the hunter’s head. Hunting scenes are frequent in Late Antique and Islamic art, but in the vast majority of cases, they represent riding hunters with a bow or, more seldom, with a spear or a sword. This is the case for one of the contemporary paintings in Qaṣr al-Ḥayr al-Gharbī12 (fig. 6.5), but also for the many Sasanian representations either on silver vessels or in stone sculpture.13
601
160
D enis G enequand
Figure 6.5 Qaṣr al- Ḥayr al- Gharbī, painting with a hunting scene (after Schlumberger 1986).
There is a riding falconry scene on one of the Abbasid-period paintings found in Nishapur and dated to the late ninth to tenth century14 (fig. 6.6). There are more numerous scenes with the same subject, but much later in date (from the eleventh and twelfth centuries onward), on various kinds of materials. In the Nishapur painting, a man is riding a galloping horse and holding a bird of prey, most likely a falcon, with his gloved left hand. They are turning to the left; both the medium and the size of the painting allowed the artist to enter into much more precise details concerning the bridle and saddlery of the horse and the dress of the hunter. The latter wears a much-ornamented tunic or coat (in silk?) over baggy trousers, and some sort of helmet with a conical upper part and a lower part in leather or cloth covering the nape and sides of the face. Two other falconry scenes should be briefly mentioned for comparison. The first is on an eleventh-century cloth from Rayy (Iran) presenting octagonal medallions, including two riding hunters facing each other and holding a bird of prey.15 The hunters have no headgear and are wearing a
6 11
Two Possible Caliphal Representations
161
Figure 6.6 Nishapur, drawing of the painting with a falconry scene (after Wilkinson 1986).
coat covering a tunic and boots. They are also both wearing a glove to hold a falcon, whose head is covered by a hood, alternatively in the right and left hand. The arm position is almost straight. The second scene figures on an Egyptian luster painted bowl dated to the twelfth century (end of the Fatimid period).16 It represents a richly dressed rider, holding the reins of his horse in his right hand and a bird of prey in his gloved left hand stretched toward the back. The clothes of the rider from Qaṣr al-Ḥayr al-Sharqī are in fact quite simple and, despite having no direct parallel, can find comparisons in several other early Islamic representations. At Khirbat al-Mafjar, several figures, among them three riders, wear knee-length dresses or coats.17 Another figure, on foot, wears a dress or a coat open on the chest and decorated by scale-like motifs.18 Despite this last detail, it is not really possible to see in this garment an armor made of leather or metal sheets. At least three of the figures (two riders and the pedestrian) are wearing baggy trousers decorated with diagonal incised lines forming squares, at the center of which are small dots. This kind of dress looks Persian in style. Only one figure (“the warrior”) wears real armor typical of the Roman-Byzantine tradition, which is a short tunic made of small rectangular metal sheets.19 In the same group, another two figures, perhaps also soldiers equipped with shoulder belts, are wearing conical-shaped helmets covering the ears and the back of the head; the form of the latter is not so different from the headgear of the rider from Qaṣr al-Ḥayr al-Sharqī.20
621
162
D enis G enequand
A series of ivory carvings, found in al-Ḥumayma (southern Jordan), but attributed to eastern workshops (Iran or Central Asia) and dated to the first half or mid-eighth century, present the same kind of low-relief carving, despite being on a different medium.21 Four panels represent a soldier holding a spear, whose body is seen frontally and the head in profile (Plate 11 in color insert). Some similarities exist between their clothes and those of the male figures on the stuccoes from Qaṣr al-Ḥayr al-Sharqī. They all wear the same uniform, comprising a long dress with a scale motif, perhaps even an armor with scale-shaped leather sheets, baggy trousers with similar motifs, and a soft helmet covering the nape and neck down to the shoulders. It is difficult to know if the scale motif corresponds to some sort of armor, as would be inferred by its shape and association with a helmet, or simply to textile clothes, as its use for the baggy trousers would imply. Two helmets are decorated by horizontal strips with zigzag motifs, very reminiscent of chain mail. A third one has a more familiar decoration with perpendicular lines forming a pattern of checks with dots in the center. It is therefore also possible that the three helmets are made of the same soft material, such as leather or cloth. Two other panels represent a seated personage in an entirely frontal position, wearing the same kind of dress or uniform, but bareheaded and rather longhaired (Plate 12 in color insert). He holds a staff or a sword in a scabbard vertically in front of him, with the hands on the pommel and possibly on the cross-guard. His rounded face, almond-shaped eyes, and nose—a vertical bridge in the continuation of the arches of the eyebrows—are all similar in terms of style and workmanship to those of the rider and other figures from Qaṣr al-Ḥayr al-Sharqī. The use of perpendicular or oblique-crossing incised lines to suggest the texture of various pieces of clothing is frequently attested on older graffiti. In Dura-Europos, this motif is diversely used for clothes made of textile (trousers, edge of tunics), for armor, and for large pieces of saddlery (saddle blanket, caparison).22 A series of five late Parthian or Sasanian graffiti from Persepolis is also worth mentioning, as it systematically uses the cross incised lines for at least one piece of clothing for each person represented and sometimes also for saddlery pieces.23 It is necessary to introduce a last comparison for the headgear of the rider. A soft helmet made of silk was found in Balkh (Afghanistan) and is dated between the seventh and ninth centuries.24 Its upper part is conical and it completely covers the back of the head, the nape, and the ears, leaving the rest of the face exposed, exactly like on the ivories
613
Two Possible Caliphal Representations
163
from al-Ḥumayma. This type of soft helmet, originating from the eastern provinces of the caliphate and here attested by a well-preserved example, is surely a model quite close to those of the soldiers from al-Ḥumayma and probably also to the headgear of the rider from Qaṣr al-Ḥayr al-Sharqī. The conception of the saddlery and bridle of the horse, despite their very simplified details, found a number of near comparisons in other representations dated to the Sasanian or early Islamic period (saddle with vertical girth, blanket, breastcollar and crupper; bit, reins, crownpiece, browband). For the Sasanian period, one might cite as examples the many hunting scenes on silver vessels and the large rock carvings from the second to fourth centuries in Bīshāpūr or Naqsh-i Rustam or from the seventh century in Ṭāq-i Bustān (deer hunt in the large cave).25 For the early Islamic period, the contemporary and geographically close example of the painted hunting scene in Qaṣr al-Ḥayr al-Gharbī comes to mind.26 This early Islamic conception of the saddlery is also not very different from the traditional military saddlery of Roman times.27 We should nevertheless notice the lack of phalera or pendants on the breastcollar in Qaṣr al-Ḥayr al-Sharqī, even though these are always present on representations of Sasanian saddlery. The incomplete saddlery of the stucco horses from Khirbat al-Mafjar, which the excavator compared to Persian models from Ctesiphon and Niẓāmābād, is roughly the same, but with more details, mostly decorative: rounded saddle, breastcollar and crupper, woven reins; only the saddle blanket is really different, with a rounded shape.28 Leaving stylistic comparisons and the rendering of details, we should conclude this discussion at a more general level by adding that hunting evidently had an important role among the various activities of Umayyad court life. This is shown, for instance, by the references to hunting in the risāla of ʿAbd al-Ḥamīd ibn Yaḥyā,29 but also, and in a much more important manner, by the oldest treatise on falconry in Arabic, the Kitāb ḍawārī al-ṭayr, written by Ghiṭrīf ibn Qudāma al-Ghassānī.30 This text was written and compiled from an older document in Baghdad during the reign of the Abbasid caliph al-Mahdī (158–169/775–785) by a Syrian who had been the master of the hunt for the Umayyad caliphs Hishām ibn ʿAbd al-Malik (105–125/724–743)—the builder of most of the monumental structures in Qaṣr al-Ḥayr al-Sharqī—and al-Walīd ibn Yazīd (125–126/743–744).31 This shows beyond any doubt that falconry was certainly widespread at the court of the Umayyads.
641
164
D enis G enequand
Panel Q: The “Standing Caliph” This panel has many lacunae but must have measured in the region of 50 by 70 centimeters (4 centimeters thick). It is bordered by a rectangular frame and is openwork. It represents a standing man, surrounded by a few vegetal elements (fig. 6.7). These are mostly visible on the left side of the panel, which is better preserved, and consist of a series of small palms turned alternatively to the right and left, closely surrounding the body of the figure; a similar alternation of small palms is found to the left of the mounted hunter described above. The man is standing in a frontal position. His arms are bent across the lower part of his chest and he holds a sword vertically in front of him. The left shoulder and arm follow a rounded line quite typical of the figures found on this group of stucco decoration. The man wears a garment in the form of a full ankle-length dress slightly flaring downward. The surface of the garment is fully covered by incised lines that are parallel and perpendicular or oblique, in order to suggest its texture. These lines were
Figure 6.7 Panel Q: the “standing caliph”.
615
Two Possible Caliphal Representations
165
not all incised in the same manner and one can distinguish four different zones corresponding to the lower or upper, right or left parts of the body. One side of the garment bears regular squares with horizontal and vertical lines, while the other is characterized by oblique lines defining diamond- shaped forms instead of squares. On the shoulders and on the arms, the incisions gradually change their orientation or are incurved and give some impression of movement. The lower part of the garment presents undulations depicting its pleated aspect; these are vertical and more or less parallel. They were executed before the incised lines to the left and at the same time to the right, explaining the regular squares. Here again, the incised lines are intended in all likelihood to represent the motifs or texture of the cloth and not some sort of coat of mail as one might think at first sight. It is possible that the garment had a V-neck, but the state of preservation of the panel at this specific place is not sufficient to ascertain it. The feet of the man are rather small, turned in opposite directions and resting on the lower frame of the panel. As the toes are not visible, he is likely wearing some sort of shoes. On each ankle, in the short space in between the shoe and the garment, the lower part of a pair of trousers is clearly discernible. The exact position of the man’s hands is not known, but one can logically think that they are gripping the sword’s hilt or resting on the pommel. Indeed, the preserved part of the hilt has three rounded elements, which can be either part of the grip itself, or the fingers of the man’s left hand gripping the hilt. In any case, the right hand was in the upper position, resting on the pommel or on top of the left one, unless it was stretching in front of the chest. Part of the cross-guard is also preserved and seems quite short compared to the total length of the sword. The blade appears to be inside a scabbard, as would indicate two low grooves decorating the side of the latter and inside which subsist some traces of red paint. There are large lacunae in and around the head of the man and only some deteriorated fragments from his face were found. The latter has an oval form. A short horizontal and deep incision represents the mouth. Part of the nose and the almond-shaped left eye are also preserved. Directly under the mouth, a small vertical goatee is indicated by a few incised lines and a thicker beard covers all the chin and part of the cheeks. By comparison with the other figurative panels, the circular edge of the palm in the upper left corner indicates that the man stood under an arch, was wearing a conical or rounded headgear, or even had more or less rounded hair. Indeed, a fragment of stucco representing the left side
6 61
166
D enis G enequand
of long and curved hair was found in the very same part of the eastern courtyard (it has been added to the previous image on fig. 6.8). The hair is situated under some sort of arch exceeding a half-circle and looking like the trunk of a palm tree. In the same manner as on the other panels, a palm leaf with three leaflets sits in the right corner of the panel. The sides of all the fragments are damaged and there are no direct connections between the piece with hair and the pieces with the face of the standing man; it is therefore not possible to assert that they belong to the same panel. Stylistically, the two fragments seems to belong to two different figures, but, on the other hand, with the exception of a few isolated fragments not attributable to any of the known panels, there is not much evidence for the existence of another figurative panel including a frontal representation. The hair in question is long and one would first think of a female figure, but, as will be clear from a few parallels presented below, long hair could also be male and would fit a standing man holding a sword. In any case, the question of whether the fragment of long
Figure 6.8 Panel Q: hypothetical reconstitution of the “standing caliph” with the fragment of long hair that were found together.
617
Two Possible Caliphal Representations
167
hair belongs or not to the standing figure will remain open, but needed to be asked.
Comparisons and Interpretation Umayyad stucco sculpture has produced two famous figures in frontal representation: one sitting figure from Qaṣr al-Ḥayr al-Gharbī and one standing figure from Khirbat al-Mafjar. Both are in the round, near to life- size, and are supposed to represent caliphs, more precisely Hishām ibn ʿAbd al-Malik. The sculpture from Qaṣr al-Ḥayr al-Gharbī is rather damaged and represents a caliph with Persian-style clothing sitting in a frontal position clearly reminiscent of Sasanian monarchs in throne scenes.32 The standing caliph from Khirbat al-Mafjar is more revealing and was originally on the façade of the bath.33 He is wearing, over baggy trousers, a long dress tightened round the waist by a belt and whose tails are decorated with beaded braids. He is holding in front of him, in his left hand, the grip of a sword partly removed from the scabbard. The upper part of the body is lost, with the exception of the head, characterized by a long beard and a moustache. The treatment in the round of the sculpture and the style are, of course, not the same as on the panel from Qaṣr al-Ḥayr al-Sharqī, but the general composition and the position of the figure are quite close: frontal position, ankle-length flaring dress, and sword in front of the body. The two major differences are the presence of a belt and the oblique holding of the sword at Khirbat al-Mafjar. There are other similar caliphal representations on coins of the “standing caliph” type, which marks the first stage of ʿAbd al-Malik’s monetary reform.34 On these coins, as far as it is possible to judge from the details, the caliph always appears as a bearded and long-haired man in a frontal position, with an ankle-or foot-length dress (sometimes tightened round the waist) and holding a sword with a large scabbard in his hands (right hand on the grip, left hand on the scabbard) (fig. 6.9). The standing caliph is intended as an iconographic representation of the empire on these coins minted in Syria.35 The two front-facing personages already described on the ivory carvings from al-Ḥumayma also present striking similarities.36 Both appear to be holding a vertical sword in a scabbard in front of them, are bareheaded, and wear what might be a diadem. They have rather long and curvy hair, but no beard or moustache (leaving some slight doubts about their being male or female?). By comparison with the soldiers seen in profile, they
681
168
D enis G enequand
Figure 6.9 Golden dīnār of the “standing caliph” type; anonymous and without mint (Damascus), dated ah 77/ad 696. Jena, Jena University, (Oriental Coin Cabinet, Inv. no. 303-A02, ex coll. F. Soret 1866).
are obviously very important persons—military commanders, princes, or rulers—but not hitherto identified. We should also add that the representation of a standing or sitting figure in a frontal position and holding a sword vertically with one or two hands in front of him is frequent on throne scenes of Sasanian monarchs, whether on silver plates or on rock carvings.37
Discussion The two panels from Qaṣr al-Ḥayr al-Sharqī discussed here are single figure representations. According to a number of similar details on the face and clothes of the figures, and to their frontal or half-frontal position, we may postulate that they represent the same personage or, at least, the same category of personage. Therefore, the main question regarding the two figures on panels H and Q is whether or not they could be caliphal representations. Indeed, some of the comparisons alluded to in the preceding lines point in this direction. The clearest case comes from the second, the standing figure, for which other caliphal representations in a similar position on coins (“standing caliph”) or in stucco (Khirbat al- Mafjar) were cited. In general, in ancient Near Eastern art, frontality is a symbol of a high status and often even of kingship. If there are several figures in a composition, usually only the principal one is represented in a frontal position. This is the case for the many Sasanian silver vessels and rock carvings,
6 91
Two Possible Caliphal Representations
169
where only the monarch and his close entourage are seen in a fully frontal position. In Islamic art, this is found for instance on ivory carvings from al-Ḥumayma, where the staff or sword holders, who obviously have an important function or status, are the only one in a frontal position, whereas the soldiers are seen in profile. The probability that the standing figure from Qaṣr al-Ḥayr al-Sharqī is a caliph is therefore quite high and, if we accept that it is the same figure on the two panels, both of them are likely to be caliphal representations. This idea is reinforced if we consider more specifically mounted hunting scenes, whether in the Persian or Islamic world, where the principal role (if there is one) is always played by monarchs, princes, or persons of high rank. Most of the time, these scenes belong to what can be defined as royal or princely iconography. It is therefore perfectly possible that panel H, with the mounted falconry scene, represents an Umayyad prince or caliph. If the two panels are caliphal representations, we should come back to Building E and briefly investigate the reasons for their presence in it. We know that the madīna, which likely included the palace, the large enclosure, and most monumental structures from Qaṣr al-Ḥayr al-Sharqī, was founded by the caliph Hishām ibn ʿAbd al-Malik in 110/728–729. But we do not know if this was a creation entirely ex nihilo or the expansion of an earlier settlement. Textual sources show that several sons of ʿAbd al-Malik, Hishām being prominent among them, had diverse rural estates in wider Palmyrena and in the Euphrates Valley.38 Most of the latter are not clearly identified and others may not necessarily have been mentioned in the sources. Building E could have been the residence of one such estate belonging to Hishām before he became caliph or possibly when he was already caliph, after 105/ 724, but before he decided to expand the settlement and transform it into some sort of new urban settlement or, as he put it in Arabic, a madīna.39 Such a hypothesis would account for the rather short occupation of Building E and its early abandonment in favor of the new palace. Having been the former princely or caliphal residence could be an explanation for the absence of reoccupation at a time when many of the surrounding houses were still settled. This could also provide some clues about the close resemblance between the somital panel of the stucco arches or window frames in Building E and those of the blind arches on top of the towers flanking the gateway of the palace (figs. 6.10 and 6.11).40 Finally, this would account for a progressive and somehow more logical development of Qaṣr al-Ḥayr al-Sharqī during the first decades of the eighth century.
701
Figure 6.10 Arch or window frame from the inner façade of Building E (western part).
Figure 6.11 Blind arches on the top of the towers flanking the entrance of the palace of Qaṣr al-Ḥayr al-Sharqī.
171
Two Possible Caliphal Representations
171
If this hypothesis turns out to be right, it would then also be easier to explain the plan of Building E. If it was built as a simple princely residence between a few years and three decades before the palace and large enclosure, this could explain why the structure does not follow more conventional or classical plans, including porticoes and apartments (buyūt), and why it has this surprising subdivision in two parts with completely different functions. The western part was the residence where Hishām ibn ʿAbd al-Malik, whether a prince, an heir apparent, or even a young caliph, was living with his entourage. That is where he ordered that the decoration must include caliphal representations indicating clearly his own status and that of the structure. The eastern part was intended to be a warehouse for local agricultural products and as a service area for the people engaged in the running of the agricultural part of the estate.41 Having these two functions combined in one single building can be seen as a precursor to what later became organized and constructed in a much more monumental manner within the madīna, for instance with the palace and the so-called southern castles.
Notes 1. Genequand 2011; see also Genequand 2012, 130–142. 2. The Syrian-Swiss mission at Qaṣr al-Ḥayr al-Sharqī was directed by the author and Walid al-Asʿad (Palmyra office of the DGAM). This is the place to thank again the General Directorate of Antiquities and Museums of Syria (DGAM); the then general director, Dr. Bassam Jamous; the director for excavations, Dr. Michel al- Maqdissi; and the Swiss-Liechtenstein Foundation for Archaeological Research Abroad (SLSA, Zürich), which funded the project from its beginning. Sincere thanks are also due to my colleagues and participants to the mission involved in the excavations of Building E, especially Elise Allaoua, Marion Berti, Erwan Dantec, Mathias Dupuis, Marcia Haldemann, and Jacqueline Studer. I warmly thank Isabelle Ruben for correcting the English text of this essay and I am grateful to Rebecca Foote and Stefan Heidemann for providing photos of the ivory carvings found in al-Ḥumayma and of the “standing caliph” dīnār and for allowing me to reproduce them. 3. Grabar et al. 1978; Genequand 2006; Genequand 2012, 95–159. 4. Genequand 2012, 154–159. 5. See preliminary field reports: Genequand et al. 2008; Genequand et al. 2010; Genequand and Studer 2011; Genequand and al-Asʿad 2012. 6. For a more detailed account on the pottery, see Genequand et al. 2010, 202–205. 7. See, for instance, some of the elongated rooms on the side of the peristyle houses in the large enclosure at Qaṣr al-Ḥayr al-Sharqī (Grabar et al. 1978, 58),
721
172
D enis G enequand
the storerooms of the oil presses in the same building (Genequand 2012, 322), or the storeroom of al-Faddayn (Genequand 2012, 276–277). 8. Sauvaget 1939; Genequand 2012, 142–149, 359–360. 9. For more details and photographs of these stucco decorations, see Genequand 2011. 10. Hamilton 1959, 237–239, pl. XXXVI 3 and 7. 11. There is, however, no indication that this panel or the others were fully painted. Apart from the upper eyelids, lips, and beard of the rider, the only other evidence for painting are the black eyelids, red lips, red nails, and red palm leaves on and around a female figure from another panel; Genequand 2011, 370–371. 12. Schlumberger 1946–1948; Schlumberger 1986, pl. 34. 13. Harper 1981, 40–98, passim. It should be noted here that some falconry scenes, on foot or riding, are attested from the second and first millennium bc in Anatolia and Mesopotamia, during Roman times in Italy and Gaul, and during the early Byzantine period on mosaics in Greece (Argos), North Africa (Carthage and Tabarqa), and the Near East (Mādabā). 14. Wilkinson 1986, 205–214. 15. Wilkinson 1986, 209, 212, fig. 2.45. 16. Wilkinson 1986, 209, 213, fig. 2.46. 17. Hamilton 1959, 237–239, pl. XXXVI 1–3. 18. Hamilton 1959, 230, pl. XXXVI 6–7. 19. Hamilton 1959, 232–233, pl. LVI 1. 20. Hamilton 1959, 228, pl. XXXVI 11. 21. Foote 1999, 425–427, figs. 5–7; Foote 2012 (Cat. 153 A, B, and C). 22. Goldman 1999, 26–27 (A.3), 29–30 (A.12), 31–33 (A.13), 40–41 (C.2), 45 (C.11). 23. Callieri 2006 (graffiti 2 to 6). 24. Wilkinson 1986, 209, fig. 2.42. 25. Godard 1962, 201–243, passim; Harper 1981, 40–98, passim. 26. Schlumberger 1986, pl. 34. 27. Bishop 1988. 28. Hamilton 1959, 237–239, pl. XXXVI 1–5. 29. ʿAbd al-Ḥamīd ibn Yaḥyā 1937 (Risāla ʿan Marwān ilā ibnihi ʿAbd Allāh). 30. Ghiṭrīf ibn Qudāma al-Ghassānī 2002 (Kitāb ḍawārī al-ṭayr). 31. See especially the introduction by F. Viré and D. Möller in Ghiṭrīf ibn Qudāma al-Ghassānī 2002, 1–39. 32. Schlumberger 1986, 15, pl. 64a. 33. Hamilton 1959, 228–232, pl. LV 1 and 5. 34. Foss 2008, 66–81; Heidemann 2010, 175–176, fig. 21–24. 35. Heidemann 2010, 175. 36. Foote 1999, 425–427, figs. 5–7; Foote 2012 (Cat. 153 B and C). 37. Harper 1981, 99–122, figs. 25–27, 43, pl. 19, 33–36. 38. Genequand 2012, 364–368.
713
Two Possible Caliphal Representations
173
39. On the meaning of the word madīna in the inscription and on Umayyad new urban settlement, see Genequand 2012, 154– 159, 242– 248; and also Northedge 1994. 40. On the details of these, see Genequand 2011, 357–359. 41. On the economic and especially agricultural function of the Umayyad aristocratic settlements, see Genequand 2012, chapter 11.
Bibliography Bishop, M. C. 1988. “Cavalry Equipment of the Roman Army in the First Century ad.” In Military Equipment and the Identity of Roman Soldiers: Proceedings of the Fourth Roman Military Equipment Conference (BAR International Series 394), edited by J. C. Coulston, 67–195. Oxford: B.A.R. Callieri, P. 2006. “At the Roots of the Sasanian Royal Imagery: The Persepolis Graffiti.” In Ērān ud Anērān. Studies Presented to Boris I. Maršak on the Occasion of His 70th Birthday, edited by M. Compareti, P. Raffetta, and G. Scarcia, 129–148. Venice: Cafoscarina. Foote, R. M. 1999. “Frescoes and Carved Ivory from the Abbasid Family Homestead at Humeima.” Journal of Roman Archaeology 12: 423–428. ———. 2012. “An Abbasid Residence at al-Humayma.” In Byzantium and Islam, Age of Transition, 7th–9th Century, edited by H. C. Evans and B. Ratliff, 221–222. New York: Metropolitan Museum of Art. Foss, C. 2008. Arab- Byzantine Coins: An Introduction, with a Catalogue of the Dumbarton Oaks Collection. Washington, DC: Dumbarton Oaks. Genequand, D. 2006. “Umayyad Castles: The Shift from Late Antique Military Architecture to Early Islamic Palatial Building.” In Muslim Military Architecture in Greater Syria: From the Coming of Islam to the Ottoman Period, edited by H. Kennedy, 3–25. Leiden: Brill. ———. 2008. “The New Urban Settlement at Qasr al-Hayr al-Sharqi: Components and Development in the Early Islamic Period.” In Residences, Castles, Settlements: Transformation Processes between Late Antiquity and Early Islam in Bilad al-Sham, edited by K. Bartl and A. R. Moaz, 261–285. Rahden, Westf.: Verlag Marie Leidorf. ———. 2011. “Les décors en stuc du bâtiment E à Qasr al-Hayr al-Sharqi.” Syria 88: 351–378. ———. 2012. Les établissements des élites omeyyades en Palmyrène et au Proche-Orient. Beirut: Institut Français du Proche-Orient. Genequand, D., R. R. Ali, M. M. Haldemann, J. J. Studer, and A. Vokaer. 2010. “Rapport préliminaire des campagnes 2008 et 2009 de la mission archéologique syro-suisse de Qasr al-Hayr al-Sharqi.” SLSA Jahresbericht 2009: 177–219. Genequand, D., H. Amoroso, M. M. Haldemann, D. Hull, M. Kühn, and J. J. Studer. 2008. “Rapport préliminaire des travaux de la mission archéologique syro-suisse à Qasr al-Hayr al-Sharqi en 2007.” SLSA Jahresbericht 2007: 141–178.
741
174
D enis G enequand
Genequand, D., and W. al-Asʿad. 2012. “Rapport préliminaire des travaux de la mission archéologique syro-suisse à Qasr al-Hayr al-Sharqi en 2011.” SLSA Jahresbericht 2011: 55–82. Genequand, D., and J. J. Studer. 2011. “Rapport préliminaire des travaux de la mission archéologique syro-suisse à Qasr al-Hayr al-Sharqi en 2010.” SLSA Jahresbericht 2010: 55–80. Ghiṭrīf ibn Qudāma al-Ghassānī. 2002. Traité des oiseaux de vol (Kitāb ḍawārī aṭ-ṭayr). Translated by F. Viré, B. Abeele, and D. Möller. Nogent-le-Roi: J. Laget. Godard, A. 1962. L’art de l’Iran. Paris: Arthaud. Goldman, B. 1999. “Pictorial Graffiti of Dura-Europos.” Parthica 1: 19–105. Grabar, O., R. Holod, J. Knustad, and W. Trousdale. 1978. City in the Desert: Qasr al- Hayr East. Cambridge, MA: Harvard University Press. Hamilton, R. W. 1959. Khirbat al-Mafjar: An Arabian Mansion in the Jordan Valley. Oxford: Clarendon. Harper, P. O. 1981. Silver Vessels of the Sasanian Period, Vol. I: Royal Imagery. New York: Metropolitan Museum of Art. Heidemann, S. 2010. “The Evolving Representation of the Early Islamic Empire and Its Religion on Coin Imagery.” In The Qurʾan in Context: Historical and Literary Investigations into the Qurʿanic Milieu, edited by A. Neuwirth, N. Sinai, and M. Marx, 149–195. Leiden: Brill. Kröger, J. 1982. Sasanidischer Stuckdekor: ein Beitrag zum Reliefdekor aus Stuck in sasanidischer un frühislamischer Zeit nach den Ausgrabungen von 1928/29 und 1931/32 in der sasanidischen Metropole Ktesiphon (Iraq) und unter besonderer Berücksichtigung der Stuckfunde von Taḫt- i Sulaimān (Iran), aus Nizāmābād (Iran) sowie zahlreicher anderer Fundorte. Mainz: P. von Zabern. Northedge, A. 1994. “Archaeology and New Urban Settlement in Early Islamic Syria and Iraq.” In The Byzantine and Early Islamic Near East II, Land Use and Settlement Patterns, edited by G. R. D. King and A. Cameron, 231–265. Princeton: Darwin Press. Sauvaget, J. 1939. “Les ruines omeyyades du Djebel Seis.” Syria 20: 239–256. Schlumberger, D. 1946–1948. “Deux fresques omeyyades.” Syria 25: 86–102. ———. 1986. Qasr el-Heir el-Gharbi. Paris: Librairie Orientaliste P. Geuthner. Vibert-Guigue, C., and G. Bisheh. 2007. Les peintures de Qusayr ʿAmra. Un bain omeyyade dans la bâdiya jordanienne. Beirut: Institut Français du Proche-Orient. Wilkinson, C. K. 1986. Nishapur: Some Early Islamic Buildings and Their Decoration. New York: Metropolitan Museum of Art.
715
7
Umayyad Palace Iconography On the Practical Aspects of Artistic Creation Nadia Ali and Mattia Guidetti*
Q U Ṣ AY R ʿ A M R A A N D K H I R B AT A L -M A F J A R
are among the most representative early Islamic sites built on the semi-arid fringes of the eastern Mediterranean during the eighth century.1 Quṣayr ʿAmra is located in the basalt desert known as the Balqāʾ, in modern Jordan. Its fame stems from the wall paintings that decorate the bathhouse and audience hall (ruins of other structures, including a mosque and caravanserai, stand nearby).2 Quṣayr ʿAmra has been convincingly ascribed to the Umayyad prince al-Walīd ibn Yazīd (708–744) before he became caliph (r. 743–744).3
This contribution is the result of collaboration. Nadia Ali is the author of the section devoted to Qusayr ʿAmra, Mattia Guidetti of the section devoted to Khirbat al-Mafjar. The Introduction and Conclusions are authored by both. This essay was planned during participation in the conference “From Influence to Translation: The Art of the Global Middle Ages,” held at the University of Edinburgh in June 2012. Previously the two authors had the chance to meet and discuss their respective works on early Islamic art thanks to the Aga Khan fellowship (2008) and more recently (2011) thanks to the projects “Art, Space and Mobility” (Kunsthistorisches Institut in Florence) and “Connecting Art Histories” (Pergamon Museum, Berlin). We are deeply grateful to all those institutions and scholars who have helped us with our respective research in recent years. Mattia Guidetti would like to thank Daniela Massara, Steven Judd, John Mitchell, Nathan Dennis, and Christopher Busuttil for their help in the course of his research. *
761
176
N adia A li and M attia G uidetti
Discovered by Alois Musil in 1898,4 the building was used as a pleasure retreat and a hunting lodge.5 Khirbat al-Mafjar is a complex of buildings and infrastructure serving different functions located near Jericho, in present-day Palestine. The monumental part of the site consisted of a palace, a mosque, a bathhouse, and a courtyard with a domed fountain.6 Around this core of buildings a vast area devoted to more modest residences and agricultural activities suggests the site was a settlement and not only an elite residence as previously thought.7 The complex has been associated with the time of Hishām ibn ʿAbd al-Malik (r. 724–743) because of the discovery of an ostracon— a ceramic fragment reused for writing—bearing his name. Even though several art historians, basing themselves on iconographical interpretations, have ascribed the site to the patronage of al-Walīd ibn Yazīd, recent archaeological surveys point to the likelihood of an earlier foundation date. Finally, although the palace was left unfinished, the site remained in use long after the fall of the Umayyads.8 From an art-historical point of view, these residences contain an invaluable corpus of images produced in the early Islamic period, with figurative and nonfigurative programs rendered with a variety of techniques ranging from stucco through mosaic and painting to stone carving. The formative period of Islamic culture in which these programs were produced and the variety of their themes and techniques have stimulated the work of numerous scholars. The general tone of the interpretations offered by art historians has recently been summarized in The New Cambridge History of Islam as follows: “The opulent decoration encountered on sites such as Khirbat al- Mafjar and Quṣayr ʿAmra offers intriguing insights into the culture of Umayyad princely pleasure known otherwise from textual sources.”9 This concise statement sums up two tendencies that have dominated modern understanding of Umayyad secular art: the focus on the princely cycle (first put forward by Oleg Grabar) and the importance of written sources in order to better define the nature of princely representation in these sites (an approach whose champion has undoubtedly been Robert Hamilton). Textual sources on the early Islamic period are, however, problematic: most texts are often dated later than the Umayyads and informed by the hostile perception of the first Islamic dynasty by later rulers. What is more, they focus only on the élite. It must be said that some of these challenges have been mentioned as caveats by way of an introduction to their study.10 At the same time, not only the nature but also the very use
717
Umayyad Palace Iconography
177
of written sources to decipher “Umayyad iconography” presents some drawbacks. The sources on the Umayyad elite are for the most part poetical compositions written during the later Abbasid period. The texts mainly focus on or are ascribed to al-Walīd ibn Yazīd (al-Walīd II, r. 743–744), describing his debauchery and love for wine and promiscuous parties organized in his palaces, as well as the disputes he had with other members of the ruling dynasty. Although al-Walīd ibn Yazīd might have been involved in the planning of one or more of these Umayyad residencies, some art historians have tended to project every single detail contained in the textual sources onto the interpretation of the iconography. This tendency could be defined as overinterpretation. Because al-Walīd ibn Yazīd emerges from the sources, to repeat the words of Robert Hamilton and Robert Hillenbrand, as “the most dedicated playboy of the age,” “extravagant, imaginative, witty, and eccentric,” “long thwarted, an aesthete, and an unrepentant hedonist,” this has often been the cue followed in the interpretation of the decorative programs.11 Despite the fact “these palaces are neither named nor precisely located in the texts,” as noted by Hillenbrand,12 the extant Umayyad residences with their bath complexes have been interpreted as the most likely “hedonistic setting” of the anecdotes and “orgiastic scenes narrated by Walīd’s biographers.”13 Art-historical literature on Umayyad complexes therefore appears to be dominated by a macroscopic paradox: despite the fact there is no certainty about the patrons of many of these sites and their precise dates of foundation, the iconographic interpretations offered by most scholars are incredibly rich in anecdotes and personal details.14 With regard to al-Walīd ibn Yazīd, it should be mentioned that his figure—as portrayed by Abbasid sources—has recently been re-scrutinized. Judd argues that al-Walīd ibn Yazīd was a much more complex figure who, among other things, contributed in developing “Islamic theories of government.”15 He did share with several other members of world elites the pleasures of courtly life (such as poetry, singing, wine, and women), yet he also stands out for his speculations on the relationship between political and religious authority. Beside an overemphasis on the controversial figure of al-Walīd ibn Yazīd, what is striking in most analyses of Umayyad palatial art is the exclusive primacy given to texts over other factors—related to the activities of craftsmen and the artistic context—which most likely have also
781
178
N adia A li and M attia G uidetti
affected the creation of the decorative programs of early Islamic palatial residences. This last point becomes clear once one expands the analysis of modern studies on early Islamic secular art beyond Islamic art scholarship. On the one hand scholars of the Roman and early Christian periods acknowledge early Islamic art as the last episode of Late Antiquity. Their evidence mainly lies in the formal qualities of the art of the Umayyads. The final chapters of their works are often devoted to Islam: the fact that they cannot properly handle written sources and historiographical issues related to the early Islamic period tends to stop them from fully including early Islamic art within their narrative.16 On the other hand Islamic art historians, despite acknowledging the debt of Umayyad art to Late Antiquity, desperately search for a new or Islamic meaning in order to attribute as much agency as possible to early Islamic patrons and conversely to Islam.17 This up to the point of completely forgetting the art-historical context of some works—as in the case of the alleged “portrait of al-Walīd’s family” at Khirbat al-Mafjar and the vignettes of Quṣayr ʿAmra’s vault, both of which will be discussed below. Important exceptions do exist: although some of his interpretations may be seen as overstretched, Garth Fowden has convincingly reassessed the evidence from Quṣayr ʿAmra both from a Late Antique and an early Islamic perspective. Similar remarks apply to the work devoted by Hana Taragan to some aspects of the decoration at Khirbat al-Mafjar, in which formal analysis plays a distinctive role.18 More generally, in the last decades scholars have increasingly addressed the theme of continuity between Late Antiquity and early Islam.19 The present essay stems from a dissatisfaction with iconographic interpretations largely based on a forced overlap between texts about the Umayyad elite, hypothetical patrons, and early Islamic art. By focusing on the paintings on the central vault of the reception hall at Quṣayr ʿAmra (Jordan) and the floor mosaics of the bathhouse attached to the settlement of Khirbat al-Mafjar (Palestine), both sections (the first by Nadia Ali, the second by Mattia Guidetti) deal with the previously underrated role of the logic of craftsmen’s practice in the making of early Islamic art.20 The main argument is that the Umayyad agenda, if any, does not fully or necessarily determine the final iconography; and that between the intended program and the final decoration, the logic of artistic practice and forms might have played a role. This reverses the approach of those who have insisted on the peculiarities or singularities of Umayyad palatial art.
791
Umayyad Palace Iconography
179
To give priority to artistic practice means to start the inquiry by posing questions such as the following: How were images and programs produced in eighth-century Greater Syria (Bilād al-Shām)? What parameters other than the symbolic ones motivated the choice and distribution of subject matter, and guided the modifications of inherited formulas in the early Islamic program? What was transmitted from one generation of craftsmen to the next? In the 1930s, the Austrian art historian Otto Pächt developed the concept of “formal opportunity” according to which the structure of a work of art might be governed in essential ways by the function it performed within a wider whole. The demands of this whole impinge on the artist in the form of a brief to be met… . To see the finished work of art as the fulfillment of a brief—this is what we do when we try to see it in terms of its formal opportunity. … Formal opportunity takes us back to the point of departure of the creative process: it makes us think about some at least of the factual circumstances in which the work came into existence.21 The notion of “formal opportunity” relates the arrangement of pictorial contents to the architectural surface and suggests to reconsider the role of artists’ habitual working practices.22 Following this, the two sections converge to assert that the architecture and function of each building offered the makers of the decoration a preestablished set of options to work with. The existence of a catalogue of sorts, a body of images regarded as suitable for a specific architectural space, makes the appearance of certain themes recognizable and to a certain extent predictable by Late Antique standards.
An Agricultural Calendar in Quṣayr ʿAmra In this section, the frescoes from Quṣayr ʿAmra’s central vault will be reexamined. The author will focus on this material because it is a key example of traditional scholarship on Umayyad palace iconography. This model assumes, first, the prevalence of the so-called princely cycle and, second, a novel and specifically Umayyad-Arab recombination of Greco-Roman motifs with Iranian themes into a new syntax. While this model certainly contains elements of truth, re-examining the iconography in light of the craftsmen’s practice will allow us to challenge both assumptions.
801
180
N adia A li and M attia G uidetti
Princely Cycles in Umayyad Art, or “the Perpetual Oriental Dinner”: Old Stereotypes and Dubious Assumptions We should first survey the prevalent position, initially formulated in the 1950s and still echoed in recent studies. In the second half of the twentieth century, a need was felt to counterbalance earlier Roman-centric and Orientalist views by defining the parameters of early Islamic art and establishing its distinct visual and cultural identity.23 The most vital contribution in this direction came from Oleg Grabar, who developed a strongly influential methodological framework. From his doctoral dissertation (1955) to “Umayyad Palaces Reconsidered” (1993),24 he attempted to reinterpret the meaning of Umayyad art by reconsidering it as a visually compelling vehicle for Umayyad needs and aspirations, and by conceptualizing the process involved in the formative stages of Islamic art. Grabar’s central assumption was that Umayyad palatial iconography was mainly dominated by the so-called princely cycle characteristic of Near Eastern aristocratic life.25 He described the cycle as follows: “It consists of a number of representations of princes… . Besides representation of the prince, we meet illustrations of the pastimes: hunting, dancing, music-making, nude or half-clad women, games, acrobats, gift bearers are themes in all three major buildings.”26 His argument is mostly based on texts and on the nature of the motifs exhibited in the central hall of Quṣayr ʿAmra, the inner and outer façades of Qaṣr al-Ḥayr al-Gharbī (728) and the bath-porch of Khirbat al-Mafjar (728–743). Grabar explains the meaning of the decor of Quṣayr ʿAmra as follows: “The art of Quṣayr ʿAmra expresses the unique, personal desires of a prince. It is through a deeper perception of the private character of early Islamic princes that the monument’s elucidation may come about.”27 Many factors may have encouraged the idea of a princely cycle, including the debate over the hedonistic character of the desert castle28 and the bias of Abbasid textual sources29 mainly emphasizing Umayyad immorality.30 Grabar’s work on Umayyad art was part of his broader inquiry into the genesis of Islamic art. The most important problem in Umayyad iconographic studies, as explained in his foundational study The Formation of Islamic Art, is “to decide how these pre-Islamic subject-matters were understood when they were made, why they were made and whether they were but accidental collections of motifs or significant and conscious accumulation of subjects in the process of creating a new aesthetic and material vision.”31 Grabar proposes that the answers to these questions will be found
8 11
Umayyad Palace Iconography
181
by identifying and explaining three elements: “the mind of the Muslim user and beholder, the meanings given to the artistic creations and the forms utilised by him.”32 Grabar’s hypothesis that a princely cycle was borrowed from Near Eastern motifs of power and luxury illustrates his thesis that there was syntactic change in the formation of Islamic art. He refers to “a series of transfers of forms and of meanings, as new sponsors and new resources contribute to the recomposition of older forms, on the one hand, and to the development of new meaning on the other. The most obvious example is the ceremonial pastime which acquires new dimensions.”33 Concentrating on the beholder or the patron’s mentality as the true engine of change and meaning, Grabar aims to establish that a Muslim onlooker might have beheld images in a new and distinctive way. Thus, he replaces the traditional assumption of passive reception by a new assumption of selective appropriation and creative reception. The need for a contextualized interpretation, and the emphasis on the “mind of the beholder,” has tended to lead to a focus on interpretation and meaning and has promoted interdisciplinary approaches in which written sources play a central role.34 In the same vein, Fowden’s recent monograph on Quṣayr ʿAmra represents the most brilliant attempt to situate the program in its Late Antique35 and Arab-Syrian context and attribute the building to al-Walīd II. Fowden’s attribution has been recently confirmed by an inscription located on the south wall of the west aisle and mentioning the name al- Walīd ibn Yazīd.36 Using a wide array of Arabic texts, he connects the wall paintings with the patron’s life and attempts to reconstruct al-Walīd’s self- representation. His classification of subjects echoes Grabar’s (women, entertainment, hunting, princely patron, and myths), but he adds new meanings. Even though the relationship between texts and images is mostly argued very convincingly,37 there are other possible interpretations, all supported by equally convincing evidence.38 There is no need to impose a single interpretation on the iconography, but taking refuge in texts and locating the program’s genesis in the realm of an “Umayyad mind” or “the patron’s personality” alone may sometimes make it less susceptible to analysis because we have no firm evidence concerning the patron’s intention or the audience’s nature and reception of the forms inherited from Late Antiquity.39 One revealing example is Fowden’s speculation on the Arab-Umayyad contribution to the Quṣayr ʿAmra program. After drawing parallels between the themes in the program (hunting, entertainment, and women) and the subjects of a qaṣīda (the classical Arabic poem), he says, “Quṣayr ʿAmra’s most characteristic features is the combining of
821
182
N adia A li and M attia G uidetti
Graeco-Roman with Iranian motifs and visual styles and the choosing of themes congenial to Arab taste and circumstances.” To illustrate his view, he observes: “Nowhere did this coincidence of Quṣayr ʿAmra’s emphases with those of the qaṣīda come out more clearly than in the chapter of the hunt; and one might add that conversely, no area is more comprehensively ignored by both [the decoration and the qaṣīda]—while abundantly alluded to in Late Roman art—than that of agricultural activity, which held no charm for Arabs. There are no calendars of months in Umayyad art.”40 The present discussion, however, aims to show that what has been taken as a “princely cycle” in fact represents an agricultural calendar, thus providing a striking case of how the explanatory power of texts has sometimes been overestimated at the expense of the continuities of artistic practice and repertoire.41
The Paintings of Quṣayr ʿAmra’s Central Vault in the Making In this first part, we re-examine the scheme of composition or layout applied to the paintings of Quṣayr ʿAmra’s central vault in the reception hall (fig. 7.1) by asking two elementary questions. First, to what extent does
Figure 7.1 View of the audience hall, Quṣayr ʿAmra, Jordan.
813
Umayyad Palace Iconography
183
Figure 7.2 Drawing of the organizational patterns of wall surfaces, Quṣayr ʿAmra (after Vibert-Guigue and Bisheh 2007, pl. 14a).
the architecture—a barrel vault (7.5 meters long and 4.5 meters wide)— determine the scheme of composition applied to the surface? And, second, to what extent does this scheme motivate the choice of subject matter or to what extent is the pictorial content subordinate to architecture and scheme of composition? The scheme of composition applied to the frescoes of Quṣayr ʿAmra’s central vault is a grid pattern made of thirty-two square panels (one meter per side) of which twenty-eight are complete and four on the northern part are incomplete,42 cut off at the sides by the north wall of the building (figs. 7.2 and 7.3).43 The grid is underlined by five horizontal and seven vertical friezes of rosettes. The grid pattern appears to be among the earliest and most common methods of maintaining the decorative unity of a decor by imposing on the surface a geometric organization in which small panels are repeated to give a regular rhythm. It is well suited to the needs of flat or barrel-vault ceilings since it avoids hierarchical emphasis on any specific areas of the surface and the orientation of the figures can be adjusted to the viewpoint of the spectator. As Alix Barbet suggests in her study of Roman painting, the grid as a compositional principle was inspired by the architecture of coffered ceiling and stucco coffering.44 Whether truly illusionist in rendering the solidity of architecture or reduced to mere outlines, the grid had often
841
Figure 7.3 Drawing of the organizational pattern applied to the ceiling decoration of the central span, Quṣayr ʿAmra (after Vibert-Guigue and Bisheh 2007, pl. 48).
815
Umayyad Palace Iconography
185
Figure 7.4 Coffered ceiling, tomb of Elahbel, Palmyra, 103 ad.
been used for the decoration of vaults and ceilings and recurs throughout Roman painting and sculpture in the Mediterranean. For example of stucco-coffered ceilings in the Levant, we can cite the multiple-storied tomb structures in Palmyra in Syria. Two of these seem particularly relevant: the tomb of Jamlichus, dated 83 ad, and the tomb of Elahbel, dated 103 ad (fig. 7.4). In both cases, the tomb chamber had stucco coffering and the design shows a framing of rosette friezes,45 recalling the framing elements of our Umayyad decorative scheme. For late Roman painting, one of the most famous examples of a gridded ceiling is to be found in a basilica belonging to the period of Constantine underlying the Romanesque cathedral of Trier and dated 326 ad. Underneath the sanctuary of the north basilica was discovered a large rectangular room with painted walls and ceilings (fig. 7.5). The ceiling imitated coffers, of which eight panels have been recovered. The surface is defined by a framework of broad red bands forming rectangular panels. To these bands is attached a twisted rope of gold with gold rosettes at the intersections. The panels are alternatively rectangular and square and contain busts of the Seasons and pairs of winged putti.46 Other evidence of the widespread use of gridlike arrangement applied on painted ceilings is found in Syria and Jordan. For instance, there are
861
186
N adia A li and M attia G uidetti
Figure 7.5 Ceiling frescoes, Trier, Aula Palatina, 326 ad.
Figure 7.6 Grid pattern applied to ceiling decoration, Tomb of Loukianos, Jordan, third century (after Barbet and Vibert-Guigue 1994, pl. 11).
a number of painted tombs in Jordan with pre-Islamic examples of barrel vaults: for example, the third-century Tomb of Loukianos, where the decorative scheme of the ceiling consists of a simulated lattice work with a system of eighteen square panels framed by simple bands (fig. 7.6).47 In the mid-third-century synagogue Dura Europos on the Euphrates, the
817
Umayyad Palace Iconography
187
Figure 7.7 Grid pattern applied to the ceiling of the synagogue, Dura Europos, third century.
basic framework of the painted decoration was established by four great pilasters in the corners that served as fictional supports for a ceiling of roof tiles, suggesting figurated coffers with cosmic imagery of fertility and abundance, land and sea animals, fruits, flowers, and zodiacal signs (fig. 7.7).48 Although seemingly far removed from our inquiry into ceiling decoration, the surviving Christian mosaics from the Bilād al-Shām provide the most interesting comparison point, not only from a geographical or historical point of view, but also from a technical one.49 Of particular interest is the grid pattern in the pavement of the Church of the Holy Apostles at Madaba in Jordan (578 ad). Other examples can be found in the mosaics of the eighth-century Church of St. Stephen at Umm al-Rasas and the fifth-century Church of St. Peter, near Jerash (fig. 7.8). In the latter, the pavement of the nave consists of a grid pattern made up of two rows of seven panels.50 It is worth noting two other similarities with the Quṣayr ʿAmra ceiling: the grid is applied to the building’s central axis and the surface is divided into panels underlined by friezes of interlaced circles and ellipses. Another example is the rich third-century Roman dwelling at al-Djem (Tunisia), which shows a floor mosaic organized by a grid pattern
881
188
N adia A li and M attia G uidetti
Figure 7.8 Drawing of floor mosaics, Church of St. Peter, Jordan, sixth century (after Piccirillo 1993, fig. 633).
Figure 7.9 Calendar, floor mosaic, House of the Months, al-Djem, Ancient Thysdrus, Tunisia, third-fourth century (Dunbabin 1999, fig. 113).
of twenty-eight square panels framed by friezes of rosettes and medallions (fig. 7.9).51 Many scholars have argued that the systems of composition used in mosaics had their origin in the painted or stuccoed decoration of vaults and ceilings.52 This allows us to use the mosaics as indirect evidence in our argument in favor of the correspondence between grid-pattern and painted-ceiling decoration.
891
Umayyad Palace Iconography
189
Figure 7.10 Calendar, floor mosaics, Villa of the Falcon, Argos, Greece, sixth century (Dunbabin, 1999, fig. 233).
It is not my intention here to discuss the architectural origins of the schemes of composition used on ceiling paintings.53 Rather, I would like to highlight a long tradition of gridded ceiling (stuccoed or painted) and a tradition of gridded mosaics imitating now lost ceiling decoration in the eastern Mediterranean. In my view, these examples add up to suggest that the gridlike arrangement of the pictorial content in Quṣayr ʿAmra’s vault was chosen for the specific surface of the vaulted ceiling because this was the pattern traditionally preferred for ceiling decoration. Did this organizational pattern influence the choice of subject matter depicted in Late Antique floor mosaics? In other words, does a certain type of frame carry with it a certain predictable content? Although many decorative themes are to be found in mosaics with a gridlike arrangement, a dominant position is occupied by the representation of the months of the year, as is apparent from examples found in North Africa (Djem) (fig. 7.9),54 Greece (Villa Argos55 and the Basilica of Tegea56) (figs. 7.10–7.11), Gaul (Saint-Romain-en-Gal), and the Levant. Of particular interest are the Christian mosaics from Jordan. For instance in Jerash, the months were personified in two mosaic pavements: one in a chapel in the sixth- century Jerash cathedral and the other in St. John’s cathedral. Although the figures in both cases have been defaced, the inscriptions giving their Macedonian names make it clear that they were personifications of the months.57 However, the important point is that, as in the other examples, the two Jerash calendars were organized in a grid pattern forming square panels. A more legible example of the grid-pattern use in
901
190
N adia A li and M attia G uidetti
Figure 7.11 Calendar, floor mosaics, Basilica of Tegea, Greece, sixth century (Maguire 2012, fig. 1.2).
the Christian mosaics of Jerash is found in the sixth-century Chapel of SS. Elias, Maria and Soreg (figs. 7.12–7.13). The mosaic pavement occupies the central nave of the church, paralleling the central position of our Umayyad decoration. The greater part of the nave is occupied by a large field subdivided into squares, set in a framework of interlinking circles and ovals, all of which are enclosed by an acanthus border. Of the twenty panels in the field, twelve contained personifications of the months with their Greek names; the rest are scenes from daily life. Each horizontal row of the calendar is to be read from north to south. Of equal interest is the great popularity of the calendar theme in Middle Eastern mosaics
9 11
Umayyad Palace Iconography
191
Figure 7.12 Drawing of the calendar in the Chapel of SS. Elias, Maria and Soreg, Jordan, sixth century (after Saller 1949, fig. 17).
Figure 7.13 Diagram indicating the position of the motifs in the calendar, Chapel of SS. Elias, Maria and Soreg, Jerash, sixth century (Saller 1949, fig. 46).
before Islam,58 as exemplified by the sixth-century mosaics of al-Awzāʿī59 and Qabr Hiram in Lebanon.60 In our view—and at the risk of our own oversimplification—this pattern type suggests that it was associated with specific content, leading one
921
192
N adia A li and M attia G uidetti
to expect a calendar in Quṣayr ʿAmra’s central vault. This probability is strengthened by the fact that the calendar is often located in the central nave in our Christian pavements. Moreover, we have noted the presence of an arch in each panel of the Umayyad grid. Although it is true that in the mosaics discussed above, none of the personifications of the months was placed under an arch, it is easy to find examples of calendars showing the allegories set in an architectural framing, for example in a Coptic sixth-century relief61 and most importantly the famous Codex-Calendar of 354 (fig. 7.14).62 In addition, a sixth-century Byzantine text by Cosmas Indicopleustes describes the painted decor of the bath constructed by
Figure 7.14 Calendar of 354, copy of Peiresc, the Vatican Library, seventeenth century (Levi 1941, fig. 1).
913
Umayyad Palace Iconography
193
the Emperor Constantine in Constantinople (now lost). According to Indicopleustes, the Byzantine bath was decorated with seven niches (enthekas) containing depictions of planets and twelve other niches with personifications of the months.63
Iconographical Analysis Thus at this point, it is necessary to establish the existence of the Umayyad calendar by making an iconographical analysis.64 For the description, I used the reproduction of the paintings published by Claude Vibert- Guigue and Ghazi Bisheh in 2007 (Plate 13 in color insert),65 which was made before the Italian campaign of restoration.66 I will focus on the most legible panels (D6, D7, C4, A1, B1, B3, D2, D3, D1, C2, D8, A8, A4, A2, A7, A6) and group them in four thematic categories that usually appear in the iconography of the calendar, namely, the official images, the seasonal activities, the zodiacal sign, and depictions of Christian saints. The main objective here is to provide sufficient clues for the identification of a calendar and to begin to examine its nature.67 Our description begins with the panel located in the northwestern part of the ceiling (D6) (Plate 13). It shows a standing male figure in a frontal pose. The costume, attribute, and gesture strongly suggest that this may be the personification of January. He wears a long-sleeved tunic covered by a mantle (pallium) draped over his left arm and reaching down to the ground. The mantle is attached to the right shoulder by a fibula. This costume recalls the traditional costume of the Byzantine consul, as frequently found on Byzantine diptychs;68 more interestingly, there are also several depictions of consuls found in the calendar iconography (Chronography of 354, Argos, Jerash). The figure raises his right hand, holding a “handkerchief,” which might be the traditional mappa raised by the consuls to signal the opening of the games.69 This type of iconography was used to personify the month of January in the calendars of Argos (figs. 7.10, 7.15) and Jerash (fig. 7.16). In addition, the left arm is almost concealed by the mantle with only the hand appearing. In Byzantine representations of consuls and in the personifications of January, this configuration usually symbolizes the consul’s sparsio (donation), which, as the tradition required, he distributes at the moment of the festivities organized in January for his nomination.70 This practice can be observed on the mosaic of Argos where January is a consul holding the mappa and having at his feet a bag of coins to be distributed.
941
Figure 7.15 Calendar of Argos, detail showing the personification of January, sixth century (Akerstrom-Hougen 1974, 75).
Figure 7.16 Calendar of Jerash, Chapel of SS. Elias, Maria and Soreg, sixth century; detail showing the personification of January compared (Saller 1949, pl. 46).
915
Umayyad Palace Iconography
195
The next compartment (D7) shows a man also in a Byzantine costume and in a frontal pose (Plate 13). Contrary to what Claude Vibert-Guigue has argued, I think that the man is not standing, but sitting. Although the seat is no longer discernible, it is clear from the widening of the lower part of the figure, which is characteristic of many representations of enthroned consuls. But more importantly, it recalls the position of January in the Jerash calendar. Moreover, the gesture of the Umayyad figure with the right hand on the chest and the left hand resting on the knee is very common to a great number of Byzantine representations of consul and emperor as it can be observed for instance in the official images of The Codex-Calendar of 354. The hypothesis of a consul or an emperor could be strengthened by the fact that the depiction of Constance enthroned under an arch is painted on the page opposite another page showing a standing consul (fig. 7.17).71 Thus, they form a collocation, that seems to have survived in the Umayyad decor in the visual contiguity of D6 and D7. The presence of these two official images in our decor and their iconographic affinities with the personification of January are a first argument in favor of the calendar hypothesis. At this point, it may be useful to keep in mind that the two destroyed calendars of Jerash both begin with the personification of January.72 This could suggest that the series of months in the Umayyad calendar may also begin with January and that the reading of the panels should begin in its northwestern part.
Figure 7.17 Detail of the Chronography of 354 showing a consul and an emperor, seventeenth century copy of Peiresc (Stern 1953, pl. I).
9 61
196
N adia A li and M attia G uidetti
Evidence to support the calendar thesis may also be provided by the presence of seasonal images. By “seasonal image,” I mean a category of subjects that includes both the personifications of the Seasons, or a related theme such as Gaea (the Earth), and the depictions of months through rural works or allegories. The compartment C4 in the western part of the ceiling shows a half- naked female figure sitting with her legs bent and wrapped in an ample garment, her head leaning on her left shoulder (Plate 13). Her right hand is raised toward the south, with the palm clearly shown, whereas with the left hand she holds a part of the garment that covers her legs, thus forming a sort of large fold at the level of the belly. This costume and gesture are typical of the personifications of Gaea shown usually with the fold of her mantle full of fruits,73 but also of the personifications of Seasons as illustrated by the mosaic of Carthage (fig. 7.18).74 This North African document shares many similarities with our panel: the square framing, the gesture of the Seasons with legs folded, the half-nudity and the plastic treatment of the breast characterized by two discs and finally the garment which covers their legs. If in the Carthage mosaic, the different seasons are recognizable through the fruits contained in their garments, we have no remaining trace of any fruits or vegetables in the Umayyad picture. However, the presence of Gaea or a Season in a calendar should not be surprising, given that these themes are visually and semantically connected with calendar iconography (Chronography of 354, Saint-Romain-enGal, Baysan-Scythopolis). Panel A1 houses a half-naked woman sitting (Plate 13). Her torso is in frontal pose while her legs are in profile and covered by a garment or a skirt decorated with a square design. The right leg is folded while the left is outstretched toward the north. With her right arm resting on the folded knee, she holds an object and seems to bring it to her nose. It suggests that the object might be a flower as one can observe on the personification of May in the Chronography of 354, which shows a woman smelling a flower (fig. 7.14). This is further confirmed by her jewels (necklace, bracelets above her elbow), among them a strap crossed over her torso. According to Claude Vibert-Guigue, the strap is made up of a series of petals.75 This conducts us to hypothesize that the flowered strap may allude to the festival of roses (Rosalia) that took place in May throughout the Roman and Byzantine empires.76 The placing of our figure and its relationship to other compartments could strengthen our hypothesis, because, as we shall see further on, the personification of May is followed by what might be an
917
Umayyad Palace Iconography
197
Figure 7.18 Floor mosaic showing Gaea or a Season, Carthage, fourth century (Parrish 1984, pl. 21).
allusion to May and June through the representation of Gemini in A4. The visual association between an allegory of the month and the zodiac signs connected to the month is not unusual as the Chronography of 354 shows.77 Let us take a look at the panels, which allude to rural works (hunting or harvesting). A standing male figure, wearing a cape that covers his back, is set in compartment B3 (fig. 7.19). His right arm is folded over his chest and seems to hold an object nestled in his arm. We can only discern the outlines of wings of birds with the heads down. In his left hand, he is holding a sort of bag or vase with a narrow top and a wider bottom. These
981
198
N adia A li and M attia G uidetti
Figure 7.19 Drawing of the calendar of Quṣayr ʿAmra, detail showing a personification of October? (Vibert-Guigue and Bisheh 2007, pl. 48).
features are comparable to the traditional personification of October.78 For instance, in the Chronography of 354 (fig. 7.20)79 and in the Carthage mosaic,80 October is personified by a nude hunter wearing a cape floating behind his back; he is holding a hare by its legs, head down, in one hand, and in the other a bag into which he is about to put the game. The wings of the birds visible on the Umayyad figure may have been inspired by another traditional iconographic type of October, which associates this month with small-bird hunting.81 The identification of October may be confirmed by the content of B1 and B4. Compartment B1 houses a standing man in a frontal pose (Plate 13). He seems to have his left hand on his chest while the right hand is holding a basket full of round fruits, which may allude to some harvesting activities. In calendar iconography, there are many months that are personified by a man holding a basket of fruits or vegetables. For instance in the Chronography of 354, May is holding sometimes a basket of flowers, September a basket of grapes or of apples, and October a basket of
9 91
Umayyad Palace Iconography
199
Figure 7.20 Chronography of 354, detail showing a personification of October, seventeenth century copy of Peiresc (Levi 1941, fig. 1).
grapes.82 But since B3 clearly represents October, it is not unreasonable to assume that B1 could be a personification of September through a reference to grape or apple harvesting. In the same row, B4 (Plate 13), though deteriorated, shows recognizable outlines of a horseman advancing toward the South and holding something in a gesture that strongly suggests a falconer, similar to the falconer found in the depiction of February in the Jerash calendar of the Chapel of SS. Elias, Maria and Soreg,83 but also in a miniature of Munich (818 ad).84 More interestingly, in the Jerash calendar, the falconer’s panel of February follows the October panel (fig. 7.13).85 It seems that this collocation has survived in the Umayyad decor, as if the Umayyad painters had read the Jerash series of months not by taking each horizontal row read from north to south, but, strangely, by considering the vertical rows of the Jerash calendar. Scenes of harvesting may have been represented in panels D2 and D3 (Plate 13). D2 contains two figures in profile and oriented toward the north, one in the foreground and the other in the middle ground (only
02
200
N adia A li and M attia G uidetti
the legs are discernible). In D2, both men are barefoot, almost naked, and wear only short loincloths. The figure in the foreground seems to be leaning; the arms are held horizontally, the right leg is bent, and the left is raised. The movement of the feet gives the false impression that the figure is a dancer. This may have led previous scholars to interpret this section as a depiction of entertainers and dancers. Another misunderstanding may have been at work in the previous reading of D3, which shows two half-naked men, face-to-face and holding each other with outstretched arms. In this case, scholars have interpreted them as wrestlers, dancers, or erotic figures.86 However, D2 and D3 may allude to grape harvesting and, more precisely, to the pressing of wine. A striking similarity is to be found between the gesture of the men in D3 (who hold each other with outstretched arms) and the gesture of the vintagers on the mosaics of Cherchel,87 Baysan-Scythopolis, and Khirbat al-Mukhayyāt, which show figures pressing the grapes over a large vat (fig. 7.21).88 If our reading is correct, we may have in D2, D3, and probably D1 the representation of vintage activities and the pressing of wine, which is usually associated with the month of September. The distribution of the rural works done in a particular month in several compartments is not unusual; it appears, for example, in the calendar of Saint-Romain-en-Gal. I have suggested earlier that the Jerash calendar organizational pattern and sequence may have had an impact on the Umayyad calendar. Thus, the location of three panels dedicated to September in a row in which we find also two panels alluding to January could be explained by the same
Figure 7.21 Floor mosaic showing vintagers pressing the wine, Church of alHammam, Baysan-Scythopolis, sixth century (after Balty 2003, fig. 5).
201
Umayyad Palace Iconography
201
principle evoked earlier. If we read the Jerash calendar by considering the vertical row read from east to west (instead of taking the horizontal row read from north to south), one realizes that the personification of September is followed by January placed in the same row (fig. 7.13).We find other allusions to harvesting in panels B8, C8, and D8. B8 houses two baskets full of grapes, which, according to rustic menologists, refer to the vintage activities usually associated with September. D8 and A8 show the outlines of two standing male figures wearing a short tunic girded/belted around the waist. In A8, the man is holding a basket while in D8, the figure seems to be throwing small objects on the ground. The short belted tunic and the barely distinguishable forms, which seem to be falling from his hands, resemble the depictions of sowers that usually serve to personify the months of November or December, as illustrated by the mosaic of Baysan-Scythopolis (fig. 7.22).89 This identification of D8 as November or December gains in probability, since the panel that immediately follows D8 is the one identified as January.
Figure 7.22 Drawing of a floor mosaic showing a calendar, Monastery of Lady Mary, Baysan-Scythopolis, sixth century (Levi 1941, fig. 12).
0 2
202
N adia A li and M attia G uidetti
Let us now turn to the compartments that allude to planets and zodiacal signs.90 Panel A4 shows two naked standing figures side-by-side and entwined (plate 13). The right-hand figure is a woman, wearing a square-patterned cape over her shoulders and back, and the left-hand figure is a man wearing a cape extending down to his knees. His right hand is raised, enclosing an undistinguishable object. The number of figures, their nudity, the capes, and their entwined position clearly indicate that we have here a traditional depiction of Gemini comparable to the Gemini represented in the zodiac mosaic discovered on the floor of the third-century synagogue of Beth Alpha in Israel.91 Interestingly, the panel A2 shows two standing figures (Plate 13).92 A woman on the left is wearing a long dress; she is oriented to the north and, with her hand, she seems to point her face. She is flanked by a man holding a rectangular object of which only a few outlines remain. The shape of this object recalls the lyre. The gesture of the woman suggest that she is a dancer, thus recalling the couple of dancers represented in the Jerash calendar, placed in the panel immediately following the personification of May (fig. 7.12).93 On the northern part of the same row, A7 shows two standing figures (Plate 13). The left one is an archer oriented to the south; he holds his bow in his left hand while the right retains the arrow. He is bare-chested and wears a square-patterned cape that falls over his back to knee level. The archer recalls Sagittarius, who is often represented as a centaur shooting an arrow and wearing a goat skin or a floating cape. On certain Late Antique depictions of Sagittarius, we see the centaur raised up on his hind legs (the planisphere of the Vatican,94 or the mosaic of Sepphoris) as in the Umayyad picture. This identification might be strengthened by the presence of a smaller female figure who is represented in the right-hand side of the panel, below the bow. She holds her right arm down beside her body, while the left hand is pulled to the level of her groin. Her nudity and gesture brings to mind Libra, which is often represented in collocation with Sagittarius in some scientific representations of the sky, such as the Vatican planisphere (fig. 7.23). The lute player in panel C2 may also refer to the iconography of the planet Venus, but this interpretation is hypothetical, because, to my knowledge, the depiction of Venus as lute player is not found before the eleventh century. The presence of zodiacal signs and planets in the calendar should not be surprising, because zodiac, calendar, and planets are tightly linked in practice and in iconography. Zodiacal signs appear not only in the Chronography of 354, but also in compositions that show
0 23
Umayyad Palace Iconography
203
Figure 7.23 Sagittarius and Libra, fragment of a miniature from the Vatican Library, 1320 ad (Vat.gr.1087; Stern 1953, pl. XXXVII, 6).
an association between the months, the zodiac, and the planets as in the famous painting of the Ptolemy manuscript.95 Before ending the description of the Umayyad decor, let us look at panel A6, which in my view houses a saint horseman (Plate 13). He is wearing a cape and is holding a lance pointing down, while his horse is rearing up with one front leg bent above an animal or a monster. This iconography is typical of the depiction of mounted saints. It could be St. George who is for instance represented on the mosaic of the sixth-century Church of Khirbat al-Mukhayyāt in Jordan,96 or St. Sisinnios who is depicted on a Coptic mural painting found in Bawit monastery (sixth century) as a horseman staving off a female demon.97 The presence of a Christian saint in the Umayyad calendar suggests that the painters may have used different cycles and sources, including Christian ones. This practice is far from being peculiar to the Umayyad painters. According to Henri Stern, in the making of the Chronography of 354, painters and authors have consulted different sources such as the
0 42
204
N adia A li and M attia G uidetti
catalogues of the Martyrs, and the catalogue of bishops, as well as the Pascal cycle. The comparison between the Umayyad decor and a wide array of pre- Islamic calendars allows us to identify some typical components of calendar iconography, thus supporting our hypothesis of an Umayyad calendar. In spite of a relative stability in the iconographic types of certain months or zodiacal signs, it remains difficult to grasp the organizing principle, which may have guided the placing of each personification within the Umayyad calendar taken as a whole. This difficulty is increased by the fact that we have not yet identified the cycle, nor the sequence of the months. Actually, some panels can be read in a chronological sequence from north to south (D8, D7, D6), while others are from south to north (A1, A2, A4). Hence, at first glance, it seems that chronological order has not affected the making of the Umayyad program. On further examination, we find that the specific placing of some personifications in the ensemble reveals the impetus of visual habits. To illustrate this point, I shall return to two examples. First, let us consider the beginning of the Umayyad calendar. From a chronological point of view, the fact that January and September are represented in the same row seems hard to explain. To grasp the process by which they have been placed in the decor, we need to consider the following points. First, judging from the Christian calendar mosaics of Jerash, the series of months could begin by January (as in the destroyed mosaic of Jerash) or by September as exemplified by the calendar of the Chapel of SS. Elias, Maria and Soreg.98 Consequently, the placing of January in the first complete compartments (D6–D7) located in the northwestern part of our Umayyad grid and near the entrance of the building is not incoherent. According to our interpretation, the second possible candidate for the beginning of the Umayyad calendar, namely September, is placed in the first three panels located in the southwestern part of our grid, which is near the entrance of the throne room (D1, D2, D3). At this point, I want to stress that the placing of January and September in panels that are located at the beginning of a row and near the main entrance or near the entrance of the alcove cannot be a mere accident. Rather, it may have been motivated by the traditional starting position of these months in the different local Christian calendars. Second, there may be a reason for the placing of January and February in the same row, thus forming an illogical sequence from the chronological point of view. The Jerash calendar and its organizing principle offers us a hypothetical explanation. In fact, the grid pattern allows two main directions of reading: on
0 25
Umayyad Palace Iconography
205
one hand, one can read each horizontal row from north to south, then following a chronological order of months; on the other hand, one can read each vertical row from east to west, thus resulting in a different and nonchronological sequence. Here, we should replace the term sequence by “collocation,” which, broadly speaking, designates a traditional visual association of motifs. In the Jerash calendar, September is followed by the panel of January situated in the same vertical row as in the Umayyad series. Similarly, we can also note that in the Jerash calendar, October is immediately followed by February on the same vertical row, thus giving us a possible reason for the placing of October (B3) near February (B4) in the Umayyad calendar. From the point of view of the craftsmen’s methods of work, the results of the comparative analysis suggest several points. First, that the Umayyad calendar has been built up through a basic savoir-faire, which consists of organizing the ceiling surface by a grid pattern and by using the traditional theme associated with the grid, namely the calendar. Second, our identification of the personification of months would not have been possible without a relative stability of the iconographic types, which is quite rare in Quṣayr ʿAmra. Third, to understand the ordering of the figures within the grid, it is necessary to consider, besides the small chronological sequences that seem to have survived in the Umayyad calendar (A1, A2, A4; B3, B1; D8, D7, D6), the impact of traditional visual collocations of months. The survival of the grid pattern and its structural link to the calendar theme, as well as the survival of some of the pre-Islamic collocations of months found in the grid calendar of Jerash, suggest that one of the craftsmen’s visual sources must have been a grid-pattern calendar read from various directions. This leaves us with several questions concerning the models used and their transmission: were the painters inspired by several local calendars or a preexisting illustrated calendar that no longer exist? Did they use a paradeigmata99 with some indications about the figures that have to be joined, but not necessarily for the whole scheme? Did they reconstruct a new type of calendar resorting to their memory and training? Did they use visual models that they were no longer able to read correctly?100 Further research is needed to answer those questions. The presence of signs of the zodiac and planets in the calendar should not come as a surprise, because the zodiac, calendar, and planets were tightly linked in practice and in iconography. Signs of the zodiac appear not only in the Chronography of 354, but also in in several manuscript paintings where the months, zodiac and planets are depicted together.101
6 0 2
206
N adia A li and M attia G uidetti
For now, let us be content to note that the identification of the Umayyad calendar supports our hypothesis that architecture and layout influenced subject-matter choice. In fact, without wishing to be overly determinist, it seems that these factors were concrete parameters in the choice and distribution of subjects, as well as one of the surest ways of accessing the craftsmen’s habitus. More importantly, the survival of a structural link between architecture and pattern on the one hand, and between a pattern and a specific theme on the other suggest that, contrary to current arguments, the Umayyad craftsmen not only inherited an artistic vocabulary to be recomposed at will, but also a syntax deeply rooted in Late Antiquity.
Interpretation of the Broader Scheme: God’s Caliph and Felicitas In the same perspective, another aspect needs to be taken into account. Umayyad iconography has often been classified into categories, such as royal iconography, hunting, entertainment, and naked women. These categories are based not only on an incorrect identification of themes but also on an arbitrary division of subjects. However, if we are to dissect the program into its main thematic building blocks, we need to agree on where to make the divisions. I suggest an approach that builds solidly upon the artists’ established conventions. If we look closely at the pre-Islamic visual artistic traditions, we observe that certain themes regularly function alongside other themes, forming “thematic structures.” These continue to function together during the Umayyad period. For instance, the calendar theme occurs alongside the Seasons and the personification of Gaea on several Late Antique mosaics (el-Djem, Carthage, and Antioch) and artifacts (Patera from Parabiago and Coptic textiles). It should come as no surprise to find the personifications of the Seasons and six busts of Gaea102 flanking the image of the caliph in the Quṣayr ʿAmra alcove (fig. 7.24),103 forming a continuity with our monumental calendar (fig. 7.25). The broader scheme of the central nave suggests an argument in favor of a caliphal iconography revolving around the agricultural calendar and seasonal cycle (as well as cosmic themes echoed in the dome of the caldarium).104 To better explain the reasons behind the subject-matter choice, two main avenues of research could be taken. We could consider the relationship between this type of agricultural calendar and the role played by the Umayyad quṣūr in the cultivation of Greater Syria.105 Or we might situate this Umayyad iconography in the context of a Late Antique visual rhetoric
0 2 7
Umayyad Palace Iconography
207
Figure 7.24 Drawing of wall paintings of the alcove showing a prince sitting above a nilotic landscape, the Seasons, Gaea, and male figures under arches, Quṣayr ʿAmra, Jordan (after Vibert-Guigue and Bisheh 2007, pl. 19b).
of power. It is well attested that in Late Roman and Early Byzantine art, Gaea, the Seasons and the calendar were often used in the decoration of rich houses (the Villa of the Falconer in Argos, Dominius Julius in Carthage II and al-Awzāʿī), churches (Tegea- Episkopi, Thebes, Petra, Jerash, and Khirbat al-Mukkhayyāt), and synagogues (Baysan-Scythopolis, Sepphoris, and Hammat Tiberias) to embody ideas of plenty, good fortune, and earthly abundance.106 As Maguire suggests, the Christian authorities converted a good life dominated by the Late Roman aristocrats into a good life controlled by monotheist authority and the divine emperor. These themes also played a significant role in Late Roman and Byzantine propaganda, especially when they occur alongside images of rulers.107 The Seasons occur frequently on monuments that served Roman
0 82
208
N adia A li and M attia G uidetti
Figure 7.25 Drawing of the wall painting from the central hall showing Gaea, the Seasons, and the calendar, Quṣayr ʿAmra, Jordan (after Vibert-Guigue and Bisheh 2007, pl. 122b).
imperial propaganda (Trajan’s triumphal arches of 114 ad, Constantine’s triumphal arch of 315 ad, and coins). According to Hanfmann, most of the Seasons shown on imperial coins were inscribed with the phrase felicitas temporum, which can be roughly translated as “happiness of the times.”108 The notion of blessings brought to earth by the divine emperor was preserved in the imperial ideology of Byzantium as we can see in the mosaic in the palace of Theodoric in Ravenna (500 ad), in which the emperor is shown as a Bellerophon surrounded by the Seasons and Gaea.109 Umayyad poetical sources often mention the caliph in connection with the Seasons, the earth, the year, and water. Poets such as al-Akhṭal (d. ca. 710 ad) turn to a range of timeworn topics, including praise of the happiness brought by the rule of the caliph, the blessings of agriculture abounding at every season thanks to the caliph’s generosity, and the caliph’s capacity to produce fertility and abundance at all times of the year and, more importantly, to provide wealth because he is God’s chosen deputy on earth and is rightly guided.110 To convey the idea of the unlimited power of the ruler,111 poets such as al-Farazdaq and Jarīr often evoke “the khalīfa of God through whom rain is sought.”112 Most of these Arabic poetical topics could find counterparts in the laudatory devices
9 0 2
Umayyad Palace Iconography
209
of Late Roman and Byzantine poets. At this point it is worth stressing the similarities between Late Roman and Byzantine conceptions of rulership and the Umayyad concept of God’s caliphate.113 This congruence may have encouraged the Umayyad reuse of this particular repertoire. It further implies that the Umayyads might have made a very conscious use of the forms and themes they inherited,114 thus encouraging a revision of the false dichotomy often established between Late Antiquity and an Arab background assumed to be isolated from the world civilization before the rise of Islam.115
The Floor Mosaics of the Bathhouse of Khirbat al-Mafjar The ubiquity of bathhouses in the early Islamic period can hardly be surprising. The sixth century is a period of revitalization of this architecture, as shown by the bath complexes of Androna and Zenobiyya. The former was lavishly decorated with opus sectile, marble panels, frescoes, wall glass mosaics, and marble statues.116 The ensuing Islamic conquest does not seem to have changed the habits of the population and elite. The bishop of Qinnasrīn, in the mid-seventh century, confessed to his pupil Basilius that he was attending a thermal complex to treat the pains given by the aging.117 During the same period Hammat Gader, one of the most famous Roman thermal complexes in the eastern Mediterranean, was still functioning.118 A Greek inscription found in a hall of the bathhouse embodies the continuity of Late Antique practices into the early Islamic period. Dated to the year 662, the inscription refers to the refurbishment of the thermal complex by the Muslim governor ʿAbd Allāh ibn Abī Hāshim, under the care of Ioannes, the Greek official of Gadara. The highest Muslim authority, the caliph Muʿāwiya, is also mentioned, though the inscription is in Greek and begins with a cross. The inscription motivates the refurbishment “for the healing of the sick,” which is one of the renowned virtues of Hammat Gader mentioned by Antoninus of Piacenza in the second half of the sixth century and al-Muqaddasī in the tenth century.119 Hammat Gader was a thermal complex connected to a natural source and as such exploited throughout different periods. However, thermal complexes unrelated to a spring also show a continuity of use. Such is the case of the complexes of Tiberias and Androna; in the latter, although a second bathhouse was added within the city, the Late Antique one remained in use.120
021
210
N adia A li and M attia G uidetti
The rise of Islam had no impact on the continuity of bathing culture. One may suppose that the disputes between religious prescriptions and social practices experienced during Late Antiquity among Jews and Christians were also common among Muslims. Bathing culture— including the presence of “licentious” images inherited from classical antiquity and disapproved of by (monotheistic) religious elites—lived on along with the beliefs and lifestyle promoted by early Islam.121 Late Antique complexes are also worthy of investigation with regard to some architectural features of early Islamic bathhouses. Androna, for instance, seems relevant for explaining the apsidal spaces of the complex of Khirbat al-Mafjar.122 The numerous apses in the main bath hall of Khirbat al-Mafjar are a common feature of Roman and Late Antique bathing buildings and appear to be related to the function of the place as the dressing room of the bathhouse.123
The Floor Mosaics of the Bath Hall According to Hamilton the floor mosaics in Khirbat al-Mafjar consist of thirty-eight different carpets in the apodyterium plus the mosaic floor of the small room located in the northwestern corner (fig. 7.26).124 Two points emphasized by Hamilton deserve particular remark. First, the disposition of the panels might reflect the superstructure and the architecture of the bath complex. The symmetrical layout of the panels around the five medallions, he suggests, is to be related to the cross vaults and the squared bays of the architecture above. Second, the geometric vocabulary that dominates the panels can be found in Late Antique mosaics. According to Hamilton, however, this vocabulary was used in Umayyad times for its linear rather than spatial qualities; in other words he argues that whereas in Roman and Christian mosaics geometry was used for creating shapes and spaces for the emblemata (i.e., the central panel in which the main subject matter was located), in the Umayyad context similar patterns were employed for the sake of their geometrical features.125 Admittedly since the publication of Hamilton’s Khirbat al-Mafjar the corpus of mosaics available to scholars has dramatically increased. The Byzantine art scholar Andrea Paribeni, for instance, has recently emphasized the extent to which Umayyad mosaic patterns are embedded in the Late Antique mosaic tradition.126 With regard to the interloped patterns visible in some panels at Khirbat al-Mafjar (fig. 7.27), a similar treatment of the surface is featured in the presbytery of the
21
Umayyad Palace Iconography
211
Figure 7.26 Plan of the bathhouse of Khirbat al-Mafjar, Palestine, mid-eighth century (Hamilton 1959, fig. 258).
Church of St. Stephen at Umm al-Rasas, dated 756. The rectangular panel in front of the altar and the square panel on the right of it display the same range of circular, straight, and broken lines composing intricate woven patterns (fig. 7.28).127 The propensity to fill the whole surface of a panel or a medallion with pure geometrical motifs started to emerge during Late Antiquity. Geometrical floors often decorate the lateral aisles of basilicas, as in the cases of the Church of Moses on Mount Nebo (597) or St. Mary and St. Paul in Rihab (late sixth century).128 In Rihab, the Church of St. Menas, dated 635, displays purely geometrical floor mosaics throughout the whole church: emblemata are also absent from the central nave. A similar prevalence of geometrical
21
212
N adia A li and M attia G uidetti
Figure 7.27 Medallions 27 and 7, Khirbat al-Mafjar, Palestine, mid-eighth century (Hamilton 1959, pl. LXX).
patterns over figurative ones is visible in the basilica of Yasilah (early sixth century).129 Single medallions dominated by intertwined circular patterns appear in some further sixth-century floor mosaics, as in the Lower Church of Kaianus at Uyun Musa (early sixth century)130 and in two panels from the monastery at Siyagha (sixth century) (fig. 7.29).131 The choice to use intertwined geometrical patterns per se and not as a frame for figural motifs also distinguishes the medallion in the north aisle of Procopius Church in Jerash (526 ad) and those in the floor mosaics of the Byzantine baths in Gadara.132 The reign of Justinian (527–565 ad) is the most likely date of another early mosaic with patterns very
231
Umayyad Palace Iconography
213
Figure 7.28 Area of the presbytery, Church of St. Stephen, Umm al-Rasas, Jordan, 756 ad (Piccirillo 1993, 220, fig. 346).
close to the interlaced compositions of Khirbat al-Mafjar: the medallions in the floor mosaic of the central nave of the Basilica of the Nativity in Bethlehem.133 The eastern side of the rectangular panel displays two large medallions filled with circular and rectilinear interloped patterns, while the same motifs are repeated on a smaller scale in two medallions in the center of the same panel (fig. 7.30). The organization of the decoration at Khirbat al-Mafjar is consistent with the development of floor mosaics during Late Antiquity: according to Ernst Kitzinger, around the late fourth century, a new perception of the floor as a plain surface appeared to replace the “classical” search for a tri- dimensional effect. The vanishing of perspective planes from floor mosaics corresponds to the diffusion of geometrical and organic elements, which start to organize the surface as a carpet.134 The near-contemporary eighth-century floor mosaics at Khirbat al-Mafjar and St. Stephen at Umm al-Rasas can be considered as the final stage of a decorative trend that started in Late Antiquity.
421
Figure 7.29 Medallion, monastery at Siyagha, Jordan, sixth century (Piccirillo 1993, fig. 285).
Figure 7.30 Drawing of the mosaics in the central nave (detail of the eastern side), Basilica of the Nativity, Bethlehem, Palestine, 565 ad (after Harvey 1935, pl. XX).
251
Umayyad Palace Iconography
215
Figure 7.31 Apse IV, Khirbat al-Mafjar, Palestine, mid-eighth century (Hamilton 1959, pl. LXXXV).
Similar parallels with Late Antique mosaics can also explain the decoration of the apses of the main hall. This is the case with apse IV, decorated with rows of small lunettes each enclosing a floral motif (fig. 7.31). The area around the font platform in the baptistery of the Old Diakonikon on Mount Nebo, dated around 530, displays an identical pattern (fig. 7.32).135 Further comparisons can be drawn with regard to apse IX (fig. 7.33): rows of little flowers fill the apsidal space and a very similar arrangement appears in the main apse of the Church of the Lions at Umm al-Rasas (late sixth century), as well as in the apse of the Theotokos Chapel on Mount Nebo (early seventh century) (fig. 7.34).136 Apse V, which lies opposite to the entrance and is slightly larger than the others, is instead decorated with a radiating motif. Multicolored scales are deployed so as to create a whirling effect, with single units increasing in size moving toward the border (fig. 7.35). The presence of a central lunette from which the entire motif seems to emanate and the spiral effect of the composition find once more a close comparison in the apse of the Church of St. Stephen at Umm al-Rasas (fig. 7.36).137 Four further apses of the apodyterium feature what Hamilton has defined as rainbow matting patterns (fig. 7.37).138 Symmetrical chevron lines rule the compositions. The first three mosaics imitate book-matched marble revetment, though, while the first apse above resembles a four marble-slab pattern, the second and the third apses are limited to a two- slab pattern. The fragmented lines displayed in the fourth apse create the illusion of waviness recalling the qualities of marble veneers139 and opus sectile pavements.140 Marcus Milwright, besides hypothesizing an aesthetic relationship between matting and marble veining, has explored the
621
Figure 7.32 Baptistery of the Old Diakonikon, Mount Nebo, Jordan, 530 ad (Piccirillo 1993, fig. 184).
Figure 7.33 Apse IX, Khirbat al-Mafjar, Palestine, mid-eighth century (Hamilton 1959, pl. LXXXV).
2 71
Umayyad Palace Iconography
217
Figure 7.34 Apse of the Theotokos Chapel, Mount Nebo, Jordan, early seventh century (Piccirillo 1993, fig. 200).
Figure 7.35 Apse V, Khirbat al-Mafjar, Palestine, mid-eighth century (Hamilton 1959, pl. LXXXVI).
medieval descriptions of marble by Muslim authors and linked them back to Late Antique ekphrasis.141 The patterns created by the veining of marble were interpreted by him as “waves of sea.” The connection between marble, water, and floor mosaics has been further discussed by Fabio Barry. He argues that Carystian marble, distinguished by wavy lines, was often selected to evoke water or considered suitable for watery spaces. In Thuburbo Maius (Tunisia), Carystian marble was used in a bathhouse
821
218
N adia A li and M attia G uidetti
Figure 7.36 Apse, Church of St. Stephen, Umm al-Rasas, Jordan, 756 ad (detail from Piccirillo 1993, 220, fig. 346).
while mosaics replicating its natural patterns were displayed in the impluvium of a nearby villa (fig. 7.38).142 In the baths located nearby, the basilica of Campanopetra at Constantia (Cyprus), opus sectile and marble veneers on the floor evoke the qualities of water (fig. 7.39).143 Opus sectile was the noblest of the flooring techniques and whenever possible preferred to mosaics that were possibly used to replicate the aesthetic effects of marble.144 It seems significant that the same chevron motif was also used to decorate the floor of the small pool of the apsidal room. It should be mentioned that, besides its hypothetical courtly functions, the apodyterium of Khirbat al-Mafjar was mainly a space with water. The entire southern side of the main hall is occupied by a pool, and the large medallion located in the very center of the hall (n. 17) is provided with a pierced disc of marble to drain water out.145 The latter feature might also help to explain the decoration of the central largest medallion. The structure follows the principle used in the decoration of apse V. The single unit, however, is not a scale divided vertically into two halves but a rhomboidal form divided horizontally into two triangles. In both cases the single units expand in size toward the margin and compose curved lines that, rotating in opposite directions, create a spiral effect. The medallion is framed by an undulating ribbon (fig. 7.40).
921
Figure 7.37 Apses VI, VII, VIII, X, Khirbat al-Mafjar, Palestine, mid-eighth century (Hamilton 1959, pl. LXXXIII).
0 2
Figure 7.38 Impluvium, villa at Thuburbo Maius, Tunisia, 317 ad (Barry 2007, fig. 17).
Figure 7.39 Opus sectile floor, baths, Constantia, Cyprus, sixth century (Michaelides 1987, pl. XXXVIII).
221
Umayyad Palace Iconography
221
Figure 7.40 Medallion 17, Khirbat al- Mafjar, Palestine, mid-eighth century (courtesy of the Palestinian Department of Antiquities and Cultural Heritage).
The medallion, which is the largest panel at Khirbat al-Mafjar, stood under the main dome of the complex. The importance of the location has been highlighted by Hillenbrand, who has also suggested that both the round shape and the effect produced by the rays emanating from the center might have been related to solar symbolism.146 The motif displayed in the medallion originates in Roman times when the whirls drew the gaze of the beholder toward an image in the center.147 The whirlpool effect was, however, considered to be fitting in relation to water: the same motif can be found in Late Antique impluvia, baptisteries, and bathhouses.148 In the case of the baths at Salamis-Constantia the same motif, based on a rhomboidal unit, is repeated, although with the opus sectile technique (fig. 7.41).149 The presence of the pierced marble disc at the center of the medallion in Khirbat al-Mafjar might suggest that it was the presence of water that dictated this specific motif. At the same time the central position of
2
222
N adia A li and M attia G uidetti
Figure 7.41 Medallion (opus sectile), baths, Constantia, Cyprus, sixth century (Michaelides 1987, pl. XXXVIII).
the panel and its relationship with the above dome might have also concurred to using the circular form. Medallions, in fact, are often put in the center of rectangular or squared compositions in pivotal spatial locations. Medallions might carry inscriptions, as in the floor mosaics of the central nave of the basilica of Khirbat Mouqa, in Syria (late fourth century), the eastern cathedral of Apamea (early sixth century), and the Church of St. George in Khirbat al-Samra (637).150 Medallions with whirling patterns appear in a central position at the foot of the steps in the Glass Court, belonging to the first nucleus of the Cathedral of Jerash (fourth century); at the center of the rectangular panel located in front of the apse of the northern chapel of a church at Zay, nearby Gadara (probably sixth century); further north in the center of a panel in the Church of Hanita (sixth century), and in the center of the opus sectile decoration of the southern aisle of the eastern basilica at Qal’at Sem’an (late fifth century).151 Finally a further comparison might be drawn with two medallions in the Basilica of the Nativity in Bethlehem juxtaposed to those with geometrical interlaced patterns described above.152 Located in the central nave just in front of the three conches of the apsidal area, the two medallions exhibit the typical arrangement of rhomboidal forms composing the whirls of a spiral (fig. 7.42). It seems, therefore, that the presence of water as well as the
23
Umayyad Palace Iconography
223
Figure 7.42 Medallion, central nave of the Basilica of the Nativity, Bethlehem, Palestine, 565 ad (Harvey 1935, fig. 94).
central location within the architectural space both contributed to dictating the selection of the form and motif of this decoration.
The Two Figurative Mosaic Panels The last two sections of the floor mosaics to be examined are the “figurative” ones, which have inspired most iconographic interpretations. In front of apse V there is a rectangular border serving as a threshold. The elongated panel is subdivided into nine units, portraying swastikas (four units), chiastic motifs (two units), a chevron rhomboidal form (two units), and a knife with a fruit (central unit) (fig. 7.43). The apotropaic nature of the threshold has been highlighted by De Loos-Dietz. The horizontal panel has been associated to other “protective” mosaics portraying lozenges and swastika motifs distributed in other parts of the floor and identified either as apotropaic elements or as place-markers.153 The chiastic motifs represented in two units during Late Antiquity were, together with knots, swastikas, and rhomboidal patterns, also widely used to decorate liminal spaces such as entrances. One of the two panels located in front of the stairs of the apse in the Basilica of the Nativity in Bethlehem shows the association of chiastic motifs and swastikas as in Khirbat al-Mafjar.154
4 2
224
N adia A li and M attia G uidetti
Figure 7.43 Threshold, apse V (detail), Khirbat al-Mafjar, Palestine, mid-eighth century (Hamilton 1959, pl. LXXXVI).
The central unit of the threshold—portraying a fruit with a knife—has triggered a vast array of interpretations including a creative reading of the composition as a family portrait in pun form.155 Other scholars have emphasized how often the motif was used in Late Antiquity: in order to make it “Islamic,” however, it has been suggested it was inspired from a Qurʾanic passage mentioning the usage of knives during a banquet (Sura 12: 31).156 The fruit depicted in the center of the threshold has been identified as an ethrog by Ettinghausen.157 Its association with a knife was quite common in Late Antique and contemporary Christian mosaics. An ethrog provided with a sprig together with a curved knife appears in one of the corners of the central panel at the Church of the Theotokos (early seventh century), and the same fruit together with both a curved knife and a dagger is also among the motifs displayed in front of the apse in the Church of the Lions at Umm al-Rasas (late sixth century). The mosaics of the Church of the Virgin Mary in Madaba, datable to the Umayyad period (early eighth century), present two symmetric trays with a knife, an ethrog, and what looks like the oblong stem of the fruit; they are located in front of the presbytery (fig. 7.44).158 Finally, the distinctive shape of the fruit in Khirbat al- Mafjar finds a close comparison with the fruits depicted in a sixth-century floor mosaic at the Kathisma church. The five curvilinear bulges and the
25
Umayyad Palace Iconography
225
Figure 7.44 Area of the presbytery, Church of the Virgin Mary, Madaba, Jordan, sixth-seventh century (Piccirillo 1993, fig. 20).
Figure 7.45 Floor mosaic from the Kathisma church, Palestine, sixth century (courtesy of Shmuel Browns).
sprout down below the fruit create a very similar profile (fig. 7.45). The ethrog was a harvest-related theme, suggesting wishes of fertility and immortality.159 The threshold of apse V, therefore, consists of a selection of recurrent elements used to decorate borders and threshold areas. Loose notions of apotropaism and auspiciousness were probably attached to the iconography of the nine units, but the main reason to have the threshold
6 2
226
N adia A li and M attia G uidetti
decorated in such a way was habitual: this was how artists and artisans used to decorate the surface of this specific space of a building. The last panel is the floor mosaic of the apse in the small room located in the northwestern corner of the apodyterium. Known as the “lion and gazelle” mosaic, this is one iconic example of early Islamic art. It consists of a central pomegranate tree flanked on the right side by a lion attacking a gazelle and on the left by two gazelles grazing bushes and shoots from the tree (fig. 7.46). In the literature, the contraposition between the peaceful left side and the violent right side has been emphasized by Ettinghausen, the centrality and the large crown of the tree by De Loos-Dietz, whereas the metaphorical reference of the tree and the animals to an erotic encounter as evoked in a poem by al-Walīd ibn Yazīd by Behrens-Abouseif.160 In a previous article I suggested a possible comparison between this panel and the floor mosaics of two Late Antique bath complexes located in Syria: Serjilla and Dibsi al-Faraj.161 In both complexes the apodyteria are decorated with rectangular panels portraying fighting between felines (a lion in Sarjilla
Figure 7.46 Apse of the “diwan,” Khirbat al-Mafjar, Palestine, mid-eighth century (Hamilton 1959, pl. LXXXIX).
2 7
Umayyad Palace Iconography
227
Figure 7.47 Drawing of the floor mosaic of the apodyterium, Serjilla, Syria, 473 ad (Butler 1904, fig. 290).
and a tiger in Dibsi al-Faraj) and gazelles.162 In the former case a majestic fruit-loaded tree divides two animal couples (fig. 7.47). The theme, however, was by no means restricted to bathhouses: compositions including a tree, peaceful animals, and aggressive or violent beasts were also part of the repertoire of ecclesiastic buildings. This ensemble of subjects was mostly located in front of the altar, or in the apse, or in the culminating side of naves and aisles. According to Pauline Doncel-Voûte spaces devoted to antithetical compositions were those marking the transition between two different spaces within one single building. It might also be that a sense of protection was conveyed through these scenes.163 The panel at Khirbat al-Mafjar displays both a sense of symmetry and contrast between the gracefulness of the gazelles and fury of the lion. This opposition is part of a general trend in Late Antique floor mosaics, which Paribeni has described as the contraposition between the wild and the
8 2
228
N adia A li and M attia G uidetti
domestic world: hunting scenes opposed to idyllic scenes of tending cattle and cultivation.164 The theme was adaptable to secular contexts, such as countryside estates, which represented the “good life” secured by the master in contrast to the dangers of savageness, as well as to religious contexts, in which Paradise and its bounties were opposed to the cruel law of the natural world. On a formal level, the bucolic aspect is often rendered through animals grazing directly from the trees (an aspect that had in turn a symbolic signification making the tree a source of life): this is the case, in Jordan, with the Old Diakonikon on Mount Nebo (530 ad), with the Church of the Lions at Umm al-Rasas (late sixth century), and in Syria and Lebanon, with Late Antique churches such as Khan Khaldé (south church), Maʿrata, and Sardje.165 In the Church of the Deacon Thomas, in the Valley of ‘Uyun Musa in Jordan (unknown date), the rectangular panel of the presbytery displays in the upper (eastern) section a ram quietly grazing the fronds above a lion attacking a zebu. The scene runs among four trees loaded with fruit (fig. 7.48).166 A small rectangular panel from the Church of Sardje, published by Doncel-Voûte, presents a similar case: in
Figure 7.48 Altar area, Church of the Deacon Thomas, ‘Uyun Musa, Jordan, sixth century (Piccirillo 1993, 188, fig. 266).
9 2
Umayyad Palace Iconography
229
Figure 7.49 Apse of the crypt of the Church of St. Elianus, Madaba, Jordan, 595 ad (Piccirillo 1993, 125, fig. 134).
the upper part, a lion is attacking its prey while below in the center of the composition two animals graze directly from the tree.167 The compositions seem to vary, according to the space available to mosaicists. In the case of a small and/or vertical rectangular panel the animals are displayed on two rows on both sides of a central tree (Sardje, Church of the Deacon Thomas). When instead the panel is larger and displayed horizontally, the trees are located at the four corners of the compositions with pairs of facing animals filling the remaining space. This is the case, for instance, with a mosaic found in Madaba and of a panel of the Church of the Holy Martyrs on Mount Nebo.168 Finally, when the space is a narrow rectangular band, the animals are displayed facing each other on either side of a tree on a single row (as at the Church of the Holy Martyrs on Mount Nebo). This is also the case in apses, where the size of the central tree is adapted to the concavity of the apse (apse of the Crypt of St. Elianus in Madaba, 596/7 ad) (fig. 7.49).169 The latter typology, to which the apse of Khirbat al-Mafjar clearly belongs, is also visible at the presbytery of the Church of the Lions, where three trees separate two lateral grazing gazelles and two central aggressive lions: a mosaic reshuffling of several elements featured at Khirbat al-Mafjar (fig. 7.50).
Mosaics and Architectural Spaces While more general conclusions will be drawn in the next section, two interrelated features of the floor mosaics of the bathhouse of Khirbat
0 32
230
N adia A li and M attia G uidetti
Figure 7.50 Area of the presbytery, Church of the Lions, Madaba, Jordan, late sixth century (detail from Piccirillo 1993, 211, fig. 337).
al-Mafjar are worth mentioning. The first is that, once replaced in a Late Antique perspective, the themes appear to have been mainly dictated by features of the architectural spaces. The function of the space might also have played a role, especially the presence of water. The second point is the level of contact with Late Antique and contemporary Christian mosaic practices. Civic and religious Late Antique architecture displays similar patterns and motifs. The area of Mount Nebo and Umm al-Rasas offers plenty of parallels and it is not accidental that Christian mosaics were produced during the Umayyad period in this area. At the same time, although the technical features of the mosaics have not been discussed here, the outstanding quality of the mosaics of Khirbat al-Mafjar—as well as those from the Umayyad residence of al-Qasṭal170—recalls Justinian’s patronage in Greater Syria in the sixth century, for example, the floor mosaics at the basilica of Bethlehem and a fragment with an antelope recovered in one of the basilicas at Rusafa.171
231
Umayyad Palace Iconography
231
These two points lead to the idea that the whole repertoire and arrangement of motifs and compositions displayed in the mosaics of Khirbat al- Mafjar obeyed to a canon of sorts, with rules deeply rooted in the artistic world of Late Antiquity. Rather than searching for an iconographic program shaped by the biography of the patron, this analysis invites us to consider how mosaicists operated within a quite standardized process of selection. The position of a panel within the whole building and the function of the space where it was located offered mosaicists a restricted range of choices within a well-established preexistent artistic tradition.
Conclusions The approach shared by the two sections of this article keeps “iconography” and “meaning” in the background and emphasizes instead the importance of the artistic practice and the transmission of forms. Techniques, motifs, and compositions were often used because of the existence of a canon of artistic rules and practices to be followed when planning the structure and the decoration of specific buildings. The existence of “rules to follow” in order to make something “proper” is one of the core principles of architecture according to Vitruvius.172 Decorum (propriety) is described as deriving from prescription, usage, and nature and it reflects the values of the time in which architecture is produced and not the whims of one single patron. Approaching iconography through the reconstruction of the craftsmen’s habits allows the analysis of early Islamic artworks and their functioning in a less impressionistic and arbitrary way than in the past. In so doing, we are aware that we leave ourselves vulnerable to criticisms. Some scholars might think that this approach leads to a facile reductionism and sterile analyses that disregard the historical context as well as the cultural specificity of Umayyad artistic production. Paradoxically, the risk is being accused of falling back into a rigid and reactionary view of early Islamic art. This is because the iconographic programs have been linked to a set of constraints such as architecture, patterns, and thematic structures and to habits that condemn Umayyad art to conformity and copy and reduce the function of images to mere space-filling. However, our main goal was to show that in addition to a text-based analysis—and perhaps even before arbitrarily pairing images with textual passages—one can draw a specific description of the program, its genesis,
23
232
N adia A li and M attia G uidetti
and regularities by looking at the mechanisms of transmission of forms and at the relationship between architecture and decoration. If we partially suspended reference to the historical context, it was not to discover dehistoricized laws of construction of iconography that could be applied in the same way in all artistic traditions. On the contrary, the aim was to bring to the foreground the position and function occupied by the artistic context in the making of Umayyad iconography. Moreover, it should be said that artistic practice or habit is by no means less relevant than patrons’ agenda in assessing why a specific form was used and how it was perceived by a patron or society.173 Artistic practice, which can be described as the stratified knowledge processed and reproduced generation after generation by artists and craftsmen, is part of what Pierre Bourdieu called the modus operandi of a period. The latter is the mental habit underpinning the actions of the components of a specific society, including the material production of culture as well as the attribution of specific functions and meanings to determinate sets of artworks.174 To further analyze the interrelation of artistic forms with the culture which produced them, two research avenues can be suggested. With regard to the transmission of motifs and the meaning of images it is worth discussing a well-known passage that scholars have used to suggest—once again—the possible meaning of the iconography of bathhouse decoration.175 It is from al-Ghuzūlī’s Maṭāliʿ al-budūr fī manāzil al-surūr, a text dealing with housing in medieval Islam. In the following passage al-Ghuzūlī (d. 1412) focuses on bath decoration: “In good baths, you also find artistically painted pictures of unquestionable quality. They represent, for instance, lovers and beloved, meadows and gardens and hunts on horseback or wild beasts. Such pictures greatly invigorate all the powers of the body, animal, physical and psychological.” Al-Ghuzūlī continues by referring to two previous authors:176 Badr al-Dīn ibn Muẓaffar and Muḥammad ibn Zakariyyā al-Rāzī (854–925/35). The former praised the importance for the soul of contemplating “artistic and beautiful pictures.” The latter links the observation of pleasant colors and right proportions of the forms to the healing of melancholy and removal of worries. Al-Ghuzūlī then carries on referring al-Rāzī’s explanation of the therapeutic function of images found in bathhouses: drawing on Neo-Platonic and Epicurean notions, he
23
Umayyad Palace Iconography
233
defines bathing as a holistic experience during which both body and spirit are elevated. A further passage by al-Rāzī, transmitted by al-Ghuzūlī, is however important in order to understand the relationship between culture and artistic practice: If one asks a discerning painter why painters use only these three subjects for the painting of baths, he cannot give a reason for this; he would not remember those three qualities (of the mind) as the reason. This is due to the fact that the earliest beginnings lie so far back, and hence the cause is no longer known. (The philosophers) have not omitted anything that is correct, nor introduced anything meaningless.177 Al-Rāzī notices that artists contemporary to him were not able to explain the reasons for having certain forms and themes depicted in bathhouses. It appears that craftsmen were replicating patterns that no longer “belonged” to their age and culture. Al-Rāzī was impregnated with classical and philosophical texts and therefore able to recall the original function of a specific iconography. The passage suggests that forms and artistic habits might remain valid and active well beyond the time of their first formulation. It is, in fact, only when a new “paradigm” sediments and spreads its vision of the world into social practices that one can assist to the rise of a new modus operandi and therefore to a shift in artistic production. With regard to bathhouses, since the Ayyubid period, a new model of architecture, with a different heating system and decorated according to new aesthetic principles, emerged as a result of radical changes within medieval Islamic society.178 The evidence from Khirbat al-Mafjar and Quṣayr ʿAmra, instead, suggests that early Muslims still operated within a Late Antique cultural paradigm, which brought patrons to desire lavish bathhouses and at the same time also provided craftsmen with a set of techniques and motifs in order to properly decorate the buildings. In the same line of thought, with regard to the specific meaning of a set of images, the evidence from Quṣayr ʿAmra’s central vault and especially the finding of a calendar has four consequences for our understanding of the mechanisms between iconography and society.
4 2 3
234
N adia A li and M attia G uidetti
First, it encourages a revision of both the assumption concerning the prevalence of the so-called princely cycle and the traditional narrative, which has been far too oriented toward the frivolous life of the supposed owner, al-Walīd ibn Yazīd. Second, it seems legitimate to posit that once the hypothesis of “a princely cycle” has been complicated, the argument in favor of an influence of Sasanian ways and ceremonial practices on the Umayyad decor is weakened as well as the thesis according to which the most characteristic feature of Quṣayr ʿAmra’s iconography is the recombining of Greco- Roman and Iranian themes.179 Third, because the “princely cycle” provided the initial stimulus to postulate an Umayyad agency in the process of selection, reappropriation, and reinterpretation of an inherited vocabulary, the discovery of the calendar provides grounds for arguing that the importance of distinctive Umayyad taste, attitude, ceremony, and mentality in the characterization of what is held to be peculiar in Umayyad iconography should be tempered. Fourth, the various indications of a thematic unity in the program of Quṣayr ʿAmra’s central hall combining the enthroned ruler and the triad of Gaea, the Seasons, and the calendar hints at a degree of coherence that has rarely been appreciated and suggests an argument in favor of a caliphal iconography revolving around an agricultural calendar and seasonal cycle comparable in its form and ideological content to the Late Antique visual expression of authority and felicitas. The rule and mechanisms of the artistic habitus and Pächt’s notion of formal opportunities have been the underlying theme of this article. The emphasis on the tight relationship between architectural space, function, and decoration has allowed us to reframe some interpretations of motifs in two of the most important early Islamic sites in the Syrian area. The consistency of the continuity of Late Antique patterns from a formal, thematic, and functional point of view implies that the relevance of the single patrons and their biographies as conveyed by later written sources should be nuanced. Early Islamic art should be visualized as the outcome of the mediation between the dynamic transposition of artistic formulae common throughout Late Antiquity into early Islamic buildings by the workshops in charge of their decoration and an Umayyad-Islamic agenda (expressed, for instance, through the addition of Arabic inscriptions). From a cultural point of view, both of our case studies concur to suggest that early Islamic society—its values, ambitions, and cultural codes—participated in a Late Antique cultural paradigm.
253
Umayyad Palace Iconography
235
Notes 1. For discussions of the Umayyad palaces and settlements, see Lammens 1930; Herzfeld 1921; Grabar 1955; Grabar 1973; Grabar 1993; Hamilton 1959; Schlumberger 1986; Hillenbrand 1982; Geyer 2000; Fowden 2004, 31– 57; Genequand 2012; Genequand 2006; Bartl and Moaz 2008. 2. For detailed descriptions of Quṣayr ʿAmra’s architecture, see Creswell 1932, 255– 272; Sauvaget 1939, 13– 16; Almagro 1975, 25– 48; Bisheh, Morin, and Vibert-Guigue 1997, 375–393; Fowden 2004, 31–57. 3. Fowden 2004, 142–174. The recent campaign of restoration, carried out under the supervision of the World Monuments Fund, has further clarified an Arabic inscription in the main hall in which al-Walīd ibn Yazīd is quoted as a prince; see http://www.wmf.org/video/re-discovering-qusayr-%E2%80%98amra-conservation-early-islamic-site-jordan (accessed November 20, 2015); the inscription is discussed at minute 32:50. 4. Musil 1907. For an illuminating account of Musil’s discovery and study of Quṣayr ʿAmra, see Fowden 2004, 1–12. 5. Regarding the debate over the function of Quṣayr ʿAmra’s hall, see Grabar 1973, 146–148; Sauvaget 1952, 124–129, 147–148; Morin and Vibert-Guigue 2000, 581–591; Fowden 2004, 46–57. 6. Soucek 1993, 109–110. 7. The two concomitant archaeological projects on Khirbat al-Mafjar investigate the physical and social landscape around the more monumental parts of the complex: http://www.jerichomafjarproject.org (accessed November 20, 2015); http://www.pef.org.uk/lectures/hisham-s-palace-in-context-the-archaeological- survey-and-excavations-in-the-hinterland-of-khirbat-al-mafjar-jericho (accessed November 20, 2015). See also the case of Qaṣr al-Ḥayr al-Sharqī (Genequand 2005). 8. Hamilton 1959, 7, pl. LVII, 1 (after Baramki); Hamilton 1969, 61–65. An earlier date in the eighth century was suggested by Donald Whitcomb in an unpublished paper at the third conference of the Historians of Islamic Art Association (October 18th–20th, 2012, Metropolitan Museum of Art, New York). 9. Milwright 2010, 667. 10. Hillenbrand 1982, 1. 11. Hillenbrand 1982, 10; Hamilton 1988, 16; Hamilton 1978, 136. 12. Hillenbrand 1982, 9. 13. Behrens-Abouseif 1997, 16. 14. On this issue with regard to history, see Morony 1981. 15. Judd 2008, 448. 16. Piccirillo 1993, 343–353; Yegül 1992, 339–349; Doncel-Voûte 1988, 463–464; Paribeni 2004. 17. A notable example is Bisheh’s interpretation of the floor mosaics of Qaṣr al- Ḥallābāt: “The important question raised by this mosaic, however, has not to
6 32
236
N adia A li and M attia G uidetti
do with its prototypes and sources, but with whether those borrowed elements had been assimilated, adapted to a new context, and given a specific meaning” (Bisheh 1993, 53). 18. Fowden 2004; Taragan 1993, 2001, 2003. 19. Shboul and Walmsley 1998; Hoyland 2001; Marsham 2009a, 2009b. 20. For an attempt to look at Umayyad palace iconography from the point of view of the craftsmen’s methods of work and the power of forms, see Ali 2008. 21. Pächt 1999, 25–29 (italics ours); Ali 2006, 2–6. 22. Louis Reau, in an article dedicated to the influence of form on medieval iconography, speaks of “la tyrannie de l’architecture sur l’iconographie” (Reau 1951, 86). In the same line of reasoning, Henri Focillon evokes “la loi du cadre” to describe how architecture conditions the forms of the sculpted programs in romanesque art (Focillon 1938, 219). Focillon’s ideas on this topic have been recently analyzed by Sauerlander (2004). 23. Rabbat 1993, 67. 24. Grabar 1955; Grabar 1977, 51–60; Grabar 1988, 75–83; Grabar 1992, 187–193; Grabar 1993, 93–108. 25. The hypothesis concerning the prevalence of the “princely cycle” was first formulated by Grabar 1955, 1–2, 14, 192–233. See also Grabar 1973, 162–164; Grabar 1993, 96; Hillenbrand 1982, 1–20, see esp. 9-10 for reference to Quṣayr ʿAmra; Vibert-Guigue and Bisheh 2007, 8; Almagro 1975, 56; Ettinghausen 1977, 50; Baer 2004, 5–6; Grabar, Ettinghausen and Jenkins-Madina, 2001, 45–48; Van Lohuizen-Mulder 1990, 125–151, especially, 128–129; Zayyadine 1979, 19–29; Fowden 2004, 60–61, 78. 26. Grabar 1973, 163. 27. Grabar 1973, 164; Grabar 1988, 83. 28. Hillenbrand 1982, 1–20; Grabar 1993, 96. 29. Donner 2010, 187–188. Concerning the bias in the Abbasid sources, see also Lewis 1970, 1:47, 65; for a recent revision of the Umayyad historiography, see Borrut 2011, 25–63; Kilpatrick 2010, 63–89. 30. One striking case is the disproportionate attention paid to the depiction of naked women in the study of Umayyad iconography. For instance, Grabar 1993, 97. See also Baer 1999, 13–24; Taragan 1997, 78–81. 31. Grabar 1973, 16. 32. Grabar 1973, 5. 33. Grabar 1993, 98; Grabar 2005, 414–415. This narrative of an inherited vocabulary recomposed in a new syntax is also found in Fowden 2004, 299, 307; Flood 2001, 203; Flood 2012. 34. This shift in the scholarship is not confined to the study of Islamic art. See, for instance, Kessler 1988, 166–187, esp. 166–167. For a critique of the Panoskian scheme, see Klein 1970, 354–375; Bruneau 1986, 249–295. 35. Along with Becker and Brown, Fowden is one of the most stimulating advocates of the continuation of Late Antique ideas into Islam. Although I shall quarrel
2 73
Umayyad Palace Iconography
237
with his interpretation of the painting, one must say that his contribution was a landmark. See Fowden 1993, 138–168 and more recently Fowden 2011, 148–176; Brown 1971; Becker 1924–1932, 1:201. 36. The recent campaign of restoration, carried out under the supervision of the World Monuments Fund, has further clarified an Arabic inscription in the main hall in which al-Walīd ibn Yazīd is quoted as a prince (see note 3). 37. Fowden 2004, 115–142, 85–106, 257–265; Fowden 2007, 29–45. 38. A revealing example is provided by the different interpretations of the panel with the reclining figure (Quṣayr ʿAmra, west aisle, south wall). Basing his argument on several theological sources, Fowden initially interpreted the picture as a representation of Sarah (Fowden 1993, 143–149). Using a different array of textual sources, he recently argued that the panel actually represents Umm al-Ḥakam, the mother of al-Walīd’s designated heir. See Fowden 2004, 184. However, the cleaning of the painting by the Italian team between 2012 and 2013 revealed that the reclining figure is actually a man, probably al-Walīd II. 39. Grabar noted the methodological problem of assigning specific meaning to Umayyad paintings, stating that, except for the princely cycle, “no clear sense of what is shown emerges”; Grabar 1993, 96. 40. Fowden 2004, 298. For the view that there are no representations of the seasons or months in Islamic art, see also Rice 1954, 2. 41. Note, however, that Fowden, like previous scholars, acknowledges the importance of the Roman and Sasanian visual heritage to Umayyad art but does not attempt to define how this heritage helps to explain how the program was made and how it functions. For a thought-provoking analysis of early Islamic art in terms of the continuity of artistic practices, see Allen 1988, 1–15. 42. A noteworthy feature in the decor is the arbitrary cutting of the terminal units. One wonders why the painters did not avoid this effect in incompleteness. Did they wish to imply a continuation of the pattern? Or, as seems more likely, were their compositions governed by a predetermined module, which was repeated as far as necessary in order to cover the surface—even if that meant cutting the last units in half? Some Late Antique mosaics also exhibit arbitrary cutting of the decor. 43. Note that a grid pattern is also used on the ceiling of the eastern barrel vault. 44. Barbet 1985, 77–79, 81–83, 215. 45. Lavin 1967, 102, figs. 17, 18, 19, 20. 46. Lavin 1967, 102. 47. Barbet and Vibert-Guigue 1994, 79. 48. Lavin 1967, 109. 49. The relevance of the Christian mosaics of Bilād al-Shām to the study of Umayyad art has already been noted by Fowden 2004, 282–304, esp. 292. See also Doncel Voûte 1988; Bowersock 2006. 50. Piccirillo 1993, 312, fig. 95, 633. 51. Dunbabin 1978, fig. 99.
8 2 3
238
N adia A li and M attia G uidetti
52. Grabar 1983, 191; Lavin 1967, 198. For Dunbabin “the origins of the schemes of composition of Greco-roman mosaics varied widely; among them are adaptations of patterns used in painting and for stuccoed ceilings” (Dunbabin 1999, 314). 53. For a recent discussion on this topic, see Vasaros 2004, 61–66. 54. Parrish 1984, 52; Dunbabin 1999, 111; Stern 1953, 117–131. 55. Akerstrom-Hougen 1974; Levi 1941, 253, 291. 56. Maguire 2007, 11, figs. 2, 3, 4, 5, 6. 57. Stern 1953, 297; Kraeling 1938, 325, 475, 480; Saller 1949, 284; Michel 2001, 233. 58. For the use of seasonal and calendar iconography in the Christian context, see Hanfmann 1951, 47, 56, 72. 59. Balty 2003, 180. 60. Levi 1941, 289; Reinach 1922, 352–353. 61. Stern 1953, 236; Levi 1941, fig. 13; Paris 2000, 169. 62. The Chronography of 354 was written for Valentinus by Furius Dionysius Philocalus. It no longer exists, but the illustrations are known through two later copies of the fifteenth century at Vienna and one of the seventeenth century at Rome, each being a copy of a ninth-century copy of the original. 63. Stern 1953, 294, 309–340. 64. To facilitate the description, I will designate each panel under analysis by a letter corresponding to each horizontal row and a number corresponding to each vertical row read from south (the alcove’s entrance) to north (the building’s entrance). 65. Vibert-Guigue and Bisheh 2007, fig. 24. 66. The Italian campaign has focused so far on the cleaning of the paintings of the west aisle of Quṣayr ʿAmra’s audience hall providing some spectacular new finds (still unpublished). They have not yet started the restoration of the central vault under study in this essay. So the results of my iconographic analysis are provisional. 67. Of the thirty-two compartments, six are completely illegible (B7, C3, C5, C7, C8, D5). In eight other panels, we hardly discern the outlines of figures about which almost nothing can be said for the moment (A3, A5, B2, B4, B5, C2, C6, D4). It is thus impossible to reconstruct the whole scheme. 68. Weitzmann 1979, 47, 50, 72, 98. 69. Levi 1941, 253–255. 70. Levi 1941, 253–255. 71. Stern 1953, pl. I. 72. Saller 1949, 288–289. 73. Weitzmann 1979, cat.28 (Diptych of Justinian, Constantinople, sixth century). 74. Parrish 1984, pl. 21. 75. Vibert-Guigue 1997, 312. 76. Stern 1953, 249–252; Levi 1941, 260–261. 77. Stern 1953, 366–369.
9 32
Umayyad Palace Iconography
239
78. Levi 1941, 268–269. 79. Stern 1953, pl. XVII, 1. 80. Reinach 1922, 222. 81. Stern 1953, 246. 82. Levi 1941, fig. 1. 83. Saller 1949, 276–277, pl. 47. 84. Strzygowski 1888, 60, pl.19. 85. Saller 1949, pl. 47. 86. Grabar 1955, 204; Fowden 2004, 77. 87. Levi 1941, 267–268, fig. 17; Dunbabin 1978, pl. 105. 88. Balty 1995, fig. 5. 89. Levi 1941, 270–274. 90. Note that there is a reversed zodiac in the caldarium. See Brunet, Nadal, and Vibert-Guigue 1998, 97–123. 91. Dunbabin 1999, 191. 92. Note that the sign of Gemini was also recognized by Martin Almagro (1975, fig. 27). 93. Saller 1949, 276, 279, pl. 51, 3. 94. Stern 1953, pl. XXXVII, 6 (Vat. gr. 1087). 95. Levi 1941, fig. 3 (Ptolemy, Astronomical Text, ninth-century copy of a third- century original). 96. Balty 2003, fig. 50. 97. Paris 2000, 103. 98. Saller 1949, 288–289. 99. Regarding the use of paradeigmata (a preestablished visual model) for ceiling painted decoration, see Ling 1991, 217–219. 100. In this respect, it is worth noting that the zodiac ceiling in Quṣayr Amra is reversed. 101. Levi 1941, fig. 3 (Ptolemy, Astronomical Text, ninth-century copy of a third-century original). See also a manuscript in the Vatican Library, Vat. gr. 1291; Stern 1953, 366. 102. Note that Gaea is also depicted in one of the floor paintings in Qaṣr al-Ḥayr al-Gharbī. 103. For a detailed analysis of the Quṣayr ʿAmra’s alcove, see Ali 2006, 1–31; Ali 2008, 256–289, 368–436. 104. Saxl 1931, 97–123; Brunet, Nadal, and Vibert-Guigue 1998, 97–123. 105. Genequand 2009, 157–177; Genequand 2012. 106. Maguire 1999, 238–257; Maguire 2012. 107. For examples of collocations between Gaea and depictions of Byzantine rulers, see the plate of Theodosius (388 ad, Egypt), the diptych of Consul Magnus (Constantinople, 518 ad), the Diptych of Justinian (Constantinople, sixth century); Weitzmann 1979, 75, 50, 35. Note that the Seasons also appear in Sasanian visual propaganda: Lehmann 1945, 24.
0 42
240
N adia A li and M attia G uidetti
108. Hanfmann 1951, 163–167. 109. Hanfmann 1951, 168, 183–184. 110. al-Akhṭal 1891–1925, poems nos. 25, v. 21; 34, v. 34; 31, v. 31; 11, v. 39; 45, v. 22. See also al-Akhṭal’s panegyrics, nos. 7, 11, 13, 18, 25, 34, 37, 43, 44, 45, 57, 58, 75, 97. For a recent study on al-Akhṭal’s panegyrics, see Bakhouch 2001, 283–314. 111. On this topic, see the contributions by Andrew Marsham and Alain George in this volume. 112. Crone and Hinds 1986, 37. 113. For a description of the Umayyad concept of God’s caliphate, see Crone and Hinds 1986, 8–9, 24–42; Khoury 1992, 97–123. 114. Regarding Umayyad familiarity with the visual language of Late Antiquity and its manipulation, see Rabbat 2003, 79–94; Treadwell 1999, 246. 115. For a revision of the importance of pre-Islamic Arabia to the development of Islamic art, see for instance Finster 1996, 287–319; King 1991, 94–102; King 1980, 37–43; Keall 1995, 11–23. More generally, see Hoyland 2001. 116. Mundell Mango 2002; Lauffray 1983. 117. Nau 1929, 337. 118. Hirschfeld 1997. 119. Hirschfeld 1997, 4–6, 235–240, 267–278. On the marble slabs of the floor of the same niche, Arabic graffiti dated to the period in between 679 and 776 have been found. The dates of the different phases of the bathhouse have been recently reassessed by Magness who argues for a longer use of the complex: Magness 2010. 120. Cytryn-Silverman 2009; Mundell Mango 2008. 121. Within both Christianity and Judaism, debates about the legitimacy of bathing and bathing culture (including attendees’ exposure to nudity and classical artworks) were quite common: Yegül 2010, 26–34; Yegül 1992, 322; Dvorjetski 1999, 127–128. On the removal of a statue from an Islamic bathhouse following the “Edict of Yazīd II,” see Vasiliev 1956, 39. 122. The main hall is most effectively described by Soucek 1993, 109–110. 123. Mundell Mango 2002, fig. 3; Yegül 1992, 339–349. Yegül explicitly mentions that the multi-apsed caldarium of Khirbat al-Mafjar aligned within Roman and Late Antique architectural bathing tradition (p. 346). 124. Hamilton 1959, 327–342. 125. Hamilton 1959, 330. 126. Paribeni 2004, 639. 127. Piccirillo 1993, 220, 238. 128. Piccirillo 1993, 148–149, 310–311. 129. Piccirillo 1993, 313, 341. 130. Piccirillo 1993, 189–190. 131. Piccirillo 1993, 192–193. 132. Kraeling 1938, pl. LXXXI; Piccirillo 1993, 328.
241
Umayyad Palace Iconography
241
1 33. Harvey 1935, pl. XX, figs. 93, 95; Pringle 1993, 138. 134. Kitzinger 1976, 64–88. 135. Piccirillo 1993, 147. 136. Piccirillo 1993, 211, 236–237, 151. 137. Piccirillo 1993, 220, 238. 138. Hamilton 1959, 334–335. 139. See the effect of marble veneers in the Dome of the Rock: Nuseibeh and Grabar 1996, 52–53. This effect was also replicated in the paintings of secular residences; besides the palatial area of the same Khirbat al-Mafjar, see Balis: Leisten 2009, 390–391. 140. See, for instance, the baptistery of the church St. John the Baptist in Jerash: Kraeling 1938, pl. XLVIII, a. 141. Milwright 2005, 211–216. 142. Barry 2007, 632–634. 143. Michaelides 1987, pl. XXXVIII, fig. 64. 144. Doncel-Voûte 1988, 450–451. 145. As also noted by Hillenbrand (1982, 34, n. 238). 146. Hillenbrand 1988, 8. 147. Pappalardo and Ciardiello 2010. 148. Impluvium of a Roman villa (House of the Dolphins) in Baetulo (mosaic now in the Museum of Badalona); Baptistery of Ayios Philon, Carpasia (Cyprus) (Michaelides 1992, pl. XXXIX, fig. 65b); tepidarium of the Roman baths of the city Ghajn Tuffieha in Malta (Zammit 1930). 149. Michaelides 1992, pl. XXXVIII, fig. 63. 150. Doncel-Voûte 1982, 159–167; 203–215; Piccirillo 1993, 306. 151. Kraeling 1938, 309–311, pl. LVIII, a–b; Piccirillo 1993, 324–325; Doncel-Voûte 1982, 464, 225–240. 152. Harvey 1935, pl. XX, fig. 94. 153. De Loos-Dietz 1990; Allen 1995. 154. Kitzinger 1970. 155. Hamilton 1959, 337. Quite interestingly, the only evidence brought in support of this interpretation is what is said in modern times by “more than one Arab visitor” in front of the artwork. 156. De Loos- Dietz 1990, 130; Cruikshank Dodd 1975, 270; Cruikshank Dodd 1984. 157. Ettinghausen 1972, 35. 158. Piccirillo 1993, 151, 211, 64–65. 159. De Loos-Dietz 1990, 127–133. De Loos-Dietz also notices the presence of the ethrog together with other products of harvest among the mosaics of the Dome of the Rock (pp. 129, 133). Ettinghausen (1972, 35–36) argued the fruit at Khirbat al-Mafjar symbolized an offer to the master of the estate who, allegedly, might have sat in the apse V.
2 4
242
N adia A li and M attia G uidetti
160. Hartner and Ettinghausen 1964, 167; Ettinghausen 1972, 43–46; De Loos-Dietz 1990; Behrens-Abouseif 1997. 161. Guidetti and Macchiarella 2007. 162. Charpentier 1994; Harper and Wilkinson 1975. 163. Doncel-Voûte 1988, 478–479. 164. Paribeni 2004, 638. 165. Piccirillo 1993, 135, 211, 213; Doncel-Voûte 1988, 388, 463, 484. Doncel-Voûte (1988, 463–464) mentions Khirbat al-Mafjar within her discussion of Late Antique Syrian mosaics. An image of the floor mosaic of the apse of the Umayyad palace is shown on the same page with an image from the Church of Maʿrata. 166. Piccirillo 1993, 187–188. 167. Doncel-Voûte 1988, 484, fig. 455. 168. Piccirillo 1993, 128, 164–165. 169. Piccirillo 1993, 124–125. 170. Bisheh 2002. 171. Ulbert 1986, fig. 1. 172. Vitruvius 1914, 14–16. 173. A similar approach informs the analysis of the decorative patterns of Samarra’s palaces by Marcus Milwright: “A twin approach is adopted in order to focus upon the balance struck between the patron’s desire that the building should reflect his own power and magnificence through the use of scale and the employment of costly materials on the one hand, and the constraints imposed by habitual building practices and the availability of man-power, materials and fuel on the other” (Milwright 2001, 80; italics ours). 174. Bourdieu 2005, 233–235. 175. Arnold 1928, 88; Fowden 2004, 316–318. 176. Rosenthal 1975, 265–266. 177. Rosenthal 1975, 265–266. 178. It is worth mentioning the examples of the Mamluk bathhouse of Bosra (Meinecke, Aalund, and Korn 2005, 101–170) and the Ayyubid and Mamluk bathing complexes of Damascus (Écochard and Le Coeur 1943). 179. Fowden 2004, 298.
Bibliography Akerstrom-Hougen, G. 1974. The Calendar and Hunting Mosaics of the Villa of the Falconer in Argos. Stockholm: Svenska Institutet i Athen. al-Akhṭal. 1891–1925. Dīwān, edited by P. A. Ṣalḥānī. Beirut: Imprimerie Catholique. Ali, N. 2008. “Processus de création de l’iconographie omeyyade: L’exemple des décors palatiaux du Bilâd ash-Shâm.” Unpublished doctoral diss., Université d’Aix-en-Provence.
2 4 3
Umayyad Palace Iconography
243
———. 2006. “Qusayr ‘Amra, la peinture du personnage trônant sur l’eau: Aspects pratiques de la fabrication d’une image.” Annales Islamologiques 40: 1–31. Allen, T. 1988. “The Arabesque, the Bevelled Style, and the Mirage of an Early Islamic Art.” In Five Essays on Islamic Art, 1–15. Sebastopol, CA: Solipsist Press. ———. 1995. A Lozenge in the Bath Pavement at Khirbat al-Mafjar. Occidental, CA: Solipsist Press. Almagro, M. 1975. Qusayr Amra. Madrid: Ministerio de Asuntos Exteriores. Arnold, T. W. 1928. Painting in Islam. Oxford: Clarendon. Bacharah, J. L. 1996. “Marwanid Umayyad Building Activities: Speculations on Patronage.” Muqarnas 13: 27–44. Baer, E. 1999. “Female Images in Early Islam.” Damaszener Mitteilungen 1: 13–24. ———. 2004. The Human Figure in Islamic Art. Costa Mesa, CA: Mazda. Bakhouch, M. 2001. “Un aspect de la poésie d’al- Akhtal: Le panégyrique.” Unpublished doctoral diss., Université d’Aix-en-Provence. Balty, J. 2003. “La place des mosaïques de Jordanie au sein de la production orientale.” In Les églises de Jordanie et leurs mosaïques, edited by N. Duval, 153–186. Beirut: Institut Français d’Archéologie du Proche-Orient. Barbet, A. 1985. La peinture murale romaine. Paris: Picard. Barbet, A., and C. Vibert-Guigue. 1994. Les peintures des nécropoles romaines du Nord de la Jordanie. Beirut: Institut Français d’Archéologie du Proche-Orient. Barry, F. 2007. “Cosmic Floors in Antiquity and the Middle Ages.” Art Bulletin 89: 627–656. Bartl, K., and A. Moaz (eds.). 2008. Residences, Castles, Settlements: Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham. Rahden, Westf.: Verlag Marie Leidorf. Becker, C. H. 1924–1932. Islamstudien, 2 vols. Leipzig: Quelle und Meyer. Behrens- Abouseif, D. 1997. “The Lion- Gazelle Mosaic at Khirbat al- Mafjar.” Muqarnas 14: 11–18. Bisheh, G. 1993. “From Castellum to Palatium: Umayyad Mosaic Pavements from Qasr al-Hallabat in Jordan.” Muqarnas 10: 49–56. ———. 2002. “Two Umayyad Mosaic Floors from Qastal.” Liber Annuus 50: 431–438. Bisheh, G., T. Morin, and C. Vibert-Guigue. 1997. “Rapport d’activités à Qusayr Amra.” Annual of the Department of Antiquities of Jordan 41: 375–393. Blazquez, J. M. 1981. “Las pinturas helenistícas de Qusayr Amra y sus fuentes.” Archivo Español de Arqueología 54: 157–202. — — — . 1983. “La pintura helenistíca de Qusayr Amra II.” Archivo Español de Arqueología 56: 169–212. Borrut, A. 2011. Entre mémoire et pouvoir: l’espace syrien sous les derniers Omeyyades. Leiden: Brill. Bourdieu, P. 1990. The Logic of Practice. Translated by Richard Nice. Stanford, CA: Stanford University Press.
42
244
N adia A li and M attia G uidetti
— — — . 2005. “Postface to Erwin Panofsky, Gothic Architecture and Scolasticism.” Translated by Laurence Petit in Bruce Holsinger, The Premodern Condition: Medievalism and the Making of a Theory, 221–242. Chicago: University of Chicago Press. Bowersock, G. by Bruce Holsinger, 2006. Mosaics as History: The Near East from Late Antiquity to Islam. Cambridge, MA: Belknap Press of Harvard University Press. Brown, P. 1971. The World of Late Antiquity ad 150–750. New York: Harcourt Brace Jovanovich. Bruneau, P. 1984. “Les mosaïstes antiques avaient-ils des cahiers de modèles?” Revue Archéologique 2: 260–272. ———. 1986. “De l’image.” Revue d’archéologie moderne et d’archéologie générale 4: 249–295. Brunet, J. P., R. Nadal, and C. Vibert-Guigue. 1998. “The Frescoes of the Cupola of Qusayr Amra.” Centaurus 40: 97–123. Butler, H. C. 1904. Architecture and Other Arts. New York: Century Co. Charpentier, G. 1994. “Les bains de Serjilla.” Syria 71: 113–142. Creswell, K. A. C. 1932. Early Muslim Architecture. Pt. 1: Umayyads, a.d. 622–750. Oxford: Clarendon. Crone, P., and M. Hinds. 1986. God’s Caliph: Religious Authority in the First Centuries of Islam. Cambridge: Cambridge University Press. Cruikshank Dodd, E. 1975. “Review of From Byzantium to Sassanian Iran and the Islamic World, Three Modes of Artistic Influence, by Richard Ettinghausen; The Formation of Islamic Art by Oleg Grabar.” Art Bulletin 57.2: 267–270. ———. 1984. “The Mosaic of the Fruit and the Knife in Khirbet al Mafjar.” In Studies in the History and Archaeology of Palestine I, edited by S. Shaath, 317–327. Aleppo: Markaz al-Āthār al-Filasṭīnī. Cytryn-Silverman, K. 2009. “The Umayyad Mosque of Tiberias.” Muqarnas 26: 38–44. De Loos-Dietz, E. P. 1990. “Les mosaïques à Khirbat al-Mafjar près de Jéricho.” BaBesch 65: 123–138. Doncel-Voûte, P. 1988. Les pavements des églises byzantines du Syrie et du Liban: Décor, archéologie et liturgie. Louvain-La-Neuve: Département d’Archéologie et d’Histoire de l’Art. Donner, F. M. 2010. “Umayyad Effort at Legitimation: The Umayyads’ Silent Heritage.” In Umayyad Legacies: Medieval Memories from Syria to Spain, edited by A. Borrut and P. M. Cobb, 187–213. Leiden: Brill. Dunbabin, K. M. D. 1978. The Mosaics of Roman North Africa: Studies in Iconography and Patronage. Oxford: Clarendon. ———. 1999. Mosaics of the Greek and Roman World. Cambridge: Cambridge University Press. Dvorjetski, E. 1999. “Social and Cultural Aspects of Medicinal Baths in Israel, Roman Baths and Bathing.” Journal of Roman Archaeology, Supplementary Series 37.1: 117–129. Écochard, M., and C. Le Coeur. 1943. Les bains de Damas. Beirut: Institut Français de Damas.
2 4 5
Umayyad Palace Iconography
245
Ettinghausen, R. 1977. Arab Painting. 2nd edition. Genève: Skira. ———. 1972. From Byzantium to Sasanian Iran and the Islamic World: Three Modes of Artistic Influence. Leiden: Brill. Finster, B. 1996. “Arabia in der Spätantike.” Archäologischer Anzeiger 2: 287–319. Flood, F. B. 2001. The Great Mosque of Damascus: Studies on the Making of Umayyad Visual Culture. Leiden: Brill. Focillon, H. 1938. Art d’Occident: Le Moyen-Âge roman et gothique. Paris: A. Colin. Fowden, G. 1993. Empire to Commonwealth: Consequences of Monotheism in Late Antiquity. Princeton, NJ: Princeton University Press. ———. 2004a. Qusayr Amra: Art and the Umayyad Elite in Late Antique Syria. Berkeley: University of California Press. ———. 2004b. “Late Antique Art in Syria and Its Umayyad Evolution.” Journal of Roman Archaeology 17: 282–304. — — — . 2007. “Greek Myth and Arabic Poetry at Qusayr ‘Amra.” In Islamic Crosspollinations: Interactions in the Medieval Middle East, edited by A. Akasoy et al., 29–45. Cambridge: Gibb Memorial Trust. ———. 2011. “Contextualizing Late Antiquity: The First Millennium.” In The Roman Empire in Context: Historical and Comparative Perspectives, edited by J. P. Arnason and K. A. Raaflaub, 148–176. Chichester Wiley-Blackwell. Fowden, G., and E. K. Fowden. 2004. Studies on Hellenism, Christianity and the Umayyads. Athens: Kentron Hellēnikēs kai Rōmaïkēs Archaiotētos. Genequand, D. 2005. “From ‘Desert Castle’ to Medieval Town: Qasr al-Hayr al-Sharqi (Syria).” Antiquity 79/304: 350–361. ———. 2006. “Umayyad Castles: The Shift from Late Antique Military Architecture to Early Islamic Palatial Buildings.” In Military Architecture in Greater Syria: From the Coming of Islam to the Ottoman Period, edited by H. Kennedy, 10–22. Leiden: Brill. ———. 2009. “Economie de production, affirmation du pouvoir et dolce vita: Aspects de la politique de l’eau sous les Omeyyades du Bilad al-Sham.” In Stratégies d’acquisition de l’eau et société au Moyen-Orient depuis l’Antiquité, edited by M. al- Dbiyat and M. Mouton, 157–177. Beirut: Institut Français du Proche-Orient. ———. 2012. Les établissements des élites omeyyades en Palmyrène et au Proche-Orient. Beirut: Institut Français du Proche-Orient. Grabar, A. 1983. “Quelques observations sur la mosaïque perdue de Carthage.” In Mosaïques: Recueil d’hommages à Henri Stern, 189–194. Paris: Editions Recherche sur les Civilisations. Grabar, O. 1954. “Paintings of the Six Kings at Qusayr ‘Amrah.” Ars Orientalis 1: 185–187. ———. 1955. “Ceremonial and Art at the Umayyad Court.” Unpublished PhD diss., Princeton University. ———. 1973. The Formation of Islamic Art. New Haven, CT: Yale University Press. ———. 1988. “La place de Qusayr ʿAmra dans les arts profanes de la haute antiquité.” Cahiers Archéologiques 36: 76–83.
6 42
246
N adia A li and M attia G uidetti
———. 1992. “L’art omeyyade de Syrie, source de l’art islamique.” In La Syrie de Byzance à l’Islam VII-VIIIe siècles, edited by P. Canivet and J. P. Rey-Coquais, 187–194. Damascus: Institut Français de Damas. ———. 1993. “Umayyad Palaces Reconsidered.” Ars Orientalis 23: 92–108. ———. 2005. Constructing the Study of Islamic Art I: Early Islamic Art, 650–1100. Burlington, VT: Ashgate/Variorum. Grabar, O., Ettinghausen, E., and M. Jenkins-Madina. 2001. Islamic Art and Architecture 650–1250. New Haven: Yale University Press. Guidetti, M. 2010. “L’editto di Yazid II’: Immagini e identità religiosa nel Bilad al- Sham dell’VIII secolo.” In L’VIII secolo: un secolo inquieto, edited by V. Pace, 69– 79. Cividale del Friuli: Comune di Cividale del Friuli. Guidetti, M., and G. Macchiarella. 2007. “Problemi di ermeneutica nell’iconografia umayyade: Qusayr ʿAmra e Khirbat al-Mafgar.” In Immagine e ideologia: scritti in onore del prof. Quintavalle, edited by A. Calzona, R. Campari and M. Mussini, 53–64. Milan: Electa. Hamilton, R. 1959. Khirbat al-Mafjar: An Arabian Mansion in the Jordan Valley. Oxford: Clarendon. ———. 1969. “Who Built Khirbat al-Mafjar?” Levant 1: 61–67. ———. 1978. “Khirbat al-Mafjar: the Bath Hall Reconsidered.” Levant 10: 126–138. ———. 1988. Walid and His Friends: An Umayyad Tragedy. Oxford: Oxford University Press. Hanfmann, G. 1951. The Season Sarcophagus in Dumbarton Oaks. Cambridge, MA: Harvard University Press. Harper, R. P., and T. J. Wilkinson. 1975. “Excavations at Dibsi Faraj, Northern Syria, 1972–1974: A Preliminary Note on the Site and its Monuments with an Appendix.” Dumbarton Oaks Papers 29: 319–338. Hartner, W., and R. Ettinghausen. 1964. “The Conquering Lion: The Life Cycle of a Symbol.” Oriens 17: 161–171. Harvey, W. 1935. Structural Survey of the Church of the Nativity, Bethlehem. London: Oxford University Press. Hawting, G. 1987. The First Dynasty of Islam: The Umayyad Caliphate ad 661–750. London: Croom Helm. Hillenbrand, R. 1982. “La Dolce Vita in Early Islamic Syria: The Evidence of Later Umayyad Palaces.” Art History 5.1: 1–35. ———. 1988. “The Symbolism of the Rayed Nimbus in Early Islamic Art.” Cosmos 2: 1–52. ———. 1994. Islamic Architecture: Form, Function and Meaning. New York: Columbia University Press. Hirschfeld, Y. 1997. The Roman Baths of Hammat Gader: Final Report. Jerusalem: Israel Exploration Society. Hoffman, E. (ed.). 2007. Late Antique and Medieval Art of the Mediterranean World. Malden, MA: Blackwell.
4 2 7
Umayyad Palace Iconography
247
Hoyland, R. 2001. Arabia and the Arabs from the Bronze Age to the Coming of Islam. London: Routledge. Jamil, N. 1999. “Caliph and Qutb: Poetry as a Source for Interpreting the Transformation of the Byzantine Cross on Steps on Umayyad Coinage.” In Bayt al-Maqdis: Jerusalem and Early Islam, edited by J. Johns, 11–57. Oxford: Oxford University Press. Judd, S. 2008. “Reinterpreting al-Walid b. Yazid.” Journal of the American Oriental Society 128.3: 439–458. Keall, E. J. 1995. “Forerunners of Umayyad Art: Sculptural Stone from the Hadramawt.” Muqarnas 12:11–23. Kennedy, H. 1986. The Prophet and the Age of the Caliphates: The Islamic Near East from the Sixth to the Eleventh Century. London: Longman. Khoury, G. 1992. “Calife ou roi: du fondement théologico-politique du pouvoir suprême dans l’islam sous les califes orthodoxes et Omeyyades.” In La Syrie de Byzance à l’islam VII-VIIIe siècles, edited by P. Canivet and J. P. Rey-Coquais, 323–332. Damascus: Institut Français de Damas. Kilpatrick, H. 2010. “‘Umar ibn ʿAbd al-ʿAziz, al-Walid ibn Yazid and their Kin: Images of the Umayyads in the Kitâb al-Aghâni.” In Umayyad Legacies: Medieval Memories from Syria to Spain, edited by A. Borrut and P. M. Cobb, 63–89. Leiden: Brill. King, G. 1980. “Some Christian Wall Mosaics in Pre-Islamic Arabia.” Proceedings of the Seminar for Arabian Studies 10: 37–43. ———. 1991. “Creswell’s Appreciation of Arabian Architecture.” Muqarnas 8: 94–102. ———. 1992. “Settlements Patterns in Islamic Jordan: The Umayyad Use of the Land.” Studies in the History and Archaeology of Jordan 4: 369–375. Kitzinger, E. 1970. “The Threshold of the Holy Shrine: Observations on Floor Mosaics at Antioch and Bethlehem.” In Kyriakon: Festschrift Johannes Quasten, edited by P. Granfield and J. A. Jungmann, 2:639–647. Münster: Aschendorff. ———. 1976. “Stylistic Developments in Pavement Mosaics in the Greek East from the Age of Constantine to the Age of Justinian.” In The Art of Byzantium and the Medieval West: Selected Studies, edited by E. Kleinbauer, 64–88. Bloomington, IN: Indiana University Press. Klein, R. 1974. La forme et l’intelligible. Paris: Gallimard. Kraeling, C. H. 1938. Gerasa: City of the Decapolis. New Haven, CT: American Schools of Oriental Research. Lammens, H. 1930. Études sur le siècle des Omayyades. Beirut: Imprimerie Catholique. Lauffray, J. 1983. Ḥalabiyya- Zenobia, place forte du limes oriental et la Haute- Mésopotamie au VIe siècle: L’architecture publique, religieuse, privée et funéraire. Paris: Geuthner. Lavin, I. 1963. “Antioch Hunting Mosaics and Their Sources.” Dumbarton Oaks Papers 17: 179–286. ———. 1967. “The Ceiling Frescoes in Trier and Illusionism in Constantinian Painting.” Dumbarton Oaks Papers 21: 97–113.
8 4 2
248
N adia A li and M attia G uidetti
Lehmann, K. 1945. “The Dome of Heaven.” Art Bulletin 27: 1–27. Leisten, T. 2009. “For Prince and Country(side)—the Marwanid Mansion at Balis on the Euphrates.” In Residences, Castles, Settlements: Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham, edited by Karin Bartl and Abd al- Razzaq Moaz, 377–394. Rahden, Westf.: Verlag Marie Leidorf. Levi, D. 1941. “The Allegories of the Months in Classical Art.” Art Bulletin 23: 251–292. Ling, R. 1991. Roman Painting. Cambridge: Cambridge University Press. L’Orange, H. P. 1953. Studies on the Iconography of Cosmic Kingship in the Ancient World. Oslo: H. Aschehoug. Magness, J. 2010. “Early Islamic Urbanism and Building Activity in Jerusalem and Hammath Gader.” In Money, Power and Politics in Early Islamic Syria, edited by J. Haldon, 153–162. Burlington, VT: Ashgate. Maguire, H. 1999. “The Good Life.” In Late Antiquity: A Guide to the Postclassical World, edited by G. W. Bowersock, P. Brown, and O. Grabar, 238–257. Cambridge, MA: Belknap Press of Harvard University Press. ———. 2007. Image and Imagination in Byzantine Art. Burlington, VT: Ashgate. ———. 2012. Nectar and Illusion: Nature in Byzantine Art. Oxford: Oxford University Press. Marsham, A. 2009a. Rituals of Islamic Monarchy: Accession and Succession in the First Muslim Empire. Edinburgh: Edinburgh University Press. ———. 2009b. “The Early Caliphate and the Inheritance of Late Antiquity (c. ad 610—c. ad 750).” In A Companion to Late Antiquity, edited by P. Rousseau, 479– 492. Chichester: Wiley-Blackwell. Meinecke, M., F. Aalund, and L. Korn. 2005. Bosra: Islamische Architektur und Archäologie. Rahden, Westf.: Verlag Marie Leidorf. Michaelides, D. 1992. Cypriot Mosaics. Nicosia: Department of Antiquities (first edition, 1987). Michel, A. 2001. Les églises d’époque byzantine et Umayyade de la Jordanie. Ve-VIIIe siècle. Turnhout: Brepols. Milwright, M. 2001. “Fixtures and Fittings: The Role of Decoration in Abbasid Palace Design.” In A Medieval Islamic City Reconsidered: An Interdisciplinary Approach to Samarra, edited by C. Robinson, 79–109. Oxford: Oxford University Press. ———. 2005. “‘Waves of the Sea’: Responses to Marble in Written Sources (9th– 15th Centuries).” In The Iconography of Islamic Art: Studies in Honour of Robert Hillenbrand, edited by B. O’Kane, 211–221. Edinburgh: Edinburgh University Press. — — — . 2010. “Archaeology and Material Culture.” In The New Cambridge History of Islam, edited by C. Robinson, 1:664–681. Cambridge: Cambridge University Press. Morony, M. G. 1981. “Bayn al-fitnatayn: Problems of Periodization of Early Islamic History.” Journal of Near Eastern Studies 40.3: 247–251. Mundell Mango, M. 2002. “Excavations and Survey at Androna, Syria: The Oxford Team 1999.” Dumbarton Oaks Papers 56: 303–311.
9 42
Umayyad Palace Iconography
249
———. 2008. “Baths, Reservoirs and Water Use at Androna in Late Antiquity and the Early Islamic Period.” In Residences, Castles, Settlements: Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham, edited by K. Bartl and A. Moaz, 73–88. Rahden, Westf.: Verlag Marie Leidorf. Musil, A. 1907. Kusejr Amra. Vienna: Kaiserliche Akademie der Wissenschaften. Nau, F. 1929–1930. “Le traité sur les ‘constellations’ écrit, en 661, par Sévère Sébokt évêque de Qennesrin.” Revue de l’Orient Chrétien 27: 327–410. Necipoglu, G. 1993. “An Outline of the Shifting Paradigms on Palatial Architecture in the Pre-Modern Islamic World.” Ars Orientalis 23: 3–24. Nuseibeh, S., and O. Grabar. 1996. The Dome of the Rock. New York: Rizzoli. Pächt, O. 1999. The Practice of Art History. London: Harvey Miller. Pappalardo, U., and R. Ciardiello. 2010. Mosaici greci e romani: Tappeti di pietra in età ellenistica e romana. San Giovanni Lupatoto: Arsenale. Paribeni, A. 2004. “Osservazioni sui mosaici pavimentali dei palazzo omayyadi.” In Medioevo: Arte Lombarda, edited by C. A. Quintavalle, 634–642. Milan: Electa. Paris. 2000. L’art copte en Egypte: 2000 ans de christianisme (exposition présentée à l’Institut du Monde Arabe). Paris: Institut du Monde Arabe. Parrish, D. 1984. Season Mosaics of Roman North Africa. Rome: Giorgio Bretschneider Editore. Piccirillo, M. 1993. The Mosaics of Jordan. Amman: American Center of Oriental Research. Pringle, D. 1993. The Churches of the Crusader Kingdom of Jerusalem: A Corpus. Volume I. Cambridge: Cambridge University Press. Rabbat, N. 1990. “The Meaning of the Umayyad Dome of the Rock.” Muqarnas 6: 12–21. ———. 1993. “The Dome of the Rock Revisited: Some Remarks on al-Wasiti’s Accounts.” Muqarnas 10: 67–75. ———. 2003. “The Dialogic Dimension of Umayyad Art.” RES: Anthropology and Aesthetics 43: 79–94. Reau, L. 1951. “L’influence de la forme sur l’iconographie dans l’art médiéval.” Journal de psychologie normale et pathologique 44: 85–102. Reinach, A. 1922. Répertoire des peintures gréco-romaines. Paris: E. Leroux. Rice, D. 1954. “The Seasons and the Months in Islamic Art.” Ars Orientalis 1: 1–19. Rosenthal, F. 1975. The Classical Heritage in Islam. Berkeley: University of California Press. Saller, S. J. 1949. The Town of Nebo. Jerusalem: Franciscan Press. Sauerlander, W. 2004. “L’art des sculpteurs romans et le retour à l’ordre.” In Catalogue de l’exposition La vie des formes: Henri Focillon et les arts, edited by C. Briend, 147– 152. Gand: Snoeck. Saxl, F. 1931. “The Zodiac of Qusayr Amra.” In Early Muslim Architecture, by K. A. C. Creswell, 424–431. Oxford: Clarendon. Schlumberger, D. 1980. Qasr al-Hayr al-Gharbi. Paris: Geuthner.
0 52
250
N adia A li and M attia G uidetti
Shboul, A., and A. Walmsley. 1998. “Identity and Self-Image in Syria and Palestine in the Transition from Byzantine to Early Islamic Rule: Arab Christians and Muslims.” Mediterranean Archaeology 11: 255–287. Soucek, P. 1993. “Solomon’s Throne/Solomon’s Bath: Model or Metaphor?” Ars Orientalis 23: 109–134. Stern, H. 1953. Le calendrier de 354. Paris: Geuthner. Strzygowski, J. 1888. Die Calenderbilder des Chronographen vom Jahre 354, Jahrbuch des kaiserlichen deutschen archäologischen Instituts, Ergänzungsheft 1. Berlin: G. Reimer. Taragan, H. 1993. “The Peopled Scrolls at the Umayyad Palace in Jericho: Some Observations.” Assaph: Studies in Art History (Section B, no 3): 93–108. ———. 2001. “A Matter of Looking: The Female Images in the Umayyad Palace at Khirbat al-Mafjar.” In The Metamorphosis of Marginal Images: From Antiquity to Present Time, edited by N. Kenaan-Kedar and A. Ovadiah, 69–78. Tel Aviv: Tel Aviv University. ———. 2003. “Atlas Transformed—Interpreting the ‘Supporting Figures’ in the Umayyad Palace at Khirbat al-Mafjar.” East and West 53: 9–29. Treadwell, L. 1999. “The Orans Drachms of Bishr ibn Marwan and the Figural Coinage of the Early Marwanid Period.” In Bayt al-Maqdis: Jerusalem and Early Islam, edited by J. Johns, 223–268. Oxford: Oxford University Press. Ulbert, Th. 1986. Die Basilika des Heiligen Kreuzes in Resafa-Sergiupolis. Mainz: P. von Zabern. Van Lohuizen- Mulder, M. 1998. “The Frescoes in the Muslim Residence and Bathhouse Qusayr Amra.” BaBesch 73: 125–151. Vasaros, Z. 2004. “Second Century Illusionary Ceiling-Painting and Its Architectural Origins.” In Plafonds et voûtes à l’époque antique: Actes du VIIIe Colloque International de l’Association Internationale pour la Peinture Murale Antique (AIPMA), edited by L. Borhy, 61–66. Budapest: Pytheas. Vasiliev, A. 1956. “The Iconoclastic Edict of the Caliph Yazid II, a.d. 721.” Dumbarton Oaks Papers 9/10: 23–47. Vibert-Guigue, C., and G. Bisheh. 2007. Les peintures de Qusayr ‘Amra. Beirut: Institut Français d’Archéologie du Proche-Orient. Vitruvius. 1914. The Ten Books on Architecture. Translated by M. H. Morgan. Cambridge, MA: Harvard University Press. Walmsley, A. 2007. Early Islamic Syria. London: Duckworth. Weitzmann, K. (ed.). 1979. Age of Spirituality: Late Antique and Early Christian Art, Third to Seventh Century. New York: Metropolitan Museum of Art. Winkler- Horacek, L. 1998. “Dionysos in Qusayr ‘Amra— Ein Hellenistisches Bildmotiv im Fruhislam.” Damaszener Mitteilungen 10: 261–290.
251
Umayyad Palace Iconography
251
Yegül, F. 1992. Baths and Bathing in Classical Antiquity. Cambridge, MA: MIT Press. ———. 2010. Bathing in the Roman World. Cambridge: Cambridge University Press. Zammit, T. 1930. “Roman Villa and Thermae at Ghajn Tuffieha, Malta.” Bulletin of the Museum, Valletta, Malta 1.2: 56–64. Zayyadine, F. 1979. “The Umayyad Frescoes of Quseir ‘Amra.” Archaeology 31: 19–29.
25
235
PART III
Historiography and Historical Memory
4 2 5
25
8
Christians in Umayyad Iraq Decentralization and Expansion (600–7 50) Philip Wood
not loomed large in the study of the Umayyad period. Its notables and its institutional structures are invisible in the Muslim Arabic histories, such as al-Ṭabarī and al-Dīnawarī, which are normally used to reconstruct the period. And, unlike the better-known Christianities of the Levant, its relationship with the Christian Roman Empire and its Byzantine successor was both more attenuated and more temporary. Iraqi Christianity does not fit into the neat narratives of Roman collapse and Islamic victory that might suit a version of history centered on states. Instead, I argue here, it makes better sense to see Iraqi Christianity as a Sasanian institution (or set of institutions) that outgrew and survived the fall of the Sasanian Empire.1 Using sources written from the central perspective of the catholicoi of Seleucia-Ctesiphon, and monastic histories and hagiographies written in the less urban north of Iraq, I suggest that we can chart the development of a set of cultural and political compromises, which saw its boundaries and internal power balance change in a way that was distinctive to the Umayyad period in general and to the experience of Umayyad Iraq in particular. IRAQI CHRISTIANITY HAS
A Sasanian Church At the end of the sixth century, the Church of the East, known to later generations as the “Nestorian” Church, was governed by a catholicos, ruling
6 52
256
P hilip W ood
from the royal capital of Ctesiphon. The catholicos’ relationship with the state was considerably weaker than that of Roman patriarchs with the emperor,2 but there is a strong correlation between the improved relationship between shah and catholicos and the ability to convene synods to summon bishops to the capital.3 The increasing number of bishops who attended these synods is, of course, no certain indication of expanding Christian numbers, but it is suggestive of Christianity’s growing prestige and organization. Equally important indicators of this prestige are the large number of official seals that deployed Christian symbols for men bearing Iranian names,4 or the issuing of provincial bronze coins bearing the sign of the cross.5 The historian al-Ṭabarī (839–923) even reports that the shah Hōrmīzd IV (r. 579–590) himself advocated religious pluralism within his realm: “My empire is supported by its different faiths just as my throne is supported by four legs.”6 But if the growth of church structures and synodal rulings are indications of the status of the catholicos, this status came at the high price. Christian bishops played an important role in passing on military information from the Roman frontier,7 acting as diplomats,8 and praying for the shah as a righteous and God-appointed sovereign. The Ctesiphon assembly of 612 declared that the new shah, Khusrau II (r. 590–628), had “lit up the world like the sun.” “We,” declared the bishops, “are the luckiest men since Adam to live in the glory of your empire.”9 The most striking example of the compromise between catholicosate and shah comes in the dramatic interference in the first election of a new catholicos of Khusrau’s reign. The Arabic Life of Sabrīshōʿ describes how the shah selected a catholicos for a divided congregation. Khusrau’s emissary informs the crowd that “this man [Sabrīshōʿ] is the chief whom God has sent you from the heavens, whom the shah has placed at your head. Celebrate his election according to your canons … and receive his blessing,” whereupon the bishops fall to the floor to kiss Sabrīshōʿ’s feet and praise Khusrau.10 Interestingly, this hagiography frames no objection to Khusrau’s interference in the holy man’s election, or to Khusrau’s co-option of Sabrīshōʿ (r. 596–604) into his retinue during his invasion of the Roman Empire in 604. But it is important to remember that Khusrau’s interference in elections was only the apex of a broader exchange of influence between secular elites of Iranian culture and the institutional church. For just as Khusrau picked a catholicos, so too lay Christian aristocrats sponsored churches, monasteries, and schools and exercised their own influence
2 75
Christians in Umayyad Iraq
257
over episcopal elections.11 And as much as the synodica complains about such interference, we can also read it as an indicator of the deepening interest of aristocrats in Christian institutional structures.12 Monasteries in particular served as economic insurance for aristocrats, and the profusion of monastic foundations in the period 580–720 may be associated with the appearance of these new sources of funding for new monastic settlements that, in spite of their protestations of poverty, were well integrated into a monetized economy.13
A Change of Masters The assassination of Khusrau II, possibly at the hands of a disgruntled Christian aristocrat, Shamṭā bar Yazdīn,14 was followed swiftly by occupation by the Roman forces of the emperor Heraclius and a series of coups against Khusrau’s short-lived successors.15 This civil war left the Sasanian realm in little position to defend itself against a new and unexpected foe from the southwest, and the Islamic invasions of Iraq and Iran were rewarded with rapid and decisive victory. With the defeat of the last shahs, the state that had supported Zoroastrianism as an official religion collapsed. This loss of official status allowed ambitious elites to adopt new kinds of religious behavior, now that the special relationship between Zoroastrianism and the laws and ethnic boundaries of the Persians had been challenged.16 One Zoroastrian apocalyptic text composed in this era, the Zāmāsp-Nāmag, complains that “ērān and anērān will be confounded so that Iranian will not be distinguished from the foreigner and Iranians will return to foreign ways.” The apex of the sign of the end-times, to this writer, is not just ethnic mixing and the loss of tradition, but also the disappearance of the orthodox hierarchies of age and wealth and the appearance of new forces of competition between different regions of the former Sasanian Empire: “the inferior and obscure will come to notice … and the districts will vie with one another.”17 The complaints of the Zāmāsp-Nāmag may be explained in part by the defection of parts of the Sasanian army to the Arab conquerors. Al- Balādhurī’s example of the defection of a unit of elite cavalry, the asāwira, is the most famous instance of this, but it is accompanied by numerous other instances of low-level aristocrats, dihqāns, taking on roles as administrators, tax collectors, and warriors for the early caliphate, often after conversion to Islam.18
8 2 5
258
P hilip W ood
Yet, at the same time, other religious groups also benefited from the collapse of the Sasanians. The bronze coinage of this period implies that Christian bishops or aristocrats stepped in to fill the lacunae left by the state, issuing their own cross-bearing coins to their communities.19 And the new political environment meant the end of Sasanian-era constraints on proselytism: the long seventh century also saw substantial gains by Christian missions and their celebration in hagiography. This was the era of John of Daylam’s missions into Fars and the foothills of the Caspian, where the saint is meant to have spread Christianity into new, Persian- speaking regions that formerly lay beyond the missionary scope of the Church.20 And monastic hagiographies, of the kind that are embedded in the tenth-century Chronicle of Seert, report an increase in the conversion of polytheists in the former Sasanian world in the aftermath of the conquest.21 The impressive extent of these missions beyond the borders of the old empire is testified by the famous bilingual inscription in Chinese and Syriac at Chang’an, which heralds the beginning of the Church’s presence in the Far East in 635.22 Notably, the advantages of this “free market” were not limited to Christians: the same era also saw the reestablishment of contacts between Manicheans in Iraq and Central Asia and attempts to reunite the two groups under a single archegos.23 The missionary expansion of the Church of the East in the Sufyanid period was closely connected to the monastic movement, which continued the growth it experienced at the end of the Sasanian period. The ninth- century Book of Governors, the monastic history of Thomas of Marga, reports the activities of the holy men of the monastery of Beth ʿĀbe in converting nearby villages and constructing churches. Its monks are praised for their role in destroying pagan holy sites, such as the holy oak of Shīrwān, and for converting Zoroastrian and pagan villages through spectacular public healings and miracles.24 This missionary movement was backed up by a spate of new literary compositions and building work to support the education of this newly Christianized population, ranging from the construction of schoolhouses to teach literacy, theology, and music, to the arrangement of a system of visitations and the composition of a standard liturgical cycle.25 The Book of Governors, and the East Syrian hagiography of the seventh century in general, is dominated by the issue of the acquisition of funding from lay patrons. The majority of cases do not discuss the higher clergy, but lay elites, many of them explicitly labeled dihqāns. The same category of Iranian aristocrats that found ready employment as Arab mawālī in the
9 52
Christians in Umayyad Iraq
259
south of Iraq and in Central Asia were able to occupy a different, and more independent, political niche in the north.26 The stories reported in the hagiographies, where such men fund and create new monasteries, and sometimes convert to Christianity if their families are not already Christian, are, in one sense, an advertisement for the possibilities of this symbiosis. But they also set out a paradigm for its limits, for the political and cultural compromises that must be made to render such patronage safe for monasteries and the Christian communities they wished to lead.
Paradigms of Behavior for Monastic Patrons The first, and most obvious, of the requirements demanded of aristocratic funders was conversion to Christianity. Hagiographies set in the aftermath of the Arab conquests assert the importance of religious conversion as a prerequisite for attaining miraculous protection. The History of the Convent of Beth Qōqā includes miracles at the level of local politics under successive abbots from the seventh to eleventh centuries, and seems to have been continuously composed under each abbot. Its stories justify the legal claims of the monastery and present patterns for the interaction of the monastery and locals. Thus vignettes tell how the holy man Sabrīshōʿ rescued local villagers from the Arab invaders or how an Arbelan magnate became a monk after being cured from leprosy and gave his “woods and fields” to the monastery.27 But, notably, among these brief scenes is a cautionary tale: a Sasanian marzbān (a high-ranking officer) flees from the Arabs to Sabrīshōʿ and asks to be saved. But Sabrīshōʿ tells him it will only be possible if he converts. The marzbān refuses and is killed. Against the background of Sabrīshōʿ’s success in protecting the villagers, the scene emphasizes the Christian nature of local solidarity, and the comparative ineffectiveness of the offices of the Sasanian state when compared to an act of God, such as the Arab invasion, or to God’s agents, such as this Christian holy man. Individuals might also consider themselves Christian while still carrying the assumptions and customs of their older religion.28 A good example of this, where the hagiographer is keen to assert “Christian” economic and political behavior, is a scene from ca. 660, reported in Thomas of Marga’s Book of Governors. Here one Ḥūgair wishes to found a new monastery, “not for a godly intention, but for boasting and pride. Acting as if he was a good man, he named it Ḥūgair-Abad, in the style of the Magians from whose race he had come.” However, Aḥā, metropolitan of Marga
0 6 2
260
P hilip W ood
and former head of the monastery, refused to consecrate it and bring his monks and scribes there, saying “The house of Ḥūgair-Abad is ruined before it is built.”29 Maria Macuch notes that the Zoroastrian “Church” made no pious foundations of its own, and instead individuals made their own foundations, which provided ceremonies for the deceased, but also gave their founder inalienable rights over their religious estates. These estates would then grow cash crops and provide loans at interest to be used for prestigious local “charity,” such as building bridges and canals.30 Ḥūgair probably made similar assumptions about funding a monastery in his newly adopted religion, namely that he could enjoy a tax-protected investment as well as making a prestigious display of his wealth. The hagiography celebrates how a protégé of the monastery of Beth ʿĀbe maintained the principles of monastic independence in the face of this offer of “tied funding” following a Zoroastrian norm, and the story itself serves as a parable for the dependence of donors on the approval of the monastery, in spite of whatever material goods they might offer. The need to control and redirect the economic behavior of former Zoroastrians is displayed in a different way in the course of the missionary monk Īshōʿzkhā, another hero of the Book of Governors. He was associated with the recently converted villages of Salākh, where he sought to publicly humiliate local aristocrats who failed to treat him with proper respect and destroy the sacred places of older pagan practice. In one of these villages, his conflict with the locals was not centered on conversion per se, but on economic provision for Christian institutions and on the inheritance of property. Here, in the village of Gōlai, the Zoroastrian inhabitants had converted to Christianity and set aside a grove of nut trees and a small plantation to provide for their new church. After the death of the first generation of converts, their children challenged the arrangement with the priests and insisted that the trees belonged to them. To reconcile the two parties, Īshōʿzkhā called on the trees to move to the church on their own, a miracle that convinced the children of the converts to give up their claim. All the more convincing, to the hagiographer, was the detail that half of the produce from one of these trees had been given to a poor old woman by the villagers, and that this tree had divided itself in two halves, one of which remained in place at the center of the village.31 Unlike Aḥā’s defiance of Ḥūgair, Īshōʿzkhā’s miracle concerns an entire village. It is possible that the nut trees that were in dispute had been the shared property of the village as a whole and that the church’s priests were asserting their ownership of them against longer established traditional
261
Christians in Umayyad Iraq
261
rights, even if the initial conversion had seen an actual gift of property. Indeed, the detail about the half nut tree that remains with the old woman may be an indication that the hagiography explained a concession that was mediated by the bishop, whereby a traditional pre-conversion charitable arrangement was recognized and normalized through a Christian miracle. This arrangement also set reasonable boundaries to the property of the church’s priests, who may well have been outsiders, given the village’s recent conversion.
The Monks and the Shahrēgān A more complex example of the control and censure of local aristocratic patronage by the monks of Beth ʿĀbe comes in the Life of Maranʿammeh, the lengthy third part of the Book of Governors set at the end of the seventh century. Maranʿammeh, like Īshōʿzkhā, seems to have been associated with a particular group of villages in the vicinity of the mountains of Izla. But, unlike Īshōʿzkhā, Maranʿammeh’s villages seem to have been dominated by a powerful upper level of aristocrats, the shahrēgān, who dominated both the local peasantry and the dihqāns as property owners and tax collectors on behalf of the Muslim caliphate. No parallel for these powerful, independent aristocrats exists in the south of Iraq in the same period, where the administrative capacity of the caliphate was much more developed.32 By contrast, these northern, upper-level aristocrats seem to have done very well out of the immediate aftermath of the conquests and may have profited considerably from the farming out of the tax system.33 In addition, the ability of the shahrēgān to maintain their roles as local landowners and caliphal tax officers was mirrored by a second sphere of “liminal behavior” in their religion. Strikingly, Maranʿammeh accuses them of “calling themselves Christians but saying that Christ was but an ordinary man.”34 The arrival of Islam allowed aristocrats of Iranian culture to place themselves on the edge of different religious traditions and, in so doing, claim leadership of Christian communities, while simultaneously challenging the dominant role of the clergy and the monks of Izla or Beth ʿĀbe. Thomas of Marga presents the monks of Beth ʿĀbe as enjoying a good relationship with the lower aristocracy: the dihqāns of Adiabene are said to have insisted on a bishop from among the disciples of Jacob of Beth ʿĀbe.35 Men of similar standing are presented as Maranʿammeh’s clients in his struggle against the shahrēgān. In their role as tax collectors, the shahrēgān
6 2
262
P hilip W ood
are said to seize “half of the corn, wine, nuts and poll-tax from the people and the dihqāns,” a situation that may combine customary rights and rents with an excessive implementation of the demands of the governor in Mosul.36 This agonistic relationship between the weaker dihqāns, who seek Maranʿammeh as an ally, and the more powerful shahrēgān stands at the background to Maranʿammeh’s condemnation of the lands of the shahrēgān for their Zoroastrian practices and for the Islamizing heresy of the shahrēgān themselves. Maranʿammeh’s tour of the region of Salākh is dominated by his condemnation of polygamy and sexual misbehavior, especially that of the shahrēgān. In one such instance, Maranʿammeh is moved to intercede on behalf of the wife of the shahrig Armenāzwai after he commits adultery with one of the nuns from the convent Beth Ṭeḥūnai. The shahrig ignores the saint’s chastisement and moves to strike him, whereupon the holy man curses him: “I trust in our Lord that you and all your village will go down alive into Sheol like Korah, Daithan and Abiram.” Following the saint’s warning, the shahrig and his village are swallowed up in the middle of the night.37 Similar scenes of destruction are performed at a number of villages on the saint’s itinerary in Salākh and Zab. The hagiographer’s chief theme here is to emphasize the fleeting nature of the wealth of the shahrēgān as sinful men who had opposed God and his agent Maranʿammeh. To the shahrig Zādhai, the owner of some seventy-two estates in the region of Nineveh, he announces, “You shall fall from all this glory, and your estates shall be taken from you and you shall die of hunger,” while another landowner, Shāpur, is slain by Arab raiders only a day after the saint’s visit.38 The disappearance of the Sasanian state had afforded great opportunities for the shahrēgān’s political and religious independence in the uplands of northern Iraq, but, as we see in Thomas of Marga, it was an independence that was challenged by their monastic neighbors, who continued to articulate their own moral leadership of the Christians of the north and the illegitimacy of the shahrēgān. This final scene, in which the saint prophesies the destruction of numerous village-estates dependent on the shahrēgān, serves as a warning to the aristocrats of the north to respect the censure of the monks, but it is also a response to the precipitous decline of the fortunes of these aristocrats after the imposition of a new central authority under the Marwanids. As in Egypt during the same period, the aftermath of the second fitna reduced the opportunities available for older, pre-Islamic elites, and it may be this change of the intensity of Arab rule
6 23
Christians in Umayyad Iraq
263
that allowed Maranʿammeh’s hagiographer to parade the saint’s victory over the pride of Armenāzwai and his fellows.
Arab Elites and Jacobite Foes: The Life of Rabban Hōrmı z̄ d Thomas of Marga describes how monks and holy men tried to assert their authority over the independent-minded native aristocracies of northern Iraq. But other hagiographers, such as the anonymous hagiographer of the seventh-century saint Rabban Hōrmīzd, did not only emphasize the relationship between the saint and local aristocrats of Iranian culture, but also the relationship with the newly arrived Arab governors of Mosul and their local appointees. Here, with the saint’s appeals to Arab elites for protection and his provision of healing miracles for these newly dominant aristocrats, the hagiographer, probably writing soon after Hōrmīzd’s demise, sought to include them in the same kind of reciprocal relationship that had been established for Iranian notables a century earlier. This quest for patronage and protection was especially fraught because, unlike Thomas of Marga’s history, Hōrmīzd was operating on the very edge of the region loyal to the Church of the East. Very close to his grotto lay villages loyal to the “Jacobites,” to Syriac-speaking Christians who followed a Miaphysite Christology and whose successful missions in Iraq and beyond challenged the dominance of the Church of the East.39 Hōrmīzd’s chief opponents throughout the hagiography are the Jacobite monks of Bezqīn. These men were loyal to the ancient monastery of Mar Mattai that had refused to follow the Dyophysite Christology of most Iraqi Christians at the end of the fifth century and entered communion with Jacobite refugees from the Roman Empire in the sixth century.40 Thus, at the background to Hōrmīzd’s attempts to win over the Arab governor, lies the fear that the Jacobites might do so first and establish their own claim to represent the indigenous, orthodox version of Christianity, approved by God and meriting official support. The Life opens by describing Hōrmīzd’s youthful asceticism and miracles, setting him up as a successor to another local holy man, Rabban bar ʿIdtā. The first of his conflicts with the Jacobites occurs when these charismatic credentials are already well established. In this first encounter he is said to be “a terror to the heretical village of Arshām and an affliction to the wasted tavern of Bezqīn.” In the aftermath of this conflict, which is left
6 42
264
P hilip W ood
without detail or explanation in the hagiography, the governor of Mosul drives the heretics out of Arshām and replaces them with a new Christian community, this time members of the Church of the East, from the village of Ḥazar.41 One suspects that, like Thomas of Marga’s treatment of the fall of the shahrēgān, the hagiographer has appropriated a political calamity that befell the Jacobite community, in which the inhabitants of Arshām were deported, and re-interpreted it as the divinely approved result of the struggle between the Jacobites of Arshām and local members of the Church of the East. The hagiographer follows the deportation from Arshām with a long scene describing Hōrmīzd’s opposition to the monks of Bezqīn and his attempts to find political redress for their misdemeanors. First the Bezqīnites seize and beat Hōrmīzd in his cell. Next they take a prostitute, who has been impregnated by one of the monks of Bezqīn, and kill her and her child in front of the cave of Hōrmīzd. After the beating, a local notable, Gabriel of al-Qōsh, the wealthy descendant of an old Persian family, offers to help Hōrmīzd but is restrained by the saint. It is only after the attempt to frame Hōrmīzd for the murder of the prostitute that Gabriel goes to the governor, who leads the local people in beating up the Bezqīnites after Hōrmīzd has resurrected the prostitute to give her testimony.42 However, the local governor is unable to get any further intervention from his superior in Mosul. This man, ʿUqba, is the subject of the next of Hōrmīzd’s miracles when his son falls sick and dies while being brought to the monastery. Hōrmīzd comforts ʿUqba and raises his son from the dead, after which both father and son ask to be baptized by the saint.43 At this moment, “as the Arabs and men of al-Qōsh rejoiced,” the Bezqīnites arrived and appeared to commiserate with the governor. Though the governor initially rejects them, as “far away from the truth of Rabban Hōrmīzd,” they attempt to bend the situation to their favor by offering to perform the baptism, since Hōrmīzd does not have the necessary equipment, stating that “baptism is the same from either one of us.” This issue of baptism is the crux of the definition of the Jacobites as beyond the pale: traditionally, only very extreme heretics, such as the followers of Paul of Samosata, were viewed to be incapable of providing the sacrament.44 Hōrmīzd voices his opinion strongly: “our baptism differs from yours as light differs from darkness.” To demonstrate this, the saint calls for two children, one “orthodox” and the other Jacobite, and places them in a font that he has blessed. The water parts around the orthodox child, since he has already been baptized, but it embraces the Jacobite,
6 25
Christians in Umayyad Iraq
265
because his earlier baptism had been invalid. This act demonstrates the truth of Hōrmīzd’s claim and convincing ʿUqba, his son, and his companions to receive baptism from Hōrmīzd.45 We can pick out two major themes from this short narrative, both of which find analogues in the rest of the text. The first is the idea of “a hierarchy of resort” in the search for aristocratic patronage: Hōrmīzd starts off by appealing to Gabriel of al-Qōsh, the same kind of local Iranian magnate whom we have encountered in Thomas of Marga. But he quickly upgrades from this to a more effective and prestigious relationship with the unnamed local governor and his superior ʿUqba.46 At the very least, the hagiographer suggests the existence of a new and wider-reaching patronage network, where political favors can be sought to defeat the Jacobites and to fund church construction. In this sense, the paradigm for patronage that the hagiographer employs is a localized forerunner of the patronage exercised by the Church of the East in Abbasid Baghdad to gain advantages over other Christian confessions.47 But there may also be a grain of truth to claims or expectations that men like ʿUqba might have converted to Christianity during the years of light Sufyanid governance in Iraq. In an era when Islam’s boundaries with other monotheisms may have remained blurred, and when local Iranian aristocrats were public patrons of Christian institutions, it is not implausible that certain Arab elites might have trumped the local ties of men like Gabriel of al-Qōsh by converting themselves.48 Thus, in a later scene in the hagiography we see ʿUqba stepping in to extend the funding for a monastery created by the local aristocrat Khūdhāhwai of Beth Qōpā and his villagers: patronage of Christian institutions allowed an outsider like ʿUqba to extend his rural influence and prestige in the same kind as the local dihqāns, and in this sense the picture given by the Life may be a plausible report, as well as an advertisement for the right kind of elite behavior.49 At a time when Arab elites were adopting the political symbols and personal retinues of the dihqāns, some, this saint’s life implies, might have also flirted briefly with their religion.50 The second theme is that Hōrmīzd’s miracles and political relationships take place against a contest for religious legitimacy. Later in the Life, the saint proves that the Jacobites are reading books of sorcery and worshipping the statue of a demon in their monastery,51 but it is clear from the accusations surrounding the death of the prostitute that both sides accused the other of presenting false piety and sorcery as holiness and divine intervention.52 Moreover, it is clear that the arrival of a new stratum
62
266
P hilip W ood
of Arab elites from outside meant that confessional loyalties could not be assumed. For instance, both ʿUqba, and his successor ʿAlī, changed their patronage to the Jacobites after being drugged with “magical cakes.”53 In a genre where confessional allegiance is explained by miraculous intervention, and where confessional boundaries are articulated very starkly indeed, only sorcery can be used to explain a change in political behavior. Here again, we can read against the grain of the text: if Arab elites did indeed wish to embed themselves into rural political and religious power dynamics, then their purposes might well be served even better by patronizing multiple religious communities. Where authors from the Church of the East have a vested interest in presenting the Jacobites as deceitful and violent, as bad as pagans, Jews, and Marcionites,54 this assertion of communal boundaries was a reaction to new opportunities for expansion in the “free market” of post-Sasanian Iraq. In addition, the extreme portrayal of heretics in the text is also a recognition of the opportunities for patronage, both from established elites and from the new conquerors, in an environment where institutions (churches, monasteries, and schools) had not yet had the opportunity to cement and defend the boundaries between the different Christian confessions.55
Conclusions: The Monasteries and the Catholicos An important feature of the hagiographies that we have used so far is that they are all produced in a monastic environment, rather than in the court of the catholicos. It is worth observing that this shift in provenance is one of the most novel features of the Umayyad period. All historical writing in the Church of the East before the early sixth century was centered on the person of the catholicos, and his relations with other bishops, with the shah, and sometimes with the Roman emperor.56 Hagiography too was closely focused on Ctesiphon and the south of Iraq: though there are occasional examples from Adiabene and the north, the vast majority of the corpus of Syriac hagiographies collected by the Chaldean scholar Paul Bedjan have a southern provenance, and several of the longest deal with catholicoi or with martyrs from the region of Ctesiphon.57 Monastic history seems to have set itself up in opposition to this older, centralized history. Perhaps the most strident of these new historians was Sergius of Beth Rāstāq, whose hagiographic collection of the monks of Beth Garmai, composed in the reign of Īshōʿyahb III (r. 649–659), was
6 2 7
Christians in Umayyad Iraq
267
included in Thomas of Marga’s collection. Thomas notes that Sergius called his work “Destroyer of the Mighty,” since he did not write about the great men of the Church “but those who were victorious in the houses of their fathers or the churches of their own villages, who were men of simple spirit.”58 It seems to have been a deliberate reaction to an older clerical tradition: now the central relationship of shah and catholicos had been removed,59 and monasteries could assert their own financial and political independence through their relationships with local lay elites. If the Sufyanid period saw catholicoi who were reliant on their personal connections as aristocrats to assert their influence, then the Marwanid period spelled the end of the road both for independent aristocratic activity and even heavier pressure on the prestige and freedom of action of the catholicoi. Gīwargīs I (r. 661–681) ended his reign in prison after refusing to assist in raising large sums of money. His career provides an indicator of how governors of Iraq viewed the catholicosate, as a means of extracting wealth from the land in the short term, but not an institution worth cultivating in itself. Similarly, the next governor, Bishr ibn Marwān, deposed Gīwargīs’ successor Ḥnānīshōʿ I (r. 686–698) following an insult against Islam and after receiving a bribe from a rival candidate.60 There followed a sixteen-year interregnum after al-Ḥajjāj suppressed the catholicosate.61 When the office was restored after his death, incumbents continued to have difficulties in asserting their authority over other parts of Iraq, or even Ctesiphon. Ṣlībā-zkhā (r. 714–728), the first of this new line of incumbents, failed to reclaim the illuminated service book of Īshōʿyahb III in an embarrassing stand-off with the monks of Beth ʿĀbe, and Abā II (r. 741–751) proved incapable of preserving the revenues of the patriarchal school at Ctesiphon from the priests who ran it.62 In sum, the Marwanid period, with rule centered on Damascus, encouraged Arab governors to maximize profits from the Christians to fund an expanding state apparatus, or for personal gain, but where Christian leaders within Iraq had no access to the highest-ranking political leaders. As such, catholicoi were very vulnerable to the accusations of rivals and eventually lost the legitimacy that had allowed their predecessors to project their authority: the young monks who opposed Ṣlībā-zkhā, it must be remembered, might have never known a catholicos, let alone one who tried to exercise his theoretical powers. But it is notable that the monasteries of the north, and the missions to the east that they sponsored, seem to have had a very different attitude to worldly
6 82
268
P hilip W ood
authority. Where al-Ḥajjāj was seen by some as an oppressive persecutor,63 just as he was by many Muslim historians, he was celebrated in the Life of John of Daylam for granting the saint a remittance from taxes for the monasteries of the east.64 Not only do monasteries seem to have shown a greater resilience to the pressures of the conquests and the dislocation of established forms of patronage, they also seem to have been more successful in securing new forms of patronage, from both old and new elite groups. It is only in the Abbasid period that the “norm” of a centralized catholicosate was restored. The creation of a new capital in Baghdad and the transfer of the seat of the catholicos to the same city as the caliph maximized the potential for royal patronage in the reign of Timothy I (r. 780– 823).65 Here, urban Christians, many of them laymen, could find great rewards for their skills as translators, doctors, pharmacists, and administrators, and outlets for their political ambitions as éminences grises for a series of catholicoi.66 The famous seventh-century missions of John in Daylam and Fars and of “Alopen” in China continued to find successors in the great missions of Timothy and his creation of new oriental dioceses, as far afield as Turkestan, China, and Tibet.67 But the critical difference here is that Timothy was personally associated with the missions, in monastic and ecclesiastical histories, as well as in his own correspondence, whereas the earlier missions are known from hagiographies and inscriptions and were not publicly identified as initiatives driven by the catholicos. If the expansion of the Church of the East continued into the Abbasid period, the resurrection of the catholicosate at Baghdad would spell a future for the church where monasticism existed once more in symbiosis with the catholicos.
Notes 1. Similar arguments have been made for Egyptian Christianity and Roman identity: see Papaconstantinou 2009. Please note that this paper was submitted for publication in 2012 and has not been substantially altered since then. 2. De Vries (1965) observes the growing importance of the office of catholicos and Ctesiphon’s patriarchal claims. 3. See the signatory lists in Chabot 1902, an early ninth-century collection of the earlier councils. 4. Gyselen 2006, esp. 20, for Pahlavi/Syriac and Pahlavi/Greek bilingualism; 27, for Pahlavi Christian names; 28, for adaptation of Zoroastrian protective formulae. Several seals bear crucifer figures (e.g., Gyselen 2006, nos. 23, 25). For a recent assessment of the place of Christians in the Sasanian world, see Payne 2015.
9 6 2
Christians in Umayyad Iraq
269
5. Longinov and Nikitin 1993, 271–272, for coins of Yazdegard I. 6. De Goeje 1901, 1:991. See now Wood 2016. 7. Scher 1919, XLII, 440, for Īshōʿyahb I’s activities while bishop of Arzun. 8. Sako 1986. 9. Chabot 1902, 563. 10. Scher 1919, LXVII, 487. 11. Yazdīn bar Shamṭā, Khusrau’s governor for Beth Garmai and Adiabene, was particularly important in this regard. See Wallis Budge 1893, I, xxiii, 47/80–82 (for Yazdīn’s family as monastic donors); Scher 1919, LXXXI (524–525) (for his prominence at court, alongside other secular figures). 12. Chabot 1902, 79, 100, 103. See, in general, Fiey 1969. 13. Villagomez 1998, 100–107, 140–143, 173; Wood 2013, 153. 14. Wallis Budge 1893, 63/113–114. 15. On Shīrōë, see Scher 1919, XCII, 551. On Shahrbarāz, see Scher 1919, XCIII, 554–556; Guidi 1903, 29–30. See also Mango 1985. 16. For the situation under the Sasanians, see Choksy 1988; for apostasy from Zoroastrianism, see Morony 1984, 301–302 and Hoyland 2015, 161, 208. For a military narrative of the Sasanian collapse see now Hoyland 2015, 115–122. 17. Bailey 1930, 56 (verse 13) and 59. 18. Hitti 1916, 272–274; De Goeje 1901, 2:458. See further examples and discussion in Zakeri 1995, 101–112; Morony 1984, 206. 19. Gyselen 2000, 140 (type 32), for a figure bearing a crown with a cross; 177 (type 90), for a coin with a cross-bearing figure on one side and a cross on the other, with the possible Pahlavi legend “episkopos”; 180 (type 97), for a local version of a Roman coin of Heraclius. 20. Brock 1982, 123–190, sections 19, 25, and 39. 21. Scher 1919, LXXVI, 516, for the Kurds near Izla; XCVII, 587– 588, for the Khurasānīs of ʿAin al-Namīr; XCIX, 597, for Rabban Hōrmīzd baptizing the children of Jacobites; CIX, 631, for two pagan villages near the highlands of Shāhrāzūr in northern Iraq. 22. Baumer 2006, 168–174 (on Central Asia) and 179–186 (on China). 23. Encyclopaedia of Islam, 2nd ed., s.v. “Zindīḳ” (by F. De Blois). 24. Wallis Budge 1893, II, xxxii, 110/242. On Thomas in general see Teule 2009. 25. Wallis Budge 1893, III, ii, 143–144/296–297. 26. Robinson 2000, 108. 27. Scher 1906, 186–188. 28. Much of this paragraph is shared with Wood 2013, 167. 29. Wallis Budge 1893, II, xliii, 136–137/282–283. On Aḥā’s election, see Wallis Budge 1893, II, xxxvi, 120/257. 30. Macuch 2004. 31. Wallis Budge 1893, II, xxxii, 110–111/242–243.
0 72
270
P hilip W ood
32. Robinson 2000, 108; Morony 1984, 205. Al- Masʿūdī’s Golden Meadows (de Meynard and de Courteille 1877, 1:327) emphasizes the preservation of Sasanian genealogies and distinctive clothing by the shahrēgān. However, Robinson is wrong to imagine a religious distinction between shahrēgān and dihqāns. In several of the cases described by Thomas of Marga both groups of aristocrats seem to be Christian. 33. For Egypt, see the Chronicle of John of Nikiu (Charles 1916, CXXI.5), which notes the excessive taxation of the aristocrat Menas, theoretically acting on behalf of the Muslims. See also Sijpesteijn 2009 on the retention of Christian pagarchs in Egypt until ca. 700. 34. Wallis Budge 1893, III, iii, 151/310. 35. Wallis Budge 1893, II, xxxvi, 120/256–257. 36. Wallis Budge 1893, III, iii, 152/311–312. 37. Wallis Budge 1893, III, viii, 161/324–325. 38. Wallis Budge 1893, III, viii, 163/329. 39. Fiey 1973, 78; Fiey 1974; Fiey 1992. 40. Fiey 1970a, 123. For the fifth-century background, see Gero 1981. 41. Wallis Budge 1902, 53–54/80. 42. Wallis Budge 1902, 60–64/90–95. Note now the analysis of Penn 2015, 136–138. 43. Wallis Budge 1902, 65–68/97–100. 44. Wood 2010, 165. The denial of Catholic baptism was a key complaint against the African Donatists in the fifth century; Brown 1967, 216–219. 45. Wallis Budge 1902, 68–72/101–105. 46. Robinson (2000, 73) identifies him as either ʿUtba ibn Farqad or ʿUqba ibn al-Walīd. 47. See, in general, Fiey 1980; Allard 1962. 48. On the blurred edges of early Islam, see the arguments of Donner 2003; Donner 2010; Penn 2015, 156–161. For Arab settlement in Iraq, see Donner 1981, chap. 5. For the alleged mass conversion of an Arab Muslim army to Christianity in Egypt in the same period, see Palmer 1983, 114. 49. Wallis Budge 1902, 79/117. 50. On private militias, see Mason 1967; Athamina 1998. Miles (1959, 94) notes the Persianized self-representation of the Muhallabids. 51. Wallis Budge 1902, 81/120, 83–84/124. 52. For the archaeological evidence for the use of “magic” by Christians, see the Aramaic incantation bowls discovered at Uruk and elsewhere: Walker 2012, 1013. 53. Wallis Budge 1902, 74–76/109–113, 98/146–147. 54. On Marcionites, see Fiey 1970b. 55. The Jacobite Life of Marutha of Takrit (Nau 1909, 65) complains at “Nestorian” schools and choirs in Beth Nuhadra winning over Miaphysite “widows and virgins.” Also note Brock 1982, parags. 26–31 (138), for ʿAbd al-Malik’s promise to John to spare monks, priests, and teachers (malpāne) from taxation.
271
Christians in Umayyad Iraq
271
56. See, for instance, the testimony of the medieval chronicles of ʿAmr ibn Mattā (Gismondi 1899) and Bar Hebraeus (Abeloos and Lamy 1877, vol. 3), which probably preserve much of the form of this earlier material. 57. On the martyrs near Ctesiphon in the fifth century, see Devos 1965. 58. Wallis Budge 1893, I, xxxiii, 61/109–110. 59. The effect of this disappearance is seen most clearly in the brevity of the accounts in the Ctesiphon-centered patriarchal histories of Mārī ibn Sulaymān and ʿAmr ibn Mattā, which become suddenly richer once more in the Abbasid period. 60. Abeloos and Lamy 1877, 3:136–140; Mingana 1908, 184. 61. Called “a persecution of the church” in the prophecy of one holy man; Scher 1919, XCVIII (593). 62. Wallis Budge 1893, II, xxvii; Abeloos and Lamy 1877, 3:152–154. 63. Scher 1919, LXXXI (525). 64. Brock 1982, parag. 26. 65. Fiey 1980, 12, 36. 66. Putman 1986, 93–108. The activities of these powerful laymen are a recurring theme in Mari and ʿAmr. For the famous family of Bokhtīshōʿ, see Baumer 2004, 155, 158. See also Encyclopaedia Iranica, s.v. “Boḵtišu” (by L. Richter-Bernberg). 67. Fiey (1993) summarizes all of his administrative reforms. See now the important study by Berti (2009). On the Chinese inscriptions and the activity of missionaries in the Far East, see Walker 2012, 1022–1023.
Bibliography Abeloos, J., and T. Lamy, ed. and trans. 1872–1877. Gregorii Barhebraei chronicon ecclesiasticum. 3 vols. Paris: C. Peeters. Allard, M. 1962. “Les chrétiens à Baghdad.” Arabica 9: 375–388. Athamina, K. 1998. “The Non-Arab Regiments and Private Militias during the Umayyad Age.” Arabica 45: 347–378. Bailey, H. 1930. “To the Žāmasp-Nāmak.” Bulletin of the School of Oriental Studies 6: 55–85. Baumer, C. 2006. The Church of the East: An Illustrated History of Assyrian Christianity. London: I. B. Tauris. Berti, V. 2009. Vita e studi di Timotheo I, Patriarca cristiano di Baghdad. Paris: Association pour l’avancement des études iraniennes. Brock, S., trans. 1981–1982. “Life of John of Daylam.” Parole de l’Orient 10: 123–190. Brown, P. 1967. Augustine of Hippo: A Biography. London: Faber. Chabot, J.-B., ed. and trans. 1902. Synodicon Orientale, ou, Recueil de synodes Nestoriens. Paris: Imprimerie Nationale. Charles, R. H., trans. 1916. The Chronicle of John, Bishop of Nikiu. Oxford: Oxford University Press.
27
272
P hilip W ood
Choksy, J. 1988. “Sacral Kingship in Sasanian Iran.” Bulletin of the Asia Institute 2: 35–53. De Goeje, M. ed. 1879–1901. al-Ṭabarī: Taʾrīkh al-rusul wa’l-mulūk. 15 vols. Leiden: Brill. (A translation of 39 volumes, prepared by various scholars, was published by the State University of New York in 1986–2007.) De Vries, W. 1965. “The College of Patriarchs.” Concilium 8: 35–43. Devos, P. 1965. “Abgar: Un hagiographe perse méconnu.” Analecta Bollandiana 83: 303–328. Donner, F. 1981. The Early Islamic Conquests. Princeton, NJ: Princeton University Press. ———. 2002–2003. “From Believers to Muslims.” Al-Abhath 50–51: 9–53. ———. 2010. Muhammad and the Believers: At the Origins of Islam. Cambridge, MA: Belknap Press. Fiey, J. 1969. “Les laïcs dans l’histoire de l’église syrienne orientale.” Proche-Orient Chrétien 14: 169–183. ———. 1970a. Jalons pour un histoire de l’Église en Iraq. Beirut: Secrétariat du Corpus. ———. 1970b. “Les Marcionites dans les textes historiques de l'Église de Perse.” Le Muséon 83: 183–188. — — — . 1973. “Chrétientés syriaques du Horāsān et Ségēstan.” Le Muséon 86: 75–104. ———. 1974. “Les diocèses de maphrianat syrien, 629–1860.” Parole de l’Orient 5: 133–165. ———. 1980. Chrétiens syriaques sous les Abbasides, surtout à Baghdad (750–1250). Leuven: Secrétariat du Corpus. — — — . 1992. “Syriaques occidentaux du pays des Perses.” Parole de l’Orient 17: 113–127. ———. 1993. Pour un Oriens Christianus Novus: Répertoire des diocèses syriaques orientaux et occidentaux. Beirut: Steiner. Gero, S. 1981. Barsauma of Nisibis and Persian Christianity in the 5th Century. Leuven: E. Peeters. Gismondi, H. 1899. Maris, Amri, et Salibae: De Patriarchis Nestorianorum Commentaria I: Amri et Salibae textus arabicus et versio Latina. Rome: De Luigi. Guidi, I., ed. and trans. 1903. “Chronicum Anonymum.” In Chronica Minora I, CSCO, 1–2. Paris: Peeters. Gyselen, R. 2000. Arab-Sasanian Copper Coinage. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. ———. 2006. “Les témoinages sigillographiques sur la presence chrétienne dans l’empire sassanide.” In Chrétiens en terre de l’Iran I: Implantation et Acculturation, edited by R. Gyselen, 17–78. Paris: Association pour l'avancement des études iraniennes. Hitti, P., trans. 1916. The Origins of the Islamic State: Being a Translation from the Arabic Accompanied with Annotations, Geographic and Historic Notes of the Kitȃb
273
Christians in Umayyad Iraq
273
futȗḥ al-buldȃn of al- Imȃm abu l'Abbȃs Aḥmad ibn-Jȃbir al-Balȃdhuri. Vol. 1. New York: Longmans. Hoyland, R. 2015. In God’s Path: The Arab Conquets and the Creation of an Islamic Empire. Oxford: Oxford University Press. Longinov, S., and R. Nikitin. 1993. “Sasanian Coins from Merv.” Mesopotamia 28: 225–318. Macuch, M. 2004. “Pious Foundations in Byzantine and Sasanian Law.” In Convegno internazionale la Persia e Bisanzio: Roma, 14–18 ottobre 2002, edited by A. Carile, L. Ruggini, and G. Gnoli, 181–195. Rome: Accademia Nazionale dei Lincei. Mango, C. 1985. “Deux études sur Byzance et la Perse Sassanide.” Travaux et Mémoires 9: 93–118. Mason, G. 1967. “Merv and the Muhallabids: A Historical Re-assessment.” Arabica 14: 191–207. Miles, G. 1959. “The History of Kirman.” In The World of Islam: Studies in Honour of Philip Hitti, edited by J. Kritzeck and R. B. Winder, 85–98. New York: Macmillan. Mingana, A. 1908. Sources syriaques I. Leipzig: O. Harrassowitz. Morony, M. 1984. Iraq after the Muslim Conquest. Princeton, NJ: Princeton University Press. Nau, F., ed. and trans. 1909. Life of Marutha of Takrit, Patrologia Orientalis 3. Paris: Firmin-Didot. Palmer, A., S. P. Brock, and R. Hoyland. 1983. The Seventh Century in the West Syrian Chronicles. Liverpool: Liverpool University Press. Papaconstantinou, A. 2009. “‘What Remains Behind’: Hellenism and Romanitas in Christian Egypt after the Arab conquest.” In From Hellenism to Islam: Cultural and Linguistic Change in the Roman Near East, edited by H. Cotton, R. Hoyland, J. Price, and D. Wasserstein, 447–466. Cambridge: Cambridge University Press. Payne, R. 2015. A State of Mixture. Christians, Zoroastrians and Iranian Political Culture in Late Antiquity. Berkeley, CA: University of California Press. Penn, M. 2015. Envisioning Islam. Syriac Christians and the Early Muslim World. Philadelphia, PA: University of Pennsylvania Press. Putman, H. 1986. L'Église et l'islam sous Timothée I (780–823): Étude sur l'église nestorienne au temps des premiers ʿAbbāsides avec nouvelle édition et traduction du Dialogue entre Timothée et al-Mahdi. Beirut: Dar el-Machreq. Robinson, C. 2000. Empire and Elites after the Muslim Conquest: The Transformation of Northern Mesopotamia. Cambridge: Cambridge University Press. Sako, L. 1986. Le rôle de la hiérarchie syriaque orientale dans les rapports diplomatiques entre la Perse et Byzance aux Ve–VIIe siècle. Paris. Scher, A., ed. and trans. 1906. “History of the Convent of Sabrisho of Beth Qoqa.” Revue de l’Orient Chrétien 11: 182–197. Scher, A., ed., and F. Graffin, trans. 1919. Patrologia Orientalis 13. Paris: Firmin-Didot. Sijpesteijn, P. 2009. “Landholding Patterns in Early Islamic Egypt.” Journal of Agrarian Change 9: 120–133.
4 72
274
P hilip W ood
Teule, H. 2009 “Thomas of Marga.” In Christian-Muslim Relations: A Bibliographical History. Vol. I, 600–900, edited by D. Thomas, 688–690. Leiden: Brill. Villagomez, C. 1998. “The Fields, Flocks and Finances of Monks: Economic Life at Nestorian Monasteries, 500– 850.” Unpublished PhD diss., University of California, Los Angeles. Walker, J. 2012. “From Nisibis to Xi’an: The Church of the East in Late Antique Eurasia.” In The Oxford Handbook of Late Antiquity, edited by S. F. Johnson, 994– 1052. Oxford: Oxford University Press. Wallis Budge, E., ed. and trans. 1893. Book of the Governors: The Historia Monastica of Thomas, Bishop of Marga, a.d. 840. 2 vols. London: Kegan Paul. ———. 1902. The Histories of Rabban Hȏrmȋzd the Persian and Rabban Bar-'idtȃ. 2 vols. London: Luzac. Wood, P. 2010. “We Have No King but Christ”: Christian Political Thought in Greater Syria on the Eve of the Arab Conquest (c. 400–585). Oxford: Oxford University Press. ———. 2013. The Chronicle of Seert: Christian Historical Imagination in Late Antique Iraq. Oxford: Oxford University Press. ———. 2016. “The Christian Reception of the Xwadāy-Nāmag: Hormizd IV, Khusrau II and their successors.” Journal of the Royal Asiatic Society series 3(26): 407–422. Zakeri, M. 1995. Sāsānid Soldiers in Early Muslim Society: The Origins of ʿAyyārān and Futuwwa. Wiesbaden: Otto Harrassowitz Verlag.
275
9
The Future of the Past Historical Writing in Early Islamic Syria and Umayyad Memory Antoine Borrut
book dedicated to Umayyad Legacies, Paul Cobb and I asked a simple question: how do we know what we think we know about the first dynasty of Islam?1 The problem is remarkably acute when one considers Syria, the former heartland of Umayyad power. As Patricia Crone once put it, “the Syria to which Heraclius bade his moving farewell seems to have vanished, not just from Byzantine rule, but from the face of the earth.”2 Indeed, early Islamic Syria is extremely poorly documented in Muslim narrative sources in comparison to Iraq or Iran, for instance. In short, early Islamic Syria largely fell into oblivion. This has mostly to do with the fact that Syrian history—as we now have it—was chiefly (re-)composed in Iraq, by scholars active under the Abbasids, the dynasty that toppled the Umayyads, the former masters of Damascus. Yet, for all the oblivion surrounding the first dynasty of Islam, it seems that early Abbasid Syria was even more thoroughly erased or forgotten: if the former seat of Umayyad power was indispensable to any narrative of Islamic history, the focus suddenly shifted to Iraq and Iran in the aftermath of the so-called Abbasid Revolution, in 132 ah/750 ad. This situation reminds us of the formidable methodological challenges offered by the lack of contemporary narrative sources for the beginnings of Islam that have been stressed by various scholars and that have generated an important debate in the field early IN A RECENT
6 72
276
A ntoine B orrut
on,3 despite significant progress through archaeology, epigraphy, numismatics, or papyrology within the past few decades.4 This situation invites us to approach early Islamic Syria from a history of memory perspective because what was to be remembered or forgotten was largely determined by editorial choices and strategies implemented by generations of scholars. Indeed, one could argue along with German Egyptologist Jan Assmann that “the past is not a natural growth but a cultural creation” elaborated by a scholarly elite of specialists of cultural memory.5 Memory is therefore a social practice and a cultural production. Layers of rewriting of the past, of remembrance and oblivion— memory’s two attributes—are thus a testimony to power (be it caliphal political power or elite cultural power) and “reflect a particular exploitation of the potential of written culture”6 in the early Islamic world. The Abbasid world arguably saw an unprecedented flowering of literacy in a context where developing new links toward the past was certainly a convenient way to give meaning to a profoundly transformed present and to conjugate power and legitimacy into the present and future tenses. In the process, new meanings were granted to the past. They deserve to be studied as such, in a perspective of what German scholars have defined as Sinngeschichte, or history of meanings.7 Such an approach can help us to unveil “alternative pasts,”8 alternative meanings granted to the past that were largely buried under layers of writings and rewritings and historiographical reconstructions; but in a sense some of these alternative pasts “refused to stay buried”9 as some echoes or whispers can still be heard in the sources. It is therefore not a history of silence that I am advocating—a truly dangerous exercise indeed—but rather, following Patrick Geary, an attempt to listen to these murmurs that still survive in the texts, revealing the tension between memory and oblivion in the medieval sources themselves.10 Let us first look more precisely at the formidable challenges posed by narrative sources for early Islam in order to clarify the situation.
Historical Writing in Early Islamic Syria Because very little historical narrative exists that is contemporary with the first two centuries of Islamic history, our knowledge of early Islam is highly dependent upon Islamic histories produced much later, in the late third/ ninth and during the fourth/tenth century, such as the famous History of the Prophets and the Kings written by al-Ṭabarī (d. 310/923), arguably the
2 7
The Future of the Past
277
most important and best-known of the early Islamic “historians.”11 This situation raises a crucial question: does the absence of preserved narrative Muslim sources from the first two centuries of Islam imply that nothing was actually written or that these early sources vanished, and, if so, why? In other words, such an inquiry invites us to ask when history was first written in the Islamic world—even if such a question goes far beyond the purpose of this paper—especially under the Umayyads, to unveil pre- Abbasid historiographical layers. Despite some pessimistic responses to these challenges, it nevertheless now seems clear that the creation of an agreed version of the early Islamic past, in other words a vulgate, composed around the end of the third/ninth century and the beginning of the fourth/tenth, was by no means the first attempt by Muslims to write the story of their origins. But it was this Abbasid-era version that succeeded when previous efforts vanished, or more aptly were reshaped, in subsequent layers of rewriting, and became enshrined as the “official” version of Islamic sacred history that we can find in mainstream chronographies. Nevertheless previous attempts to impose different historical orthodoxies did affect the making of the Abbasid-era version, as history had to be rewritten with whatever materials were available, and this was itself partly determined by editorial decisions made by earlier generations of Muslim historians whose own works do not survive. As one scholar of the medieval West pertinently formulated it, “The starting point [of historiography] is not silence (by now irretrievable), but what has been said already.”12 This remark has far-reaching implications, in the sense that it invites us to look for a now-lost early Islamic historiography and its impact on subsequent historiographical developments. There are strong arguments, in my view, to prove that history was written early on, at least under the Umayyads in the late first/seventh and early second/eighth century, if not earlier.13 For instance, I have tried to show elsewhere that a robust culture of historical writing existed in second-/eighth-century Syria, largely sponsored by the Umayyad caliphate, and that this practice continued in Syria under the first Abbasids despite the transfer of the capital to Baghdad, as evidenced by a sizable number of scholars actively engaged in historical writing and/or the transmission of historical material in early Islamic Syria.14 This development of historiography in Umayyad Syria has to be understood in the larger framework of the use of literacy under the first dynasty of Islam that certainly needs to be reevaluated in the light of countless inscriptions, papyri, or coins. In fact, it may well be the case,
8 72
278
A ntoine B orrut
in broader terms, that the entire literary culture that existed under the Umayyads has been underestimated, if we follow George Saliba’s recent suggestion that the dīwān was a large body of texts, including perhaps scientific texts, rather than a simple payroll.15 When one thinks about historical writing under the Umayyads, a few names come immediately to mind, such as the likely legendary character ʿUbayd (or ʿAbīd) ibn Sharya al-Jurhumī, who allegedly composed a history at Muʿāwiya’s request.16 Even if this specific episode could be regarded as doubtful,17 the sources are unanimous to stress Muʿāwiya’s interest for history.18 This anecdote also reveals that at an uncertain point in time, Muʿāwiya was seen as one of the founding fathers of historical writing in Islam, an initiative that would thus have taken place in Syria. A much more historically grounded scholar of early Islam is ʿUrwa ibn al-Zubayr (d. ca. 94/712–713) who has attracted a good deal of scholarly attention, especially given his pivotal role as a transmitter of historical material about the Prophet Muḥammad.19 But the most obvious scholar that immediately stands out is, of course, the famous al-Zuhrī (d. 124/742), active from the time of ʿAbd al-Malik (r. 65–86/685–705) to the caliphate of Hishām (r. 105–125/724–743). He is in fact credited with so much material that he has sometimes been identified as the “common link” of many traditions, to borrow from the terminology of isnād studies.20 Al-Zuhrī’s relationships with Umayyad rulers have generated some debate in modern scholarship, but despite attempts to demonstrate that he was a truly independent scholar, it seems on the contrary that he was working in close collaboration with the caliphs.21 His efforts are best seen during Hishām’s caliphate and al-Zuhrī clearly represents a major “historiographical filter,” possibly reflecting Umayyad efforts to control the tradition and to define a Marwanid orthodoxy.22 Beyond such well- known names, scholars such as Fred Donner, Amikam Elad, and lately Nancy Khalek and myself, have tried to document historical writing during the first two Islamic centuries. And the results of all these studies are quite clear: history was being written as early as the first century of the hijra, under the Umayyads, by a sizable number of scholars. More than a dozen of them (and even more if we include early Abbasid Syria) can be clearly identified, such as Khālid ibn Maʿdān al-Kalāʿī al-Ḥimṣī (d. 104/722–723),23 ʿUbāda ibn Nusayy al-Kindī (d. 118/736–737), or Sulaymān ibn Mūsā (d. between 115/733 and 119/737). Although such scholars were not strictly speaking “historians,” and history
9 72
The Future of the Past
279
was likely not their main field of activity, they transmitted historical material related to the conquest of Syria or Umayyad military campaigns such as Maslama ibn ʿAbd al-Malik’s siege of Constantinople.24 Interestingly enough, these three scholars, among others, were undoubtedly connected with Umayyad power and served the first dynasty of Islam in some official capacities. Yet not all “historians” were clearly connected with Umayyad rulers: some, particularly in the late Marwanid period, seem to have been working independently, especially when they appear to have shown Qadarite tendencies, for example, al-Waḍīn ibn ʿAṭāʾ al-Dimashqī (d. 149/ 767).25 Other scholars continued to flourish in early Abbasid Syria in the decades following the downfall of the Umayyads, such as Hishām ibn al- Ghāz (d. between 153/770 and 159/776), Saʿīd ibn ʿAbd al-ʿAzīz al-Tanūkhī (d. 167/784), and al-Walīd ibn Muslim al-Umawī al-Dimashqī (d. 194/ 810).26 And so we do see a sizable group of scholars emerging, actively engaged in historical writing and the circulation of historical information, and constituting scholarly networks, either families of scholars, perhaps best exemplified by some members of the Banū Ghassān,27 or master- student networks. Such an impetus to write history could be explained by simple needs, such as the necessity to preserve a basic chronology of rulers. The Asnān al-khulafāʾ attributed to al-Zuhrī were maybe just that: a chronology of the caliphs. In fact, it is quite likely that such lists survive in Syriac sources that do preserve several of them, and they seem to reflect pro- Umayyad views. It is striking for instance to note that ʿAlī’s caliphate is systematically ignored in such Syriac lists and that the period traditionally associated with his caliphate is simply referred to either as a fetnā, a word obviously borrowed from the Arabic fitna, or as a period with no ruler.28 This enthusiasm for lists was actually not restricted to caliphal succession, if one is to judge by the abundance of other lists for almost every single year in later sources (governors, judges, military and ḥajj leaders, etc.). It has been argued that such lists “may have functioned as the sequential scaffolding upon which historians of the later eighth and ninth century erected their dynastic chronographies,” thus playing a significant role “in sparking the chronographic tradition.”29 The bottom line is that, as argued by Assmann, such lists acted at the same time as inhibitors limiting the historian’s options and should thus be seen as “a tranquilizer and not an incentive when it comes to the writing of history.”30
0 82
280
A ntoine B orrut
But moreover, we should also keep in mind, with Fred Donner, that “the creation of historical narratives is always, ultimately, an exercise in legitimation.”31 And therefore, there may well be a strong link between historical writing and state building. The lists we just mentioned may, in fact, be seen as part of this process, as an attempt to produce a record, a memory, of the young Umayyad state, its rulers, and its bureaucracy. We should also consider that Umayyad-era scholars had to give meaning to a deeply transformed present, especially in the aftermath of the dramatic second fitna (ca. 61–72/680–692) so instrumental in the shaping of identities (and arguably of a discrete Muslim identity),32 not to mention the need to make sense of the rise of Islam and of the expansion of the first Muslim empire. All of these elements suggest that a vigorous culture of historical writing did actually exist in Umayyad Syria, especially in the second/eighth century.33 How this production should be ranked compared to what was produced in Medina or in Iraq remains debatable, even if examples such as al-Zuhrī, frequently traveling between Syria and the Hijaz, reveal that those centers were much more connected than often is assumed. In other words, this Syrian historiographical production should not be understood in the old framework of Julius Wellhausen’s theory of historical schools, duly rejected by Albrecht Noth among others: it rather reminds us of Donner’s wise comment that historical writing was restricted to a limited number of places in the early Islamic world.34 It is to be noted, however, that early Syrian production appears to have been more focused on the Prophetic period and on Islamic expansion rather than on the Umayyads themselves. Indeed, there seems to only have been a limited interest for “contemporary” history, judging by the titles of lost books.35 Some likely exceptions include a Kitāb sīrat Muʿāwiya wa-Banī Umayya, attributed to ʿAwāna ibn al-Ḥakam (d. 147/ 764) or, slightly later on, a Kitāb akhbār al-Umawīyyīn, authored by ʿAlī ibn Mujāhid (d. 182/798).36 The scholar Maslama ibn Muḥārib (active in the early Abbasid period) also seems to have devoted most of his attention to Umayyad history, with a clear Sufyānid bias easily explicable, given that he was the great-grandson of a famous governor of Iraq, Ziyād ibn Abīhi (d. 53/673).37 But more importantly, two essential questions need to be addressed: (1) What was the impact of these early historiographical layers on “classical” Abbasid-era sources? and (2) Can we retrieve some of this now-lost historiography?
281
The Future of the Past
281
Early Historiographical Layers and “Classical” Abbasid Historiography I think that early Islamic historiography is best understood as a stratified process of historical writing, made up of layers of writing and rewriting, layers of interpretation and reinterpretation. I have tried to identify the most important moments of such re-composition of the past in a recent book, and I have argued that the abandonment of Samarra and the return of the caliphate to Baghdad, in 279/892, marked the most decisive of these layers.38 A vulgate, an agreed-upon version of the past, was imposed, thereby organizing early Islamic history into a very rigid framework and providing a new meaning of the past with a view to making sense of a profoundly altered present.39 Indeed, by the late third/ninth century, the “evidence of … continuity had become unintelligible.”40 Consider the rapid succession of dramatic changes that occurred during the first 150 years of Abbasid rule: the so- called Abbasid Revolution itself, of course; the fall of the Barmakids; the fourth fitna (193–198/809–813) and the challenge of regicide;41 the miḥna; the move to Samarra (221/836); the rise of the Turks who would not just become the new backbone of the military but would soon make and unmake caliphs; and so on.42 Rewriting the past was therefore more than a choice; it was a need to give meaning to a much-transformed society. Cultural actors played a decisive role in the process, offering responses required by a profound political and, above all, social crisis. Times of crisis are indeed always favorable to developments centered on memory and oblivion. This sensible rewriting would prove to be a long- lasting one, largely accepted by medieval and modern scholars alike. In the process of these various layers of writings and rewritings, “alternative pasts” were forgotten. For instance, if we largely see the Umayyads through an Abbasid lens, what should be said of Zubayrid memory that was largely erased already by the first dynasty of Islam?43 It is worth noting that Chase Robinson recently suggested that Ibn al-Zubayr should probably be regarded as the legitimate caliph in the late 680s and early 690s, thus limiting ʿAbd al-Malik to the role of a mere rebel, before the restoration of Marwanid authority in 72/692.44 Of course, this is not the image transmitted by the sources, which likely echo, in this, the Umayyad vision of history, for history is ultimately rewritten by winners.45 The very nature of today’s available sources and these successive rewritings lock the first centuries of Islam into a dichotomy between
2 8
282
A ntoine B orrut
remembrance and oblivion. Of course such a situation is by no means an Islamic exception and the same is true in many other areas, as brilliantly studied for instance by Patrick Geary for the Medieval West around the turn of the first millennium. Allow me to paraphrase Geary in Islamic terms: what we think we know about [early Islam] is largely determined by what people of the [third/ninth and fourth/tenth centuries] wished themselves and their contemporaries to know about the past… . These people went about remembering their past in relationship to their own present. This process, which involved creating new systems and structures, both social and textual, within which the past was preserved, necessarily changed the very nature of what would be preserved through elimination, elaboration, and reinterpretation. For the most part, this process took place for very local and specific reasons and often followed regional forms [i.e., we mainly have the Iraqi/Abbasid version]. Moreover, … the effects were broad and long-lasting. The filter of the [third/ninth and fourth/ tenth centuries] determined in a fundamental way the access that all future generations would have to alternative pasts.46 Historiography, then, is like archaeology, and we must begin with the most recent layers. It is therefore necessary to study first the use of the past made by the Abbasid chroniclers and the different meanings granted to a given period by various reinterpretations. An interesting example of this process is how the Abbasids confronted and managed the Umayyad past.47 Because the Abbasids violently overthrew the Umayyads in 132/750,48 it is not surprising that they generally depicted the caliphs of Damascus in a rather demeaning light. But the strategies of Abbasid-era historians toward the Umayyad century were in fact much more complex, revealing changing approaches since a full oblivion of the Umayyad century was not an option. Umayyad history was actually reduced by Abbasid chroniclers to a limited number of “sites of memory” (lieux de mémoire), to borrow Pierre Nora’s famous terminology: a limited number of events ranging from ʿUthmān’s assassination to the final defeat on the Upper Zāb; the image of some caliphs or of a limited set of administrative reforms; the dichotomy between Kalb and Qays; the Dome of the Rock or the Mosque of Damascus; and so on.49 This core data became common to almost every source while many other aspects of the same historical reality faded from
2 8 3
The Future of the Past
283
memories. Another strategy developed by the Abbasid chroniclers consisted in enclosing and limiting the memory of the first dynasty of Islam to Syria, the Umayyad heartland of power. Syria was thus presented as a rebellious province associated with the Umayyads while the Abbasids linked their origins to the eastern part of the caliphate. This is somehow surprising as the Abbasid family was settled in Syria for much of the Umayyad period. But in its course to legitimacy, the young Abbasid regime tried to minimize—or even erase—its Syrian background. For similar reasons, they stressed their links with Iraq and the East, Khurasan in particular. The result was, in some ways, the development of an Iraqi- Abbasid memory, to be opposed to a Syrian-Umayyad one. In this reconstruction, the past was geographically enclosed: the Euphrates became a mirror and a dividing line between both dynasties, thus depriving us of a more comprehensive view of Umayyad and Abbasid projects at the scale of the empire. Obviously, some alternative Umayyad and Abbasid pasts were forgotten in the processes of redaction in the third/ninth and fourth/ tenth centuries.50
Retrieving a Now-Lost Early Islamic Historiography Can we retrieve parts of this now-lost historiography? The situation is not hopeless, and several approaches are possible to tackle this vexing issue of the sources. One of the most promising is arguably to study what Lawrence Conrad has termed “intercultural transmission,”51 that is, the circulation of historical information between the various historiographies at work in the early Islamic Middle East. Indeed, unlike their Islamic counterparts, several non-Muslim sources have the advantage of being contemporary with the first centuries of Islam (mostly Christian sources for the Middle East during the period under consideration, most notably Syriac sources, Byzantine, Armenian, Coptic, and so on). While some scholars have tended to reject such sources,52 others (such as the skeptics) have found them to be much more reliable than the Muslim narratives.53 This is due to the fact that the non-Muslim sources were seen as “external” sources, to be opposed to the Islamic “internal” sources. Such a dichotomy is quite misleading and I have argued elsewhere that what we have are in fact Middle Eastern sources—that is, common narratives produced in the different languages of the historiographies at work in the Middle East but based (at least for Islamic history) on a shared core of data, circulated through “intercultural transmission.”54
4 8 2
284
A ntoine B orrut
In other words, Syriac chroniclers, for instance, depended on Muslim and/or Arabic informants for their knowledge of the history of Islam and of the caliphate, when they were not themselves important actors at the Muslim court, as with Theophilus of Edessa (d. 785), who ended his career as court astrologer for the caliph al-Mahdī (r. 158–169/775– 785).55 This being so, and since we have Syriac sources from the seventh and eighth centuries but no Islamic ones, can we use these Syriac sources as repositories that preserve those “lost” layers of early Islamic historiography that were effaced by the “official” Abbasid-era version of Islamic history? I think we can, and I think that, ironically, we can have access to earlier layers of Muslim historical writings through these non- Muslim sources than with the Muslim sources. Studying these layers of rewritings is important in order to establish the sequence by which specific themes or clusters of information appeared, revealing the various political, ideological, or historiographical projects behind the writing down of these early attempts at Islamic history. Thanks to the different rhythms of the various historiographies at work in the Middle East (and the variable rhythms of the processes of transmission, including intercultural transmission), we are thus sometimes able to gain access to fragments of this “buried past [that] refused to stay buried.”56 In other words, non-Muslim sources provide access to different moments of historiographical sedimentation. Indeed, we can trace some elements back to the Umayyad period itself, well before the Abbasids even appeared on the scene. Of course we cannot pretend to have any access to the original (literary) form of this information, but we might have access to the content itself and be thus able to date the circulation of historical information. We are thus facing a multilayered historiography and to understand it properly we have to detect the several stages of writing and rewriting. We then have to identify in the Christian sources which elements in a given layer come from a now-lost Islamic historiography and eventually what subsequent generations chose to do with this material. This last point is especially critical when dealing with Abbasid sources narrating the Umayyad past. If some Muslim sources could definitely be situated at the margins of such a vulgate,57 non-Muslim sources offer a unique opportunity to access some of these lost early layers and alternative pasts, thus revealing early Islamic historiography in the making, and to shed light on various editorial strategies through which specific “memorabilia were deemed memoranda.”58
2 8 5
The Future of the Past
285
One of the most problematic events for the Abbasids to deal with was paradoxically their own rise to power during the so-called Abbasid Revolution in 132/750. Indeed their victory over the Umayyad armies was followed by the massacre of numerous members of the Umayyad family at Nahr Abī Fuṭrus (the Antipatris or Yarkon River, in modern Israel) as well as in other places.59 The bloody Nahr Abī Fuṭrus episode became embarrassing in the aftermath of the revolution as it was clearly not in accordance with the high moral Islamic principles promoted by the Abbasids through their propaganda against the “impious” Umayyads.60 Because reports of the massacre were circulated too early and too widely to simply be suppressed, it is possible for us to trace in the sources how the report of the massacre was reworked to appear justified, as proven by the wide diffusion of this episode in more or less contemporary Christian texts.61 In what became the standard narrative of the event itself, as abundantly related in Muslim sources,62 the public remembrance of some infamous episodes of Umayyad history—chiefly the martyrdom of several key proto-Shiite figures and most notably the Prophet’s grandson al-Ḥusayn (d. 61/680)—plays a decisive role in the massacre, justifying the right of the Abbasids, new holders of the monopoly on legitimate violence, as theorized by Max Weber, to kill the kinsmen of the former masters of Damascus, in an attempt to fulfill a duty toward Islamic historical memory. Most of the narratives run as follows: about eighty members63 of the Umayyad family were invited under false pretexts for a dinner by ʿAbd Allāh ibn ʿAlī (d. ca. 147/764–765), the uncle of the first Abbasid caliph and one of the leading military figures of the fight against the Umayyads. During this encounter, a poet starts declaiming verses against the Umayyads. Thereafter, ʿAbd Allāh ibn ʿAlī speaks, recalling the death of al-Ḥusayn at the hands of Umayyad forces at Karbalāʾ. Then, with this memory still hanging in the air, he makes a sign, sending the order to his soldiers to massacre the Umayyad guests.64 He then throws carpets over the still-moving bodies, we are told, and settles down to enjoy his dinner amidst the sounds of the dying Umayyads. It is in the name of remembrance that the Umayyads were massacred.65 This presentation of the events is very easy to understand: in so doing, this episode is not to be associated with a simple vendetta, but rather to be included in the active process of legitimation of the new Abbasid regime. The narrative of these massacres is generally followed in the sources by accounts of the violation of the graves of the Umayyad caliphs, whose
6 82
286
A ntoine B orrut
remains were flogged and burned, with their ashes scattered in the wind. The veracity or the legendary character of these events is not so important, in the framework of a history of memory. The destruction of the graves, places of commemoration par excellence, shows clearly that, beyond the Umayyads themselves, it is their memory that is to be fought. As ʿAbd Allāh ibn ʿAlī tried a few years later to seize the caliphate for himself, after the death of the first Abbasid caliph in 136/754, and was defeated by the one who became the second Abbasid caliph, al-Manṣūr (r. 136–158/754–775), he was portrayed as the bloody organizer of the killing of the Umayyads. The main actor in the massacre, ʿAbd Allāh ibn ʿAlī, was thus isolated in Abbasid historiography: the hero of the victory against the Umayyads became a blood-shedding rebel.66 As he became governor of Syria immediately after the fall of the Umayyads, he was in fact “Syrianized” in Abbasid narratives: as the former Umayyad heartland was a rebellious province, so was its first Abbasid governor. Umayyad iniquity, it seems, was contagious. If this presentation of ʿAbd Allāh ibn ʿAlī conveniently allowed shifting the blame of the massacre of the Umayyads away from the rest of the Abbasid family, this successful strategy was not the only such attempt in Muslim narratives. Other sources, indeed, reflect clear efforts simply to silence this embarrassing event. This tactic is probably best exemplified by al-Ṭabarī, who mentions the massacre in a single line, almost in passing.67 The same is true with other sources that are similarly silent on the episode, such as Khalīfa ibn Khayyāṭ (d. ca. 240/854), Ibn Qutayba (d. 276/ 889), and al-Masʿūdī (d. 345/956).68 Al-Ṭabarī and others’ quasi-amnesia is quite puzzling and suggests an effort to bury a problematic past or even to suppress “controversial history.”69 Through Christian sources, on the other hand, we see Abbasid-era historiography in the making and evidence for an early circulation of the historical information about this bloody episode. This is notably true through the lens of sources borrowing material from the lost chronicle of Theophilus of Edessa, who was an eyewitness to the Abbasid Revolution70 before serving the new rulers in official capacities. As such, Theophilus offers a good example of the false dichotomy between external and internal sources, being one of the many Christians working at the Abbasid court. His lost history has attracted a good deal of scholarly attention and generated many debates that cannot be reviewed here.71 Suffice it to say that his account of the Abbasid Revolution and its immediate aftermath is arguably the material attributed the most securely to him, granted
8 2 7
The Future of the Past
287
that he was a direct observer of the events. Deriving their material from Theophilus, Agapius of Manbij (wr. ca. 942) and the anonymous Syriac chronicle of 1234 both relate a version of the massacre quite consistent with Muslim narratives even if some details vary, thus testifying for an early diffusion of the relevant historical material.72 Such an initial dissemination of information largely explains why the efforts of the apostles of silence, anxious of presenting the Abbasid coming to power under a less polemical light, were doomed to failure. Among others, this example shows how one of the main episodes of the first centuries of Islam, the so-called Abbasid Revolution, was the object of huge efforts of rewriting by later chroniclers in order to give an appropriate meaning to this founding event. To transpose the ideas of François Furet about the French Revolution, the Abbasid Revolution was no longer “a transition, [ … but] an origin and a fantasy of origin.”73 Making sense of these reconstructions would have been difficult only with the Muslim narratives, but it is fortunate that the Christian sources provide access to earlier stages of interpretation, allowing us to shed some light on the historiographical manipulations at work.74 If forgetting was simply impossible in such a case, as the memory of the event was too inescapable to be repressed, remembrance proved helpful in other cases to serve the needs of the new Abbasid regime. Indeed, despite a complicated relationship with the Umayyad past, oblivion of the whole Umayyad century was not an option for the chroniclers or for the Abbasid caliphs themselves. It was necessary to affirm the political continuity of the caliphate and to find useful models of government in the past, much as the Umayyads themselves did with the Sassanid and Byzantine precedents, for instance. In other words, when Abbasid power was more firmly established, it became possible and, to a certain degree, even necessary to redeem Umayyad memory.75 It is the control over the past that made this redemption possible, moving the Umayyads from adversity to otherness.76 This shift from adversity to otherness is perhaps best exemplified by the image of the pious caliph ʿUmar II, truly a paradigmatic example of an early Islamic site of memory.77 Indeed, despite a very short caliphate (99–101/717–720), this caliph obviously occupies a highly distinctive place in the Islamic tradition. He is usually presented as the fifth rightly guided caliph, and as a new version of his namesake, the pious caliph and companion of the Prophet Muḥammad, ʿUmar ibn al-Khaṭṭāb (ʿUmar I). He is also depicted as the redeemer (mahdī or mujaddid), given his place as
82
288
A ntoine B orrut
the ruling caliph in the year 100 of the Hijra, a time connected with strong messianic expectations. Accordingly, he was seen as an example of piety and justice for the Muslims, as a “pious” caliph, and he became the subject of biographical works early on. I have previously tried to shed some light on the making of this image not only because ʿUmar was an Umayyad caliph, a dynasty that was generally portrayed in a rather demeaning light in Abbasid historiography, but also because this constitutes an indigenous process for the making of a truly Islamic hero (or model), unrelated to Abbasid political concerns.78 I was able to show that ʿUmar II’s image in the sources was simply too early and too widely circulated to be later deleted. Thanks to the Syriac sources, we can actually prove that the making of ʿUmar II’s image as a “pious caliph” was early, going back at least to the 730s, only a decade after his death, and before the Abbasids came to power. In any case this construction was an Umayyad one, largely diffused among the historiographies at work in the Middle East and also through Islamic legal texts, most notably through Malikism. Therefore the Abbasid chroniclers had to somehow confront this pious figure and so they developed strategies to link ʿUmar II—an Umayyad caliph—with the success of the Abbasid regime. They stressed, for instance, ʿUmar II’s decisive role in authorizing the controversial wedding-alliance of two rival Arabian tribes, since the offspring of the wedding in question would eventually become the first Abbasid caliph. The chroniclers also emphasized the fact that the secret movement (daʿwa) that led to the so-called Abbasid Revolution was launched precisely during ʿUmar II’s reign, linking again the caliph with the very birth of the Abbasid regime and at the same time reinforcing the messianic dimension of the second dynasty of Islam. It is thus not surprising that, contrary to the graves of the other Umayyad caliphs, ʿUmar II’s tomb was spared by the Abbasids. The paradigmatic example of ʿUmar II is of course quite exceptional but it is by no means an isolated one. Other main characters of Umayyad history had to be reintegrated into the “official” Abbasid-era version of the past. For instance, Maslama ibn ʿAbd al-Malik, the great Umayyad general who conducted the siege of Constantinople in 97–99/715–717 and fought the Turks in the Caucasus, became a useful hero for the Abbasids as well.79 Indeed both locales, Byzantine Constantinople and the Caucasus Mountains, are clearly connected with messianic expectations. The fall of Constantinople was supposed to announce the End of Times, as well
9 82
The Future of the Past
289
as the invasion of Gog and Magog, often associated with the Turks in Islamic tradition.80 Thus Maslama was not only defining the boundaries of the Islamic empire, but also a warrior of the apocalypse. These functions stimulated various interpretations. Through Syriac contemporary sources, we can see Umayyad ideology at work: Maslama’s attributes as the great conqueror and his eschatological dimension make him particularly relevant for assimilation with Alexander the Great. The use of such mythical figures was common practice for the dynasty, as I have tried to show elsewhere with the example of Solomon.81 Such a glorious parallel between the son of Philip II and the son of ‘Abd al-Malik was, of course, of little interest to the Abbasids and they simply tried to suppress it. But Maslama was useful in another way as his failure to realize the conditions of the End of Times was essential for the success of the Abbasids, who had their own messianic claims to make. Through the Christian sources, it is again clear that Maslama’s image is early, going back, at least for some elements, to the Umayyad period. Moreover, some essentials were already known in the 730s, during Maslama’s own lifetime. This fits well with al-Ṭabarī’s affirmation that scholars and traditionists were campaigning beside Maslama on the battlefield. Hence, in a khabar ultimately placed under Sulaymān ibn Mūsā’s authority in al-Ṭabarī’s history, we learn that Khālid ibn Maʿdān al-Kalāʿī al-Ḥimṣī was present at the siege of Constantinople.82 Maslama was thus conducting his campaigns with those in charge of promoting his glory.
Conclusion These few examples illustrate, I hope, the complicated historiographical situation of the first centuries of Islam. Oblivion could at times be impossible and so the strategies of silence were doomed to failure when a piece of historical information was circulated too widely or too early, as revealed by the unsuccessful attempt to conceal the dinner-massacre of the Umayyads. On the other hand, remembrance might prove necessary to assert the political continuity of the caliphate, for instance, or to support other claims, such as those of a messianic nature, thus making the redemption of Umayyad memory indispensable. In the process, fragments (or murmurs) of Umayyad-era historiography and ideology were preserved in later Muslim narratives as demonstrated by the cases of ʿUmar II and Maslama.
0 9 2
290
A ntoine B orrut
Yet, if for long the debate was limited to the reliability of Islamic sources, I would insist that one of the most important roles of non-Muslim sources is to provide access to different moments in the stratigraphy of a Middle Eastern historiography. This is of paramount importance as the earliest Islamic narrative sources preserved are mainly from the third/ ninth century onward, so any potential access to earlier stages of Islamic historiography needs to be explored. What such a methodological approach shows is the complicated relation between remembrance and oblivion in the medieval sources themselves. It demonstrates as well that any attempt to eliminate a given theme or cluster of information depends upon the diffusion of the information. A large, and if possible early, transmission (such as the pious image of ʿUmar II or the military hero Maslama) forbids complete oblivion. The creation of a new past under the early Abbasids, as part of their process of legitimation, became “an enduring creation: its central outlines, accepted and elaborated upon by subsequent medieval generations, have been largely accepted by modern historians.”83 One of the best ways to renew our knowledge of the first centuries of Islam is certainly “to marry history and historiography”84 and to try to trace back elements of a now largely lost early Islamic historiography. Non- Muslim sources have an important role to play in this endeavor. In this perspective we must definitely follow the famous German Egyptologist Jan Assmann: before studying “the past as such,” we must study it “as it is remembered.”85
Notes 1. Borrut and Cobb 2010, 2. I am grateful to Paul M. Cobb and Sarah Bowen Savant for their helpful comments on earlier drafts of this essay. My discussion is largely based on Borrut 2011, which offers a much more extensive analysis of historical writing in early Islamic Syria and the construction of historical memory. 2. Crone 1980, 11. 3. For a convenient discussion of modern historiographical debates, see Donner 1998, 1–31. 4. See, for instance, Sharon 1997; Walmsley 2007; Sijpesteijn et al. 2007; Foss 2008; Haldon 2010; Borrut et al. 2011; Holum and Lapin 2011; Genequand 2012. 5. Assmann 2011, 33, 40 (where the English translation is not really doing justice to the German original, see Assmann 1997b, 55). 6. McKitterick 2004, 22.
291
The Future of the Past
291
7. On the concept of Sinngeschichte, see in particular Assmann 1996, 1997a. 8. Geary 1994, 177. 9. Geary 1994, 180. 10. Geary 1994, xiv; Monnet 2003, 627. 11. By historian, I mean a scholar actively engaged in historical writing. It is indeed worth remembering that a discrete category of “historians” is actually absent in the early and classical periods of Islam, as best evidenced by the lack of such a classification in the Fihrist of Ibn al-Nadīm (d. 385/995). See the useful discussions of Leder 1992; Robinson 2003, 3–17; Cheddadi 2004, 71–93. 12. Rigney 1992, 86. 13. A point already made by Elad (2002, 2003). See also Donner 1987; Donner 1998, 276–282; Khalidi 1994, 17–30; Schoeler 2009, 54–56; Robinson 2003, 18–30; Borrut 2011, 27–33. 14. Borrut 2011, 33–60. 15. Saliba 2007, 49ff., esp. 55–56. See also the stimulating discussion by MacDonald (2009) of literacy in a predominantly oral society, though the false opposition between the oral and the written should be largely rejected after Schoeler (2009). A generous survey of Arabic literature in the Umayyad period can be found in Beeston et al. (1983). 16. Ibn al-Nadīm 1871–1872, 132; 1970, 194. 17. Some Akhbār ʿUbayd are preserved in Ibn Hishām’s (d. ca. 216/831) Kitāb al- Tijān. Nabia Abbott (1957, 9ff.) has argued in favor of the authenticity of the text while most scholars adopted a much more cautious approach, following the old reservations of Krenkow (1928). See especially Encyclopaedia of Islam, 2nd ed., s.v. “Ibn Sharya” (by F. Rosenthal); Khoury 1993; Cheddadi 2004, 36–70; Elad 2002, 270–271; Humphreys 2006, 129–130; Crosby 2013. 18. See, for instance, al- Masʿūdī 1965–1979, 5:77–78; al-Masʿūdī 1962–1997, 3:726–727. 19. See, most recently, Görke and Schoeler 2008; Shoemaker 2011; Görke, Motzki, and Schoeler 2012. 20. A theory supported by Schoeler (1989, 226, 230) and criticized by Cook (1997, 465). For a more general critique of the Schachtian methodology on the dating of traditions, see Cook 1992. 21. Pace Duri 1957, see especially Lecker 1996; Borrut 2011, 45–48, 57–58; Judd 2014, 52–59. 22. Borrut 2011, 73–76. 23. On his confusing death date, see most recently Al-Qāḍī 2009. 24. Donner 1987, 4–12; Elad 2002, 260–261, 263; Borrut 2011, 42–44. 25. Borrut 2011, 52. 26. Borrut 2011, 53–54. 27. Khalek 2011, 43–52, 66–72.
9 2
292
A ntoine B orrut
28. See the short Syriac chronologies of 724 (ed., 155), 705, and 775. See also, in Greek, 818. On these various texts, see Hoyland 1997, 393–399, 434–437 (including English translations); Palmer 1993, 43–44, 49–52; Palmer 2009; and Borrut 2014a, 48–51. 29. Robinson 2003, 23. 30. Assmann 2011, 56. 31. Donner 1998, 114. 32. Donner 2010. 33. Pace Dahan’s statement (1962, 109): “During the 2nd/8th century not a single historian in Syria is known who was working on the history of his country.” 34. Wellhausen 1902; Noth 1973; Noth and Conrad 1994; Donner 1998, 216. 35. See the list compiled by Donner (1998, 297–306) and the remarks of Robinson (2003, 94). 36. Donner 1998, 195, 305; Borrut 2011, 58. 37. Madelung 2002. Ziyād ibn Abīhi himself became the subject of discrete works early on, and scholars such as Abū Mikhnaf (d. 157/774), Hishām al-Kalbī (d. 206/821), or ʿAbd al-ʿAzīz ibn Yaḥyā al-Jalūdī (d. 332/944) are credited with the compilation of Akhbār Ziyād bin Abīh. See Sezgin 1967, 262; Encyclopaedia of Islam, 2nd ed., s.v. “Ziyād b. Abīhi” (by I. Hasson). 38. On these various historiographical layers, see Borrut 2011, 61–108. 39. Borrut 2011, 97–108; Borrut 2014a, 61. 40. Geary 1994, 25. 41. On the historiographical implications of the regicide of al-Amīn, see in particular El-Hibri 1995; 1999, 59–94. 42. Borrut 2011, 98–99. 43. On Zubayrid memory, see especially Campbell 2003, and the remarks of El-Hibri 2002, 250–251; Borrut 2011, 159; Scheiner 2012. 44. Robinson 2005, 31–48. 45. This point raises important issues about the agreed periodization of early Islamic history, on which see Borrut 2014a. 46. Geary 1994, 177. 47. Borrut 2011, 184–217. 48. Robinson 2010. 49. Borrut 2011, 179–184. See also Borrut 2014a. 50. See Borrut 2011, in particular c hapters 4 and 7. 51. On the circulation of historical information between the various historiographies at work in the early Islamic Middle East, see in particular Conrad 1988; Hoyland 1997; Borrut 2011, 137–166; Debié 2015. 52. See, for instance, Sauvaget 1967, 18, or, more recently, Benkheira 2005. 53. The most radical example being of course Cook and Crone 1977. See also most recently Shoemaker 2012. 54. Borrut 2009; Borrut 2011, 137–166.
9 23
The Future of the Past
293
5 5. On Theophilus of Edessa, see Borrut 2014b, 477–481 and below. 56. Geary 1994, 180. 57. See my discussion in Borrut 2011, 109–137. 58. Geary 1994, 9. 59. Elad 1995; Robinson 2010. 60. Robinson 2010, 226. 61. Borrut 2011, 191–194. 62. In particular by al-Balādhurī 1978, 3:103–104; al-Yaʿqūbī 1883, 2:425ff. 63. Or in many sources, seventy-two victims, to precisely mirror the number of martyrs fallen at Karbalāʾ with al-Ḥusayn. See Borrut 2015; Elad 1995, 92. 64. The question of the weapons used for this bloody massacre has generated some debate in modern scholarship but it seems that the soldiers used iron staffs (ʿumud al-ḥadīd) rather than wooden weapons (kāfir kūbāt) as suggested by Crone (2000). See Elad 2000, 286–289. 65. The lexical field of memory (dhakara, tanāsī, etc.) is well attested in the Arabic sources that relate the episode. See, for instance, al-Yaʿqūbī 1883, 2:425; Ibn ʿAbd Rabbih 2001, 4:453, 5:81. 66. A point already noted by Lassner (1977; 1980, 19–38). See also Robinson 2010, 232; Marsham and Robinson 2007; Borrut 2011, 354–368. In the process, as I have tried to demonstrate, ʿAbd Allāh ibn ʿAlī was even deprived of his original nickname. He was actually the “blood shedder,” al-Saffāḥ, a laqab eventually confiscated by the first Abbasid caliph Abū al-ʿAbbās and subsequently reinterpreted, thus judging from the Arabic, turning the blood shedder into the generous. See Borrut 2011, 369–381. 67. al-Ṭabarī 1879–1901, 3:51; al-Ṭabarī 1985–2007, 27:175. 68. Khalīfa ibn Khayyāṭ 1967, 2:612; Ibn Qutayba 1960, 372; al-Masʿūdī 1965–1979, 6:75–76; al-Masʿūdī 1962–1997, 4:931; al-Masʿūdī 1894, 329; al-Masʿūdī 1896, 425. See Robinson 2010. 69. Robinson 2003, 41. 70. Agapius 1911, 525. 71. These debates are likely to continue with the recent attempt by Robert Hoyland (2011) to reconstruct Theophilus’s lost chronicle. See most recently the skeptical views of Conterno (2011) and Debié (2015), as opposed to the more optimistic approach of Van Bladel (2009, 174–175). See my discussion in Borrut 2014b, 477–481. On the broader issue of non-Muslims within the early Islamic State, see Borrut and Donner 2016. 72. Agapius 1911, 529; 1234 (ed., 333; trans., 260). 73. Furet 1978, 130. 74. For a detailed discussion of what non-Muslim sources have to offer on the topic, see Borrut 2011, 338–351. 75. El-Hibri 2002; Borrut 2011, 179–228. 76. Borrut 2011, 198–200.
9 42
294
A ntoine B orrut
77. On the concept of sites (or realms) or memory (lieux de mémoire), see the fundamental work of Nora (1996–1998). 78. Borrut 2005; Borrut 2011, 283–320. 79. Borrut 2011, 229–282. 80. Van Donzel and Schmidt 2010. 81. Borrut 2003; Borrut 2011, 217–228. 82. al-Ṭabarī 1879–1901, 2:1315; al-Ṭabarī 1985–2007, 24:39–40. 83. Geary 1994, 23. 84. Robinson 2000, viii. 85. Assmann 1997a, 9.
Bibliography P r im a ry Sour ce s 705 = Land, J. P. N., ed. 1868. Anecdota Syriaca, vol. 2:11. Leiden: Brill. 724 = Brooks. E. W., ed. 1904 (1955). “Chronicon miscellaneum ad annum domini 724 pertinens.” In Chronica Minora II, CSCO vol. 3, Scriptores Syri t. 3, 77–155. Leuven: Secrétariat du Corpus; Chabot, J.-B. trans. 1955 (1904). “Chronicon miscellaneum ad annum domini 724 pertinens.” Chronica Minora II, CSCO vol. 3, Scriptores Syri t. 3, 61–119. Leuven: Secrétariat du Corpus. 775 = Brooks, E. W., ed. 1905 (1960). “Expositio quomodo se habeant generationes et familiae et anni ab Adamo usque hunc diem.” Chronica Minora III, CSCO vol. 5, Scriptores Syri t. 5, 337–349. Paris: E Typographeo Reipublicae; Brooks, E.W., trans., 1960 (1905). “Expositio quomodo se habeant generationes et familiae et anni ab Adamo usque hunc diem.” Chronica Minora III, CSCO vol. 6, Scriptores Syri t. 6, 265–275. Leuven: Secrétariat du Corpus. 818 = Schoene, A. ed., 1875. Eusebi chronicorum libri duo, vol. 1. Berlin: Weidmann. 1234 = Chabot, J.-B., ed., 1920. Chronicon ad annum Christi 1234 pertinens. 2 vols. Paris: E Typographeo Reipublicae; Chabot, J.-B. trans. 1965 (1937). Chronicon ad annum Christi 1234 pertinens. Leuven: Secrétariat du Corpus. Agapius of Manbij. 1911. “Kitāb al-ʿunwān.” In Patrologia Orientalis 8, edited and translated by A. Vasiliev, 399–547. al-Balādhurī. 1978. Ansāb al-ashrāf, vol. 3. Edited by ʿA.ʿA. al-Dūrī. Wiesbaden: F. Steiner. Ibn ʿAbd Rabbih. 2001. Al-ʿIqd al-farīd. 7 vols. Edited by M. al-Tawnajī. Beirut: Dār Ṣādir. Ibn Hishām. 1347/1928. Kitāb al-tījān. Edited by F. Kenkrow. Hyderabad: Dāʾirat al- Maʿārif al-ʿUthmāniyya al-Kāʾina. Ibn al-Nadīm. 1871–1872. Kitāb al-fihrist. 2 vols. Edited by G. Flügel. Leipzig: Vogel. ———. 1970. The Fihrist: A 10th Century ad Survey of Islamic Culture. 2 vols. Translated by B. Dodge. New York: Columbia University Press.
9 25
The Future of the Past
295
Ibn Qutayba. 1960. Kitāb al-maʿārif. Edited by T. ʿUkāsha. Cairo: Maṭbaʿat Dār al-Kutub. Khalīfa ibn Khayyāṭ al-ʿUṣfūrī. 1967. Ta’rīkh. 2 vols. Edited by A.Ḍ. al-ʿUmarī. Najaf: Maṭbaʿat al-Ādāb. al-Masʿūdī. 1965–1979. Murūj al-dhahab wa-maʿādin al-jawhar. 7 vols. Edited by C. Pellat. Beirut: al-Jāmiʿa al-Lubnāniyya. ———. 1962–1997. Les prairies d’or. Translated by C. Pellat. Paris: Société Asiatique. ———. 1894. Kitāb al-tanbīh wa’l-ishrāf. Edited by M. J. De Goeje. Leiden: Brill. ———. 1896. Le livre de l’avertissement et de la révision. Translated by C. De Vaux. Paris: Société Asiatique. al-Ṭabarī. 1879–1901. Taʾrīkh al-rusul wa’l-mulūk. 15 vols. Edited by M. J. De Goeje. Leiden: Brill. ———. 1985–2002. The History of al-Ṭabarī. 39 vols. Albany: State University of New York Press. al-Yaʿqūbī. 1883. Taʾrīkh al-Yaʿqūbī. 2 vols. Edited by M. T. Houtsma. Leiden: Brill.
S e c onda ry Sour ce s Abbott, N. 1957. Studies in Arabic Literary Papyri I: Historical Texts. Chicago: University of Chicago Press. Assmann, J. 1996. Ägypten: Eine Sinngeschichte. Munich and Vienna: Hanser. ———. 1997a. Moses the Egyptian: The Memory of Egypt in Western Monotheism. Cambridge: Harvard University Press. ———. 1997b. Das kulturelle Gedächtnis: Schrift, Erinnerung und politische Identität in frühen Hochkulturen. Munich: C. H. Beck. ———. 2011. Cultural Memory and Early Civilization: Writing, Remembrance, and Political Imagination. Cambridge: Cambridge University Press. Beeston, A. F. L., T. M. Johnstone, R. B. Serjeant, and G. R. Smith. 1983. Arabic Literature to the End of the Umayyad Period. Cambridge: Cambridge University Press. Benkheira, M. 2005. “L’analyse du ḥadīṯ en question: À propos de A.-L. de Prémare et G. H. A. Juynboll.” Arabica 52/2: 294–306. Borrut, A. 2003. “La Syrie de Salomon: l’appropriation du mythe salomonien dans les sources arabes.” Pallas 63: 107–120. ———. 2005. “Entre tradition et histoire: genèse et diffusion de l’image de ʿUmar II.” Mélanges de l’Université Saint-Joseph 58: 329–378. ———. 2009. “La circulation de l’information historique entre les sources arabo- musulmanes et syriaques: Élie de Nisibe et ses sources.” In L’historiographie syriaque, edited by M. Debié, 137–159. Paris: Geuthner. ———. 2011. Entre mémoire et pouvoir: l’espace syrien sous les derniers Omeyyades et les premiers Abbassides (v. 72–193/692–809). Leiden: Brill.
6 9 2
296
A ntoine B orrut
———. 2014a. “Vanishing Syria: Periodization and Power in Early Islam.” Der Islam 91/1: 37–68. — — — . 2014b. “Court Astrologers and Historical Writing in Early ʿAbbāsid Baghdād: An Appraisal.” In The Place to Go: Contexts of Learning in Baghdād, 750–1000, edited by J. Scheiner and D. Janos, 455–501. Princeton: Darwin Press. ———. 2015. “Remembering Karbalāʾ: The Construction of an Early Islamic Site of Memory.” Jerusalem Studies in Arabic and Islam 42: 249–282. Borrut, A., and P. Cobb, eds. 2010. Umayyad Legacies: Medieval Memories from Syria to Spain. Leiden: Brill. Borrut, A., M. Debié, A. Papaconstantinou, D. Pieri, and J.-P. Sodini, eds. 2011. Le Proche-Orient de Justinien aux Abbassides: Peuplement et dynamiques spatiales. Turnhout: Brepols. Borrut, A., and F. Donner, eds. 2016. Christians and Others in the Umayyad State. Chicago: The Oriental Institute. Campbell, S. S. 2003. “Telling Memories: The Zubayrids in Islamic Historical Memory.” PhD. diss., University of California, Santa Barbara. Cheddadi, A. 2004. Les Arabes et l’appropriation de l’histoire: émergence et premiers développements de l’historiographie musulmane jusqu’au iie/viiie siècle. Paris: Actes Sud. Conrad, L. I. 1988. “Theophanes and the Arabic Historical Tradition: Some Indications of Intercultural Transmission.” Byzantinische Forschungen 15: 1–44. Conterno, M. 2011. “Palestina, Siria, Constantinopoli: La ‘Cronografia’ di Teofane Confessore e la mezzaluna fertile della storiografia nei ‘secoli bui’ di Bisanzio.” PhD. diss., University of Florence. Cook, M. A. 1992. “Eschatology and the Dating of Traditions.” Princeton Papers in Near Eastern Studies 1: 23–47. ———. 1997. “The Opponents of the Writing of Tradition in Early Islam.” Arabica 44: 437–530. Cook, M. A., and P. Crone. 1977. Hagarism: The Making of the Islamic World. Cambridge: Cambridge University Press. Crone, P. 1980. Slaves on Horses: The Evolution of the Islamic Polity. Cambridge: Cambridge University Press. ———. 2000. “The Significance of Wooden Weapons in al-Mukhtār’s Revolt and the ʿAbbāsid Revolution.” In Studies in Honour of Clifford Edmund Bosworth. Vol. 1: Hunter of the East: Arabic and Semitic Studies, edited by I. R. Netton, 174–187. Leiden: Brill. Crosby, E. W. 2013. The History, Poetry, and Genealogy of the Yemen: The Akhbar of Abid b. Sharya Al-Jurhumi. Piscataway: Gorgias Press. Dahan, S. 1962. “The Origin and Development of the Local Histories of Syria.” In Historians of the Middle East, edited by B. Lewis and P. M. Holt, 108–117. Oxford: Oxford University Press. Debié, M. 2015. L’écriture de l’histoire en syriaque: transmissions interculturelles et constructions identitaires entre hellénisme et islam. Leuven: Peeters.
9 2 7
The Future of the Past
297
Donner, F. M. 1987. “The Problem of Early Arabic Historiography in Syria.” In Proceedings of the Second Symposium on the History of Bilad al-Sham during the Early Islamic Period up to 40 a.h./640 a.d., edited by M. A. al-Bakhit and I. ʿAbbās, 1–27. Amman: University of Jordan. ———. 1998. Narratives of Islamic Origins: The Beginnings of Islamic Historical Writing. Princeton: Darwin Press. ———. 2010. Muhammad and the Believers: At the Origins of Islam. Cambridge, MA: Harvard University Press. Duri, A. A. 1957. “Al-Zuhrī: A Study on the Beginnings of History Writing in Islam.” Bulletin of the School of Oriental and African Studies 19/1: 1–12. Elad, A. 1995. “Aspects of the Transition from the Umayyad to the ʿAbbāsid Caliphate.” Jerusalem Studies in Arabic and Islam 19: 89–132. ———. 2000. “The Ethnic Composition of the ʿAbbāsid Revolution.” Jerusalem Studies in Arabic and Islam 24: 246–326. ———. 2002. “Community of Believers of ‘Holy Men’ and ‘Saints’ or Community of Muslims? The Rise and Development of Early Muslim Historiography.” Journal of Semitic Studies XLVII/1: 241–308. ———. 2003. “The Beginning of Historical Writing by the Arabs: The Earliest Syrian Writers on the Arab Conquests.” Jerusalem Studies in Arabic and Islam 28: 65–152. El- Hibri, T. 1995. “The Regicide of the Caliph al- Amīn and the Challenge of Representation in Medieval Islamic Historiography.” Arabica 42: 334–364. ———. 1999. Reinterpreting Islamic Historiography: Hārūn al-Rashīd and the Narrative of the ʿAbbāsid Caliphate. Cambridge: Cambridge University Press. ———. 2002. “The Redemption of Umayyad Memory by the ʿAbbāsids.” Journal of Near Eastern Studies 61/4: 241–265. Foss, C. 2008. Arab- Byzantine Coins: An Introduction, with a Catalogue of the Dumbarton Oaks Collection. Washington, DC: Dumbarton Oaks Research Library and Collection. Furet, F. 1978. Penser la Révolution française. Paris: Gallimard. Geary, P. J. 1994. Phantoms of Remembrance: Memory and Oblivion at the End of the First Millennium. Princeton: Princeton University Press. Genequand, D. 2012. Les établissements des élites omeyyades en Palmyrène et au Proche- Orient. Beirut: Institut Français du Proche-Orient. Görke, A., and G. Schoeler. 2008. Die ältesten Berichte über das Leben Muhammads: Das Korpus ʿUrwa ibn Az-Zubair. Princeton: Darwin Press. Görke, A., H. Motzki, and G. Schoeler. 2012. “First Century Sources for the Life of Muḥammad? A Debate.” Der Islam 89/2: 2–59. Haldon, J., ed. 2010. Money, Power and Politics in Early Islamic Syria: A Review of Current Debates. Farnham: Ashgate. Holum, K., and H. Lapin. 2011. Shaping the Middle East: Jews, Christians, and Muslims in an Age of Transition, 400–800 C.E. Bethesda: University Press of Maryland.
9 82
298
A ntoine B orrut
Hoyland, R. G. 1997. Seeing Islam as Others Saw It: A Survey and Evaluation of Christian, Jewish and Zoroastrian Writings on Early Islam. Princeton: Darwin Press. ———. 2011. Theophilus of Edessa’s Chronicle and the Circulation of Historical Knowledge in Late Antiquity and Early Islam. Liverpool: Liverpool University Press. Humphreys, R. S. 2006. Muʿawiya ibn Abi Sufyan: From Arabia to Empire. Oxford: Oneworld. Judd, S. C. 2014. Religious Scholars and the Umayyads: Piety-Minded Supporters of the Marwānid Caliphate. London: Routledge. Khalek, N. 2011. Damascus after the Muslim Conquest: Text and Image in Early Islam. Oxford: Oxford University Press. Khalidi, T. 1994. Arabic Historical Thought in the Classical Period. Cambridge: Cambridge University Press. Khoury, R.G. 1993. “Kalif, Geschichte und Dichtung: Der jemenitische Erzähler ʿAbīd Ibn Šarya am Hofe Muʿāwiyas.” Zeitschrift für arabische Linguistik 25: 204–218. Krenkow, F. 1928. “The Two Oldest Books on Arabic Folklore.” Islamic Culture 2: 55– 89, 204–236. Lassner, J. 1977. “Did the Caliph Abu Jaʿfar al-Manṣūr Murder His Uncle ʿAbdallāh b. ʿAlī, and Other Problems within the Ruling House of the ʿAbbasids.” In Studies in Memory of Gaston Wiet, edited by M. Rosen-Ayalon, 69–99. Jerusalem: Institute of Asian and African Studies, Hebrew University of Jerusalem. ———. 1980. The Shaping of Abbassid Rule. Princeton: Princeton University Press. Lecker, M. 1996. “Bibliographical Notes on Ibn Shihāb al-Zuhrī.” Journal of Semitic Studies 41: 26–63. Leder, S. 1992. “The Literary Use of the Khabar: A Basic Form of Historical Writing.” In The Byzantine and Early Islamic Near East I: Problems in the Literary Source Material, edited by A. Cameron and L. I. Conrad, 277– 315. Princeton: Darwin Press. MacDonald, M. 2009. “Literacy in an Oral Environment.” In Literacy and Identity in Pre-Islamic Arabia, edited by M. MacDonald, 49–118. Farnham: Ashgate. Madelung, W. 2002. “Maslama b. Muḥārib: Umayyad Historian.” In Proceedings of the 20th Congress of the Union Européenne des Arabisants et Islamisants, edited by K. Dévényi, 203–214. Budapest: Eӧtvӧs Loránd University Chair for Arabic Studies and Csoma de Kőrӧs Society, Section of Islamic Studies. Marsham, A., and C. F. Robinson. 2007. “The Safe-Conduct for the Abbasid ʿAbd Allāh b. ʿAlī (d. 764).” Bulletin of the School of Oriental and African Studies 70/2: 247–281. McKitterick, R. 2004. History and Memory in the Carolingian World. Cambridge: Cambridge University Press. Monnet, P. 2003. “Conclusion.” In Les tendances actuelles de l’histoire du Moyen Âge en France et en Allemagne, edited by J.-C. Schmitt and O. G. Oexle, 625–644. Paris: Publications de la Sorbonne. Nora, P. 1996–1998. Realms of Memory. 3 vols. New York: Columbia University Press.
92
The Future of the Past
299
Noth, A. 1973. Quellenkritische Studien zu Themen, Formen und Tendenzen frühislamischer Geschichtsüberlieferung. Bonn: Selbstverlag des Orientalischen Seminars der Universität. Noth, A., and L. I. Conrad. 1994. The Early Arabic Historical Tradition: A Source- Critical Study. Princeton: Darwin Press. Palmer, A. 1993. The Seventh Century in the West-Syrian Chronicles. Liverpool: Liverpool University Press. ———. 2009. “Les chroniques brèves syriaques.” In L’historiographie syriaque, edited by M. Debié, 57–87. Paris: Geuthner. Al-Qāḍī, W. 2009. “A Documentary Report on Umayyad Stipends Registers (Dīwān al-ʿAṭāʾ) in Abū Zurʿa’s Tārīkh.” Quaderni di Studi Arabi 4: 7–44. Rigney, A. 1992. “Time for Visions and Revisions: Interpretative Conflict from a Communicative Perspective.” Storia della Storiografia 22: 85–92. Robinson, C. F. 2000. Empire and Elites after the Muslim Conquest: The Transformation of Northern Mesopotamia. Cambridge: Cambridge University Press. ———. 2003. Islamic Historiography. Cambridge: Cambridge University Press. ———. 2005. ‘Abd al-Malik. Oxford: Oneworld. — — — . 2010. “The Violence of the Abbasid Revolution.” In Living Islamic History: Studies in Honour of Professor Carole Hillenbrand, edited by Y. Suleiman, 226–251. Edinburgh: Edinburgh University Press. Saliba, G. 2007. Islamic Science and the Making of European Renaissance. Cambridge: MIT Press. Sauvaget, J. 1967. “Châteaux omeyyades de Syrie: Contribution à l’étude de la colonisation arabe aux ier et iie siècles de l’hégire.” Revue des Études Islamiques 35: 1–52. Scheiner, J. 2012. Review of Die ältesten Berichte über das Leben Muḥammads: Das Korpus ʿUrwa ibn az-Zubair, by A. Görke, and G. Schoeler. Jerusalem Studies in Arabic and Islam 39: 437–442. Schoeler, G. 1989. “Mündliche Thora und Hadīṯ: Überlieferung, Schreibverbot, Redaktion.” Der Islam 66: 213–251. ———. 2009. The Genesis of Literature in Islam: From the Aural to the Read. Edinburgh: Edinburgh University Press. Sezgin, F. 1967. Geschichte des Arabischen Schrifttums. Vol. 1. Leiden: Brill. Shoemaker, S.J. 2011. “In Search of ʿUrwa’s Sīra: Some Methodological Issues in the Quest for ‘Authenticity’ in the Life of Muḥammad.” Der Islam 85/2: 257–344. ———. 2012. The Death of a Prophet: The End of Muhammad’s Life and the Beginnings of Islam. Philadelphia: University of Pennsylvania Press. Sharon, M. 1997–. Corpus inscriptionum Arabicarum Palaestinae. Leiden: Brill. Sijpesteijn, P., L. Sundelin, S. Tovar Torallas, and A. Zomeño. eds. 2007. From al- Andalus to Khurasan: Documents from the Medieval Muslim World. Leiden: Brill. Van Bladel, K. 2009. The Arabic Hermes: From Pagan Sage to Prophet of Science. Oxford: Oxford University Press.
03
300
A ntoine B orrut
Van Donzel, E., and A. Schmidt, 2010. Gog and Magog in Early Christian and Islamic Sources: Sallam’s Quest for Alexander’s Wall. Leiden: Brill. Walmsley, A. 2007. Early Islamic Syria: An Archaeological Assessment. London: Duckworth. Wellhausen, J. 1902. Das arabische Reich und sein Sturz. Berlin: Georg Reimer.
301
10
Caliphs and Conquerors Images of the Marwanids and Their Agents in Narratives of the Conquest of Iberia Nicola Clarke
and Christian historical traditions, narratives of the Islamic conquest of Iberia tend to share the same protagonists.1 First, there are the conquerors: Mūsā ibn Nuṣayr, governor of North Africa and son of a Persian convert to Islam, and Ṭāriq ibn Ziyād, Mūsā’s Berber client (mawlā), who is portrayed as carrying out the bulk of the campaign, including the conquest of Toledo. Second, there is Ṭāriq’s opponent, Roderic, impetuous and uninspiring king of Visigothic Iberia. Finally, back in the Muslim capital of Damascus, there are two caliphs of the Umayyad dynasty (661–750): al-Walīd (r. 705–715), the son of ʿAbd al-Malik (r. 692–705),2 and al-Walīd’s brother and successor Sulaymān (r. 715–717). The latter half of al-Walīd’s reign saw three separate military campaigns take place on the frontiers of the Umayyad caliphate; as well as Iberia in the west, over in the east Muslim forces moved into Central Asia (Transoxania) and northern India (Sind). Upon Sulaymān’s accession, however, all three of these campaigns were brought to a swift and rather acrimonious end. While this essay will concentrate on events on the western frontier, the three campaigns must in some senses be viewed as contiguous. By this, I do not mean to imply that they formed part of some consciously formulated grand strategy on the part of al-Walīd. On the contrary, in two of the three cases I would argue that the campaigns are an expression of the autonomy of frontier commanders, rather than of the centralized authority and policy-making of the caliph.3 Rather, I mean that I N M E D I E VA L M U S L I M
0 23
302
N icola C larke
the literary and historical themes that emerge from the way these campaigns are treated in the historiographical tradition give them much in common. Politically, these conquests tell us much about relations between center and periphery—both in the Umayyad period and, through their framing in the later sources, after it—and about how policy could shift dramatically with the accession of a new caliph. Socially, they point to tensions between the conquering elite and the rank and file of the Umayyad armies. Culturally, these conquests are remembered by the later tradition for what they have to say about the Umayyad dynasty and, by extension, about those who came after them: how the regime related to non-Muslims; how individual caliphs could become emblems of particular aspects of the Umayyad polity; and how, after the family was driven from imperial power in 750 and forced to seek refuge in al-Andalus (Muslim Iberia), the Umayyads who made a new home in Cordoba reimagined their links with their Iberian Peninsula, in order to cement their rule there. The year 711, in other words, was extremely important in the life of the Umayyad caliphate, in both its Damascene (661–750) and its Cordoban (emirate 756–929; caliphate 929–1009) incarnations. I shall begin with an outline of the traditional narrative of the conquest of Iberia, highlighting the roles of the main Muslim characters. I will then move on to discuss four different, but related, ways to read this narrative, in terms of events in al-Andalus and in the wider Islamic world, both at the time of the conquest and in the decades after 711. First, I will look at the question of succession in the Umayyad caliphate, focusing in particular on tensions between our two caliphs, al-Walīd and Sulaymān. Second, I shall set these tensions in the context of broader political currents, contemporary with the events of the Andalusī conquest, which centered on tribally based factionalism and whose effects rippled out to all three of the 711 frontiers. Third, I will look at the structures of political and social power portrayed in the conquest accounts: the conflict that emerged between patron (Mūsā) and client (Ṭāriq) on this western frontier, and the role of the two caliphs in resolving it. Finally, I will discuss what the narratives of the conquest of Iberia tell us about how the Umayyads were presented in the historical tradition after their overthrow in the Abbasid coup of 750, and during their subsequent recovery in al-Andalus: that is, how memories of conquest-era caliphs like al-Walīd and Sulaymān were used to reflect upon the proper role and status of a caliph, and how the events of 711 were used to create personal ties between the Umayyad family and the Iberian Peninsula, particularly through—as Maribel Fierro and
0 3
Caliphs and Conquerors
303
Ann Christys have noted—the portrayal of negotiation and accommodation with the local Christian population.4
The Traditional Narrative The invasion of Iberia in 711 was part of a second wave of Islamic imperial expansion, following the rapid seventh-century conquests and a fitful campaign in North Africa hampered by concerted Berber resistance.5 It seems that the longer the expedition went on, the more the Muslim army came to consist of Berbers who had allied themselves with the Arab and Persian Muslims. Mūsā ibn Nuṣayr, the first of our four main Muslim characters, was especially active in such recruitment; he established ties of walāʾ (clientage) with Berbers on such a large scale that he was later credited by some—perhaps rather precipitously—with the complete “conversion” of North Africa.6 After taking Tangiers in 708, Mūsā installed his client Ṭāriq ibn Ziyād as its governor, left him there with a Berber force, and returned to his seat of power in Qayrawān (modern-day Tunisia).7 Two years later, the Iberian Visigothic King Witiza died, leaving behind him a political crisis. Dynastic succession being far from automatic among the Visigoths—if anything, quite the opposite—the kingdom’s elite divided, some supporting Witiza’s designated heir, Akhila, and some preferring a rival candidate, Roderic.8 It is in the resultant context of generalized political unrest and disunity that we can place the Muslim conquest.9 According to the bulk of the Muslim tradition, whose earliest surviving examples date from the ninth century, the idea of an invasion was placed in Ṭāriq’s head by disaffected elements from within Iberia itself. Beginning with the account of the Egyptian jurist Ibn ʿAbd al-Ḥakam (d. 870), we are usually told that someone named “Count Julian”—supposedly a Byzantine or a Visigothic potentate on the coast,10 but in reality likely fictional—sought revenge for Roderic’s rape of his daughter by contacting Ṭāriq in Tangiers and inviting him to invade.11 Other, slightly later authors— notably, Andalusīs—bestow the role of traitorous, legitimizing plot device on the sons of King Witiza, left out in the political cold by Roderic’s seizure of power.12 The Latin Christian tradition, for its part, places the blame for the conquest variously on Witiza’s sons, on vacillating ecclesiastical figures, or else on punishment from God for a variety of sins.13 Whether intended as legitimizing explanation, or as excuse, the image that emerges from both traditions is of a peninsula divided.
0 43
304
N icola C larke
In the Islamic tradition, Ṭāriq needs little persuasion, hastening across the Straits with a force from Tangiers and—in some versions of the story—barely pausing to notify Mūsā, much less ask for his approval.14 Ṭāriq’s force landed at Gibraltar and fought Roderic’s army, which, already beleaguered from fighting the Basques in the north and exhausted from the forced march south, duly crumbled.15 Thereafter, Ṭāriq apparently met with little resistance, capturing city after city and taking vast amounts of plunder—including, in most accounts, a suitably legendary artifact called the Table of Solomon, on which more below. The following year, a displeased Mūsā elected to join the fray; he carried out his own conquests and caught up with his recalcitrant mawlā at Toledo in 713. Most accounts have Ṭāriq being arrested, reprimanded for acting alone, and stripped of his loot by his patron; in some, Mūsā also beats Ṭāriq, in a way that seems calculated to humiliate him and undermine his manhood (striking him with his whip and/or shaving off his beard).16 It is at this point that the story, in the medieval Islamic sources, divides. In one set of accounts, Ṭāriq is vindicated, and Mūsā punished, by the caliph al-Walīd. With both of the conquerors recalled to Damascus, Ṭāriq appeals to the caliph for redress against his patron’s unjust actions. Earlier in this version of events, Ṭāriq is said to have secretly switched one of the legs of the Table of Solomon for a dummy leg, just before handing it over to Mūsā. In the audience with the caliph at the climax of the tale, we see why: producing the true leg of the Table enables Ṭāriq to prove his case to al-Walīd; al-Walīd then strips Mūsā of the stolen profit of the conquest, and restores it to Ṭāriq. The only “complete” version of this narrative—that is, one that connects Ṭāriq and Mūsā’s quarrel, the dummy leg, and al-Walīd’s arbitration—to appear before the mid-tenth century is that of Ibn ʿAbd al-Ḥakam.17 Elements of the story can be seen in the eastern Islamic historical tradition—that is, in the books of classical writers working within the orbit of Baghdad, like al-Balādhurī (d. 892), al-Yaʿqūbī (d. c. 900), and al-Ṭabarī (d. 923)—but these accounts lack the climactic moment of Ṭāriq’s triumph before the mediating presence of al-Walīd.18 Evidently the “full” story had not yet traveled eastward by the end of the ninth century, whether because it had only recently been formulated or because it was not a significant part of Muslim historical memory beyond the west. Furthermore, the tale only entered the Andalusī canon during the tenth century; the Andalusī Mālikī Ibn Ḥabīb (d. 852), although he sourced all of his narrative in Egypt, as Maḥmūd Makkī and Jorge Aguadé have demonstrated,19 refers briefly to al-Walīd
0 35
Caliphs and Conquerors
305
ordering the Table broken up, but makes no mention of an arbitration episode of this nature.20 The second, alternative narrative turns up less frequently, but when it does, it tends to be used in chronicles written within al-Andalus—notably Ibn Ḥabīb’s—or those, like easterner Ibn al-Athīr’s (d. 1233), that drew on Andalusī works. It should be emphasized, however, that this is not a tradition that arises in al-Andalus, but rather one from Egypt that found a home in al-Andalus, and indeed Ibn ʿAbd al-Ḥakam includes both versions.21 Here, Ṭāriq plays only a minor role, and the focus is instead on the injustices suffered by Mūsā.22 Al-Walīd’s brother Sulaymān writes to Mūsā, when the latter is traveling back to Damascus, to urge him to slow his journey. Al-Walīd is dying, Sulaymān says, and so if Mūsā would be so kind as to take his time, he will arrive when Sulaymān is on the throne, at which point Sulaymān, not al-Walīd, will receive the customary caliphal share of the loot from Iberia. Mūsā ignores this request, but to no avail; by the time he reaches Damascus, Sulaymān is the caliph, and he is not best pleased with Mūsā. As in the first version of the narrative, Mūsā loses his wealth at the hands of a caliph; here, however, the focus is not on an ill-treated client finding justice through the caliph, but on the caliph’s own injustice toward Mūsā. The easterner al-Yaʿqūbī has a variant of this narrative. Here, Ṭāriq profits from Mūsā’s downfall, but he does so because he exploits Sulaymān’s existing (unexplained) antipathy toward Mūsā, denouncing his patron (“saʿā Ṭāriq … bi-mawlāhi ilā Sulaymān”) to an all-too-receptive audience, rather than because of a clever trick with a specific piece of loot, or a gesture of caliphal benevolence toward an unjustly treated underdog.23 The contrast between the two versions’ characterization of the two caliphs, and the messages they are used to convey about relations between caliph and subject, is marked. In some iterations of this second narrative, Sulaymān explicitly accuses Mūsā of not sending home the khums—the “fifth” share of any loot obtained by the Muslim armies that was supposed to be given to the caliph for the benefit of the community—upon completion of his campaign.24 Debate over the khums is a common motif in accounts of all three of the conquest campaigns that began in 711. Often, it may have been a just charge; given the circumstances, as I shall discuss below, it seems likely that frontier conquerors were rarely entirely honest about the proceeds from their activities, and moreover that the early caliphs rarely had the chance to do much about it. Nevertheless, in this version of events
0 63
306
N icola C larke
Sulaymān does not simply take the correct share from Mūsā; he humiliates him, forcing him to stand in the sun until he collapses from heat exhaustion, and then strips him of as much of his wealth as can be found.25 Interestingly, in the Latin Christian Chronicle of 754, which was composed in Muslim-ruled Iberia and bears some of the hallmarks of later Arabic historiography—such as praise of the caliph ʿUmar II—the antagonistic role is fulfilled by al-Walīd rather than Sulaymān, who ties a rope around Mūsā’s neck.26 (This swapping of caliphs is mirrored in the anonymous post-tenth-century Akhbār majmūʿa, which contains the Ṭāriq-Mūsā Table showdown described above, but places Sulaymān in al- Walīd’s role.)27 Perhaps what we have here—both in the Chronicle of 754 and in Ibn Ḥabīb’s Egyptian tales—is the earliest layer of the tradition: one of antagonism between center and periphery, in the context of frontier campaigning. I will return to this point below.
Authority and Delegation in the Umayyad Period So much for the story, or the stories. The first of the ways that we can read all this is in terms of debates over authority in the early Islamic state. In the seventh and eighth centuries, there were several competing models of how the early Islamic community should be led and of how authority could be passed on from one individual to another. At various times during the first century of Islam, claims to power were staked upon activist piety, clan prestige, descent from the Prophet Muḥammad, and the consensus of the community’s leading figures. Dynastic succession was far from being the most obvious or most popular way of doing things, and primogeniture still less so. While the Umayyads did opt for a dynasty, they held no clear principle of succession, and the question of which branch of the clan might control the throne, and thus the resources of Syria, was never a settled matter.28 Contemporary poetry gathered by Andrew Marsham demonstrates that tension between caliphs’ sons and their uncles over the succession was a recurring theme in later Umayyad politics.29 In light of this, it comes as no surprise to read in the eastern historiographical tradition—although not, notably, in Andalusī chronicles— that our caliphs al-Walīd and Sulaymān differed over more than just who was entitled to the spoils of the Iberian conquest. Al-Walīd, according to al-Balādhūrī, sought to undo his father ʿAbd al-Malik’s succession
0 37
Caliphs and Conquerors
307
arrangements and replace heir-designate Sulaymān with his (that is, al- Walīd’s) son ʿAbd al-ʿAzīz; al-Ṭabarī also quotes several poems by Jarīr ibn ʿAṭiyya (d. 729) championing ʿAbd al-ʿAzīz’s candidacy.30 In this matter, al-Walīd was apparently supported by the commanders of two campaigns that were being carried out on the eastern frontiers of the Islamic world at the same time as the conquest of Iberia: Qutayba ibn Muslim, in Central Asia, and Muḥammad ibn al-Qāsim in northern India. Also along for the ride was al-Ḥajjāj, the governor of Iraq and the direct patron of these two conquerors. I have found no indication in the historical tradition from either half of the medieval Islamic world that Mūsā himself was involved in this conspiracy. He need not have been; the simple fact that he held a position of authority, and amassed a considerable amount of wealth, under al-Walīd would have been sufficient to attract Sulaymān’s antipathy. The second theme running through the accounts of the conquest of Iberia, and indeed through accounts of contemporary conquests on the eastern frontier too, is precisely this: who controlled the resources—and the armies—on the frontiers? Whose loyalty did a new caliph reward, or win, with an appointment to a prime post in a rich land? Despite various caliphs’ efforts to the contrary, succession remained not only disputed but a matter for factionalism, fought out within the Umayyad clan through the support of the leading Arab elites and their tribal followings, conventionally referred to as the broad confederations of Qays and Yaman.31 Sulaymān’s main supporters when he came to power were the Yamanī-aligned Muhallabid family, but they gave that support in expectation of reward;32 they, in turn, fell foul of another changing of the guard a few years later, with Yazīd ibn al-Muhallab being stripped of his position by the caliph ʿUmar II (r. 717– 720) on similar charges of failing to dispatch the khums.33 Seen this way, it is entirely plausible that the eastern conquerors, Qutayba and Muḥammad, might have sought to engineer the succession of al-Walīd’s son, since—they being his supporters—that was the scenario most likely to result in continued favor and careers for them. Both were members of the Qays tribe, and therefore likely to be out in the cold under the new Yamanī-leaning regime. But regardless of the truth or otherwise of the putative conspiracy, what is striking about Sulaymān’s accession is how quickly he initiated a changing of the guard on the frontiers. Qutayba and Muḥammad were both replaced with Sulaymān’s allies and (in some versions) their heads sent back to Damascus.34 Mūsā,
0 83
308
N icola C larke
already in the capital, was disgraced and stripped of his wealth, and the son (another ʿAbd al-ʿAzīz) that he had left behind as governor in Iberia was soon killed too;35 al-Ṭabarī adds that ʿAbd al-ʿAzīz’s head was sent back to Sulaymān, thus paralleling the fates of al-Walīd’s conquerors more closely.36 There was a lot of money to be made on the frontiers, and ambitious individuals could carve out autonomy there with a military following of kinsmen, tribesmen, and clients, so far from central oversight.37 It was only natural, therefore, that the accession of a new caliph would create (or exacerbate) tensions between center and periphery, as the caliph sought to wrest back control of frontier resources, in order to bestow these resources upon his own supporters; it was to be expected, also, that the caliph’s efforts in this regard would have been hampered by distance and the limited reach of the state, and thus reliant on alliances made with magnates who had tribal backing and a thirst for gold and glory, and who barely needed the excuse of acting on the caliph’s behalf to bring down their rivals. In those accounts of the conquest of Iberia in which Mūsā and Sulaymān clash, therefore—the ones with (just about) the earliest roots, Ibn Ḥabīb (d. 852) and his sources—these are the dynamics at work. Mūsā’s fate is typical of the period. Here it must also be noted that the importance of the year 711 and its three conquests does not lie in grand caliphal strategizing, in a deliberate plan to expand the empire simultaneously to east and west, however tempting it might be to assume this, looking at a modern map. Centralizing elements are visible after the fact: within a year or so of the initial expedition into Iberia, gold coins were issued in the peninsula bearing shortened versions of the shahāda in Latin,38 and governors were sent out from (and recalled by) Damascus on a regular basis, as both surviving lead seals39 and the Chronicle of 754 attest.40 But the caliph replacing, every few years, an individual ally and agent of his authority with another individual ally and agent of his authority does not in and of itself constitute careful caliphal oversight, any more than co-opting preexisting local mints to pay the army in itself implies a functioning bureaucracy. Even as the Chronicle of 754 catalogues al-Andalus’s governors, it also notes their autonomy from the center and the frequent unauthorized changes of power.41 The lead seals, meanwhile, are exciting material evidence for the names and authority of some of our governors, but what they mainly attest to is a series of local peace agreements42 and some concern for licit loot (fayʾ).43 That al-Walīd or Sulaymān—or their immediate successors—sent someone to a territory
0 93
Caliphs and Conquerors
309
in order to secure a share of the spoils from a successful raid does not mean that either of them ordered that raid in the first place. Nor do the lead seals necessarily indicate anything more than an ad hoc oversight of tribute extraction; the most frequent term that appears on them is qism,44 which (along with other iterations of the root q-s-m) appears frequently in conquest narratives to denote the correct division of spoils after a successful raid, not organized taxation.45 Nevertheless, parts of the medieval historiographical tradition try very hard to present these conquests as centrally directed and legally watertight. As might be expected, a tendency to tidy up the history of the Iberian campaign is present, especially in chronicles written in al-Andalus during the tenth century and later, in a courtly milieu that was—as Gabriel Martinez-Gros and Eduardo Manzano Moreno have explained—infused with the assumptions and priorities of the reconstituted Umayyad caliphate of Cordoba.46 Unlike the earliest Arabic sources of the ninth century,47 these later accounts abound in letters supposedly exchanged between caliph and conquerors, in an apparent effort to confirm, at every step of the way, that patrons had control of their clients and that the conquest was carried out under properly designated authority.48 By the tenth century, legal standards for conquest had definitively changed, and the chroniclers collected their information and structured their accounts with those standards in mind: the tax status of a territory or the cities within it depended on how it had been conquered, and caliphs were assumed to have always had the same sort of authority that they enjoyed at the time of writing in the tenth century.49 In fact, this was not the real story: the conquest of Iberia in 711, in a similar way to the conquests in the east, was not so much ordered by the caliph as recognized by him after the fact. In reality, raiding and skirmishes— which sometimes resolved into conquests— were simply something that ambitious individuals on the frontiers did, regardless of whether or not the center wanted them to. There may have been elements of religious duty at work; jihād had yet to be fully co-opted as a function of the state rather than an obligation upon (unpredictable and unruly) individuals.50 But more importantly—and not necessarily in contradiction to any religious impulse—there existed frontier war bands with time on their hands, and commanders whose careers depended on military prestige and the loot they could distribute to their kin and their clients.51 In a sense, the conquest of Iberia was carried out just to give the Berber rank and file something to do.
031
310
N icola C larke
Ethnicity and Social Status The third and fourth themes that I want to consider in the conquest narratives are more closely tied to the figure of al-Walīd. The third concerns ethnic and cultural differences—and concomitant differences in social status—both at the time of the conquest and as they were later accommodated (or not) within the political and social life of al-Andalus. The most obvious opposition of this nature in the conquest accounts is between the conquerors, Mūsā and Ṭāriq. Mūsā’s father was a Persian convert to Islam, but the Arabic chronicles—few of which mention his heritage52—unquestionably position him as being among the conquest elite, in a position of social and military authority over a great many clients. Ṭāriq, for his part, is almost always explicitly identified as a Berber and as a client:53 the junior partner who, it is made clear, is expected to defer to Mūsā. The fact that he does not, and indeed raids Iberia under his own steam, may be, as I have noted, a genuine reflection of the circumstances of how the Islamic frontiers worked; it clearly caused discomfort in chroniclers of the tenth century, who made sure to state that Ṭāriq had written to Mūsā for permission before starting to raid. Ṭāriq going it alone, however, also enabled later historians to set up a conflict between patron and client, in which the patron was greedy and unreasonable, and the client was the put-upon figure who did all the real work while the patron took the credit. Thus Mūsā arrives in Iberia in high dudgeon, not only rebuking Ṭāriq but also hitting him with a whip and stealing his hard-won wealth (including the Table of Solomon, as discussed above) to boot. Happily for all concerned, at this point the conquest narrative could offer an Umayyad caliph swooping in to the rescue, demonstrating the proper operation of hierarchical power relations. A good patron ensures the welfare of his client, rather than abusing him; a good caliph, moreover, is responsible for seeing that justice is done and the correct ordering of society is maintained.54 More concretely, it is essential that a good caliph ensures the profits of conquest are properly distributed among those entitled to a share. Thus al-Walīd investigates the matter and upholds Ṭāriq’s rightful claim to the movable property. This, I would suggest, is the reason why—despite its absence from Ibn Ḥabīb’s account—this version of the conquerors’ interactions with their caliphs eventually found its way into the Andalusī tradition, as reflected in a pair of anonymous compilations, the Akhbār majmūʿa and the Fatḥ al-Andalus.55 These collections postdate
31
Caliphs and Conquerors
311
the fall of the Umayyad caliphate at the beginning of the eleventh century but preserve segments of now-lost tenth-century material.56 The story of al-Walīd judging between Ṭāriq and Mūsā shows an early Umayyad caliph embodying justice and right order, and it fits with the Cordoban Umayyads’ renewed interest in their self-presentation as caliphs during the tenth century.
Cordoba: The Umayyads after 750 In this showdown between patron and client, therefore, the fourth and final of my themes is also at work: how the presentation of the Umayyad caliphs in the Iberian conquest narratives reflects the complicated history of the Umayyads after the year 750. As Ann Christys has demonstrated, the image of the Umayyads (or certainly of al-Walīd) as restorers of the social and political order, seen in the Table of Solomon story, is also visible in the treatment of Christian characters within the conquest narratives.57 Indeed, this trend begins as far back as the Christian Chronicle of 754, whose anonymous author reflects the same preoccupation with rightful authority, protection of property rights, and justice, when he says that a governor appointed over Iberia by the Umayyads, named al-Ḥurr (r. 716– 718), restored to the Christians property that had been wrongfully taken from them at the conquest.58 The author who was most obviously interested in this issue is Ibn al- Qūṭiyya (d. 977), who begins his chronicle of the history of al-Andalus with an account of his own Visigothic ancestors’ involvement in the events of the conquest.59 He first portrays the sons of Witiza, who have been shut out of power by the usurpation of Roderic (whom he calls “the tyrant”), turning to Ṭāriq for help. Having given symbolic consent to the Muslim conquest of Iberia, the princes are passed up the chain of Islamic authority, from Ṭāriq to Mūsā to the caliph al-Walīd. In later years, the daughter of one of these collaborators, Sara, loses some of her property to a relative and travels to Damascus to meet with the Umayyad caliph of the day, Hishām (r. 724–743). Hishām duly ensures that her property is restored to her, in memory of the ties of clientage established between the two families at the time of the conquest, and he sees her married to another of his clients. During this visit to Damascus, Sara also encounters the young Umayyad prince ʿAbd al-Raḥmān, who will later be the sole survivor of the massacre of his family during the Abbasid coup in 750 and who will re-found the dynasty in al-Andalus.
231
312
N icola C larke
By the time Ibn al-Qūṭiyya was writing this account, in the mid-tenth century, a third Umayyad ʿAbd al-Raḥmān had sat on the throne in Cordoba and had once again adopted the title of caliph to express his authority.60 Even before the proclamation of the renewed caliphate in 929, he and his predecessors had poured considerable time and money into broadcasting a message of continuity with the Umayyads’ Damascene past. Chief among the media for this image was the Great Mosque of Cordoba, which blended the style of al-Andalus with that of Umayyad Damascus, including a minbar built to the old Umayyad template (which was itself, Maribel Fierro has argued, designed to hark back to the Prophet’s)61 and a qibla wall that indicated the direction of Mecca from the old Umayyad capital rather than from the new one.62 When ʿAbd al-Raḥmān’s successor al-Ḥakam II rebuilt the qibla after a further expansion, he followed his Damascene forebears in seeking out Byzantine mosaicists to create its decoration.63 The link with the past is also present in the historical tradition: tenth-century chroniclers’ accounts of the mosque’s history drew parallels between the Church of St. John in Damascus, which had been shared between Muslims and Christians before being converted in a mosque, and the Church of St. Vincent that underlay the mosque in Cordoba.64
Conclusion As a strategy for the legitimization of Umayyad rule in al-Andalus, the message of Ibn al-Qūṭiyya’s version of the conquest of Iberia, centered as it is on the Umayyads and their native Christian clients, could not be clearer. Here is an appeal to Islamic history, in the form of the Umayyad family’s past glory in Damascus. Here is also an appeal to the conquered Christian people of the peninsula, or perhaps more importantly to their converted descendants, many of whom now formed the core of the Umayyad court in Cordoba: their aristocracy had chosen to collaborate with the conquerors and the caliphs, not out of cowardice or sin, as it is presented in the Christian tradition, but out of a desire to see justice done and order restored. In the story of Sara, we are shown that only the Umayyads could function as the proper authorities in the peninsula, protecting the property of a noble Christian woman wronged by her uncle. Ibn al-Qūṭiyya’s account may have been a minority report, but its themes and goals are shared by the rest of what became the Andalusī tradition: in the tale of the Table of Solomon, we are shown that the Umayyads ensured justice was done in response to events in al-Andalus,
31
Caliphs and Conquerors
313
even before they came to make their capital there. The image of the rapacious Sulaymān stealing the somewhat ill-gotten gains of the conquest from Mūsā came to be balanced, and even overshadowed, in the Andalusī historical tradition by al-Walīd’s good judgment over the Table of Solomon, coming to the aid of wronged non-Arab client Ṭāriq. Ṭāriq had been an unruly adventurer, in the early Muslim tradition, a typical autonomous frontier commander; now he had become the grateful recipient of caliphal oversight and benevolence. The Umayyads were and always had been caliphs, whether in Cordoba or in Damascus, and the narratives surrounding the events of 711 were a vital part of reinforcing that idea.
Notes 1. The primary accounts drawn upon for this essay are, in probable order of composition, Lopez Pereira 1980 (tr. in Wolf 1999) and Fernández 1985 (tr. in Wolf 1999); Ibn ʿAbd al-Ḥakam 1922; Fernández 1985 (tr. in Wolf 1999); al-Balādhurī 1866; al-Yaʿqūbī 1883; al- Ṭabarī 1879– 1901, 15 vols.; Ibn al- Qūṭiyya 1926; Lafuente y Alcántara 1867; Molina 1994; al-ʿAbbādī 1971; Ibn al-Athīr 1965– 1967; Ibn ʿIdhārī 1948–1951. 2. Taking the beginning of ʿAbd al-Malik’s reign proper to be his defeat of Ibn al-Zubayr. 3. See section “Authority and Delegation in the Umayyad Period”, and also Clarke 2012, chap. 7, esp. 127. 4. Fierro 1989, 502, 511; Christys 2002, 179–180. 5. Abun-Nasr 1987, 28; Manzano Moreno 1990, 410, 417; Manzano Moreno 2010, 583–584; Brett and Fentress 1996, 83–87. 6. Djaït 1967, 106–107; Abun-Nasr 1987, 34–35; Lewicki 1970, 209–210. Whether either clientage or entry to the army necessarily “presupposed” conversion to Islam (Abun-Nasr 1987, 34) is doubtful; for further discussion on this point, see Clarke 2012, 51–52. 7. Ibn ʿAbd al-Ḥakam 1922, 205; Encyclopaedia of Islam, 2nd ed., s.v. “Ṭāriḳ b. Ziyād” (by L. Molina). 8. Lopez Pereira 1980, parag. 52. The crown had, in theory, been elective since the previous century: Grierson 1941, 114–116. Such conflict and confusion over the Visigothic throne were not unusual, as the succession to Wamba shows (Fernández 1985, parag. 2), and there are numerous examples of coins issued by “kings” not in the conventional regnal lists: Miles 1952, 30 (Iudila), 37–38 (Suniefred), 40–42 (Akhila). 9. Frez 2002, 621–624. 10. Vallvé 1989, 121; Clarke 2012, 108–111.
431
314
N icola C larke
11. Ibn ʿAbd al-Ḥakam 1922, 205–206; Ibn Ḥabīb 1991, 136–137, writing a little earlier, opts—characteristically—for prophecy as a motivator for conquest, rather than factionalism. 12. E.g., Ibn al-Qūṭiyya 1926, 2–4. 13. Lopez Pereira 1980, parags. 53–55; Fernández 1985, parags. 7–8. 14. In Ibn ʿAbd al-Ḥakam 1922, 205, he does not even do that much; cf. Molina 1994, 15–16. 15. Reilly 1993, 52. 16. Ibn Ḥabīb 1991, 138; Ibn ʿAbd al-Ḥakam 1922, 207; al-Balādhurī 1866, 230–231 (less dramatic); al-Ṭabarī 1879–1901, ii.2:1253–1254; Lafuente y Alcántara 1867, 18–19; Molina 1994, 24–25; al-ʿAbbādī 1971, 149–150; Ibn al-Athīr 1965–1967, 4:565; Ibn ʿIdhārī 1948–1951, 2:16. Ibn al-Qūṭiyya (1926, 10) skips over this incident. 17. Ibn ʿAbd al-Ḥakam 1922, 211. Other versions of this: Lafuente y Alcántara 1867, 29–30 (on which more below); Molina 1994, 36; Ibn al-Athīr 1965–1967, 4:566. 18. al-Balādhurī 1866, 231, says only that Ṭāriq found the Table, which Mūsā then presented it to al-Walīd; al-Ṭabarī 1879–1901, 2.2:1253–1254, has Ṭāriq finding the Table only after he has already placated the angry Mūsā, and no mention is made of a showdown in front of either caliph; al-Yaʿqūbī 1883, 2:341, 353, mentions that Ṭāriq replaced a Table leg but leaves out the “punchline” of him producing it for al-Walīd. 19. Ibn Ḥabīb 1991, 71–72, 104–107; Makkī 1998, 205. 20. Ibn Ḥabīb 1991, 147. 21. Ibn ʿAbd al-Ḥakam 1922, 210–211; for further discussion of the origins of these accounts, see Clarke 2011, 45–48. 22. Ibn al-Qūṭiyya 1926, 10–11. 23. al-Yaʿqūbī 1883, 2:353. 24. Ibn ʿAbd al-Ḥakam 1922, 211; Molina 1994, 36. 25. Ibn Ḥabīb 1991, 147; Ibn ʿIdhārī 1948-1951, 2:20 (essentially the same account). 26. Lopez Pereira 1980, parag. 56. 27. Lafuente y Alcántara 1867, 29–30. 28. Marsham 2009, 118–119. 29. Marsham 2009, 113–114, 122–124. 30. al-Balādhurī 1997–2004, 7:28, 44 (the statement that the scheme was the suggestion of al-Ḥajjāj and Qutayba is in the second reference); al-Ṭabarī 1879–1901, ii.2:1283–1284. 31. Ibn Aʿtham (1968–1975, 7:205) notes of Qutayba’s appointment that “there was not a Qaysī in Khurasan who was not overjoyed …, nor a Yamanī who did not hate it.” In both Ibn Aʿtham’s (1968–1975, 7:252–279) and al-Ṭabarī’s (1879– 1901, 2.2–3:1283–1305) somewhat different versions of the events surrounding Qutayba’s downfall, the Yamanī-aligned Azd tribe emerge as particular antagonists of Qutayba—and, ultimately, custodians of his severed head (Ibn Aʿtham
351
Caliphs and Conquerors
315
1968–1975, 7:276). For further discussion of these groupings, see Crone 1994, 1–57. 32. They sometimes acted as the caliph’s agents in removing his brother’s appointees, e.g., al-Yaʿqūbī 1883, 2:356. Ibn Ḥabīb 1991, 147, puts Yazīd ibn al-Muhallab by Sulaymān’s side when the latter abuses Mūsā, too, although regrettably for the present argument he is not portrayed as being evilly delighted by this turn of events. 33. al-Ṭabarī 1879–1901, 2.3:1334–1335, 1350. 34. Qutayba: Ibn Aʿtham 1968– 1975, 7:276– 279; al- Balādhurī 1866, 424. Muḥammad: al-Balādhurī 1866, 441; al-Yaʿqūbī 1883, 2:356. 35. Ibn al-Qūṭiyya 1926, 11; Lafuente y Alcántara 1867, 20. 36. al-Ṭabarī 1879–1901, 2.3:1306. 37. One example is Ibn al-Ashʿath’s Peacock Army: al-Ṭabarī 1879–1901, 2.2:1046, 1052–1077, 1085–1125. 38. See, for example, Manzano Moreno 2006, 57, fig. 2.2 (dated to the year 94/712– 713; the slogan is In nomine Domini non Deus nisi Deus solus). 39. See, for example, the images collected for the first time by Ibrahim 2011, 147–161. 40. Lopez Pereira 1980, parags. 59, 75, 77 (via the North African governor), 78, 81. 41. Lopez Pereira 1980, parags. 59, 69, 74, 78. 42. Ibrahim 2011, 153–156. 43. Ibrahim 2011, 157. 44. Ibrahim 2011, 148–150. 45. See Clarke 2012, 131–133, for more. 46. Martinez-Gros 1992, 19–20 and passim; Manzano Moreno 1997, 74–75, 83–84. 47. Ibn Ḥabīb (1991, 136–137), has Ṭāriq write to Mūsā for permission to carry out the conquest, but this is confirmed by an omen rather than Mūsā’s inherent authority as Ṭāriq’s superior; Ibn ʿAbd al-Ḥakam (1922) gives no indication of the chain of command until p. 207, when Ṭāriq formally hands over the command of the conquest to Mūsā, long after the fact. 48. Ibn al-Qūṭiyya 1926, 3–4; Molina 1994, 16–17. The tenth-century author al- Rāzī’s work does not survive, although based on two independent quotations of his text, he showed Mūsā giving permission in writing prior to the raid being launched: Ibn ʿIdhārī 1948–1951, 2:5–6; al-Maqqarī 1968, 1:259. 49. Clarke 2012, 42–46, 126–131. 50. Cf. the discussion in Robinson 2000, 109–126. 51. Qawm tends to be the term used for such a war band; Clarke 2012, 127; Crone 1980, 55. There are examples of its use to describe some or all of the invaders in the Transoxania narratives (al-Balādhurī 1866, 424; al-Yaʿqūbī 1883, 2:354–355; Ibn Aʿtham 1968–1975, 7:273), and those for al-Andalus (Ibn al-Qūṭiyya 1926, 7; Lafuente y Alcántara 1867, 8). 52. Ibn Ḥabīb 1991, 138– 139, notes Nuṣayr’s exalted position with the caliph Muʿāwiya, but not his heritage.
631
316
N icola C larke
53. Ibn ʿIdhārī (1948–1951, 2:5) catalogues some minor differences of opinion on this. 54. Crone and Hinds 1986, 26–42. 55. Lafuente y Alcántara 1867, 29 (with, as noted above, Sulaymān playing al-Walīd’s role); Molina 1994, 29–30, 36. 56. Molina 1989, 513; Manzano Moreno 1992, 43. Molina (1994, 120) gives a terminus ad quem for that work’s compilation with a reference to Yūsuf ibn Tashufīn (d. 1106) as if he is still alive, and the text’s editor, pp. XXVIII–XXX, argues that it is a redaction of an eleventh-century work by Ibn Muzayn, prince of the ṭāʾifa of Silves, who drew upon Ibn Ḥayyān (d. 1076), who in turn had used Ibn Ḥabīb, al-Rāzī, and others. 57. Christys 2002, chap. 8. 58. Lopez Pereira 1980, parag. 64; this commentary is lent considerable weight, now, by the lead seals discussed above. For the circulation of historical knowledge between Christian and Muslim sources, see Borrut (this volume). 59. Ibn al-Qūṭiyya 1926, 1–6. 60. The text of the address appears in Lévi-Provençal and García Gómez 1950, 79, and the surviving fragment of ʿArīb ibn Saʿīd (d. 980): Castilla Brazales 1992, 205–206. 61. Fierro 2007, 153–161. 62. Khoury 1996, 80, 83–84; Fernández-Puertas 2009, 296–299. This was also common practice in early North African mosques, for example, in Fez; Cornell 1998, 297, n. 32. I am indebted to attendees at the Muslim-Christian Encounters seminar at the Khalili Research Centre, University of Oxford (February 15, 2011), for bringing this wider context to my attention. 63. Fernández-Puertas 2009, 380; Ibn ʿIdhārī 1948–1951, 2:237 (al-Ḥakam sent to Constantinople for decorative materials, again as had been done for the Damascus mosque); Fierro 1991, 1:135–139, 143–144. On the Byzantine mosaicists at the Prophet’s Mosque in Medina, see George (this volume). 64. Ibn ʿIdhārī 1948–1951, 2:229 (quoting al-Rāzī).
Bibliography al-ʿAbbādī, A. M., ed. 1971. Taʿrīkh al-andalus li-ibn al-kardabūs wa-waṣfuhu li-ibn al- shabbāṭ. Madrid: Maʿhad al-Dirāsāt al-Islāmiyya. Abun- Nasr, J. M. 1987. A History of the Maghrib in the Islamic Period. Cambridge: Cambridge University Press. al-Balādhurī. 1866. Futūḥ al-buldān. Edited by M. J. Goeje. Leiden: Brill. ———. 1997–2004. Ansāb al-ashrāf. Edited by S. N. Mārdīnī. Damascus: Dār al-Yaqazah. Brett M., and E. Fentress. 1996. The Berbers. Oxford: Blackwell.
371
Caliphs and Conquerors
317
Castilla Brazales, J., trans. 1992. La Crónica de ʿArīb sobre al-Andalus. Granada: Impredisur. Christys, A. 2002. Christians in al-Andalus (711–1000). Richmond: Curzon. Clarke, N. 2011. “Medieval Arabic Accounts of the Conquest of Cordoba: Creating a Narrative for a Provincial Capital.” Bulletin of the School of Oriental and African Studies 74(1): 41–57. ———. 2012. The Muslim Conquest of Iberia: Medieval Arabic Narratives. London: Routledge. Collins, R. 2004. Visigothic Spain, 409–711. Malden: Blackwell. Cornell, V. J. 1998. Realm of the Saint: Power and Authority in Moroccan Sufism. Austin: University of Texas Press. Crone, P. 1980. Slaves on Horses: The Evolution of the Islamic Polity. Cambridge: Cambridge University Press. ———. 1994. “Were the Qays and Yemen of the Umayyad Period Political Parties?” Der Islam 71(1): 1–57. Crone, P., and M. Hinds. 1986. God’s Caliph: Religious Authority in the First Centuries of Islam. Cambridge: Cambridge University Press. Djaït, H. 1967. “La wilāya d’Ifrīqiya au IIe/VIIIe siècle: Étude institutionelle.” Studia Islamica 27: 71–121. Fernández, J. G., ed. and J. L. Moralejo, trans. 1985. Crónicas Asturianas. Oviedo: Universidad de Oviedo. Fernández-Puertas, A. 2009. Mezquita de Córdoba: Su studio arqueológico en el siglo XX. Granada: Universidad de Granada. Fierro, M. I. 1989. “La obra histórica de Ibn al-Qūṭiyya.” al-Qanṭara 10(2): 485–512. ———. 1991. “En torno a la decoración con mosaicos de las mezquitas omeyas.” In Homenaje al Prof. Jacinto Bosch Vilá, 1, 131–144. Granada: Universidad de Granada. ———. 2007. “The Mobile Minbar in Cordoba: How the Umayyads of al-Andalus Claimed the Inheritance of the Prophet.” Jerusalem Studies in Arabic and Islam 33: 153–161. Frez, A. I. 2002. “Conflictos internos y externos en el fin del reino visigodo.” Hispania 62(211): 619–636. Grierson, P. 1941. “Election and Inheritance in Early Germanic Kingship.” Cambridge Historical Journal 7(1): 1–22. Ibn ʿAbd al-Ḥakam. 1922. Kitāb futūḥ Miṣr wa-aẖbārihā. Edited by C. C. Torrey. New Haven, CT: Yale University Press. Ibn Aʿtham. 1968–1975. Kitāb al-futūḥ. 8 vols. Hyderabad: Maṭbaʿat Majlis Dāʾirat al-Maʿārif al-ʿUthmāniyya. Ibn al-Athīr. 1966. al-Kāmil fī al-taʾrīkh. Edited by C. J. Tornberg. Beirut: Dār Ṣādir. Ibn Ḥabīb. 1991. Kitāb al-taʾrīkh. Edited by J. Aguadé. Madrid: Consejo Superior de Investigaciones Científicas.
831
318
N icola C larke
Ibn ʿIdhārī. 1948–1951. Kitāb al-bayān al-mughrib fī akhbār al-Andalus wa’l-Maghrib. 2 vols. Edited by G. S. Colin and E. Lévi-Provençal. Leiden: Brill. Ibn al-Qūṭīyya. 1926. “Taʾrīkh iftitāḥ al-Andalus.” In Historia de la conquista de España del abenalcotía el Cordobés: Seguida de fragmentos históricos de Abencotaiba, etc, edited by J. Ribera. Madrid: Tipografía de la “Revista de Archivos.” Ibrahim, T. 2011. “Nuevos documentos sobre la conquista Omeya de Hispania: Los precintos de plomo.” In 711: Arqueología e Historia entre dos mundos, 1: 145–162. Alcalá de Henares: Museo Arqueológico Regional. Khoury, N. N. N. 1996. “The Meaning of the Great Mosque of Cordoba in the Tenth Century.” Muqarnas 13(1): 80–98. Lafuente y Alcántara, E., ed. 1867. Akhbār majmūʿa. Madrid: Rivadeneyra. Lévi-Provençal, E., and E. García Gómez, eds., and trans. 1950. Una Crónica anónima de ʿAbd al-Raḥmān III al-Nāṣir. Madrid: Consejo Superior de Investigaciones Científicas. Lewicki, T. 1970. “Les origines de l’islam dans les tribus berbères du Sahara occidental: Mūsā ibn Nuṣayr et ʿUbayd Allāh ibn al- Ḥabḥāb.” Studia Islamica 32: 203–214. Lopez Pereira, J. E., ed. and trans. 1980. Estudio crítico sobre la Crónica Mozarabe de 754. Zaragoza: Anubar. Makkī, M. A., 1998. “Egypt and the Origins of Arabic Spanish Historiography: A Contribution to the Study of the Earliest Sources for the History of Islamic Spain.” In The Formation of al-Andalus, Part 2: Languages, Religion, Culture and the Sciences, edited by M. Fierro and J. Samsó, 173–233. Aldershot: Ashgate Variorum. Manzano Moreno, E. 1990. “Beréberes de al-Andalus: Los factores de una evolución histórica.” Al-Qanṭara 11: 397–428. ———. 1992. “Oriental ‘Topoi’ in Andalusian Historical Sources.” Arabica 39: 42–58. ———. 1997. “El ‘medio cordobés’ y la elaboración cronística en al Andalus bajo la dinastía de los omeyas.” In Historia social: Pensamiento historiográfico y Edad Media, edited by G. I. García, 59–85. Madrid: Ediciones del Orto. ———. 2006. Conquistadores, emires y califas: Los omeyas y la formación de al-Andalus. Barcelona: Crítica. ______. 2010. “The Iberian Peninsula and North Africa.” In The New Cambridge History of Islam, edited by C. F. Robinson, 1: 581–621. Cambridge: Cambridge University Press. al-Maqqarī. 1968. Nafḥ al-ṭīb. 8 vols. Edited by I. Abbas. Beirut: Dār Ṣādir. Marsham, A. 2009. Rituals of Islamic Monarchy: Accession and Succession in the First Muslim Empire. Edinburgh: Edinburgh University Press. Martinez-Gros, G. 1992. L’idéologie omeyyade: La construction de la légitimité du Califat de Cordoue (Xe–XIe siècles). Madrid: Casa de Velázquez. Miles, G. C. 1952. The Coinage of the Visigoths of Spain: Leovigild to Achila II. New York: American Numismatic Society.
931
Caliphs and Conquerors
319
Molina, L., ed. 1994. Fatḥ al-Andalus: La conquista de al-Andalus. Madrid: Consejo Superior de Investigaciones Científicas. Molina, M. 1989. “Los Ajbār Maŷmūʿa y la historiografía árabe sobre el periodo omeya en al-Andalus.” Al-Qanṭara 10: 513–542. Reilly, B. F. 1993. The Medieval Spains. Cambridge: Cambridge University Press. Robinson, C. F. 2000. Empire and Elites after the Muslim Conquest: The Transformation of Northern Mesopotamia. Cambridge: Cambridge University Press. ———, ed. 2010. The New Cambridge History of Islam, vol. 1. Cambridge: Cambridge University Press. Samsó, J., and M. Marin, eds. 1998. The Formation of al-Andalus, Part 2: Languages, Religion, Culture and the Sciences (Formation of the Classical Islamic World, 47). Aldershot: Ashgate. al-Ṭabarī. 1879–1901. Taʾrīkh al-rusul wa’l-mulūk. 15 vols. Edited by M. J. de Goeje. Leiden: Brill. Vallvé, J. 1989. “Nuevas ideas sobre la conquista árabe de España: Toponimia y onomástica.” Al-Qanṭara 10: 51–150. Wolf, K. B., trans. 1990. Conquerors and Chroniclers of Early Medieval Spain. Liverpool: Liverpool University Press. al-Yaʿqūbī. 1883. Taʾrīkh. 2 vols. Edited by M. T. Houtsma. Leiden: Brill.
321
11
The Umayyads in Contemporary Arab TV Drama ʿUmar ibn ʿAbd al-ʿ Azīz and al-Ḥ ajjāj ibn Yūsuf in Musalsalāt Jakob Skovgaard-Petersen
Arabic historiography the Umayyads may be l’ancien régime, modern Arabic writings have taken a more nuanced, if no less ideological, view of the first dynasty of Islam. The aim of this essay is to see how the historical and fictional treatment of the Umayyads, which began in the early twentieth century, has been taken up by the musalsal (pl. musalsalāt), or Ramadan TV drama, which reaches huge audiences in the Arab world and beyond. The Ramadan musalsalāt are clearly the place to look if one wants to investigate how the Umayyad period is known and understood by contemporary Arab Muslims. I study the figures of al- Ḥajjāj ibn Yūsuf and ʿUmar ibn ʿAbd al-ʿAzīz, as these have been the main characters of two different musalsalāt, one Egyptian and the other Syrian, that provide quite different understandings of these two characters. WHILE TO CLASSICAL
Modern Arabic Historiography and the Umayyads Founding and expanding a vast empire, amassing great wealth and bringing together Arabic-speaking tribes and other peoples to forge what is now the Arab world, the Umayyads had obvious attraction to twentieth-century
23
322
J akob S kovgaard - P etersen
Arab nationalists, not least those concerned with the Arabic language as the vehicle of an “Arab spirit.” Confronting their negative treatment in classical sources, theirs was a task of distilling a pure Umayyad history from the murky concoctions of Abbasid historiography. Werner Ende, who studied the modern Arab historians’ treatment of the period, concluded that twentieth-century writers went about this challenge in principally three ways:1 a. They searched for the positive details and judgments in the classical literature and emphasized them. b. They reinterpreted events and items that early readers would have condemned in a more positive light. c. They applied a vocabulary of source criticism to dismiss very negative treatments as partisan and substituted them with their own more positive versions. Ende also identified certain recurring themes in the historiography; unsurprisingly, there was an interest in the nature of the rule of the Umayyad caliphs, in their clashes with the Shiʿites, and in their treatment of the mawālī (clients) and non-Muslims. More curiously, mid-twentieth-century historians developed an interest in class and social consciousness during the Umayyad period. Ende concentrated his study on nonfiction, and especially on the first half of the twentieth century, when a methodical historical profession was still in its infancy. However, right from its beginning, the modern interest in Arab history also gave rise to fictional treatments. The historical novels of Alexandre Dumas, Eugene Sue, and Walter Scott appeared in Arabic translations in the late nineteenth century and inspired a few Arab, mainly Christian, writers to try their hand at Arabic historical material.2 The most famous of these was Jurji Zaydan who between 1891 and 1914 published twenty-one historical novels, which are still available and read today. After a few scattered attempts, from 1896 he began to publish them following a chronology of Arabic history that he simultaneously traced in his five-volume History of Islamic Civilization (Tārīkh al-tamaddun al- islāmī). One interesting feature of this series is that it prepares a role for history in modern Arab mass culture. It is striking that Zaydan’s themes and personalities for his historical novels have resurfaced in the twenty- first century within television dramas. The Umayyad themes selected by Zaydan can be glanced from his titles: Ghādat Karbalāʾ (“The Young
2 3
The Umayyads in Contemporary Arab TV Drama
323
Lady of Kerbala,” 1901), Al-Ḥajjāj ibn Yūsuf (1902), Fatḥ al-Andalus (“The Conquest of al-Andalus,” 1903), Shārl wa-ʿAbd al-Raḥmān (“Charles and ʿAbd al-Raḥmān,” 1904), and Abū Muslim al-Khurāsānī (1905). As a publicist Zaydan realized the popularity of the genre of the historical novel, which he serialized in his journal al-Hilāl. But he was also well aware of the didactic benefits of creating a historical consciousness in his modern Arab readers; in both the History of Islamic Civilization and the historical novels certain moral lessons are put forward. Writing on the Abbasids, for example, Zaydan would stress how the Muslims thrived in periods when they allowed themselves to learn from foreign civilizations and to enjoy poetry and arts. In the preface to one of his novels, Al-Ḥajjāj ibn Yūsuf, he outlined his general approach; the historical material should not, he wrote, merely function as a stage of event, as would be the case with the novels of Dumas. Rather, it should be the other way around, in that elements of drama must be added to the narration of historical events, which must be edifying. Modern Western scholars have evinced little admiration for this choice. “A certain set of persons appear in most of his plots,” asserts Thomas Philipp: There is the hero and the heroine, a young girl well educated and intelligent, who is in love with the hero. She is supported by an equally well educated and intelligent mother. A totally devoted male or female servant generally completes the list of representatives of the good element. On the evil side we find usually the father of the girl who is cruel, uneducated and tyrannical and tries to force his daughter to marry the villain of the plot, who does not love the girl but wants to marry her for his own personal advantage.3 Much of this recurs in the ordinary TV drama of a hundred years later. Jurji Zaydan was no Arab nationalist, and he wrote his novels just before the revaluation of the Umayyads set in. He reproduced the classical historical verdict on the Umayyads and was attacked for it by Rashid Rida and Rafiq al-ʿAzm. Zaydan deplores the “chauvinism of the Arabs” and the “Umayyad contempt towards the mawālī”: “Islam should have raised the status of the client, but the Umayyads thought it fit to treat him with contempt on the ground of non-Arab descent.”4 There are, however, instances where he distances himself from this negative image, especially in his fairly positive assessment of the Umayyads’ tolerance and employment
2 4 3
324
J akob S kovgaard - P etersen
of Christians and Jews. Although himself a Syrian Christian, Zaydan is, on the other hand, very defensive of Islam and in his judgment of the various caliphs, their personal piety seems to indicate their level of morality and justice. Particular criticism is directed at al-Ḥajjāj ibn Yūsuf (ca. 661–714), who, although a capable administrator, was even worse than his corrupt masters ʿAbd al-Malik and al-Walīd. Zaydan provides a list of his misdeeds:5 • Directing his catapults against the Kaʿba, cutting the head of Ibn al- Zubayr and kindling a fire within its precincts. • Vesting the title caliph with the meaning of God’s viceregent • Extremely cruel torture before execution • Devastating Iraq This image of al-Ḥajjāj as a harsh and tyrannical ruler can still be found today, as we shall see, but it was seriously challenged by a number of later Arab nationalist historians and even a couple of novelists such as Mahmud Taymur. They pointed to al-Ḥajjāj’s rhetorical skills and political wisdom, and from the 1930s onward there was also a certain admiration for his political ruthlessness, which led to the “reunification of the Arab empire” after years of infighting. Ende provides a long list of modern biographers who considered al-Ḥajjāj’s every move as serving the higher interest of the state.6 It was, after all, the Umayyads, and not the Abbasids, who expanded the Arab empire and that required the pacification of its interior. ʿUmar II is, generally speaking, a less divisive figure. But precisely his relationship to al-Ḥajjāj and his reversal of the policy toward the mawālī mean that the authors may either consider him fortunate to be able to build on what had been achieved in terms of unity and Arabization, or they may join classical historiography in seeing him accomplish a rectification of policies gone astray, by introducing the rule of Islamic law, respecting the fuqahāʾ (religious scholars), and accepting the mawālī as fully legitimate Muslims with the same rights as everyone else. In the majority of modern Sunni Arab treatments of al-Ḥajjāj, then, he comes across as a ruthless but visionary Arab nationalist who relentlessly served his masters in their struggle to expand and pacify the great Umayyad state. Ende’s study is from 1977 and thus prior to political events in which the Arab nationalist rulers of Syria and Iraq, Hafez al-Asad and Saddam Hussein, brutally crushed internal dissent in the name of Arab unity and the need to confront non-Arab enemies. How has their rule
2 35
The Umayyads in Contemporary Arab TV Drama
325
influenced the perception of al-Ḥajjāj ibn Yūsuf and ʿUmar ibn ʿAbd al-ʿAzīz?
The Musalsal Let us move on to television. For various reasons, television is by far the most important medium in the Arab world today, and with the rise of satellite television, a great number of channels (now more than six hundred) are competing for an audience that even includes the many Arabs in other parts of the world. In television fiction, the Arab specialty is the Ramadan series, musalsal ramaḍān, which runs in thirty episodes in the evenings after iftar throughout the month. Here too the competition is fierce. From its weak beginnings in the 1970s, Egypt was the all-dominant provider of both state-and privately produced series, but with the rise of satellite television, private Syrian companies, frequently with financial backing from Gulf investors, secured almost half of the market, often with more upscale productions. Historical drama has been part of the fare from the very beginning. These productions are, however, fairly expensive to produce and rarely among the most successful, so in the 2000s they numbered less than 10 percent of an annual production that would typically reach between sixty and a hundred musalsalāt. Since the early 1990s there has been a marked rise in quality of these series, with the best Syrian and Egyptian productions being written by established authors and directed by some of the best directors. Gulf money has also ensured that the most ambitious can depict major battles or lavish court scenes. But thirty episodes of typically forty minutes each are still challenging, for authors and directors alike, and scriptwriters are once again forced to insert maids, servants, friends, and other fictional characters to add personal and emotional dimensions to historical figures that they cannot find in sufficient detail from the work of Arab historians. This is certainly also the case with the Umayyads. The best-known musalsalāt on the Umayyads are a Syrian trilogy from the early 2000s on the Spanish Umayyads ʿAbd al-Raḥmān I and II and Ṭāriq ibn Ziyād. As with so many other treatments of Muslim Spain, they perpetuate the theme of confessional tolerance and scientific achievement, and it is interesting that these figures, like al-Ḥajjāj ibn Yūsuf, were also treated in dramas by Mahmud Taymur around the middle of the twentieth
6 23
326
J akob S kovgaard - P etersen
century. There is, however, an added theme, as John Shoup has pointed out, namely a lash at the material and scientific underperformance of Arab societies today.7 On the eastern Umayyads and early Abbasids, there are series about Abū Jaʿfar al-Manṣūr (Syria, 2008) and Mūsā ibn Nuṣayr (Egypt, 1990).
ʿUmar ibn ʿAbd al-ʿAzı̄z ʿUmar ibn ʿAbd al-ʿAzīz qualifies as a historical series, but also as a religious one because it comes close to hagiography. Produced by the Egyptian state production company Ṣawt al-Qāhira, and with the popular star Nur al-Sharif as ʿUmar II, it was screened in Ramadan in 1995 and very well received. The script was written by ʿAbd al-Salam Amin, a veteran of Egyptian television productions who had written shows for the October victory celebrations and the Ramadan quiz show with the eternal star Nelly.8 He had also written several Ramadan musalsalāt about Dhū al-Nūn al-Miṣrī and Rābiʿa al-ʿAdawīyya and about the rightly guided caliphs. The subtitle of the series about ʿUmar II was The Fifth of the Rightly Guided. One striking feature about this series is its length and disposition. It spans not thirty, but thirty-eight episodes, and ʿUmar only becomes caliph in episode 35. In fact, the first nineteen episodes take place during the reign of ʿAbd al-Malik, who dies in 705 when ʿUmar is just twenty-three. The reason is not only that we need to see ʿUmar as a child prodigy, deeply devoted to religion and justice, but also that the series is about the moral decay of the Umayyads and al-Ḥajjāj ibn Yūsuf as the main factor in this development. It is the juxtaposition of al-Ḥajjāj and ʿUmar that is at the core of the narrative. In the very first scene of the first episode, probably in 685, we meet al-Ḥajjāj attending a meeting between Marwān and his two sons ʿAbd al-Malik and ʿAbd al-ʿAzīz (ʿUmar’s father). And in the second scene we see five-year-old ʿUmar ask his mother what the caliphate is about, and her answering that al-khilāfa wilāyat umūr al-nās bi-iqāmat al-ʿadl baynahum (“the caliphate is rule over people’s affairs by bringing justice between them”). The meaning of the caliphate is a recurring theme until the end, and ʿUmar is always considering it as a vocation and a burden. Like Harry Potter, he has a scar (he was kicked by a horse as a child), which leads every knowledgeable person to consider him an heir to the caliphate and to the style of his famous namesake and maternal grandfather, ʿUmar ibn al-Khaṭṭāb.
237
The Umayyads in Contemporary Arab TV Drama
327
ʿUmar II is about eleven years old when he confronts al-Ḥajjāj for the first time; he is coming to Medina to study fiqh and is courteously invited by al-Ḥajjāj to visit the governor’s house. Before showing up, however, ʿUmar insists that he will visit the tomb of the Prophet and the house of his grandfather (episode 3: 32–37). When studying in Medina, ʿUmar is escorted by a fellow student who is spying for al-Ḥajjāj and can report how ʿUmar is asking questions to Anas ibn Mālik and others about the caliphate and proper government and is listening to Anas’s complaints about al-Ḥajjāj. Upon hearing these reports, al-Ḥajjāj has Anas arrested for conspiracy, but ʿUmar goes to confront him and makes sure that a transcription of their meeting is sent to the caliph in Damascus. ʿAbd al- Malik believes ʿUmar and transfers al-Ḥajjāj to Iraq (episode 4: 10–25). During the following episodes ʿUmar discovers widespread corruption in the state treasury, related to al-Ḥajjāj who has agents working there, and he manages to convince ʿAbd al-Malik that it is time for the empire to expand again, this time in North Africa. Unsurprisingly, it is also ʿUmar who comes up with the idea of introducing a caliphal currency, with the argument that relying on the enemy’s hard currency might be fatal as they can manipulate their distribution and value (episode 8: 18–25). Just like it turns out that it is ʿUmar who demands that a major mosque be built in Damascus (episode 21: 17). This is in the first weeks of the reign of al-Walīd, in 705, and a new confrontation is taking place with al-Ḥajjāj, as the new caliph has called both of them home to Damascus to advise him. Al-Ḥajjāj is trying to have ʿUmar sentenced to death and produces evidence of mutiny in the northern town of al-Khanasir, where ʿUmar has for some years been the governor. The evidence, it turns out, consists of a couple of verses written by a blind man about how injustice reigns on earth. ʿUmar knows the man as he has granted a pension to him—and to all blind and lame inhabitants of the city. ʿUmar answers with the Qurʾanic verse that God made man His khalīfa on earth, so the caliph must judge, and if there is any truth in the blind man’s claim then he is not to be punished. The blind man is harshly interrogated by al-Ḥajjāj, and although he answers calmly, with ample use of Qurʾanic quotations, al-Ḥajjāj succeeds in making him state that Umayyad rule is mere mulk (kingship). This is an affront to the caliph, and although the blind man does not confess to sedition he is sent to prison (episodes 21, end, and 22, beginning) and ʿUmar struggles in vain to get him released. Al-Walīd is slowly developing a complete dependence on
2 8 3
328
J akob S kovgaard - P etersen
al-Ḥajjāj, so much so that he is deeply affected by al-Ḥajjāj’s death in 714 and dies himself shortly afterward. In the meantime, ʿUmar has been the governor of Medina for some years. He himself had suggested to al-Walīd that the appointment of the governor should be up to the fuqahāʾ of that city, and they elected him. There is a discussion between al-Walīd and ʿUmar about the requirements of rule, and while al-Walīd stresses military prowess, ʿUmar insists that a ruler must be just and make no distinction between rich and poor, and seek guidance in the Book of God and the sunna of His messenger (episode 25: 0–5). As governor, ʿUmar institutes a council of fuqahāʾ to advise him, and he displays clemency toward his enemies and generosity toward the poor. Needless to say, this is all repeated on a larger scale when ʿUmar finally and reluctantly assumes the caliphate in the last four episodes; he commissions a collection of prophetic Hadith, institutes a welfare system throughout the Umayyad domains, and orders a halt to the squandering of public money by Umayyad princes. He himself leads such a puritan lifestyle that his old nurse surprises him by repairing a wall in his humble house. He dies poor, but happy. Just before his death he has visited the shepherd—a recurring figure throughout the musalsal who always seems to know what is going on and comments upon it in a philosophical language, while also stressing the truth of the Qurʾan and the sunna—and learned that he would be given the name “the Fifth of the Rightly Guided.” As can be seen from this résumé, Jurji Zaydan’s technique of introducing figures around historical persons—or adding psychological substance especially to female figures only briefly mentioned in the sources—is here employed to create drama. The shepherd is important not just as an oracle and commentator, but also because ʿUmar’s grandmother was the daughter of a shepherd—a social fact employed against him not so much by al-Ḥajjāj as by ʿAtīqa, the evil scheming sister of ʿAbd al-Malik and ʿAbd al-ʿAzīz. Finally, the theme of the shepherd is often evoked by ʿUmar to explain the true role of the caliph, who must tend to his subjects like a flock. Al-Ḥajjāj only meets the shepherd once, in episode 32, which is almost fully devoted to his death. Feeling death approaching, al-Ḥajjāj has set out to seek advice and some consolation in the fact that he has erected mosques in Iraq; the shepherd, however, rejects him and states that everything is in the Book of God. Al-Ḥajjāj is then consumed by fear of meeting God; he tries to atone by giving money to one of those he had abused earlier, but in vain, and amidst nightmares and praying he finally breathes
9 23
The Umayyads in Contemporary Arab TV Drama
329
his last. ʿUmar, of course, leaves judgment to God, but the rest of us can see that al-Ḥajjāj was only serving the Umayyads insofar as it served him. He had his own spy network, sometimes collaborating with agents of the enemy; he was personally corrupt, cruel, and insatiably ambitious. The one major good deed that is mentioned is his commissioning of a Qurʾan with diacritical marks. But when he presents it, he is only trying to avoid the rage of ʿAbd al-Malik because ʿUmar has disclosed his graft at the treasury (episode 11: 25). The ʿUmar ibn ʿAbd al-ʿAzīz series evinces no trait of admiration for the powerful and visionary; the truly good ruler is the one who does not want to rule but who quotes the Qurʾan and Hadith at every turn and leaves it to the fuqahāʾ to tell him what is the law and how it can best be applied. When (in episode 11) ʿAbd al-Malik asks his family how to spend the new wealth of the state, the others answer that it needs fortifications and more cavalry, but ʿUmar proposes to secure the hajj with stations and build mosques in Mecca and Medina.
Al-Ḥ ajjāj ibn Yūsuf The other series dealing with the same material is Al-Ḥajjāj ibn Yūsuf (2003). This series was important in that it heralded a new and more ambitious treatment of historical material, shot partly on location (as Syrian musalsalāt do much more than Egyptian ones), partly in the reconstructed cities of Mecca, Kufa, and Basra. This very expensive décor was made possible by the investment of the producer, Qatar Television. In the role of al-Ḥajjāj was the Syrian ʿĀbid Fahd, who won a prize for best actor that year and went on to become the lead actor in a number of historical films, such as Al-Ẓāhir Baybars, and is today an absolute star of Syrian drama. The series was directed by Muhammad al-ʿAziziyya, a Jordanian who has specialized in expensive historical dramas such as Khālid ibn al- Walīd and Suqūṭ al-Khilāfa (2010) about the reign of the Ottoman sultan ʿAbd al-Hamīd II. The playwright, Jamal Abu Hamdan, is also Jordanian, better known for his novels and short stories than for his TV scripts. He was praised for trying to depict a character of complexity in the portrait of al-Ḥajjāj.9 I will not go through Al-Ḥajjāj ibn Yūsuf in the same detail. In the beginning we meet him working as a teacher in Taʾif, stern but loved by his students, who are very sorry to learn that he will migrate as a result of a humiliation perpetrated by the soldiers of the anticaliph Ibn al-Zubayr
0 3
330
J akob S kovgaard - P etersen
who has established himself in Mecca. Al-Ḥajjāj sets out for Damascus where he is employed in the police and soon becomes known as a man who can deliver discipline—both in his troops and in society. He works himself up the ranks of the army of the caliph ʿAbd al-Malik, and one day, in 692, is put in charge of troops laying siege to Mecca and his personal enemy Ibn al-Zubayr. The bombardment of Mecca is depicted as violent but not gruesome, and its religious dimensions do not occupy center stage. ʿAbd al-Malik in Damascus knows that it is wrong, but he still hopes for a victory, and al-Ḥajjāj bribes his messenger and threatens him never to reveal that the caliph had in fact forbidden the attack. ʿUmar, back in Damascus, is against it. Ibn al-Zubayr is abandoned, even by his sons (episode 19). When at the end Ibn al-Zubayr kneels in prayer at the Kaʿba, al-Ḥajjāj is even directing his manjanīq (stone-throwing siege machines) to hit it, consoling his worried soldiers that hitting an enemy of God cannot be a sin. The musalsal leaves little doubt that al-Ḥajjāj is transgressing religious norms here. But that does not seem to be all there is to be said about it; in other scenes he appears God-fearing. And he is a patriot: it is he who turns the administration to Arabic and introduces the Arab dinar. First of all, of course, he is a statesman who wages wars to pacify Iraq and Khurasan, not for his own gain but as a loyal servant of ʿAbd al- Malik and al-Walīd. An interesting scene to be found is both series is his famous khuṭba to the Muslims of Kufa, where he initially appears with his face covered by a red silk turban.10 The ʿUmar ʿAbd al-ʿAzīz series (episode 4) shows a tiny fragment of the speech to demonstrate his violent temper and tyrannical ways. The Al-Ḥajjāj ibn Yūsuf series shows him no less intemperate, but it displays the whole speech as a brilliant act of taking control over a restive town with amazing eloquence and willpower (episode 24). As in the series on ʿUmar ibn ʿAbd al-ʿAzīz, al-Ḥajjāj’s death is narrated over most of an episode and here, too, he is absorbed with memories and doubts. But this time there is no weakness; he is still in control and can express his faith in God with confidence. Preachers all over the country are ordered to pray for him. As he dies at the end of episode 35, a narrator takes over and enumerates his achievements: Between the rainbow and the sword did al-Ḥajjāj inscribe his name in the annals of history, in a mixture of blood and ink. Al-Ḥajjāj ibn Yūsuf al-Thaqafī died in Ramadan in the year 95 ah, at the age of 52. He left behind strong imprints, with the quelling of civil strife,
331
The Umayyads in Contemporary Arab TV Drama
331
the unification of the country, conquests, construction, and reform. He presided over the most extended conquests of any governor in history. But his image as a tyrant is the part which endures. As pointed out by the critic Yusuf Abu al-Khayl, this was one of the first series to portray a significant figure of sacred Islamic history as a complex personality, with tyrannical traits but also achieving great feats.11 The humanity of the portrait should make it more interesting, but perhaps also more inspiring. The lead actor, ʿĀbid Fahd, pointed out that the series was made before the invasion of Iraq, but it seemed relevant to Iraq as al-Ḥajjāj was a “school of politics, organization and army.”12 The hero was no saint, but great achievements and impressive will seemed to count for more in this world than the good governance of ʿUmar ibn ʿAbd al-ʿAzīz.
Conclusion History writing, as we know, is always about the present as much as it is about the past. This is particularly true when it comes to expensive historical television fiction produced by intellectuals, funded by state companies, censored by other states, and broadcast to huge audiences. The Umayyads, with their nationalist appeal and complex Islamic legacy, are once again at the center of the struggle over Arab historical consciousness, now in the form of one huge public sphere, covering much of the territory that the Marwanids once opened for Islam. A part of that struggle is reflected in the remarkable differences between the musalsal production in Egypt, which in the 1990s was allowed to introduce piety as a theme in the characterization, and the more recent Syrian production, which very much tends to treat religion and piety as accidental factors in the production of plots and characters. As we have seen, the recent dramatic treatment of the Umayyads within Arab TV dramas generally replicates the Arab historical novels of Jurji Zaydan and their cast of characters. They can be critical of the received knowledge about a period or a character, but their aim is certainly not historical accuracy. On the contrary, the figures of history are vested with the personal traits needed to make them characters of a drama, as well as paragons of rulers: the dramatic interest in figures such as ʿUmar ibn ʿAbd al-ʿAzīz and al-Ḥajjāj ibn Yūsuf stems from their suitability to embody values and orientations that are relevant today. Here we can note a difference between the Egyptian series, concerned as it is with Islamic
2 3
332
J akob S kovgaard - P etersen
moralism and populism, and the Syrian series that aims at portraying a complex character who is, in the end, considered a great statesman and nationalist. As with the later series on Baybars, there is an emphasis on the main character’s humanity, but also a tacit acceptance of the cruelty needed to realize higher political aims.
Notes 1. Ende 1977, 291–294. 2. Dupont 2006, 327. 3. Philipp 1979, 235. 4. Zaydan 1907, 117. 5. Zaydan 1907, 102–112. 6. Ende 1977, 225. 7. Shoup 2005. 8. Nelly Artine Kavian (b. 1949). 9. Idrīs 2003. 10. Ṭabarī 1989, 13–16. 11. Abū al-Khayl 2006. 12. Taha 2004.
Bibliography Abū al-Khayl, Y. 2006. “al-Ṭaʾrīkh al-islāmī bayn maṭraqat al-taqdīs wa-sanadāt al- naqd.” Al-Riyāḍ, December 4, 2006. Di-Capua, Y. 2009. Gatekeepers of the Arab Past: Historians and History Writing in Twentieth-Century Egypt. Berkeley: University of California Press. Dupont, A.-L. 2006. Gurgi Zaydan (1861–1914): écrivain reformiste et témoin de la Renaissance arabe. Damascus: Institut Français du Proche-Orient. Ende, W. 1977. Arabische Nation und Islamische Geschichte: Die Umayyaden im Urteil Arabischer Autoren des 20. Jahrhunderts. Beirut, Wiesbaden: Orient-Institut der Deutschen Morgenländischen Gesellschaft, F. Steiner. Idris, M. 2003. “al-Mushāhidūn yarawn an al-musalsalāt al-ʿarabiyya tatashābah maʿa al-maksīkiyya fi’l-ratāba wa-suṭḥiyyat al-mawāḍīʿ.” al-Sharq al-awsaṭ, December 5, 2003, http://archive.aawsat.com/details.asp?section=25&issueno=9138&artic le=206063&feature=1#.WWDjM4qx9E5 (accessed July 10, 2017). Philipp, T. 1979. Gurgi Zaidan: His Life and Thought. Beirut: Orient Institut der Deutschen morgenländischen Gesellschaft. Shoup, J. 2005. “As It Was, and as It Should Be Now: Al Andalus in Contemporary Arab Television Dramas.” Transnational Broadcasting Studies 15, http://tbsjournal.arabmediasociety.com/Archives/Fall05/Shoup.html (accessed December 8, 2015).
3
The Umayyads in Contemporary Arab TV Drama
333
al-Ṭabarī. 1989. The History of al-Ṭabarī. Vol. 22: The Marwanid Restoration. Translated by E. Rowson. Albany: State University of New York Press. Taha, A. M. 2004. “al-Fannān al-sūrī ʿĀbid Fahd [Interview].” al-Sharq al-awsaṭ, June 25, 2004, http://archive.aawsat.com/details.asp?issueno=9165&article=241225#. WWDkKIqx9E4 (accessed July 10, 2017). Zaydan, J. 1907. Umayyads and ʿAbbásids: Being the Fourth Part of J. Zaydán's History of Islamic Civilization. Translated by D. S. Margoliouth. Leiden/London: Brill/ Luzac.
4 3
Photography credits: Nadia Ali (fig. 7.1), Marion Berti (fig. 6.10, pl. 10), Sheila Blair and Jonathan Bloom (fig. 2.4), Hans-Caspar Graf von Bothmer (figs. 2.9, 2.10), Denis Genequand (figs. 6.3, 6.7, 6.8, 6.11, pl. 9), Alain George (figs. 2.7, 2.8), Robert Hoyland (fig. 51, pl. 7, 8), Said Nuseibeh (figs. 2.1, 2.2), John Oleson (pl. 11), Dick Osseman (pl. 1), Christian Robin (fig. 5.1). Every effort has been made to trace copyright holders and to credit them appropriately, but if any have been omitted or cited incorrectly, the authors will be pleased to make the necessary arrangement at the first opportunity.
35
Index
Note: Italic page numbers represent figures and Pl. represents plates. Abā II (Catholicos) 267 ʿAbbās, Iḥsān 4 Abbasid chroniclers 282–3, 288 copper coin 141 coup 302, 311 (see also Abbasid Revolution) court 286 era xiv, xv, xvii, xviiin3, 5, 141, 153, 177, 268, 271n59, 280 sources 177, 180, 280, 284 (see also historiography) Abbasids xiv, xviiin4, 26, 46, 121n42, 163, 275, 276, 277, 282, 283, 285–7, 288, 289, 323, 324, 326 Abbasid-period paintings 160, 161 Abbasid Revolution xiv, xviiin3, 115, 275, 281, 285–7, 288. See also Abbasid: coup; historiography ʿAbd Allāh ibn Abī Hāshim 209 ʿAbd Allāh ibn ʿAlī 285–6 ʿAbd al-ʿAzīz ibn Mūsā ibn Nuṣayr 308 ʿAbd al-ʿAzīz ibn al-Walīd ibn ʿAbd al-Malik 307 ʿAbd al-Ḥamīd (ibn Yaḥyā al-Kātib) 8, 163 ʿAbd al-Ḥamīd II (Ottoman Sultan) 329
ʿAbd al-Malik xiv, 6–7, 19, 46, 49, 52, 136, 169, 270n55, 278, 281, 289, 301 chancery reform 76 coinage and monetary reform 7, 9, 10, 20–23, 26, 27, 32n72, 153, 167 and Dome of the Rock 7, 23, 40, 46 as “God’s caliph” 6–7, 24, 27 and Ibn al-Zubayr 7, 22, 23, 24, 46, 281 as intercessor 48 leading prayer 56 milestones of 75, 76 poetry about 24, 48, 56 succession plans 306–7 in television serials 324, 326, 327, 328, 329, 330 ʿAbd al-Raḥmān (I) 311, 323, 325 ʿAbd al-Raḥmān (II) 325 ʿAbd al-Raḥmān (III) 312 ʿAbd al-Raḥmān ibn Muḥammad 46 ʿAbīd ibn Sharya al-Jurhumī. See ʿUbayd ibn Sharya al-Jurhumī Abu Hamdan, Jamal 329 Abū Jaʿfar al-Manṣūr. See al-Manṣūr Abū ʿUbayda ibn al-Jarrāḥ 136
6 3
336
Index
Adam 10–11, 13, 14–19, 15, 16, 23, 26, 27, 256, Pl.1 Adamnan 143n10 Adiabene 261, 266 Agapetus 12, 13 Agapius of Manbij 287 Aght’amar 100 agricultural activity 176, 182 estates xvi, 109 fertility 109, 208 manor 101 products 171 See also calendars, agricultural agriculture 48, 105, 134, 148, 176, 208 Ai Khanum 89 al-Akhṭal 23–4, 28, 48, 208 Alexander the Great 289 Amin, ʿAbd al-Salam 326 amīr al-muʾminīn 8, 9, 21, 139–40 amr Allāh 26 ʿAmr ibn Mattā 271n59 Anas ibn Mālik 327 Anasartha. See Khanāṣira Anatolian highlands xiv, 133 al-Andalus 302, 305, 308, 309, 310, 311, 312 Andarīn xvii, 133–42, 209 Androna. See Andarīn angels 15, 17, 41, 43, 54, 62n38 Antioch, mosaics of 103, 206 Apamea 222 Arab nationalists 322, 324 Arabic inscriptions. See inscriptions Arabic, as official language of the administration 76, 330 Arabic-speaking tribes xiv, 19, 28, 321 Argos 172n13, 189, 194 calendar of 189, 193, 194, 207 aristocrats, Christian, as patrons of churches and monasteries 257, 265
army Muslim, in Iberia 303 payment of 308 Sasanian, defection to Arab conquerors 257 Umayyad xiii, 302 Arshām 263, 264 artists 160, 179, 206, 226, 232, 233. See also copyists; craftsmen; mosaicists; painters; sculptors al-Asad, Hafez 324 Asia Minor. See Anatolian highlands Avestan 41 ʿAwāna ibn al-Ḥakam 280 al-Awzāʿī 191 Ayyubids 233, 242n178 Badr al-Dīn ibn Muẓaffar 232 Baghdad 117, 119, 163, 265, 268, 277, 281, 304 al-Balādhurī 257, 304, 306 Balqāʾ 175 Banū Ghassān 279 baptism 264–5 Barmakids 281 barracks 134, 140. See also kastron Basilica of Campanopetra, Constantia 218, 220 Basilica of the Nativity, Bethlehem 213, 214, 222, 223, 230 Basilica of Tegea 189, 190, 207 Basques 304 bathhouses Andarīn 134, 137, 141, 209 continuity of bathing culture 210, 233, 240n119, 121 iconography 221, 226–7, 232–3 inscription at Hammat Gader 27 Khirbat al-Mafjar 176, 178, 209, 211, 229–30 Mamluk 242n178 pre-Islamic 209–10, 217, 221
37
Index Quṣayr ʿAmra 175 See also Dibsi al-Faraj; Hammat Gader; Khirbat al-Mafjar; Salamis- Constantia; Serjilla; Thuburbo Maius, Tunisia; Tiberias; Zenobiyya, bath complex Bawit (monastery) 203 bayʿa 26–7 Baysan-Scythopolis 196, 200, 201, 207 bedouin 134. See also pastoralists Berbers 301, 303, 309, 310 Beth ʿĀbe 258, 260, 261, 267 Beth Alpha 202 Beth Garmai 266 Beth Ṭeḥūnai 262 Bezqīn 263, 264 Bible 11, 16, 24, 51, 77 Bīshāpūr 163 bishops 12, 134, 256, 257, 258, 261, 266. See also catholicos Bishr ibn Marwān 267 Book of the Cave of Treasures 17, 19 Book of Governors 258, 259, 260, 261 bridle 156, 160, 163. See also saddlery Byzantine Empire xiii, 196, 255 art and architecture 43, 45, 47, 83–6, 88, 100–101, 112, 116, 118–19, 161, 192–3, 195–6, 207–8, 212 as model for Abbasids 287 See also craftsmen; mosaicists; Roman Empire calendars, agricultural 182, 188, 189, 190–1, 206, 233, 234, Pl.13 Christian saints depicted in 203 collocation of 205 grid patterns of 189–93, 205 iconography of official images 193–95 organizing principle of 204–5 planets and zodiacal signs depicted in 202–3, 203, 205
337
regional variants 204 seasonal images depicted in 196–200 caliph of God. See God's caliph calligraphers. See copyists calligraphy 58, 75 Cappella Palatina, Palermo 100 Carthage 196, 197, 198, 206, 207 catholicoi. See catholicos catholicos 255, 256, 266, 267, 268 Caucasus xiv, 288. See also Transcaucasus ceiling frescoes. See Quṣayr ʿAmra chancery. See dīwān; dīwān al-rasāʾil Chapel of SS. Elias, Maria and Soreg 190, 191, 194, 199, 204 Chapel of the Theotokos. See Church of the Theotokos Cherchel 200 Christ 10–13, 17, 19, 22, 45, 124n85, 126n120, 261 Christian missionary work 135. See also monastic movement Christian property 260, 261, 311, 312 Christianity, Iraqi xvii, 255, 256. See also Church of the East Christology, Dyophysite and Miaphysite 263 Chronicle of 754 306, 308, 311 Chronicle of Seert 258 chroniclers Abbasid 282–3, 287–8 Andalusī 309, 310, 312 Syriac 284 See also historiography Chronography of 354 192, 193, 195, 196, 197, 198, 199, 202, 203, 205 Church of Aght’amar 100 Church of the Deacon Thomas, Jordan 228–9, 228 Church of the East 255, 256, 258, 263, 264, 265, 268 historiography of 266
8 3
338
Index
Church of the Holy Apostles, Madaba 187 Church of the Holy Martyrs, Mount Nebo 229 Church of Kaianus, Uyun Musa. See Lower Church of Kaianus, Uyun Musa Church of the Lions, Umm al-Rasas 215, 224, 228, 229, 230 Church of Moses, Mount Nebo 211 Church of St. Elianus, Madaba 229 Church of St. George, Khirbat al-Samra 222 Church of St. John, Damascus 312 Church of St. Mary, Rihab 211 Church of St. Menas, Rihab 211 Church of St. Paul, Rihab 211 Church of St. Peter, near Jerash 187, 188 Church of St. Stephen, Umm al-Rasas 187, 211, 213, 215, 218 Church of St. Vincent, Cordoba 312 Church of the Theotokos 215, 217, 224 Church of the Virgin Mary, Madaba 224, 225 civil war xiv, 22, 330–1. See also fitna clients 310, 311, 312. See also mawālī; walāʾ clothing 91, 157, 160, 161–2, 164–5, 167, 193, 195, 202, 270n32 Codex-Calendar of 354. See Chronography of 354 coinage Abbasid copper 141 ʿAbd al-Malik’s reform of (see ʿAbd al-Malik) Arab-Sasanian 43 Iberian Peninsula 308 iconography of Islamic 40, 167–8 “miḥrāb and ʿanaza” type 21 minted by Christian bishops 258 “post reform” epigraphic type 22 Roman 20–2, 25 Sasanian 20–1, 25, 40 “standing caliph” type 20, 168
Umayyad copper 21, 140, 153 commander of the faithful. See amīr al-muʾminīn conquests xiv, xvi, 20, 259, 261, 268, 301, 302, 309–10, 331 narratives 311 (see also historiography) North Africa 303 See also al-Andalus; Iran; Iraq; Mūsā ibn Nuṣayr; Syria Constans II 12, 13 Constantine (I) 6, 10, 11, 13, 185, 193 arch of 88, 208 Constantinople 7, 11, 14, 19, 22, 23, 43, 44, 100, 193, 279 attempted conquest of 112, 279, 288–9 currency circulating in 22, 23 great palace at 103 consuls 193, 195 conversion to Christianity 258, 259, 260, 261, 265, 270n48 to Islam 257, 261–2, 301, 303, 310, 313n6 copyists handwriting style of 74–5 training of 77 Cordoba 302, 309, 311, 312, 313 Great Mosque of 312 Corippus 11 Cosmas Indicopleustes 192 craftsmen Byzantine 44, 53, 111 Christian 85 practices of 178, 179, 205, 226, 231–2, 233, 234, 236n20 at Qaṣr al-Ḥayr al-Gharbī 85, 101, 111, 114 repertoires and conventions of xvii, 178, 203, 206, 231, 233 Sasanian 85 See also artists; copyists; mosaicists; painters; sculptors
9 3
Index Crone, Patricia 4–9, 10, 26, 28, 275 crowns 11, 12, 40–4, 41, 42, 46–7, 119, 269n19, Pl.2 Ctesiphon 116, 163, 255, 256, 266, 267 assembly of 256 Cyrrhus 138, 139 Damascus 46, 86, 139, 140, 267, 275, 282, 285, 301, 304, 305, 307, 308, 311, 312, 313, 327, 330. See also Great Mosque of Damascus David, King 7, 11, 12, 13, 18, 23 daʿwa 288 Dayr Ḥanīnāʾ 140 Dayr al-Zaʿfarān 47 deputy of God. See God's caliph Dibsi al-Faraj 226, 227 dihqāns 257, 258, 261, 262, 265 al-Dīnawarī 255 dīwān 76, 278 dīwān al-rasāʾil 4 al-Djem 187, 188, 189, 206 Dome of the Rock xvii, 7, 23, 39, 40– 50, 41, 42, 57, 59, 75, 111, 112, 113, 117, 119, 282 mosaics 41, 42 Donner, Fred 6–7, 27, 278, 280 Dura Europos 108, 109, 162, 186–7, 187 Egypt xiv, xvi, xviii, 53, 262, 304, 305, 325, 326, 331 ekphrasis 217 Elahbel, tomb of 185 Ende, Werner 322 Euphrates 23, 24, 51, 108, 116, 139, 141–2, 169, 186, 283 Eusebius 10 Fahd, ʿĀbid 329, 331 falconry depiction of 155, 158, 159–61, 161, 169, Pl.10 treatise on 163
339
al-Farazdaq 55, 56, 208 Fars 258, 268 Fierro, Maribel 302, 312 fitna 262, 279, 280, 281. See also civil war Fowden, Garth 13–16, 178, 181–2 France. See Merovingian France, Umayyad raiding of frescoes at Andarīn 134, 136 at Qaṣr al-Ḥayr al-Gharbī 85, 89, 103–11, 104–8, 110, 117–18, 119 at Quṣayr ʿAmra 14–16, 14, 179–209, 182, 183, 184, 198, 207, 208, Pl.13 frontiers 134, 256, 301–2, 305, 306–9, 310, 313 Fustat 52 Codex of (see Qurʾanic manuscripts; Qurʾanic scripts) Gabriel of al-Qōsh 264, 265 Gadara 209, 212, 222. See also Hammat Gader Gaea 107, 108, 110, 196, 197, 206, 207, 208, 234 Ge, Gea. See Gaea Geary, Patrick 276, 282 Gemini 197, 202 Genesis 51 story of Adam in 16–18 George, St. 203 Ghassān, son of Muṭarrif 138 Ghassanids. See Banū Ghassān Ghiṭrīf ibn Qudāma al-Ghassānī 163 al-Ghuzūlī 232–3 Gibb, Hamilton 116 Gibraltar 304 Gīwargīs I (Catholicos) 267 God’s caliph 3–28, 48, 56, 208, 209 ʿAbd al-Malik and 20–25 on coinage 20–3 Crone and Hinds on 5–6
4 03
340
Index
God’s caliph (cont.) and law 24–5, 26 modern historians’ interpretations of 5–7 parallels with Roman titles 8–9, 10–13 al-Walīd II and 4–5, 13–19 Gog and Magog 289 Gōlai 260 Grabar, Oleg xixn6, 45, 147, 176, 180–1 graffiti 136, 142n8, 157, 162, 240n119, 136 Great Mosque of Damascus xvii, 39, 40, 52–7, 52, 54, 59, 101, 113, 119, 282 grids, in composition, 75, 183–92, 204–5 Guarrazar 43–44 Hagia Sophia 44 hagiography 255, 256, 258–68, 326 al-Ḥajjāj ibn Yūsuf 76–77, 267, 268, 307 Jurji Zaydan on 323, 324 modern Sunni Arab treatment of 324–5 in television serials 321, 325, 326–332 al-Ḥakam II 312 Hakim, Avraham 6, 7 Hamilton, Robert 98, 176, 177, 210, 215 Hammat Gader 27, 209. See also Gadara Ḥassān ibn Thābit 28, 29n5 Ḥawīrtah, floor mosaic of 14, 15, 16 headgear 91, 106, 158, 160, 161, 162–3, 165. See also clothing; crowns heaven and earth, in cosmology 4, 50, 51, 57, 60 hell 55 Heraclius 10, 11, 12, 257, 275 Hijazi script. See Qurʾanic scripts Hillenbrand, Robert xviiin4, 177, 221 Hinds, Martin 4–8, 26, 28 Hishām ibn ʿAbd al-Malik xvii, 24, 86, 108, 110, 112, 116, 117, 119, 140, 142, 163, 169, 171, 278, 311 accession of 142
inscription mentioning 111 letter to, on marble 138 miserliness of 105 as patron 112–114, 115–116, 148, 169, 176 possible representations of 167 Hishām ibn al-Ghāz 279 Hishām ibn al-Kalbī 50 historiography Abbasid xvii, 277, 280, 283, 284 Abbasid representations of Umayyads 282–3, 284, 285–6, 302 Andalusī xvii, 304–5, 309, 310, 312 chronologies and lists 279–80 non-Muslim sources 283–4, 287–9, 290 problems of xiv–xv, 176–7, 177, 276, 277, 284, 289 representation of conquest of Iberian peninsula 302, 309, 310 Syriac xvii, 17, 19, 27, 46, 141–2, 279, 283, 284–5, 288, 289 twentieth century writers on Umayyads 322 Umayyad 277, 278, 279, 280 See also chroniclers; hagiography Ḥnānīshōʿ I (Catholicos) 267 Holy Sepulchre 44 Ḥoms 139 horses 90, 106, 155–6, 157, 160, 161, 163, 199, 203, 232, 326, Pl.10 Hoyland, Robert xvii, 6, 7, 27 al-Ḥumayma 162, 163, 167, 169, Pl.11–12 hunting 105, 109, 159, 160, 163, 169, 176, 180, 181–2, 197, 198, 206, 228, 232. See also falconry al-Ḥurr 311 al-Ḥusayn (ibn ʿAli) 285, 293n63 Hussein, Saddam 324 hyparchos 10–11, 19. See also titulature
341
Index Iberian peninsula. See al-Andalus; historiography; Ibn al-Qūṭiyya Ibn ʿAbd al-Ḥakam 303, 304, 305 Ibn al-Athīr 305 Ibn Ḥabīb 304, 305, 306, 308, 310, 315n32 Ibn Hishām 291n17 Ibn al-Muqaffaʿ 25, 27 Ibn al-Murajjā 51 Ibn al-Nadīm 291n11 Ibn Qutayba 286 Ibn al-Qūṭiyya 311, 312 Ibn Shākir al-Kutubī 54 Ibn Zabāla 53, 54 Ibn al-Zubayr 7, 22, 23, 24, 46, 278, 281, 324 iconoclasm 110, 112 India, northern 307. See also Sind inscriptions xv, xvii, 27, 75, 111, 113n98, 134, 135, 136, 137, 148, 181, 189, 209, 222, 234, Pl.7–8 bilingual, Chinese and Syriac 258 bilingual, Greek and Syriac 135 calligraphic style of 75 Greek 136 on marble xvii, 137, 138, Pl.7–8 See also coinage; Dome of the Rock; graffiti; Great Mosque of Damascus Iran xiv Abbasid Revolution in xiv, 275 art and architecture 42, 45, 87, 101, 112, 115, 116, 160, 162 historiography 275 numismatics 23, 40 See also Sasanian: Empire Iraq 23, 24, 115, 255, 258, 262, 265, 266, 267, 275, 280, 307, 324 Abbasid xvi Abbasid links with 283 American invasion of 331 Arabian settlers in 19
341
coinage of 40 congregational mosques of 24 hagiographies of 280 Islamic invasion of 257 northern 262, 263 Sasanian xiv, 85, 116 southern 259, 261, 266 See also Mesopotamia Īshōʿyahb III 266, 267 Īshōʿzkhā 260, 261 Islamic art, in scholarship 178–9, 180–2 ivory carvings 162, 167, 169, Pl.11–12 Jabal Says 143n18, 153 Jacob of Beth ʿĀbe 261 Jacobites 263, 264, 265, 266 Jamlichus, tomb of 185 Jarīr 55, 208 al-Jazīra. See Mesopotamia: northern Jerash 187, 190, 191, 193, 194 calendar of 189, 194, 195, 199, 200, 201, 202, 204, 205 cathedral of 212, 222 Jericho 176 Jerusalem 17, 23. See also Dome of the Rock; Holy Sepulchre Jesus. See Christ jewelry 43 jewels 40–7 Jews 62n38, 210, 266, 324. See also Judaic tradition jihād 309 John of Daylam 258, 268 Judaic tradition 26. See also Jews Justin II 11 Justinian (I) 10, 11, 12, 43, 61n14, 134, 212, 230 Justinian II 7, 22 Kalb (tribe) 282 Karbalāʾ 285, 293n63, 322–3
4 2 3
342
Index
kastron 134, 136, 142n9. See also barracks Kathisma church, Palestine 224, 225 Khālid ibn Maʿdān al-Kalāʿī al-Ḥimṣī 278, 289 Khālid al-Qaṣrī 116, 123n78 khalīfa, in the Qurʾan and pre-Islamic texts 18, 27. See also God’s caliph Khalīfa ibn Khayyāṭ 286 khalīfat Allah. See God’s caliph khalīfat rasūl Allah 4, 5, 6, 26, 28, 29n12 khān 86, 111, 124n98 al-Khanasir. See Khanāṣira Khanāṣira xvii, 327 inscriptions of xvii, 134 late Roman period 133–5 Umayyad period 136–7 Khirbat al-Mafjar xvii, 98, 101, 103, 109, 111, 113, 115, 117, 159, 167, 175, 180, 233 apodyterium of 218 apses of 210, 215, 216, 217, 219, 224, 226 mosaics of 210, 212, 213, 221, 223, 224, 226, 227, 230, 231 Oleg Grabar on 180 plan of 176, 210, 211 stucco sculptures of 98, 157, 159, 161, 163, 167, 168 text on marble from 138 See also mosaics Khirbat al-Mukhayyāt 200, 203 khums 305, 307. See also conquests Khurasan xiv, 24, 115, 283, 330 Khusrau II 256, 257 khvarnah 41, 61n10 Kiev 100 Kish, Sasanian palace of 42 Landau-Tasseron, Ella 6, 7, 27 law xv, 5, 24–5, 26, 31n53, 257, 324, 329 al-Layth ibn al-Dhiyāl 138, 141
Lebanon 55, 191, 228 Leo the Isaurian 110, 112 Libra 202, 203 lieux de mémoire 282, 294n77 Life of Maranʿammeh 261 Life of Sabrīshōʿ 256 Loukianos, tomb of 186 Lower Church of Kaianus, Uyun Musa, mosaics of 212 Madaba 187, 224, 225, 229, 230 madina (Arabic term) 148, 153, 169 Magians 259. See also Zoroastrianism; Zoroastrians al-Mahdī (Abbasid Caliph) 163, 284 Manicheans 258 al-Mansūr 117, 286, 326 Mar Mattai, monastery 263 Maranʿammeh 261, 262, 263. See also Life of Maranʿammeh marble 47, 52, 88, 101, 103, 118, 134, 137, 138, 140, 209, 215, 217–8, 220, 221, 222, Pl.7–8 Mārī ibn Sulaymān 271n59 marzbān 259 Maslama ibn ʿAbd al-Malik 279, 288–9, 290 Maslama ibn Muḥārib 280, 289 al-Masʿūdī 286 mawālī 258, 301, 304, 322, 323, 324. See also clients mawlā. See mawālī Maximus the Confessor 12–13 Mecca 22, 23, 46, 52, 312, 329 Medina 52, 53, 57, 280, 327, 328, 329 Mediterranean Sea xiv, xvi, 23, 85, 89, 105, 115, 175, 185, 189, 209 Merovingian France, Umayyad raiding of xiii Mesopotamia 172n13 lower 116 northern 47
4 3
Index Midrash Rabbah 17–18 miḥna 281 minbar 56, 57, 312 Miskīn al-Dārimī 29n5 monasteries funding of 257, 258, 260 patrons of 259–61 See also Bawit (monastery); Mar Mattai, monastery; Monastery of Lady Mary; Siyagha, monastery Monastery of Lady Mary 201 monastic movement (Church of the East) 258, 259, 260–1, 263, 265 Monza 43 mosaicists 43, 53–4, 229, 231, 312 mosaics 12, 14, 15, 16, 40–8, 41, 42, 43, 50, 51, 52–6, 52, 54, 57–60, 75, 101, 103, 109, 110, 118, 134, 176, 178, 187–93, 196, 198, 200, 201, 202, 203, 204, 206, 208, 209–31, 212–19, 221, 223–30, Pl.1 Mshatta 111, 117 Muʿāwiya (I) ibn Abī Sufyān xiii, xiv, 29n5, 32n73, 112, 143n19, 280 attestation of titles in documentary evidence 8, 27 inscription in bathhouse at Hammat Gader 27, 209 interest in history 278 Muʿāwiya (II) ibn Ḥishām 138, 139–40, Pl.7–8 Muhallabids 270n50, 307 Muḥammad (Prophet) xiii, xiv, 3, 5, 19, 22, 51, 278, 287, 306 Muhammad al-ʿAziziyya 329 Muḥammad ibn al-Qāsim 307 Muḥammad ibn Zakariyyā al-Rāzī 232, 233 Mujāhid 50 mukūs 139. See also taxes mulk 327 Munich 199
343
al-Muqaddasī 54, 209 Mūsā ibn Nuṣayr establishing of walāʾ with Berbers 303 heritage of 310 invasion of Iberia and consequences 301, 304, 305, 307–8, 310, 311 and succession conspiracy 307 television serial on 326 musalsal. See television serials musalsalāt. See television serials muṣḥaf 76–78, 79n35. See also Qurʾanic manuscripts al-Nābigha al-Dhubyāni 24, 28 al-Nābigha al-Shaybānī 55, 57 Nahr Abī Fuṭrus, massacre of Umayyad family at 285. See also historiography Naqsh-i Rustam 163 “Nestorian” Church. See Church of the East Nishapur, paintings of 160, 161 Niẓāmābād 163 nomads. See bedouin; pastoralists Noth, Albrecht 280 novels, historical 322, 323, 331 Old Diakonikon, Mount Nebo 215, 216, 228 Old Testament 50. See also Bible orthography. See Qurʾanic manuscripts Oxus 51, 89 painters 16, 199, 203, 205, 233, 237n42. See also artists; craftsmen Palestine (jund) 139 Palestine (modern) 176, 178 Palmyra 86, 88, 115, 118, 142, 147, 185 Parthenon 89
43
344
Index
pastoralists 133. See also bedouin patriarchs in Bible and Qurʾan 12, 13, 17, 19, 23 Roman Christian 256 schools of, in Church of the East 267 Philip II (of Macedon) 289 Philo of Alexandria 17 poetry 5, 28, 31n53, 32n60, 112, 177, 181, 182, 208, 285, 306, 323 panegyrics 10, 11, 13, 23–4, 25, 28, 48, 55, 56, 208–9 polysemy of 48, 52 polysemy. See poetry; Qurʾan, text pottery 137, 153 precious objects. See jewels; jewelry princely cycle 100–101, 176, 179, 182, 234, 236n25 Oleg Grabar on 176, 180–1 Procopius Church, Jerash 212 proto-Shiites 285 Ptolemy manuscript 203 Qabr Hiram 191 qaṣīda 181, 182. See also poetry Qaṣr al-Ḥayr al-Gharbī 83–119, 86, 159, 180, Pl.3–6 classicizing character of 84–5 east and west, theme of 111, 117–19 frescoes 89, 103–11, 104–10, 117–18, 119, 159, 160, 163, Pl.6 “Byzantine” northern floor fresco 107, 108, 109–11, 110, 117 “Sasanian” southern floor fresco 105–6, 109–11, 117, 160 gateway, decoration of 84, 87–9, 87–8, 98, 99, 100, 118–19 iconographic program of xvii, 111, 112–15, 115–18, 119 lunettes of 100–103, 102 Oleg Grabar on 180 patronage of 85, 113–14
plan of 86 portico sculptures of 100, 167 princely cycle of 100, 180 sculptures animal 90, 91, Pl.3 human 91– 8, 92–7, 99–100, 167 Pl.4–5 rulers 99–101, 99–100, 167 vegetal 84, 98 stucco panels 88, 90 See also Hishām ibn ʿAbd al-Malik Qaṣr al-Ḥallābāt 235n17 Qaṣr al-Ḥayr al-Sharqī 87, 142, 147–71, 147, 154, Pl.9–10 “Building E” 150, 152, 161, 170 as caliphal residence 169–70 date of “Building E” 153 plan of “Building E” 147–54, 150, 171 inscription of 124n98 location near Palmyra 147 plan of 149, 154 See also stucco decoration Qays (tribe) 282, 307 qibla 55, 56, 57, 58, 312 Qinnasrin 138, 139, 140 bishop of 209 Qurʾan, text and fertility 48 and Khirbat al-Mafjar 224 paradisiac imagery of 40, 47, 50, 53 and political thought 7, 13, 18–19, 23, 26, 27, 28, 58–9 polysemy in 48–9, 60 and Quṣayr ʿAmra 15, 16, 19, 24, 28 Qurʾanic manuscripts Damascus Codex 70–8 Fustat Codex 49–52, 69–78 illuminations of 40, 49–52, 49, 58–9, 70, 71–2, 71 layout and production 70–7 Sanaa Codex 57–9, 58–9, 69 sura headings 49, 71
4 35
Index Umayyad patronage of xv, xvii, 58–9, 69, 75–8 Qurʾanic scripts 73–78 Hijazi 72, 73, 74, 77 letter forms of 73–5 relationship to Umayyad inscriptions 74–5 Umayyad style of xvii, 73, 75 Quṣayr ʿAmra xvii, 24, 101, 175, 176, 178, 180, 182, Pl.13 calendrical program of 101, 192–3, 198, 205, 233, 234, Pl.13 central vault ceiling frescoes of 179, 182–9, 182–4, 192, Pl.13 enthroned ruler iconography at 13–19, 24, 206–7, 207, 234 frescoes 14–16, 14, 28, 159, 179–209, 182–4, 198, 207–8 Garth Fowden on 181–2 Late Antique cultural paradigm of 233, 234 Oleg Grabar on 13–16, 180–1 organizational patterns of ceiling frescoes 182, 183, 184, Pl.13 Qutayba ibn Muslim 307 Raʿbān al-Awwal 138–9 al-Rabaʿī 57 Rabban bar ʿIdtā 263 Rabban Hōrmīzd 263–5 Rajā ibn Ḥaywa al-Kindī 124–5n99 Ramadan TV drama. See television serials Ravenna, mosaics of 43, 44, 50, 208 Recceswinth, King 44, Pl.2 rivers of paradise 47, 50, 51, 53 rock reliefs, Sasanian 105 Roderic, king of Visigothic Iberia 301, 303, 304, 311 Roman Empire xiii, 6, 8, 9, 25, 255, 263 Khusrau II’s invasion of 256 See also Byzantine Empire
345
Roman painting 183, 185 Rusafa 116, 140, 142, 230 Sabrīshōʿ 259 sacral kingship. See sacral monarchy sacral monarchy xvi, 6, 9, 25 Armenian concept of 11 Biblical archetypes 11, 16–19 Hellenistic concept of 10–13 saddlery 156, 160, 162, 163. See also bridle Sagittarius 202, 203 Saʿīd ibn ʿAbd al-ʿAzīz al-Tanūkhī 279 Saint-Romain-en-Gal 189, 196, 200 saints, Christian. See calendars Salākh 260, 262 Salamis-Constantia 220, 221, 222 Saliba, George xixn10, 278 Sālim (Abū al-ʿAlāʾ), letters of 8 Samarra 242n173, 281 San Vitale 43, 44, 50 Sant’ Apollinare Nuovo 43 Sasanian art and architecture xvii, 14, 40, 41, 42–3, 42, 45, 47, 83, 84, 85, 87, 99, 101, 105–9, 111, 112, 113, 115, 116, 118, 119, 157, 159, 162, 163, 167, 168 ceremonial 234 Christianity 255–7, 258 coinage 20–1, 22, 25–6, 40 conquest/collapse of empire xiv, 257, 258, 259, 262 Empire 255, 257–8 rock reliefs 105 See also crown; Iran; Iraq; shah Ṣawt al-Qāhira 326 scribes. See copyists sculptors 89, 159 sculptures. See Qaṣr al-Ḥayr al-Gharbī seals 139–40, 308, 309 Seasons, the 185, 196, 206, 207, 208, 234 Sebeos (Pseudo-) 11 Sepphoris 202, 207
4 63
346
Index
Sergiopolis. See Rusafa Sergius of Beth Rāstāq 266 Sergius, St., martyrium of 135 Serjilla 226, 227 shah 257, 266, 267 relationship with catholicos 256 shahrēgān 261, 262, 264 Shamṭā bar Yazdīn 257 al-Sharif, Nur 326 Shiʿites 322. See also proto-Shiites Shīrwān 258 silver 41, 106, 118, 159, 163, 168. See also coinage Sinai 12 Sind xiv, 301. See also India, northern Sinngeschichte 276, 291n7 Sisinnios, St. 203 Sistan xiv Siyagha, monastery 212, 214 Ṣlībā-zkhā (Catholicos) 267 Solomon 11, 27, 289. See also Table of Solomon spolia 84, 119 stucco decoration xvii, 42, 154, 155–8, 160–1, 164–72, Pl.3–5, 10 “standing caliph” at Qaṣr al-Ḥayr al-Sharqī 164, 166 See also Qaṣr al-Ḥayr al-Gharbī; sculptors succession, in Umayyad caliphate 24, 25, 302, 306–7 Sulaymān ibn ʿAbd al-Malik 301, 302, 305, 306–7, 308, 313 Sulaymān ibn Hishām 140 Sulaymān ibn Mūsā 278, 289 Sulaymān ibn ʿUbayd 124n98 sura headings. See Qurʾanic manuscripts Syria (Greater) xiii, xvii, 14, 20, 22, 23, 47, 49, 85, 86, 88, 103, 107, 112, 115, 116, 133, 134, 136, 139, 141, 142, 167, 179, 185, 206, 222, 226, 227, 275, 283, 286, 306
churches of 14, 15, 16, 222 conquest xiii, 279 early Abbasid 275, 277, 278, 279 early Islamic 275 historiography of early Islamic 275–7, 278, 280 Justinian’s patronage 230 tribes of xiv, 19, 28, 135 Syria (modern) xvi, xviii, 14, 148, 228, 324. See also television serials Syriac sources. See historiography Syrians (Late Antique/early Islamic) xiv, 15, 19, 46, 77, 163 Syrian-Swiss archaeological mission 147 al-Ṭabarī 50, 256, 276, 286, 289, 304, 307, 308 Table of Solomon 304, 310, 311, 312, 313 Tang Empire xiv Tangiers 303, 304 Taq-i Bustan 14, 45, 163 Ṭāriq ibn Ziyād 301, 303, 304, 310, 311, 313, 325 taxes 137–9, 257, 260, 261–2, 268, 309 Taymur, Mahmud 324, 325 television serials containing themes from Jurji Zaydan 322–3, 328, 331 representation of al-Ḥajjāj ibn Yūsuf 321, 329–31 representation of ʿUmar ibn ʿAbd al-ʿAzīz 321, 326–9 Temple Mount 50, 51 Theodelinda (Lombard queen) 43 Theodora (Empress) 43, 44, 50 Theodoric, palace of 208 Theophilus of Edessa 46, 284, 286 Theotokos Chapel, Mount Nebo 215, 217, 224 Thessaloniki 50
437
Index Thomas of Marga 258, 259, 261, 262, 263, 264, 265, 267 Thuburbo Maius, Tunisia 217, 220 Tiberias 27, 207, 209 Timothy I (Catholicos) 268 titulature on coins 20–3 Roman imperial 10–11, 25 theoretical perspectives on 8–9 See also God’s caliph Toledo 44, 301, 304 tombs 53, 185–6, 288, 327 Transcaucasus 23. See also Caucasus Transoxania 301 Trier, cathedral of 185, 186 Turgesh xiv Turks 281, 288, 289 ʿUbāda ibn Nusayy al-Kindī 278 ʿUbayd ibn Sharya al-Jurhumī 278 ʿUmar (I) ibn al-Khaṭṭāb 6, 124n96, 287 ʿUmar (II) ibn ʿAbd al-ʿAzīz 53, 136, 142, 307 in Arabic historiography 287–8, 289, 290, 306 in the Latin Chronicle of 754 306 See also television serials Umayya ibn ʿAbd Shams xiii Umm al-Rasas 187, 211, 213, 213, 215, 218, 224, 228, 230 ʿUrwa ibn al-Zubayr 278 ʿUthmān ibn ʿAffān xiii, xiv, 28, 282 Uyun Musa 212, 228 Vatican planisphere 202, 203 Visigothic ancestors of Ibn al-Qūṭiyya 311 Iberia 301 kings 301, 303 treasury of Guarrazar 43, Pl.2 Visigoths 43, 44, 303, 311 Vitruvius 231
347
al-Waḍīn ibn ʿAṭāʾ al-Dimashqī 279 walāʾ 303. See also clients; mawālī al-Walīd (I) ibn ʿAbd al-Malik 52, 73, 301, 308, 310, 324 Christian workers, employment of 53 court poets of 55–7, 307 as God’s deputy 56 in historiography (Andalusī and Latin) 306–7, 310–13 imperial mosques of 53 military campaigns during reign of 301 Mūsā ibn Nuṣayr and Ṭāriq ibn Ziyād, role in disputation between 302, 304, 310–11, 313 patron of the Great Mosque of Damascus 52–4 protection of Christian property 310–11 succession, attitude to 306–7 Sulaymān, tensions with 302, 307 See also conquests al-Walīd ibn Ḥammād al-Ramlī 46, 50 al-Walīd ibn Muslim al-Umawī al-Dimashqī 279 al-Walīd (II) ibn Yazīd 3, 13, 142, 163, 177, 226 and “God’s caliph” 9, 13, 27 in historiography 177–8 letters of 3–5, 24–5 love of wine, debauchery 177 as patron of Khirbat al-Mafjar 175, 176, 178 as patron of Quṣayr ʿAmra 13–19, 24, 25, 27, 175–6, 181, 234 reappraisal of 177, 234 al-Wāsiṭī 46 Watt, William Montgomery 5 Weber, Max 285 Wellhausen, Julius 280 wine 177, 200, 262 Witiza (King) 303, 311
4 83
348
Index
Yaman (confederation) 307 al-Yaʿqūbī 304, 305 Yasilah, basilica 212 Yazīd (II) ibn ʿAbd al-Malik 110, 112 Yazīd ibn al-Muhallab 307 Yazīd ibn Sallām 124–5n99 Yazīd (III) ibn al-Walīd 32n62 Yūsuf ibn ʿUmar 116 Zab 262, 282 Zabad. See Zebed Zāmāsp-Nāmag 257 Zāmyād yasht 41 Zay, church 222
Zaydan, Jurji 322–4, 331 Zebed, inscription 135 Zenobiyya, bath complex 209 Ziyād ibn Abīhi 280 Zoroastrian apocalyptic text 257 “Church” 260 practices of shahrēgān 262 protective formulae 268n4 villages 258, 260 Zoroastrianism 257, 269n16 Zoroastrians 260. See also Magians Zubayrids 23, 281. See also Ibn al-Zubayr al-Zuhrī 278, 279, 280