The archaeology of medieval Islamic frontiers from the Mediterranean to the Caspian Sea 9781607328773, 1607328771, 9781607328780, 160732878X

Citation preview

THE ARCHAEOLOGY OF MEDIEVAL ISLAMIC FRONTIERS

THE ARCHAEOLOGY OF MEDIEVAL ISLAMIC FRONTIERS From the Mediterranean to the Caspian Sea

ED I T ED B Y

A. Asa Eger

UNIVERSIT Y PRESS OF COLORADO

Louisville

© 2019 by University Press of Colorado Published by University Press of Colorado 245 Century Circle, Suite 202 Louisville, Colorado 80027 All rights reserved Printed in the United States of America



The University Press of Colorado is a proud member of the Association of University Presses.

The University Press of Colorado is a cooperative publishing enterprise supported, in part, by Adams State University, Colorado State University, Fort Lewis College, Metropolitan State University of Denver, University of Colorado, University of Northern Colorado, Utah State University, and Western State Colorado University. ∞ This paper meets the requirements of the ANSI/NISO Z39.48‐1992 (Permanence of Paper). ISBN: 978‐1‐60732-­878-­0 (cloth) ISBN: 978‐1‐60732-­877-­3 (ebook) DOI: https://​doi​.org/​10​.5876/​9781607328773 Library of Congress Cataloging‐in‐Publication Data Names: Eger, A. Asa, editor. Title: The archaeology of medieval Islamic frontiers from the Mediterranean to the Caspian Sea / edited by A. Asa Eger. Description: Louisville : University Press of Colorado, [2019] | Includes bibliographical references and index. Identifiers: LCCN 2019000682 | ISBN 9781607328780 (cloth) | ISBN 9781607328773 (ebook) Subjects: LCSH: Islamic Empire—­Antiquities. | Islamic Empire—­Boundaries. | Borderlands—­Islamic Empire. Classification: LCC DS38.3 .A74 2019 | DDC 909/.097492707—­dc23 LC record available at https://​lccn​.loc​.gov/​2019000682 The University Press of Colorado gratefully acknowledges the generous support of the University of North Carolina Greensboro toward the publication of this book. Cover photograph by the author.

Contents

1 The Archaeology of Medieval Islamic Frontiers: An Introduction A. Asa Eger  3 Part I. The Western Frontiers: The Maghrib and The Mediterranean Sea 2 Ibāḍī Boundaries and Defense in the Jabal Nafūsa (Libya) Anthony J. Lauricella  31 3 Guarding a Well-­Ordered Space on a Mediterranean Island Renata Holod and Tarek Kahlaoui  47 4 Conceptualizing the Islamic-­Byzantine Maritime Frontier Ian Randall  80 Part II. The Southern Frontiers: Egypt and Nubia 5 Monetization across the Nubian Border: A Hypothetical Model Giovanni R. Ruffini  105 6 The Land of Ṭarī’ and Some New Thoughts on Its Location Jana Eger  119

Part III. The Eastern Frontiers: The Caucasus and Central Asia 7 Overlapping Social and Political Boundaries: Borders of the Sasanian Empire and the Muslim Caliphate in the Caucasus Karim Alizadeh  139 8 Buddhism on the Shores of the Black Sea: The North Caucasus Frontier between the Muslims, Byzantines, and Khazars Tasha Vorderstrasse  168 9 Making Worlds at the Edge of Everywhere: Politics of Place in Medieval Armenia Kathryn J. Franklin  195 About the Authors  225 Index  229

vi

Contents

THE ARCHAEOLOGY OF MEDIEVAL ISLAMIC FRONTIERS

1 In the last decade, archaeologists have increasingly focused their attention on the frontiers of the Islamic world, partly as a response to the political conflicts in central Middle Eastern lands. In response to this trend, a session on “Islamic Frontiers and Borders in the Near East and Mediterranean” was held at the American Schools of Oriental Research (ASOR) Annual Meet­ ings, from 2011 through 2013. The main goal of this session was to initiate comparative and interpretive dialogues between archaeologists who work on different regions and time periods. A second goal was to bring to the foreground the importance of approaching the theoretical concept of the frontier, constructed or deconstructed, as it applies in an archaeological context. Third, the session examined closely both similar and dissonant processes within Islamic frontiers. As the session progressed over the years, what was intended as a series of conceptual frontier types became organized by key geographic borders of Islamic territories with non-­Islamic lands. This volume, based on three consecutive years of talks, is constructed similarly. The chapters all use historically assumed political and religious boundaries as starting points: the western frontier (Mediterranean and Maghrib) from ASOR 2011, the northeastern frontier (Caucasus) in 2012, and the southern frontier (Egypt and Nubia) in 2013 (figure 1.1).1 These frontiers, according to an article by Haug, can all be considered the minor frontiers, the major ones being the northern Byzantine frontier, Iberian

The Archaeology of Medieval Islamic Frontiers An Introduction A. Asa Eger

DOI: 10.5876/9781607328773.c001

3

Figure 1.1. Map of places discussed in this volume (courtesy of Anthony Lauricella, CAMEL, University of Chicago).

Andalusī frontier, and Central Asian/east Iranian frontier.2 These minor frontiers feature less frequently in contemporary primary-­source texts, in part as they were not characterized strongly as zones of conquest or annual campaigning. Furthermore, for these minor frontiers there is no established opposing power of comparable status to the Islamic Caliphate or even sub-­caliphates and dynasties. So too, much work has been done and is currently being conducted on the major frontiers and so these minor frontiers are also minor in terms of modern scholarship. As such, their focus here is deliberate. Several key questions emerged from the panel discussions: how did Islamic political or religious ideology play a role in delimiting real or imagined spaces in the shaping of frontiers? What other forces were integral in characterizing Islamic frontiers? How were these frontiers manifest through archaeological evidence, and how was their development affected over time? Despite the shift in focus within Islamic archaeology, there are significant gaps in scholarship. Frontiers and borders, once seen as divided and contested landscapes delimiting not only political space but ethnicities and religious 4

A . A S A E G ER

groups, have become a category of inquiry by many Western medieval scholars, who see these spaces as varied, complex, and dynamic transitional zones of cultural interaction and ambiguity that can show processes of assimilation, acculturation, or ethnogenesis (the creation of new societies).3 The study of frontiers in the medieval Islamic world is slight; few important recent titles are text based.4 The study of the archaeology of these same frontiers is almost nonexistent. This volume brings the Islamic world into the ongoing dialogue on medieval frontiers. Since most work on medieval Islamic frontiers (generally termed al-­thughūr) has been text based, its focus on archaeological methods gives voice primarily to perspectives that are non-­mainstream, non-­ urban, and non-­orthodox.5 Archaeological research on the frontiers provides evidence for the presence of heterogeneous Islamic and non-­Islamic societies and the complexity of their engagement with one another and with a more central ruling or orthodox authority. Even if the frontiers are not obvious (whether through textual mention or geographic location), archaeology can reveal or test internal frontiers within Islamic society that have not been well understood before. Moreover, the inclusion of archaeology fosters examination of frontiers in terms of those who live within them, and as social spaces and processes.6 The discursive theoretical framework of frontiers to locate analysis provides a way for scholars to explore more precisely the points of interaction/disconnection and conflict/community that more accurately comprised the periods of Islamic rule following the seventh century. CORE-­P ERIPHERY

The standard work on Islamic frontiers is that of Brauer, who established, using mainly the works of medieval Islamic geographers, that there is no concept of the frontier in cartography, no political boundaries in legal texts, and no agreement on what is the thughūr by the geographers; the frontier is a zone rather than a sharply defined border.7 The thughūr often had mixed populations and an active trading economy, with or without military engagement and religious motivation. This premise has been accepted widely, and is reflected, for example, in early work on the Islamic-­ Byzantine frontier, Mamlūk-­ Mongol frontier, and early Ottoman frontier.8 However, Brauer concludes that the frontier is best represented by a core-­periphery model, which establishes that inhabitants of core areas (urban polities, populated homelands) have a hierarchical and structural relationship to inhabitants on the periphery. In most cases, peripheral communities are colonies or outposts of core populations. The periphery provides economical (resources, labor, raw materials, and T he A rchaeology of M edieval I slamic F rontiers

5

basic goods) and territorial (boundary maintenance) support for the core in an unequal exchange; those on the periphery are typically exploited, weak, and poor populations lacking in technological advances, cultural production, and other forms of agency.9 Corollary to this relationship, the core exerts a dominant social, religious, and cultural ideology over its periphery. The core-­periphery model has been challenged by scholars from many disciplines as top-­down, colonialist, and flawed.10 In studies on frontiers using Islamic geographical literature, the division between core and periphery was not universally fixed. The capital was only given prominence and value through the ideologies and myths created by the ruler (or his propagandists) as specific situations and challenges arose.11 This was created when political sovereignty frequently assumed a prior existence or claim to the land by rewriting history, retroactively imposing new boundaries on the past, or perpetuating old boundaries in the present, while at the same time using the concept of a border to contain, uplift, and thereby necessitate their own civilization.12 In Islamic geographies and fadā’il literature (books in praise of certain cities), cities assume primacy as powerful homelands that elicited longing and a sense of ownership, and are described in detail with names, locations, and associated narratives of foundation or conquest.13 Maps showed cities first and foremost, perhaps fueling the often assumed urbanity of Islamic culture and religion. Medieval Islamic maps, following the Persian tradition of organizing the world by climes or regions (the kishvar system), always depicted the central clime, the caliph’s own heartlands as the best and most temperate. Thus, in these literary imaginings of the world, anything beyond the city and its own immediate hinterlands was a place where authority did not extend; that is, it was the frontier. Despite how place was articulated, connectivity and boundedness without religious or political overtones are suggested in the same medieval maps that frequently label the edges with regional names, thus showing how they connect to other maps and a wider world.14 Thus the frontier as peripheral is created by the central state, and is accordingly a matter of perspective. The frontier was also a center, with its own agency and influence. Ellenblum has argued against any real division of lands between Crusader enclaves and the Islamic world, instead maintaining that the intersections were frontiers and centers in their own right, and spheres of influence that competed with one another.15 Khurāsān, located on the northeastern Islamic frontier between Iran and Central Asia, was a frontier province populated with a mix of Arab soldiers, preexisting Persian families, and many other religious or ethnic groups (Manichaeans, Buddhists, Sogdians, Hephthalites). It was precisely the frontier-society blend of religious, ethnic, and linguistic groupings that 6

A . A S A E G ER

proved to be so strong so as to initiate powerful processes of change that rippled back to the heartlands throughout Islamic history.16 It was in Khurāsān that the ‘Abbāsid “Revolution” began, paving the way for their rise to power in 750. One hundred and fifty years later, Khurāsān was the epicenter of a series of breakaway provincial autonomies that eventually were responsible for undoing ‘Abbāsid power. A third example is provided by the southeastern frontier with India. Originally seen as a one-­way dominant Islamic cultural influence over its southeastern neighbor, this region has recently been recast as a place that was certainly influenced from central Islamic lands, but that generated new systems of meaning through rich cultural production that rippled back to Baghdād.17 Ellenblum also shows how models of core-­periphery were constantly being dynamic and deliberately being altered, for example, with the construction of new castles to shift not only administrative, political, and military power away from existing cities or other centers but also to break up economic holds on lands.18 These new “cores,” or centers, were often placed on an old center’s periphery, thus engaging in a continuous reimagining of a core-­ periphery relationship. These arguments align with recent studies on borders and frontiers such as the group of studies edited by Zartman, which show that borders are constantly in flux diachronically and spatially, and are not fixed places. Further, core-­periphery relationships are but one dynamic that can also encompass “relations between neighboring peripheries or by relations within the autonomous periphery.”19 In other words, frontiers can exert power that at various points can exceed that of a traditional “core.” The rise of provincial autonomies in the tenth century on the ‘Abbāsid Empire’s eastern and northern frontiers are a case in point. Since the eighth century, raids past its borders were a projection of central power, a way to keep the enemy off balance, and a dynamic locus of economic ventures, but within a complicated interwoven series of processes—­as ‘Abbāsid central authority waned, local dynasties arose, enacting greater power and influence in the frontiers themselves. These studies move our understanding of Islamic frontiers as imagined and ideological landscapes, not fixed but relatively located, and spaces that exert their own political, social, and cultural capital, changing over time. Yet, the implications of accepting a core-­periphery model as defined primarily by a relationship of core dominance, unequal economic exchange, and lack of technological and cultural production on the periphery is best critiqued with physical evidence on the ground and material culture, a challenge well s­ uited to archaeology. Of what little has been published, most archaeological studies of Islamic frontiers have been on the major ones: the Islamic-­Byzantine/ Syro-­Anatolian frontier and the Islamic-­Christian/Iberian-­Andalusī frontier.20 T he A rchaeology of M edieval I slamic F rontiers

7

Among the earliest, Redford showed how communities in southeast Anatolia in the twelfth to fourteenth centuries carved out autonomous territories among these blended frontier societies through shifts in settlement patterns and localized production of material culture. These territories both contributed to and were a result of a less stable landscape and the decentralization of central ‘Abbāsid authority.21 My own work on this same region, though chronologically earlier, shows that the Islamic-­Byzantine frontier, or thughūr and ‘awāṣim provinces, developed from the seventh to tenth centuries as an important agricultural and commercial region inhabited by coexisting diverse communities. This argument is supported by archaeological evidence, mainly recent, from surveys and excavations that provide evidence for non-­urban settlement types, including rural sites and waystations and their land-­use initiatives, urban settlement, and both locally produced and imported commodities. Further, this evidence gives insight into the life, interactions, and exchanges of mainly non-­urban and less literate groups that inhabited the frontier. On the one hand, this was an external frontier between Islamic and Byzantine lands dictated by seasonal transhumance and competition for resources; on the other, it was an internal frontier between the central state and peripheral frontier societies containing a mix of heterodox Muslims, Christians, rebels, insurgents, and independent warlords. The frontier was also a religious one, appearing in the pages of manuals of jihād and apocalyptic narratives that created an imagined barrier. Inhabitants of these multivalent frontiers were agents of their own space, as frontier towns, villages, monasteries, and waystations interacted with one another economically and culturally. This zone transformed at its own pace, not directed by or synchronized with the fate of the central state.22 Using as a point of departure Brauer’s core-­periphery model of frontiers and the subsequent critiques it stimulated, this volume significantly advances our understanding of Islamic frontiers both by viewing them through the lens of archaeology and by expanding them geographically to include more frontiers. In nearly every case, the contributions in this volume on the “minor” frontiers of the Islamic world deconstruct historically assumed frontiers, focusing rather on the interaction between differently perceived religious and ethnic groups. Also in every case, there are no physical frontiers built in the Islamic period (such as walls). This is unlike the Sasanian Empire, where we actually have texts describing built frontiers and archaeological evidence of walls between the Sasanians and non-­Persian/nomadic societies in the fifth and sixth centuries, such as the Gurgān Wall, the Darband Wall (Caspian Gates), and the fortifications of Ultan Qalasi and Ören Qala in the Mil Plain, 8

A . A S A E G ER

discussed by Alizadeh in chapter 7. Here, it seems that the Islamic rulers were opportunistic in the maintaining of these walls against ongoing Khazar incursions. However, were these walls defining the limits of empires or were they utilized as a display of imperial power that in actuality projected well past the wall?23 For many of the frontiers in this study, a natural boundary is present, such as the Caucasus Mountains or Nafūsa Range, the Mediterranean Sea, and the Sudanese desert, but this is never an actual boundary for movement and settlement; rather it is secondary to the creation of a complicated frontier. What emerge as common themes of the archaeology of Islamic frontiers is perhaps not so distinct or surprising, but addresses and moves well beyond the unequal exchange posited by the core-­periphery model. First, frontiers were never borders between two entities but were porous zones of interaction and exchange. Second, frontiers were never political arenas between two groups, but were sites of local expression, cultural production, and human agency. Before examining these two themes more closely as they relate to this volume, it is necessary to discuss the methodologies employed. Archaeological methods are well suited to reveal economic interconnections or lack thereof through trade and distribution. However, on the ground, were frontiers prime zones for economic exchange and were they truly dictated by these processes, or is this examination rather tautological in nature? Reframing the question, is the economic frontier synonymous with the archaeological frontier, because of the inherent value of material objects? As these studies show, frontiers were multivalent spaces where many forms of interaction played out. Despite the inherent archaeological bias, economic exchange was undeniably one of the most influential of these. The studies included here show evidence of economic exchange on all levels: not only movement of money and goods, but the building of waystations and caravanserais, the presence of individual merchants, texts detailing exchange, and treaties concerned with the legality of all these exchanges. METHODOLOGIES

The eight studies in this volume investigate three frontier areas. In the western frontier of the Mediterranean Sea and the Maghrib (part I), Renata Holod and Tarek Kahlaoui (chapter 3) present a diachronic archaeology of the small island of Jerba off the Tunisian coast, based on survey evidence from the eighth to eighteenth centuries. Ian Randall in chapter 4 delves into the experience of living on and traveling between Mediterranean islands, complicated spaces floating somewhere between Byzantine and Islamic territories in T he A rchaeology of M edieval I slamic F rontiers

9

the seventh to tenth centuries. Anthony Lauricella examines the Jabal Nafūsa Range in Libya in chapter 2, which differs from the previous studies as it presents an internal frontier within Islamic lands and not with the Mediterranean. As such it has some connection to Egypt, but nevertheless, it is included in the section on the western frontier as it relates strongly with the study of Jerba and heterodox Islamic communities. In the southern frontier (part II), Giovanni Ruffini (chapter 5) hypothesizes a specific model of monetary exchange between Islamic Egypt and Christian Nubia by the twelfth century. Jana Eger in chapter 6 reports on a survey of a monastery in Sudan, dating possibly from the sixth to eighth centuries, on the frontier of Christian Nubia and African kingdoms farther south, yet mentioned in an Islamic geographical text. In the eastern frontier, the three chapters of part III focus on the Caucasus region. Karim Alizadeh (chapter 7) offers a theoretical perspective on the material evidence of borders, utilizing surviving physical Sasanian boundaries in the Islamic period. In chapter 8, Tasha Vorderstrasse reanalyzes the eighth/ninth-­century remains of a burial with Chinese and Buddhist grave goods at the northwest Caucasus site of Moschevaja Balka. Finally, in chapter 9, Kathryn Franklin considers how Armenia was perceived as a frontier by looking at both textual and material evidence in the context of her excavations of the thirteenth-­and fourteenth-­century Arai-­Bazarjuł caravanserai in Armenia. The methodologies employed in these studies vary considerably, but all fit well within an archaeological study of frontiers. At the very core of such research is the problem of sources. As stated at the outset, archaeology of these periods is not robust; frontier archaeology is decidedly even less conspicuous. As such, material evidence needs to be balanced by the textual sources. In some cases, excavations produce texts that are also artifacts. For instance, Ruffini’s study on the monetization of the Egyptian-­Nubian frontier is not well-­enough supported by extant evidence of coins excavated. It is precisely the open economy of wide distribution of money into Nubia from Islamic lands, he argues, that caused coins to quickly scatter and be absorbed into a wide system of exchange. This is why coin hoards are not found in key excavated sites. He analyzes texts excavated at the twelfth-­century site of Qaṣr Ibrim to fill out the picture. Likewise, Vorderstrasse analyzes a preserved fragment of Buddhist text in a grave in the context of a site on the Islamic frontier (with other artifacts bearing Arabic inscriptions). These chapters show how the relationship between artifacts and texts to either or both of the built environment and landscape and the imagined environment is an important methodology for Islamic archaeology. 10

A . A S A E G ER

Inclusion of textual analysis in many of these studies reflects an attempt to begin with a historically assumed frontier as a problem, and to reconcile texts with material evidence. Non-­artifact textual sources often provide only a perception of the frontier from one point of view that is typically retroactive and, in almost every case, stems from the Islamic side of the frontier. Toggling between text and artifact, as studies such as Randall’s, Franklin’s, and Alizadeh’s do, also maintains the importance of considering that the frontier is as much about perception as it is about reality on the ground. The physical evidence is only one side to what a frontier was or how it may have been perceived or felt. While all the studies utilize texts to varying degrees, the archaeological method is varied. Lauricella employs an analysis of place and placement of settlement rather than material culture, augmented by GIS studies and viewshed analysis to articulate the relationships between settlements. Jana Eger undertook an excavation of the church/monastery, relied on remote sensing to establish its connectivity and relative isolation, and used Islamic texts to aid in identifying the site and situating it in a wider context. Holod and Kahlaoui conducted surveys on Jerba and augment their interpretation, particularly of later historical periods, with textual accounts. Vorderstrasse uses “legacy” archaeological data, reanalyzing the site of Moschevaja Balka of the northwest Caucasus in Russia. FRONTIERS AS ECONOMIC ZONES

From a textual perspective, frontiers were frequently spaces created to reflect an administrative division of space—­that is, a separate province or district. This was the case in the southern frontiers of Egypt with the three regions of Nubia, Makuria, and Alwa. The Islamic-­Byzantine thughūr/‘awāṣim division, established in the early ‘Abbāsid period, also shows the same administrative redistricting on the part of the central state. Yet, in reality, these sub-­frontiers were likely not perceived on the ground. Confusion also is evident from sources, as not all authors acknowledge the division or establish clearly which settlement or place name belongs where. In all likelihood, frontiers were divided for taxation and other economic purposes. While disguising economic frontiers as territorial ones is plausible administratively, it is precisely archaeological analysis that reveals a much more complicated economic reality. Furthermore, “international” trade between regions was rarely hampered even when those regions were in conflict. Concerning the same infamous Islamic-­ Byzantine frontier, which by 1200 CE had become an arena of constant and T he A rchaeology of M edieval I slamic F rontiers

11

complex shifting wars and alliances between different Crusader factions, different Muslim dynasties, Armenians, and Byzantines, merchants seemed to carry out business as usual. The traveler Ibn Jubayr (1145–­1216) in his Rihla, captures this dynamic, albeit by simplifying the players: “one of the astonishing things that is talked of, is that though the fires of discord burn between the two parties, Muslim and Christian, two armies of them may meet and disperse themselves in battle array, and yet Muslim and Christian travelers (rifāq) will come and go between them without interferences.”24 Indeed, his own caravan, moving through Muslim and Crusader lands, is a testament to this fact. In this volume, the southern frontiers with Nubia best show a frontier of economy, namely trade and business between these Islamic and Christian lands. It is this process of exchange that actually changes the nature of the frontier. Ruffini advances this complicated frontier further with analysis of texts and goods in twelfth-­century Qasr Ibrim. He shows how the Nubians were active agents in aligning their monetary system with the more dominant one of Islamic Egypt to the north. Nubians, in seeking to attain status and prestige at the border, help create an open-­flow system in which gold and silver coming from Fatimid Egypt was not held but distributed to the countryside, and was returned as materials, namely wine, slaves, and exotica from Africa, like ivory. Goods from Islamic lands not mentioned in texts—­such as the glazed wares (often the silent markers of trade that go unnoticed in texts) that spread wide in Nubia—­show a richer trade and consumption of taste. All this points to a non-­physical frontier between Muslims and Christians that was quite interconnected and permeable. So much so, in fact, that evidence points to exchange and travel between Nubia and the Mediterranean world, including Italy.25 Southern Christian Nubia’s involvement in international trade essentially explodes the frontier outward with wider economic implications. These frontiers connect with one another. Thus, McCormick’s seminal Origins of the European Economy, which shows a Mediterranean crisscrossed with people (pilgrims, slaves, pirates, travelers, merchants) and goods (for trade and for worship, as in the case of relics) is part of an even wider stage.26 Jerba and the other Mediterranean islands in Holod and Kahlaoui’s and Randall’s studies are at the center of this density of Mediterranean traffic. Jerba, for example, frequently appears in itineraries for grain shipments to Sicily, nearly 600 km away. Vorderstrasse’s study shows more evidence than just local or regional economic exchange in the eastern part of the Islamic world. Goods arrived to the Caucasus from China and Central Asia as well as Byzantine and Islamic lands. Darband, as Alizadeh writes, often described as a fortress with a long-­fortified Sasanian wall between Islamic and Khazar lands, was 12

A . A S A E G ER

frequently conquered. However, its lower town became an increasingly important manufacturing and economic center. Perhaps these Sasanian walls, reused in the Islamic period, were used to traffic the movement of goods and people and control nomadic and sedentary interactions, much as has been argued for the Great Wall of China27 Can one speak of an early medieval globalism or an interconnected world with numerous points of contact and departure? Certainly, such evidence would be presumed on the Silk Road and can be extrapolated to other such routes, including the Gold Road in West Africa and the perhaps less noble Slave Road in West and East Africa. Yet putting aside assumptions about larger premodern economic systems, the actual physical evidence of such long-­distance international trade must not be ignored. This is not least because of what these goods suggest. The physical evidence stands as a marker in place of human presence and interaction. It is important to remember that people moved these goods around: local tribes (like the Banū Kanz) that facilitated exchange, merchants (like Muslim traders in Nubian cities from tombstones and the Chinese merchant buried in the Caucasus) that traversed distances, and elites that distributed money and goods locally. CULTURAL EXCHANGES

The frontier was not, of course, neatly balanced between both sides ad infinitum but was a dynamic space of interaction between groups, which transformed over time. Ruffini argues that the process of monetary exchange and mutual partnership led to a power differential, as Islamic Egypt became more prosperous than Nubia and Islam spread southward as a byproduct (as the late arrival of mosques may suggest). In Jana Eger’s study, Ṭarī’, between Nubia and African lands farther south, was a border region and contact zone. Although not a border with Islam, it is mentioned in Islamic guidebooks by its distance from Cairo (al-­Qāhirah). One interpretation for the isolated church/ monastery was its economic role in controlling nomads and facilitating tolls and customs, perhaps as an outpost for Egyptian caravans. Though far from the Nile, Jana Eger’s excavations of graves show burial practices similar to those of the Nile valley and Nubia, providing evidence for sociocultural links among inhabitants of the site. Franklin’s analysis of her own excavated caravanserai in Armenia provides little ambiguity for these structures as standing evidence for economic frontiers, at once connected to the world and reflective of a mixing of international style in art and architecture. Vorderstrasse’s study reassessing a Chinese burial in the Caucasus also shows sociocultural links on several levels. The discovery of Buddhist texts speaks to T he A rchaeology of M edieval I slamic F rontiers

13

a tangible evidence for an archaeology of ideas on the border. Religious ideas spread regardless of borders, and the infiltration of non-­Islamic thought into Islamic lands can be considered as equally as the spread of Islam into “infidel” lands. Further, the Buddhist text parallels several narratives in Persian literature, reinforcing a more complex cultural mixing that resulted from the flow of ideas beyond religious propaganda or the physical movement of the objects of religious use themselves. These economic processes are powerful, inextricably linked to the passing of social, cultural, and religious information. They point to varying forms of contact across the frontier, and ultimately demonstrate the fluidity of frontiers and their relativity to one another. FRONTIERS AS INTERNAL MIRRORS

The frontiers presented in this volume are all necessarily frontiers of boundary with the outside world. All of them are located on peripheries: Jerba protrudes into the western Mediterranean as a mixed zone of interaction among groups vying for control, merchants traveling on trade routes and exchanging slaves, and pilgrims journeying to Jerusalem, Mecca, and Medina. Jerba also has been the site of literary or propagandistic venues for Christian-­Islamic jihād based on conflict between the Christian kingdoms of Iberia and the Normans. Similarly, the Jabal Nafūsa, near the coast, was also at the edges of the Islamic world and embroiled in conflict, first as a site of Berber/Arab discord. The most salient characteristics of these case studies are not the traditional frontiers one assumes in negotiating space in the Islamic world, namely Islam’s frontiers with its outside neighbors. Rather, the studies reveal the various internal workings of (and divisions within) the Islamic Empire, a plurality of frontiers, and more complicated processes of settlement and interaction. Mullin has discussed how archaeology as a discipline is frequently concerned with ideas of boundedness, whether looking for discrete, spatially articulated units or measuring temporally and culturally defined units. The former makes an archaeological study of frontiers compelling and relevant. The latter raises the question of ethnicity. Both intersect on the subject of territoriality and control. Do archaeologists prefer one ethnic group over another in categorizing cultures? Or do they conflate large cultural groups with specific territories?28 Stein argued that this concept of a monocultural “horizon,” a concept often employed by archaeologists, is in reality a thin veneer covering regions with a variety of differing communities who shared some characteristics.29 More useful is an examination of these regions as local manifestations that differed significantly from one another and were, according to Mullin, “free-­flowing, 14

A . A S A E G ER

heterogeneous, and flexible.”30 This perspective has been extensively discussed for the Roman and Byzantine worlds.31 Local ceramic traditions and artistic and architectural styles abounded throughout the Islamic world, as did linguistic and religious divergences. Scholarship in Islamic archaeology, a discipline well suited to examining local manifestations, is addressing these variances. Proceeding from this, how can archaeologists address local regionalism on Islamic frontiers, which are frequently characterized as a dichotomous relationship between two groups, as described in texts with an overarching Islamic political or religious dominant ideology? An examination of internal frontiers functions both as a process and heuristic tool—­a mirror—­that reveals types of frontier interaction within the Islamic world: heterodox religious communities, settlements in environmentally marginal areas, and rural landscapes far from urban centers. The focus on communities, settlements, and landscapes provides a tangible view from within frontier societies themselves. Demographic Frontiers One significant category of internal frontiers is based on population movement, whether deliberate or otherwise. There are certainly many instances of population movement as control and ethnic exclusion. Communities were resettled to the frontiers for a number of reasons: to move them away from central lands because they were perceived as threats; to prevent already present frontier societies from becoming too homogenized and resistant to centralized control; or to repopulate newly acquired territories with loyal subjects. Population movements on the Islamic-­Byzantine thughūr show even more specific movements. The ‘Abbāsids settled loyal supporters, mainly Persians from Khūrāsān, on the frontier to weaken the Umayyads’ last power base. Muslims from surrounding Islamic lands came frequently on their own (though aided with incentives and by propaganda) to volunteer in the seasonal raids against the Byzantines. Alizadeh strongly argues that such demographic manipulations, mainly state-­directed deportations, were powerful tools for the creation of frontiers. What is mostly absent from these examples is any deliberate process of conversion. Religious Frontiers In the case studies on Jerba by Holod and Kahlaoui and the Jabal Nafūsa by Lauricella, we have physical manifestations of a religious frontier. We know heterodox religious communities existed within the Islamic world not even twenty years following the Prophet Muḥammad’s death in 632 CE. The Khawārij split from mainstream Islamic groups after the Caliph ‘Alī’s capitulation at the Battle T he A rchaeology of M edieval I slamic F rontiers

15

of Ṣiffīn in 657 CE and spread to the fringes of the Islamic world, practicing a militant form of Islam. The Ibāḍīs split around the same time, also spreading to the margins, yet adopting a more quietist approach to Islam. Subsequently, the supporters of ‘Alī, the Shī‘a, also settled widely, or in many instances were settled by the ‘Abbāsids when they came to power in 749 CE. They went to similar peripheries: North Africa, the southern Arabian Peninsula/Gulf region, and India. Aside from two main instances in Islamic history when Shī‘a groups rose to power on a grand scale (the Faṭimids in the tenth century and the Safavids in the fourteenth), we know very little of the nature of these groups. How does one look for archaeological signs of heterodoxy? Holod and Kahlaoui’s and Lauricella’s case studies take steps toward identifying Ibāḍīs regions and internal frontiers between Ibāḍīs and the rest of the Islamic world. The former study even shows subdivision between the two main Ibāḍī groups, the Wahbī and Nukkār on Jerba, through the appearance of discrete clusters of mosque/estate units and separate markets. For both the Jerba and Nafūsa regions, mosques reveal some shared signatures: they were rural, small scale with little ornament, and fitted only with slit windows. There were no central congregational mosques or madrasas, implying no hierarchy of space; they were built by and for each individual community. Further, the mosque often was associated with tower features for defense and communication. Was this a purely Ibāḍī mechanism or a local manifestation, or are the two inseparable? These structures also echo the North African ribāṭ, fortified enclosures built along the coasts in the ninth and tenth centuries, which were outfitted with mosques and towers. What is the relationship between Ibāḍī mosques and the ribāṭ in the same region and time period? On the scale of landscape, the Jabal Nafūsa, as an inaccessible and enclosed high space and hidden valley, and the island of Jerba both embody the closed, non-­hierarchical aspect of Ibāḍī society. In Nafūsa, the mosques were on mountaintops, while in Jerba, the mosques could be underground; either served as a place for refuge/danger and water storage and as a good marker of life on the periphery, whether within Islam society or between Islam and Christianity. Comparison with surveyed or excavated settlements in Oman or in Morocco, such as Sijilmasa in the Tafialt Oasis—­both regions known as centers of Ibāḍīsm—­would be important for observing whether there is an overarching Ibāḍī architecture, or if trade and connectivity with the outside world or local regionalism are visible in the material culture, function, and aesthetic of these sites. Did the Khawārij, known for example to also have lived on Jerba, also have similarly identifiable communities? Or is our inability to easily locate the materiality of heterodoxy in itself significant? In the Islamic world, sectarianism or religious pluralism need not manifest solely as a religious landscape of 16

A . A S A E G ER

“Islams” rather than Islam, to cite a popular phrase. Non-­Muslims certainly lived under Islamic authority. Nafūsa was also a place for Christian minority groups, such as Donatists, while Jerba had a community of Karaite Jews. Non-­Islamic evidence shows up on other frontiers as well, such as at Moschevaja Balka in the Caucasus, where Vorderstrasse analyzes Buddhist material culture from the burial of a Chinese merchant. Was the Buddhist monastic text, sutra, banner, and prayer flag at Moschevaja Balka for personal use, for distribution, or as a souvenir of exotic consumption? The church/monastery of Ṭarī’ between the mid-­sixth and mid-­eighth centuries was an isolated Christian outpost in a nomadic desert world, yet Islamic texts suggest it was one of many monasteries. The frontier locates, even permits, non-­orthodox Islamic and non-­Islamic settlements. Even further, the interaction of these spaces with mainstream Islam suggests an exchange of religious ideas. Environmental Frontiers In some cases, the frontiers were in relatively isolated environments, such as the Jabal Nafūsa, the Mediterranean islands of Jerba, Crete, Cyprus, and the Aegean, or the desert of the land of Ṭarī’, far from the Nile. Yet this isolation was never absolute and connections were always made with the outside world. Factors of environment and human agency controlled the degree to which this happened. Trade routes connected sites in the Nafūsa, which were all on the uplands, as well as in the Mediterranean, where many island coastal sites moved inland after the sixth and seventh centuries and even later.32 Sites interacted with the outside world whenever a ship came in, regardless of its affiliation, which was often dictated as much by weather and season as by commercial intent. In Jerba, access was not simple; good anchorages were scarce for large ships. Interestingly, in Jerba, inhabitants at one point in its history cut the Roman causeway connecting it to the mainland, further isolating the island. In the Jabal al-‘Ayn in the western Sudan, Jana Eger identified relic routes linking the desert site, rather indirectly (and perhaps intentionally so), to the Nile valley. The Jabal Nafūsa presents a case of highland–­lowland interaction.33 Yet the Nafūsa is not just an upland, but also served to shield outsiders from those who dwell in valleys. Natural resources were among the most important things that were protected: fortified granaries were key buildings of importance in Nafūsa and throughout North Africa and al-­Andalūs. Water resources were protected in Jerba within walled mosque complexes. Other examples can be seen in the Byzantine settlements of Cappadocia, where underground mosques had attendant granaries, as noted by Islamic geographers.34 In North Africa, the famous fourteenth-­ century ksar in T he A rchaeology of M edieval I slamic F rontiers

17

Tripolitania and southern Tunisia were granaries that feature a number of rooms around a central courtyard and vaulted roof. The aqrar in northern Morocco and the agadir in southern Morocco were similar fortified granaries with units assigned to families. In al-­Andalūs, the Cabezo de  la  Cobertera was a granary built on a steep mountain.35 On the islands where fresh water was scarce, reservoirs were the treasured commodity. Marginal areas within the Islamic world were frontiers of settlement and often the home of rebels, insurgents, or enemies of the state. Such was certainly the case for the island of Jerba—which functioned as a refuge or place of exile, and even housed a possible pirate base—as it was for Crete. The ‘Abbāsid slave revolt occurred in the swamp settlements of southern Iraq, and similar frontier wetlands can be seen near Antioch. Here, environmental frontiers also become sites of rebellion and conflict, homes to groups who wished to live literally outside the reaches of ruling or central authorities. One famous example in Islamic history is the Hashāshiyūn, the Nizarī Ismā‘īlī who in the late eleventh and twelfth centuries lived in the mountainous Alborz region in the nearly inaccessible fortification of Alamut.36 Geography and religious frontiers converge and are intertwined in these processes. Rural and Local Frontiers A well-­known and recognized trap, and one into which scholars still consistently fall, is the act of categorizing peoples on either side of the frontier into one or the other of two homogenous groups. In contrast, several of the studies in this volume highlight heterogeneity. The Alans and Khazars, in Vorderstrasse’s study of the Caucasian frontier, were not a unified group. As Ruffini shows, the Nubians were not either, and their frontier was peppered with local tribes acting on their own, controlling trade south of ‘Uswān, and intermarrying. The plurality of groups negates any notion of a dichotomous frontier. One exception may be in the Mediterranean islands, where Randall shows that circumstance and perceived isolation may have led “islanders” to self-­identify as a distinct group. Social processes such as intermarriage made sharply divided identities difficult to parse out. The Ibāḍīs also represented a plurality within Islamic society. In the Mediterranean, Randall’s study shows how Byzantine and Islamic shipping lanes were frequented by unaffiliated pirates and privateers negotiating interaction and exchange, much as did the Banū Kanz in Nubia. Whether pirates or nomads, the frontier was a complex landscape of many communities and individuals carving out space. At the same time, these examples give us a view into rural societies within the Islamic world. This view is rare: Islamic studies typically have focused 18

A . A S A E G ER

on the monumental and urban, the elite and literate. Traditional excavations have focused on the central lands of the empires and on urban, monumental, artistic, and religious sites. Thus one frontier was between rural and urban space. Newer field projects in Islamic archaeology, as a consequence of political instabilities, have moved to the peripheries of urban landscapes, focusing on rural, military, and trading sites: the farmstead or village, the fortified outpost, or the khan (inn).37 There were no cities on Jerba and Nafūsa; they were far removed from any large urban center. Before the Ottoman period, Jerba had no major military presence or significant fortifications. In Nubia, towns were fortified with walls,38 but possessed no fortification as such (castle, etc.). In both Jerba and Nafūsa, the towns were laid out for protection and defense with an ad hoc system of warning built from local efforts. Satellite mosques and watchtowers perhaps played a strategic role, to maximize visibility as they communicated with one another. Fortified granaries may have also played a role. A similar interpretation might explain the satellite mosques in the seminomadic Negev settlements in the early Islamic period.39 Efforts at defense and protection on the Mediterranean islands were also local and not a result of central or state initiatives. Islands largely ruled themselves as local and rural communities; indeed, they were often called “a landscape of villages.” Left to their own devices, they instituted an internal warning system.40 So, too, shipping became increasingly privatized. Franklin’s study in Armenia shows how, besides its international influence in connecting to the outside world, the caravanserai was loaded with local authority and agency of patronage, with strong traditions of hospitality. An important consequence of a rural frontier is the political, economic, and sociocultural interactions that occurred between the urban and the rural, or between a government and its hinterland. Walker’s study on the Mamlūk frontier in Jordan, made use of archaeology, survey, and waqf textual records, to show precisely how rural inhabitants assumed local autonomy.41 This agency was manifest when locals exhibited an avoidance of state control and implemented taxation through private or cooperative entrepreneurship. The archaeology of these rural local frontiers shows us the micro-­interactions as streams that both derive from and contribute to larger historical processes of caliphal or dynastic rise and fall, the failings of the central state to pay the military or governors, the ‘iqṭa system of distributing lands in exchange for service, the failings of ‘iqṭa, and the inexorable rise in local power and expansion of privatization. The Mamlūk hinterland in Jordan, an internal frontier, developed only as a result of imperial decline. Future studies might reveal and address some limitations: what do communities look like with state backing as opposed to private backing? How T he A rchaeology of M edieval I slamic F rontiers

19

can we differentiate between private and cooperative field systems over time, and can these shift back and forth? And how can we consider dynamic fluctuation between sedentarists who shifted to a nomadic existence and vice versa? An archaeology of frontiers shows us not necessarily the ruptures and political and military successes or failures of ruling groups but continuities of local groups and traditions, building styles and crafts, within these regions. These varied frontiers—­ economic, religious, environmental, rural, and local—­are not easily mutually exclusive. Many of the studies here possess some or all of these attributes. A recent study by Mahoney perfectly encapsulates this multilayered space by showing an enclosed insular valley in highland Yemen, housing an Ibāḍī community with a history of rebellion against central Islamic lands, with its own local traditions of architecture and ceramics (utilizing its own clay source).42 Cognitive Frontiers Beyond the tangible material-­based realm of archaeology, frontiers are a metaphysical state of mind as imaginary places, feared or desired, and often political constructs. Several studies in this volume show the frontier experience was a cognitive one; as an imagined space, the frontier was relative to anywhere. In Randall’s study, the inhabitants of the Mediterranean islands perceived themselves as isolated and disconnected from the major cities. For Franklin, the Caucasian frontier was as much a place of imagination and exoticism as it was wild and dangerous. The frontier featured prominently in the pages of Islamic geographies, which described routes and itineraries as well as the delicacies and unique products of each of town, thus increasing the exoticism of those living there. Yet, not too far away was the fantastical border of Gog and Magog, separating the uncivilized nomadic hordes from the lands of Islam by a very real wall, which Alizadeh discusses in his chapter. It was of Sasanian construction but continues into the early Islamic period, at least by its reference in the pages of Islamic historians.43 While Randall’s island inhabitants frequently complained of their situation, Franklin’s embraced it. Jana Eger’s monastery in Ṭarī’’ is far from the Egyptian-Nubian frontier, yet its placement in an anonymous Islamic geographical text, Ḥudūd al-­ʿĀlam min al-­Mashriq ila-­l-­Maghrib, “The Boundaries of the World from the East to the West,” written in Persian in the late tenth century, connects it to Islamic lands in two ways: metaphysically as a location included within the pages of a text on frontiers, and situated at one edge of an imagined Islamic world. For this Persian traveler and for those for whom this book was written, whether administrative officials who used the text to delineate frontiers, or merchants 20

A . A S A E G ER

and pilgrims, this was a Christian border, remote but on some itinerary. All three frontiers intentionally connected to both local and international customs and styles. The exoticism of the frontier is turned around and for these frontier societies, it is the major cities inland, those of the Islamic world, that are distant and lofty, at the edges of imagination. Perhaps the thirteenth-­century poet Amīr Khusraw, a Turk who wrote in Persian but lived in India, expressed it best in elevating his town of Delhi, at the edges of the Islamic world, over Mecca, its very heart: Excellent Delhi, the protector of religion and justice It is the Garden of Eden; may it flourish forever. It is like an earthly paradise in its qualities—­ May God guard it from all calamities! . . .  If Mecca but heard of this garden It would circumambulate Hindustān The city of the Prophet takes oaths by it; The city of God became deafened from its fame.44 CONCLUSIONS

All of the studies in this volume suggest that the old core-­periphery model be complicated and rethought. In its place, they offer the sense that the frontier, lying at “the edge of everywhere,” is itself a center. It is here where interaction between people and goods, moving back and forth and following numerous trajectories, not unlike McCormick’s Mediterranean Sea, is as frequent and relevant as that which takes place in the city or moves from the city to the frontier. Thus, these frontiers connect to the Islamic world and exhibit influence from it, but also show that the frontier is less singularly defined as “Islamic” culturally when viewed from the outside in. The study of frontiers, whether on the edges of settlements or empires, raises crucial questions for the field of Islamic studies, since it addresses interactions with other religious and ethnic groups and local variations within material culture and architecture. These new approaches have redefined outmoded traditional views of Islam and have certainly dismantled any notion of a monolithic Islam, whether defined through universally shared religious beliefs, cultural cohesion, or wide-­reaching political authority. Whether these frontiers actually existed and/or whether we as scholars heighten their contours matters little. As an analytical tool, they allow us to view alternative histories, and interactions and processes that we might otherwise neglect, or worse, assume never changed and thus leave unchallenged. T he A rchaeology of M edieval I slamic F rontiers

21

NOTES

1. All the chapters in this volume arose from those meetings and two were solicited to round out the discussion: Ian Randall’s chapter (also presented at ASOR in 2013) and Karim Alizadeh’s (published in an altered and earlier form in Persian in the Journal of the Society of Iranian Archaeology 2016). 2. Haug 2011, 635. What is missing is the West African frontier of the Sahara/ Sahel; see Insoll 2003, which addresses the relationship of commerce and conversion (not part of the chapters in this volume), and Heddouchi 2012. Islamic merchant communties in West and East Africa, and Southeast Asia and China, are also important to consider. 3. Mullin 2011b. See Curta 2005; Abulafia and Berend 2002; Powers and Standen 1999; Bartlett and MacKay 1989. 4. Antrim 2012; Zadeh 2011. 5. Or archaeology of frontiers in general, for that matter, see Mullin 2011a, 1. 6. Eger 2015, 313; Zartman 2010, 1–­2. 7. Brauer 1995, 1–­73. 8. Although Kaegi views this as a fluid and porous frontier characterized as much by commercial activities and gift and prisoner exchanges as by military expeditions. He first published this before Brauer: Kaegi 1986; Kaegi 1996, 83–­92. See also Amitai-­ Press 1999, 128–­152; Heywood 1999, 228–­250. 9. Lightfoot and Martinez 1995, 472–­473; Wallerstein 1974. 10. Lightfood and Martinez 1995, 228ff.; Mullin 2011a, 5–­6 for critiques; see also Stein 1999. 11. As such, Smith argues against any universalizing mythology in Islam but an opportunistic value assigned to space (1993, 289–­310). 12. Rooke 2000. See also Eger 2015, 9–­12; Zartman 2010, 13. 13. Antrim 2012, 33ff. 14. Ibid., 121. 15. Ellenblum 1999. 16. Luce 2009. 17. Bekker 1967, 71–­86; Gabbay 2010. 18. Ellenblum 2007, 143–­144. 19. Zartman 2010, 2–­3. 20. For al-­Andalūs, see Curta 2011; Fábregas and Sabaté 2015. New work is coming out of the frontier of Central Asia/northeast Iran; see Genequand and Northedge 2012; Frachetti 2016; Wordsworth 2016. 21. Redford 1998. 22. Eger 2015. 23. As suggested by the Roman limes; see Isaac 1990 and Ellenblum 2007, 122. 22

A . A S A E G ER

24. Ibn Jubayr 1907, 287; 300–­301. See also Eger 2015, 275–­276; Ellenblum 2007, 138. 25. In the fourteenth century, the search for the mythical wealthy king Prester John would also connect this region (and even more so, Ethiopia) with west European travelers, cartographers, and merchants. 26. McCormick 2002. 27. Di Cosmo 2006. 28. Mullin 2011a, 2–­3, 5. Though the question of ethnicity is avoided in Mullin’s volume. 29. Stein 2011. 30. Mullin 2011a, 3. 31. Among the many recent works, see, for example, Mattingly 2011; vanDommelen and Terrenato 2007. 32. Recent evidence at Naxos, largest of the Aegean islands, and Paros shows that in the eighth/ninth century, new churches were built inland, not on the coast and there were fewer settlements in general until the tenth century. This was also seen in other Cyclades Islands, such as Amorgos and Thira (Santorini). The main inland fortification on Naxos, at Kastro Apalirou, dating from the sixth/seventh to the ninth centuries, shows exactly this settlement shift. The fortification had likely functioned as a place of refuge and had cisterns, presses, and a threshing floor. See Ødegård et al. 2017. 33. The nature and process of this environmental frontier has begun to receive more scholarly attention recently in Near Eastern archaeology. See, for example, Gatz 2016. 34. Indeed, one name for this frontier was al-­Matāmīr, “the granaries.” See Eger 2015, 253. 35. De Meulemeester 2005, 609–­615. 36. See also the twelfth-­century Isma’ili Shrine of Nasir Khusraw in the mountainous hidden valley of Yumgan; Schadl 2009. 37. The number of excavations of such sites is rapidly growing and too many to enumerate. For a good recent example of work published in one volume, see McPhillips and Wordsworth 2016 and the following chapters within it: Macumber 2016; Bartl 2016; Jones 2016; Thomas and Gascoigne 2016. Of course, the volume itself and its focus on rural landscapes and environmental analysis is an essential new direction for the field. 38. By late antiquity, many of these towns had churches inside. Serra East was an Iron Age fortress that transformed into a fortified town. In the early Islamic period, it featured animal pens, workshops, and kilns. See Williams and Tsakos 2013. 39. Avni 1996. 40. The monasteries on Naxos follow a line from coast to center, toward Kastro Apalirou. Some of these were tower monasteries. It is possible that these also served as T he A rchaeology of M edieval I slamic F rontiers

23

a network of warning stations from coastal incursions or attacks on the harbor. For a discussion on the monasteries, see Sigala 2016. http://​byz2016​.rs/​OkrugliStolovi/​U TO RAK/​03​-RT​_Byzantine​%20Naxos​.pdf. 41. Presented as part of the original ASOR frontier panels, see Walker 2011. 42. Mahoney 2016. 43. See also Jaritz and Kreem 2009. 44. As translated by Gabbay 2010, 111n61, from Amīr Khusraw, 1918, Qirān al-­ sa‘dayn, Mawlavī Muḥammad Ismā‘īl Meraṭhī, ed., 28–­29, Aligarh, India: Maṭba‘i Instītūt. REFERENCES

Abulafia, D., and N. Berend. 2002. Medieval Frontiers: Concepts and Practices. Burlington, VT: Ashgate. Amitai-­Press, R. 1999. “Northern Syria between the Mongols and Mamluks: Political Boundary, Military Frontier, and Ethnic Affiliation.” In Frontiers in Question: Eurasian Borderlands, 700–­1700, edited by D. Powers and N. Standen, 128–­152. New York: St. Martin’s Press. Antrim, Z. 2012. Routes and Realms: The Power of Place in the Early Islamic World. New York: Oxford University Press. Avni, G. 1996. Nomads, Farmers, and Town-­Dwellers: Pastoralist-­Sedentist Interaction in the Negev Highlands, Sixth–­Eighth Centuries CE. Jerusalem: Israel Antiquities Authority. Bartl, K. 2016. “Water Management in Desert Regions: Early Islamic Qasr Mushash.” In Landscapes of the Islamic World: Archaeology, History, and Ethnography, edited by S. McPhillips and P. Wordsworth, 50–­68. Philadelphia: University of Pennsylvania Press. Bartlett, R., and A. MacKay. 1989. Medieval Frontier Societies. New York: Oxford University Press. Bekker, K. 1967. “Historical Patterns of Culture Contact in Southern Asia.” In Beyond the Frontier: Social Process and Cultural Change, edited by P. Bohannan and F. Plog, 71–­86. New York: The Natural History Press. Brauer, R. 1995. “Boundaries and Frontiers in Medieval Muslim Geography.” Transactions of the American Philosophical Society 85.6: 1–­73. Broadhurst 1952. Curta, F., ed. 2005. Borders, Barriers, and Ethnogenesis: Frontiers in Late Antiquity and the Middle Ages. Turnhout, Belgium: Brepols. Curta, F. 2011. “The Centrality of the Periphery: the Archaeology of al-­Andalus.” Early Medieval Europe 19.4: 377–­384. 24

A . A S A E G ER

Di Cosmo, N. 2006. “The Origins of the Great Wall.” The Silk Road 4.1: 14–­19. van Dommelen, P., and N. Terrenato, eds. 2007. “Articulating Local Cultures: Power and Identity under the Expanding Roman Republic.” Journal of Roman Archaeology, Supplement 63. Eger, A. 2015. The Islamic-­Byzantine Frontier: Interaction and Exchange among Muslim and Christian Communities. London: I.B.Tauris. Ellenblum, R. 1999. “Were There Borders and Borderlines in the Middle Ages? The Example of the Latin Kingdom of Jerusalem.” In Frontiers in Question: Eurasian Borderlands, 700–­1700, edited by D. Powers and N. Standen, 105–­119. New York: St. Martin’s Press. Ellenblum, R. 2007. Crusader Castles and Modern Histories. Cambridge, UK: Cambridge University Press. Fábregas, A., and F. Sabaté. 2015. Power and Rural Communities in al-­Andalus: Ideological and Material Representations. Turnhout, Belgium: Brepols. Frachetti, M. 2016. “High Altitude Towns of the Qarakhanid Empire: New Archaeology at the Medieval Center of Tashbulak, Uzbekistan.” Fondation Max van Berchem Bulletin 30: 1–­3. Gabbay, A. 2010. Islamic Tolerance: Amīr Khusraw and Pluralism. New York: Routledge. Gatz, C. 2016. “Of Highland-­Lowland Interaction Borderlands: Local Societies and Foreign Power in the Zagros-­Mesopotamian Interface.” Journal of Anthropological Archaeology 44 (Part A): 127–­147. Genequand, D., and A. Northedge. 2012. “Dehistān: une ville médiévale au sud-­ouest du Turkmenistan.” Fondation Max van Berchem Bulletin 26 (December): 1–­3. Haug, R. 2011. “Frontiers and the State in Early Islamic History: Jihād between Caliphs and Volunteers.” History Compass 9/8: 634–­643. Heddouchi, C. 2012. “For a Darislamic Archaeology: Egypt and West Africa and the Limits of Islamic Archaeology.” PhD diss., Department of Near Eastern Languages and Civilizations, University of Chicago. Heywood, C. 1999. “The Frontier in Ottoman History: Old Ideas and New Myths.” In Frontiers in Question: Eurasian Borderlands, 700–­1700, edited by D. Powers and N. Standen, 228–­250. New York: St. Martin’s Press. Ibn Jubayr. 1907. Riḥla. Leiden, Netherlands: Brill. Ibn Jubayr. 1952. The Travels of Ibn Jubayr. Trans. R.J.C. Broadhurst. London: Jonathan Cape. Insoll, T. 2003. The Archaeology of Islam in Sub-­Saharan Africa. Cambridge, UK: Cambridge University Press. Isaac, B. 1990. The Limits of Empire: The Roman Army in the East. Oxford, UK: Oxford University Press. T he A rchaeology of M edieval I slamic F rontiers

25

Jaritz, G., and J. Kreem. 2009. The Edges of the Medieval World. Budapest: Central European University Press. Jones, I. 2016. “Beyond Iron Age Landscapes: Copper Mining and Smelting in Faynan in the Twelfth to Fourteenth Centuries CE.” In Landscapes of the Islamic World: Archaeology, History, and Ethnography, edited by S. McPhillips and P. Wordsworth, 111–­128. Philadelphia: University of Pennsylvania Press. Kaegi, W. 1986. “The Frontier: Barrier or Bridge?” The 17th International Byzantine Congress: Major Papers. New Rochelle, NY: A. D. Caratzas. Kaegi, W. 1996. “Reconceptualizing Byzantium’s Eastern Frontiers in the Seventh Century.” In Shifting Frontiers in Late Antiquity, edited by R. Mathisen and H. Sivan, 83–­92. Brookfield, VT: Variorum. Lightfoot, K. G., and A. Martinez. 1995. “Frontiers and Boundaries in Archaeological Perspective.” Annual Review of Anthropology 24: 471–­492. Luce, M. 2009. “Frontier as Process: Umayyad Khurasan.” PhD diss., University of Chicago. Macumber, P. 2016. “The Islamic Occupation of Qatar in the Context of an Environmental Framework.” In Landscapes of the Islamic World: Archaeology, History, and Ethnography, edited by S. McPhillips and P. Wordsworth, 34–­49. Philadelphia: University of Pennsylvania Press. Mahoney, D. 2016. “Ceramic Production in the Central Highlands of Yemen during the Islamic Period.” In Landscapes of the Islamic World: Archaeology, History, and Ethnography, edited by S. McPhillips and P. Wordsworth, 129–­142. Philadelphia: University of Pennsylvania Press. Mattingly, D. J. Imperialism, Power, and Identity: Experiencing the Roman Empire. Princeton, NJ: Princeton University Press, 2011. McCormick, M. 2002. Origins of the European Economy: Communications and Commerce AD 300–­900. New York: Cambridge University Press. McPhillips, S., and P. Wordsworth, eds. 2016. Landscapes of the Islamic World: Archaeology, History, and Ethnography. Philadelphia: University of Pennsylvania Press. de Meulemeester, J. 2005. “Granaries and Irrigation: Archaeological and Ethnological Investigations in the Iberian Peninsula and Morocco.” Antiquity 79: 609–­615. Mullin, D. 2011a. “Border Crossings: The Archaeology of Borders and Borderlands.” In Places in Between: The Archaeology of Social, Cultural, and Geographical Borders and Borderlands, edited by D. Mullin, 1–­12. Oxford, UK: Oxbow Books. Mullin, D., ed. 2011b. Places in Between: The Archaeology of Social, Cultural, and Geographical Borders and Borderlands. Oxford, UK: Oxbow Books. Ødegård, K., D. Hill, and H. Ingvaldson. 2017. “Kastro Apalirou at Naxos: Preliminary Report on the Investigation of an Early Byzantine Town.” New Cities in Late Antiquity. Bibliothèque de l’Antiquitié tardive 35. Turnhout, Belgium: Brepols. 26

A . A S A E G ER

Powers, D., and N. Standen, eds. 1999. Frontiers in Question: Eurasian Borderlands, 700–­1700. New York: St. Martin’s Press. Redford, S. 1998. The Archaeology of the Frontier in the Medieval Near East: Excavations at Gritille. Philadelphia: University Museum Publications, University of Pennsylvania. Rooke, T. 2000. “Writing the Boundary: Khitat al-­Shām by Muhammad Kurd ‘Ali’.” In The Concept of Territory in Islamic Law and Thought, edited by Y. Hiroyuki, 165–­186. London: Kegan Paul International. Schadl, M. 2009. “The Shrine of Nasir Khusraw: Imprisoned Deep in the Valley of Yumgan.” Muqarnas 26: 63–­93. Sigala, M. 2016. “Discussing Byzantine Churches with Monastic Use on Naxos (and the Neighboring Islands).” Paper given at the 23rd International Congress on Byzantine Studies, Belgrade, Serbia. Smith, J. Z. 1993. “Map Is Not Territory.” In, 289–­310. Chicago: University of Chicago Press. Stein, G. 1999. Rethinking World Systems: Diasporas, Colonies, and Interaction in Uruk Mesopotamia. Tucson, AZ: University of Arizona Press. Stein, G. 2011. “Plenary Address.” American Schools of Oriental Research Annual Meeting, San Francisco, CA. Thomas, D. C., and A. Gascoigne. 2016. “The Architectural Legacy of the Seasonally Nomadic Ghurids.” In Landscapes of the Islamic World: Archaeology, History, and Ethnography, edited by S. McPhillips and P. Wordsworth, 169–­183. Philadelphia: University of Pennsylvania Press. Walker, B. 2011. “Carving out a Space for Us: Religion, Land, and the Transformed ‘Frontiers’ of the Late Mamluk State.” Paper presented at the American Schools of Oriental Research Annual Meeting, San Francisco, CA. Wallerstein, I. 1974. The Modern World System I: Capitalist Agriculture and the Origins of the European World-­Economy in the Sixteenth Century. New York: Academic Press. Williams, B., and A. Tsakos. 2013. “Serra East and the Twelfth Century Nubian Frontier.” Paper given at the American Schools of Oriental Research Annual Meeting, Baltimore, MD. Wordsworth, P. 2016. “Sustaining Travel: The Economy of Medieval Stopping-­Places across the Karakum Desert, Turkmenistan.” In Landscapes of the Islamic World: Archaeological, Historical and Ethnographic Approaches, edited by S. McPhillips and P. Wordsworth, 219–­236. Philadelphia: University of Pennsylvania Press. Zadeh, T. 2011. Mapping Frontiers across Medieval Islam: Geography, Translation, and the ‘Abbasid Empire. London: I.B.Tauris Zartman, I. W. 2010. “Introduction: Identity, Movement, and Response.” In Understanding Life in the Borderlands: Boundaries in Depth and in Motion, edited by I. W. Zartman. Athens: University of Georgia Press. T he A rchaeology of M edieval I slamic F rontiers

27

Part I

The Western Frontiers The Maghrib and the Mediterranean Sea

Crete and Greek Islands; Idrisī, Nuzhat al-mushtāq f ī ikhtirāq al-āfāq, MS Pococke 375, fol. 204b-205a, courtesy of Bodleian Library, University of Oxford

2 The Maghrib (Libya, Tunisia, Algeria, and Morocco) is often treated as a coherent unit, although it may better be conceived of as an assemblage of regions.1 Before the arrival of Islam, the Maghrib had been incorporated into larger empires (Roman, Vandal, and Byzantine) that centralized their power on the Mediterranean coast. The challenge for such controlling powers was how and whether to extend dominance beyond the fertile agricultural coastline and into the mountainous and inhospitable interior. Populations not living on the coast naturally sought to avoid imperial power and the concomitant threats of conscription and taxation. The Jabal Nafūsa region, in the western part of Libya, perennially resisted external power.2 The inhabitants, indigenous North Africans, included members of the Donatist sect of Christianity, who had been the target of anti-­Donatist repression before Islam.3 Imazighen (that is, indigenous Berber) populations throughout the Maghrib opposed the Arab armies, both at the time of the initial invasions at the end of the seventh century and through rebellions directed at the Umayyad capital Qayrawān in 740 and 742. The Imazighen, many of whom were Christian or Jewish, were treated poorly, even when they served in Arab armies.4 Taxation was often taken in the form of slaves and was particularly brutal. When the inhabitants of the Jabal Nafūsa did adopt Islam they, like many other indigenous groups in the Maghrib, became followers of the Ibāḍī sect.

Ibāḍī Boundaries and Defense in the Jabal Nafūsa (Libya) Anthony J. Lauricella

DOI: 10.5876/9781607328773.c002

31

Ibāḍī Islam is located at the fringes of the Islamic word, both doctrinally and geographically. Today it is the official religion of Oman and has historical roots in the Mzab region of Algeria, Zanzibar, the island of Jerba (Tunisia), and the remote Jabal Nafūsa mountain range in Libya. Unlike the violence of the Khārijite schism at the Battle of Ṣiffīn (656) that originally led to its formation in the seventh century, Ibāḍīs rejected violence. Non-­believers, that is non-­Ibāḍī Muslims, were to be ignored rather than fought. Instead, the focus was on individual piety and the righteousness of the community as a whole. The most virtuous member of the community should lead it regardless of tribal affiliation or ethnicity. This openness made Ibāḍī Islam attractive to the indigenous populations of North Africa and gave religious voice to their resistance against enslavement at the hands of the Muslim regime centered at Qayrawān.5 In the Jabal Nafūsa of Libya, landscape and the Ibāḍī faith produced a unique set of defensive strategies and borders between Ibāḍī space and the outside world. Considerable evidence demonstrates that the Jabal Nafūsa before Islam was a religiously plural region, a frequent home to dissident or minority religious movements. The topographic and geographical features made the area a sanctuary for those opposing domination from outside. The Jabal Nafūsa was home to both Christian and Jewish populations, and regional toponyms preserve memory of the region’s identification with the Donatist schism of Christianity, which, in its focus on the need for religious leaders to themselves be pious, bears similarities to Ibāḍī Islam.6 After the arrival of Islam, the area retained traces of its pre-­Islamic religious character. J. W. Allan’s architectural survey of the region notes that the names of several mosques in the Jabal Nafūsa have a Christian flavor. Mosques are called ḥawāriyīn (that is, connected to the Arabic word for Jesus’s disciples), kanīsa, or taghlis, which is derived from the Latin ecclesia.7 The Jabal Nafūsa was geographically, religiously, and ethnically marginal, a distinct region marked off by natural geographic features, and socially defined by the differences between its inhabitants and those outside it. Frontiers in the Islamic world are typically thought of in terms of the Byzantine-­ Islamic border in Anatolia and Syria.8 Regions like the Jabal Nafūsa, however, sought to set up boundaries not only with the opponents of Islam but also with central Islamic authority. What kind of geographic and ideological frontiers served to delimit the spaces occupied by this spiritually distinct part of the early Islamic world? The landscape of the region and the religious ideology within that landscape were bound up with the negotiation of borders both inside and outside the Jabal Nafūsa. This chapter intends to accomplish two things. First, it aims to describe the settlement patterns, the 32

A N T H O N Y J. L AU R I C ELL A

defensive strategies, and the nature of the various borders within the Jabal Nafūsa region of Libya in the early Islamic period. Drawing heavily on the work of Tadeusz Lewicki, I provide a spatial description of the region and an analysis of strategic visibility using a Geographic Information System (GIS). Second, I suggest that certain aspects of Ibāḍī regional organization seen elsewhere may have been originally conceived here, and are defined, at least in part, by the conflation of religion and defensive considerations, expressed through strategic control of the visible landscape. This analysis sheds light on the ways in which Ibāḍī authority was extended through the region in the Islamic periods. THE LANDSCAPE

Tripolitania (roughly corresponding to the western portion of modern Libya) can be divided into three main geographic zones: the Mediterranean coastal plain, a range of foothills and mountains set back from the coast, and the beginning of the desert to the south of the mountains. The Jabal Nafūsa constitutes the western portion of the second mountainous zone between coast and desert and extends eastward from what is today the border with Tunisia (figure 2.1). The roughly 180-­km-­long range is cut by several deep wadis. Elevations range from 500 to around 900 m above sea level. Rainfall generally varies between 150 and 250 mm per year.9 The economy is varied. Grain crops are grown below the escarpment and watered using captured runoff, while olives and other tree crops are cultivated at higher elevations. Both agriculturalists and seminomadic pastoralists have inhabited the region.10 As McLachlan notes “human settlement of the [ Jabal] Nafūsa has historically been associated with groups seeking refuge.”11 Movement in an east–­west direction would require crossing many deep valleys. Access from the plain below is very difficult since intruders could be easily detected from the hilltops. REGIONAL-­LEVEL ORGANIZATION

In the absence of archaeological excavations, insight into regional-level organization of space can be gleaned from textual sources. Lewicki’s analysis of the Tasmīya Shuyūkh Jabal Nafūsa wa-­Qurāhum (The Names of the Sheikhs of the Jabal Nafūsa and Their Villages), a list of prominent Ibāḍī teachers operating in the region during the eighth through twelfth centuries, incorporates his study of other Ibāḍī textual sources. Unfortunately this analysis of the region does not incorporate Prevost’s recently published study of the Jabal Nafūsa, which I BĀ Ḍ Ī B O U N DA R I ES A N D D EF EN S E I N T H E JA BA L N A F Ū S A ( LI B YA )

33

Figure 2.1. The Jabal Nafūsa (Libya).

provides updated and precise geographic information.12 According to Lewicki, the Jabal Nafūsa was divided into two zones: a Western region known as the Amīnāj, and “Jādū and its villages” ( Jādū wa-­nāhītihi), the eastern region.13 This two-­part division appears to predate Islam, and may represent a distinction between the Nafūsans and neighboring related groups.14 The western portion contains twenty-­two named towns. The capital, Nālūt, is located at the western end of the Amīnāj. Nālūt is the largest settlement in the region today, and was a major stopping point for goods being traded between the coast and the Sahara.15 The boundary between the Amīnāj and the region of Jādū does not seem to correspond to any salient geographical feature, and must have been regionally understood. Lewicki notes that the Amīnāj is divided into three additional zones in the Kitāb Sīyar al-­Shammākhī. In this western part of the Jabal Nafūsa, several locations stand out as particularly significant. Sharūs, located roughly in the center of the Amīnāj, was another important stopping point for caravans traveling north–­south. The town is said to have had a mosque near its fortified granary, and a sūq near its Jewish quarter.16 The town of Buqṭūra was close by and included a church reused as a mosque. The town of Wīghū had an underground mosque and was a stopping 34

A N T H O N Y J. L AU R I C ELL A

point on a route passing through the Jabal Nafūsa. Wīghū seems to have been a major rival of Sharūs, and apparently destroyed the town sometime in the twelfth century, although Warfalli mentions evidence of settlement there until the fifteenth century.17 The town of Fursaṭā was located along a road leading to Jādū. This settlement is said to be very old and there was also a church later reused as a mosque.18 Warfalli notes the similarities between Fursaṭā and Nālūt and with the Mzab region of Algeria, another Ibādī stronghold.19 Interestingly, the names of the villages Anīr and Kabāw refer to defensive features of the landscape, meaning “front” (most likely referring to its position at the front of the escarpment) and “refuge,” respectively, in local Amazigh dialect.20 The eastern zone of the Jabal Nafūsa, Jādū and its environs, was similarly dominated by a major city that was also an entrepôt for trade plying a north–­south route. One of the smaller towns under the jurisdiction of Jādū, Ijnāwūn, may have been named for a population from the Sudan that settled there. An Ibāḍī governor of the early ninth century is said to have spoken Kanembu, a language of the region around the Lake Chad basin.21 By the middle of the eighth century, Imazighen Ibāḍīs had become involved with the trade in slaves across the desert from the south.22 These trade routes traversed and articulated a separate economic frontier with sub-­Saharan Africa. As with the economic frontiers described in Nubia by Ruffini (chapter 5, this volume), and in the Caucasus by Vorderstrasse (chapter 8, this volume), the exchange of goods promotes connections across (and despite) geographical distance. The town of Fusāṭū, like Fusṭāṭ in Egypt, takes its name from the Latin fossatum, meaning “ditch.” For Lewicki, this indicates a history of fortification and a connection to the limes system.23 The eastern border of the Jabal Nafūsa is marked by Tigharmīn. This town also demarcates the western border of the territory of Zintān.24 Overall, some broad patterns of settlement in the Jabal Nafūsa can be described. Multiple Ibāḍī sources saw the region as divided into an eastern and a western part. Both segments were further subdivided. Two important settlements on north–­south trade routes were located at both the east and west limits of the region. This last point is particularly interesting if we step back and consider what is known of the Roman limes in the region. Tripolitania was dotted with forts and enclosures, particularly along trade routes. The main charge of the units policing Tripolitania was to protect trade and travel, and to keep peace between settled farmers and nomadic groups.25 The old Roman road through the Jabal Nafūsa appears to have begun at the coast, near the island of Jerba, then crossed the escarpment somewhere near the fort of Dehibat. From there it followed along the southern side of the mountain range until I BĀ Ḍ Ī B O U N DA R I ES A N D D EF EN S E I N T H E JA BA L N A F Ū S A ( LI B YA )

35

it reached Zintān, where another north–­south route crossed the escarpment. Defensive sites were found to the west of Nālūt (at Dehibat), and to the east of Jādū at Zintān.26 Although Nālūt and Jādū were inhabited before the arrival of Islam to the Maghrib, and before the rise of the Jabal Nafūsa as an Ibāḍī stronghold, texts describing these periods show little interest in the edges of the mountain range where it would have been more practical to interface with people from the lowlands. DEFENSIBILIT Y: SETTLEMENT PLACEMENT AND SETTLEMENT STRUCTURE

The first line of defense is obviously the landscape itself. In this and other regions defined by topography, the mountain range provides a foundation for built systems of defense and territorial control (walls, gates, or watchposts, for example). For a different utilization of natural barriers see Alizadeh (chapter 7, this volume). The mountain range is most vulnerable to attack from the south, from the back of the escarpment. The desolate terrain there, however, makes attack from that direction unlikely. Typically, settlements are placed above valleys, between ravines, or on the tops of peaks. Isabella Sjöström, in her analysis of Tripolitania as surveyed in the UNESCO Libyan Valley Survey, notes that “villages in the Jabal are mostly perched on the crests of hills, on top of the escarpment that rises above the Gefara plain, often hidden from view from below. Most of the villages are not visible from the plain unless they are truly inaccessible, e.g., Tirmisa. They are well hidden, and do not advertise their location, a location that tends not to be as difficult of access when approached from the plateau in the south.”27 In many locations, defensive position is favored over access to water. According to Despois, towns in the Jabal Nafūsa are not fortified with walls. Rather, defense is provided by a system of factors. Beyond the natural defensive character of the landscape, there is the actual layout of the towns themselves. The core of settlement in the Jabal Nafūsa is the mosque and the fortified granary (gasr).28 These two structures are frequently located together. The gasr, which has a long history both before and after the arrival of Islam to the Maghrib, as well as a wide distribution throughout the southern Mediterranean,29 is commonly associated with fortified towers. Despois describes these towers as being around 10 m high, with a square base, and narrowing toward the top. They are often equipped with a terrace, or meurtiere, possibly for repelling attackers. In some cases, defensive towers are located at a distance, and presumably served as outposts that signaled back to the main settlement.30 36

A N T H O N Y J. L AU R I C ELL A

The mosques of the region are often built fully or partially underground. Construction began by digging a large hole, building the mosque, and then piling the removed soil around the new structure.31 While underground construction is not specific to any particular period in this region, it is in keeping with the broader agenda of defense seen in the Jabal Nafūsa. Not all mosques in this region are underground, but the fact that they are not uncommon highlights another way in which local Ibāḍī Islam drew on previous local traditions. No earlier external models outside the Jabal Nafūsa present themselves as clear candidates. Rather, defensible structures like the gasr or other subterranean and semi-­subterranean buildings may have served as the inspiration for the underground mosques. In addition to reflecting a construction style that fit the environmental and material needs of the region, the underground mosque fused the pre-­Islamic traditions of local building with the needs of the newly formed Islamic community seeking to ensure that their region remained a refuge. The core of a Nafūsa settlement in the Islamic period consists of the circular fortified gasr, one or more defensive towers, and a mosque nearby. This pattern is described at Sharūs and Wuryūrī.32 To this we can add the additional mosques often located at the edges of the various towns in the Jabal Nafūsa. Allan’s survey notes that mosques often “[appear] at first sight to be in the middle of nowhere . . . [with] no visible connection with past communities” and records that many mosques are located on hilltops.33 Thus, much like the watchtowers, some mosques may have been placed at locations peripheral to settlements to signal threats back to the center (although perhaps also to emphasize the private nature of Ibāḍī religious practice). Mosques in the Jabal Nafūsa, in some cases somewhat fortified themselves, were part of a system of surveillance and control of the visible landscape. The question of how visibility of a landscape might structure settlement is complicated. Visibility analysis with GIS has the potential to inform this discussion. Global Digital Elevation data freely available for download from the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) mission34 can be used to assemble a 30-­m-­resolution digital elevation model of the region. Coordinates taken from the GeoNames​.org database of toponyms, and features visible on Google Earth imagery allowed for the identification of five mosques and five towns in the Jabal Nafūsa mentioned in Tasmīya Shuyūkh Jabal Nafūsa wa-­Qurāhum and Allan’s architectural survey (figure 2.2). Unfortunately, precise dating for these structures is unavailable, and conclusions must be read as broadly characteristic for the region, rather than for a specific period. The calculation of landscape visible I BĀ Ḍ Ī B O U N DA R I ES A N D D EF EN S E I N T H E JA BA L N A F Ū S A ( LI B YA )

37

Figure 2.2. Towns discussed and analyzed (after Lewicki 1955).

from each location is straightforward. This metric, called a viewshed, displays the pixels with an unobscured line of sight from the specified observer point.35 Viewsheds were also calculated from the five mosques and towns, and visibility from points randomly generated from “highland” locations (at elevations above 500 m). Visibility was calculated using an assumed observer height of 1.5 m and an outer limit at 10 km. The results of this analysis confirm what common sense would tell us: visibility is greatest from locations along the edge of the ridge. Visibility from the settlements analyzed (the mosques and towns) is greater than from the average of a random sample (3.652427 km2). As we saw above, visibility from the mosques (table 2.1) appears more extensive than from the towns (table 2.2). In the instances where visibility from both a town and its associated mosques can be analyzed (at Fursaṭā and Tmizda), the mosques are able to survey a more extensive part of the landscape than can the town itself (figure 2.3). In both areas, none of the sites around the towns is visible from any of the other sites, and perhaps at Fursaṭā mosques were located to parcel out areas to be monitored. This provides tentative evidence that the location of satellite mosques served a strategic purpose. They may have 38

A N T H O N Y J. L AU R I C ELL A

T able 2.1. Mosques Site

Visibility (km 2)

Fursaṭā-­Taghlis Mosque

Fursaṭā-­Ḥawāriyīn Mosque

Tmizda-­Ḥawāriyīn Mosque Ineri Mosque

Jami Ṭarī’gh Mosque

20.72426 7.8189

Direction

East; West

Southeast; Southwest

6.54666

East

5.46258

Northeast; Southwest

18.42592

Northeast

T able 2.2. Towns Site

Nālūt

Fursaṭā Jādū

Kabāw

Tmizda

Visibility (km 2)

Direction

6.82786

Southeast

17.16363

East

6.21574

West; South

4.80206

Northeast

3.28136

North; South

been deliberately placed to maximize visibility from the highland. Very high visibility was possible from sites at the very forward edge of the mountain range, but as we saw above, this was avoided because they could be spotted from the lowlands. Additionally, we can note that the two most important towns within the Jabal Nafūsa, Nālūt and Jādū, the east and west centers of the region, have strategically oriented viewsheds. Visibility from Nālūt is the most extensive, which suits a regional capital. The visibility from Jādū is lower, but is oriented to the west, while Nālūt’s commanding visibility is trained toward the towns of the eastern Jabal Nafūsa. Thus, these two “bookends” of the Jabal Nafūsa region appear to feature visibility directed toward each other, effectively enclosing the region. A more elaborate understanding of these dynamics in the Jabal Nafūsa would require a deeper picture of settlement in the region, including accurate locations of mosques, watchtowers, and fortified granaries. While it is no substitute for traditional survey, systematic study of satellite imagery has the potential to locate additional sites if fieldwork remains unfeasible. These basic analyses are only a preliminary step toward understanding surveillance and defense in the region. Further research incorporating Prevost’s authoritative new study will obviously problematize these results. I BĀ Ḍ Ī B O U N DA R I ES A N D D EF EN S E I N T H E JA BA L N A F Ū S A ( LI B YA )

39

Figure 2.3. Visibility from both the town of Tmizda and the Ḥawāriyīn mosque associated with Tmizda. CONCLUSIONS

Borders in the Jabal Nafūsa are of two types: internal and external. The landscape provided a natural barrier to the north, while the south was bounded by the natural contours of the escarpment. The east and west boundaries correspond to major economic centers on trade routes connecting the Jabal with the oases of the Sahara and sub-­Saharan Africa. The definition of eastern and western boundaries suggests a contraction away from the Roman forts. Connections to outside powers, while beneficial at certain moments in the past, offered little to the Jabal Nafūsa during the early Islamic period. This is perhaps explained by the fact that, at this time, defensive organization was internally directed, rather than being planned and coordinated by outside rulers. Even if the limes system had largely declined in effectiveness before the arrival of the Vandals, the Jabal Nafūsa was uniquely distant from power centers both before and after the arrival of Islam. The region had been entangled in longstanding antagonism from the more cosmopolitan and Romanized coastal populations. After the Islamic expansion into North Africa, the Jabal Nafūsa came into conflict with the mainline Muslim authorities in Qayrawān. 40

A N T H O N Y J. L AU R I C ELL A

Rates of taxation were high and payment was often mandated to be in the form of slaves. The Jabal Nafūsa faced political isolation. With even less central authority to appeal to in the early Islamic centuries, the contours of the landscape provided both regional boundaries and sites for defensive structures. Internal borders existed on two scales: settlement and subregional. While evidence is limited, we can describe a town structure consisting of a fortified gasr and a nearby mosque. Mosques were placed at strategic vantage points, often along the edge of the highland ridge. The watchtowers described by Despois (and unfortunately not locatable today without an updated survey) likely served to demarcate the edge of a town’s territory. Finally, the region was divided into east and west, a division that likely predated the beginning of the Islamic period. This division corresponds to no clear geographic or topographical boundary, meaning it must have been understood in a purely abstract and locally defined way. Defensive strategy in the Jabal Nafūsa was determined by the desire for both religious and political autonomy and was shaped by the landscape. It is possible to argue that the system of boundary definition and defense was a product of the Ibāḍī community grafted onto local cultural forms and the preexisting system of defense—­that is, influenced by the particularities of the landscape and Roman limes system. The Ibāḍī system as seen in the Jabal Nafūsa possibly served as the template for the similar patterns of defense seen in other regions where Ibāḍīsm took hold. Attempts to identify Islamic sectarianism archaeologically has focused, in the case of the Ibāḍī, on the form and placement of mihrabs (prayer niches).36 Understanding Ibāḍī mosques as loci for surveillance and defense broadens our conception of how these mosques may have functioned. It may be that what is distinctive emerges (if at all) at the level of landscape. Jerba, another Ibāḍī stronghold off the coast of Tunisia, is another naturally defensible landscape. The island is famous for its extraordinary number of mosques: according to a 1941 count, there were 288 for an island measuring 512 km2. The majority of these structures are very small rural mosques located in the countryside and associated with small villages. The Archaeological Survey of Jerba suggests a dispersed, mostly rural pattern of settlement on the island.37 Mosques identified in the survey—­for example, the mosque of al-­Hara—­included towers, cisterns, and other structures aside from the prayer hall. As in the Jabal Nafūsa and the Mzab, Jerban mosques also play a role in the defense of the island. Prevost describes the coast as ringed by a series of small mosques, lacking minarets but with spaces on roofs for observation, tasked with watching for the approach of enemies. These mosques, like the I BĀ Ḍ Ī B O U N DA R I ES A N D D EF EN S E I N T H E JA BA L N A F Ū S A ( LI B YA )

41

ribāṭs built by the Aghlabids, are frequently dedicated to a saint or holy man. When a coastal mosque sees a threat, it could signal a warning to the interior of the island with fire.38 Farther inland, the warning message would alert the island’s inhabitants to seek refuge in one of the larger, fortified mosques set back from the coast. These mosques are complementary to the coastal mosques and are more elaborate, with meurtrières and machicoulis to repel attackers, as well as cisterns and storage areas for food in case of a prolonged siege.39 The findings of the Archaeological Survey of Jerba supports the picture of how mosques functioned in a defensive system. The findings suggest mosques suited for defense (or those built underground) are located near the coast, while others are built inland. This system of defense is based on a distinction between border and center, which are linked through fortified religious structures. Perimeter watchtower structures are also associated with religious figures and strategic visibility of the landscape. The mosques of the island share many characteristics seen in the Jabal Nafūsa and Mzab (an isolated Ibāḍī stronghold in Algeria)—­ for example, the use of a subterranean construction style.40 Defense of any landscape necessarily involves some amount of surveillance, and architectural styles need not be transmitted by intentional vectors. Furthermore, mosques with defensive features are not unique to Ibāḍī landscapes. Nevertheless, if we accept a system of defense associated with mosques in the Jabal Nafūsa, it is tempting to see a distinctly Ibāḍī “fingerprint,” perhaps expressed in terms of visibility, that influenced the distribution of mosques on the island. The boundaries seen in the Jabal Nafūsa were outward expressions of religious and material concerns. With the integration of Ibāḍī Islam into the region, religious architecture became a new avenue through which defensive concerns, and a local desire for autonomy, could be realized. Evidence discussed here suggests that while the Jabal Nafūsa was isolated and insular, its inhabitants were participating in long-­distance trade. Furthermore, there was conflict within the region itself. The town of Sharūs was destroyed in 1100, apparently by the inhabitants of a neighboring town, and visibility analysis provides tentative evidence that settlements may have had an interest in monitoring each other. The competition for control of terminal points on the trade routes may be one explanation for this internal conflict. It is also possible that the particularities of Ibāḍī Islam led to conflict between cities. The faith upheld the idea that the most virtuous member of the community should lead, and this may account for the nested nature of territories and the boundaries between them in the region. The frontier between the Ibāḍī Jabal Nafūsa and the outside world existed to uphold the autonomy of the Nafūsans. High up on the escarpment they were 42

A N T H O N Y J. L AU R I C ELL A

safe from the tax collectors and untrusted and uninvited religious authorities, who were the perennial threat posed by the political entities on the coast. The inhabitants of this remote region were well versed in what James C. Scott has called “the art of not being governed.”41 Much like Scott’s “shatter zones” outside the range of central authority, the Jabal Nafūsa was topographically difficult to access, and sheltered a range of religious dissenters.42 While the highlands themselves articulated a border with the lowlands, and deep valleys separated different settlements, the mosques of this region served as focal points of surveillance. In addition to serving as places of worship, mosques established observable borders that separated and enclosed Ibāḍī spaces. NOTES

Thanks to Donald Whitcomb and Fred Donner who long ago encouraged me to consider what archaeology could say about the Ibādī. I am grateful to Asa Eger for his invaluable comments and suggestions. 1. For detailed geographical description of the region as related to the time of Islamic expansion see, Kaegi 2010, 41–­64. 2. The tension between centralized coastal power and the Jabal Nafūsa is a longue durée pattern and one that most recently manifested during the Libyan Revolution of 2011 when rebel groups based in the region attacked government forces in the foothills (Al-­Jazeera 2011). 3. See Savage 1997, 97–­100. 4. Abun-­Nasr 1987, 33–­35. As fellow monotheists and “people of the Book” (ahl al-­Kitāb), indigenous populations could have expected to be spared persecution, as Christians and Jews were in Syria and Palestine. Abun-­Nasr describes negative beliefs about the Berbers circulating at the time of the conquests that may account for this. 5. For general historical surveys of Ibādī Islam, see especially Lewicki 1971. More recently: Hoffman 2012; Prevost 2010; Gaiser 2010. 6. For example, the town of Idunat. See Lewicki 1955, 51. 7. Ibid., 57. See also Savage 1997, particularly 89–­112. While it is possible that some mosques in the region actually are reused churches, Prevost argues that in most cases the toponyms simply represent a memory of the pre-­Islamic Christian past. See Prevost 2012, 325–­347. 8. Mostly thoroughly described by Eger 2015. 9. Sjöström 1993, 4–­19. 10. Mattingly 1994, 15. 11. McLachlan, “Djabal Nafusa.” 12. Prevost 2016. I BĀ Ḍ Ī B O U N DA R I ES A N D D EF EN S E I N T H E JA BA L N A F Ū S A ( LI B YA )

43

13. Lewicki 1955, 83–­85. 14. Ibid., 88. 15. Ibid., 126. See also Warfalli 1981, 54–­63. 16. Lewicki 1955, 44–­45. 17. Ibid., 47. For Sharūs, see Lowick 1974, 14–­19; Warfalli 1981, 114–­142. For Wīghū, see Warfalli 1981, 142–­147. 18. Lewicki 1955, 72. 19. Warfalli 1981, 82. 20. Lewicki 1955, 64. 21. Lewicki 1955, 96. See also Martin 1969. 22. Savage 1997, 79. 23. Lewicki 1955, 109. 24. Ibid., 111–­112. 25. Mattingly 1994, 88–­89. 26. Ruprechtsberger 1993, 39. 27. Sjöström 1993, 103–­104. 28. See Despois 1953, 82–­85. 29. For discussion of the use of fortified granaries in different regions and time periods in the Maghrib and Europe, see De  Meulemeester 2015; De  Meulemeester 2005, 609. 30. Despois 1935, 209–­210. 31. Allan 1972–­1973, 167. 32. Lewicki 1955, 44, 81. 33. Allan 1972–­1973, 167. 34. ASTER data are produced through a collaboration between the Japanese Ministry of Economy, Trade, and Industry and NASA (the US National Aeronautics and Space Administration). 35. This approach necessarily assumes vision and simple visibility as the dominant sense in both surveillance and experience of the defensive landscape. Other factors, for example the distance the sound of the adhān could travel, might have also been considered. 36. Insoll 1999, 45. 37. See Holod and Kahlaoui, chapter 3, this volume. 38. Prevost 2006, 9–­10; Prevost 2009, 220. 39. Prevost 2006, 10–­11. 40. Prevost 2009, 221–­222, 226. 41. Scott 2009. 42. Ibid., 7–­8.

44

A N T H O N Y J. L AU R I C ELL A

REFERENCES

Abun-­Nasr, Jamil M. 1987. A History of the Maghrib in the Islamic Period. New York: Cambridge University Press. Al-­Jazeera. 2011. “Fighters Launch Assault in Libya’s West.” July 28. http://​www​ .aljazeera​.com/​news/​africa/​2011/​07/​20117287598328864​.html. Allan, J. W. 1972–­1973. “Some Mosques of the Jabal Nafūsa.” Libya Antiqua 9.10: 147–­169. De Meulemeester, Johnny. 2005. “Granaries and Irrigation: Archaeological and Ethnological Investigations in the Iberian Peninsula and Morocco.” Antiquity 79: 609–­615. De Meulemeester, Johnny. 2015. “Ethno-­Archaeological Approaches to Medieval Rural Settlement in Spain and Morocco.” In Studies in the Archaeology of the Medieval Mediterranean, edited by James G. Shryver, 155–­173. Boston, MA: Brill. Despois, Jean. 1935. Le Djebel Nefousa. Paris: Larouse-­Editeurs. Eger, A. Asa. 2015. The Islamic-­Byzantine Frontier: Interaction and Exchange among Muslim and Christian Communities. London: I.B.Tauris. Gaiser, Adam. 2010. Muslims, Scholars, Soldiers: The Origin and Elaboration of the Ibadi Imamate. New York: Oxford University Press. Hoffman, Valerie J. 2012. The Essentials of Ibadi Islam. Syracuse, NY: Syracuse University Press. Insoll, Timothy. 1999. The Archaeology of Islam. Malden, MA: Blackwell. Kaegi, Walter. 2010. Muslim Expansion and Byzantine Collapse in North Africa. New York: Cambridge University Press. Lewicki, Tadeusz. 1955. Études Ibadites Nord Africaines. Warsaw: Panstwowe Wydawnictwo Naukowe. Lewicki, Tadeusz. 1971. “The Ibadites in Arabia and Africa.” Cahiers d’Histoire Mondial 13.1: 51–­81. Lowick, N. M. 1974. “The Arabic Inscription on the Mosque of Abu Mahruf at Sharwas.” The Society for Libyan Studies Fifth Annual Report: 14–­19. Martin. G. B. 1969. “Kanem, Bornu, and the Fezzan: Notes on the Political History of a Trade Route.” Journal of African History 10: 15–­27. Mattingly, David J. 1994. Tripolitania. Ann Arbor: University of Michigan Press. McLachlan, K. S. Encyclopaedia of Islam, Second Edition. s.v. “Djabal Nafusa.” Norris, H. T. 1953. “Cave Habitations and Granaries in Tripolitania and Tunisia.” Man 53: 82–­5. Prevost, Virginie. 2006. “L’urbanisme des Ibadites maghrébins medievaux et le cas particulier de Djerba.” Rocznik Orientalistyczny 59: 5–­16. Prevost, Virginie. 2009. “Les mosquées ibadites du Maghreb.” Revue des Mondes Musulmans et de la Méditerranée 125 Juillet: 217–­232. I BĀ Ḍ Ī B O U N DA R I ES A N D D EF EN S E I N T H E JA BA L N A F Ū S A ( LI B YA )

45

Prevost, Virginie. 2010. Les Ibadites: De Djerba á Oman, La Troisième voie de l’Islam. Turnhout, Belgium: Brepols. Prevost, Virginie. 2012. “Des églises byzantines converties á l’islam? Quelques mosquées ibadites de djebel Nafûsa (Libye).” Revue de l’histoire des religions 3: 325–­347. Prevost, Virginie. 2016. Les Mosquées Ibadites du Djebel Nafūsa. London: Society for Libyan Studies. Ruprechtsberger, Erwin M. 1993. Die romische Limeszone in Tripolitanien und der Kyrenaika Tunisien-­Libyen. Stuttgart: Limesmuseum Aalen. Savage, Elizabeth. 1997. A Gateway to Hell, a Gateway to Paradise: The North African Response to the Arab Conquest. Princeton, NJ: The Darwin Press. Scott, James C. 2009. The Art of Not Being Governed: An Anarchist History of Upland Southeast Asia. New Haven, CT: Yale University Press. Sjöström, Isabella. 1993. Tripolitania in Transition: Late Roman to Early Islamic Settlement. Brookfield, VT: Avebury. Warfalli, Muhammad Salim Muqayyid. 1981. “Some Islamic Monuments in Jabal Nafusa.” PhD diss., School of Oriental and African Studies, University of London.

46

A N T H O N Y J. L AU R I C ELL A

3 For Ifrīqiya, the Maghrib, and the eastern coast of al-Andalūs, a ribāṭ can be defined as a coastal point of defense meant to host ghazīs, warriors for the faith or volunteers on the basis of religious commitment. From the eighth century onward, ribāṭs stretched along the seashore of Islamic Maghrib from Tripoli (Ṭarāblus) up to the east coast of al-Andalūs. These were typically connected to another urban structure with an immediately supporting rural environment, or they could be isolated structures supplied from the sea. In one sense, they functioned in much the same way as stations along the ‘Abbāsid-­Byzantine land frontier (thaghr) in southern Anatolia-­northern Syria.1 In another, they could be seen as a continuation of the sea frontier of the eastern Mediterranean. The best-­ studied Maghribī frontier is that of the Ifrīqiya coast and its ribāṭ culture. It has been characterized as a series of fortified establishments, supported by a surrounding countryside, from which raids would be launched upon Byzantine military naval traffic or merchant shipping. These bases in turn served to anchor and expand nascent Muslim communities in the area.2 In addition to the mainland coastal frontier, the central and western Mediterranean is filled with islands, large and small, from Sicily, Sardinia, Malta, and Majorca and Minorca to Pantelleria. Any island has a ready-­made geographical frontier—­its coasts—­and yet is still readily accessible by sea. These islands also played a crucial role in the drawn-­out struggle for hegemony, control of trade,

Guarding a Well-­ Ordered Space on a Mediterranean Island Renata Holod and Tarek Kahlaoui

DOI: 10.5876/9781607328773.c003

47

Figure 3.1. The Central Mediterranean and North Africa (map: Michael Frachetti).

and territory in the central Mediterranean for the several polities of the medieval and early modern periods: the Byzantines and their successors, the Normans, the Kingdom of the Two Sicilies, and the Spanish, and the various successive Islamic polities of Ifriqīya, the Aghlabids (800–­909 CE),3 the Zirids (972–­1148), and the Hafsids (1229–­1574), up to the Ottomans. This sea frontier, and the specific details of its island-­by-­island dynamics forms a complex picture in the history of the central Mediterranean, and deserves fuller, updated treatment.4 For the present study, we center our attention on Jerba alone (figure 3.1), where matters appear to have been more complex because of its geography, and because its historical narrative for the medieval and early modern centuries must encompass not only the main events as presented by the Sunni-­majority texts but also an alternate sectarian (Khārijī/Ibāḍī) narrative and interpretations of events.5 THE ISLAND OF JERBA

Lying directly off the coast of present-­day southern Tunisia, the island of Jerba is geologically simply an extension of the mainland. It has an area of 198 square miles and is separated from the mainland both by the Gulf of Bu Ghara, an underwater trench, or wadi, running along the island’s southern 48

R EN ATA H O L O D A N D TA R EK K A H L AO U I

Figure 3.2. Topography of Jerba (map: Michael Frachetti).

coast, and by treacherous shallows.6 Over the course of its history, the island was indeed accessible from the mainland by lighters, if not by large boats or ships. In fact, the location of a ferry point at Ajīm apparently served the island through all periods. Further, a causeway (al-­qantara) first constructed during the Roman (Flavian) period made the island more accessible from the mainland to the south. This connection was dismantled periodically during later periods. Perhaps more surprising, close by the causeway, along the south coast, was also another access point, known only to the local inhabitants. It was a ford, called the tricky passage (majāzun khabīthun), where the channel could be crossed on horseback (figure 3.2).7 From the north, there was no good deep-­water anchorage or port, and ships had to dock at a jetty. Thus, the very geography of the island allowed it to function as a refuge. The present study is limited mainly to a detailed consideration of the island’s frontiers through the medieval and early modern periods. It was then that the natural defenses were utilized and strengthened by the new social order instituted across the territory of the island. Settlement patterns recorded during our archaeological survey brought to light yet other frontiers that had developed on the island itself for ideological, social, political, or military reasons. G uarding a W ell - ­O rdered S pace on a M editerranean I sland

49

Figure 3.3. Probabilistic and purposive survey with transects indicated (map: Michael Frachetti).

The archaeological survey of Jerba recorded all phases of occupation from prehistoric periods through the end of nominal Ottoman control in the middle of the nineteenth century.8 The survey was carried out over the territory of the entire island, primarily as a stratified random sampling, with additional purposive sampling and test trenches, and was further supported by textual study and archival sampling (figure 3.3). For all three medieval periods (700 through 1500 CE) up to 354 sites were identified, surveyed, recorded, and studied: domestic sites large and small, olive presses and ceramic kilns, mosques and shrines, forts, lookouts and landings. For the early modern period (1500–­1700 CE), some 130 sites of similar categories were noted. Dating of sites was based on ceramic chronotypologies,9 the available epigraphic record, and published and archival sources. The periods in question are the medieval periods—­ EM (700–­ 1050), MM (1050–­ 1300), LM (1300–­1500)—­and the early modern—­Mod 1 (1500–­1700). For them, a markedly different landscape can be reconstructed in the type and location of 50

R EN ATA H O L O D A N D TA R EK K A H L AO U I

T able 3.1. Sites Recorded from the Probabilistic Survey, 500–­1900 CE Period

N

Late Antique 2 / LA2 (500–­700)

Early Medieval / EM (700–­1050)

85

98

Middle Medieval / MM (1050–­1300)

137

Modern 1 / Mod 1 (1500–­1700)

130

Late Medieval / LM (1300–­1500) Modern 2 / Mod 2 (1700–­1900)

119 90

sites (table 3.1). The earlier (Roman and late antique) settlement order of villas served by cities and towns appears to have been attenuated, or became defunct.10 The recurrence of the Justinian plague raging throughout the Mediterranean from the 540s and periodically to 750 was then followed by the impact of the Islamic invasions.11 Major city sites of the Roman and late antique periods on the island, and in particular Meninx with its port, ceased functioning as full-­fledged urban centers, remaining only as fields for spolia and loci of suburban villages and weekly markets.12 By the 660s, according to Islamic conquest narratives, the island of Jerba was considered a place of rural habitation (qariya) rather than the locus of cities.13 While it is difficult to date any dramatic shifts or changes in local ceramic production for the next two centuries—­the kilns were still in the same locations and were likely worked by a continuity of artisanal labor—­new settlement appears to have taken place in a different manner. A new population of settlers begins to utilize the spaces emptied of villas and towns. Their living patterns make the space of Jerba a landscape of farm communities surrounding their own mosques. And by the end of the ninth century, this space could be designated as decidedly sectarian: mainly Ibāḍī Muslim but also Jewish.14 The fact that the newcomers settled differently helps to identify these spaces cumulatively as an internal frontier—­a well-­defended and controlled countryside. The forts on the island, and there were two by the end of the periods under review (EM to Mod 1), likely never belonged to this order of settlement. These were not controlled by the Ibāḍī settlers, and can be definitely understood as intruders into the reconfigured island space. Without major forts and towns tied to these rural settlements, the area of the island, then, cannot be classified as the classic frontier (thaghr) of ribaṭs from which to launch attacks on the enemies of Islam, as was the case of the Ifriqīya sea frontier nearby. Nonetheless, one must consider it a frontier, albeit one along sectarian as well as confessional lines. G uarding a W ell - ­O rdered S pace on a M editerranean I sland

51

Jerba as Ibā ḍ ī Space

Ibāḍī sectarian separatists—­a branch of the Khārijī, the group that had separated itself from the Islamic community about 685 CE—­launched a polity based on an interpretation of leadership different than the interpretation of both the Sunni majority and the Shī‘a minority.15 As one of the factions of the Khārijī group, the Ibāḍīs found their way in early Islamic times to North Africa, where they were able to control several territories in the transitional climatic zone between the Mediterranean coast and the pre-­Sahara Desert. There, they established the Rustamid dynasty (767–­909 CE), centered in Tāhart (Tiaret) in central modern-­day Algeria. The time was highly tumultuous for Ibāḍīs as they took part in revolts and divided into two sub-­factions, the Wahbī and the Nukkārī. The events describing the arrival on the island of Jerba of new settlers who belonged to the Nukkārī faction of Ibāḍī are mentioned in the main Ibāḍī chronicle for the island, Kitāb Siyar al-­Aʾimma wa Akhbārihim. Writing his history during the eleventh to twelfth centuries CE, Abū Zakariyā b. Abū Bakr could be considered as the main Ibāḍī chronicler of medieval Jerba through this work, also known as Tārīkh Abū Zakariyā.16 Granted, the author belonged to the Wahbī faction, the other and competing faction of the Ibāḍī who settled on the island, probably in the tenth century CE, and would, therefore, not have focused much on the processes of Nukkār infiltration into the island’s territory in the late eighth–­ninth century. Yet, these events and associated power struggles are mentioned by this Ibāḍī text alone, and not by any standard (Sunni) historical works describing this period. Such an omission would indicate that the island space was already out of mainstream consideration, and existed mainly within the sectarian world of the Ibāḍī by the time of his writing, if not earlier. Abū Zakariyā’s account allows us to reconstruct some aspects of early medieval life on Jerba. The island came to be settled mainly by some of the Zwāgha Berbers who were Ibāḍī, but Nukkārī and, therefore, anti-­Wahbī. Their arrival does not seem to have been opposed actively by the much-­diminished local population of the late eighth–­ninth century. In fact, the new arrivals took over at least one of the preexisting Byzantine forts, whose ruins are still visible today. Standing at the first rise from sea level, the two sites (today called Tala and Ghardāya) were built to the same plan. The structure mentioned as “fort of Ghardāya” is still identifiable today in a southern part of the island known as Ḥumat Ghardāya (figure 3.4).17 Since the leader of the Zwāgha, the first Ibāḍī group arriving on Jerba, was referred to by Abū Zakariyā as a “settler” (maʿqil), likely this arrival marked the influx of a new population to the island. The 52

R EN ATA H O L O D A N D TA R EK K A H L AO U I

Figure 3.4. Transitional sites from LA 2 to EM with forts indicated (map: Michael Frachetti).

current toponym of Banī Maʿqil may still echo this new settlement, and might be referring to the old core of the Nukkārī part of the island.18 The island, then, functioned as a place of refuge, or even exile, for these rebels. Ironically, this faction was rebelling against the Ibāḍī Wahbī state of Tāhart, itself considered a rebel state by the official (Sunni) historians. It should be stressed, however, that there are no available textual accounts from the Nukkārī side of this or other events. Thus, available Ibāḍī narratives of events are actually Wahbī, first and foremost. Other medieval sources really do not address the order of settlement on the island. For the later periods, and especially for the beginning of early modern period, Muḥammad b. Zakariyā al-­Bārūnī (d. 1589) gives us an idea of the island territory. A scholar from Jabal Nafūsa who became a student of Abū Sulaymān al-­Tlātlī, one of Jerba’s main religious figures in the mid-­sixteenth century, G uarding a W ell - ­O rdered S pace on a M editerranean I sland

53

al-­Bārūnī, wrote Nisbat al-­Dīn as an appendix to al-­Shammākhī’s Kitāb al-­ Siyar.19 It is a rare Ibāḍī source for the sixteenth century, covering especially the unstable relations between the local Jerban religious and political elites, and the Ottoman forces led by Dargūth Bāshā—­the infamous Turgut Reis (1485–­1565—­during the 1550s). The source shows that the Ottomans succeeded far more quickly than previous intruding powers in hemming in their opponents on the island. Largely, they were able to ensure local support critical for future shifts in the religious and political structures, thereby rendering the island less and less independent from outside power. Sa‘īd al-­Barunī, writing in the middle of the eighteenth century, in his quintrain (mukhammas) laments the fact that many of the Ibāḍī/Nukkārī turned Sunni, or as he says, “have diverted.” This “sunnifcation” of half the territory of the island, then, allowed for an easier extension of Ottoman control.20 Finally, and mostly important for the present study, the Epistles (Rasa’il) of al-­Ḥīlātī al-­Jirbī (d. 1688–­1689?) provides a list of the Wahbī mausoleums and memorials (mashāhid and adriha) on seventeenth-­century Jerba21 as well as a survey of the major mosques functioning along the island coasts and of specific interest to the Wahbī.22 In addition, the Epistles affirms the function and the political role of fortifications as outposts by outsiders during the sixteenth and seventeenth centuries, notably of Qashtīl al-­Wādī, the southern fortress, newly constructed by the Sicilian/Spanish forces.23 JERBA IN A REGIONAL SETTING

The island of Jerba begins to appear in non-­Ibāḍī Arabic sources more frequently and extensively by the eleventh century, perhaps as a function of its involvement with the Faṭimid/Zirid world of Ifriqīya and maritime traffic. For the geographer, al-­Bakrī, the earliest mention of Jerba was as a base for pirates and thus as chiefly a threatening place.24 Zirid attempts launched out of Ifrīqiya to curb Jerban piracy—­and to extend Zirid control over the Jerba Sea (the Gulf of Bu Ghrara), if not over the territory of the island itself—­were ongoing throughout the second half of the tenth to the early eleventh centuries.25 Taking into consideration the commercial network detailed by al-­Bakrī, and therefore the importance of Jerba’s southern and northern harbors and anchorages, it becomes clear why the Zirids sought to secure these sea lanes. Two major campaigns for control of the island space are reported by al-­Tījānī (citing the Zirid source, Abū al-­Ṣalt)26 and by Ibāḍī sources.27 In parallel, and connected with the trade in wool, in textiles, and often in dyestuffs throughout the medieval and early modern Mediterranean regions, 54

R EN ATA H O L O D A N D TA R EK K A H L AO U I

and beyond, was the far-­flung network of Jewish communities. The activities of this diaspora have been reconstructed through disparate documents, ranging from letters to court records, found in the Cairo Geniza.28 Several instances dealing with the Jews of Jerba can be cited from these materials. First, the Norman conquest of Jerba in 1135, and the resulting capture of many inhabitants, is recorded through the case of a Jew from Jerba who was ransomed from this captivity by a benefactor from Egypt in 1136.29 The impact of the Norman intrusion on the regular rhythms of life on the island is perhaps also reflected in the mention of a famine in 1136.30 The close ties to Egypt are evident in a record where Jerban Jews receive clothing from a community chest there, while in another case clothing is ordered from Egypt.31 Still, trading continued along the island’s coasts and across to Sicily and Italy (and beyond) parallel to sieges, invasions, and other warring activities on the island and its surrounding waters.32 As discussed by Jehel, the volume of this trade is reflected in several archival collections of customs and tax lists, ship registers, and the like, located in Sicily, Genoa, and Venice. The Sicilian notarial records mention the transportation of cereals from Palermo to Jerba, and deal specifically with the cost of transportation (nolis).33 Catalan vessels traded through Sicily: for example, the Sanctus Nicolaus was a grain ship to Jerba in 1297–­1298. The register of Adamo de Citella (1290s–­1310) also records that in 1298 wheat was sent to Jerba.34 Brunschvig, a key modern authority writing about the Maghrib, turns to Bartholomaei de  Neocastro’s Historia Sicula in particular. Written in Latin before 1293 CE,35 the work speaks of a devastating Sicilian campaign against the island of Jerba. A slightly later source, Ramón Muntaner’s Crónica Catalana, is an account written in Catalan by the military commander of the fort on the island, and later a representative of Frederic I, King of Sicily and Aragon.36 It corroborates the fact that it was the Sicilians who undertook the building of a fort, most likely the northern fortress, called Burj al-­Kabīr but subsequently renamed Burj al-Ghazī Muṣṭafā by the Ottomans.37 This chronicle also independently bears witness to the Hafsid siege of this fort in 1306–­1307. That siege, as well as the entire military expedition from Tunis to Jerba, was described by al-­ Tījānī. The latter accompanied Ibn al-­Liḥyānī, the vizier of the Hafsids, and his troops on this expedition. Al-­Tījānī describes the stay of the army on the island for more than two months. The expedition had as its aim to dislodge the Sicilians from their stronghold on the island in this great northern fort, but apparently failed.38 By the fifteenth century, there were other attempts by the Kingdom of the Two Sicilies to extend their control over the central Mediterranean, the North African shores, and Jerba. The naval G uarding a W ell - ­O rdered S pace on a M editerranean I sland

55

campaign sent out by Alfonso V to Jerba between 1425 and 1432 was one such attempt. It ended disastrously, beaten back as much by the island’s geographical defenses—­especially the shallows—­as by opposing forces. The defeat was reported by Bartolomeo Facio and has been summarized in modern studies such as Brunschvig’s La Berbérie orientale.39 Ottoman expansion into the waters of the central Mediterranean and along the shores of North Africa brought Jerba into a different and wider trade and political orbit, and at the same time opened its territory to new Sunni authorities both in Istanbul and Tunis. The Kitāb-­ī Baḥriye, written by the Ottoman sea captain Piri Reis between 1520 and 1526, includes a chapter on Jerba, with a map, and is a primary Islamic source about the island at the beginning of the sixteenth century. It provides a unique insight from an Ottoman point of view about the internal conflicts and the defensive strategies on the island. The Nukkārī population seems to have had their own shaykhly institution (mashaykhat), and Banī Smūmin shaykhs of the Wahbī were not necessarily recognized across the island territory. Piri Reis makes note of the defensive strategy and commercial impact of the Jerbans’ decision to cut the old Roman causeway (al-­qantara). According to him, the Wahbī shaykh, Yaḥyā, dismantled the causeway in its center in order to prevent easy access to the island from the mainland, and also to allow the movement of small vessels along the entire southern coast of the island. In fact, the reconstruction of the bridge by the Hafsids, which had been undertaken earlier in the fifteenth century, seems to have prevented east–­west sailing along the southern coast, the center of the Nukkārī settlements, including the location of their major market, Sūq al-­ Qiblī (the Southern Market).40 Nonetheless, by the middle of the seventeenth century, Ottoman control over the northern and southern forts on Jerba comes to be well established, and urban life begins to return to the territory of the island.41 Specifically, the northern market, Ḥumat al-­Sūq, eventually becomes the locus of a new congregational mosque (masjid al-­jami‘), and therefore can be rightly called a city (madinah), while the southern market, Sūq al-­Qiblī, is less and less frequented. Finally, Ibn Maqdīsh wrote Nuzhat al-­ʾAnzār f ī ʿAjāʾib al-­Tawārīkh wa al-­ʾAkhbār42 at al-­Zāwīya al-­Jimniya on Jerba sometime in the mid-­eighteenth century. Since the zāwīya was the major center from which the Sunni rulers of Tunis coordinated their policy of the conversion of the Nukkārī of Jerba to Sunnism, especially during the critical period of the eighteenth century, the report of Maqdīsh is a Mālikī witness to the tense religious factionalism on the island.43 The source links the introduction of Sunnism into the island with the extension of political power by the Ottoman rulers of Tunis.44 These 56

R EN ATA H O L O D A N D TA R EK K A H L AO U I

years then mark the beginning of the waning of a true internal frontier: the sectarian space that had come to define most of the surface of the island for the previous ten centuries. ORDER OF SETTLEMENT

Although the sources give us glimpses of the history of the island, it was our survey that allowed for a detailed mapping of the territory. For the present purposes, we present different types of sites, both domestic and religious, as a way of identifying the frontiers, both internal and external, operating within the frame of the medieval and early modern centuries on Jerba. Domestic Sites

Early aerial photographs of Jerba, and even those as late as the 1990s give us a view of the cellular structure of rural Jerba, with every house (dar) or (hūsh) and estate (manzil) surrounded by its own irrigated garden(s) (saniya) and walled by a boundary wall of mixed rubble and mud (ṭābiya) overplanted with bramble or cactus.45 Narrow lanes connected to the main roads of the island. Such a dense, enclosed landscape was also referred to by al-­Tījānī when he recounted the march of the Hafsid army through the island from south to north: “We continued walking between palm trees and other kinds of trees until we reached the site of al-­Qashtīl” (i.e., the northern fort).46 Walking (or riding) between the high walls would have been truly disconcerting, as the maze of lanes provided no vantage point from which to see the way ahead. Few complete plans of domestic sites of the three medieval periods (EM, MM, LM) were recovered during our survey; chief among those that were recovered is a farm (site K085) near Masjid Zayd, datable by its ceramic to as early as MM (figure 3.5). There, its surrounding boundary wall has not survived. The residential unit itself would have contained four units (bayt) arranged around an internal courtyard, while outside, its gardens would have been watered from a well, whose water was pulled up by draft animals into a basin and then emptied into channels irrigating plots. We are better served for the later periods, for which a fuller idea of domestic sites has survived, and more particularly in the center areas of the island with better access to the aquifer ( Jerba has no surface water at all). An estate at Qishi’in, datable from about the seventeenth century through the 1990s, can serve as an example, and shows the same reliance on enclosing domestic spaces with surrounding walls (figure 3.6). G uarding a W ell - ­O rdered S pace on a M editerranean I sland

57

Figure 3.5. Domestic Site (hūsh/dar) in close proximity to the Zayd mosque, datable from MM (drawing: E. B. Fentress and E. Tapp).

In this case, two residential multiunits (hush) form an estate (manzil). Areas were set aside for irrigated plots with their wells and basins, as well as for olive presses. In sum, while boundary walls of this estate could not have been defined as fortifications, they did mark out very clearly the frontiers between the extended family inhabiting the estate and outsiders.47 FUNCTION OF A MOSQUE SITE

All mosques on Jerba are small-­scale buildings. The building module seems to have been about 2 m, or an arm span. In general, there are two basic types of mosque plans: type 1, based on a hypostyle plan, and type 2, based on a nine-­ bay, four-­support plan. A smaller four-­bay, single-­support variant appears to be a scaled-­down version of the latter type (figure 3.7). Every mosque site itself encompassed more than the prayer hall. It is clear from the sites recorded in our survey, as well as from legal digests such as that discussed by al-­Shammākhī, that a variety of additional structures were 58

R EN ATA H O L O D A N D TA R EK K A H L AO U I

Figure 3.6. Qishi’in estate (manzil), datable Mod 1 and 2.

considered part of a mosque site, although they were differentiated from the prayer hall itself.48 Distinctions were drawn legally between the prayer hall itself (masjid)—­defined as a building built with walls, columns, and vaults—­and the rest of the mosque site. Funds intended for the upkeep of the prayer hall could not be used for any other areas of the site. The mosque site was understood as belonging to a particular group, legally defined as the congregation ( jama‘at al-­muṣallin). It was the congregation that built it, used it, and owned it. Thus, depopulation of an area and/or migration of the congregation away from the mosque would mean that the mosque site became abandoned, because that particular group was no longer present to maintain it. Abandonment within the Jerban landscape meant that their congregation had moved elsewhere and/ or probably turned Sunni. In this context, the poem written in the middle years of the eighteenth century by Saʿīd al-­Bārūnī is particularly evocative of the major political and ideological shifts on the island. It laments the wholesale abandonment of mosques and estates, and decries the establishment of a Sunni mosque in the new town.49 G uarding a W ell - ­O rdered S pace on a M editerranean I sland

59

Figure 3.7. Mosque typology: all periods.

Two of the several mosques recorded and studied by our team can serve as examples through which the function of a mosque site can be understood. The first site is the mosque site of al-­Hara (figure 3.8). It consisted of a nine-­bay mosque with a staircase minaret attached on the exterior, next to its entrance.50 The tower in the northeast corner of the interior was inserted some time after the initial structure had been erected. A watchtower, it would have formed part of the chain of such lookouts along the first rise from the southern shore. The mosque building stands alone on a platform raised several steps above the rest of the site. South of the mosque itself stood another group of structures—­four single cells and a rectangular hall—­by all accounts the remnants of a school or even a madrasa. The entire ensemble was surrounded by a perimeter wall and sat on a paved surface (an impluvium) with four well or cistern openings. The latter would have served as an emergency water supply for the community of this mosque. Separately walled and to the northwest lay a cemetery with its graves marked only by stones without any inscriptions. In this particular case, no epigraphy has been noted anywhere on the mosque site. In fact, there is no ornament whatsoever on any surface. Absent these, the dating of this site, as in the case of others like it, must come from a mosque chronotypology, and even more so from the surrounding domestic sites. The latter have to be dated by the ceramics picked up by our survey. In this case, however, the site is mentioned as one of the mosques in the itinerary of the Wahbī shaykhs in the seventeenth century, discussed below. From its location along the southern line, it could have been functioning earlier, probably already in the fourteenth century.51 The second site, that of Bani Maghzil, is also part of this same defensive line, and is similar in all main structural and layout features to the al-­Hara site 60

R EN ATA H O L O D A N D TA R EK K A H L AO U I

Figure 3.8. Al-Hara Mosque Site (Drawing: Elizabeth B. Fentress).

(figure 3.9). It seems to have had a longer history of use, however. Its renaming as a zawiya (religious school or retreat) would indicate that its community turned from Ibāḍī Nukkārī to Malikī/Sunni at some moment in the eighteenth to nineteenth centuries. THE RESPONSE OR DEFENSIVE LINE

The mosques of Jerba were first utilized to maintain the local Ibāḍī political, social, and administrative structure at a distance from the rule of the invaders. The defensive nature of mosques is not new and not unique to Jerba, however. Religious architecture in the Maghrib (and in Egypt) is characterized by architectural elements that projected a powerful defensive appearance. Whether in the case of the Great Mosque of Qayrawān, the Mosque of Mahdiyya, and even that of al-­Ḥakim in Cairo, the square and sometimes multiple towers (burj) and projecting portals suggest a resemblance to the ribāṭs such as those built along the North African coasts. Yet, such resemblances cannot be explained necessarily from a defensive point of view alone.52 In some cases, such as in Sousse, the proximity and the resemblance between the ninth-­century ribāṭ and mosque are striking, pointing to complexes built with the possible intention of using all buildings as refuge locations.53 G uarding a W ell - ­O rdered S pace on a M editerranean I sland

61

Figure 3.9. Mosque site of Bani Maghzil, with inserted watchtower in the prayer hall (drawing: E. Tapp and E. B. Fentress).

Still, the Jerban case is even more particular in the Maghribī context. The dominant mosque typology is not only vernacular and small, but also related to a social context that is dispersed and rural. The type of the Great or Congregational Mosque (masjid al-­jami‘) was completely absent on the island territory as a marker of the center of urban structure. The Ibāḍī ideological stand against the Friday prayer because of the absence of a righteous imam (al-­‘imam al-­‘adil) had been legitimized on the island territory through the early modern period (Mod 1) by the absence of nucleated settlements with more than “40 able men.”54 More important for this study of frontiers, Jerbans used mosques (and shrines) to constitute a defensive line along the first rise up from the coast. About thirty of them are located right on the coast and thirty-­nine are located on a second line of defense inside the island.55 Some of them can be simply attributed the function of a “watch or look-­out.”56 In the Jerban seventeenth-­ century source, al-­Ḥīlātī’s Epistles, we have a chapter titled “The visit to the 62

R EN ATA H O L O D A N D TA R EK K A H L AO U I

Figure 3.10. Rounds of the Shaykhs to check the physical defenses of the island and to mark their territory.

island’s beach mosques [masājid shuṭuṭ al-­jazira] and the inspection of the coastal frontiers fortifications [thughūr al-­sawāhil] so that they would not be surprised by the enemy.” The author describes a visit conducted by the religious leaders and the elders of the community, called the shuyukh to mosques and shrines located along the coasts (figure 3.10). There were, in fact, two rounds of visits, one starting from the north shore, along the western coast to the middle of the southern coast; the other beginning there on the south, proceeding along the eastern coast, and ending with the northern coast. Even though this is a late text, the surviving mosques not only support al-­Ḥīlātī’s claim of vigilance, but they also serve to establish a defensive line that may go back to as early as the eleventh century, when some of the sites where built. While the rounds of the elders of the two communities (or moieties) might be understood as inspection rounds to ascertain the physical preparedness of G uarding a W ell - ­O rdered S pace on a M editerranean I sland

63

individual structures in the “early-­warning” defense line, they can also be seen as a physical inscription and re-­inscription of the territory considered to be Ibāḍī on the island. Of note also are the paths taken, where these look out onto the sea and when they veer inland. In effect, the shaykhs were marking out the frontiers (thughūr) of their communities and their authority in the face of constant outside threats. Three types of mosque sites could be identified as bases for a local defensive system: (1) mosques with inserted towers along the coast or inland in isolated areas, and with clear defensive features such as arrow slits, like the two presented above; (2) underground mosques, especially along the coast; and (3) mosques located away from any habitats along the coast or on the small islands near the main island (masājid al-­shaṭṭ) (appendix 3.1).57 Among the “underground mosques,” that of Sīdī Zikrī was built in 1264, according to al-­Ḥīlātī. An underground structure with 3.5-­m-­high walls on average, it was located on the eastern coast of the island at Mizrāya. It seems to have been named after Abū Zakkāriyyā, an unknown “Jihādī figure,” but who certainly came from the Banū Yahrāsin.58 Likewise, the Zāwīyat Sidi Būsa‘īd functioned as a shrine but also as a mosque. Likely built in the sixteenth century on the northwestern part of the island, it was named after a murābiṭ (religious teacher) of the same century and consists of an underground mosque cut out of living rock.59 Another example may be M. Ghār Mijmāj (also known as Bin Biyān), certainly built before 1040. Located at a countryside mosque in the southern part of Wādī Zabīb, it may be better designated as a retreat for scholars.60 The entire mosque site functioned as a major Wahbī learning center while the underground retreat space was where seven Ibāḍī scholars wrote the Diwan al-­‘Azzaba’.61 While the concept of the thughūr as practiced by the elders of the two Ibāḍī communities on Jerba may not fit into the more normative frontier descriptions as we have come to understand them from the Central Asian, Anatolian, or even Ifrīqiya frontiers, at the microlevel of this island the physical reinscription of their territory by the elders gives us an idea where they thought the lines were drawn. Interestingly, they seem to have skirted the great northern fort while ignoring the southern one completely. FORTS: OUTSIDERS’ INTRUSIONS

In contrast to the network of lookouts, early warning posts and hideaways, full-­fledged fortifications on Jerba seem to have been largely alien to the local Ibāḍī system of defense. Apart from the two Byzantine fort-­towers, discussed 64

R EN ATA H O L O D A N D TA R EK K A H L AO U I

above, which were abandoned by the MM period (the mid-­eleventh century), no new forts can be identified with the EM period. We have no solid evidence to suggest that newly built fortifications were raised immediately after the Islamic conquest.62 The only possible early example is the pre-­Ottoman core of the northern fort, presently named Burj al-Ghāzī Muṣṭafā. Its date has been debated—­whether this part of the fort goes back to the Aghlabids or only to the Sicilians in the thirteenth century—­and we opt for the Sicilian version. Nonetheless, its very history provides the best example of how such fortifications and their control were usually the goal of non-­Ibāḍī forces, outsiders on the island. The early section of the fortress makes the inner quadrilateral, which consists of a square courtyard structure with four corner towers, three rectangular midwall towers, and a rectangular gate tower. While the earlier form of the fortress resembles a ribaṭ prototype known mainly from the Ifrīqiya shores, its very organization and layout within the territory was not part of an Ibāḍī landscape.63 Ribaṭs generally consist of a quadriform structure, often flanked by round towers at the corners and semiround or polygonal interval towers along their sides, one of which often serves as a gate tower.64 Archaeological comparisons aside, this quadrangular section of Burj al-­ Ghāzī Muṣṭafā was certainly extant during al-­Tījānī’s trip to the island in 1306–­1307, when he witnessed the army of Hafsid troops besieging the Sicilian forces that had been established in the fort since 1284. In one of the most important passages describing the island, al-­Tījānī wrote: the fort [ḥiṣn] has a quadrilateral form and there is a tower [burj] in each one of its corners. So two of them are circular and the other two are octagonal, and there is a small quadrilateral tower [burj] between each two of them in the middle of the wall. A small curtain [fasīl] surrounds the fort and a large moat [hafr] surrounds all of that.65

The ḥiṣn (also known as the qashtīl, from the Spanish castillo or Latin castellum) of Jerba has been mentioned in few other medieval sources. The account of al-­Tījānī was reproduced almost verbatim by Ibn Khaldūn, writing in the late fourteenth/early fifteenth century.66 In his history, he mentioned al-­Qashtīl as one of the major coastal defensive fortifications of the Hafsids, and as the main focal point of the political and military conflict over the island from the late thirteenth century to the beginning of the fifteenth century. Ibn Khaldūn’s history is the first source in Arabic to mention that “the Sicilians built al-­Qashtīl after they conquered [ Jerba] in H688 [1284 CE].”67 The dating of the fort’s creation may be corroborated by medieval European G uarding a W ell - ­O rdered S pace on a M editerranean I sland

65

sources. For example, historians generally credit the Sicilians and, more precisely, Roger of Loria, with building a fort on Jerba after capturing the island in 1284. Most notably, this construction is mentioned in Brunschvig’s68 history of the Hafsids, where he cites briefly the different sources that mention this event, although he does not quote any specific text.69 Still, Brunschvig turns to Bartholomaei de Neocastro’s Historia Sicula, written in Latin before 1293.70 Speaking of a devastating Sicilian campaign against the island of Jerba, de Neocastro states:71 The island is captured and it is ruled by the Christians. They have pillaged gold and silver, various possessions and even people; therefore, many people are enriched. And more than six thousand people are bestowed from Jerba to Sicily; any one of them who has been captured is bought by a Sicilian, and more than four thousand are killed. The survivors have escaped, every one of them has hidden and secured himself, instead of serving the king, and they are considering surrender. The major part of the prisoners of war has bought themselves out; afterwards, anyone who is a dignified person and faithful to the kingdom has returned back to his home country [ Jerba]. And there afterwards, his respected Admiral has built [fecit] a castle [castrum] of good endurance, from where he keeps the above-­mentioned island under the Christian rule.72

This is the best textual evidence for dating the building of the fort, which was to be used as a principal residence for the commander during the Sicilian rule of the island after 1284. Another contemporary Christian source, Ramón Muntaner’s Crónica Catalana, corroborates the building of a fortification by the Sicilians and independently notes that this was the very one besieged by the Hafsid troops in 1306–­1307 and described by al-­Tijānī. The account, written by the military commander who witnessed the siege, and who later was the representative of Aragonese rule on the island, also states that after the siege ended: And I had a fosse [ditch] with a wall of stone and earth made at once outside the castle, and within this fosse and wall I had many huts made of planks and matting and branches. And all of Moabia [Wahbī] came to me in the night with their wives and children, and I gave, at once, rations of flour and vegetables and cheeses, which were coming in abundantly. Likewise, I sent to tell the traitor who was chief of Miscona [Mistawa, i.e., Nukkār], namely Alef, to come to me; but he would not. However, two old men of Miscona came to me, but their followers would not leave the others; and of these two one was Amar Ben Buceyt and the other Barquet. What shall I tell you? I had not been in Jerba

66

R EN ATA H O L O D A N D TA R EK K A H L AO U I

a month before I had in my power, altogether, three hundred men of Moabia with their wives and children.73

Although the Catalan chronicle does not clearly mention the building of a second curtain wall, as reported by Ibn Khaldūn, the textual sources indicate that the fort was subject to architectural intervention right after al-­Tijānī’s description had been written. Additional medieval references show that the fort then functioned during the fifteenth century as a political residence for the Hafsid administration.74 After the Spanish defeat in the waters surrounding Jerba in 1560, the Ottomans commissioned Ghazī Muṣṭafā to modify the fort (which now bears his name) for cannons in 1560 to 1567.75 The later history of Jerba and Burj al-­Ghāzī Muṣṭafā involved its eventual attachment to the pashalik of Tunis during the reign of Hamuda Bey (1631–­1659).76 From the seventeenth to early nineteenth centuries, the fort was used to control the island politically and militarily,77 especially with respect to the growing European threat.78 In 1805, the island and fort passed into the hands of the Tunisian authorities.79 Subsequently, it passed into French control in 1880. Finally, Burj al-­Ghāzī Muṣṭafā was returned to the Tunisian government in 1903, becoming a historical monument on March 15, 1904.80 Although it sometimes could serve to defend Jerba from attack, most often this fort, the Burj al-­Ghāzī Muṣṭafā, was the main strongpoint for outsider occupying forces and was used more or less successfully to control the island and its inhabitants. This is most clearly evidenced by the direction of its major cannon embrasures—­in every direction except toward the sea to the north. Such an emplacement would indicate that it was not primarily concerned with defending Ḥumat al-­Sūq and/or the island Jerba from attack from the sea, but rather with defending itself and its occupants from a land-­based attack. Given this, a strong case can be made that this fort, Burj al-­Ghāzī Muṣṭafā in all its phases, was a structure built and used by non-­Jerbans to control and administer the island. The same goes also for the other known fortifications, all of them located at strategic points on the coastline of the island.81 CONCLUDING REMARKS

Close inspection of the territory of the island of Jerba through archaeological survey and sources during the medieval and early modern centuries has shown that it was functioning both as a refuge and as a frontier at the same time. As a refuge, its territory became an important locus of Ibāḍī settlement and a center for intellectual life, with every mosque serving also as a place G uarding a W ell - ­O rdered S pace on a M editerranean I sland

67

of learning. The territory into which both branches of the Ibāḍī, the Nukkār and the Wahbī, expanded also became the place for new settlement patterns: estates clustered around their own mosques (masjid), with no urban centers in the absence of a righteous imam, no city (madina), and, therefore, no right for or necessity of a congregational mosque (masjid al-­jami‘). This rural order created its own boundaries of estate, mosque site, and mosque. In turn, all this was imagined and enacted as a bounded territory with carefully maintained “frontiers” of chains of lookouts, retreats, and signal points. The borders were also enforced because of the sectarian nature of the inhabitants. Their sense of purpose and mission in pursuing the third way of Islam, the Khārijī/Ibāḍī way, was constantly challenged throughout the centuries in question. And the challenges came from two frontiers: the “outside”—­from the thughūr—­and the “inside”—­from the Sunni powerholders in the Maghrib. Opponents in the shape of the Normans, Sicilians, and Spanish could be understood simply as the continuators of the classic frontiers with the Byzantines, and as such they fit into the maritime frontiers’ narrative of Ifrīqiya. The planting of their fort(s) on Jerban territory is indicative of how constant and active this frontier continued to be in the central Mediterranean’s waters and on its shores. Still, at the same time, challenges to the Ibāḍī-­held territory on the island of Jerba came from the inside also, from every Sunni powerholder in the area, from the Aghlabids, Zirids, and Hafsids to the Ottomans. The two most successful of these, first the Hafsids and then especially the Ottomans, secured these forts/strongholds, expanded them, and over time managed to reshape the earlier order of settlement in a sectarian territory to one where that order was much challenged and diminished by the mid-­eighteenth century. APPENDIX 3.1

Mosques as part of the defensive system Mosque

Locationa

Type b Notes

Jāmi’ Al-­Ḥashānī

148001001

2

Jāmi’ Sīdī Sālim

148004001

2

Jāmi’ Sīdī Zāyid

148010001

2

Unknown date; named after a 16th-­century shaykh; destroyed; on the little peninsula of Ḥashānī (Ṣidghiyān) (al-Mrābit 2002, 1).

Datable in the 16th century; along the coast in Shaṭṭ Khanfūs (Tāwrīt); named after a “Jihādī figure” Sālim al-­Baghdādī aka Ādhrūm.

Uncertain; named after a “Jihādī figure,” Sīdī Zāyid; in Fātū (Ṣidghiyān); used as a coastal lookout until the 17th century, according to al-Mrābit 2002, 45. continued on next page

68

R EN ATA H O L O D A N D TA R EK K A H L AO U I

APPENDIX 3.1—continued Mosque

Locationa

Typeb Notes

Jāmi’ Būshaddākh

148010005

1?

Jāmi’ Sīdī Mihriz

148011001

2

Jāmi’ Sīdī Ismā́īl

148012001

2

Jāmi’ Al-­Mistīrī

148013001

2

Jāmi’ ammī Ya ́mir

148013002

2

Jāmi’ Arwāy

148013003

2

Jāmi’ Tlākīn 148013004 (Banī Lākīn)

1

Jāmi’ Būzīd

148016001

1

Jāmi’ Midrājin

148018001

1

Jāmi’ Sīdī Zikrī

148020001

3

Uncertain date; Fātū (Ṣidghiyān); tower similar to Jāmi ‘al-­Māy’s tower (al-Mrābit 2002, 53). Uncertain date; destroyed; named after an unknown murabiṭ aka Sīdī Mihriz (al-Mrābit 2002, 54).

Possibly 15th century; on the coast in Fātū (Ṣidghiyān); attributed to an Ibāḍī Wahbī shaykh (al-Mrābit 2002, 55).

Uncertain; on the coast at Ghīzin (Ṣidghiyān); most likely an outsider murābiṭ; mentioned in Bardu map as “lookout” (al-Mrābit 2002, 56–­57).

Uncertain date but mentioned in the 17th-­century itinerary of the Ibāḍī Shaykhs as its starting point; on the coast at Ghīzin (Ṣidghiyān) (al-Mrābit 2002, 58–­59).

Possibly 17th century; mentioned in Bardu map (al-Mrābit 2002, 59).

Early 13th century (expanded in the middle of the 13th century and a tower was inserted in the 16th century); a countryside mosque with a tower and possibly indications of arrow slits along the roof parapet and “saqqāṭa”; it is believed to be a major center of the Diwan al-­‘Azzaba’ especially beginning from the middle of the 15th century; an Ibāḍī judicial court met there in the 17th and 18th centuries (al-Mrābit 2002, 60–­64). Exact date on inscription: 601 AH/1203–­1204 CE; original tower; a countryside mosque at al-Zāwīya (Ṣidghiyān); named after the family of Abū Zayd (al-Mrābit 2002, 65–­69).

Uncertain but already standing by the 17th century; in the countryside of Mizrāya (Ṣidghiyān) famous for its tall tower from which the coast can be seen; its Berber name, midrājin, suggests “staircases”; its defensive features include a second floor as well (the only Jerban mosque with such a feature); al-Mrābit suggests that this location would relay any news of attacks to inland mosques (al-Mrābit 2002, 71–­75).

Mid-­thirteenth century, according to al-Ḥīlātī 1998 (663 AH/1264 CE); an underground mosque, with 3.5-­m-­high walls on average, on the coast of Mizrāya (Ṣidghiyān); named after an unknown “Jihādī figure” called Abū Zakkāriyyā but certainly from Banū Yahrāsin. continued on next page

G uarding a W ell - ­O rdered S pace on a M editerranean I sland

69

APPENDIX 3.1—continued Mosque

Locationa

Typeb Notes

Jāmi’ Tājdīt

148022001

1

Jāmi’ Bin Ghurbāl

148022002

1

The Great Mosque of Ḥashshān

148026001

1

a

Datable certainly before the end of the 15th century and Abū Rās suggests an even earlier date (10th century) but without referencing any medieval sources; mosque clearly went through different phases of renovation; a countryside mosque in Fātū with a fairly tall tower along with arrow slits and saqqāṭa at the entrance (al-Mrābit 2002, 82–­87). Unknown date, but could be early; Fātū (Ṣidghiyān) Nukkārī in a Wahbī area; a countryside mosque with a tower (al-Mrābit 2002, 87–­89).

Early 10th century; believed to be the oldest Ibāḍī (Wahbī) mosque on the island; a countryside mosque in Ḥashshān, includes a tower but unclear if in the original building; the center of learning of Ibāḍī Wahbīsm (al-Mrābit 2002, 98–­106).

GPS coordinates; see Frachetti 2009, 31–­34.

On the typology of mosques, see Figure 3.7 (p. 60). Type 1 was a hypostyle plan, and type 2 a nine-­ bay, four-­support plan. b

NOTES

1. As described now by Eger 2015. Whether ribāṭs existed as lone structures or attached to other structures actually has not been determined for the Byzantine frontier. In most cases there, frontier fortifications were actually cities or waystations. 2. For studies of the Ifrīqiya frontier, see El-­Bahi 2004 and Bahri 2004; for the term ribāt and its different temporal and regional uses, see Chabbi and Rabbat 2007. 3. On the Aghlabids now, see Anderson et al. 2017. 4. Anderson et al. 2017. For a study of earlier periods in the central Mediterranean, see Horden and Purcelle 2000. 5. General introductions to the Ibāḍī/Khārijī are Lewicki 1959, 2013 and Rebstock 1983, as well as the more recent Prevost 2010. 6. For full results of the survey, see Fentress et al. 2009; vol. 2 is in preparation. For fuller details of the island’s geography and topography, see Frachetti 2009, 55–­70. 7. This term is recorded for eighth-­century H (fourteenth-­century CE) Jerba; see al-­Tijānī 1958. The account is fully analyzed in Holod and Kahlaoui (in preparation). 8. The survey, carried out in five seasons, 1996–­ 2000, was a joint Tunisian-­ American project, supported by the University of Pennsylvania (the University of Pennsylvania Museum of Archaeology and Anthropology, the History of Art Department, and the University Research Foundation), the American Academy in Rome, 70

R EN ATA H O L O D A N D TA R EK K A H L AO U I

and the Institut National du  Patrimoine (INP), Tunis. Additional support was provided by The 1984 Foundation, by Fondation Van Berchem, Geneva, and Mrs. Merle Smith. 9. For a discussion of the transition from late antique to medieval ceramic chronotypologies, see Holod and Cirelli 2011 and Cirelli (in preparation). 10. See Little 2007 for a recent consideration of the impact of this plague cycle on the late antique world, as well as Ellenblum 2012 for a more detailed review of the eastern Mediterranean regions, up to and including Ifrīqiya. See also Thébert and Biget 1990. 11. This impact is discussed nowadays. See Fenwick 2013. 12. Specifically on the decline in population, see Fentress 2009. 13. See our discussion of eighth/ninth-­century CE Jerba in Holod and Kahlauoi 2017. 14. For the most comprehensive discussion of the Jewish community and settlement on Jerba, see Udovich and Valensi 1984. It is not clear whether the two identified modern neighborhoods of settlement, al-­ḥara al-­kabīra and al-­ḥara al-­ṣaghīra, reflect the only Jewish settlements of the late antique or medieval periods on Jerba. Certainly, the latter could be considered the earlier of the two. See also Fentress 2018. 15. On the Tāhart Imamate see Lewicki 1959 and Savage 1997. The leader of the community, the imam, had to be known as a just man (al-­imām al-­ʿādil). 16. Abū Zakariyā 1985. 17. For a full discussion of the two forts, see Fentress 2009, 201–­205, figures 13.1 and 13.2. 18. This is the most probable explanation for the origin of the name Banī Maʿqil. 19. Al-­Shammākhī 1992. 20. As transcribed and edited by al-­Jaʿbīrī 1975, 331–­338. We differ in dating this poem to the mid-­eighteenth century and not the seventeenth; see the full argument in Kahlauoi and Holod, in preparation. 21. Al-­Ḥīlātī 1998, 73–­85. 22. Ibid., 86–­87. 23. Ibid., 5, 9, and 51–­52 24. Al-­Bakrī 1965, 668. 25. Al-­Idrīsī 1990, 1: 317–­318. 26. Al-­Tījānī 1958, 125. On the problem with Jerban piracy, see also Johns 2002, 260. 27. Al-­Shammākhī as cited in Idris 1962, 1: 165–­166. Also see Mathieu and Kahlaoui (in preparation). 28. A geniza is a dead storage for disused documents that could not be destroyed according to Jewish tradition because they bore the name of God written upon them. See Goitein 1967–­1993. G uarding a W ell - ­O rdered S pace on a M editerranean I sland

71

29. Goitein 1967–­1993, 4: 439, where the Norman invasion is mentioned. In Goitein 1967–­1993, 3: 30 and 117, the same ransomed Jerban later marries out of the house of his bride. 30. Goitein 1967–­1993, 4: 283. Because of the famine and debts, in 1136 a man sells his large house on Jerba for 90 dinars. 31. Goitein 1967–­1993, 2: 444, appendix b: documents listing the beneficiaries of the community chest # 25, and two lists of recipients of clothing. Jerba is mentioned among the places from which foreigners received clothes (in Egypt). 4: 400, n.94: as part of a bridal outfit, a blue costume with gold threads and two white wimples ordered from Cairo, and a cloak and shoes. The note is written on vellum, early eleventh century CE. The addressee is Khallūf b. Faraḥ al-­Zarbī. 32. For example, trade with Byzantine territories, as indicated by the jars with signatures from Jerba found on Cyprus and reported by Michailides and Bakirtis 2003; see also Cirelli (in preparation). 33. The register of Giovanni Maiorana (1297–­1300  CE) is preserved in Trapani, edited in 1982, cited by Abulafia 1994, 145. See the statement found in Peri 1990, 71, which also mentions the grain trade cereali da Palermo a Tunisi per 2.15 tari, . . . , a Gerba 3.1 tari; but stating (72) in 1309 that it is forbidden to export grain into the land of the infidels (a la terra de infideli). Ṭarī’ (plural for tareno) are small gold coins (around 1 gram) that were introduced by the Muslim rule in Sicily in the eleventh century CE when they are thought to have been equal to a quarter of a dinar. 34. Preserved in the Archivo del commune di Palermo, unpublished but cited by Peri 1990, 75, 1298–­1299 CE: frumento (wheat) to Tunis 3 tari, . . . , to Gerba 3.1 tari. 1308–­1309 . . . to Gerba 3.1 tari. 35. De  Neocastro 1922, 63, and see below. This text has been rarely consulted directly by contemporary researchers who prefer to cite Brunschvig’s reading of it (e.g., Gharīb 1975, 89–­90). 36. Muntaner 1860. 37. Despite the fact that the original plan of the fort was ribāt-­like in its general features, it was not raised by the Aghlabids but by these later intruders into Jerban space; see below, and also Holod and Kahlauoi 2017. 38. Al-­Tījānī’s Risālah, dated 706–­708 AH, was published in 1981, edited by Hasan Hosni Abdulwahab, and based on manuscripts in Tunis, Algiers, Tlemcen, and private libraries. In his introduction, the editor states that no full fair copy exists. He has had to base himself upon the work of al-­Wazīr al-­Sarrāj (d. 1149 AH/1736–­1737 CE), al-­Ḥulal al-­Sundusiyya, as the latter had copied many paragraphs from existing manuscripts of al-­Tījānī into his own work. The project of reconstruction and edition was begun in 1923 with W. Marçais, which Abdulwahab was able to complete many years later. One should be wary, therefore, of misreadings, omissions, etc. H. R. Idris, whose 1962 monograph, 72

R EN ATA H O L O D A N D TA R EK K A H L AO U I

Berbérie Orientale has been used most by subsequent scholars in citing al-­Tījānī, also based his book on work of the same al-­Sarrāj. There are important differences between these two versions of the al-­Tījānī narrative. 39. Facio 2004; Brunschvig 1940, 1: 230–­232: “Gerba, isola abitata da molti contadini, distava della terraferma circa quattromila passi e, nella parte occidentale, ricca de insenature [inlets?], era ancora piu vicina (la distanza in quel punto era di mille passi): un ponte roccioso congiungeva l’isola al continente. Per fronteggiare l’assalto dei barbari, occorreva distruggere quel ponte: in questo modo gli abitanti di Gerba, privati della speranza di un aiuto, si sarebbero arresi più facilmente. Avvicinandosi lungo la rotta all’isola, Alfonso ordino prima alle navi pesanti di dirigersi verso il porto dove si trovava la torre di  Valguarnera, cosi chiamata dal nome del fondatore, poiché, a causa delle Sirti, non potevano avvicinarsi al ponte. Poi, divise in due gruppi le galee, assegnandone le meta a Gutierre e poiche aveva deciso di attaccare il ponte da due parti, questi gli chiesi il permesso di accostare al ponte da occidente: era convinto di riuscire a raggiungere per primo il ponte e ad occuparlo. Benche ad Alfonso non sfuggisse il motivo di quella richiesta, lo accontento. Gutierre infatti era assai esperto di mare. Cosi Alfonso da oriente e Gutierre da occidente si diressero verso il ponte, ma u Alfonso ad arrivare per primo: Gutierre non molto pratico di secche, perse tempo in giri e regiri e alcune nave che lo seguivano, impigliate nelle secche, si fermarono.” We thank Enrico Cirelli for his translation of this passage. 40. Soucek 1973, 251–­255. 41. See Kahlaoui 2009 for a discussion for the new urban centers. 42. Ibn Maqdīsh 1988, II: 28. 43. Ibn Maqdīsh 1988, I: 140–­143. 44. Ibn Maqdīsh 1988, II: 436–­448, 462–­467. 45. Frachetti 2009, 66–­68. 46. Al-­Tījānī 1981, 128. 47. For a mapping of the area of Mahbubin, to which our project did not have access, see Wilder 1980. 48. Al-­Shammākhī 1992. There was some disagreement among scholars whether any space could be a mosque, including even a cave or reed enclosure. The first two could be considered a masjid but al-­Shammākhī considers that only a structure of permanent materials such as bricks (tubah), or cut stone, or a mixture of clay and water (e.g. rammed earth) could suffice. 49. His quintrain (mukhammas) was transcribed and edited by al-­Jaʿbīrī 1975, 331–­338. Our translation and discussion are found in Kahlaoui and Holod (in preparation). 50. On staircase minarets as an earlier alternative type for a call-­to-­prayer platform, see Schacht 1961. G uarding a W ell - ­O rdered S pace on a M editerranean I sland

73

51. Holod and Cirelli 2011; and Cirelli (in preparation). On the rounds of the Wahbī shaykhs, see al-­Ḥīlātī 1998, 86, and below. el-Mrabet dates the site in the eighth AH/ fourteenth CE century (al-Mrābit 2002: 583–­586). el-Mrabet's date follows al-Ja‘bīrī 1975. 52. See for example Jonathan Bloom’s (1983) discussion of Maghribī and Faṭimid mosques, and notably al-­Ḥakim Mosque, where he emphasizes symbolic and ceremonial explanations for the multiple minarets and the projected portal structures. 53. Lézine 1971. 54. After the fall of the Tāhart imamate in 296 AH/909 CE, neither of the Ibāḍī factions practiced the Friday congregational prayer in the absence of one who was considered a righteous imam (al-­‘imām al-­ʿādil) to lead it. 55. Hbaieb 2013, 307. 56. Prevost 2006, 11. 57. The use of the terms masjid and jami’ in a Jerban context is a complex question. Clearly, in the periods under review here, there was an avoidance of the term jami‘ with its connotation of a congregational mosque. Later, however, and perhaps under pressure from the Ottoman administration, many mosque sites changed their names from masjid to jami‘, with the result that, at the time of our survey, all were named by the latter term. See Kahlaoui and Holod (in preparation) for an extended discussion. 58. al-­Ḥīlātī 1998, 74 59. al-Mrābit 2002, 205–­207. 60. Ibid., 236–­241. 61. On the Diwan al-­‘Azzaba’, see al-Ja‘bīrī 1975; al-Mrābit 2002, 60–­64; and Holod and Kahlaoui (in preparation). 62. For a full description of the forts and other redoubts on the island, see Mathieu and Kahlaoui (in preparation). 63. On Jerba and the Aghlabids, now see Holod and Kahlauoi 2017. 64. Al-­Ḥīlātī 1998, 44n17; Djelloul 1999, 39, 51; Pringle 1981, 168; cf. Fentress 1979; Mattingly 1994. These have been argued to be Islamic derivations of earlier Romano-­ Byzantine forts, whether in the Near East (Creswell 1952) or locally in North Africa (Pringle 1981, 168). 65. Al-­Tījānī 1981, 128. 66. Ibn Khaldūn 1959, 6: 645–­650, 697–­698, 730, 820, 874, 908–­909. 67. Ibn Khaldūn, 1959, 5: 429 and 6: 250. 68. Brunschvig 1940, 1: 93–­94, 319. 69. Brunschvig 1940 erroneously suggests that Roger of Loria built the “Qashtīl” of the south rather than the northern one. 70. Interestingly, this text is rarely consulted directly by medieval historians, who instead prefer to cite Brunschvig’s reading of it; see for example, Gharīb 1975: 89–­90. 71. Brunschvig 1940, 1: 93–­94, 319; De Neocastro 1922. 74

R EN ATA H O L O D A N D TA R EK K A H L AO U I

72. De Neocastro 1922, 63 (translation: Tarek Kahlaoui). 73. Muntaner 1967, 604–­605. 74. Brunschvig 1936, 36, 210. 75. Gharīb 1975, 62. 76. Julien and Le Tourneau 1970, 304. 77. Al-­Hīlātī 1998, 9–­51. 78. Ibid., 1998, 103–­104. 79. Gharīb 1975, 62, who mistakenly says 1605. 80. Ibid. 81. For a full description of the fortifications on the island, see Mathieu and Kahlaoui (in preparation). REFERENCES

Abulafia, David. 1994. A Mediterranean Emporium: The Catalan Kingdom of Majorca. Cambridge, UK: Cambridge University Press. Abun-­Nasr, Jamil M. 1987. A History of the Maghrib in the Islamic Period. New York: Cambridge University Press. Abū Zakariyā, Ibn Abi Bakr. 1985. Kitāb al-­Sīrah wa-­akhbār al-­a’immah. Tunis: al-­ Dār al-­Tūnisīyah lil-­Nashr. Al-­Tījānī 1981. Anderson, Glaire D., Corisande Fenwick, and Mariam Rosser-­O wen, with Sihem Lamine, eds. 2017. The Aghlabids and Their Neighbors: Art and Material Culture in Ninth-­Century North Africa. Leiden: Brill, HdO series. Bahri, F. 2004. “De certaines techniques de construction d’époque fatimide à Qasr al-­‘Aliya.” Africa 20: 37–­60. al-­Bakrī, Abū ʾUbayd. 1965. Description de l’Afrique septentrionale [Al-­Masalik wa’l Mamalik], edited and translated by M. De Slane. 2nd ed. Paris: Adrien Maisonneuve. Bloom, Jonathan. 1983. “The Mosque of Al-­Hakim in Cairo” Muqarnas 1: 15–­36. Brunschvig, R. 1936. Deux récits de voyage inédits en Afrique du Nord au XVe siècle. Paris: Larose. Brunschvig, R. 1940. La Berbérie orientale sous les Hafsides: Des origines a la fin du XVe siècle. 2 vols. Paris: Adrien Maisonneuve. Chabbi, J., and N. Rabbat. 2007. “Ribāṭ.” In Encyclopaedia of Islam, edited by P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs. Brill: Brill Online. Cirelli, E. In preparation. “The Ceramic Production: 700–­1800.” In An Island through Time: Jerba Studies, volume 2, edited by R. Holod, A. Drine, and E. Fentress. JRAS Supplement. Portsmouth, RI: Cambridge University Press. G uarding a W ell - ­O rdered S pace on a M editerranean I sland

75

Creswell, K. A. C. 1952. “Fortification in Islam before AD 1250.” Proceedings of the British Academy 38: 89–­125. de Neocastro, B. 1922. “Historia Sicula.” In Raccolta Degli Storici Italiani dal cinquecento al millecinquecento, edited by L. A. Muratori, 13. Bologna: N. Zanichelli. Djelloul, Néji. 1999. Al-­Ribātāt al-­bahrīyah bi-­Ifrīqīyah f ī al-­‘asr al-­wasīt. Tunis: Markiz al-­Dirāsāt wa al-­Buh ūth al-­Iqtis ādiyya wa al-­Ijtimā iyya. Eger, A. A. 2015. The Islamic-­Byzantine Frontier. London-­New York: I.B.Tauris. El-­Bahi, Ahmed. 2004. Sūsah wa-­al-­Sāḥil f ī al-­‘Ahd al-­Wasīṭ, Q [i.e. Qarn] I-V H/VII­XI M: muḥāwalah f ī al-­jughrāfiyā al-­tārīkhīyah. Tūnis: Markaz al-­Nashr al-­Jāmi‘ī. Ellenblum, Roonie. 2012. The Collapse of the Eastern Mediterranean, 950–­1072. Cambridge, UK: Cambridge University Press. Facio, B. 2004. Rerum gestarum Alfonsi regis libri, translated from Latin to Italian by Daniela Pietragalla. Alessandria: Edizioni dell’Orso. Fentress, Elizabeth B. 1979. Numidia and the Roman Army: Social, Military, and Economic Aspects of the Frontier Zone. BAR International Series 53. Oxford: BAR. Fentress, Elizabeth B. 2009. “The Towns and Ports.” In An Island through Time: Jerba Studies, volume 1: The Punic and Roman Periods, edited by E. Fentress, A. Drine, and R. Holod, 75–­85. JRAS Supplement. Portsmouth, RI: Cambridge University Press. Fentress, Elizabeth. 2018. “An Island in Transition: Jerba between the Fifth and the Ninth Centuries.” In Entre civitas y madīna: El mundo de las ciudades en la Península Ibérica y en el norte de África (siglos iv–­ix), edited by Sabine Panzram and Laurent Callegarin, 241–­252. Collection de la Casa de Velázquez 167. Madrid. Fenwick, C. 2013. “From Africa to Ifriqiya: Settlement and Society in Early Medieval North Africa (650–­800).” Al-­Masaq 25.1: 9–­33. Frachetti, Michael. 2009. “Geography.” In An Island through Time: Jerba Studies, volume 1: The Punic and Roman Periods, edited by E. Fentress, A. Drine, and R. Holod, 55–­71. JRAS Supplement. Portsmouth, RI: Cambridge University Press. Gaiser, Adam. 2010. Muslims, Scholars, Soldiers: The Origin and Elaboration of the Ibadi Imamate. New York: Oxford University Press. Gharīb, R. 1975. “Le Borj Ghazi Mustapha a Houmt-­Souk.” In Séminaire pour la Sauvegarde de l’Architecture et de l’Environment de Djerba (23–­26 Janvier 1975): Communications et Recommendations, 61–­67. Tunis: Association de sauvegarde de Djerba. Goitein, Shlomo. 1967–­1993. A Mediterranean Society: The Jewish Communities of the Arab World as Portrayed in the Documents of the Cairo Geniza. 6 vols. Oakland: University of California Press. Hbaieb 2013. al-­Hīlātī, S. 1998. Rasa’il, edited by Muhammmad Gouja. Beirut: Dar al-­Gharb al-­Islami. 76

R EN ATA H O L O D A N D TA R EK K A H L AO U I

Holod, R., and E. Cirelli. 2011. “Islamic Pottery from Jerba (7th–­10th c.): Aspects of Continuity?” In La Céramique Maghrébine du Moyen Age (VIIIe–­Xe Siècles), edited by P. Cressier and E. B. Fentress, 165–­186. Rome: École Française de Rome. Holod, Renata, and Tarek Kahlaoui. 2017. “Jerba in the Ninth Century: Under Aghlabi Control?” In The Aghlabids and Their Neighbors: Art and Material Culture in Ninth-­Century North Africa, edited by Glaire D. Anderson, Corisande Fenwick, and Mariam Rosser-­O wen, 451–­469. Leiden: Brill, HdO series. Holod, Renata, and Tarek Kahlaoui. In preparation. “Medieval Sources.” In An Island through Time: Jerba Studies, volume 2, edited by R. Holod, A. Drine, and E. Fentress. JRAS Supplement. Portsmouth, RI: Cambridge University Press. Horden, Peregrine, and Nicholas Purcell. 2000. The Corrupting Sea: A Study of Mediterranean History. Oxford; Malden, MA: Blackwell Publishers. Ibn Khaldūn. 1959 [1852]. Histoire des Berberes. 4 vols. Translated by M. DeSlane. Algiers: Institut d’Etudes orientales de la Faculté des lettres d’Alger. Ibn Maqdīsh, Mahmūd. 1988. Nuzhat al-­anẓār f ī ‘ajā’ib al-­tawārīkh wa-­al-­akhbār. 2 vols. Beirut: Dār al-­Gharb al-­Islāmī. Idris, H. R. 1962. La Berbérie orientale sous les Zīrd̄es, Xe–­XIIe siècles. 2 vols. Paris: Adrien-­Maisonneuve. al-­Idrisi, al-­Sharif Muhammad Ibn Muhammad. 1990. Kitab Nuzhat al-­Mushtaq fi Ikhtiraq al-­Afaq. 2 vols. Cairo: Maktabat al-­Thaqāfah al-­Dīnīyah. Insoll, Timothy. 1999. The Archaeology of Islam. Malden, MA: Blackwell. al-Ja‘bīrī, F. 1975. Niẓām al-­ʾAzzāba ʾind al-­Ibāḍiyya al-­Wahbiyya. Tunis: Wizarat al-­ Shuun al-­Thaqafiyah, al-­Ma‘had al-­Qawmi lil-­Athar wa-­al-­Funun. Johns, J. 2002. Arabic Administration in Norman Sicily: The Royal Diwan. Cambridge, UK: Cambridge University Press. Julien, C.-­A., and R. Le Tourneau, R. 1970 [1952]. History of North Africa, Tunisia, Algeria, and Morocco from the Arab Conquest to 1830, 2nd ed., translated by J. Petrie. New York: Praeger. Kaegi, Walter. 2010. Muslim Expansion and Byzantine Collapse in North Africa. New York: Cambridge University Press. Kahlaoui, Tarek. 2009. “Tracing Urbanization in Early Modern Jerba.” In The Mediterranean Medina, edited by Ludovico Micara, Attilio Petruciolli, and E. Vadini, 406–­414. Rome: Laterza. Kahlaoui, Tarek, and Renata Holod. In preparation. “Chapter 2: Early Modern Sources.” In An Island through Time: Jerba Studies, volume 2, edited by R. Holod, A. Drine, and E. Fentress. JRAS Supplement. Portsmouth, RI: Cambridge University Press. Lewicki, Tadeusz. 1959. Les Ibadites en Tunisie au Moyen âge. Rome: A. Signorelli.

G uarding a W ell - ­O rdered S pace on a M editerranean I sland

77

Lewicki, T. 2013. “al-­Ibāḍiyya.” In Encyclopaedia of Islam, 2nd ed. Brill Online, University of Pennsylvania. Accessed March 1, 2013. Lézine, Alexandre. 1971. Deux villes d’Ifriqiya: Sousse, Tunis, études d’archéologie, d’urbanisme, de démographie. Paris: P. Geuthner. Little, Lester K., ed. 2007. Plague and the End of Antiquity: The Pandemic of 541–­750. Cambridge, UK: Cambridge University Press. Mathieu, James, and Tarek Kahlaoui. In preparation. “Fortifications.” In An Island through Time: Jerba Studies, volume 2, edited by R. Holod, A. Drine, and E. Fentress. JRAS Supplement. Portsmouth, RI: Cambridge University Press. Mattingly 1994. Michailides, D., and C. Bakirtsis. 2003. “Arabiko’ Amphorei Kypro.” VII Congrès International sur la Céramique Médiévale en Méditerranée, Thessaloniki, 11–­16 Octobre 1999, 125–­136. Athens: Caisse des Recettes Archéologiques. al-­Mrābit (el-­Mrabet), Riadh. 2002. Mudawwanat Masājid Jirba. Tunis: al-­ Jumhuriyah al-­Tunisiyah, Wizarat al-­Thaqafah, al-­Ma‘had al-­Watani lil-­Turath. Muntaner, Ramón. 1860. Crónica Catalana. Antonio de Bofarull, Trad. in Les Quatre Grans Cròniques. Revisió del text, Pròlegs i notes, Ferran Soldevila. Barcelona: Editorial Selecta. Muntaner, Ramón. 1860. Crónica Catalana. Antonio de Bofarull, Trad. in Les Quatre Grans Cròniques. Revisió del text, Pròlegs i notes, Ferran Soldevila. Barcelona: Editorial Selecta. Peri, Illuminato. 1990. La Sicilia dopo il Vespro: Uomini, città e campagne, 1282–­1376. Rome: Laterza. Prevost, Virginie. 2006. “L’urbanisme des Ibadites Maghrebins Medievaux et le Cas Particulier de Djerba.” Rocznik Orientalistyczny LIX: 9–­10. Prevost, Virginie. 2009. “Les Mosquées Ibadites du Maghreb.” Revue des Mondes Musulmans et de la Méditeranée 125 ( July): 220. Prevost, Virginie. 2010. Les Ibadites de Djerba à Oman, La Troisième vie de l’Islam. Turnhout, Belgium: Brepols. Pringle, Denys. 1981. The Defense of Byzantine Africa from Justinian to the Arab Conquest: An Account of the Military History and Archaeology of the African Provinces in the Sixth and Seventh Centuries. BAR International Series 99i. Oxford: BAR. Rebstock, U. 1983. Die Ibaditen im Maghrib. Berlin: K. Schwarz. Savage, Elizabeth. 1997. A Gateway to Hell, A Gateway to Paradise: The North African Response to the Arab Conquest. Princeton, NJ: The Darwin Press. Schacht, J. 1961. “Further Notes on the Staircase Minaret.” Ars Orientalis 4: 137–­141. al-­Shammākhī, Abu’l ‘Abbas. 1992. Kitab al-­Siyar, Sultanate of Oman. Soucek, S. 1973. “Tunisia in the Kitāb-­i Bahriye by Piri Reis.” Archivum Ottomanicum 5: 129–­296. 78

R EN ATA H O L O D A N D TA R EK K A H L AO U I

al-­Tijānī, A. 1958. Rihla de Abdallah Tidjani: Relation de voyage en Tunise et en Tripolitaine (de 1306 à 1308 J. C.). Tunis: Kitabat al-­Dawlah lil-­Ma‘arif. al-Tijānī, ‘Abdullah. 1981. Rihla al-Tidjani. Tunis: Jami’al-Huquq Mahfuzah. Thébert, Yvon, and Jean-­Louis Biget. 1990. “L’Afrique après la disparition de la cité classique: Cohérence et ruptures dans l’histoire maghrébine.” In L’Afrique dans l’Occident romain, 575–­602. Actes du colloque de Rome, 3–5 decembre, 1987 [CEFR CXXXIV]. Rome: École Française de Rome. Udovitch, A. L., and L. Valensi. 1984. The Last Arab Jews: The Communities of Jerba, Tunisia. New York: Harwood Academic. Warfalli, Muhammad Salim Muqayyid. 1981. “Some Islamic Monuments in Jabal Nafusa.” Unpublished PhD diss., University of London. Wilder, M. P. 1980. “The Djerban Diaspora: A Tunisian Study of Migration and Ethnicity.” PhD diss., University of Pennsylvania.

G uarding a W ell - ­O rdered S pace on a M editerranean I sland

79

4 Conceptualizing the Islamic-­Byzantine Maritime Frontier Ian Randall

DOI: 10.5876/9781607328773.c004

80

As noted in the introduction to this volume, it is safe to say that in the study of frontier dynamics the utility of core-­periphery models has by now been called seriously into question.1 The notion of passive peripheries absorbing cultural, political, and economic directives from their distant metropoles has been seen to be largely incorrect, with an understanding of more bivalent relationships now à la mode.2 Additionally, the frontiers themselves are often seen as what Richard White3 has called the “Middle Ground,” zones rather than strict demarcations in which a dialectic relationship emerges between differing sets of conceptual orders, habitus, and material culture to produce something altogether new. Archaeology has proven particularly adept at outlining the local manifestations of these relationships, as the case studies in this volume illustrate for the other frontiers of Islam. They also continually show how borders are conceptual constructs with political and ideological baggage; nowhere is this truer than in the perceived division between the dār al-­ḥarb, the house of war, and the dār al-­Islām, the house of Islam.4 A great deal of work has been done in recent years on the thughūr, the major frontier or border towns of Syro-­Anatolia, problematizing the textual accounts of wasteland, raid, and counterraid between the caliphate and Byzantium, and drawing a far more complex picture with the help of archaeology.5 There has been considerably less attention devoted to the islands of

the eastern Mediterranean and the early maritime frontier of Islam, however, and there the sea offers an interesting case in which to consider the role of this natural “boundary” in the conceptualization of frontier spaces. Much like the Jabal Nafūsa, the Sudanese desert, and the Caucasus, this environment acts as a major structuring element in how interaction and disconnection articulate to create the local manifestations of frontier life.6 Getting at how people in the late antique and early medieval Mediterranean conceived of the spaces in which they lived and traveled is no easy task. Texts help us somewhat, but for this period in particular they are few and far between, and often do not speak explicitly about seafaring, travel, or the concept of space for the lay individual. The Sea is also a complex medium to discuss, with any purported all-­encompassing characterization but a fool’s errand for the historian and archaeologist. In Horden and Purcell’s The Corrupting Sea,7 for example, microregionalism is taken as the guiding principle even as various uniting themes are discussed in Mediterranean history. Here, however, I attempt to outline some of the overall factors that likely structured the conceptualization of this maritime frontier for the people who lived there, the islanders of the Eastern Mediterranean. First I outline a general framework for discussing insularity, isolation, and connectivity, before delving into the material and textual evidence for shifts in these issues taking place at the time of the emergence of Islam. I focus primarily on the inhabitants of Crete, Cyprus, and the islands of the Aegean from the seventh to the tenth centuries, but will also stray somewhat before and after this period, particularly as the textual evidence is so scarce (figure 4.1). Without any texts detailing the specific self-­identification or even external identification of the islanders beyond Orthodox Christians and sometime Byzantine citizens, this study differs somewhat from the emphasis on various known ethnic or religious groups in other chapters, although a focus on frontier lifeways, perception, and identity remains. Material evidence forms the bulk of my inquiry, although textual evidence is incorporated where available. Shifts in connectivity are addressed first, looking at the historical context and ceramic data for continued mercantile activity and travel. Next some of the textual evidence for said connectivity, which is largely hagiographic, is also examined. This is followed by an in-­depth discussion of the technological limitations on sea travel and sea power at the time, and the architectural evidence for changes in piracy, raiding, and the dangers of coastal habitation and sea passage in general. I close by looking at traveling specialists themselves and then outlining a framework for considering an emic conceptualization of the maritime frontier for the inhabitants of the Conceptualizing the I slamic - ­B yzantine M aritime F rontier

81

Figure 4.1. The eastern Mediterranean Sea.

islands, and its significance for their material entanglements with the lands of Islam. INSULARIT Y AND CONNECTIVIT Y: A FRAMEWORK

Islands often have an ambiguous place in the scholarship on frontiers and borderlands.8 Their insularity is often taken for relative isolation, and in treating the emergence of Islam they have generally been discussed as outposts of Byzantium, subject to the raids of the caliphate and retreating into the same sort of abject devastation and relative emptiness ascribed in textual accounts to the mainland thughūr.9 Indeed in many of the major works on the archaeology and history of early Islam, relying as they do on the works of early Muslim historians, islands, such as Crete, Rhodes, and even nearby Cyprus, are mentioned only in the context of early raids and then never mentioned again, with their potential connections to or contacts with the caliphate being completely omitted.10 Even the conquest of Crete by Muslims from al-­Andalūs is treated largely as a footnote. Below I show that the material evidence speaks against this implicit characterization of total isolation. 82

I A N R A N DA LL

Insularity can, and often does, however, mean many different things at different times. Christy Constantakopoulou in The Dance of the Islands,11 discussing the Athenian thalassocracy, has pointed to this quality as having the possibility of connoting notions of backwardness, connectivity, safety, danger, disconnectedness, and of naval power, to name just a few, at the same and various moments, depending on how the islands are discussed. Cyprian Broodbank,12 in considering islands, has described them as existing on a continuum between completely connected or totally isolated at any given time, with the possibility to move in either direction, depending on context. It cannot be denied that islands have the singular characteristic of being completely surrounded by water. This qualitative difference in environmental situation—­of being both separated by, and bordering, the sea—­gives islands a circumscribed area that can be easily grasped cognitively.13 Moreover, travel across the medium of water is fraught with its own unique dangers and difficulties, and requires a set of specialized skills and knowledges to undertake successfully.14 This set of constraints changes the character of travel and habitation for islanders in a way that mountains or desert environments, which can at times be even more isolating, do not, and that condition is worthy of special consideration. In examining the Byzantine-­Islamic maritime frontier, two sets of poles should be taken into account when discussing how this singular environmental situation may have affected the perception of space and meaning for late antique and early medieval islanders. Isolation and connectivity have already been mentioned, to which the continuum of safety and vulnerability can be added (figure 4.2). The factors that influence movement along these poles, which the material and textual evidence address below, provide the framing conditions for perceptions of both distance and travel, and can be seen to have both epistemological and material implications for island inhabitants. The changes that occurred with the emergence of Islam onto the Medi­ter­ ranean stage in terms of trade and the movement of goods and people to and from islands allow for the characterization of relative isolation and connectivity for the seventh through tenth centuries, particularly when considered next to the situation in the hundred years prior. The ceramic information is by and large better for the sixth and seventh centuries, due to the still relatively nascent study of early medieval pottery, but the character of these changes can nonetheless be assessed, particularly when the few textual sources are also taken into account. CONNECTIVIT Y: THE MATERIAL EVIDENCE

The islands of the eastern Mediterranean are confidently considered to have been highly connected by trade to their surrounding landmasses, and each Conceptualizing the I slamic - ­B yzantine M aritime F rontier

83

Figure 4.2. The poles of interpretation for conceptualizing the Byzantine-­Islamic frontier.

other, during the sixth and early seventh centuries.15 Large quantities of late Roman Amphorae 1 of diverse local and imported provenience found off the coasts of Cyprus and Crete, as well as their appearance far inland alongside North African, Cypriot, and Phocaean Red Slip wares, point to the situation of these islands within a complex web of interaction and exchange that stretched from Egypt and Constantinople to Syria, the Levant, Sicily, and North Africa.16 The annona system of grain shipments from Egypt that fed the capital sailed past Agios Georgios on the southwestern coast of Cyprus and then passed on through the islands of the Aegean until finally heading through the Marmara to Constantinople.17 Initially state controlled, but largely in the hands of private merchants by the seventh century, these grain ships and their owners were given significant tax breaks and insurance protection by Byzantium and carried with them ceramic goods and agricultural produce for private sale on the side.18 These passed through the large emporia visited on the journey to and from the capital, and were then distributed around the islands and coasts by a network of local merchants operating out of small, independent anchorages that were often less than a day’s sail apart.19 Much of the commerce conducted would have consisted of this small-­scale cabotage, and indeed after the beginning of Arab raids in the seventh century there is a notable shift in the ceramics of the eastern Mediterranean toward a larger percentage of more local products, indicating a greater emphasis on this level of activity.20 These small ships and crews would also have been more likely to pivot around environmental conditions that would otherwise have made travel difficult or impossible.21 84

I A N R A N DA LL

The first Arab raids by sea, which began on Cyprus in 649 and then spread to Crete and the Aegean, were previously considered to have largely ended this period of high interconnectivity and commerce. These raids were thought to have been so pervasive and devastating that maritime life was brought to a practical standstill, followed by swift declines in settlement, agriculture, and even ceramic production around the northern half of the eastern Mediterranean.22 This view is now considered to be a gross exaggeration, having been based on a combination of a poor understanding of the shifts in ceramic forms and an overreliance on textual sources that often had an interest in portraying the conditions following the emergence of Islam as apocalyptic.23 Recent reexaminations of the ceramic material have, incrementally but also continually, pushed the dates of fine ware and amphorae forms formerly thought to have ended around 700 well into the eighth century, and in some cases beyond, and trade can be seen to continue.24 The accumulated evidence for this is too great to detail here, but suffice it to say there has been an overreliance on Hayes’s 1972 work25 on late Roman pottery, which stated in particular a circa 700 cutoff for Cypriot Red Slip Form 9, an overall lack understanding of local and handmade forms, which largely continues, and a general belief in the Pirenne Thesis and the complete collapse of trade at the end of the seventh century due to Arab raiding.26 Raids certainly did occur, as can be seen from the following, fragmentary, mid-­seventh-­century inscription from the church at Soloi, Cyprus: At that time, in Year 7 of the indiction, 365 of the era of Diocletian [August 648–­August 649], because of our sins, there took place the attack on the island. Many were killed, and about 120,000 were led away as prisoners. Again, subsequently, the island suffered a new invasion, more lamentable than the preceding one, in the course of which a great number of people fell under the dagger and were led away prisoner . . . except 50,000 . . . destroyed by fire . . . the whole of the episcopal [palace] . . . of the saints . . . of good appearance . . . in great haste . . . reconstructed the buildings which had collapsed, had made a roof, had raised them again on their foundations, and finished the work in Year 13 of the indiction, 371 of the Diocletian era.27

It is notable though that the plaque from which this inscription comes was erected as part of the rebuilding of the church, and many other churches that show destruction layers dating to the same period were also reconstructed with relative rapidity.28 By the mid-­eighth century the situation in the eastern Mediterranean appears to have stabilized somewhat, likely correlating, at least Conceptualizing the I slamic - ­B yzantine M aritime F rontier

85

in part, with the ‘Abbāsid revolt of 750 that toppled the Umayyads and moved the capital of the caliphate to Baghdad.29 There was some urban contraction and abandonment, with sites such as Constantia on Cyprus significantly reduced, and others such as Kourion and Gortyn on Crete abandoned at some point during this period; by and large, however, cities and settlements endured on the islands.30 Trade also continued with areas that were now under Muslim domination. Palestinian lamps and amphorae, as well as Egyptian Red Slip ware appear on sites, if in small numbers, on both Crete and Cyprus as well as on islands that featured large emporia, such as Chios.31 CONNECTIVIT Y: TEXTUAL EVIDENCE

The Greek textual sources from this period, primarily hagiographies and pilgrims’ accounts, also add to a picture of relatively high connectivity. Saints and holy men are described traveling considerable distances and with some frequency. Pirates, bad weather, and other hazards are mentioned in the texts, but they are often depicted as being deftly managed by the protagonists, and there is a sense of confidence about travel that comes through strongly.32 The Life and Conduct of Our Blessed Mother Theodora of Thessalonike, for example, mentions the depredations of Arab pirates on the island of Aegina, forcing the saint to travel, after some deliberation, to Thessaloniki, a not insubstantial distance away.33 By the eleventh and twelfth centuries, letters describe monks traveling between monasteries in the Aegean simply to borrow books, and trips to Constantinople are discussed as normal embassies of no great discomfiture.34 A ninth-­century text, the Life of St. Demetrianos of Chyrtoi, describes the passage of the ecclesiastic from Cyprus to the Levant, and from there on to Baghdad to meet with the caliph to discuss a treaty abrogation, but the majority of the text concerns itself with matters other than the actual journey, which receives only passing reference.35 Indeed, the picture painted by these sources is one of networks and well-­trodden routes that would not have been unfamiliar to a Roman of the first or second century. The aforementioned evidence for connectivity, however, must be balanced against our sources of information for the relative safety and vulnerability of passage over the sea, and other considerations. For example, the movement described in the texts of ecclesiastics and holy men across vast distances and with relative ease, can be ascribed at least in part to literary trope.36 The dangers of travel discussed earlier—­pirates, weather, and so on—­are easily overcome within the texts by virtue of the status of the travelers as holy men. 86

I A N R A N DA LL

Just as Jesus calmed the Sea of Galilee, saints and monks dispel demons, calm frightened sailors, and display their prowess in a way that makes the destination and purpose of the voyage often secondary.37 Early Muslim chroniclers are far less sanguine. The historian al-­Ṭabarī, writing in the ninth century CE, compares sailing to “that of a small creature riding on the back of a great one, as a worm on a twig. If it bends, he is drowned, and if spared he is astounded.”38 Even late writers such as Ibn Jubayr (1145–­1217) speak tremulously about sea passage, and the ḥadīth of al-­Muttaqī (1472–­1567) states that “a campaign by sea is like ten campaigns by land, and he who loses his bearings at sea is like one who sheds his blood in the path of God,” and “if a campaigner by sea is seasick, he has the reward of a martyr; if drowned, of two martyrs.”39 TRAVEL BY SEA: DANGERS AND DIFFICULTIES

Travel by sea was fraught with real difficulties and dangers, and specialized knowledge was required to achieve it. The proximity of islands in the Aegean mitigated this somewhat, with ships able to move quickly from landfall to landfall, but along the trunk routes from Egypt to the Levant and from thence to Cyprus, Crete, and further west was a different matter.40 The winds in the eastern Mediterranean tend overwhelmingly to come from somewhere between the northwest and northeast, while currents move counterclockwise around the basin.41 Warships of the period were unable to sail effectively into the wind, due to their shallow draft and penchant for heeling dangerously, and would have been forced to travel under oars in adverse conditions, severely limiting their speed and range.42 Merchant vessels, while slightly better off in this regard, would still have been unable to make considerable headway, and while passage north and west around the basin, hugging the shore for safety, was generally easier during the spring and summer of the year, traveling back could be more difficult, and significant portions of the fall and winter were considered unsafe for travel.43 Moreover, storms could alter this general pattern at any time, and it is not unusual for strong northern winds to isolate communities from the sea for significant periods of time even now.44 Dangerous shoals and offshore rocks could also easily sink the vessel of an incautious captain. In a letter to his brother Euoptios in the early fifth century, Synesios describes being becalmed, and then shipwrecked, while other contemporaneous correspondence speaks of running out of water, piracy, and other dangers.45 Numerous shipwrecks dating to late antiquity and the early Middle Ages have Conceptualizing the I slamic - ­B yzantine M aritime F rontier

87

been found in the eastern Mediterranean, including those at Yassi Ada, Tantura, and the thirty-­seven recently excavated ships at Yenikapı in Istanbul.46 All of them are eloquent testimony to the dangers of traveling by sea. As mentioned above, small-­scale travel, such as cabotage, would not have been necessarily constrained in the same manner as larger merchant vessels or warships, and the variability of naval travel gets at the crux of the safety and vulnerability continuum. The technological limitations of ships at this time meant that the Byzantine state had a difficult time in projecting sea power into the wider Mediterranean, a central element in the security afforded to island environments.47 In the sixth century, trade and connectivity were safeguarded by the general absence of human threats beyond low-­level piracy, but by the late seventh century this had changed drastically.48 Many island sites on Crete, Cyprus, and in the Aegean were coastal, and fortifications were either nonexistent or in disrepair.49 Indeed, as the Persian wars were getting underway at the beginning of the seventh-­century, Justinianic fortifications of the type that existed in Syria, for example, were largely absent, with the majority of building effort of the previous century having been directed toward ecclesiastical constructions.50 When Arab raiding operations in the Mediterranean began, most island communities were unprepared. These raids, some of which were relatively large scale, with tens of thousands of participants, took place nearly annually and their depredations were a real concern for coastal communities and shipping51 Fortifications appeared at Constantia in Cyprus at this time in a rushed and slapdash fashion, chopping up the classical city and protecting its cultural heart, the church of Agios Epiphanios.52 Coins from the reigns of Heraclius (610 to 641) to Constantine IV (668 to 685) were found in the fill of that wall, pointing to a construction in the years shortly after the first raids began.53 Walls were built somewhat earlier than the Arab raids at Amathous, likely due to a short-­lived threat from Egyptian rebels, but were seriously upgraded toward the end of the seventh century.54 The acropolis of Gortyn on Crete was fortified at this time as well, and defenses existed in the Aegean on Aegina and at Emporio on Chios, while other coastal cities, such as Kourion on Cyprus, were abandoned sometime during this period.55 The foundation of sites such as Episkopi in Cyprus and the late seventh-­or eighth-­century settlement of Vizari in the upland Amari valley in Crete indicates that there was also some movement away from the coast.56 NAVAL FORCES AND RAIDING

The fact of the matter was that a Byzantine naval expedition was a serious undertaking that could really only be directed against another major fleet.

88

I A N R A N DA LL

After the Battle of the Masts in 653, in which the Byzantine navy was virtually annihilated and Emperor Constans II barely escaped with his life, the Byzantine and Arab fleets were largely on equal footing, even with the advantage offered by Byzantium’s Greek fire, which seems to have been effective only under very specific conditions.57 The standard Byzantine warship was the dromon or chelandia. This was a two-­decked bireme of about 31.5 m in length that housed a hundred oarsmen, as well as additional officers and soldiers who engaged in combat.58 John Pryor, who has undertaken exhaustive studies of these ships, estimates that the crew of each of them would require over a ton of water each day while at sea, thus severely limiting their range.59 Moreover, their maximum speed is estimated to have been about three to four knots under oars, a rate that would have been increased by favorable sailing conditions, but that was used as a minimum for purposes of calculation.60 Islands and coastal bases were essential as stepping-­ stones for these expeditions, and thus the loss of an island, such as that of Crete in 827 to a marauding band of Muslim exiles from al-Andalūs, was disastrous for naval operations.61 Needless to say, logistical support while in enemy territory would not have been forthcoming. There are also indications that the Byzantine state, or at least several emperors, had a severely ambivalent attitude toward the navy. While naval districts had been formed in the Aegean and along the southern coast of Asia Minor by Heraclius during the final wars with the Sasanians in the early seventh century, the commanders of these naval forces were not accorded high rank or station, particularly when compared with those of the army.62 In the tenth-­ century De Ceremoniis of Constantine VII Porphyrogenitus, the admirals of these naval districts and the overall commander of the fleet are listed as the very last imperial attendants for the coronation ceremony in Hagia Sophia.63 The navy of the caliphate would have operated under many of the same constraints as the Byzantine. Early on, the previously Byzantine shipyards at Alexandria and Klysma in Egypt appear to have been utilized, and Coptic shipwrights and crews were employed, the latter proving disastrous at the second siege of Constantinople (717–­718).64 Under the Umayyads, Syrian ports were also refurbished and developed, although these were apparently more vulnerable to Byzantine attack.65 Overall while a dizzying array of terms in Arabic came to be used for warships and merchant vessels, it appears that the caliphate made use of essentially the same types of ships and sailing expertise as the Byzantines throughout the period in question, with the dromon and chelandia being eventually replaced by the Western galea sometime in the eleventh century.66 Conceptualizing the I slamic - ­B yzantine M aritime F rontier

89

All of this allows for the general perception that for both Mediterranean powers, given the vastness of the sea and the technology available, fleets would primarily have acted as an amphibious arm of the army, generally avoiding major engagements and shuttling troops to a place where they could be the most effective.67 Island residents would have been left to their own devices, and raiding parties and pirates would have struck with little or no warning. Only one system of fire beacons is known to have existed, on Acrocorinth in the northern Peloponnese, and would likely have only indicated to civilians they should take cover in fortifications or head inland, rather than summoning any form of help.68 As in many of the other case studies in this volume, fortification and frontier strategies were often decidedly local, with little to no input from central authorities. Raids, however, whether organized by the caliphate or undertaken by bandits and pirates, would likely have been sporadic. The number of locations listed in the Arabic chronicles indicates a scattered pattern of attacks, rather than returning to the same sites year after year.69 The fact that the site of Soloi, whose destruction at the hands of Arab raiders was recorded in the church inscription cited above, was not fortified after being rebuilt is indicative of the nature of the raiding menace.70 With the greater normalization of trading relations in the eighth century there also likely came additional measures for protection. Convoys are not unknown, and while piratical attacks at sea are a major trope in twelfth-­century Byzantine novels, there are also records of the purchase of insurance against the loss of a ship to piracy in Islamic juridical documents of the ninth century. This insurance was sponsored by trading associations within the Arab world such that if an attack did occur, legal proceedings would be targeted at those responsible for raiding the ship in question and then putting in to a Muslim port with the proceeds.71 SPECIALIST TRAVELERS

This complex picture of connectivity, yet also insecurity, produces the possibility of travel by sea, as well as the benefits to be accrued by doing so, but also increases the relative difficulty of effecting maritime passage. In terms of the perception of distance in the cognitive geography of the traveler, or of those residing next to the sea, this increase in the difficulty of passage has the effect also of increasing the subjective distance of a voyage.72 This is significant in assessing the shifts in the cachet of imported objects and the social contexts of their deployment. 90

I A N R A N DA LL

By and large, travel would have been undertaken by specialists. Passengers were not unknown, and we have many accounts of Christian pilgrims, Muslim travelers, ecclesiastics, and diplomats making voyages, but these would have constituted a minority in terms of people on the move in the eastern Mediterranean during this period.73 Some of our best archaeological evidence for maritime specialists comes from the ancient coastal city of Korykos in Cilicia, where there are the remains of around 450 funerary inscriptions from the late fifth to late sixth centuries that list the occupation of the deceased. Of these, around seven percent are in seafaring professions, including harbor masters, murex processors, net repairers, and other non-­sojourners.74 As mentioned above, by the seventh century the days of state-­run shipping had largely disappeared in the Byzantine Empire, and private merchants were now contracted to transport the annona grain to Constantinople and supply the army.75 Private enterprise also moved between the islands and the surrounding areas, with a recent study of late Roman Cyprus showing the major port of Amathus conducting long-­distance trade, while less risky cabotage from smaller ports trafficked significantly with nearby Cilicia and Lebanon.76 With the seizure of Crete by the exiled Andalusī adventurer Abū Ḥafṣ in 824, long-­distance private trade also blossomed between that island and the dār al-­ Islām, particularly with Egypt and the Maghrib.77 This type of shipping would have been high risk, but, if completed, also high in returns. SPACE AND MEANING

Having considered the factors that likely structured perception of the frontier seascape and travel across it, I shall now consider the effect that the shifts in these factors, which took place starting in the seventh century, would have had on cognitive distance, space, and meaning. In many societies space and distance are not neutral concepts, but are charged with social, political, and often cosmological meaning.78 While Wallerstein’s world-­systems theory,79 and the core-­periphery models born from it, often serve as a poor representation of economic flows or culture change, they can be interesting when used to examine the overall perception of the periphery from the core. Casting out from a political or perceived ideological center to the limits of control or cultural affiliation, the periphery in some societies can then be defined as a land of possibility, of unfamiliar and powerful things, and of beings that are charged with the mysticism that distance, and strangeness, brings.80 In considering the Islamic-­Byzantine frontier, the centers have often been perceived as being Constantinople and Damascus/Baghdad, the seats of Conceptualizing the I slamic - ­B yzantine M aritime F rontier

91

imperial political control. In order to approach the conceptualization of space and distance in the minds of those who lived on the periphery’s islands, however, it is important to flip this model on its head. The frontier now becomes the center, and as some scholars have recently argued, the seat of culture process, the point at which culture contact, and ethnogenesis, occurs.81 Seeing the frontier in this way is a common theme throughout this volume, and an essential consideration for avoiding the selfsame mischaracterization of frontier societies often found in the writings of central Islamic and Byzantine elites, and earlier generations of historians and archaeologists. For the islands of the eastern Mediterranean, this then turns Constantinople and Damascus/ Baghdad into those distant origin-­points of cosmological significance, that mystical borderland of strange and wondrous things (figure 4.3). As the home of the caliph, head of the umma, the Muslim community, both Damascus and Baghdad would have had spiritual significance, while Constantinople, for an Orthodox Christian, housed both the patriarchate and God’s anointed ruler on Earth, the emperor. In the fifth-­century Life and Works of St. Daniel the Stylite, St. Theodore of Sykeon has Daniel cry out to a pilgrim bound for Jerusalem that Constantinople is the new epicenter of the Christian cosmos.82 After the failed Muslim attempts to capture Constantinople in 668 and 717, the Christian capital also takes on an eschatological importance for Islam.83 If we are to follow this line of thinking, however, and dive into the workings of Byzantine Christian and Arab Muslim cognitive and cosmological geography, there are other places that also possess these characteristics: Jerusalem and Mecca. The directionality of this focus is apparent from the daily Muslim prayer, ṣalāh, both Christian and Muslim pilgrimage, and even the mappa mundi of later Christian cartographic expression: Jerusalem as the center of the world. While the inhabitants of Cyprus, Crete, and the Aegean islands would have been largely Christian during this period, the conquest of Crete would have injected a substantial Muslim population into that island. That Baghdad was still a focus for the emirate on this far-­flung shore of the caliphate, notwithstanding its Andalusī origin, can be seen from the coinage of the emirs of Crete, on which they pledge their allegiance, as Sunnis, to the ‘Abbāsid caliph.84 While overtly political, this relationship could not help but have had significant religious undertones given the role of the caliph as the stated leader of the umma. This situated cosmographic focus can also be seen in Franklin’s chapter (9) on medieval Armenia, acting there as a similar frame for material negotiations and political action. With the insecurity of travel and the increase in cognitive distance between the center (here, the islands) and the periphery (Constantinople/Baghdad 92

I A N R A N DA LL

Figure 4.3. The islands of the eastern Mediterranean as a center, and its peripheries of meaning in the seventh through tenth centuries CE.

and Jerusalem/Mecca), these places take on even more meaningful positions within local ideological geography. Moreover, the specialists who travel there, pilgrims and sailors, or those who originate therefrom, imperial or caliphal officials, take on added social power as controllers of, in a sense, esoteric knowledge. Given this framework, goods from overseas, particularly Jerusalem/ Mecca and Constantinople/Baghdad, would have held greater significance in certain social deployments. Relics are a good example of this phenomenon, as are other overtly religious items, but this social cachet would also have extended to more mundane objects from great distances. OBJECTS ON THE MOVE

The material entanglements created between the islands and their cognitive peripheries has only recently begun to receive attention, and a fuller examination of their contextual social deployment has yet to be completed.85 As can be seen above, the conceptualization of the maritime frontier provides a framework in which to consider those objects that moved great distances across Conceptualizing the I slamic - ­B yzantine M aritime F rontier

93

this frontier and the social meaning that may have been attached to them in their new contexts. Objects from the dār al-­Islām appear in some numbers on both Crete and Cyprus, pointing to a frontier zone of interaction, similar to some other cases in this volume, as opposed to the textual accounts of depredation and backwardness. Coupled with the aforementioned forms of Byzantine material that continued to move, the following objects from the Muslim world require reconsideration in view of the nature of the maritime frontier, and the special place of some of the islands within it. Due to a treaty negotiated between Byzantium and the caliphate in 688, Cyprus was designated a neutral, demilitarized province of the Byzantine Empire, the taxes from which would be shared between the two empires.86 To Muslim jurists, this placed Cyprus firmly in the realm of a tributary state, a dār al-­‘ahd, a house of covenant, or a dār al-­ṣulḥ, a house of peace, in which Muslim traders were free to operate.87 Little is known about how this “condominium” actually functioned on the ground, however. Some early historians posited a continuing Muslim administration, split with a Byzantine counterpart, but there remains little evidence for this and more recent scholars have largely dismissed this notion.88 Trade clearly did occur, however, with the following forms appearing on the island, particularly on the south coast. Late Roman Amphora 5, which continued to be produced in Palestine after the Arab conquest, has been found on Cyprus at Kourion and Kopetra, with several examples showing morphological similarities to “Umayyad” varieties from Khirbet al-­Mafjar.89 Additionally, oil lamps of eighth-­century Syro-­ Palestinian type appear at Amathous, Kourion, and Paphos, large coastal sites that would have seen extensive trade.90 The presence, or influence, of Muslim traders and craftsmen is also indicated by a late eighth-­century tombstone excavated in Paphos, and a kiln located at Dhiorios of an “Islamic type.”91 Umayyad and ‘Abbāsid coins and the occasional seal also appear, although most lack provenience.92 Despite being an emirate between 824 and 961, Crete has produced surprisingly little material culture that can be linked to the dār al-­Islām. Only a single “Arab-­style” house has been published, from the Knossos area, with no attendant portable material culture.93 George C. Miles undertook an extensive study on the existing numismatic material in the 1970s, the vast majority of which was stray finds, or lacked provenience, but to date no significant study of ceramics or architecture has been published.94 This material however does exist. In 2013 an exhibition catalog for a show at the J. Paul Getty Museum and the National Gallery Museum contained a section referring to recent excavations by the twenty-­eighth Ephorate of Byzantine Antiquities in the 94

I A N R A N DA LL

city of Iraklion.95 These excavations, located in the northwest of the city, in an area that would have been occupied by the Muslim capital of Chandax, exposed a series of private residences, a hydraulic system, and a large craft-­ production center dating to the late ninth and early tenth centuries, including large amounts of pottery, glass, coins, and metal objects. This material has yet to be published in detail. CONCLUSIONS

Given these material movements, the framework presented here for considering the conceptualization of the Islamic-­Byzantine maritime frontier and its material implications is meant to function primarily as a starting point for future examinations. The role of the particular geographical situation of the islands, and the simultaneously connective and isolating medium of the sea, argues against the view of these locations as completely insular outposts of Byzantium. Rather, supported by the material and textual evidence, they can be seen as zones of interaction much like the Caucasus or the Egyptian-­ Nubian frontier, whose inhabitants likely possessed a unique relationship with space, distance, and the meaning attached to objects traveling from distant lands. The social deployment of these objects, when further understood, can help fully characterize the nature of these interactive zones, and situate the Islamic-­Byzantine maritime frontier, like its mainland counterpart, at the center of culture process and identity negotiation in the early years of Islam. NOTES

1. Lightfoot and Martinez 1995; Rowlands 1987. 2. N. Thomas 1991; Whittaker 1994. 3. White 1991. 4. Eger 2015, 4; Straughn 2007. 5. Bonner 1996; Eger 2015; Haldon, 1997; Kennedy 2004. 6. Lauricella (chapter 2), Franklin (chapter 9), and J. Eger (chapter 6), all in this volume. 7. Horden and Purcell 2000. 8. Braund 1996; Constantakopoulou 2007; N. Thomas 1991; Veikou 2015. 9. Eger 2015, 3; Metcalf 2009. 10. Kennedy 2004; Insoll 1999; Esposito 1999; de  Groot 2016; D. Thomas et al. 2006. 11. Constantakopoulou 2007, 1. Conceptualizing the I slamic - ­B yzantine M aritime F rontier

95

12. Broodbank 2000, 10. 13. Constantakopoulou 2007, 119; Veikou 2015, 359; Vionis 2017, 167. 14. Trombley 2001, 145. 15. Haldon 2012, 100; Vogt 2000, 97. 16. Leidwanger 2015, 157; Trombley 2001, 133. 17. Constantakopoulou 2007, 20; Dalché 2000, 61; Leonard 2005, 618; Metcalf 2009, 230; Papacostas 2000, 120. 18. Armstrong 2009, 158; McCormick 2001, 89; Trombley 2001, 137. 19. Leidwanger 2015, 159; see Mullett 2002, 268, for textual accounts of daily sailing in the Aegean during the middle Byzantine period. 20. Horden and Purcell, 2000, 158; McCormick 2001, 103. 21. Horden and Purcell 2000, 142. 22. Christides 1984, 172; Hayes 1980, 376; Metcalf 2009, 240; Poulou-­Papadimitriou 2001, 232. 23. Metcalf 2009, 399. 24. Ibid., 265; Armstrong 2009; Haldon 2012; Hayes 1980; Poulou-­Papadimitriou 2001. 25. Hayes 1972. 26. Pirenne 1939. 27. Metcalf 2009, 223. 28. Stewart 2010, 163. 29. Haldon 2012, 99; Stewart 2010, 163. 30. Haldon 2012, 99; Trombley 2001, 168; Poulou-Papadimitriou 2001, 232; Wickham 2005, 628. 31. Hayes and Kossyva 2012, 173; Rautman 2003; Poulou-­Papdimitriou 2001, 259, figure 20; Romeo and Portale 2004, 973; Vogt 2000, 55. 32. Mullet 2002, 263. 33. Talbot 1996, 168. 34. Mullet 2002, 263. 35. Jenkins, 1949, 9. 36. Mullet 2002, 263. 37. Ibid., 267. 38. al-Tabari 1990, 26. 39. al-­Muttaqī, quoted in Lewis 1974, 211; Christides 1984, 99. 40. Vionis 2017, 167. 41. Horden and Purcell 2000, 137–­138. 42. Pryor 1988, 72. 43. Constantakopoulou 2007, 4; Pryor and Jeffreys 2006. 44. Vionis 2017, 167. 96

I A N R A N DA LL

45. Mullet 2002, 273. 46. Pulak et al. 2015, 39–­73; Trombley 2001, 135. 47. Constantakopoulou 2007, 119; Pryor 2003, 98; Sicking 2014, 494. 48. Horden and Purcell 2000, 153; McCormick 2001, 113; Trombley 2001, 133–­134; Wickham 2005, 716. 49. Vionis 2017, 169. 50. Stewart 2014, 289. 51. See Pryor and Jeffreys 2006, 25–­76, for a synopsis of Byzantine and Muslim naval operations. 52. Stewart 2014, 296. 53. Metcalf 2009, 278; see also Dikigoropoulos 1962, appendix. 54. Stewart 2014, 287. 55. Trombley 2001, 149, 162. 56. Ibid., 149. 57. Pryor and Jeffreys 2006, 383; Theophanes 1997, 482. 58. Pryor 2003, 85–­86. 59. Ibid., 89. 60. Ibid., 92. 61. Sicking 2014, 507. 62. Pryor 2003, 103–­104. 63. Porphyrogenitous 1935; Papadapoulos 2011, 125. 64. Fahmy 1966, 23. 65. Ibid., 52–­56. 66. Ibid., 120, 126; Pryor and Jeffreys 2006, 423. 67. Pryor 2003, 99; Pryor and Jeffreys 2006, 389–­390. 68. Trombley 2001, 156. 69. Ibid., 168–­169. 70. Ibid. 71. Mullet 2002, 269; Trombley 2001, 152. 72. Helms 1988, 10. 73. See McCormick 2001 for a study of early medieval non-­Muslim travelers. 74. Trombley 2001, 139. 75. McCormick 2001, 87; Trombley 2001, 137. 76. Leidwanger 2013, 221–­243. 77. Christides 1984, 117. 78. Helms 1988, 4. 79. Wallerstein 1974. 80. Helms 1988, 4. 81. White 1991; Thomas 1991. Conceptualizing the I slamic - ­B yzantine M aritime F rontier

97

82. Papadapoulos 2011, 121. 83. El Cheikh 2004, 65. 84. Christides 1984, 114. 85. Armstrong 2009; Demesticha 2004, 170–­ 178; Poulou-­ Papadimitriou 2001; Touma 2001. 86. Christides 1984, 170. 87. Ibid., 169. 88. See Metcalf 2009, 478, for a synopsis of the argument. 89. Hayes 2007, 438; Rautman 2003, 172; Touma 2001, 269. 90. Armstrong 2009; Hayes 2007, 438; Touma 2001, 275. 91. Armstrong 2009, 165; Megaw 1950, 108–­109. 92. Metcalf 2009. 93. Georgopoulou 1994, 98. 94. Miles 1970. 95. Andrianakis 2013, 252–­263. REFERENCES

Andrianakis, M. 2013. “Herakleion in Crete.” In Heaven and Earth: Cities and Countryside in Byzantine Greece, edited by J. Albani and E. Chalkia, 252–­263. Athens: Hellenic Ministry of Culture and Sports, Benaki Museum. Armstrong, P. 2009. “Trade in the East Mediterranean in the 8th Century.” In Byzantine Trade, 4th–­12th Centuries: The Archaeology of Local, Regional and International Exchange: Papers of the Thirty-­Eighth Spring Symposium of Byzantine Studies, St John’s College, University of Oxford, March 2004, edited by M. Mundell Mango, 157–­178. Farnham, UK: Ashgate. Bonner, M. 1996. Aristocratic Violence and Holy War: Studies in the Jihād and the Arab-­ Byzantine Frontier. New Haven, CT: American Oriental Society. Braund, D. 1996. “River Frontiers in the Environmental Psychology of the Roman World.” In The Roman Army in the East, edited by D. Kennedy. Journal of Roman Archaeology Supplemental Series 18. Ann Arbor, MI: Journal of Roman Archaeology. Broodbank, C. 2000. An Island Archaeology of the Early Cyclades. Cambridge, UK: Cambridge University Press. Christides, V. 1984. The Conquest of Crete by the Arabs (ca. 824): A Turning Point in the Struggle between Byzantium and Islam. Athens: Ακαδημία Αθηνών. Constantakopoulou, C. 2007. The Dance of the Islands: Insularity, Networks, the Athenian Empire and the Aegean World. Oxford, UK: Oxford University Press.

98

I A N R A N DA LL

Dalché, P. G. 2000. “Portulans and the Byzantine World.” In Travel in the Byzantine World, edited by Ruth Macrides, 59–­7 1. Aldershot, UK: Ashgate. Demesticha, S. 2004. “Amphora Typologies, Distribution, and Trade Patterns: The Case of the Cypriot LR1 Amphorae.” In The Transport Amphorae and Trade of Cyprus, edited by M. Lawall and J. Lund, 170–­178. Aarhus, Denmark: Aarhus University Press. Dikigoropoulos, A. I. 1962. “Cyprus ‘Betwixt Greeks and Saracens,’ A.D. 647–965.” DPhil thesis, Department of History, Oxford University. Eger, A. A. 2015. The Islamic-­Byzantine Frontier: Interaction and Exchange among Muslim and Christian Communities. London: I.B.Tauris. El Cheikh, Nadia Maria. 2004. Byzantium Viewed by the Arabs. Cambridge, MA: Harvard University Press. Esposito, J., ed. 1999. The Oxford History of Islam. Oxford: Oxford University Press. Fahmy, A. M. 1966. Muslim Naval Organization in the Eastern Mediterranean. Cairo: National Publication and Printing House. Georgopoulou, M. 1994. “The Topography of Chandax, Capital of Crete in the Second Byzantine Period (961–­1204).” Cretan Studies 4: 91–­136. de Groot, A. H. 2016. “Kubrus.” In The Encyclopedia of Islam, 2nd ed., edited by P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs. Leiden: Brill Online. Accessed March 7, 2016. http://​referenceworks​.brillonline​ .com/​entries/​encyclopaedia​-of​-islam​-2/​kubrus​-COM​_0534. Haldon, J. 1997. Byzantium in the Seventh Century: The Transformation of a Culture. Revised Edition. New York: Cambridge University Press. Haldon, J. 2012. “Commerce and Exchange in the Seventh and Eighth Centuries.” In Trade and Markets in Byzantium, edited by Cécile Morrisson, 99–­122. Washington, DC: Dumbarton Oaks. Hayes, J. W. 1972. Late Roman Pottery. London: British School at Rome. Hayes, J. W. 1980. A Supplement to Late Roman Pottery. London: British School at Rome. Hayes, J. W. 2007. “ ‘Pottery’ and ‘Clay Lamps.’ ” In Kourion: Excavations in the Episcopal Precinct, Dumbarton Oaks Studies 38, edited by A. H. S. Megaw, 435–­484. Washington, DC: Dumbarton Oaks. Hayes, J., and A. Kossyva. 2012. “Roman Pottery.” In An Archaeological Survey of the Gournia Landscape: A Regional History of the Mirabello Bay, Crete, in Antiquity, edited by L. V. Watrous, 167–­174. Philadelphia, PA: INSTAP Academic Press. Helms, M. W. 1988. Ulysses’ Sail: An Ethnographic Odyssey of Power, Knowledge, and Geographical Distance. Princeton, NJ: Princeton University Press. Horden, P., and N. Purcell. 2000. The Corrupting Sea: A Study of Mediterranean History. Oxford, UK: Blackwell. Conceptualizing the I slamic - ­B yzantine M aritime F rontier

99

Insoll, T. 1999. The Archaeology of Islam. Malden, MA: Blackwell Publishers. Jenkins, R. J. H. 1949. “The Mission of St. Demetrianus of Cyprus to Baghdad.” Mélanges Henri Grégoire, I (Annuaire de l’Institut de Philologie et d’Histoire Orientales et Slaves) 9: 267–­275. Kennedy, H. 2004. The Prophet and the Age of the Caliphates. 2nd ed. Harlow, UK: Pearson/Longman. Leidwanger, J. 2013. “Opportunistic Ports and Spaces of Exchange in Late Roman Cyprus.” Journal of Maritime Archaeology 8.2: 221–­243. Leidwanger, J. 2015. “Early Byzantine Cyprus: A View from the Sea.” In Maritime Studies in the Wake of the Byzantine Shipwreck at Yassiada, Turkey, edited by D. N. Carlson, S. M. Kampbell, and J. Leidwanger, 157–­166. College Station: Texas A&M University Press. Leonard, J. R. 2005. “Roman Cyprus: Harbors, Hinterlands, and ‘Hidden Powers.’” PhD diss., State University of New York at Buffalo. Lewis, Bernard, ed. and trans. 1974. Islam, from the Prophet Muhammad to the Capture of Constantinople, Volume 1, Politics and War. New York: Harper and Row. Lightfoot, K., and A. Martinez. 1995. “Frontiers and Boundaries in Archaeological Perspective.” Annual Review of Anthropology 24: 471–­492. McCormick, M. 2001. Origins of the European Economy. Cambridge, UK: Cambridge University Press. Megaw, A. H. S. 1950. “A Muslim Tombstone from Paphos.” Journal of the Royal Asiatic Society of Great Britain and Ireland 3.4: 108–­109. Metcalf, M. 2009. Byzantine Cyprus 491–­1191. Nicosia: Cyprus Research Center. Miles, G. C. 1970. The Coinage of the Arab Amirs of Crete. Numismatic Notes and Monographs 160. New York: The American Numismatic Society. Mullett, M. E. 2002. “In Peril on the Sea: Travel Genres and the Unexpected.” In Travel in the Byzantine World, edited by Ruth Macrides, 259–­284. Aldershot, UK: Ashgate. Papacostas, T. 2000. “The Economy of Late Antique Cyprus.” In Economy and Exchange in the East Mediterranean during Late Antiquity: Proceedings of a Conference at Somerville College, Oxford, 29th May, 1999, edited by S. Kingsley and M. Decker, 107–­128. Oxford, UK: Oxbow. Papadapoulos, A. 2011. “Exploring Byzantine Cartographies: Ancient Science, Christian Cosmology, and Geopolitics in Byzantine Imperial-­Era Mapping.” Essays in Medieval Studies 27: 117–­131. Pirenne, Henri. 1939. Mohammed and Charlemagne. New York: Norton. Porphyrogenitous, Constantine VII. 1935. Le Livre des Cérémonies. Translated by A. Vogt. Paris: Société d’Édition “Les Belles Lettres.”

100

I A N R A N DA LL

Poulou-­Papadimitriou N. 2001. Βυζαντινή κεραμική από τον ελληνικό νησιωτικό χώρο και από την Πελοπόννησο (7ος–­9ος αι.): μία πρώτη προσέγγιση. In Οι Σκοτεινοί Αιώνες του Βυζαντίου (7oς—­9ος αι.), edited by E. Kountoura-­Galake, 231–­266. Athens: Εθνικό Ίδρυμα Ερευνών. Pryor, J. H. 1988. Geography, Technology, and War. Cambridge, UK: Cambridge University Press. Pryor, J. H. 2003. “Byzantium and the Sea: Byzantine Fleets and the History of the Empire in the Age of the Macedonian Emperors, c. 900–­1025 CE.” In War at Sea in the Middle Ages and the Renaissance, edited by J. B. Hattendorf and R. W. Unger, 83–­104. Woodbridge, UK: The Boydell Press. Pryor, J. H., and E. M. Jeffreys. 2006. The Age of the Dromon: The Byzantine Navy ca 500–­1204. Leiden: Brill. Pulak, C., R. Ingram, and M. Jones. 2015. “Eight Byzantine Shipwrecks from the Theodosian Harbour Excavations at Yenikapı in Istanbul, Turkey: An Introduction.” The International Journal of Nautical Archaeology 44.1: 39–­73. Rautman, M. 2003. A Cypriot Village of Late Antiquity: Kalavasos-­Kopetra in the Vasilikos Valley. Supplemental Series Number 52. Portsmouth, RI: Journal of Roman Archaeology. Romeo, I., and C. Portale. 2004. “Gortina e il Commercio Mediterraneo: le Anfora da Trasporto tra l’Eta di Augusto e la Conquista Araba.” In Creta Romana e Protobizantina: Atti del Congresso Internazionale (Iraklion, 23–­30 Settembre 2000), edited by Monica Livadiotti, III.i: 959–­973. Padua: Bottega d’Erasmo. Rowlands, M. 1987. “Centre and Periphery: A Review of a Concept.” In Centre and Periphery in the Ancient World, edited by M. Rowlands, M. Larsen, and K. Kristiansen, 1–­11. New Directions in Archaeology. Cambridge, UK: Cambridge University Press. Sicking, Louis. 2014. “The Dichotomy of Insularity: Islands between Isolation and Connectivity in Medieval and Early Modern Europe, and Beyond.” The International Journal of Maritime History 26.3: 494–­511. Stewart, C. A. 2010. “The First Vaulted Churches in Cyprus.” Journal of the Society of Architectural Historians 69: 162–­189. Stewart, C. A. 2014. “Military Architecture in Early Byzantine Cyprus.” In The Archaeology of Late Antique and Byzantine Cyprus (4th–­12th centuries AD): Recent Research and New Discoveries, edited by M. Parani and D. Michaelides, 287–­306. Cahiers du Centre d’Études Chypriotes 43. Paris: De Boccard. Straughn, I. 2007. “Beyond Territorial Sovereignty: Archaeology and the Spiritual Dimension of the Arab-­Byzantine Frontier.” In Social Orders and Social Landscapes, edited by Laura Popova, Charles Hartley, and Adam Smith, 168–­180. Newcastle upon Tyne: Cambridge Scholars Press. Conceptualizing the I slamic - ­B yzantine M aritime F rontier

101

al-­Tabari. 1990. The History of al-­Ṭabarī (Ta’rīkh al-­rusul wa’l-mulūk). Volume 15. Translated by Stephen Humphreys. Albany: State University of New York. Talbot, Alice-­Mary, ed. 1996. Holy Women of Byzantium: Ten Saints’ Lives in English Translation. Washington, DC: Dumbarton Oaks. Theophanes. 1997. The Chronicle of Theophanes Confessor. Translated by C. Mango and R. Scott. Oxford, UK: Clarendon Press. Thomas, D., M. Swanson, and E. Grypeou, eds. 2006. The Encounter of Eastern Christianity with Early Islam. Leiden: Brill. Thomas, N. 1991. Entangled Objects: Exchange, Material Culture, and Colonialism in the Pacific. Cambridge, MA: Harvard University Press. Touma, M. 2001. “Chypre: Céramique et Problèmes.” In Οι Σκοτεινοί Αιώνες του Βυζαντίου (7oς–­9ος αι.), edited by E. Kountoura-­Galake, 267–­291. Athens: Εθνικό Ίδρυμα Ερευνών. Trombley, F. 2001. “Mediterranean Sea Culture Between Byzantium and Islam c. 600–­850 A.D.” In Οι Σκοτεινοί Αιώνες του Βυζαντίου (7ος—­9ος αι.), edited by E. Kountoura-­Galake, 133–­169. Athens: Εθνικό Ίδρυμα Ερευνών. Veikou, M. 2015. “One Island, Three Capitals: Insularity and the Successive Relocations of the Capital of Cyprus from Late Antiquity to the Middle Ages.” In Medieval Cyprus: A Place of Cultural Encounter, edited by Sabine Rogge and Michael Grünbart, 357–­387. Münster, Germany: Waxmann Verlag GmbH. Vionis, A. 2017. “Imperial Impacts, Regional Diversities and Local Responses: Island Identities as Reflected on Byzantine Naxos.” In Imperial Lineages and Legacies in the Eastern Mediterranean: Recording the Imprint of Roman, Byzantine and Ottoman Rule, edited by R. Murphey, 165–­196. New York: Routledge. Vogt, C. 2000. “The Early Byzantine Pottery.” In Πρωτοβυζαντινή Ελεύθερνα, Τομέας 1, edited by Petros Themelis, 37–­205. Rethymno, Greece: Πανεπιστήμιο Κρήτης. Wallerstein, I. M. 1974. The Modern World-­System. New York: Academic Press. White, R. 1991. The Middle Ground: Indians, Empires, and Republics in the Great Lakes Region, 1650–­1815. Cambridge, UK: Cambridge University Press. Whittaker, C. R. 1994. Frontiers of the Roman Empire. Baltimore, MD: Johns Hopkins University Press. Wickham, C. 2005. Framing the Early Middle Ages. Oxford, UK: Oxford University Press.

102

I A N R A N DA LL

Part II

The Southern Frontiers Egypt and Nubia

Source of the Nile, Idrisī, Nuzhat al-mushtāq f ī ikhtirāq al-āfāq, MS Pococke 375, fol. 18b-19a, courtesy of Bodleian Library, University of Oxford

5 In the medieval period, Nubia’s border with Egypt was a Muslim-­Christian border. Yet neither in this period nor at any other time did religious considerations seriously affect the course of Egyptian and Nubian relations. There was just too much else at stake. Economic relations were always the key, for it was in this domain alone that the two lands were genuinely interdependent. New evidence now requires a reassessment of this interdependence. Contrary to what has been argued in earlier generations, we increasingly see indications that the Nubian economy was monetized.1 Nubians used currency for private exchanges. This was certainly true in northern Nubia, and most likely true throughout the Nubian heartland as well. In his definitive discussion of Nubian trade, Derek Welsby recognized that the “big question” about this trade was “what were the people of Nubia able to give in return?”2 Welsby is cautious with his answers, but proposes gold as one of the crucial options. The documentary evidence now available gives us strong reason to believe that some of the gold in question was cash money. But scholars of Nubian studies raise the reasonable objection that Nubian archaeology provides no physical evidence for a monetized economy anywhere in Nubia. The following discussion presents a hypothetical model demonstrating how the Nubian border with Egypt might have served as a revolving door for that money. This model might both explain the archaeological silence about currency in Nubia

Monetization across the Nubian Border A Hypothetical Model Giovanni R. Ruffini

DOI: 10.5876/9781607328773.c005

105

and enrich our understanding of the economic activity across the Egyptian-­ Nubian border. This model fits comfortably within the larger model of Islam’s “minor frontiers” presented throughout this volume. Taking Brauer’s (1995) observations about Islamic frontiers as our starting point, we recognize the Nubian frontier almost at once to be a zone of contact rather than a sharply defined border.3 Brauer’s core-­periphery model saw the core as exerting a dominant social, religious, and cultural ideology over the periphery.4 In the Nubian case, we would speak rather of Egypt’s economic dominance over the periphery. But recognizing the flaws in Brauer’s model helps us make even more sense out of the Nubian case. Frontiers are not simply the sites of interaction between two powers, but are also the sites of “local expression, cultural production, and human agency.”5 In Nubia, this agency adopts Egypt’s economic standards but fits them to local ideological needs, and uses them as local expressions of power. The fact that archaeology uncovers the textual evidence for this process, but not the numismatic evidence, is part of the challenge we seek to address. First, a word about the Nubian people is in order. This chapter makes several statements about the motives and actions of Nubians in a most general sense. Following the model set out by one of the great archaeologists of Nubia a generation ago, this chapter takes the Nubian people to be “a culturally and ethnically homogeneous population of largely African ancestry.”6 Today’s Nubians, in southern Egypt and northern Sudan, are predominantly Muslim. While Muslims were certainly present in medieval Nubia, the process through which Nubians became a Muslim people is still poorly understood.7 Generally, this chapter’s statements about the Nubian people refer to medieval Christian Nubians, who spoke earlier forms of today’s Nubian languages. A considerable range of commodities crossed the Nubian-­Egyptian border.8 By the medieval period the Christians of northern Nubia no longer controlled the sources of ivory, ebony, ostriches, and the like; access to these resources had retreated too far to the south. The one precious export resource that they did still control was that of dark-­skinned slaves, and they were very much at the core of medieval trade relations. The slave trade generally leaves no traces in the archaeological record.9 That record tells us much more about commodities moving in the other direction, particularly wine, pottery, and textiles. Let us take wine as an example. Greeks that they were, the Ptolemaic rulers of Egypt had in an earlier period converted a considerable part of the country’s cropland into vineyards. In the end they converted both the Egyptians and the Nubians from beer drinkers into wine drinkers. The volume of wine importation into Nubia, especially from Byzantine Egypt, reached what can only be 106

G I OVA N N I R . R U F F I N I

called staggering proportions, as attested by the thousands of amphora fragments that make up such a large part of our potsherd collections from the fourth to the seventh centuries.10 We do not know as much as we would like about how importation worked, either for wine or for other goods. We do however have written records from the excavation of the Nile’s west bank, which include shipping notices hinting at the process.11 One such ostrakon contains a shipment notice of 50 artabas, or some 1,350 liters of wheat loaded onto “the skiff of Rhobia.”12 This might be a reference to the Rabī‘a Arabs, of which the Banū Kanz who controlled Aswan (Uswān) were a branch.13 If this guess is right, we get an interesting sense of the permeability of the border. These Arab tribal groups often caused considerable unrest in Upper Egypt, but their primacy in the area gave them control of trade south of Aswan, and thus the ability to move back and forth from Egypt to Nubia. Indeed, this facility of motion gives the Bānū Kanz an ultimately pivotal role in the collapse of Dongola, the capital of medieval Nubia, and the Arabization of some part of the Nubian people. That is a separate story, one unrelated to the main point here: this border permeability plays a crucial role in this chapter’s hypothetical model for the monetization of the Nubian economy. The bulk of the relevant evidence here appears in Nubian accounts from around the twelfth century. We have roughly a dozen texts from Lower Nubia, all excavated at Qasr Ibrim, detailing hundreds of payments in gold, silver dirham (the contemporary term for silver coin in the Islamic Mediterranean), and various goods: wine, grain, slaves, and so on.14 These units of exchange are also the same units of exchange appearing in Nubian legal documents. Qaṣr Ibrim has given us land sales and other legal devices documenting the exchange of property for gold, silver dirham, and slaves.15 Despite the claims of conventional historiography, nowhere in this picture do we see the Nubian king claiming ownership over all land.16 Nor do we see any suggestion that the Nubian people are his slaves, by legal fiction or otherwise. Scholars intent on preserving that conventional picture, obtained largely from the Arab literary evidence, may argue that our legal texts and accounts come primarily from Lower Nubia, where even Arab sources admit that some coinage circulated. But this is an accident of the archaeology: if we had comparable documents from Upper Nubia, they would likely reveal a similar picture. Property lists from Qaṣr Ibrim record plots of land in Lower Nubia owned by churches in Dongola, the Nubian capital.17 Similarly, we see royal decrees issued at Dongola addressing property concerns in Qasr Ibrim.18 Some of the major landowners in and around Qaṣr Ibrim were court officials appointed M onetization across the N ubian B order

107

from Dongola.19 It is hard to imagine a unified state tolerating partial monetization and privatization in a northern province while somehow forbidding or preventing it in its southern provinces. The evidence from the Nubian legal texts and accounts seems to provide convincing proof of a complex economy throughout the country, as do the texts demonstrating economic links between Dongola and Ibrim, between center and periphery. They also tell us something about the mentality hiding behind that economy. Earlier studies of these accounts drew attention to something strange in their structure: the decision to organize the entries in paragraphs rather than in columns makes it very difficult to add the entries to reach a total.20 Indeed, none of the accounts has any totals, implying that such addition was not part of the purpose of these accounts. But further elaboration is possible. These accounts include no attempt to reach a total sum for the amounts collected because that information was not useful. The total amount collected did not matter. This situation makes sense only if total incomes need not bear any relation to total expenses. So, for Nubian accountants, what mattered was who paid what. The relationship between the totals received and any future expenses on behalf of the estate owners was essentially irrelevant. Several conclusions follow logically from this fact. First, Nubian estates—­the presumptive source of most of these accounts—­had no set budget. They had no need to plan individual expenditures in advance or in detail. This explains so many of the Nubian letters from this period. We rarely see personal content of any kind in these letters.21 Quite the opposite: Nubian letters are most often instructions for the payment of certain amounts of grain, wine, or gold. These payment orders rarely come with explanations, but if they were part of a regular or budgeted cycle of payments, no orders would even be needed. Instead, these payment orders all have an ad hoc feel to them, precisely what we would expect from a system lacking any budgetary structure to detail specific expenses in advance. Second, the profit motive was not a prime consideration. Some Nubians, presumably limited to members of the elite, could buy land as a form of investment, and use that land as a source of wealth to cover necessary expenses both public and private.22 But this does not mean that they tried to accumulate capital or engaged in speculative buying and selling to aid such accumulation. To the contrary, Nubians seem less interested in what profit they can make from their property than in what public impression they can make in spending their wealth. This is why Nubian legal documents so thoroughly describe the public feasts held after land sales for all of the participants; why Nubian church inscriptions so thoroughly describe donations made to those churches; 108

G I OVA N N I R . R U F F I N I

and why Nubian church texts warn these donors to expect no more from their donations than the good of the deed itself.23 Nubians wanted to impress their peers with their wealth, not accumulate it. A third conclusion is that price was fluid. If cash expenses were not fixed in advance, payments must have been inherently more negotiable. We do see recurring payment amounts: the dozens of payments of 12 silver dirham across multiple accounts suggests that 12 dirhams was a standard rate of payment for something. But intermingled with these recurring payments is a wide range of unpredictable amounts, payments in cash mixed with payments in kind: the wine, grain, and slaves we have already mentioned.24 When an economy is monetized, but not completely, it is easy to imagine scenarios in which prices fluctuate not simply in response to market factors, but also in keeping with what goods were on hand, in keeping with one’s ability to pay. This partially monetized economy did not operate in a vacuum, but through the mediation of the Egyptian border. Nubians traveled to and received travelers from the wider world, remained aware of developments in that world, and reacted and planned for the future according to what they knew of that world. The Arab literary tradition provides more proof of this than we have previously realized. A rarely noticed story in the works of Ibn al-­Āthīr (1160–­1234) describes the palace intrigue in Muslim Egypt against the great counter-­Crusader, Salaḥ al-­Dīn.25 Court officials in Cairo had agreed to invite Crusaders from Sicily and the Syrian coast to Egypt. Several of the plotting group are described as prominent among the blacks (al-­sūdān), a term here almost certainly referring to Nubians. We can imagine high-­ranking Nubian expatriates in communication both with Europeans overseas and Nubians back home. Other documentary and literary evidence confirms this picture. The Polish excavations from Banganarti discovered a church graffito from the second half of the thirteenth or early fourteenth century written in Provençal, a Romance language from Southern France.26 The Qaṣr Ibrim finds include a piece of paper apparently mixing Old Nubian with an Italian plea to Santa Maria (the Virgin Mary) and Santa Simeouon (Saint Simeon), aïouta mi: “help me.”27 A Crusader narrative from 1203 records the presence of a Nubian king in Constantinople after an earlier pilgrimage to Jerusalem.28 All of these examples taken in isolation are hard to interpret. Taken together, they form the outline of a considerable network of travel in and out of Nubia, across the Egyptian border and into the wider Mediterranean world. Indeed, this travel and communication all but demand a wider level of economic integration between Nubia and the wider world than scholars have previously understood. M onetization across the N ubian B order

109

Claims about the role of money in medieval Nubia, summarized earlier in this chapter, are perhaps less controversial than subsequent speculation about where that money actually came from. Nubians did not mint their own coinage, but instead employed money coming to them across their northern border. As we have seen, Nubian accounts and Nubian land sales detail payments in both gold and silver, the latter explicitly described as dirham. Where gold and silver exist side by side as units of exchange, we might look for some rate of exchange between them.29 Nubian accounts—­most of which date to the late twelfth century—­record over 40 payments in gold and over 120 payments in silver dirham. Almost all of these payments in silver (119 out of 121) are for amounts under 40 dirhams. Absence of higher amounts suggests a value for gold: payments of (say) 50 or 60 dirham would appear in these accounts as one gold piece and 10 or 20 dirham. This suggests an exchange rate in medieval Nubia of one gold piece to slightly less than 40 silver dirham. It is surely no coincidence that this is the exact exchange rate that held in contemporary Islamic Egypt. We have ample data on the gold-­to-­silver exchange rates in Egypt in this period, thanks in part to the rich documentary data of the Cairo Geniza.30 These data attest to an exchange rate of 40 silver dirham to one gold piece in the 1190s, the approximate period of several of our Nubian texts. More generally, the exchange rates in mid-­to late twelfth-­ century Egypt range from 34.3 to 42 dirham to one gold piece. The data from the Nubian accounts fit this approximate range of exchange rates. This not only supports our conclusion that the Nubian economy was monetized. It also suggests that its monetary system was connected in some way to that of its northern neighbor. The Nubian border is thus less an economic barrier than an arbitrary political marker inside a larger region of more or less freely circulating gold and silver currency. Further evidence for this currency zone comes from an account of wine and slaves excavated at Qaṣr Ibrim in 1964 (P.QI 4.74). The text is in Old Nubian, but its paleography suggests that it is somewhat older than the bulk of the Old Nubian texts published from Qasr Ibrim.31 It contains two explicit references (one partially restored) to ngapir alhekkimou b, that is, apparently, two pieces of al-­Ḥakim gold. It is not obvious what should be special about Egyptian gold coins minted specifically in the caliphate of al-­Ḥākim (995–­1021). Inflationary crises under al-­ Ḥākim spurred reformation of the relationship between Faṭimid silver and gold, which apparently reassured the Egyptian public.32 But more research on Islamic numismatics is needed to explore the point in full. Al-­Ḥākim was the longest-­reigning of the Faṭimid caliphs for the first century after the death of the dynasty’s founder, and in this period, the vast 110

G I OVA N N I R . R U F F I N I

Figure 5.1. Model of currency circulation, Egypt and Nubia.

majority of the gold coins fell into a very narrow weight range, averaging slightly over four grams.33 We can imagine that at some point in the early Faṭimid period, awareness of the stability of Egypt’s gold coinage crossed the border into Nubia, and with that awareness came a tendency to call that coinage by the name of one of its long-­lived monarchs. This model works on a theoretical level. The numbers create a convincing picture. But we do not yet have an explanation for the origins of that picture. If Nubians used Egyptian currency, or some form of money pegged to that currency, we need a historical explanation for this situation. Understanding the Nubians to be active agents in the creation of their economic environment, we should look for the factors exposing Nubia to Egyptian currency and leading them to conclude that adopting that currency made the most sense for their economy. Any model we produce should take into account the chief archaeological objection to the role of money in medieval Nubia: the relative lack of coinage among archaeological finds. Figure 5.1 presents a theoretical model for currency circulation in Egypt and Nubia. It is a seven-­step cycle in which Egyptian currency enters Nubia, that currency leaves the hands of the Nubian elite and enters wider circulation, that currency returns to the elite in the form of rent and taxes, and leaves the hands of the elite to return to circulation in Egypt and the wider world. Let us take those steps in order. First, Egyptians bought Nubian goods and services. M onetization across the N ubian B order

111

We know that this happened both in Egypt and in Nubia. A letter from a Nubian named Petro “suffering a great loss” in Egypt records his sale of a slave to Egyptians for dates, barley, and gold coin from the reign of the caliph al-­ Ḥākim.34 We know also from the testimony of al-­Mas‘ūdī that Muslims resident in Egypt bought Nubian land; while al-­Mas‘ūdī does not specify, gold coin is the purchase device most commonly named in Nubian land sales.35 The inevitable result of this process is the entry of Egyptian money into Nubia in the hands of the Nubian elite. The next step is money leaving the hands of the Nubian elite and circulating more widely within Nubia. We know that this happened from roughly twenty land sales, many of which give sums of money in gold coin as the explicit sales price.36 We also know this from correspondence in which the eparch, or governor, of Nobadia (Lower or Southern Nubia) gave explicit orders to sell grain for gold.37 All of this implies the circulation of money outside of the hands of the elite, unless we imagine Nubia’s upper classes existing in a closed economy, conducting transactions only with each other, and only through means of exchange they kept unavailable to others. But once that money left elite hands, it inevitably returned to it. The chief way this took place was through the payment of rents and taxes. The Nubian documentary evidence tells us nothing about rent and tax rates. My own study of the subject used payment patterns in the Nubian accounts to guess what those rates may have been.38 Nubian accounts have a number of disproportionately frequent payments: six dirhams and twelve dirhams stand out in particular. These amounts may have related to the standard units of land, plots organized around the amount of land irrigated by a single saqiya, or waterwheel, drawing water from the Nile. These recurring payments suggest that after currency circulated into the hands of smaller landowners and tenant farmers, it may have returned to the larger landowners and the state in the form of six dirhams or twelve dirhams rent or tax payments. The final step is the most challenging: the return of Egyptian currency from the Nubian elite to Egypt. A number of avenues for this final step can be found, including blood money, trade, and travel. We know that the Egyptian government threatened to impose blood money on Nubia. In 759, they instructed the Nubian king to pay a thousand dinars (of gold coin) should the merchant of a man named Muḥammad ibn Zayd indeed have been killed in Nubia.39 We may suppose these demands only made sense if Nubia had cash with which to pay. Should they have ever done so, it would have represented a considerable return of cash to the Egyptian economy. The baqt treaty, which governed relations between the Muslim world and Nubia from the seventh century to the twelfth, stipulated the annual payments 112

G I OVA N N I R . R U F F I N I

made by Nubia to the Muslims and vice versa.40 The Nubians paid—­or were expected to pay—­over 400 slaves to the Muslims each year. In return, the Muslims paid to the Nubian king and his delegates considerable sums of food, drink, clothing, and other items. We may suppose that this exchange represented supply and demand: the Muslim world had a shortage of slaves, and Nubia had a shortage of food and drink. We may further suppose that when Muslim merchants went to Nubia, they addressed this shortage, providing food and drink and receiving cash in exchange for it. We know that Islamic glazed wares were used as far away as southern Nubia, and these wares might have been purchased with coin as easily as obtained through barter.41 Similarly, the tiraz shroud fragments found in the Qaṣr Ibrim cathedral cemetery (dating from the early 900s to the early 1000s) bear the names of Egyptian caliphs, and may have been purchased with coin as easily as obtained through barter.42 Travel would have been another major way our proposed currency transfer took place. We see members of the Nubian royal family in Egypt on several occasions. In the 830s, the Nubian king Zacharias sent his son George on an embassy first to Egypt and then to Baghdad.43 In the eleventh and twelfth centuries, King Solomon and King George abdicated the throne of Dotawo and retired to Egyptian monasteries.44 In the thirteenth and fourteenth centuries we see a depressing series of Nubian civil wars in which one king after another falls at the hands of a close relative with the help of military aid from Cairo.45 That help came most often after one claimant or another made the trip to Cairo in person to secure that aid. All of these trips would have come at considerable expense, and brought considerable Egyptian currency back to Egypt from Nubia. This return trip brings us full circle. We have brought Egyptian currency into Nubia through the purchase of Nubian goods and services. We have circulated that money through Nubian hands in the sales and rent-­tax transactions documented in Nubian legal texts and accounts. We have walked that currency back across the border into Egypt. All of this has obvious archaeological implications. If Nubia had employed a closed monetary system, we might well be puzzled by the relative scarcity of coinage in Nubia’s archaeological record. But if Nubia’s monetary system was part of a larger area of currency circulation, or a revolving door with neighboring Muslim Egypt, that relative scarcity is at least partly explained by the constant flow of coinage both into and back out of Nubian territory. We can also imagine—­as the political chaos and economic decline of Christian Nubia’s final generations began—­a model in which all of Nubia’s cash reserves fled the country. As state structures and security declined, Nubian M onetization across the N ubian B order

113

goods and services became both less available and less desirable, ensuring that those cash reserves would never be replaced. This decline in state structures is to some degree inextricably intertwined with the permeability of Nubia’s border. By the 1300s, the Bānū Kanz are not simply the mediators of that border, an influential force in Upper Egypt, but are penetrating as far as Dongola. They are intermarrying with the Nubian royal family, sparking multiple civil wars, and beginning the creation of a new branch of Nubian people, speakers of modern Dongolawi-­Kenzi. As with so much else in the historiography of medieval Islam, Ibn Khaldūn is at the heart of the traditional picture of the Islamization and collapse of medieval Nubia. In this picture, Arab tribes made the Nubian state “a place of pillage and disorder” and Nubian kings—­in desperate attempts to restore order—­married their daughters in alliance with the invaders. Nubia’s traditions of matrilineal succession meant that power soon fell into the hands of Arab nomads, who proved unable to maintain the centralized state.46 The archaeology of this transition is not widely studied. Indeed, archaeologists have argued that Islamization in the Sudan is an ongoing process of gradual transformation.47 Our model of monetization across the Nubian border does not challenge this traditional picture as much as supplement it. Nubia’s border with the Islamic world is not a firm, fixed border, disappearing the moment Nubia is Islamized. It is, instead, the site of an inherently unstable feedback loop, in which the permeability of the border in the medieval period creates a complex economic relationship between Nubia and Egypt, an economic relationship opening the gates to forces that ultimately destabilize one of the two powers and permanently alter the nature of the frontier. This chapter’s model of monetization across the Nubian frontier requires—­through Nubian use of Egyptian currency—­a certain degree of Nubian dependence upon the Egyptian economy. Perhaps that dependence on Egypt, and concurrent economic forces sending Nubian wealth north to Egypt, help explain the gradual loss of Nubian independence and the collapse of Christian Nubia at the end of the medieval period. These economic forces are working side by side with, or are embedded in, the social relations found in Ibn Khaldūn’s traditional model of medieval Nubia. NOTES

The spoken version of this chapter was developed in conjunction with Bill Adams, to whom I continue to be grateful for his support of my exploration of Nubian studies. 114

G I OVA N N I R . R U F F I N I

While I retain with his permission some of the material he wrote for that spoken paper, he neither supports the argument about nor shares the conclusions I draw on the nature of the Nubian economy. Abbreviations for published Nubian texts follow the standards proposed in “A Guide to the Texts of Medieval Nubia” online at www​.medievalnubia​.info.

1. I keep citations in this chapter to a minimum, in part because of its speculative nature, and in part because it generally builds upon the arguments I have made on the nature of the Nubian economy elsewhere, where more detailed citations are provided: see Ruffini 2012, 171–­206, and Ruffini 2014a, 27–­30. 2. Welsby 2002, 211. 3. A. Eger (chapter 1), this volume. 4. Ibid. 5. Ibid. 6. Adams 1979, 727. 7. Ruffini 2012, 233–­235. 8. For Nubia’s international trade, see Welsby 2002, 204–­215. 9. Ibid., 209–­212. 10. Adams 2013, 151. 11. Ruffini 2010, 232–­234. 12. Ibid., 232: at the time, I was unhappy with the name, and could only suggest it to be a variant of the female Arabic name Râbi̓a. 13. Vantini 1975, 634. 14. Most of these texts are only recently published. See in particular P.QI. 4.71, 4.72, 4.73, 4.75 and 4.76. 15. P.QI 3.32, 3.34, 3.37, 3.38, 3.39, P.QI 4.63, 4.64, 4.65, 4.67, and 4.69. 16. For further discussion of the origins of this historiography, and my rebuttal, see Ruffini 2012, 61–­75. 17. P.QI 3.60. 18. P.QI 3.30 and 4.66. 19. Gourresi, Eparch of Nobadia: P.QI 4.64 and 4.65. Marturokoudda, great eparch (of Nobadia): P.QI 4.67 and 4.69. 20. Ruffini 2012, 187–­188. 21. For an introduction to the study of Old Nubian letters, see Ruffini 2014b, 961–­970. 22. Ruffini 2012, 202–­206. 23. Feasts: e.g. P.QI. 3.32. Church inscriptions: Ochała 2014, 971–­976, figures 1–­2. Church texts: Browne 1983, 97–­112. See also the discussion on these issues in Ruffini 2012, 90–­111. 24. See e.g. P.QI 3.61, 3.62, 4.71, 4.72, 4.73, 4.74, 4.75, 4.76, and 4.80.

M onetization across the N ubian B order

115

25. Vantini 1975, 355–­359. 26. Łajtar and Płóciennik 2011, 95–­120. 27. Unpublished P.QI.inv. 74.1.29.7A, a photograph of which is held in the British Museum Qasr Ibrim Archive. 28. Rostkowska 1982, 113–­118. 29. Ruffini 2012, 174–­178. 30. Goitein 1967, appendix D. 31. P.QI 4.74, introd. 32. Walker 2009, 216. 33. Ehrenkreutz 1956, 179–­180. 34. Hagen 2009, 117; Hagen 2010, 722; Adams 2010, 245. 35. Vantini 1975, 134. For gold coin in land sales, see note 14 above. 36. Note 11 above with discussion at Ruffini 2012, 22–­31. 37. P.QI 3.46. 38. Ruffini 2012, 193–­202. 39. Adams 2010, 248; Burstein 2009, 152–­153. 40. Welsby 2002, 70. 41. Islamic glazed wares: Welsby 2002, 194 and 213. 42. Crowfoot 2011, 15–­16. 43. Vantini 1975, 194–­195. 44. Solomon: Vantini 1975, 331–­332. George: van Gerven Oei 2011, 225–­262. 45. Welsby 2002, 243–­255; Ruffini 2013, 179–­191. 46. Welsby 2002, 249. 47. Edwards 2004, 258. REFERENCES

Adams, William Y. 1979. “On the Argument from Ceramics to History: A Challenge Based on Evidence from Medieval Nubia.” Current Anthropology 20.4: 727–­744. Adams, William Y. 2010. Qasr Ibrim. The Earlier Medieval Period. London: Egypt Exploration Society. Adams, William Y. 2013. Qasr Ibrim. The Ballaña Phase. London: Egypt Exploration Society. Brauer, R. 1995. “Boundaries and Frontiers in Medieval Muslim Geography.” Transactions of the American Philosophical Society 85.6: 1–­73. Browne, Gerald M. 1983. “Griffith’s Nicene Canons.” The Bulletin of the American Society of Papyrologists 20: 97–­112. Browne, Gerald M. 1991. Old Nubian Texts from Qasr Ibrim III. Texts from Excavations 12. London: Egypt Exploration Society. 116

G I OVA N N I R . R U F F I N I

Burstein, Stanley Meyer. 2009. Ancient African Civilizations: Kush and Axum. Prince­ ton, NJ: Markus Wiener. Crowfoot, Elisabeth. 2011. Qasr Ibrim: The Textiles from the Cathedral Cemetery. London: Egypt Exploration Society. Edwards, David N. 2004. The Nubian Past: An Archaeology of the Sudan. Oxon and New York: Routledge. Ehrenkreutz, Andrew S. 1956. “The Crisis of Dīnār in the Egypt of Saladin.” Journal of the American Oriental Society 76: 178–­184. Goitein, Shlomo D. 1967. A Mediterranean Society: The Jewish Communities of the Arab World as Portrayed in the Documents of the Cairo Geniza. Vol. 1. Berkeley: University of California Press. Hagen, Joost L. 2009. “Districts, Towns and Other Locations of Medieval Nubia and Egypt, Mentioned in the Coptic and Old Nubian Texts from Qasr Ibrim.” Sudan and Nubia 13: 114–­119. Hagen, Joost L. 2010. “ ‘A City that Is Set on a Hill Cannot Be Hid.’ Progress Report on the Coptic Manuscripts from Qasr Ibrim.” In Between the Cataracts. Proceedings of the 11th Conference for Nubian Studies, Warsaw University, 27 August–­2 September 2006, II.2, Polish Archaeology in the Mediterranean Supplement Series 2.2/2, edited by W. Godlewski and A. Łajtar, 719–­726. Warsaw: Warsaw University Press. Łajtar, Adam, and Tomasz Płóciennik. 2011. “A Man from Provence on the Middle Nile in the Second Half of the Thirteenth/First Half of the Fourteenth Century. A Graffito in The Upper Church at Banganarti.” In Nubian Voices. Studies in the Christian Nubian Civilization, Journal of Juristic Papyrology Supplement 15, edited by Adam Łajtar and Jacques van der Vliet, 95–­120. Warsaw: Warsaw University Press. Ochała, Grzegorz. 2014. “Old Nubian Lists of Goods and Money: A Preliminary Presentation.” In The Fourth Cataract and Beyond. Proceedings of the 12th International Conference for Nubian Studies, British Museum Publications on Egypt and Sudan 1, edited by Julie R. Anderson and D. A. Welsby, 971–­976. Leuven, Belgium: Peeters. Rostkowska, Bożena. 1982. “The Visit of a Nubian King to Constantinople in A.D. 1203.” In New Discoveries in Nubia, edited by Paul Van Moorsel, 113–­118. Leiden: Nederlands instituut voor het nabije oosten. Ruffini, Giovanni R. 2010. “Nubian Ostraka from the West Bank Survey.” Zeitschrift für Papyrologie und Epigraphik 175: 231–­238. Ruffini, Giovanni R. 2012. Medieval Nubia: A Social and Economic History. Oxford, UK: Oxford University Press. Ruffini, Giovanni R. 2013. “Newer Light on the Kingdom of Dotawo.” In Qasr Ibrim, between Egypt and Africa: Studies in Cultural Exchange, edited by J. van der Vliet and J. L. Hagen, 179–­191. Leiden: Nederlands instituut voor het nabije oosten. M onetization across the N ubian B order

117

Ruffini, Giovanni R. 2014a. The Bishop, the Eparch, and the King: Old Nubian Texts from Qasr Ibrim (P.QI IV). Warsaw: University of Warsaw Press. Ruffini, Giovanni R. 2014b. “May God Increase Your Years: Unpublished Old Nubian Correspondence from Qasr Ibrim.” In The Fourth Cataract and Beyond: Proceedings of the 12th International Conference for Nubian Studies, edited J. R. Anderson and D. A. Welsby, 961–­970. Leuven, Belgium: Peeters. van Gerven Oei, Vincent. 2011. “The Old Nubian Memorial for King George.” In Nubian Voices: Studies in Christian Nubian Culture. JJP Supplement XV, edited by Adam Łajtar and Jacques van der Vliet, 225–­262. Warsaw: University of Warsaw Press. Vantini, Giovanni. 1975. Oriental Sources Concerning Nubia. Heidelberg and Warsaw: Heidelberger Akademie der Wissenschaften. Walker, Paul E. 2009. Caliph of Cairo: Al-­Hakim bi-­Amr Allah, 996–­1021. Cairo and New York: American University in Cairo Press. Welsby, Derek. 2002. Medieval Kingdoms of Nubia: Pagans, Christians and Muslims in the Middle Nile. London: British Museum Press.

118

G I OVA N N I R . R U F F I N I

6 The frontiers of the medieval Nubian states often evoke the riverine landscape of the Nile. For example, in a 2002 overview, Derek Welsby wrote, “virtually no sites [of medieval Nubia] are known outside the Nile valley.”1 This chapter presents new evidence concerning the presence of medieval Nubia west of the Nile, which in turn bears on locating the enigmatic land (or province) of Ṭarī’ mentioned in medieval texts. In general, the medieval Nubian states were not a part of the contemporary Muslim world but did maintain relations with neighboring Islamic states.2 The border between Muslim Egypt and Christian Nubia lies in lower (northern) Nubia, with the First Cataract as the main line of demarcation. This natural rocky barrier of granite forces the Nile into a narrow bed with several rapids, making navigation of the river difficult or impossible. It had defined a natural border zone from pharaonic times on, and fulfilled that role also during the times when it marked the border between Muslim Egypt and the Nubian states of Nobatia and later Makuria. The importance of this rocky frontier zone is indicated by its prominent depiction in many Arab geographical works, as for example in a map illustrating a work of al-Idrīsī, where the cataracts as border zone between Egypt and Nubia are indicated by a large jabal across the river course of the Nile (south or Assuan). Thus, Makuria’s southwestern desert frontier, the topic of this chapter, cannot be regarded as a border

The Land of Ṭarī’ and Some New Thoughts on Its Location Jana Eger

DOI: 10.5876/9781607328773.c006

119

zone of the Muslim world in a narrow sense of the word; nevertheless, Arabic and Persian scholarly works (such as the geographical handbook Ḥudūd al-­ ʿĀlam)3 mention the southern and western border situations of the medieval Nubian states, indicating the general interest of the Muslim world in the contact zone of medieval Nubia with its southern and south­western neighbor cultures. For that, it can be assumed that the outer periphery of Makuria and the other Christian states of medieval Nubia were seen by the Muslim authors as lying at the frontier of their own sphere of interest. This idea is also supported by the general role of Nubia not only as a target of Egyptian trade and commerce in antiquity as well as the medieval period, but also as a corridor to the south.4 Thus, Lower Nubia south of the First Cataract of the Nile might have functioned not only as the primary frontier between the Muslim Egyptian and the Nubian territories—­that is, as a direct contact zone into Nubia5—­but also as an indirect frontier toward sub-­Saharan Africa. Only a few traces of medieval Nubian presence were found in the western Sudan. Since the 1920s, the large, ruined, fortified settlements of Zankur and Abū Sufyān in Kurdufān had been dated to the medieval period.6 But, after French explorations of the early 2000s, they were redated to the late Iron Age.7 Other traces of connections of medieval Nubia to the west were scarce but included sherds from pilgrim bottles in the Wādī Malik, which resemble some examples found in Dongola, the capital of the medieval Nubian state of Makuria. Some bowls, closely related to ceramic wares found around Dongola and dating to the Christian medieval period (seventh to tenth centuries) were also found in the ruins of ‘Ayn Farah in Darfur; they are generally attributed to the medieval Tunjur chiefdom.8 It remains unclear how those finds could fit into a trade and communication system around the Nubian states of Makuria, Nobatia, and Alwa. The border organization of these three states—­especially the borders to the deserts of the west and east—­had until recently not been well investigated. The border between Makuria/Nobatia and Islamic Egypt is relatively well known due to the works of Arabic authors, but the border zones in the southwestern and southeastern deserts are known only by indirect implications from the written sources and some solitary archaeological and epigraphical records. A medieval manuscript, the Ḥudūd al-­ʿĀlam, provides some information about a land (or province) named Ṭarī’ in a border situation between the Nubian-­speaking world and the Sudanic countries. This chapter presents some indications that this land of Ṭarī’ can be identified with the area of the Jabal al-­‘Ayn (situated between the modern Sudanese provinces of North Kurdufān and Northern), and represents the southwesternmost known outpost and 120

JA N A E G ER

Figure 6.1. Satellite image of the possible monastic complex of Jabal al-­‘Ayn. The church appears within the dark circular structure (debris heap) in the upper-­left quadrant. Above and to the left (northwest) of it is the L-shaped, dry-­masonry structure. The dark line running below the structures, from the middle-­left edge to the center of the photograph, is a wall (© Google).

border of Nubian Christianity (and thus of the Mediterranean zone of influence). The fact that the Wādī Malik forms a convenient line of communication from the Jabal al-­‘Ayn to some urbanized centers of the ancient and medieval Kurdufān cultures (mainly Zankur and Abū Sufyān) might lead to the conclusion that the Jabal al-‘Ayn region, or the land of Ṭarī’, was a contact zone between Makuria and Kurdufān. Since Derek Welsby wrote the lines quoted at the beginning of this chapter, much has changed concerning our knowledge of the Nubian west. In 2011, during a field campaign of the University of Cologne in the Wādī Huwār, the author had the opportunity to organize a small survey of the Jabal al-­‘Ayn area north of the Wādī Malik. During that survey, an architectural structure was discovered inside a cirque of the mountain (figure 6.1).9 The structure consisted of a sandstone church (figure 6.2) and another building of dry-­stone masonry based on an L-shaped ground plan, with several box graves and T H E L A N D O F Ṭ A R Ī ’ A N D S O M E N EW T H O U G H T S O N I T S L O CAT I O N

121

Figure 6.2. Ruins of the sandstone church at Jabal al-­‘Ayn.

tumulus graveyards (figure 6.3). The buildings are separated from the outer landscape by a wall (figure 6.4). The church has a three-­nave ground plan with an outside apsis, and the main entrance on the eastern side. The overall measurements are 11 m × 7 m. Disregarding its rather small size, the facts that it was built of sandstone blocks and in the style of a basilica makes it a major Christian center in such a remote area. In fact, no other church in Sudan is known so far away from the Nile. Inside the church, a broken lintel with cross-­shaped and floral decoration was found. There was a graffito of a human figure and geometric motifs at the eastern wall of the church. Additionally, on the northern wall we found a poorly conserved inscription. The state of preservation of the script has so far not allowed identification of the language, but since some Greek letters are still preserved it could be Greek, Coptic, or (most probably) Old Nubian. Scattered across the whole area were found large quantities of medieval ceramics comparable to the wares of the Nile valley and the Bayuda. According to the classification of Adams, the church (despite its relatively small size) is a type 1a structure, since the entrance is situated in the line of the building’s main axis, and the church’s ground plan includes a narthex. Thus, 122

JA N A E G ER

Figure 6.3. Looking west: the church (foreground), the L-shaped structure (beyond the church), and the wall (running into the distance at left) at Jabal al-­‘Ayn.

according to comparable examples from Lower Nubia, the church could be dated to the early Christian period between 550 and 750.10 But it has to be stated that this assumption is based mainly on the position of the entrance, which in this case could be also simply due to topographical necessity. Because of that, the dating of the building has to be seen as preliminary so far. Northwest of the church, another building is built on an L-shaped ground plan from dry-­stone masonry. It was roughly 20 m × 16 m. The entrance is situated at the longer, inner side of the L of the ground plan, pointing to the east. Behind the entrance are traces of one room. Besides that, the former room structure remains unclear, due to the ruined situation of the building and the amount of debris. North of the church a clay-­lined pit, perhaps dug for storage purposes, resembled a stationary bucket. A similar storage facility was, for example, documented close to church Sur 22a at the Fourth Cataract.11 The facts that such an elaborated and massive sandstone church is located in a remote area without any traces of a larger settlement and that the complex is surrounded by a wall, could indicate that the structures were a monastic complex. Surrounding walls are often regarded as a typical component of Nubian T H E L A N D O F Ṭ A R Ī ’ A N D S O M E N EW T H O U G H T S O N I T S L O CAT I O N

123

Figure 6.4. The enclosure wall at the monastery of Jabal al-­‘Ayn.

monasteries, either for reasons of fortification12 or (more likely) as symbolic separation of the secular from the religious realm.13 A quite remarkable detail is that large box graves are situated inside as well as outside the walled area. But, so far, it remains completely unclear whether there are functional or simply chronological reasons for that. The box-­grave cemetery inside the wall contains eighteen tombs. Outside the “monastic” complex are at least thirty box graves, but the amount of debris around them might indicate that in former times their quantity was higher, assuming that some graves have already been destroyed (figure 6.5). In addition to the box graves, at least eight tumuli are located outside the wall, with diameters ranging from about 1.5 m to 2.5 m. Although they lack the characteristic “nose” of many post-­Meroitic grave sites,14 their size, their topographical position, and their proximity to the box-­grave cemetery could be seen as evidence for a date in the post-­Meroitic period. West of the structure, just outside the basin, in a solitary rock, we found the remains of at least two cleft burials; both were completely looted but their characteristic structure was still recognizable. Both cleft burials show some similarities with those of the Fourth Cataract.15 Since no characteristic pottery in the immediate 124

JA N A E G ER

Figure 6.5. Medieval Christian box graves at Jabal al-­‘Ayn.

area of the cleft burials was recovered, a chronological approach can be only very preliminary; however, because comparable structures in other areas of the Sudan were dated to the late Kerma or Napatan periods in the first millennium BCE,16 a similar date could probably also be assumed here. As already noted above, it seems strange that such stone buildings, especially an elaborated church, are found without any traces of a regular settlement, especially in such a remote area. The existence of a monastery unknown so far to the scientific community,17 and at such a distance apart from the Makurian heartland around the Nile, raises several questions. First, to which of the Christian Nubian states was it connected? Since the distance to the territory of Makuria is much less than to Alwa, it seems most probable to regard the monastery of Jabal al-­‘Ayn as an outpost of the former state. In addition to this medieval Nubian outpost in the western desert, other features like tumuli and alamat along a natural passage through an extended dune field found during the survey indicate a former trade route parallel to the Nile through the western desert. An old British map from the Sudan Survey Office in the 1920s shows a line labeled old track, leading from the Wādī Qa‘ab southwards to the Wādī Huwār (figure 6.6). Beside the attribute old there is no indication about age or function of the track. On the map, it ends a little north of Wādī Huwār. But further along the same course is a chain of archaeological remains that might indicate an ancient pathway leading southwards to the Jabal al-­‘Ayn, perhaps T H E L A N D O F Ṭ A R Ī ’ A N D S O M E N EW T H O U G H T S O N I T S L O CAT I O N

125

Figure 6.6. Path remains west of the Nubian Nile, satellite photo as indicated in the British map, marked by the gray line in the satellite image (© Google Earth; Map Data © 2013 and Image Landsat © 2013 Google).

dating back to the Middle Ages or, even more probably, to the Napatan period in the first millennium BCE. This ancient pathway (figure 6.7) leads through a narrow natural passage across an extended field of barchan dunes, which makes the area east and west of that passage almost impassable for donkey caravans, modern cars, and even camels. Since the main direction of sand movement in that area is north–­south, it can be assumed that this passage has existed for a long time, and the ancient (probably Napatan) tumuli following it are marking an ancient track leading through. A station along that track is marked by the ledge of a long, linear rock ridge with several old animal tracks leading around it. Those ancient tracks and the topographical situation at the southern mouth of the barchan passage make it most probable that the ridge was used as a landmark along the ancient track, visible from far away and leading caravans like a lighthouse into the entrance of the natural passage. Some remains of ancient camp structures possessing stone rings, presumably to support round huts, are not datable at the moment, but could be associated with the path remains. At the eastern 126

JA N A E G ER

Figure 6.7. Path remains in the desert west of the Nubian Nile, ground level. Note the clear one in the upper right and several in the center and left. The material in the foreground is a natural gneiss outcrop.

edge of the ridge, beside the ancient camp, we found also several traces of recent camps with modern trash (like empty tin cans), proving that the spot is extensively used as a campsite by truckers and camel nomads today. The landmark has provided its rest-­stop function for many centuries. South of that passage and the ridge is a large, empty, and flat plain, where not many traces of ancient traffic have survived. Nevertheless, Kababīsh nomads travel across that plain quite often to the more fertile area of the Jabal al-­‘Ayn, and it could be assumed that in ancient times as well that region was the destination of the route that most probably continued along the Wādī Malik to such Iron Age and medieval centers in Kurdufān as Zankur and Abū Sufyān. The Jabal al-­‘Ayn area contains two possible destinations of the ancient traffic road: the neighborhood of the well of Bir al-­‘Ayn, with an ancient round-­hut settlement, and the aforementioned monastery at the western flank of the Jabal. Around the monastery were detected many ancient donkey-­path remains similar to those on the rock ridge at the north; at least two cleft burials close to the T H E L A N D O F Ṭ A R Ī ’ A N D S O M E N EW T H O U G H T S O N I T S L O CAT I O N

127

Figure 6.8. A cleft burial at the Jabal al-­‘Ayn.

monastery might indicate that the area has been inhabited since the late Bronze Age (figure 6.8). The different archaeological records do not allow a precise dating of the path remains themselves,18 and thus a connection of these lines of communication with the monastery cannot be proven stratigraphically, but the geostatistical coincidence of that site with a dense pattern of ancient pathways indicates that the monastery was in any case somehow connected to the ancient traffic route that could be reconstructed from the scarce archaeological record. Even if the northern destination of this route cannot be confidently determined, it probably connected to other north–­south desert routes that bypassed the Nile River within the western desert and that were active up to the First Cataract region (and thus the southern administrative border of Muslim Egypt) from ancient times. These desert routes are documented especially for the pharaonic and Hellenistic-­Roman period,19 but since they offer the opportunity of trade and traffic off the inhabited (and thus easier controlled) Nile valley, it can be assumed that they kept their importance also during the medieval period, when trade and traffic at the regular border between Egypt and Nubia was regulated rather strictly. Other parts of north–­south routes through the western desert did not avoid any other administrative border (at least once Nobatia was incorporated into the Makurian state), but also had the potential to bypass areas under the direct control of the Nile-­based state authorities and thus avoid taxes or import-­export regulations on trade goods.20 128

JA N A E G ER

Another ancient traffic route to the southwestern border of Makuria most probably led through the Wādī Malik directly from the Nile, wadi-­upstream from its mouth into the river near the modern town of al-­Dabba. This route is not documented by specific archaeological investigation but is widely assumed to be a travelway because of several archaeological features along its bed, like the well station of Kufriyāt al-­Attash21 or the aforementioned settlement at the Bir al-­‘Ayn, which is situated at the southeastern flank of the Jabal al-­‘Ayn and is protected by a small wall blocking a valley. It was described briefly by Gleichen after an unpublished report by Carey.22 If such a route along the wadi bed existed, the monastery of Jabal al-­‘Ayn is not connected directly to it, as it would be to the route to the Wādī Huwār, but it is clear that the Jabal al-­‘Ayn area in general was integrated into an ancient traffic system that connected northern Kurdufān with the Nile valley, and the monastery marked the connection of that comparably fertile region to the wadis situated to the north. The exact reasons why a monastery was erected so far away from the heartland of Makuria in such a remarkably remote place are still unclear, but, it can be hoped, will be subject to further investigation in the future. Establishing some amount of control over the Jabal al-­‘Ayn would have made sense for the Makurian state. Despite the fact that the area between the bend of the southern Nile around al-­Dabba and Kurdufān was not known for its cultural and archaeological richness, the results of a remote-­sensing survey,23 mainly based on Quickbird, and SPOT satellite images showed that the Jabal al-­‘Ayn area contains almost two hundred larger archaeological sites.24 Most of them are cemeteries consisting of post-­Meroitic tumuli and typical medieval box graves. Comparisons of the number of cemeteries and graves with other densely surveyed areas like the Fourth Cataract or the large wadis of the Bayuda Desert25 reveal that at least in those periods the area of Jabal al-­‘Ayn, as well as the whole lower Wādī Malik, must have been quite densely populated. In such a case, control over that region would have made sense also for a mainly Nile valley–­based state like Makuria to make use of the resources produced by the primary pastoral economy of the steppe region. In addition, control over the region would allow such a state to establish a forward outpost to gain political control or at least to gather some intelligence about the activities of the nomadic groups of that region, which had to be seen as a possible threat for the settlements of the Nile valley. A third reason for controlling that area might have been to control desert tracks used by those non-­sedentarists to bypass the Nile valley, and thus the toll and custom authorities of the Makurian state. The indications for the route leading from the Jabal al-­‘Ayn monastery northwards at least to the Wādī Qa‘ab could probably be seen in such a context. T H E L A N D O F Ṭ A R Ī ’ A N D S O M E N EW T H O U G H T S O N I T S L O CAT I O N

129

Of course, the urban centers of the upper Wādī Malik—­Zankur and Abū Sufyān—­date to the Iron Age and form a cultural complex of their own, independent from the contemporary Nile valley cultures. Nevertheless, large graveyards indicate that the period of a quite dense population in the upper Wādī Malik did not end until the late medieval period and ceramics like the aforementioned pilgrim bottles indicate contacts of that area with the Nile valley. Thus, the Jabal al-­‘Ayn with its monastery and a population whose funerary culture, using the well-­known box graves, shows many similarities with that of the Nile valley, may have functioned as a linkage between Dongola and Kurdufān. That crucial function of the medieval Jabal al-­‘Ayn is not only supported by the archaeological evidence but also by some medieval texts. The late tenthcentury Persian geographical handbook Ḥudūd al-­ʿĀlam, which is most probably based on the Kitāb al-­Masālik wa’l-Mamālik written by al-­Iṣṭakhrī, mentions an enigmatic region, read by the editor Minorsky as Ṭarī’ or Ṭahī.26 That land is described as situated “in the desert between the Nūba and the Sudan,” approximately eighty day-­rides south of al-­Qāhira. Here, the anonymous author writes about two remote monasteries in the middle of the desert. The information is rather limited, but the most relevant statements communicated by the anonymous author are the following: • “between the Nūba and the Sudan”: The land of Ṭarī’ with its monasteries must be situated at the fringe of one of the Nubian states, in a border or contact zone with another non-­Nubian Black African culture. (For the geographical knowledge of Nūb[i]a and Sudan in medieval Arabic and Persian texts, see also the map page 103.) • “80 day-­rides south of al-­Qāhira”: The different speeds of a single rider and a camel caravan have to enter into the calculus, but even if the (slower) regular speed of a caravan is assumed, the land of Ṭarī’ should be found at least around 1,600 km south of Cairo. • “remote in the desert”: not too close to the Nile river.27

The text passage of the Ḥudūd al-­ʿĀlam was interpreted by Derek Welsby and Julie Anderson. Welsby identifies the monastic communities of the land of Ṭarī’ with some medieval remains at Salīma oasis,28 while Anderson associates this account with the well-­known monastery of al-­Ghazāl in the Wādī Abū Dūm.29 Neither interpretation is completely likely. Salīma seems, first, to be situated too much to the north to need eighty days of riding from Cairo, even if one took into consideration a slow caravan and used the longer segment along the Nile rather than the shorter oasis bypass. Second, Salīma is not 130

JA N A E G ER

really situated in a border or contact zone between a Nubian state and other non-­Nubian Black African cultures. The identification of one of the Ṭarī’ monasteries with al-­Ghazāl in the Wādī Abū Dūm, as suggested by Anderson, would make more sense concerning the distance from Cairo, but that monastery is within only a few hours of human walking distance from the Nile, which can hardly be seen as “a remote place in the desert.” All of the requirements implied by the text are best fulfilled by the Christian complex at Jabal al-­‘Ayn. Almost 200 km west of the Nile, it is quite a remote place for people used to the riverine Nile culture; its distance from Cairo matches quite closely an eighty-day ride. Most important of all, the Jabal al-­ ‘Ayn obviously seems to form some sort of westernmost border of Makuria’s influence, and a frontier and contact zone toward the non-­Nubian cultures of early Kurdufān. There is (theoretically) only one weak point about the identification of the Jabal al-­‘Ayn area with the Ḥudūd ’s Land of Ṭarī’, as there is with the other aforementioned theories as well: the text explicitly mentions “two” monasteries, not just one. Additionally, the number of monks described in the text is a problem, because any of the suggestions for the location of Ṭarī’ (including the one presented in this chapter) are not able to accommodate twelve thousand people. But this problem seems to be of little relevance concerning the identification of the Jabal al-­‘Ayn area with the land of Ṭarī’, since it likely involves an inconsistency in the text itself. The description of Ṭarī’ as an area “remote in the desert” and the huge number of inhabitants seem not to fit with each other, since twelve thousand monks (according to the text) are scarcely conceivable logistically in such a remote area, and an inflation of the number of monks living in a monastery is not uncommon in Oriental Christianity. The other possible identifications of the land of Ṭarī’ mentioned above would also be unable to support such a large number of monks, so this information does not contribute much to the possible identifications of the location of Ṭarī’. The indication of two monasteries suggests that there might still be other, unexplored monastic installations within the territory of the Libyan Desert and Kurdufān. Some scarce evidence might indicate unidentified sites the area. According to Derek Welsby,30 there are “Christian remains” marked on an old map issued by the Sudan Survey Office of the Anglo-­Egyptian Condominium. These remains are situated between the Wādī Malik and the Wādī Muqattam, roughly 80 km west of the modern town of Tamtam.31 Even if the exact character of this record remains unclear, at least it is an indication that the areas south and southwest of the Dabba Nile bend probably T H E L A N D O F Ṭ A R Ī ’ A N D S O M E N EW T H O U G H T S O N I T S L O CAT I O N

131

contain more evidences of a wide-­ranging perimeter of control established by the Kingdom of Makuria out in the desert. Even if the identification of the area around the Jabal al-­‘Ayn with the Land of Ṭarī’ is accepted as correct, little more information about its integration into north–­south relations can be extracted from the source. But at least its mere mention in the Ḥudūd al-­ʿĀlam would, in that case, indicate some significance for the area and its role as a contact zone not only for Makuria itself, but also for the Islamic world. But since the Ḥudūd al-­ʿĀlam gives no details about economic relations, any connection of that account with the aforementioned route to or from the north must remain quite speculative. After having investigated the area between the Wādī Huwār and the Wādī Malik by using satellite-­based remote sensing, and by studying old maps and the notes of early travelers in the region, only scarce information about its connection to the Nile valley cultures of different periods can be found. But, the existence of a monastery as a Makurian outpost that far from the Nile is noteworthy. The complex and the cemeteries around it, as well as a large number of archaeological sites (mostly cemeteries) discovered from remote-­sensing data in the close and wider vicinity show that, during the medieval period at least, the western section of the Jabal al-­‘Ayn was relatively densely inhabited.32 Its population made extensive use of box graves, which proves their close cultural connection to the Christian Nile valley. Immediately to the south, the area of the early Kurdufān cultures begins, where influences from Makuria or the other Christian cultures of the Nile valley can as yet be seen only as indirect and scarce. The Jabal al-­ ‘Ayn area with its close cultural relations to the Nile valley–­based Christian culture of the Makurian state on the one hand, and its convenient connection to ancient Kurdufān via the Wādī Malik, is most likely to have functioned as a direct frontier and contact zone. Even if until now cultural ties known to us seem closer to the Nile valley, pure topographical reasons make it highly presumable that in the other direction, to Kurdufān, links were comparably strong. More problematic is the connection to the north. Some traces of an old desert route leading from the Jabal al-­‘Ayn over the Wādī Huwār to the Wādī Qa‘ab were found. But until a more detailed ground survey is conducted, only some rather speculative assumptions about a secondary role for this Makurian outpost in a north–­south desert route system bypassing the Nile valley can be made. But, even if such a large-­scale north–­south trade-route system away from the Nile routes cannot be proven from the material available until now, this does not really touch the identification of the Jabal al-­‘Ayn area with the 132

JA N A E G ER

land of Ṭarī’, since combined long-­distance trade routes leading partly along the Nile river banks, and partly through the wadis of the desert (as they are indicated for later periods)33 could also have been in use during the medieval era. Thus, the integration of a route between Makuria and sub-­Saharan Africa via the Jabal al-­‘Ayn would fit into several different scenarios of long-­distance trade-route organization, which cannot be reconstructed in detail with current knowledge but has to be assumed according to the distribution of imported goods within the Makurian state.34 The connection of that archaeological evidence with historic sources is also still problematic, but there are at least some good arguments to identify the wider area around the Jabal al-­‘Ayn with the Land of Ṭarī’. This land is, as indicated above, also described in the Ḥudūd al-‘Ālam as a frontier and contact zone between Nubia and the more southern African cultures. This (brief ) mention of this border situation and the outpost character of the archaeological record described in this chapter can be linked together, supporting the hypothesized identification. Of course, this periphery of the Christian kingdom of Makuria is not a real “Islamic” border, but its mention in a Persian geographical handbook indicates its significance also for the Islamic world. Since the state of Makuria was in many respects influenced by Faṭimid Egypt, this border zone could be interpreted as part of the very periphery of the Egyptian-­influenced region. But the exact character and organization of this border zone, the role of religious installations for its administration, and the amount of Egyptian influence on this remote periphery still remain quite a mystery, and possible future investigations bear a lot of potential. A closer investigation of the area, not only from satellite images but on the ground as well, will reveal much more information about the borders of the Makurian state, its range of political and economic control, and its interactions with the African Sahel cultures to the south. NOTES

I wish to express my gratitude to the head of the University of Cologne expedition to Gala Abu Ahmed, Friederike Jesse, for giving me the opportunity to conduct a survey of the Jabal al-­Ain area. Additionally, I want to thank the survey team Wolfgang Mackowiack, Natascha Mathyschock, and Tim Karberg. Beside the survey, I thank Stephan Seidlmayer and Angelika Lohwasser for the opportunity to publish the results of my MA thesis, and Artur Obluski, Wlodzimierz Godlewski, and Derek Welsby for their comments. T H E L A N D O F Ṭ A R Ī ’ A N D S O M E N EW T H O U G H T S O N I T S L O CAT I O N

133

1. Welsby 2002, 86. 2. For example, on forwarding trade goods from the south and southwest toward the Muslim world, see Welsby 2002, 78. 3. Minorsky 1937, 165–­167. 4. Adams, 1977. 5. As stated for the medieval period (Welsby 2002, 68–­73) as well as for late antiquity (Török 2009, 523–­525). 6. Monneret de Villard 1935, 278–­279. 7. Gratien 2010, 237–­246. 8. Arkell 1959, 115–­116. 9. The complex was given the temporary survey number FJE2010–­1, and first published in Eger 2011, 115–­120. 10. Adams 1965, 103–­105. 11. Billig 2007, 94. 12. Anderson 1999, 76. 13. Adams 1977, 479; Jeuté 1994, 93–­94. 14. Thus, for example, described by Welsby 2005, 4; classified there as type IV or type IVa. 15. Welsby 2005, 5–­6. 16. For example: Wolf and Nowotnick 2005, 25; Näser 2005, 53–­54. 17. This site is listed neither in older compilations (Monneret de Villard 1935; Hinkel 1979, map sheet 26) nor in recent works about either the border regions of the empires of the Nubian Middle Ages (Welsby 2002, 85–­88; Vantini 1999) or the monastic systems in Nubia (Anderson 1999, particularly footnote 4; Jeuté 1994). 18. This is not uncommon: Ancient path relics are in most cases very hard to date. 19. Storemyr et al. 2013, 399. 20. See Ruffini, chapter 5, this volume. 21. Welsby 2002, 87. 22. Gleichen 1905, 211; Hinkel 1979, 154–­155. 23. For more information, see Eger 2013b, 127–­128. 24. Eger 2013a, 164–­292. 25. For example, see Lohwasser et al. 2014. 26. Minorsky 1937, 165–­167. 27. Quoted from Vantini 1975, 174. 28. Welsby 2002, 86. 29. Anderson 1999, 71n1. 30. Personal communication. 31. So far, the author has not been able to visit the place personally. 32. Eger 2018, 873–­878. 134

JA N A E G ER

33. Förster et al. 2013, 193–­194, 203. 34. See Ruffini, chapter 5, this volume. REFERENCES

Adams, B. 1965. “Architectural Evolution of the Nubian Church 500–­1400 AD.” Journal of the American Research Center in Egypt 4: 87–­139. Adams, B. 1977. Nubia: Corridor to Africa. Princeton, NJ: Princeton University Press. Anderson, J. 1999. “Monastic Lifestyles of the Nubian desert: Seeking the Mysterious Monks of Makuria.” Sudan and Nubia 3: 71–­83. Arkell, A. J. 1951. “An Old Nubian Inscription from Kordofan.” American Journal of Archaeology 55: 353–­354. Arkell, A. J. 1959. “A Christian Church and Monastery at Ain Farah, Darfur.” Kush 7: 115–­119. Billig, D. 2007. “H.U.N.E. 2007: Die Kirche SR022.A.” Der Antike Sudan. Mitteilungen der Sudanarchäologischen Gesellschaft zu Berlin e.V. 18: 89–­98. Eger, J. 2011. “Ein mittelalterliches Kloster am Jabal al-­Ain?” Der Antike Sudan. Mitteilungen der Sudanarchäologischen Gesellschaft zu Berlin e.V. 22: 115–­120. Eger, J. 2013a. “Spuren im Sand. Ein antikes Wegesystem im Sudan, rekonstruiert anhand von Fernerkundungsdaten.” Unpublished master’s thesis, Free University Berlin. Eger, J. 2013b. “Ancient Traffic Routes in the Sudanese Western Desert: An Archaeological Remote Sensing Project.” In Archaeological Prospection: Proceedings of the 10th International Conference, Vienna, May 29th–­June 2nd 2013, edited by W. Neubauer, I. Trinks, R. B. Salisbury, and C. Einwögerer, 127–­128. Vienna: Ludwig-­Bolzmann-­Institute. Eger, J. 2018. “Archaeological Satellite Imagery-­Based Remote Sensing in the Bayuda and the Western Desert (Sudan).” In Nubian Archaeology in the XXIst Century: Proceedings of the 13th International Conference for Nubian Studies, Neuchatel, 1st–­6th September, 2014, edited by M. Honegger, 873–­878. Leuven, Belgium: Peeters. Förster, F., H. Riemer, M. Mahir, and F. Darius. 2013. “Donkeys to El-­Fasher or How the Present Informs the Past.” In Desert Road Archaeology in Ancient Egypt and Beyond, edited by F. Förster and H. Riemer, 193–­218. Cologne: Heinrich-­Barth-­Institute. Gleichen, A. E. W. 1905. The Anglo-­Egyptian Sudan, volume I, Geographical, Descriptive, and Historical. London: His Majesty’s Stationeary Office. Gratien, B. 2010. “Prospections dans le Kordofan occidental, autour de Zankor et d’Abou Sofyan (2002–­2005).” In Between the Cataracts. Proceedings of the 11th

T H E L A N D O F Ṭ A R Ī ’ A N D S O M E N EW T H O U G H T S O N I T S L O CAT I O N

135

Conference of Nubian Studies, Warsaw, edited by W. Godlewski and A. Łajtar, 237–­246. Warsaw: Warsaw University Press. Hinkel, F. W. 1979. The Archaeological Map of the Sudan, volume II, The Area of the South Libyan Desert. Berlin: Akademie-­Verlag. Jeuté, P. 1994. “Monasteries in Nubia. An Open Issue.” Nubica III/I: 59–­97. Lohwasser, A., J. Eger, and T. Karberg. 2014. “Das Projekt Wadi Abu Dom Itinerary (W.A.D.I.), Kampagne 2014.” Der Antike Sudan. Mitteilungen der Sudanarchäologischen Gesellschaft zu Berlin e.V. 25: 177–­188. Minorsky, V. ed. 1937. Ḥudūd al-­ʿĀlam. The Regions of the World: A Persian Geography, 372 A.H.–­982 A.D., translated and explained by V. Minorsky. London: Oxford University Press. Monneret de Villard, U. 1935. La Nubie Medioevale, volume I, Inventario dei monumenti. Cairo: Imprimerie de l’Institut français d’archéologie orientale. Näser, C. 2005. “Die Humboldt University Nubian Expedition 2005: Arbeiten im Bereich der Inselkonzession.” Der Antike Sudan. Mitteilungen der Sudanarchäologischen Gesellschaft zu Berlin e.V. 16: 49–­67. Storemyr, P., E. Bloxam, T. Heldal, and A. Kelany. 2013. “Ancient Desert and Quarry Roads on the West Bank of the Nile in the First Cataract Region.” In Desert Road Archaeology in Ancient Egypt and Beyond, edited by F. Förster and H. Riemer, 399–­423. Cologne: Heinrich-­Barth-­Institute. Török, L. 2009. Between Two Worlds. The Frontier Region between Ancient Nubia and Egypt 3700 BC—­AD 500. Leiden/Boston, MA: Brill. Vantini, G. 1975. Oriental Sources concerning Nubia. Heidelberg/Warsaw: Society for Nubian Studies. Vantini, G. 1999. “The Remotest Places Reached by Nubian Christianity in Sudan.” Nubica IV/V: 347–­350. Welsby, D. 2002. The Medieval Kingdoms of Nubia. London: British Museum Press. Welsby, D. 2005. “The Merowe Dam Archaeological Salvage Project: Survey in the Vicinity of ed-­Doma (AKSE) 2004–­2005.” Sudan and Nubia 9: 2–­8. Wolf, P., and U. Nowotnick. 2005. “The Second Season of the SARS Anglo-­German Expedition to the Fourth Cataract.” Sudan and Nubia 9: 23–­31.

136

JA N A E G ER

Part III

The Eastern Frontiers The Caucasus and Central Asia

Armenian, Aran and Azerbaijan; Iṣṭakhrī, Kitab al-Mamālik wa al-Masālik. MS Ouseley 373, f.80a, courtesy of Bodleian Library, University of Oxford.

7 We live in physical and metaphorical compartments that are defined by boundaries. Some boundaries—­ranging from state borders to social and virtual frontiers—­may be fading, or are becoming permeable and easy to cross, but at the same time many new boundaries are being created and strengthened.1 Boundaries endure despite the flow of people across them.2 Some of them may appear in physical form as “borders” with material expressions such as walls and fences. It is believed that borders developed in modern times because they represent the perfect overlap between territorial integrity, ethnic/national uniformity, and long-­lasting borderlines that presumably lead to permanence.3 Borders with demarcating lines that bring people, territories, and sovereignty together are said to have developed only after the Peace of Westphalia in 1648 in Europe.4 Instead, premodern states were bounded by frontier zones rather than borders, and when they built walls it was just for protecting state territories against outsiders and not for bordering.5 Territorial differentiations were based on the logic of cores/centers and peripheries, without clear boundaries or edges.6 It is further believed that in the premodern world, political boundaries often remained porous and unclear, and led to formation of “frontier” zones or borderlands. Frontiers between two states could be areas of interaction ranging from confrontations to cooperation.7 The focus of this chapter is one of the historical examples, the Darband complex and its fortification

Overlapping Social and Political Boundaries Borders of the Sasanian Empire and the Muslim Caliphate in the Caucasus Karim Alizadeh

DOI: 10.5876/9781607328773.c007

139

wall in the Caucasus, which could also have functioned as a bordering mechanism. When these structures are historically contextualized, it becomes clear that there were many factors that contributed to their formation and maintenance in the Sasanian and the early Islamic period. Analyzing two of these factors—­functionality of the fortification wall and population settlement in the area—­points to some of the characteristics of borders in the modern sense. In this essay I first argue that the Sasanian Empire created a series of physical barriers along the Caucasus Mountains that together could have functioned as border installations. I also show that these installations were kept and reinforced as such during the Muslim Caliphate. I suggest that such bordering mechanisms should be considered as precedents for later attempts to develop fixed borders and mutually exclusive sovereignties that developed in the early modern period. Modern borders developed after a long historical process of attempts by states and empires to organize their territories sociospatially in order to decrease the fuzziness and permeability of their “frontiers” and turn them into more fixed, physically hard, and restrictive barriers. Second, I maintain that state-­directed population transformation, and particularly mass deportation of people to frontiers or borderlands, should be considered boundary-­making practice and part of a bordering process. State-­ directed demographic transformations in frontier areas point to the desire of the premodern states to overlap or approximate their political boundaries with ethnic and sociocultural boundaries. Previously, I have discussed the mass deportation and colonization of the region by the Sasanians.8 Here, I contend that their successors, the Muslim Caliphate, also followed and practiced a similar policy. FRONTIERS AND BORDERS

What is a border, and what makes a border “the border”? There is a disciplinary division of tasks while studying frontiers and borders. Among scholars there is a consensus that frontiers are historical phenomena and modern political boundaries between nation-­states are all borders. Likewise, it is widely agreed upon that borders are modern phenomena and there were no borders in the premodern world. Thus, traditionally, frontiers were addressed by disciplines such as history, archaeology, and anthropology, and borders by the social sciences such as political geography, sociology, and economics. Before I discuss how these two concepts are perceived in different disciplines, it is important to clarify terminologies and concepts. There are some terminological inconsistencies and obscurities that complicate border and frontier studies. 140

K A R I M A LI Z A D EH

For instance, relations and interconnectedness of frontier, borderland, and border are still debated. In the social sciences, the concepts of frontier and borderland are sometimes used interchangeably and generically refer to the region behind a border. However, a border is the physical location or space where the borderline is drawn, such as a fence or a wall in linear form. While borders and borderlines/ frontiers share the same physical space, they are not the same. A border is a line that separates, encloses, and excludes at a number of spatial and social scales; it is a line demarcated and implemented by the state and the area within which the polity’s sovereignty is exercised. Whereas the concept of frontier9 is usually associated with permeability and fuzziness of boundaries, borders, in contrast, form a physically linear, hard, static, restrictive, and supposedly impenetrable barrier. A frontier is an area or region in close proximity to the border within which human life and development patterns are influenced by their proximity to the border.10 In other words and without making any equations, the clearer, more linear, and more physically restrictive frontiers become, the more they functionally become like borders. Otherwise, none of them—­whether frontier or border—­is absolutely impermeable. We can argue that based on their physical and functional aspects, modern borders and frontiers have some characteristics in common, but they also differ. They both signify political boundaries of states and at the same time may represent multiple lines of differences and separation. They are products of boundary-­making practices in a broad sociopolitical process. However, borders are physical, clear, and linear barriers that define edges of state territory where the control, management, and power of the state is exercised, while frontiers are transitional zones or “regions of continuity”11 and not necessarily physical and clear. In frontiers, the state control is usually unsteady. Although both are permeable and restrictive, they differ in the degree of permeability and restriction that they create. Compared to frontiers, borders are more restrictive and less permeable since they are represented by a clear physical barrier. In addition, restriction and permeability in borders are managed via passes and gates. Fredrik Barth’s studies12 demonstrated how boundaries are both relational and socially constructed. Borders are part of ongoing boundary-­making practice in human societies. As such, they are dynamic rather than passive, and are to be considered as a process rather than a static fixture.13 Borders are complex social phenomena, and understanding them requires interdisciplinary study and breaking disciplinary boundaries. You have to break the boundaries to study the boundaries! The process through which borders are demarcated and Overlapping S ocial and P olitical B oundaries

141

managed are more important than the border by itself.14 Borders are places where territoriality, boundary lines, and identities meet and they are expressions of multiple identities.15 They produce and are products of “the separation line not only between states and geographical spaces,” but also between “us” and “them,” the “here” and the “there,” and the “insiders” and the “outsiders.”16 Borders are not only determined, but also determine the identity of people on both sides and their ways of life.17 In other words, borders usually represent multiple lines of boundaries and lines of differences or separation.18 Traditionally, border studies were contained within the political geography discourse and discussed mostly from a security perspective. In this perspective, which primarily addresses the physical and functional aspects, borders are expected to create barriers to movement of people and are considered fixed lines, stable, static, and restrictive.19 Due to the close relationship between the notion of security and the use of force by the state apparatus to ensure its borders, frontiers were mostly understood as areas of military threat, thus prompting control over border ­crossings as a necessity.20 In the premodern world, generally, states and empires relied on natural barriers to protect and define their territories. Natural features served as physical barriers more than anything else in human history. For instance, the Danube River formed both a natural and political border between the Roman Empire and the “barbarians,”21 and the Euphrates and Tigris Rivers served as a border between the Roman and Persian Empires during the Parthian and Sasanian dynasties.22 For the state, securing territories and protecting its resources were not only achieved by natural barriers or building fences and walls, but also by controlling flows of people, goods, and ideas in the frontiers. This necessity was based on the notion that, first, the density of interactions and transactions was an integration process, and, second, it leads to identity transformation in frontiers.23 As such, from the states’ perspectives, social and demographic dynamics were very important elements of borderland policies, and demographic manipulation was and is practiced by some states. For instance, studies show that demographic manipulation was an integral part of the frontier policies of the Assyrians in northern Mesopotamia,24 the Sasanians in their northern borderlands and frontiers with the Roman and the Byzantine Empires,25 the Safavid Empire in the Caucasus and frontiers with the Ottomans,26 and, in the modern world, the frontier policies of the Russian Empire in the Far East, Central Asia, and the Caucasus.27 However, historical studies show that frontiers were complex, dynamic, and interactive zones that were not only defined by states and empires of the time but mostly by their inhabitants.28 142

K A R I M A LI Z A D EH

In the modern era, some of the old mechanisms like population transformation or peopling of frontiers have been abandoned29 as no longer necessary, due to more capable technologies of control, modes of political organizations, international treaties, and rules and norms developed to protect territorial practices. Ironically, in modern times, the security of frontiers sometimes requires mechanisms opposite to population deportation into the borderlands, such as, for example, removing (rather than importing) inhabitants and dampening or suppressing economic activities. This is based on a notion that a high density of people in the frontiers may lead to a high number of interactions that facilitate social integration and identity transformation, which might be perceived as a threat by the central state. Thus, control over borders is easier if fewer inhabitants live in frontiers. Consequently, frontiers sometimes become economically backward and sparsely populated in many countries today.30 Several indicators suggest that states such as the Sasanian Empire did not favor porosity and permeability of political boundaries and did not want to leave them unclear. Thus, sometimes they applied various mechanisms to meet their needs in securing territories. When possible and necessary, the states built walls and fortifications to control, minimize, or even cease all contact with a threatening neighbor. Walls were built by some states in the past such as Hadrian’s Wall by the Roman Empire and the Great Wall of Gurgān built by the Sasanian Empire.31 States also engaged in building gates and forts in natural passes to control the movement of people, as exemplified by the boundary-­makings of the Sasanian Empire in the Caucasus. The Sasanian Empire also practiced demographic manipulation in their frontiers to further reinforce borders. After the fall of the Sasanian Empire and the rise of the caliphate, the Muslim rulers also maintained and reinforced these border installations and followed similar policies (see below). BORDERS OF THE SASANIAN EMPIRE AND THE ISLAMIC CALIPHATE IN THE CAUCASUS

According to Abū al-­Rayhān al-­Bīrūnī, an early eleventh-­century polymath, the Sasanian kings were “in need of a definite settlement of the demarcation lines between the kingdoms of others” and their own. Partitions of the Sasanian Empire “were called kishvar, which is derived from the Persian word (keshettah) for line, to signify that each partition is distinguishable from the others, like what is bounded by lines. [The Sasanian king] also needed a prevailing influence to get what he wanted, and to instill in all those surrounding him a state of fear of his might and a desire for his friendship.”32 What we Overlapping S ocial and P olitical B oundaries

143

Figure 7.1. Map of the Caucasus (modified from https://​en​.wikipedia​.org/​wiki/​Caucasus​ _Mountains).

can understand through Bīrūnī’s statement is that the Sasanian kings desired a clear sense of a border that was located in a particular space, controlled, and monitored. As Zayde Antrim33 accurately underscores, Bīrūnī maintained that this partitioning (or kishvar) represented a political decision and had nothing to do with natural or climatic setting. The Sasanian Empire engaged in the construction of walls, moats, and fortifications around their political boundaries. They built a massive wall with fortifications on the Gurgān Plain that stretched about 200 km eastward from the Caspian Sea;34 Dag-­Bary, or the Darband Wall, in the Caucasus;35 the Tāzigān Wall; and a complex system of walls, moats, and forts against the Arabs in the west and southwest of the imperial territory.36 These walls and moats served as borderlines enclosing the territory of the Sasanian Empire, Īrānshahr.37 Due to archaeological surveys and excavations during the past few decades, our knowledge of the northern frontiers of the Sasanian Empire in the Caucasus and to the east of the Caspian Sea has considerably increased.38 The 144

K A R I M A LI Z A D EH

Figure 7.2. Sasanian walls and fortifications in the Caucasus (after Aliev et al. 2006, 146).

Caucasus (figure 7.1) was a strategic region that played a crucial role in relations between the Sasanian and Roman Empires, the Huns, and the Turks.39 The Greater Caucasus range with its high summits, some of which exceed 5,000 m, particularly the Glavny Ridge, acts as a natural barrier dividing the region into two parts: the North and the South Caucasus.40 In the Caucasus, in addition to large fortifications such as Besh-­Barmak, Toprak-­Qala, Darband,41 and Çirax-­kala,42 gates and forts on the strategic passes of the Caucasus mountain range signify further boundary-­making practices by the Sasanian Empire. The functionality of these physical barriers was not confined to defense. Among them, Darband served as a border fortress in the Sasanian period.43 The fortification of Darband44 marked the northern territorial boundary of the Sasanian Empire in the Caucasus (figure 7.2). The defensive system in Darband consisted of two major elements: (1) Darband city and its fortification; and (2) the Darband defensive wall, or Dag-­Bary.45 The city consisted of a citadel46 and lower town,47 comprising about 30 ha in area (figure 7.3). The Overlapping S ocial and P olitical B oundaries

145

Figure 7.3. Top: plan of Darband in Dagestan, Russia (after Khan-­Magomedov 1979, 23). Bottom: Darband in 1796 (from E. Eykhvalda, published in 1834).

city grew in size during the early Islamic period, and the current town of Darband is built on top of Sasanian and early Islamic towns.48 The Sasanian city of Darband was formed between two parallel stone walls with projected towers and gates. The twin walls of the lower town from the sea shores to the Caucasus mountain range filling the pass between these two natural barriers. These parallel walls were about 4 m thick, 18–­20 m high, and 350–­450 m apart. The northern wall was about 3,650 m long and the southern wall about 3,500 m (figure 7.4). They both stretched from the Caspian shore westward and joined the mountain wall, filling the pass between these two natural barriers. There were seven gates, seventy-three massive projecting towers along the northern wall, and twenty-seven towers along the southern wall.49 The second element of the defensive system was the Darband Wall, which formed an extra barrier and stretched west about 45 km into the virtually impenetrable mountain range of the Caucasus with multiple forts and fortresses (figure 7.5).50 Based on archaeological evidence as well as the early Islamic texts51 and the thirty-two Middle Persian inscriptions on stones of the fortification wall, archaeologists have attributed the foundation of the town to the Sasanian period and particularly to the reign of Khusraw I Anushirwan (r. 531–­579). 146

K A R I M A LI Z A D EH

Figure 7.4. Darband northern wall in Dagestan, Russia (photo by Dimitri Eermakov, ca. 1890).

It should be noted that according to the late Sasanian text, Šahrestānīhā-­ī Ērānšahr, Darband is referred as Wērōy-­pahr, the “Gruzinian Guard”52 or the “K’art’velian Guard.”53 Since the 1970s, the Darband complex has been excavated by archaeologists. Further excavations carried out during the late 1980s and early 1990s, especially stratigraphic excavations in the citadel and along the walls, provided more information on the foundation of Darband. These excavations suggest that the stone fortification was erected on top of a mud-­brick fortification, which represents the first phase of construction at the site and, like the upper fortification, stretched from the sea up to the mountain. Archaeological evidence associated with the first phase facilitated the dating of the earlier fortification, and numismatic evidence from the reign of Yazdgird II (438–­457) points to the foundation of the fortification during the second quarter or middle of the fifth century.54 The Romans and Sasanians agreed to cooperate on the protection of the Caucasus against the Huns; however, from the fourth century, the Romans withdrew from the Caucasus and the Sasanians became the major power in the region.55 Several passes along the Caucasus Mountains were guarded by Overlapping S ocial and P olitical B oundaries

147

Figure 7.5. Darband Wall and fortifications (after Khan-­Magomedov 1979, 208).

gates built by the Sasanian Empire and were controlled by it from the third century until its fall in the seventh century.56 The pass at Dariali was controlled by the Alan Gates,57 and the Caspian pass,58 controlled by the Darband gated barriers and fortifications, were traffic routes from the northern steppes of the Caucasus to the Near East. Numerous sources refer to these gates.59 In addition to Darband complex, which includes the city, wall, and its fortification, in 2002 an Azerbaijani-­Daghestani-­American team conducted a survey that identified two more fortification walls further south, in the northeast of the Republic of Azerbaijan. These two walls, the Ghilghilchay and Besh-­Barmaq (figures 7.1 and 7.2), with more than three hundred towers and several fortified settlements along their length, also represent the significant investment of the Sasanians in the protection of the eastern passages of the Caucasus (figure 7.6). Scholars have dated these defensive walls and associated fortified settlements to the second half of the fifth century.60 Moreover, the appearance of many Sasanian fortified settlements—­such as ancient Ganja in the Ganja-­Gazakh Plain and Ören Qala61 in the northeastern and eastern piedmont of the Lesser Caucasus range,62 Ultan Qalasi, and many other fortified settlements in the Mughan Steppe—­is reminiscent of agricultural colonization63 practiced earlier by the Assyrian Empire in its frontiers in northern Mesopotamia.64 It is possible that the Sasanian Empire even expanded its territory to the north of the Darband Wall and, as some historical texts suggest,65 they possibly founded cities such as Shabiran, Balanjar, and Samandar. However, it seems that after the Khazar Turks settled and became the dominant power in the north Caucasus, the Sasanian Empire had already lost control over the areas to the north of the Darband Wall,66 mostly from the sixth century 148

K A R I M A LI Z A D EH

Figure 7.6. Portions of the Ghilghilchay long wall (after Aliev et al. 2006, 162).

onward. Thus, it seems that Darband with its wall and fortifications became a borderline between these two states. Elsewhere I have discussed that borderland policies of the Sasanian Empire in its northern borderlands were multipurpose and intertwined with their economic ambitions.67 In order to achieve their goals, such as building defensive walls and fortifications, digging irrigation canals, and founding cities, the Sasanian Empire needed a huge laboring population. The empire often transferred people from the borderland areas to heartlands and vice versa. In Overlapping S ocial and P olitical B oundaries

149

his inscription at Naqsh-­e Rustam, Shapur I (r. 239–­270) stated that he had brought Roman captives and “non-­Iranians” to Fars, Parthia, Khuzistan, and Mesopotamia and resettled them in areas under direct government control,68 and transferred “people from the East” to settle in Hisar-­Shapur and Shad-­ Shapur.69 Shapur II (r. 309–­379) brought twelve thousand Iranians from Fars to Nisibis, a frontier and strategically important city, after a peace treaty with the Roman emperor Jovian in 363, and Roman captives were systematically resettled during the wars of the third through sixth centuries.70 In case of the Caucasus, for instance, after the Lazic War between the Sasanian and Roman Empires, in which the Huns served the Sasanians as auxiliaries, Khusraw I (r. 531–­579) constructed several fortifications around Tiflis-­ Gardman, Sughdabil, and al-­Lal, and settled huge numbers of Iranians in this region. Moreover, he secured the frontiers across the Caspian coastal plain by settling some transplanted tribes, as well as using local ones.71 According to Mas’udi, a tenth-­century historian and geographer, Khusraw I settled people in the city of Darband and at the gates to guard the border with the Khazars.72 The collapse of the Sasanian Empire and the conquest of Iran by Arab Muslims in the mid-­seventh century were turning points in the history of the Near East. During the last decades of the Sasanian Empire, the Khazars, a confederacy of nomadic Turkish tribes in the northern areas of the Caucasus, were an annoyance and frequently invaded its northwestern frontiers. The Turkish Khazars became allies with the Byzantines and founded a powerful state to the north of the Caucasus range.73 In the Christian and Muslim traditions, the Khazars were sometimes associated with Gog and Magog, who lived in the northern regions and were enclosed behind a barrier, gate, or wall.74 After the Byzantine emperor Heraclius defeated the Sasanians in the Caucasus in 622, the Khazars destroyed the Sasanian army and the fortifications of Darband. Then the Khazars took control of the region and the Sasanian army was unable to reconquer the Caucasus.75 With the final defeat of the Sasanian army in the key battle of Nahāvand in western Iran in 642, Ḥodhayfa b. al-­Yamān with his Arab troops moved into Azerbaijan and defeated the marzbān (governor) of Ardabil.76 Under Caliph ‘Uthmān, Arabs campaigned more in Azerbaijan to impose or raise tributes on cities and towns.77 In 655, the Arab army headed further north and went beyond the fortification and city of Darband78 to subdue the Khazars. According to al-­Tabari,79 when the Arab army appeared in Darband this frontier area was under the control of a Persian ruler, Shahrbaraz. However, according to Balādhurī, when the first Arab army appeared in the Caucasus, Darband was ruled by the Turkish Khazars.80 By the end of seventh century Muslim Arabs had taken control 150

K A R I M A LI Z A D EH

of all the Sasanian territories and infrastructures. In 701, under Muhammad ibn Marwān, the governor of Azerbaijan, they formally annexed the region to the caliphate. Cities or major centers were now controlled directly by Muslim forces but the countryside was controlled indirectly through local elites.81 However, during the end of the seventh and beginning of the eighth centuries, Arab Muslim control over Azerbaijan and the eastern part of the Caucasus was challenged, due to interventions and attacks by Turkish Khazars. Although the Caucasus range was a natural border that worked almost as an impenetrable wall, Muslim leaders could not fully protect the region against the Khazars. The Muslims regarded them as “a people living on their borders.”82 As time passed, Caucasian frontiers remained important during the ‘Abbāsid period.83 Security of the Caucasus was one of the main concerns of Muslim leaders, especially because of the proximity of the Khazars to the heart of the caliphate. Like the Sasanians who struggled to protect their frontiers,84 Muslims also struggled to control the passes along the Caucasus range. The Khazars appeared to be a difficult rival for the Muslims in the region. Although the Arab army had successes against them, they were also defeated several times by the Khazars.85 In 726, the Khazars launched a campaign deep into Azerbaijan, where they besieged Warthān—­possibly modern Ultan Qalasi on the south bank of Araxes River86—­and sacked some other cities and towns such as Baylaqān,87 Bājarwān,88 and Ganja.89 The year 730 marked a catastrophic defeat for the caliphate when the Khazars took the upper hand in the war and penetrated into the heart of Azerbaijan to besiege and sack Ardabil, the capital city of Azerbaijan.90 At this point, despite some victories, attempts by the Muslim leaders to take full control of the region frequently failed. The Khazars repeatedly crossed the Darband and Alan Gates and invaded the regions south of the Caucasus Mountains. Muslim forces attempted to conquer Balanjar91 in Khazar territory several times, first in 642 under the rule of the Caliph ͑ Umar and again under the rule of Caliph ͑ Uthmān.92 But none of their campaigns could expand the caliphate territory beyond the Caucasus Mountain range and Darband fortifications.93 It is very important to note that during the midst of these back-­and-­forth military attacks, both the Arabs and the Khazars settled their own people at the passes, particularly in Darband. Historical texts indicated that when the Arab army under Maslama ibn ‘Abd al-­Malik94 defeated and expelled the Khazar army in Azerbaijan and Albania,95 they found thousands of Khazars settled in the city of Darband. After driving out the Khazars from Darband, Maslama settled twenty-four thousand troops there and resettled many Syrian Arabs and Kufans as well, including craftsmen and laborers.96 Overlapping S ocial and P olitical B oundaries

151

Archaeological excavations indicated that, compared with the Sasanian settlement, although its citadel remained relatively similar, the lower town in Darband grew in size and became twice as large by the thirteenth century.97 Moreover, recovered manufacturing evidence such as ceramic slags, metallurgical remains, and glass production points to a concentration of production activities in southeastern, northwestern, and northeastern parts of the lower town during the sixth to eighth centuries. Overall, the archaeological evidence suggests that Darband became also one of the significant economic centers during the early Islamic period98 and kept its function as “frontier post” (thaghr) against the Turkish Khazars.99 For instance, according to the author of the Hudud al-­Ālam, “between the town and the sea a huge chain has been stretched so that no ship can enter there without permission (dastūri).”100 Muslim Arabs also engaged in rebuilding and reinforcing the fortification wall and the fortresses.101 The resettling of Arabs on the frontier area also continued into the ‘Abbāsid dynasty, especially under Hārūn al-­ Rashīd.102 In the eighth century, the ‘Abbāsids sent more Arabs from Baṣra in southern Iraq to settle in Darband.103 Darband, on the northern border of the caliphate, became one of the most significant concentrations of Arabs in western Iran. Both the caliphs and the regional governors engaged in population resettlement.104 Resettling Arabs in hinterland cities or towns, and others in the frontiers, was sometimes aimed at overcoming insurgents and internal problems.105 However, the concentration of Arabs in Darband, where the Khazar Turks lived behind the fortification walls, could be associated not only with forestalling or suppressing insurgencies but also with border policy.106 The Muslims protected the gates and narrow passes of the Caucasus for a few centuries by using small garrisons in the forts and strongholds in Dariali and Darband Passes.107 Recent excavations in one of these Sasanian-­period gates, known as the “Alan Gate” in the Dariali gorge, demonstrated a multiperiod stronghold that was guarded and controlled by the Muslim state from the eighth century until at least the 980s and possibly beyond. It should be noted that evidence from excavation of some graves does not clarify whether the buried bodies were Muslim or Christian. Results indicated that Dariali Pass and its strongholds were strategically significant for the Muslims and received much investment.108 To Sauer and his colleagues,109 the considerable size of the garrison, dense occupation of the strongholds, and simultaneous occupation of the two strongholds in the Dariali gorge indicated that much effort and investment were devoted to defending this strategic pass. All this evidence suggests that control over traffic was very important for the Muslim rulers. This evidence also 152

K A R I M A LI Z A D EH

corroborates Mas‘ūdī’s account in the tenth century, according to which it was impossible to cross Dariali Pass without the permission of its garrison.110 The caliphate’s territory in the eastern side of the Caucasus, however, was more vulnerable, as it was easier to enter and cross.111 In Darband, the fortification walls and gates were not enough to protect against invasions from the north, while in the western Caucasus they relied heavily on the mountain ranges as natural barriers. In order to reinforce their borders effectively, they followed the border policy of the Sasanian Empire and their predecessors in the Near East, with mass deportation and compulsory resettlement.112 Overall, the political boundaries of the caliphate in the Caucasus functioned like a border rather than a frontier. A combination of natural and artificial linear barriers formed the political boundary of the caliphate and allowed them to prevent military threats from the northern regions of the Caucasus. Although in the past the concept of borders may have been perceived differently than today,113 it appears that in the Caucasus the Sasanians, the Muslim Caliphate, and their agents possibly desired impenetrable and strict borders. The following passage from al-­Tabari demonstrates this desire and may shed some light on the Muslims rulers’ perception of their political territory and borders with the Khazar Turks in the Caucasus: So ‘Abd al-­Rahman set out with his army and passed through al-­Bab. Shahrbaraz114 asked him what he intended to accomplish, and he replied that he intended [to take] Balanjar.115 [Shahrbaraz] said, “We are indeed happy that [the people of Balanjar] leave us with al-­Bab.” But [‘Abd al-­Rahman] replied, “But we are not happy with this [situation] in our dealings with them until we get at them in their own territory.”116

This historical information indicates that although the principal concerns of Muslim leaders in the Caucasus were to prevent military threats and retain their political territory, they had also a clear understanding of their borderlines in the Caucasus. Thus, we can assume that the immediate and short-­ term solution was to defend and reinforce their border with passes, gates, and fortification walls. The Darband Wall and fortifications defended a borderland rather than strictly demarcating the end of the state territory. Historical sources117 suggest that the Darband fortifications, walls, and gates were restrictive barriers in the frontiers of the caliphate.118 Although both the Sasanians in the sixth to seventh centuries and the Muslim Caliphate may have exercised power beyond their borders,119 these barriers likely marked their political boundary or border. Interestingly, the gates could have functioned as border-­ control installations to regulate the flow of goods and people.120 Overlapping S ocial and P olitical B oundaries

153

The Muslim rulers also added a new dimension to the bordering process by resettling groups of Muslim Arabs to the borders. Transplanting large populations with different cultural, ethnic, and religious backgrounds to the borders in the Caucasus was a complementary strategy to create an extra-­restrictive boundary. Indeed, they created a social boundary that overlapped their political boundary. It was not always possible to establish political boundaries matching other boundaries or set them to coincide perfectly with ethnic and cultural boundaries. Boundaries may or may not have been “congruent,”121 but, in general, they mutually enforced the formation of ethnic identity.122 Settling Arab ethnic groups in frontiers, particularly in Darband, overlapped social and political boundaries. There is not much detailed information about who these settlers were. However, historical accounts suggest that there were settlers from Kufa and Baṣra, and also Syrian Arabs.123 The author of the Hudud al-­ Ālam (early tenth century) also points out that “slaves (bandagān) are brought there [Darband] of every kind of infidel living close to it.”124 Muslim leaders settled the border with populations of Arabs who were presumably loyal to the caliphate and at the same time signified a distinct ethnic group with sharp cultural contrasts with Turkish Khazars. The following passage from al-­Tabari highlights some levels of this ethnic contrast and how effective the Darband Wall was in separating people on the two sides: Then Shahrbaraz said [to ‘Abd al-­Rahman], “Commander, do you know where this man has come from? I sent him some years ago to the wall125 to examine what it and the people on the other side of it were like.”126

After a long period of time this frontier could become an area of assimilation, identity transformation, and adaptation. Initially, however, sociocultural and ethnic differences between newly settled Arabs and Turkish Khazars in Darband could have created the dichotomy of “others” as strangers. One may argue that possibly this “otherness” and protection by walls against threats from people living to the north were fused and reinvigorated with new meanings that are reflected in later Syriac sources such as Pseudo-­Methodios, Apocalypse of Daniel,127 and Book of the Bee.128 Stephen Rapp129 argues that in these texts three ideas of “unclean peoples” of the North, the apocalyptic Gog and Magog, and the myth of Alexander’s Gate are very prevalent. It might also be true that these walls resided in the imaginations and cartographies of Islamic writings as Alexander’s Wall kept Gog and Magog out in the north. Ethnic and cultural differences imply a discord in cultural values that in turn may lead to a decrease in interaction between two ethnic groups and an emphasis on social boundaries.130 Therefore, it seems that overlapping social 154

K A R I M A LI Z A D EH

and political boundaries were intended to restrict possible interactions, occupy the territory and confirm its consolidation, decrease the flow of people and ideas, and make the political boundary stricter and less permeable.131 Muslim leaders expected these functionalities from their borders in the Caucasus, and these are characteristics we find in modern borders today. The story of the borders of the caliphate in the Caucasus indicate that the bordering process was deeply rooted in the basic desire of the states for territorial integrity, and this is a significant issue that is embodied in a fixed political boundary, or “the border.” Thus, the ideal border was the one that could also reflect and be reinforced by multiple boundaries. Borders and frontiers of the caliphate in the Caucasus cannot be fully understood without taking social and ethnic boundaries into account. CONCLUSIONS

Although not all modern political boundaries and borderlines are like borders,132 in this essay I have tried to demonstrate that, equally, not all premodern political boundaries were like frontiers. The Sasanian Empire in some areas had relatively clear borders in both legal and practical senses, particularly with the Byzantine Empire, even though boundaries sometimes shifted due to hostilities. There are accounts of some disputes over the location of fixed borders between these two empires.133 After the fall of the Sasanian Empire, the caliphate did not (or could not) make fixed, strict, restrictive borders possible along all their territorial boundaries. In the Caucasus, however, they took advantage of existing circumstances—­such as natural barriers as well as fortifications, walls, and gates that had already been built by the Sasanian Empire—­and set their own borders. These circumstances did not exist everywhere, so their territorial and political boundaries remained mostly ill-­defined zones, or “frontiers,” particularly those boundaries with the Byzantine Empire.134 Some scholars argue that defensive walls did not mark the end of state territory or the limits of their power execution. They contend that these walls were “indeterminate.”135 Although this argument might be true in some cases, examining the Darband Wall and other bordering mechanisms along the Caucasus range suggests that they possibly differed in their functionality and complexity. These elements as physical barriers not only prevented attacks, decreased penetrations, and allowed states to control interactions, they also created exclusion and inclusion. Approximating various boundaries (natural, political, and ethnic) signified a bordering process that led to the formation of borders on the northern edge of the Sasanian Empire and then the Muslim Caliphate. Overlapping S ocial and P olitical B oundaries

155

All these bordering elements suggest that strict, restricted, and impenetrable borders were desired, particularly in the Caucasus, although this might not be the case in other frontiers of the Islamic Empire. To sum up, analyzing political boundaries of the Sasanian Empire and the Muslim Caliphate in the Caucasus suggests that the concept of “border” is not confined to modern states and “frontier” is not confined to the premodern world. It may also remind us that modern fixed borders did not appear without precedents and background. Political borders of states in the past were not all the same, and could vary from frontiers to borders. Long-­term practices of inclusion/exclusion and boundary-­making ultimately led to the development of modern borders. Moreover, examining population deportation to the frontiers and the deliberate overlapping of boundaries suggests that demographic transformation was part of bordering processes frequently practiced by states. This further underscores that modern borders developed after a long process of attempts by states and empires to organize territories sociospatially. NOTES

I would like to thank Asa Eger for his gracious invitation and the opportunity to contribute to this volume. I have benefited greatly from his insights and comments, for which I am very grateful. My special thanks go to Jason Ur, who inspired and assisted me in the early stages of this research. This chapter is based on research funded by the Iranian Center for Archaeological Research (ICAR). I thank the ICAR and its late director, Dr. Massoud Azarnoush, for funding and for permission to carry out surveying and excavations in the Mughan Steppe in Iranian Azerbaijan. Without that funding and support, this work would not have been possible. Thanks to my wonderful wife, Kerry, for her support and devotion. Of course, the author is solely responsible for any shortcomings. 1. Newman 2003, 2006. 2. Barth 1969a: 9. 3. Rabinowitz 2012, 304; Zartman 2010, 3. 4. Coplan 2012, 509; Diener and Hagen 2012; Green 2012, 577. 5. Diener and Hagen 2010, 4–­5. 6. Brauer 1995, 5–­7. See also Lightfoot and Martinez 1995. 7. Anderson 1996, 1; Kristof 1959; Newman 2003, 127; Parker 2006. 8. Alizadeh 2014. 9. In some social science literature, scholars prefer the notion of borderland over frontier as “a less evocative term” (Newman 2003, 127), considering that frontier can imply “confrontation with other powers” (O’Dowd 2012, 161).

156

K A R I M A LI Z A D EH

10. Kristof 1959; Baud and Van Schendel 1997; Newman 2003; Parker 2006; Feuer 2016, 11–­23. 11. Eger 2015, 311. 12. Barth 1969a. 13. Paasi 1996, 1998, 2011; Newman 2003, 2006. 14. Newman 2006, 148. 15. Paasi 1996. 16. Newman 2006, 148. 17. Newman and Paasi 1998, 191. 18. Newman 2006, 148. 19. Paasi 1998: 72; Kolossov 2005, 620–­621; Newman 2006; Parker 2006. 20. Kolossov 2005, 620–­624. 21. Curta 2005. 22. Elton 1996, 97–­99; Lieu 1998. 23. Kolossov 2005, 620–­624. 24. Oded 1979, 48–­54; Parker 2002. 25. Christensen 1993, 68–­69; Kettenhofen 1994b. 26. Perry 1975, 62–­64; Abisaab 2004. 27. Khodarkovsky 2002; Breyfogle 2005. 28. Eger 2015, 10–­12. 29. It should be noted that demographic manipulations as a bordering mechanism are not completely abandoned, as states like Israel still appeal to such mechanisms in the West Bank. 30. Kolossov 2005, 620–­624; 2006, 13–­15. 31. Sauer et al. 2013. We can also add to the list some modern walls such as the Berlin Wall that divided Berlin into western and eastern sectors, and the wall between Israel and the Palestinians, built by the Israeli government. 32. Bīrūnī 1967, 101–­107. 33. Antrim 2012, 91–­92. 34. Sauer et al. 2013. 35. Kettenhofen 1994a; Braund 2000; Aliev et al. 2006; Gadjiev 2008. 36. Frye 1977, 8–­11; Mahamedi 2004, 152–­157. 37. Frye 1977, 7–­8. Īrānshahr (or Iranzamin) was a historical and political entity roughly comparable with the Roman Empire that stretched from “the Euphrates up to the Amu Darya (Oxus), including the modern states of Iraq and Iran as well as parts of Afghanistan and Turkistan” (Christensen 1993, 15–­16). 38. Aliev et al. 2006; Nokandeh et al. 2006; Sauer 2006; Omrani Rekavandi et al. 2007; Gadjiev 2008; Omrani Rekavandi et al. 2008; Omrani Rekavandi et al. 2010; Sauer et al. 2013. Overlapping S ocial and P olitical B oundaries

157

39. Dignas and Winter 2007, 188. 40. Volodicheva 2002. 41. Kudriavtsev 1993, 65. 42. Gadjiev 2008, 10; Rapp 2014, 136–­137. 43. Kettenhofen 1994a. 44. In Persian texts: dar “gate” and band “knot, band, dam”; in some Islamic period texts: Bab al-­Abwab, “the gate of gates.” 45. Gadjiev 2008, 1–­2. 46. Kuhandez, or present-­day Narin-­Qala. 47. Shahristan. 48. Minorsky 1958; Khan-­Magomedov 1979; Kudriavtsev 1993; Gadjiev 2008. The so-­called The Life of the Kings, a Georgian historiographical text from late antiquity, associates the construction of Darband with an Iranian general called Ardavān or Ardawān (Rapp 2014, 201). 49. Kettenhofen 1994a. 50. Gadžev and Kudrjavcev 2001, 357; Gadjiev 2008, 1–­2. 51. See, for example, Ibn Haukal 1800. 52. Daryaee 2002, 4. 53. Rapp 2014, 123. 54. Kudriavtsev 1993, 23; Gadjiev 2008, 2. 55. Dignas and Winter 2007, 188–­192. 56. Sauer et al. 2015, 886. 57. al-­Tabari 1999, 151–­152. 58. Ibid., 162. 59. See Kettenhofen 1994a; Braund 2000; Sauer et al. 2015. 60. Aliev et al. 2006. 61. Ancient city of Beylaqan. 62. Ahmadov 1997, 21–­22. See also Ricci et al. 2012 for the most recent surveys. 63. Ur and Alizadeh 2013. 64. Parker 2002. 65. Ibn Khordādbeh 1992, 122; Gadjiev 2008, 12–­13. 66. al-­Tabari 1999, 151–­153. 67. Alizadeh 2014. 68. Christensen 1993, 68–­69. 69. Ibid., 68. 70. Kettenhofen 1994b. 71. Brunner 1983, 764–­765. 72. Mas‘ūdī 1841, 401–­402.

158

K A R I M A LI Z A D EH

73. According to Mas‘ūdī (1841, 407), the Turkish Khazar state consisted of many Muslims, Christians, Jews, and pagans. However, the Khazar king and his army converted to Judaism. Many Jews also fled to the Khazar land from the Byzantine Empire and the Muslim Caliphate (see also Golden 1980; Golden et al. 2007; Koestler 1976; Brook 2006). Yāqut al-­Hamawi (2001, 293) also states that Judaism was practiced among the Khazar ruling class. However, according to Ibn Fadlān (2010, 47), another tenth-­century geographer, all Khazars were Jewish. 74. The so-­called Alexander’s wall, Alexander’s gate, or Sadd-­i Iskandar (see van Donzel and Schmidt 2009). It is interesting that the Great Wall of Gurgān in northeastern Iran was sometimes called Alexander’s barrier or Sadd-­i Iskandar (see Kiani 1982). 75. Ghodrat-­Dizaji 2011. 76. Balādhurī 1924, 19–­20; Daniel 1986; Morony 1986. 77. Morony 1986. 78. Hoyland 2015, 113. 79. al-­Tabari 1994, 34–­35. 80. Balādhurī 1916, 324–­325; Kettenhofen 1994a, 16. 81. Minorsky 1958; Blankinship 1994; Garsoïan 1997, 107; Lang 2000, 524. 82. Wasserstein 2007, 373. 83. See Vorderstrasse (chapter 8), in this volume. 84. Kettenhofen 1994a. 85. Wasserstein 2007, 375. 86. See Alizadeh 2011. 87. Modern day Ören Qala; see Ahmadov 1997. 88. Modern place is unknown. 89. Near modern-­day Ganja. 90. Balādhurī 1916, 323; Blankinship 1994, 148–­150; Hoyland 2015, 188. 91. Modern place is not clear, however, Gadjiev (2008, 12–­13) points out that Balanjar is identified with the early medieval city of Upper Chir-­Yurt. 92. al-­Tabari 1994, 38–­40. 93. Koestler 1976, 7; Blankinship 1994, 121–­125. 94. Mas‘ūdī 1841, 404; Balādhurī 1916, 323–­325. 95. Or Arrān, in the Islamic sources (Barthold 2014, 225). 96. Mas‘ūdī 1841, 436; Daniel 1986; al-­Tabari 1994, 40. 97. Kudriavtsev 1993: 48–­52. The outline of the city uncovered through archeological investigations indicated that Darband did not grow beyond its fortification walls until eighteenth century, possibly because it was confined within a fortified area and topographic circumstances (ibid., 63). However, its divisions into quarters and its outline from the eighth-­century Islamic period was preserved until modern times (ibid., 64).

Overlapping S ocial and P olitical B oundaries

159

98. Ibid., 63; see also Vorderstrasse (chapter 8), this volume, for the significance of the northern Caucasus for trade. 99. Anonymous 1982, 454. 1 00. Ibid., 145. 101. Balādhurī 1916, 324–­325; Daniel 1986; Blankinship 1994, 151, 154; Hoyland 2015, 189. 1 02. al-­Yaqubi 2003, 437. 1 03. Daniel 1986. 1 04. See, for example, al-­Yaqubi 2003. 1 05. See Bonner 1996, 58, 141. 1 06. Another example of demographic modification in the early Islamic period frontiers can be found in Balādhurī’s account (1916, 231), according to which Persians, Slavs, and Nabateans were settled in frontier towns in the time of Marwān ibn Muhammad (see Bonner 1996: 58). 1 07. Mas‘ūdī 1841, 435. 1 08. Sauer et al. 2015, 901–­902. 1 09. Ibid. 1 10. Mas‘ūdī 1841, 435. 111. Hoyland 2015, 190. 112. Christensen 1993, 67–­72; Kettenhofen 1994b; Parker 2002. 113. Eger 2015, 310. 114. A Persian military commander (al-­Tabari 1994, 35). 115. Balanjar was one of the major cities in the Khazar state and possibly in modern-­ day Daghestan on the northwestern Caspian coast (Gadjiev 2008, 12–­13). 116. al-­Tabari 1994, 38. 117. See Hudud al-­Ālam (Anonymous 1982), 145 and Mas’ūdi 1841, 401–­402, 435. 118. Kettenhofen 1994a. 119. al-­Tabari 1994, 34–­43. 1 20. Sauer et al. 2015. 121. Elton 1996. 122. Barth 1969b. 123. Mas‘ūdī 1841, 436; Daniel 1986; al-­Tabari 1994, 40. 1 24. Anonymous 1982, 145. Aside from these state-­directed population deportations, political rebels and their families, refugees, and those escaping from taxation and misgovernment probably settled in the frontiers as well (Daniel 1986). 125. Possibly the great wall that stretched from Darband westward over the Caucasus mountain range (G. R. Smith in al-­Tabari: 1994, 38). 1 26. al-­Tabari 1994, 41. See also Yāqut al-Hamawi 2001, 1: 31–­32; van Donzel and Schmidt 2009, 191. 160

K A R I M A LI Z A D EH

127. See van Donzel and Schmidt 2009, 29, 48–­49. 128. See Solomon 1886, 127–­129. 1 29. Rapp 2014, 133–­138. 1 30. Barth 1969b, 15–­16. 131. See Zartman 2010, 11. 132. This is possibly best exemplified by the detention of three American hikers by Iranian border guards on July 31, 2009, when they “crossed” the unmarked Iran-­Iraq border and stepped into Iranian territory (http://​abcnews​.go​.com/​GMA/​american​ -hikers​-held​-iran​-crossing​-border/​story​?id​=​8368888). 133. Kaegi 1996, 84–­85. 134. See Brauer 1995; Eger 2015. 135. Diener and Hagen 2010, 4–­5. REFERENCES

Abisaab, Rula Jurdi. 2004. Converting Persia: Religion and Power in the Safavid Empire. London and New York: I.B.Tauris. Ahmadov, Qara. 1997. Qadim Beylaqan. Baku: Azerbaijan Dovlat Nashriyati. Aliev, Asker A., Murtazali S. Gadjiev, M. Gaye Gaither, Philip L. Kohl, Rabadan M. Magomedov, and Idris N. Aliev. 2006. “The Ghilghilchay Defensive Long Wall: New Investigations.” Ancient West and East 5.1–­2: 143–­177. Alizadeh, Karim. 2011. “Ultan Qalası: A Fortified Site in the Sasanian Borderlands (Mughan Steppe, Iranian Azerbaijan).” Iran 49:55–­77. Alizadeh, Karim. 2014. “Borderland Projects of Sasanian Empire: Intersection of Domestic and Foreign Policies.” Journal of Ancient History 2.2: 93–­115. Anderson, Malcolm. 1996. Frontiers: Territory and State Formation in the Modern World. Cambridge, UK: Polity Press. Anonymous. 1982. Hudúd al-­‘Ālam, the Regions of the World: A Persian Geography, 372 A.H. (982 A.D.). Translated by V. Minorsky. Cambridge, UK: Cambridge University Press. Antrim, Zayde. 2012. Routes and Realms: The Power of Place in the Early Islamic World. Oxford, UK, and New York: Oxford University Press. Balādhurī, Aḥmad ibn Yaḥyá. 1916. The Origins of the Islamic State, Part I. Translated by P. K. Hitti. New York: Columbia University Press. Balādhurī, Aḥmad ibn Yaḥyá. 1924. The Origins of the Islamic State, Part II. Translated by F. C. Murgotten. New York: Columbia University Press. Barth, Fredrik, ed. 1969a. Ethnic Groups and Boundaries: The Social Organization of Culture Difference. Boston, MA: Little, Brown and Company.

Overlapping S ocial and P olitical B oundaries

161

Barth, Fredrik. 1969b. “Introduction.” In Ethnic Groups and Boundaries: The Social Organization of Culture Difference, edited by F. Barth, 9–­38. Boston, MA: Little, Brown and Company. Barthold, Vasilii Vladimirovich. 2014. An Historical Geography of Iran. Translated by S. Soucek. Princeton, NJ: Princeton University Press. Baud, Michiel, and Willem Van Schendel. 1997. “Toward a Comparative History of Borderlands.” Journal of World History 8.2: 211–­242. Bīrūnī, Abū al-­Rayhān Muḥammad ibn Aḥmad. 1967. The Determination of the Coordinates of Positions for the Correction of Distances between Cities: A Translation from the Arabic of Kitāb taḥdīd nihāyāt al-­amākin litaṣḥīh masāfāt al-­masākin. Translated by J. Ali. Beirut: American University of Beirut. Blankinship, Khalid Yahya. 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. Bonner, Michael D. 1996. Aristocratic Violence and Holy War: Studies in the Jihad and the Arab-­Byzantine Frontier. New Haven, CT: American Oriental Society. Brauer, Ralph W. 1995. “Boundaries and Frontiers in Medieval Muslim Geography.” Transactions of the American Philosophical Society 85.6: 1–­73. Braund, David. 2000. “The Caspian Gates in Roman-­Persian Relations in Ancient Transcaucasia.” Archäologische Mitteilungen aus Iran und Turan (AMIT) 32: 37–­41. Breyfogle, Nicholas B. 2005. Heretics and Colonizers: Forging Russia’s Empire in the South Caucasus. Ithaca, NY: Cornell University Press. Brook, Kevin Alan. 2006. The Jews of Khazaria, 2nd ed. Lanham, MD: Rowman and Littlefield. Brunet-­Jailly, Emmanuel. 2005. “Theorizing Borders: An Interdisciplinary Perspective.” Geopolitics 10.4: 633–­649. Brunner, Christopher. 1983. “Geographical and Administrative Divisions: Settlements and Economy.” In The Cambridge History of Iran: The Seleucid, Parthian and Sasanian Periods, edited by E. Yarshater, 747–­777, vol. 3, part 2. Cambridge, UK: Cambridge University Press. Christensen, Peter. 1993. The Decline of Iranshahr: Irrigation and Environments in the History of the Middle East, 500 B.C. to A.D. 1500. Copenhagen: Museum Tusculanum Press and University of Copenhagen. Coplan, David B. 2012. “Border Show Business and Performing States.” In A Companion to Border Studies, edited by T. M. Wilson and H. Donnan, 507–­521. Malden, MA: Wiley-­Blackwell. Curta, Florin. 2005. “Introduction.” In Borders, Barriers, and Ethnogenesis: Frontiers in Late Antiquity and the Middle Ages, edited by F. Curta, 1–­9. Turnhout, Belgium: Brepols. 162

K A R I M A LI Z A D EH

Daniel, E. L. 1986. “ʿARAB iii. Arab Settlements in Iran.” In Encyclopedia Iranica, edited by E. Yarshater, 210–­214, vol. 2, fasc. 2. New York: Columbia University. Daryaee, Touraj. 2002. Šahrestānīhā ī Ērānšahr: A Middle Persian Text on Late Antique Geography, Epic, and History. Costa Mesa, CA: Mazda Publishers. Diener, Alexander C., and Joshua Hagen. 2010. “Introduction: Borders, Identity, and Geopolitics.” In Borderlines and Borderlands: Political Oddities at the Edge of the Nation-­State, edited by A. C. Diener and J. Hagen, 1–­14. Lanham, MD: Rowman and Littlefield. Diener, Alexander C., and Joshua Hagen. 2012. Borders: A Very Short Introduction. Oxford, UK: Oxford University Press. Dignas, Beate, and Engelbert Winter. 2007. Rome and Persia in Late Antiquity: Neighbours and Rivals. Cambridge, UK: Cambridge University Press. Eger, A. Asa. 2015. The Islamic-­Byzantine Frontier: Interaction and Exchange Among Muslim and Christian Communities. London and New York: I.B.Tauris. Elton, Hugh. 1996. Frontiers of the Roman Empire. London: B. T. Batsford. Feuer, Bryan. 2016. Boundaries, Borders and Frontiers in Archaeology: A Study of Spatial Relationships. Jefferson, NC: McFarland. Frye, Richard N. 1977. “The Sasanian System of Walls for Defense.” In Studies in Memory of Gaston Wiet, edited by M. Rosen-­Ayalon, 7–­15. Jerusalem: Institute of Asian and African Studies, The Hebrew University of Jerusalem. Gadjiev, Murtazali S. 2008. “On the Construction Date of the Derbend Fortification Complex.” Iran and the Caucasus 12.1: 1–­16. Gadžev, Murtazali S., and Alexandr A. Kudrjavcev. 2001. “Steinmetzzeichen des 6. Jahrhunderts n. Chr. in Darband.” Archäologische Mitteilungen aus Iran und Turan (AMIT) 33: 357–­390. Garsoïan, Nina. 1997. “The Arab Invasions and the Rise of the Bagratuni (640–­884).” In The Armenian People from Ancient to Modern Times, Volume 1, The Dynastic Period: From Antiquity to the Fourteenth Century, edited by R. G. Hovannisian, 117–­142. New York: St. Martin’s Press. Ghodrat-­Dizaji, Mehrdad. 2011. “Disintegration of Sasanian Hegemony over Northern Iran (AD 623–­643).” Iranica Antiqua 46: 315–­329. Golden, Peter B. 1980. Khazar Studies: An Historico-­Philological Inquiry into the Origins of the Khazars. Budapest: Akadémiai Kiadó. Golden, Peter B., Haggai Ben-­Shammai, and András Róna-­Tas, eds. 2007. The World of the Khazars: New Perspectives, Selected Papers from the Jerusalem 1999 International Khazar Colloquium hosted by the Ben Zvi Institute. Leiden and Boston, MA: Brill. Green, Sarah. 2012. “A Sense of Border.” In A Companion to Border Studies, edited by T. M. Wilson and H. Donnan, 573–­592. Malden, MA: Wiley-­Blackwell.

Overlapping S ocial and P olitical B oundaries

163

Hoyland, Robert G. 2015. In God’s Path: The Arab Conquests and the Creation of an Islamic Empire. Oxford, UK, and New York: Oxford University Press. Ibn Fadlan, Ahamad. 2010. Ibn Fadlan Seyahatnamesi ve Ekleri. Translated by R. Şeşen. Istanbul: Yeditepe [Turkish Translation]. Ibn Haukal. 1800. The Oriental Geography of Ebn Haukal, an Arabian Traveller of the Tenth Century. Translated by W. Ouseley. London: The Oriental Press by Wilson & Co. for T. Cadell, and W. Davies. Ibn Khordādbeh. 1992. Masālik w’al Mamālik (c. 820—­912 CE). Translated by S. Khakrand. Tehran: Mirās-­e Melal Publishing Co. [Persian Translation]. Kaegi, Walter Emil. 1996. “Reconceptualizing Byzantium’s Eastern Frontiers in the Seventh Century.” In Shifting Frontiers in Late Antiquity, edited by R. W. Mathisen and H. S. Sivan, 83–­92. Aldershot, UK, and Brookfield, USA: Variorum. Kettenhofen, Erich. 1994a. “Darband.” In Encyclopedia Iranica, edited by E. Yarshater, 13–­19, vol. 7, fascicle 1. Costa Mesa, CA: Mazda Publishers. Kettenhofen, Erich. 1994b. “Deportations II: In the Parthian and Sasanian Periods.” In Encyclopedia Iranica, edited by E. Yarshater, 297–­308, vol. VII, fascicle 3. Costa Mesa, CA: Mazda Publishers. Khan-­Magomedov, Selim Omarovich. 1979. Derbent: Gornai︠a︡ stena, Auly Tabasrana. Moscow: Izdatelstvo Iskusstvo. Khodarkovsky, Michael. 2002. Russia’s Steppe Frontier: The Making of a Colonial Empire, 1500–­1800. Bloomington: Indiana University Press. Kiani, M. Yusuf. 1982. “Excavations on the Defensive Wall of the Gurgān Plain: A Preliminary Report.” Iran 20: 73–­79. Koestler, Arthur. 1976. The Thirteenth Tribe: The Khazar Empire and Its Heritage. London: Hutchinson. Kolossov, Vladimir. 2005. “Border Studies: Changing Perspectives and Theoretical Approaches.” Geopolitics 10.4: 606–­632. Kolossov, Vladimir. 2006. “Theoretical Limology: Postmodern Analytical Approaches.” Diogenes 53.2: 11–­22. Kristof, Ladis K. D. 1959. “The Nature of Borders and Boundaries.” Annals of the Association of American Geographers 49.3: 269–­282. Kudriavtsev, Alexandr Abakarovich. 1993. Feodal’nyĭ Derbent: puti i zakonomernosti razvitii︠a︡ goroda v VI-­seredine XIII v. Moscow: Nauka. Lang, David M. 2000. “Iran, Armenia and Georgia.” In The Cambridge History of Iran: The Seleucid, Parthian and Sasanian Periods, edited by E. Yarshater, 505–­536, vol. 3, part 1. Cambridge, UK: Cambridge University Press. Lieu, Samuel N. C. 1998. “Euphrates.” In Encyclopedia Iranica, edited by E. Yarshater, 69–­70, vol. 9, fasc. 1. New York: Columbia University Press.

164

K A R I M A LI Z A D EH

Lightfoot, Kent G., and Antoinette Martinez. 1995. “Frontiers and Boundaries in Archaeological Perspective.” Annual Review of Anthropology 24: 471–­492. Mahamedi, H. 2004. “Wall as a System of Frontier Defense during the Sasanid Period.” In The Spirit of Wisdom [Mēnōg ī Xrad]: Essays in Memory of Ahmad Tafazzoli, edited by T. Daryaee and M. Omidsalar, 145–­159. Costa Mesa, CA: Mazda Publishers. Mas‘ūdī, Abī ‘Alī ibn al-­Ḥusayn ibn ‘Alī. 1841. Murūj al-­dhahab El-­Mas’udi’s: El-­ Mas’udi’s Historical Encyclopaedia: Entitled “Meadows of Gold and Mines of Gems,” Volume I. Translated by A. Sprenger. London: The Oriental Translation Fund of Great Britain and Ireland. Minorsky, Vladimir. 1958. A History of Sharvān and Darband in the 10th–­11th Centuries (Taʾrīkh Bāb al-­Abwāb). Cambridge, UK: W. Heffer & Sons Ltd. Morony, Michael G. 1986. “ʿARAB ii. Arab Conquest of Iran.” In Encyclopedia Iranica, edited by E. Yarshater, 203–­210, vol. 2, fasc. 2. New York: Columbia University. Newman, David. 2003. “Boundaries.” In A Companion to Political Geography, edited by J. Agnew, K. Mitchell, and G. Toal, 123–­137. Malden, MA: Blackwell Publishing Ltd. Newman, David. 2006. “The Lines that Continue to Separate Us: Borders in Our ‘Borderless’ World.” Progress in Human Geography 30.2: 143–­161. Newman, David, and Anssi Paasi. 1998. “Fences and Neighbours in the Postmodern World: Boundary Narratives in Political Geography.” Progress in Human Geography 22.2: 186–­207. Nokandeh, Jebrael, Eberhard W. Sauer, Hamid Omrani Rekavandi, Tony Wilkinson, Ghorban Ali Abbasi, Jean-­Luc Schwenninger, Majid Mahmoudi, David Parker, Morteza Fattahi, Lucian Stephen Usher-­Wilson, Mohammad Ershadi, James Ratcliffe, and Rowena Gale. 2006. “Linear Barriers of Northern Iran: The Great Wall of Gorgan and the Wall of Tammishe.” Iran 44: 121–­173. Oded, Bustenay. 1979. Mass Deportations and Deportees in the Neo-­Assyrian Empire. Wiesbaden: Dr. Ludwig Reichert Verlag. O’Dowd, Liam. 2012. “Contested States, Frontiers and Cities.” In A Companion to Border Studies, edited by T. M. Wilson and H. Donnan, 158–­176. Malden, MA: Wiley-­Blackwell. Omrani Rekavandi, Hamid, Eberhard W. Sauer, Jebrael Nokandeh, and Tony J. Wilkinson. 2010. “At the Frontiers of the Sasanian Empire: The Gorgan and Tammishe Walls in Northern Iran.” In Proceedings of the 6th International Congress of the Archaeology of the Ancient Near East: 5 May–­10 May 2009, “Sapienza,” Università di Roma, edited by P. Matthiae, F. Pinnock, L. Nigro, and N. Marchetti, 599–­611. Wiesbaden: Harrassowitz Verlag.

Overlapping S ocial and P olitical B oundaries

165

Omrani Rekavandi, Hamid, Eberhard Sauer, Tony Wilkinson, and Jebrael Nokandeh. 2008. “The Enigma of the Red Snake: Revealing One of the World’s Greatest Frontier Walls.” Current World Archaeology 27: 12–­22. Omrani Rekavandi, Hamid, Eberhard Sauer, Tony J. Wilkinson, Esmail Safari Tamak, Roger Ainslie, Mahmoudi Majid, Seren Griffiths, Mohammad Ershadi, Julian Jensen Van Rensburg, Morteza Fattahi, James Ratcliffe, Jebrael Nokandeh, Amin Nazifi, Richard Thomas, Rowena Gale, and Brigitta Hoffmann. 2007. “An Imperial Frontier of the Sasanian Empire: Further Fieldwork at the Great Wall of Gorgan.” Iran 45: 95–­136. Paasi, Anssi. 1996. “Inclusion, Exclusion and Territorial Identities: The Meanings of Boundaries in the Globalizing Geopolitical Landscape.” Nordisk Samhällsgeografisk Tidskrift 23: 3–­17. Paasi, Anssi. 1998. “Boundaries as Social Processes: Territoriality in the World of Flows.” Geopolitics 3.1: 69–­88. Paasi, Anssi. 2011. “A Border Theory: An Unattainable Dream or a Realistic Aim for Border Scholars?” In The Ashgate Research Companion to Border Studies, edited by D. Wastl-­Walter, 11–­31. Surrey, UK, and Burlington, VT: Ashgate Publishing. Parker, Bradley J. 2002. “At the Edge of Empire: Conceptualizing Assyria’s Anatolian Frontier ca. 700 BC.” Journal of Anthropological Archaeology 21.3: 371–­395. Parker, Bradley J. 2006. “Toward an Understanding of Borderland Processes.” American Antiquity 71.1: 77–­100. Perry, John R. 1975. “Forced Migration in Iran during the Seventeenth and Eighteenth Centuries.” Iranian Studies 8.4: 199–­215. Rabinowitz, Dan. 2012. “Identity, the State and Borderline Disorder.” In A Companion to Border Studies, edited by T. M. Wilson and H. Donnan, 301–­317. Malden, MA: Wiley-­Blackwell. Rapp, Stephen H. 2014. The Sasanian World through Georgian Eyes: Caucasia and the Iranian Commonwealth in Late Antique Georgian Literature. Surrey, UK, and Burlington, VT: Ashgate. Ricci, Andrea, Barbara Helwing, and Tevekkül Aliyev. 2012. “The Neolithic on the Move: High Resolution Settlement Dynamics Investigations and Their Impact on Archaeological Landscape Studies in Southwest Azerbaijan.” Journal for Ancient Studies 3: 369–­375. Sauer, Eberhard W. 2006. Linear Earthwork, Tribal Boundary and Ritual Beheading: Aves Ditch from the Iron Age to the Early Middle Ages. British Series 402. Oxford, UK: BAR. Sauer, Eberhard W., Hamid Omrani Rekavandi, Tony J. Wilkinson, and Jebrael Nokandeh. 2013. Persia’s Imperial Power in Late Antiquity: The Great Wall of

166

K A R I M A LI Z A D EH

Gorgān and Frontier Landscapes of Sasanian Iran. Oxford, UK, and Oakville, CT: Oxbow Books. Sauer, Eberhard W., Konstantin Pitskhelauri, Kristen Hopper, Anthi Tiliakou, Catriona Pickard, Dan Lawrence, Annamaria Diana, Elena Kranioti, and Catherine Shupe. 2015. “Northern Outpost of the Caliphate: Maintaining Military Forces in a Hostile Environment (the Dariali Gorge in the Central Caucasus in Georgia).” Antiquity 89.346: 885–­904. Solomon, Bishop of Baṣra. 1886. The Book of the Bee: The Syriac Text. Translated by E. A. W. Budge. Oxford, UK: Clarendon Press. al-­Tabari. 1994. The History of al-­Tabari (Ta’rikh al-­rusul wa’l-muluk), Volume XIV, The Conquest of Iran. Translated by G. R. Smith. Albany: State University of New York Press. al-­Tabari. 1999. The History of al-­Tabari (Ta’rikh al-­rusul wa’l-muluk), Volume V, The Sasanids, the Byzantines, the Lakhmids, and Yemen. Translated by C. E. Bosworth. Albany: State University of New York Press. Ur, Jason A., and Karim Alizadeh. 2013. “The Sasanian Colonization of the Mughan Steppe, Ardebil Province, Northwestern Iran.” Journal of Iranian Archaeology 4: 98–­110. van Donzel, Emeri, and Andrea Schmidt, eds. 2009. Gog and Magog in Early Syriac and Islamic Sources: Sallam’s Quest for Alexander’s Wall. Leiden and Boston, MA: Brill. Volodicheva, Natalya. 2002. “The Caucasus.” In The Physical Geography of Northern Eurasia, edited by M. Shahgedanova, 350–­376. Oxford, UK, and New York: Oxford University Press. Wasserstein, David J. 2007. “The Khazars and the World of Islam.” In The World of the Khazars: New Perspectives, Selected Papers from the Jerusalem 1999 International Khazar Colloquium hosted by the Ben Zvi Institute, edited by P. B. Golden, H. Ben-­ Shammai, and A. Róna-­Tas, 373–­386. Leiden and Boston, MA: Brill. al-­Yaqubi. 2003. Tārikh, Volume 2. Translated by M. E. Ayati. Tehran: Elmi Farhangi Publishing Co. [Persian translation]. Yāqut al-­Hamawi. 2001. Mu’jam al-­Buldān, 2 vols. Translated by A. Monzavi. Tehran: Iranian Cultural Heritage Organization [Farsi translation]. Zartman, I. William. 2010. “Introduction: Identity, Movement, and Response.” In Understanding Life in the Borderlands: Boundaries in Depth and in Motion, edited by I. W. Zartman, 1–­18. Athens: University of Georgia Press.

Overlapping S ocial and P olitical B oundaries

167

8 Buddhism on the Shores of the Black Sea The North Caucasus Frontier between the Muslims, Byzantines, and Khazars Tasha Vorderstrasse

DOI: 10.5876/9781607328773.c008

168

The Caucasian frontier (figure 8.1) is an important region in the early Islamic period, acting as a flashpoint between the Byzantines and Muslims in south Caucasia and the Khazars and Alans in north Caucasia. However, the significance of this frontier has tended to be overlooked by those who have studied Islamic boundaries and frontiers, despite the fact that historian al-­Maqrizi stated that the ‘Abbāsid caliph al-­Mansur (r. 754–­7 75) described the Caucasian frontier as al-thaghr al-‘ẓam, “the greatest frontier.”1 This is in part because the region is not well documented historically, since there are no sources for the history of the area from the northCaucasian point of view. While the Byzantines and Muslims as well as south-Caucasian writers commented on the region, it was not their main interest. Our knowledge of the area is largely limited to stories about raids into the north Caucasus on behalf of the Muslims, raids into the south Caucasus by various nomadic groups and the Khazars, Byzantine concerns about the Khazars, and some limited ethnographic information about the strange peoples living to the north of the Caucasian Mountains and their religion. The archaeology provides hints of the rich trade that took place over these mountains between the lands to the north and the Muslims and the Byzantines, and indicates that the region was an area that played a vital role in trade along the western part of the Silk Road.

Figure 8.1. Map showing location of sites and trade routes in the Caucasus (after Ierusalimskaia 2012, illus. 55).

This chapter is similar to others in this volume that focus on regions lying outside the Islamic world but that, by virtue of bordering it, are closely connected to it. The historical sources that inform us about the border relations predominantly come from the Islamic rather than the non-­Islamic worlds, and these chapters provide a glimpse into the complexity of trying to reconstruct the border relations of these regions through their material culture. Therefore, this chapter sees the north-Caucasus region as both a political frontier, whose exact borders were not necessarily obvious even to the inhabitants of the area or the travelers who visited the region, as well as a zone of commercial and religious exchange. While the movement of goods has received considerable attention, particularly the trade of silk textiles found in the cemeteries of the north Caucasus and the spread of Byzantine and Islamic silver coins through this corridor, what has received less attention is the spread of people and ideas. While the spread of western religions such as Judaism, Christianity, and Islam to China from the Middle East, the spread of Buddhism to China from northwest India and Central Asia, and the contact between Chinese and Central Asian and Indian Buddhists have been well studied, the spread of Buddhism beyond Central Asia to western areas such as the Caucasus and the role of the Chinese in the spread of Buddhism has received little attention. There have been some studies of the spread of Buddhism to Iran and beyond in the Ilkhanid period,2 but the Buddhist text at the site of Moschevaja Balka in the northwest Caucasus has largely escaped attention. Although the text found there has been published and recognized B uddhism on the S hores of the B lack S ea

169

Figure 8.2. Buddhist text on paper (after Ierusalimskaia 2012: illus. 226a).

as a Buddhist sutra, precisely which Buddhist text it was had not been identified until recently. Further, the significance of this material has been known to scholars of the northern Caucasus and Central Asia, but has largely escaped scholars of Byzantium and the Middle East. This chapter focuses on the spread of Buddhism via one Chinese-­speaking merchant on the north-Caucasian frontier and the implications of the participation of Chinese speakers in this trade on what is the northern part of the Silk Road. The evidence for the Chinese-­speaking merchant and his texts, which include a Buddhist text (figure 8.2), have profound implications for our understanding of long-­distance trade along the borderlands in the early Islamic period. The chapter first discusses the evidence for the Caucasian frontier in the seventh and eighth centuries before turning to the material evidence for the frontier and its trade. It focuses in particular on the Chinese material, notably a Buddhist text now identified as being a vinaya text translated from Sanskrit into Chinese by Yijing. It concludes with a discussion of how this 170

TA S H A VO R D ER S T R A S S E

material helps us better understand the significance of the Caucasian frontier as a place of exchange of goods as well as ideas. It demonstrates that, despite being in the center of a frontier that was an active zone of conflict, trade continued to flourish, according to the material uncovered in archaeological excavations. Further, it provides valuable evidence for the westward movement of Chinese-­speaking merchants, who were often seen as being passive participants in the Silk Road trade. THE NORTHERN SILK ROAD AS A FRONTIER: ZONES OF TRADE AND CONFLICT IN THE CAUCASUS

The Alans, who lived along the northern Caucasian frontier, are the main group important for our understanding of the trade in people and ideas in the Caucasian region. Classical sources suggest that the Alans were a nomadic group who start appearing in the Black Sea region in the first century BCE, associated with the Scythians and Sarmatians, thus suggesting that they were of Iranian origin and part of the great migration of peoples from Central Asia to the lands to the west. Nevertheless, the concept of precisely what ethnicity the Alans were remains the subject of some controversy. There are no records written by the Alans themselves, meaning that we have very little information about them except from the point of view of outsiders, who had their own interests and concerns.3 It is possible that the term Alan may have described a confederation of different groups rather than one single group, who seem also to have controlled other people living in the region. The Alans were located on the western part of the Silk Road, specifically the northwestern part of the route (see below) that skirted the frontier between the north Caucasus and the Islamic world. The term Silk Road, or Seidenstrasse, was adopted in the nineteenth century for the trade networks that crisscrossed Central Asia from China and other parts of East Asia, including Japan, ultimately reaching western Europe via the eastern Mediterranean and the Black Sea. The evidence for the exchange can be seen in the travel of both goods and ideas between the various regions and peoples at different places along the route. Despite the fact that the Silk Road is seen as a series of trade arteries that linked these regions together, it nevertheless also crossed through boundaries of regions that were not necessarily at peace with each other. The Caucasus region was clearly one of these areas, where both trade and conflict were able to coexist. The area was not simply characterized by unceasing military conflict and raiding between the different areas, it was also an area through which commercial goods passed. The Caucasus was not a single, unified region, B uddhism on the S hores of the B lack S ea

171

separated as it was by mountains (see below), but rather a series of different polities that were closely connected to each other despite the challenges of geography and politics. One needs to rethink how lands such as the Caucasus were conceptualized at the time, notably through the factors of “borders, ethnicity, and loyalty.”4 This then leads to a better understanding of the Caucasus as a frontier. As Eger has noted, frontiers were not no-man’s-lands that simply existed between two or more opposing polities, but rather complex landscapes populated by sedentary, nomadic, and seminomadic peoples alike.5 Unfortunately, it can be difficult to untangle the complex web of interrelationships that were present in the Caucasus. One of the reasons for this is the fact that the region is seen from the point of view of the surviving textual sources, which derive from the area to the south of the Caucasus. The peoples of the north Caucasus did not leave behind their own histories, meaning that we must see them through the lens of individuals who were not always sympathetic to them.6 The north Caucasus was often conceived by these sources as being an exotic and strange world, where Gog and Magog lived, for example, and whose inhabitants wanted to attack the settled societies to the south.7 It is clear that in the Islamic period the peoples beyond the borders of the Byzantine and Islamic world in the north Caucasus were seen to be frightening enemies who might attack them at any moment. Nevertheless, the land beyond the Caucasian Mountains was also visited by Sallam the Interpreter in the ninth century, who was ordered by the ‘Abbāsid caliph al-­Wāthiq to look for Alexander the Great’s walls, legendarily built to keep the uncivilized northerners out of the south Caucasus. He noted that the Alans had been partly Christianized and were friends with the Georgians and other groups.8 The Byzantines were not particularly concerned with describing the region until the tenth century, when they made a concerted effort to convert the Alans. Subsequently, the Alans intermarried with the Byzantine imperial family in the Comnenian period.9 The concern about the barbarians to the north is not reflected in Islamic maps, however, where the Alans are merely noted as being to the north of the Byzantines and to the west of Arrān (Azerbaijan), and where Gog and Magog are said to be north of the Caucasus.10 Still, the description of the people of the north Caucasus as barbarians in many texts emphasizes the “otherness” of them and emphasizes their threat to the “civilized world.”11 It is the Caucasus mountain range, however, that acted as a natural frontier.12 The Caucasus mountain range, with its mountains up to 5,000 m high, is a formidable physical barrier that divides Anatolia, Mesopotamia, and Iran to the south from the Volga and the steppe to the north. The region is highly 172

TA S H A VO R D ER S T R A S S E

strategic because it controls access between the steppes of Central Asia, the eastern Black Sea and the western shore of the Caspian, as well as north–­south trade routes between the steppes and Mesopotamia and Syria.13 Further, precisely who the Alans were is unclear, but so is the exact area controlled by the Alans throughout their history, since the non-­Alan authors describing them preferred to focus on issues of who the Alans were allied with, rather than their actual territory.14 It is likely that the Alans were not a single group but rather groups that cannot be now disentangled from the historiographical writings about them written by those who observed them.15 While the precise area that the Alans themselves controlled is unclear, the empires that controlled the south Caucasus, such as that of the Sasanians, took steps to fortify the mountains against invaders from the north with a complex series of walls in the southeast Caucasus, much as other empires tried to keep out invaders elsewhere. This was not successful however. The Byzantines, on the other hand, were generally in control of the southwest Caucasus, which controlled access to the eastern Black Sea ports. Since the Alans occupied the passes, they allied with both the Byzantines and the Sasanians to the south. These differing loyalties could once again indicate that the Alans were not a single group. The Alans themselves came under the control of the empire of the western Turks, which stretched from the borders of China to the Black Sea. Ultimately, the western Turks collapsed around the same time as the Sasanians, and the Turkic Empire was replaced by the that of Khazars.16 After the Islamic conquest, the Muslims reinforced and maintained the frontier, notably the earlier Sasanian walls and fortifications. By the 640s and the 650s, the states of Iberia and Armenia had come under Islamic control.17 Because the initial Islamic Empire was an empire of conquest and expansion, their goal was to expand and conquer as many regions and peoples as possible for religious and practical reasons. They pushed onward into Central Asia and tried to do the same in the north Caucasus, while at the same time fortifying the frontier. The mountains continued to act as an effective barrier, however. This meant that the region of the south Caucasus was once again a buffer zone for the Muslims against invaders from the north, this time under the Khazars, particularly in the seventh and eighth centuries. Much as the Sasanians had before them, they controlled the central and eastern part of the Caucasus while the Byzantines still remained in control of the southwest Caucasus. The Khazars controlled the north Caucasus, as the Alans presumably did also, although the extent of this control remains unclear. Again, the Alans were caught between two powers to their south and once more seemed to ally themselves with both groups. But these alliances seem to have been B uddhism on the S hores of the B lack S ea

173

continually shifting, as is confirmed not only in the secondary sources but also in the Schechter letter in the Cairo Genizah. Some of the Alans seem to have taken a direct role in invading the southwest Caucasus when it came under Islamic control, thanks to this group of Alans’ alliance with the Byzantines in the early eighth century. At the same time, the Khazars’ relationship with the Byzantines became increasingly tense in the late eighth century, thanks to the territorial ambitions of both empires.18 It is therefore perhaps not surprising that the archaeology of Alan settlements consists of a system of fortified sites in the sixth/seventh century to the ninth/tenth century. It has been suggested that the Alan rulers would need to have been strong in order to provide security and that the population increased in this period, meaning that the northern part of the silk route continued to operate into the early Islamic period.19 While a number of Alan cemeteries have been excavated, very little attention has actually been given to urban centers.20 On the other hand, Arzhantseva argues that some of the fortresses were captured by the Turkic peoples who were likely to be Kuban Bulgars under the control of the Khazars.21 It is evident, however, that regarding the region as simply a frontier where war and raids took place needs to be rethought. A FRONTIER OF TRADE: COMMERCIAL EXCHANGE IN AND THROUGH THE NORTH CAUCASUS

Despite the long catalog of conflict and war that has been outlined above, the Caucasus Mountains constituted an important route of trade between the Muslims and their neighbors along the frontier as they had in the Byzantine/Sasanian period. The Caucasian route between the Near East from Mesopotamia and Iran, through the south and north Caucasus to the areas to the north had three branches with three main choke points. One route went from Persia through Caucasian Albania and northwards via Derbent, and led up the Volga to the southern Urals region. Another route started in Armenia and eastern Georgia and crossed the Caucasus Mountains at the Dariali Pass and then up the Volga. A third route, which is of greatest interest here, crossed through western Georgia and then went along the eastern Black Sea. Tracing the trade routes has often been done through finds of silver coins, metal ware, and textiles, pointing to trade between the Byzantines and the Persians and the south Urals in exchange for fur. Later in the Islamic period, goods such as silks, Islamic coins, and metalwork found along this route were apparently exchanged for fur, slaves, and precious stones.22

174

TA S H A VO R D ER S T R A S S E

Archaeological evidence makes it clear that the Caucasus region was important for trade for the Byzantines, Sasanians, Turks, Khazars, and Muslims, and explains why all the powers fought over this region. A number of burial sites excavated in the north Caucasus demonstrate the importance of this trade, and coins found in the north Caucasus show that the routes led directly over the Caucasus Mountains. It has been suggested that some of the Byzantine coins found in the region could have been used to pay mercenaries, although Sasanian coins have been found as well, and a limited number of Islamic coins have also been found in this area.23 The trade in textiles along this route is also indicated by finds at a number of archaeological grave sites in the region, located either on or near mountain passes, that date to the eighth to ninth centuries, arguing that the route, was used frequently.24 The Khazars controlled the northern silk route, and as their empire expanded they were able to control larger sections of it and the fur route.25 However, it was not only the Khazars who, as intermediaries between the Islamic world and the Rus, wanted to take advantage of the trade. Muslim merchants in the north Caucasus also travelled northwards for trade purposes.26 The region shows signs of cultural mixture, which, given its presence on the crossroads of trade, is not surprising.27 Menander the Protector provided a detailed, late sixth-­century Byzantine Greek description of the changing trade that took place in this period—­an account that has remained unchallenged by historians. According to Menander, the western Turks wanted to trade silk with the Sasanian Empire, with the Sogdians acting as the middlemen. The Sasanians, however, refused the western Turkish embassy and as a result they began trading directly with the Byzantines. This explains why one starts finding textiles along the northern silk route, which rose into prominence in this period as the alternative to the southern route.28 The difficulty with this reconstruction is that it relies heavily on one Byzantine text, that of Menander, which details the story of the western Turkish embassy but is obviously sympathetic to the Byzantine point of view. The theory that the northern route suddenly became prominent in this period does not take into account the fact that Chinese goods have been found along this trade route to the Black Sea since the first century BCE, nor the presence of Greco-­Bactrian coins in the south Caucasus in the second century CE,29 which had to have gone along these routes. Further, the presence of large numbers of Sasanian coins in western China30 argues that the southern Silk Road continued to function,31 which also casts the veracity of Menander’s account into doubt. According to Noonan, the wars between the Khazars and the Umayyads between 650 and 750 disrupted the trade between the two regions only for it to B uddhism on the S hores of the B lack S ea

175

recover in the ‘Abbāsid period, while trade with Byzantium was insignificant. The ‘Abbāsids consumed large amounts of fur, which had been unfashionable under the Umayyads, and needed many slaves, leading to a substantial trade in the two.32 This reconstruction relies heavily on coin evidence, however, and while coins are a useful indicator of trade, they are hardly the entire story. The fact that silver coins do not seem to have been traded in the Umayyad period is taken by Noonan to indicate that trade between the regions did not exist, which seems unlikely.33 Nevertheless, the trade in silver clearly did intensify, starting at the end of the eighth century, when dirham hoards start appearing in the lands beyond the Islamic world, and particularly between 800 and 1015, as substantial numbers of Islamic silver coins were exported. These first went through the Caucasus.34 MOSCHEVAJA BALKA AT THE CROSSROADS OF TRADE

The site of Moschevaja Balka is located at the Sanchur Pass, which is on the Bolshaya Laba River, a tributary of the Kuban in the foothills of the Caucasus. Those individuals such as merchants who were traveling along this route would then go over the Sanchur Pass and into Abkhazia via the Byzb River to the port of Pitsunda and then via boat to Constantinople,35 Trebizond, or to the Islamic lands to the east via land routes. Dating to the eighth to tenth centuries, the site is best known for its exceptional collection of silks, which have been published by Ierusalimskaia in several publications.36 The preservation at the site, thanks to its exceptional microclimate, has meant that researchers have had the unique opportunity to study not only well-­preserved textiles but also wood, leather, and other materials that have been highlighted in Ierusalimskaia’s publications on the site. While there have been criticisms of her work,37 she has made the objects from the site more widely available. The material that has been published thus far includes a wide variety of textiles, one of which was lined with squirrel fur.38 Squirrel fur was still considered a luxury item in the Mamlūk period, when Ibn Nubātah claims to be wearing the “fur of the white-­spotted squirrel (sinjāb),” which the commentator states would have been a Russian or Caucasian import.39 Fur has also been found in the lining of other textiles from the north Caucasus, such as headgear trimmed with weasel fur,40 but in another instance it has not been identified.41 According to the analyses of the silk found at Moschevaja Balka by Ierusalimskaia, the fragments came from not only China and Central Asia but also Byzantium, together with local imitations of Sogdian clothing.42 There have been attempts to determine where the silks were produced,43 but this 176

TA S H A VO R D ER S T R A S S E

type of provenance study in regards to textiles is problematic, since textile patterns seem to have been widespread across regions and were frequently copied. The provenance study of the textiles as well as the dating of the site to the eighth to ninth centuries should be accepted with circumspection.44 Therefore, Noonan’s conclusion that the silks provide evidence for trade between the Sogdians and Byzantines45 must be taken with a certain amount of caution. Further, it has been argued that the silks were at the site and used by the individuals buried there because they had been used to pay tolls.46 While this is possible, it is also possible that silk was cheaper for the population there because it was being brought into the region in large quantities, similar to the presence of large amounts of pepper imported to Egypt from India found at the site of Berenike.47 Some of the material found at the Moschevaja Balka site was clearly imported into the north Caucasus, even if its exact provenance is not always clear. However, some cases provide more clues to provenance. For example, a certain eighth/ninth-­century inscribed Greek textile fragment named an individual with a Byzantine title. While it has been suggested that the fragments were lost before it reached its destination,48 this is unclear, and recent studies have reread the fragments. In addition, they have sugested that the fragments may not be from the same piece, as was argued by Ierusalimskaia, but rather represent separate clothing made for different individuals.49 Other signs of international trade at the site include a silk textile with a Kufic inscription and another with a pseudo-­Kufic inscription, as well as glass from Syria with an Arabic inscription, and an amulet inscribed in Arabic.50 Among the graves that were found at the site, material from what is suggested to be the grave of a Chinese-­speaking merchant was found, although the lack of a good archaeological context51 means that this material could suggest the possibility of multiple Chinese speakers at the site, not just one. Further, it is not clear when the fragments were excavated. Ierusalimskaia initially claimed that the material was found in 190152 but later suggested it had been found in 1905.53 The Chinese material dated to the eighth/ninth century included a number of objects: (1) a paper sutra fragment; (2) a piece of papier mâché used in a bookbinding; (3) two accounts (including another one that she suggests is being used as papier mache as well); (4) a silk fragment of a Buddhist prayer flag depicting a man riding a horse in the mountains; (5) a leather triangle with silk gauze, which she interprets as a prayer flag and associates with three fragments of papier mache lotus-­flower appliqués that are glued in layers with silk; and (6) two fragments of the flag.54 While the sutra, the two documentary B uddhism on the S hores of the B lack S ea

177

fragments, and the painting are well known,55 the entire group of material has never been fully published (though it has been described) and therefore it is difficult to assess the entire group. The bookbinding piece of papier mache and leather triangle have been published only recently, but the flowers and additional flag fragments have never been published at all. This has led to some confusion on the part of individuals treating the texts and other objects in detail. The Chinese texts have not been discussed in detail except by Tong56 and Chen,57 because of the problems of their interpretation. Although Ierusalims­ kaia recognized that one of the texts was a Buddhist sutra, she did not know which one it was.58 As determined by this author, as well as independently by Tong,59 followed by Chen,60 the text was authored by the Chinese Buddhist pilgrim and translator Yijing (635–­713). Since the text is fragmentary, it could be one of two Chinese translations of a Sanskrit original by Yijing that date to the early eighth century and that were made in conjunction with a translation school, many of whose members were non-­Chinese:61 the Mūlasarvāstivāda-­ vinaya (Genben shuo yiqie you bu pinaiye), which were rules for monks, translated in 703 at the Ximing Monastery, or Mūlasarvāstivāda-­bhikṣuṇī-­vinaya (Genben shuo yiqieyou bu bichuni pinaiye), which were rules for nuns, translated in 710 at the Da Jianfu Monastery. The two texts closely resemble one another in the preserved passage, and the one character that would make the difference clear is very difficult to read. Tong, followed by Chen, argues that the traces of this character suggest that the text is the Mūlasarvāstivāda-­bhikṣuṇī-­vinaya.62 The character is very poorly preserved and while the traces tend to suggest it, one would want to examine the text in person to be certain of this identification. But whichever text it is, it would have been a code of monastic behavior for monks or nuns. The proper behavior of Buddhist monks and nuns was of considerable concern to scholars of Chinese Buddhism in the fifth to eighth centuries due to the fact that there were varying texts with different rules that had been imported to China in this period, and their translations into Chinese from the Sanskrit or Prakrit originals were not necessarily accurate.63 It is interesting that one of the Mūlasarvāstivāda-­vinaya should be found in the north Caucasus, given the fact that the introduction of these new texts by Yijing was not successful. Yijing had preferred the Mūlasarvāstivāda-­vinaya, which he introduced from India, to other vinaya texts that had previously circulated in China, not only because he was an expert on the texts but also because they had not been corrupted by previous poor Chinese translations.64 Nevertheless, the fact that they were not popular with the majority of the population is supported by the limited number of surviving text fragments from this early period. While some texts translated by Yijing were popular, the 178

TA S H A VO R D ER S T R A S S E

finds of Mūlasarvāstivāda-­vinaya texts are not as numerous and do not seem to have circulated as widely. The find spots of most of his Mūlasarvāstivāda-­ vinaya texts were in Dunhuang, as well as at a number of sites in the Turfan region (Murtuq, Tuquq, Bezeklik caves at Shengjinkou/Sengim and Gaochang/Khara-­Khoja). The finds of texts argue that Yijing was more popular in the region of Turfan than in Dunhuang, but the number of texts is not large. Most of the texts lack any date, but one was dated to the ninth/tenth century.65 It is interesting to compare the thirteenth-­century Letter of Tansar, which despite being preserved only in a later version, purportedly dates to the Sasanian period.66 Whether or not the text really was written in the Sasanian period, it is significant that one of the stories appended to the end of the letter not only has parallels with other existing Persian literature, namely the Sinbad-­nameh, but is also very close to a Chinese version. The story is known already from an earlier fifth-­century Chinese version, while the closest parallel with the story in the Letter of Tansar also comes from the Mūlasarvāstivāda-­ vinaya text translated by Yijing and also a Tibetan version.67 All of this could argue that there was another translation of the Mūlasarvāstivāda-­vinaya that was brought to the Middle East, in addition to the Chinese translation that was found in the northwest Caucasus. The Chinese texts and objects demonstrated that in addition to trade goods Buddhist texts and objects also travelled along the Silk Road. Neelis’s work on Buddhist trade networks has demonstrated how these religious and economic networks were interrelated with one another and allowed for the movement of goods and ideas across different cultural zones. As he notes, Buddhist objects provide the best evidence for the movement of Buddhism along these routes.68 The presence of a Buddhist text in addition to the other Chinese texts also argues that there were Chinese speakers in the area who were Buddhist. It therefore suggests that North Caucasia was part of the Buddhist trade networks outlined by Neelis and shows that Buddhism was an important part of trade, even in this far western part of the route. It does point to what he has observed, however, which was that the objects that are preserved along various parts of the route indicate that the transmission of Buddhist texts was a complex process.69 However, it is unclear why the individual had brought a Mūlasarvāstivāda-­ vinaya to the region because these were texts that were meant for a monastic community. Chen assumed that the text and other Buddhist items must have been brought to the north Caucasus by a monk, perhaps a monk who had renounced his vows visiting the region for either trade or on pilgrimage. According to his theory, this monk could have brought the text for personal B uddhism on the S hores of the B lack S ea

179

use or with the idea of instructing others. These items were then placed in the grave of a Chinese merchant or possibly a grave of the monk himself.70 It seems more likely to this author that the grave was that of a merchant who happened to have a Mūlasarvāstivāda-­vinaya text. As Chen also notes, vinaya texts were owned not only by individuals in monasteries but also among non-­ monastic individuals.71 The silk painting seems to be part of a Buddhist banner72 that has parallels to silk banner paintings from Dunhuang of the eighth/ninth century now in the British Library.73 It also looks similar in style to eighth-­and ninth-­century cave paintings in Dunhuang at the Mogao Caves.74 In this case, it seems to be a Buddhist banner, perhaps depicting the Life of Buddha, that can be associated with the presumed Chinese-­speaking merchant. The second apparent banner fragment is more problematic, however. Although the shape is correct for the top of a Buddhist prayer flag, the fact that it is made from leather and not silk or paper is in contrast to other surviving Buddhist prayer flags known to this author, but this would need to investigated further. There is also some confusion in the discussion of the prayer flags by Tong, who is the only author to discuss the prayer flags from the site in detail. He mistakenly thought that the leather fragment and two additional pieces should actually be identified with a piece that was not identified by Ierusalimskaia as being Buddhist at all but is in fact a pillow.75 The flowers are clearly paralleled by similar ones from Turfan and Dunhuang, although the fact that no pictures have ever been published makes this problematic.76 All of these Buddhist objects are in contrast to another Buddhist artifact found far to the west of China, this time a metal statue that appears in Helgö in Sweden. The statue seems to have been buried sometime around the ninth century, although the exact date of this Buddha, apparently made in the Swat valley, is problematic.77 Unlike the finds from the northwest Caucasus, however, there is no evidence that the Buddha was associated with anyone who was actually Buddhist travelling to Scandinavia. While there was some interest in Buddhism in the early Islamic period, largely as a result of encounters between Muslims and Buddhists in places such as Bactria, and interest in some Buddhist texts,78 there is no indication of a large engagement with Buddhists texts and thought by either the Byzantines or the Muslims in this period. While earlier authors such as Clement of Alexandria mention Buddha, and prior to that the Mauryan emperor Ashoka claims he sent Indian embassies to the Hellenistic courts,79 knowledge of Buddhism remained relatively vague until later periods.80 The Chinese had knowledge of Islam and the Khazars, but knew little about the Alans until the Mongol period.81 180

TA S H A VO R D ER S T R A S S E

While the exact ethnic identity of the individual(s) who had the texts, flowers, and banners is unclear, they do suggest that there were Chinese-­speaking merchants involved in the north-Caucasian trading network along the frontier with the Islamic world.82 The banner and flowers by themselves would point to trade only in Chinese objects, but it is the texts, particularly the documentary accounting texts, which themselves are not of high value except to those who could read them, that argue for the presence of Chinese speakers. While Chinese-­inscribed objects already appear in the Near East in the Roman period, these were on high-­value objects such as silk83 and were not accompanied by other objects that would point to actual Chinese-­speaking individuals in the region. De la Vaissière had concluded that there were Sogdian settlements on the Black Sea,84 but the presence of the Chinese documents and other pieces of material is exceptional, which makes it difficult to place those items within our overall understanding of trade in the region. It seems to be premature to conclude, as Kovalev does, that there is no reason to believe that merchants who were from Sogdia or China operated to the area west of the Caucasus.85 There have been other reports of Chinese material in the Crimea, including a second/ third-­century tombstone of a Chinese individual.86 However, an examination of the original publication of the relief published shows that there is absolutely no evidence for Chinese characters on this gravestone.87 CONCLUSION

The implications of the archaeological evidence from Moschevaja Balka clearly demonstrate how the Caucasian frontier functioned in the seventh to ninth centuries. The Caucasus to the north of the mountains was a contested frontier between the Byzantines, Sasanians, and nomadic peoples, including the Alans. The Byzantines and the Sasanians saw the people to the north as dangerous and as potential enemies, which is not surprising, given the fact that they frequently invaded the settlements to the south. In the sixth century, the region transformed as a result of the expansion of the western Turks, but the evidence indicates that the northern Silk Road, which traversed the Caucasus region, was not a new phenomenon. Further, the archaeological evidence suggests that it did not replace the southern Silk Road but rather operated in conjunction with it, whatever a single Byzantine source might say. In the seventh century, the Byzantine and Sasanian hegemony over the region was replaced by the Byzantines and Muslims to the south and the Khazars to the north, with the Alans remaining, as they had in previous periods, caught between the great powers. Regardless, the material from Moschevaja Balka and other cemeteries B uddhism on the S hores of the B lack S ea

181

indicates that the region remained an important trade route in the eighth/ninth century, even though some scholars feel that the route lost importance in this period. Further, the cemetery also indicates that the region was highly international, with Greek-­inscribed textiles and Chinese documents. The impact of Buddhism in the region remains unclear, however, but there are indications from other texts that the Mūlasarvāstivāda-­vinaya, or at least parts of it, may have been read in the west, perhaps as a result of merchants such as the one(s) buried in the cemetery. The region was part of the Buddhist trade networks that stretched across Asia, indicating that we should look not only to Central Asia as part of these networks but also to areas beyond it. Whoever owned these texts, it demonstrates that Chinese-­speaking Buddhists were involved in trade networks in the west at a far earlier period than one might have expected and that we should not see China and the Caucasus as merely a place that Sogdian, Persian, and Arab merchants visited to obtain valuable trade goods, but rather as an active actor in trade routes along the frontiers of Islam. NOTES

Different versions of this chapter were presented at the ASOR meeting in Chicago in 2012 and the Late Antique and Byzantine Studies workshop the same year. The author would like to thank all the participants for their helpful comments and Dr. B. Muhs of the University of Chicago for his help.

1. See al-­Maqrīzī, Kitāb al-­Muqaffa al-­Kabīr, fol. 92b, quoted in Noonan 1984, 217. The Caucasian frontier is not mentioned in Brauer 1995, 1–­73, for instance. For more on this neglect see Kovalev 2005, 55. 2. Seldeslachts 2007, 144; Azad 2011; Prazniak 2014, 661–­662. 3. Noonan 1984, 288–­289; Alemany 2000; Kouznetsov and Lebedynsky 2005, 31. 4. Standen 2007, 18. 5. Eger 2014, 13–­14, 150. 6. See Alemany 2000 for an extensive collection of sources for the Alans. 7. Shapira 2007, 348–­349; Mako 2010, 53; van  Donzel and Schmidt 2010, 10–­11, 13–­22; Shingiray 2011, 38; Rapp 2014, 134, 136–­138. 8. Van Donzel and Schmidt 2010, 131, 152, 187. 9. Alemany 2000, 170; Garland and Rapp 2006; Arzhantseva 2011–­2012, 3; Jessee and Iasenko 2013, 107–­132. 10. Miller 1986, taf. 37; Galichian 2007, 82. 11. For this phenomenon in China see Bennett and Standen 2011, 80. 12. See Eger 2014, 16, for a discussion of how natural features form frontiers in the Islamic period. See also Dagron 2006, 312. 182

TA S H A VO R D ER S T R A S S E

13. Knauer 2001, 127, 129; Shingiray 2011, 46, 48. For the geography of the region in general, see Volodicheva 2002. 14. See Standen 2007, 18, who observed a similar phenomenon in descriptions of the tenth-­century Liao in China. 15. For this phenomenon in general, see Mullin 2011, 5. For other examples, see Bennett and Standen 2011, 85. 16. Boyce 1968a, 12; Boyce 1968b, 35; Noonan 1984, 173–­176; Noonan 1992, 122; Alemany 2000, 278; Harmatta 2000, 251–­252; Greatrex and Lieu 2002, 57, 81–­82, 91, 100, 113, 115–­121, 128–­132, 170–­171, 209, 272–­274; Arzhantseva 2002a, 444; Alemany 2003, 5–­7; Huward-­Johnston 2007, 191–­192; Kouznetsov and Lebedynsky 2005, 142; de la Vaissière 2005, 249; Arzhantseva 2007, 59, 61; Arzhantseva and Ruzanova 2010, 210; Mako 2010, 51–­52; Shingiray 2011, 6, 38, 46, 138, 142, 282, 409–­411; Orfinskaya and Arzhantseva 2013, 86; Banaji 2014, 27, 30, 34–­38; Rapp 2014, 28. For a summary of the evidence for the dates of the construction of the walls, see Gadjiev 2008, 1–­15. See also Standen 1999 for frontiers in China. 17. Tourmanoff 1954, 158; Alizadeh (chapter 7), this volume. 18. Canard 1971, 353–­357; Golb and Pritsak 1982, 104, 112–­115, 134–­136; Noonan 1984, 172–­173, 176–­178, 181–­197, 219–­222, 227–­229, 248–­249; Toynbee 1973, 507–­508; Blair 1992, 17–­18; Noonan 1992, 113, 122–­123; van Esbroeck 1996, 214; Mango and Scott with Greatrex 1997, 542–­544; Alemany 2000, 173–­174, 176, 244, 250–­254, 270–­272, 294; Ierusalimskaia 2000, 128; Arzhantseva 2002a, 444; Arzhantseva 2002b, 17–­18; Arzhantseva 2007, 59–­61; Huward-­Johnston 2007, 164, 167; Noonan 2007, 240; Shapira 2007, 308, no. 6, 348; Wasserstein 2007, 373–­375, 377; Zuckerman 2007; Vachkova 2008, 351; Mako 2010, 45–­48, 55–­56; Shingiray 2011, 286–­294; Zadeh 2011, 75; Banaji 2014, 38. 19. Arzhantseva 2007, 62–­63; Arzhantseva and Ruzanova 2010, 210. 20. Arzhantseva et al. 2000, 211. 21. Arzhantseva 2007, 62. 22. Noonan 1980, 401–­469; Mundell Mango 2000, 268, 274; Noonan 2000, 285–­286, 291, 293–­294; Kovalev 2005, 64; de la Vaissière 2005, 249–­254; Noonan 2007, 224; Somogyi 2008, 123–­124, no. 138; Arzhantseva and Ruzanova 2010, 208–­209; Tezcan 2012, 129. 23. Noonan 1984, 158–­159, 166, 252–­253, 266; Flyorov 2000, 134, fig. 336; Higham et al. 2010, 656; Somogyi 2008, 209–­211. 24. Noonan 1984, 260–­261; Knauer 2001, 127, 147–­148, no.  28; Stavisky 2002, 765. Kouznetsov and Lebedynsky 2005, 165; Kovalev 2005, 78–­79; Arzhantseva and Ruzanova 2010, 209. 25. Kovalev 2005, 77. 26. Noonan 1985, 179; Noonan 2000, 290; Noonan 2006–­2007, 116; Noonan 2007, 224, 230. 27. Dode 2012, 12–­13. B uddhism on the S hores of the B lack S ea

183

28. Noonan 1984, 257, 259; Blockley 1985, 71; Noonan 1992, 121; Noonan 2000, 286–­288; Kovalev 2005, 60–­61; de la Vaissière 2005, 39, 227–­228, 230–­231, 233–­237; Shingiray 2011, 282. 29. Mielczarek 1997; Simonenko 2001, 53–­59; Tezcan 2012, 137–­138. 30. See for example Thierry 1993; Skaff 1998, 67–­115; Tsumura and Yamauchi 2003. 31. Noonan 2000, 289. 32. Noonan 2007, 33; Golden 2007, 43–­44; Kovalev and Kaelin 2007, 569–­571. 33. Noonan 2007, 233. 34. Noonan 1984, 152–­156, 158; Kovalev 2005, 80; Noonan 2007, 223–­224; Mako 2010, 49; Kovalev 2015, 68–­69. 35. Noonan 1984, 252–­253, 255; Ierusalimskaia 1996, 17; Knauer 2001, 126, 131; Kovalev 2005, 61, 79; de la Vaissière 2005, 237; Arzhantseva and Ruzanova 2010, 209. For maps of the trading route and Moschevaja Balka, see Ierusalimskaia 2012, 88–­89. 36. Ieruslaimskaia 1992, 1996, 2012; Ierusalimskaia and Borkopp 1996. 37. Roth 1999, 523–­529; Dode 2012, 384 s.; Dode 2013, 642–­653; Izbitser 2015, 299–­313. 38. Kazanski 2003, 176; Kovalev 2005, 61. 39. Bauer 2013, 17. 40. Dode 2012, 8. 41. Kajitani 2001, 105. 42. Ierusalimskaia 1967, 55–­78; Noonan 1984, 252–­254; Ierusalimskaia 1996, 115–­122; Knauer 2001, 130–­131,134–­135; Dode 2012, 8, 14. 43. Kovalev 2005, 61. 44. Knauer 2001, 126; Izbitser 2015, 310; Compareti 2015, 40–­41. 45. Noonan 1984: 252–­254. 46. Noonan 1984: 256; Knauer 2001: 131; Kovalev 2005: 61; Arzhantseva and Ruzanova 2010: 209. 47. Cappers 2006; van der Veen 2011, 41, 44–­45. 48. Ierusalimskaia 1996, 250–­251; Ierusalimskaia 2000, 125–­128; Ierusalimskaia 2012, 133–­136, fig. 68. 49. Vinogradov 2013; Sorochan and Filirov 2015, 110–­119; Vinogradov 2015, 133–­140. 50. Ierusalimskaia 1992, 22–­23, cat. nos. 51–­52; Ierusalimskaia 1996, 130–­131, abb. 162, 242–­243, abb. 213, cat. nos. 12–­13; Ierusalimskaia and Borkopp 1996, 100–­101, figs. 116–­119; Ierusalimskaia 2012, 352–­354, 359. 51. Ierusalimskaia 2012, 363. 52. Ierusalimskaia 1992, 30. 53. Ierusalimskaia and Borkopp 1996, 101; Ierusalimskaia 2012, 363. See Izbitser 2015, 307. 54. Ierusalimskaia 1996, 127–­129, abbs. 211–­212; Ierusalismskaia 2012, 363–­368; Ierusalimskaia 2014, 241. 184

TA S H A VO R D ER S T R A S S E

55. Ierusalimskaia 1992, 37, cat. nos.  110–­114; Ierusalimskaia and Borkopp 1996, 101–­102. Fvor the mentions of this material in the secondary literature see Noonan 1984, 257; Knauer 2001, 135; Kovalev 2005, 62; de la Vaissière 2005, 240; Tong Tao 2011, 110–­114; Chen Ming 2015, 229–­268. 56. Tong 2011. 57. Chen 2015. 58. Ierusalimskaia 1996, 128; Ierusalismskaia 2012, 364; Ierusalimskaia 2014, 241. 59. Tong 2011. 60. Chen 2015: 244–­245. 61. Chen 2015, 230–­232, tables 1–­2. For more on his translations in general and their historical circumstances see Gnoli 1977, xvi, xxiii; Panglung 1981, xi–­xiii, xvii; Hirakawa 1982, 11–­12; Chen 2015, 234–­236. 62. Tong 2011, 110, 113; Chen 2015, 245. 63. Yuyama 1979; von Hinüber 1995; Heirman 2008, 257–­258, 260, 266–­271. See in general Heirman 2007. 64. Heirman 2007, 179, 197; Heirman 2008, 260–­261. 65. Chen 2015, 244–­245, 247–­248, table 4. 66. See Boyce 1968a and 1968b. 67. Boyce 1955, 56–­57; Boyce 1968a, 14–­15; Boyce 1968b, 55–­59; Semenov 1996, 89. 68. Neelis 2011, 38, 54. 69. Ibid., 319. 70. Tong 2011, 110–­114; Chen 2015, 245–­247. 71. Chen 2015, 246–­247. 72. Tong 2011. 73. Whitfield 1982, nos. 28–­29, 33–­35. 74. Silva-­V igier 1967, 136; Whitfield and Whitfield 2015, 25, 27. 75. See Tong 2011, 111, 113–­114. Piece originally published in Ierusalimskaia 1996, abb. 14, and in more detail in Ierusalimskaia 2012, 39–­40. 76. Tong 2011, 111, 114. For similar flowers see Whitfield 1985, pl. 45. 77. Zavyalov 1995, 138, 141; Gyllensvärd 2004, 12, 17, 23; de la Vaissière 2005, 250–­251. 78. Prazniak 2014, 656. 79. Compareti 2007; Seldeslachts 2007, 136–­137, no. 19. 80. Berzin 2010, 187–­203. 81. Alemany 2000, 339–­401; Lin Ying 2000–­2001, 354–­359; Kouznetsov and Lebedynsky 2005, 36; Shirota 2005, 231–­261. 82. Noonan 1984, 257; de la Vaissière 2005, 240. 83. For Chinese silk textiles from Palmyra, including those with Chinese characters on them, see Stauffer 1996, 425–­430; Schmidt-­Colinet et al. 2000; Żuchowska 2013a, 133–­154; Żuchowska 2013b, 381–­387; Żuchowska 2014, 143–­162; Żuchowska 2016. B uddhism on the S hores of the B lack S ea

185

84. de la Vaissière 2005, 241, 246, 248. 85. Kovalev 2005, 96. 86. Stavisky 1995, 192; Simonenko 2001, 54; Stavisky 2002, 766, with no references; Knauer 2001, 154, no. 87, references Stavisky 1995 but he states that he could not find the original article; see also Tezcan 2012, 138. 87. Shul’c and Choref 1972, 137, 140, fig. 3. REFERENCES

Alemany, A. 2000. Sources on the Alans: A Critical Compilation. Leiden: E. J. Brill. Alemany, A. 2003. “Sixth Century Alania: Between Byzantium, Sasanian Iran and the Turkic World.” In Ērān ud Anērān. Webfestschrift Marshak: 1–­8. Arzhantseva, I. 2002a. “Alans: Between Byzantium and Khazaria.” In Centre-­Region-­ Periphery: Medieval Europe, Basel 2002 Conference, Preprinted Papers, Volume 1, edited by G. Helmig, B. Scholkmann, and M. Untermann, 441–­445. Hertingen: Folio-­Verlag. Arzhantseva, I. 2002b. “The Christianization of North Caucasus: Religious Dualism among the Alans.” In Die Christianisierung des Kaukasus/Christianization of Caucasus (Armenia, Georgia, Albania), edited by W. Seibt, 17–­36. Vienna: Verlag der Österreichischen Akadie der Wissenschaften. Arzhantseva, I. 2007. “The Alans: Neighbors of the Khazars in the Caucasus.” In The World of the Khazars: New Perspectives, edited by P. Golden, H. Ben-­Shammai, and A. Róna-­Tas, 59–­74. Leiden: Brill. Arzhantseva, I. 2011–­2012. “The Cult of Saint Eustace in the North Caucasus.” Nāmē-­ye Irān-­e Bāstān 11: 1–­12. Arzhantseva, I., D. Deopik, and V. Malashev. 2000. “Zilgi: An Early Alan Proto-­City of the First Millennium AD on the Boundary between Steppe and Hill Country.” In Les sites archéologiques en Crimée et au Caucase durant l’antiquité tardive et le haut moyen-­age, edited by M. Kazanski and V. Soupault, 211–­250. Colloquia Pontica 5. Leiden: Brill. Arzhantseva, I., and S. Ruzanova. 2010. “The Problems of the Beginning of Medieval Towns in the Northern Caucasus.” In Making a Medieval Town: Patterns of Early Medieval Urbanization, edited by A. Buko and M. McCarthy, 203–­222. Warsaw: Institute of Archaeology. Azad, A. 2011. “Three Rock-­Cut Cave Sites in Iran and their Ilkhanid Buddhist Aspects Reconsidered.” In Islam and Tibet: Interactions along the Musk Routes, edited by A. Akasoy, C. Burnett, and R. Yoeli-­Tlalim, 209–­230. Surrey, UK: Ashgate Publishing.

186

TA S H A VO R D ER S T R A S S E

Banaji, J. 2014. “On the Identity of Shahrālānyōzān in the Greek and Middle Persian Papyri from Egypt.” In Documents and the History of the Early Islamic World, edited by P. Sijpesteijn and A. Shubert, 27–­42. Leiden: Brill. Bauer, T. 2013. “‘Ayna hādhā min al-­Mutannabī!’ Toward an Aesthetics of Mamluk Literature.” Mamlūk Studies Review 17: 5–­22. Bennett, G., and N. Standen. 2011. “Historical and Archaeological Views of the Liao (10th–­12th Centuries): Borderlands in Northeast China.” In Places in Between: The Archaeology of Social, Cultural and Geographical Borders and Borderlands, edited by D. Mullin, 80–­98. Oxford, UK: Oxbow Books. Berzin, A. 2010. “Historical Survey of the Buddhist and Muslim Worlds’ Knowledge of Each Other’s Customs and Teachings.” The Muslim World 100: 187–­203. Blair, S. 1992. The Monumental Inscriptions from Early Islamic Iran and Transoxiana. Leiden: Brill. Blockley, R. C. 1985. The History of Menander the Guardsman: Introductory Essay, Text, Translation, and Historiographical Notes. Liverpool: F. Cairns. Boyce, M. 1955. “Indian Legends in the Letter of Tansar.” Asia Major New Series 5: 50–­58. Boyce, M. 1968a. “Introduction.” In The Letter of Tansar, 1–­24. Rome: Istituto Italiano per il medio ed estremo oriente. Boyce, M., trans. 1968b. The Letter of Tansar. Rome: Istituto Italiano per il medio ed estremo oriente. Brauer, R. W. 1995. “Boundaries and Frontiers in Medieval Muslim Geography.” Transactions of the American Philosophical Society 85: 1–­73. Canard, M. 1971. “L’aventure caucasienne du spathaire Léon, le future empereur Léon III.” Revue des études armeniennes 8: 353–­357. Cappers, R. T. J. 2006. Roman Foodprints at Berenike: Archaeobotanical Evidence of Subsistence and Trade in the Eastern Desert of Egypt. Berenike Reports 6. Costen Institute of Archaeology Monograph 55. Los Angeles: University of California. Chen Ming. 2015. “Vinaya Works Translated by Yijing and their Circulation: Manuscripts Excavated at Dunhuang and Central Asia.” Studies in Chinese Religions 1: 229–­268. Compareti, M. 2007. “Buddhist Activity in Pre-­Islamic Persia According to Literary Sources and Archaeology.” Transoxiana 12. http://​www​.transoxiana​.com​.ar/​12 /​compareti​-iranian​_buddhism​.php. Compareti, M. 2015. “Ancient Iranian Decorative Textiles: Evidence from Archaeological Investigations and Private Collections.” The Silk Road 13: 36–­44. Dagron, G. 2006. “Byzance et la frontière. Idéologie et réalité.” In Frontiers in the Middle Ages, edited by O. Merisalo with P. Pahta, 303–­318. Louvain-­la-­Neuve, Belgium: Fédération internationale des instituts d’études médiévales. B uddhism on the S hores of the B lack S ea

187

Dode, Z. 2012. “Costume as Text.” In Dress and Identity, edited by M. Harlow, 7–­18. BAR International Series 2356. Oxford, UK: BAR. Dode, Z. 2013. “ ‘Iranskiy’ khalat Khosrova i ‘adygskaya struya.’ (Ierusalimskaya A. A. Moshchevaya Balka: neobychnyy arkheologicheskiy pamyatnik na Severokavkazskom shelkovom puti. SPb.: GE, 2012. 384 s.).” Rossiyskiy Arkheologischeskiy Yezhegodnik: 642–­653. 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. Eger, A. 2014. The Islamic-­Byzantine Frontier: Interaction and Exchange among Muslim and Christian Communities. London: I.B.Tauris. van Esbroeck, E. 1996. “Lazique, Mingrélie, Svanéthie et Aphkhazie du IVe au IXe siècle.” In Il caucaso: cerniera fra cultura del Mediterraneo all Persia (VI–XI), 20–­26 aprile 1995, 95–­221. Settimane di studi sull’Alto Medioevo 43. Spoleto, Italy. Flyorov, V. S. 2000. The Alans of the Central Caucasus’ Foothills in the 5th–8th Centuries: Prophylactic Rite Directed Against the Dead. Moscow: Polimedia. Gadjiev, M. 2008. “On the Construction of the Derbend Fortification Complex.” Iran and the Caucasus 12: 1–­15. Galichian, R. 2007. Countries South of the Caucasus in Medieval Maps: Armenia, Georgia, and Azerbaijan. London: Gomidas Institute. Garland, L., and S. Rapp. 2006. “Mary of ‘Alania’: Woman and Empress between Two Worlds.” In Byzantine Women: Varieties of Experience, 800–­1200, edited by L. Garland, 91–­124. Aldershot, UK: Ashgate. Gnoli, R. 1977. The Gilgit Manuscript of the Saṅghabhedavastu: Being the 17th and Last Section of the Vinaya of the Mūlasarvāstivādin. Part I. Istituto italiano per il medio ed estremo oriente vol. XLIX. Rome. Golb, N., and O. Pritsak. 1982. Khazarian Hebrew Documents of the Tenth Century. Ithaca, NY: Cornell University Press. Golden, P. B. 2007. “Khazar Studies: Achievements and Perspectives.” In The World of the Khazars: New Perspectives, edited by P. Golden, H. Ben-­Shammai, and A. Róna-­Tas, 7–­58. Leiden: Brill. Greatrex, G., and S.N.C. Lieu. 2002. The Roman Eastern Frontier and the Persian Wars, Part II: AD 363–­630. A Narrative Sourcebook. London and New York: Routledge. Gyllensvärd, B. 2004. “The Buddha Found at Helgö.” In Excavations at Helgö XVI: Exotic and Sacred Finds from Helgö, edited by H. Clarke and K. Lamm, 11–­28. Stockholm: Motal. Harmatta, J. 2000. “The Struggle for the ‘Silk Route’ between Iran, Byzantium and the Türk Empire from 560–­630 A. D.” Kontakte zwischen Iran, Byzanz und der Steppe im 6.-­7. Jahrhundert, edited by C. Balint, 249–­252. Budapest: Paulus-­Publishing Verlag. 188

TA S H A VO R D ER S T R A S S E

Heirman, A. 2007. “Vinaya: From India to China.” In The Spread of Buddhism, edited by A. Heirmanand and S. P. Bumbacher, 167–­202. Leiden: Brill. Heirman, A. 2008. “Indian Disciplinary Rules and their Early Chinese Adepts: A Buddhist Reality.” Journal of the American Oriental Society 128: 257–­72. Higham, T., R. Warren, A. Belinskii, H. Härke, and R. Wood. 2010. “Radiocarbon Dating, Stable Isotope Analysis, and Diet Derived Offsets in 14C Ages from the Klin-­Yar Site, Russian North Caucasus.” Radiocarbon 52: 653–­70. von Hinüber, O. 1995. Selected Papers on Pāli Studies. Oxford, UK: Pali Text Society. Hirakawa, A. 1982. Monastic Discipline for the Buddhist Nuns: An English Translation of the Chinese Text of the Mahāsāṃghika-­Bhikṣuṇi-­Vinaya. Tibetan Sanskrit Works Series No. XXI. Patna, India: Kashi Prasad Jayaswal Research Institute. Huward-­Johnston, J. 2007. “Byzantine Sources for Khazar History.” In The World of the Khazars: New Perspectives, edited by P. Golden, H. Ben-­Shammai, and A. Róna-­Tas, 163–­194. Leiden: Brill. Ierusalimskaia, A. A. 1967. “O severokavkazskom ‘shelkovom puti’ v rannem srednevekov’e.” Sovietskaia arkheologiia: 55–­78. Ierusalimskaia, A. A. 1992. “Kavkaz na Shelkom Puti: katalog vremennoi vystavki.” St. Petersburg, Russia: Gosudarstvennyi Ermitazh. Ierusalimskaia, A. A. 1996. Die Gräber der Moščevaja Balka. Frühmittelalterliche Funde an der nordkaukasien Seidenstrasse. Munich: Editio Maris. Ierusalimskaia, A. A. 2000. “Un chef militaire byzantine au Caucase du Nord? Le ruban de soie de Moščevaja Balka.” In λιθόστρωτον. Studien zur byzantinischen Kunst und Geschichte. Festschrift für Marcel Restle, 125–­130. Stuttgart: A. Hiersemann. Ierusalimskaia, A. A. 2012. Moshchevaia Balka: Neobychyni arkheologischeskii pamiatnik na Severokavkazskom shelkovom puti. St. Petersburg, Russia: Gosudarstvennyi Ermitazh. Ierusalimskaia, A. A. 2014. “Early Medieval Burial Grounds in the North Caucasus.” In Expedition Silk Road: Journey to the West, edited by Hermitage Amsterdam, 240–­247. Amsterdam: Hermitage Amsterdam. Ierusalimskaia, A. A., and B. Borkopp.1996. Von China nach Byzanz: Fruhmittelalterliche Seiden aus der Staatlichen Ermitage Sankt Petersburg. Munich: Bayerischen Nationalmuseum und der Staatlichen Hermitage. Izbitser, E. 2015. “Professional’naya etika i sub’yektivnost’: Ob odnoy monografii dvukh retsenziyakh.” Tyragetia: Arheologie Istorie Anticǎ Serie Nouǎ 9: 299–­313. Jessee, S. and A. Iasenko. 2013. “The Military Effectiveness of Alan Mercenaries in Byzatium.” Journal of Military History 11: 107–­132. Kajitani, N. 2001. “A Man’s Caftan and Leggings from the North Caucasus of the Eighth to the Tenth Century: A Conservator’s Report.” Metropolitan Museum of Art Journal 36: 85–­124. B uddhism on the S hores of the B lack S ea

189

Kazanski, M. 2003. Les peuples du Caucase du Nord: le début de l’histoire (Ier–­Viie s. apr. J.-­C.). Paris: Errance. Knauer, E. 2001. “A Man’s Caftan and Leggings of the Eighth to Tenth Century: A Genealogical Study.” Metropolitan Museum of Art Journal 36: 125–­154. Kouznetsov, V., and I. Lebedynsky. 2005. Les Alains: cavaliers des steppes, seigneurs du Caucase: Ier–XVe siecles apr. J.-C. Paris: Errance. Kovalev, R. K. 2005. “Commerce and Caravan Routes along the Northern Silk Road (Sixth-­Ninth Centuries), Part I: The Western Sector.” Archivum Eurasiae Medii Aevi 14: 55–­106. Kovalev, R. K. 2015. “When and What Regions of the Islamic World Exported Sasanian and Arab-­Sasanian Silver Coins to Early Viking-­Age Northern Lands.” In Myntstudier: Frestskrift till Kenneth Jonsson, edited by T. Talvio and M. Wijk, 68–­83. Stockholm: Svenska Numismatika Föreningen. Kovalev, R. K., and A. C. Kaelin. 2007. “Circulation of Arab Silver in Medieval Afro-­ Asia: Preliminary Observations.” History Compass 5: 560–­580. Lin Ying. 2000–­2001: “Some Chinese Sources on the Khazars and Khawarazm.” Archivum Eurasiae Medii Aevi 11: 339–­364. Mako, G. 2010. “The Possible Reasons for the Arab-­Khazar Wars.” Archivum Eurasiae Medii Aevi 17: 45–­57. Mango, C., and R. Scott with G. Greatrex, trans. 1997. The Chronicle of Theophanes the Confessor: Byzantine and Near Eastern History, A. D. 284–­813. Oxford, UK: Clarendon. Mielczarek, M. 1997. “Remarks on the Numismatic Evidence for the Northern Silk Route: The Sarmatians and the Trade Route Linking the Northern Black Sea Area with Cenral Asia.” In Studies in Silk Road Coins and Culture: Papers in Honour of Professor Ikuo Hirayama on His Sixty-­Fifth Birthday, 131–­148. Kamakura, Japan: Institute of Silk Road Studies. Miller, K. 1986. Mappae Arabicae. Wiesbaden: L. Reichert. Mullin, D. 2011. “Border Crossings: The Archaeology of Borders and Borderlands: An Introduction.” In Places in Between: The Archaeology of Social, Cultural and Geographical Borders and Borderlands, edited by D. Mullin, 1–­12. Oxford, UK: Oxbow Books. Mundell Mango, M. 2000. “Byzantine, Sasanian and Central Asian Silver.” In Kontakte zwischen Iran, Byzanz und der Steppe im 6.-­7. Jahrhundert, edited by C. Balint, 267–­284. Budapest: Paulus-­Publishing Verlag. Neelis, J. 2011. Early Buddhist Transmission and Trade Networks: Mobility and Exchange within and beyond the Northwestern Borderlands of South Asia. Dynamics in the History of Religion 2. Leiden: Brill. Noonan, T. 1980. “When and How Dirhams First Reached Russia [A Numismatic Critique of the Pirenne Theory].” Cahiers du monde russe et soviétique 21: 401–­69. 190

TA S H A VO R D ER S T R A S S E

Noonan, T. 1984. “Why Dirhems First Reached Russia: The Role of Arab-­Khazar Relations in the Development of the Earliest Islamic Trade with Eastern Europe.” Archivum Eurasiae Medii Aevi 4: 151–­282. Noonan, T. 1985. “Khazaria as an intermediary between Islam and Eastern Europe in the Second Half of the Ninth Century.” Archivum Eurasiae Medii Aevi 5: 179–­204. Noonan, T. 1992. “Byzantium and the Khazars: A Special Relationship?” In Byzantine Diplomacy: Papers from the Twenty-­Fourth Spring Symposium of Byzantine Studies Cambridge, March 1990, edited by J. Shepard and S. Franklin, 109–­132. Aldershot, UK: Variorum. Noonan, T. 2000. “The Fur Road and the Silk Road: The Relations between Central Asia and Northern Russia in the Early Middle Ages.” In Kontakte zwischen Iran, Byzanz und der Steppe im 6.-­7. Jahrhundert, edited by C. Balint, 285–­302. Budapest: Paulus-­Publishing Verlag. Noonan, T. 2006–­2007. “Nomads and Sedentarists in a Multi-­Ethnic Empire: The Role of the Khazars in the Khazar Khaganate.” Archivum Eurasiae Medii Aevi 15: 107–­124. Noonan, T. 2007. “The Economy of the Khazar Khaganate.” In The World of the Khazars: New Perspectives, edited by P. Golden, H. Ben-­Shammai, and A. Róna-­Tas, 207–­244. Leiden: Brill. Orfinskaya, O., and I. Arzhantseva. 2013. “The Cut of Clothes of North Caucasian Alans.” Archaeological Textiles Review 54: 86–­96. Panglung, J. L. 1981. Die Erzählstoffe des Mūlasarvāstivāda-­vinaya Analysiert auf Grund der tibetischen Übersetzung. Studia Philogica Buddhica Monograph Series III. Tokyo: Reikuyai Library. Prazniak, R. 2014. “Ilkhanid Buddhism: Traces of a Passage in Eurasian History.” Comparative Studies of a Passage in Eurasian History 56: 650–­680. Rapp, S. H. 2014. The Sasanian World through Georgian Eyes: Caucasia and the Iranian Commonwealth in Late Antique Georgian Literature. Aldershot, UK: Variorum. Roth, H. 1999. “Review of A. Iersualimskaja. Die Gräber der Mošcevaja Balka. Frühmittelalterliche Funde an der nordkaukischen Seidenstraße (München 1996).” Byzantinische Zeitschrift 97: 523–­529. Schmidt-­Colinet, A., A. Stauffer, and K Al ‘Asad. 2000. Die Textilien aus Palmyra: Neue und alte Funde. Damaszener Forschungen 8. Mainz: P. von Zabern. Seldeslachts, E. 2007. “Greece, the Final Frontier? The Westward Spread of Buddhism.” In The Spread of Buddhism, edited by A. Heirman and S. P. Bumbacher, 131–­166. Leiden: Brill. Semenov, G. L. 1996. Studien zur sogdischen Kultur an der Seidenstrasse. Wiesbaden: Otto Harrassowitz Verlag.

B uddhism on the S hores of the B lack S ea

191

Shapira, D. D. Y. 2007. “Armenian and Georgian Soruces on the Khazars: A Re-­ Evaluation.” In The World of the Khazars: New Perspectives, edited by P. Golden, H. Ben-­Shammai, and A. Róna-­Tas, 307–­352. Leiden: Brill. Shingiray, I. 2011. “On the Path through the Shadow Empire: The Khazar Nomads at the North-­Western Frontier and the Islamic Caliphate.” Unpublished PhD diss., Archaeology Program, Boston University. Shirota S. 2005. “The Chinese Chroniclers of the Khazars: Notes on Khazaria in Tang Period Texts.” Archivum Eurasiae Medii Aevi 14: 231–­61. Shul’c, P. N., and M. J. Choref 1972. “Novy rel’ef sarmatskogo kruga.” Sovetsaja archeologija 1: 135–­45. Silva-­V igier, A. de. 1967. The Art of Chinese Landscape Painting: In the Caves of Tun-­ Huang. New York: Crown Publishers. Simonenko, A. 2001. “Chinese and East Asian Elements in Sarmatian Culture of the North Pontic Region.” Silk Road Art and Archaeology 7: 53–­72. Skaff, J. K. 1998. “Sasanian and Arab-­Sasanian Silver Coins from Turfan: Their Relationship to International Trade and the Local Economy.” Asia Major 11: 67–­115. Somogyi, P. 2008. “New Remarks on the Flow of Byzantine Coins in Avaria and Walachia during the Second Half of the Seventh Century.” In The Other Europe in the Middle Ages: Avars, Bulgars, Khazars, and Cumans, edited by F. Curta with R. Kovalev, 83–­150. Leiden: Brill. Sorochan, B., and V. S. Filirov. 2015. “Lenta iz Moschevoy Balki: Problema tolkovaniya istochnika.” Rossiyskaya Arkheologiya 110–­19. Standen, N. 1999. “(Re)Constructing the Frontiers in Tenth-­Century North China.” In Frontiers in Question, edited by D. J. Power and N. Standen, 55–­79. New York: St. Martin’s Press. Standen, N. 2007. Unbounded Loyalty: Frontier Crossings in Liao China. Honolulu: University of Hawai‘i Press. Stauffer, A. 1996. “Textiles from Palmyra: Local Production and the Imports and Imitation of Chinese Silk Weavings.” Annales archéologiques de Syrie 42: 425–­30. Stavisky, B. Y. 1995. “Central Asian, Mesopotamia, and Roman World Evidence of Contacts.” In In the Land of Gryphons: Papers on Central Asian Archaeology, edited by A. Ivernizzi, 191–­202. Florence: Le Lettere. Stavisky, B. Y. 2002. “The Silk Road and Its Importance in History.” In The Turks, Volume I, Early Ages, edited by H. C. Güzel, C. Cem Oǧuz, and O. Karatay, 763–­773. Ankara: Yeni Türkiye. Tezcan, M. 2012. “Eastern Trade during the Hellenic and Roman Periods and the Struggle for East-­West Trade in the Caucasus.” The Caucasus and Globalization 6: 125–­139.

192

TA S H A VO R D ER S T R A S S E

Thierry, F. 1993. “Sur les monnaies sassanides trouvées en Chine.” In Circulation des monnaies, des marchandises et des biens, edited by R. Gyselen, 89–­139. Recherches Orientales 5. Leuven, Belgium: Peeters. Tong Tao. 2011. “Beigaojiasuo de sichou zhi lu.” [The Silk Road in the Northern Caucasus]. In Sichou zhi lu shang de kaogu, zongjiao yu lishi, edited by Luo Feng, 102–­114. Beijing: Wenwu chubanshe. Tourmanoff, C. 1954. “Christian Caucasia between Byzantium and Iran: New Light from Old Sources.” Traditio 10: 109–­89. Toynbee, A. J. 1973. Constantine Porphyrogenitus and His World. Oxford, UK: Oxford University Press. Tsumura M., and K. Yamauchi. 2003. “Sasanian and Arab-­Sasanian Silver Coins from Xinjiang: Sasanian Silver Coins in the Xinjiang Museum.” Silk Roadology 19. In Japanese with English summary. Vachkova, V. 2008. “Danube Bulgaria and Khazaria as Part of the Byzantine Oikoumene.” In The Other Europe in the Middle Ages: Avars, Bulgars, Khazars, and Cumans, edited by F. Curta with R. Kovalev, 339–­362. Leiden: Brill. de la Vaissière, E. 2005. Sogdian Traders: A History. Leiden: Brill. van der Veen, M. 2011. Consumption, Trade and Innovation: Exploring the Botanical Remains from the Roman and Islamic Ports at Quseir al-­Qadim, Egypt. Frankfurt: Africa Magna Verlag. Vinogradov, A. Y. 2013. “Vizantiyskaya lenta protospafariya iz Moshchevoy balkii.” In III po arkheologii Zapadnogo Kavkaza Pamyatniki rannego khristianstva na Zapad nom Kavkaze K 1025-­letiyu kreshcheniya Rusi Materialy mezhdunarodnoy arkheologischeskoy konferentsii, edited by T. A. Pavlenko, R. B. Skhatum, and V. V. Ulitin, 21–­24. Anfimovskie chteniia po arkheologii Zapadnogo Kavkaza. Krasnodar: KGIAMZ im. E.D. Felitsyna. Vinogradov, A. Y. 2015. “Grecheskiye teksty iz Moshchevoy balki i poeticheskaya struktura vizantiyskikh akklamatsiy.” Indoevropeyskoye yazykoznaniye i klassicheskaya filologiya 18: 133–­140. Volodicheva, N. 2002. “The Caucasus.” In The Physical Geography of Northern Eurasia, edited by M. Shahgedanova, 350–­376. Oxford, UK: Oxford University Press. Wasserstein, D. 2007. “The Khazars and the World of Islam.” In The World of the Khazars: New Perspectives, edited by P. Golden, H. Ben-­Shammai, and A. Róna-­Tas, 373–­386. Leiden: Brill. Whitfield, R. 1982. The Art of Central Asia: The Stein Collection in the British Museum. Paintings from Dunhuang, Volume 1. Tokyo: Kodansha International. Whitfield, R. 1985. The Art of Central Asia: The Stein Collection in the British Museum. Textiles, Sculptures and Other Arts, Volume 3. Tokyo: Kodansha International.

B uddhism on the S hores of the B lack S ea

193

Whitfield, R., and S. Whitfield. 2015. Cave Temples of Mogao at Dunhuang: Art and History on the Silk Road. 2nd ed. Los Angeles: Getty Conservation Institute. Yuyama A. 1979. A Systematic Survey of Buddhist Sanskrit Literature. Wiesbaden: Franz Steiner Verlag. Zadeh, T. E. 2011. Mapping Frontiers across Medieval Islam: Geography, Translation, and the ‘Abbāsid Empire. London: I.B.Tauris. Zavyalov, V. A. 1995. “K voprusu o proiskhozhdenii Statuetki Buddy iz Khelgo.” Arkheologicheskie Vesti 4: 137–­142. Zhivkov, B. 2015. Khazaria in the Ninth and Tenth Century. Leiden: Brill. Żuchowska, M. 2013a. “From China to Palmyra: The Value of Silk.” Światowit 21: 133–­154. Żuchowska, M. 2013b. “Palmyra and the Far Eastern Trade.” Studia Palmyreńskie 12: 381–­387. Żuchowska, M. 2014. “‘Grape Picking’ Silk from Palmyra. Han Dynasty Chinese Textile with Hellenistic Decoration.” Światowit XII: 143–­162. Żuchowska, M. 2016. “Palmyra and the Chinese Silk Trade.” In Palmyrena: City, Hinterland and Caravan Trade between Orient and Occident: Proceedings Held in Athens, December 1–­­3, 2012, edited by J. C. Meyer, E. H. Seland, and N. Anfinset, 29–­38. Oxford UK: Archaeopress. Zuckerman, C. 2007. “The Khazars and Byzantium: The First Encounter.” In The World of the Khazars: New Perspectives, edited by P. Golden, H. Ben-­Shammai, and A. Róna-­Tas, 399–­432. Leiden: Brill.

194

TA S H A VO R D ER S T R A S S E

9 This chapter considers the medieval Islamic frontier from the perspective of Armenia, situated on the spatial limits of the world of Islam during the tenth to fourteenth centuries. However, rather than attempt to determine what type of frontier Armenia constituted during this time period, relative to the Islamic center, or to assess whether or to what extent it functioned as a perphery, the chapter examines the ways in which the “frontier-­ness” of Armenia was always a component within cosmographic projects, or material and historical strategies for understanding the world. As was observed in this volume’s introduction, the limits of power are always a matter of perspective, and such was certainly the case at the northern or Caucasus edge of the Islamic world.1 Ultimately, this chapter’s discussion posits both edges and centers as aspects of the imagined worlds produced by medieval people for framing and following courses of political action. When the Armenian highlands are considered from the perspective of historiographical discourses dealing with the medieval period, it seems that medieval Armenians were situated at “the edge of everywhere” (see figure 9.1). Whether the kings of Armenian Cilicia were serving as a buffer for the Rum Saljūq Sultanate against the Mongols,2 or the Caucasus region provided a remote topography at the verge of tenth-­century Arab geography,3 the lands of the Armenians in the Middle Ages seem to have continued the role they played in the late Byzantine period as an apple of discord thrown between

Making Worlds at the Edge of Everywhere Politics of Place in Medieval Armenia Kathryn J. Franklin

DOI: 10.5876/9781607328773.c009

195

Figure 9.1. Relief map showing major urban centers (black circles) of thirteenth/fourteenth-­ century Armenia, as well as the locations of known caravanserais (black diamonds). Note that, as signified by the asterisk to its name, the Jrapi caravanserai has since been moved from its original location in advance of the damming of the Akhurian River in 1975.

the Roman and Persian Empires:4 the Armenians and the Caucasus were construed as an interstitial zone at the edges of world powers. Primary sources from the medieval period hardly help: European travelers such as William of Rubruck or Marco Polo looked upon Armenia as a mountainous expanse that featured fiery lakes and proud ancient houses but little in the way of civilization as they knew it.5 On the one hand, Armenia emerges from historical narrative as a persistent frontier—­a stubborn line of mountain passes and peculiar peoples—­that haunted the periphery through the classical, medieval, and modern periods.6 On the other hand, traditions of history written in Armenia by Armenians cast this frontier quality in a different light, evoking the highlands as a holding line of civilized life and comprehensible rule, or even as a center in itself. Archaeological problematizations of the concept of frontier7 have pointed out that such seeming rigidity of frontiers in the past is often an artifact of 196

K AT H RY N J. F R A N K LI N

the scale of analysis, and that at different scales of engagement what was apparently a line drawn in the sand may reveal itself as a complex zone of practice, interaction, and exchange. More acutely, this chapter would suggest that discussions of frontier at the macroscale are self-­perpetuating, because explorations of practice at the local level reveal centers where edges were expected. Whether a hard line, a zone of interaction, a buffer, or a wasteland, the frontier is first and foremost an imaginary—­a version of the world that is constructed as a place for social thought and action.8 This argument applies particularly to scholarly approaches to medieval geopolitics, but also opens a route to thinking about the ways in which people in the medieval period (roughly the fifth to fifteenth centuries for Armenia) created the worlds they acted within. Focusing in particular on trade practice and the institution of the road inn or caravanserai, this chapter presents a brief examination of such practices of “world making” in the medieval Near East and Armenia, through which not only frontiers but central places as well were created in text, material, and practice. MEDIEVAL ARMENIA: A BRIEF INTRODUCTION

The medieval Armenia that was called Mec’ Hayk’ by locals, Armenia by Byzantines, ‘rmny by Sasanian rulers,9 and al-­Arminiya by Arab administrators was a shifting territory of princedoms, kingdoms, and emirates in the geographical area now contained within the Republic of Armenia, the eastern portion of the Republic of Turkey, the northern provinces of the Islamic Republic of Iran, and the western half of the Republic of Azerbaijan. Armenian historians of the early medieval period (fourth to tenth centuries) emphasize the social and political salience during this era of the naxarar system, a quasifeudal, Iranian-­influenced structure of dynastic houses and inherited estates and offices.10 The medieval period in Armenian history is dominated by the fissive tendencies of the naxarar houses, which would negotiate terms of allegiance with neighboring non-­Armenian states and polities more often than joining together in the name of centralized Armenian sovereignty. The exception to this trend, the historical importance of which suggests the nation-­state orientation of much Armenian historiography, is the Bagratid Kingdom, which unified the highlands briefly in the ninth to tenth centuries, only to dissolve in the advance of the Saljūq invasions. According to predominant historiographical accounts, at this point in time the traditional naxarar “system” was also dismantled, to reconfigure in the guise of “merchant princes” (mec’atun išxan’ner) and their lineages in the high medieval M aking Worlds at the Edge of Everywhere

197

period (late twelfth to fourteenth centuries, called the Developed Medieval period in Armenian archaeology).11 This chapter centers exclusively on highland Armenia (rather than Cilicia), which during the period in question was contained within a shifting constellation of political configurations, frequently centered far outside the Caucasus. These shifts in medieval geopolitics have interesting ramifications for a study of frontiers across this period, as Armenia presents itself simultaneously as a persistent place (the mountain ranges and valleys of the Caucasus are themselves resolutely immobile) but also as a case study in the ways that frontiers are produced through social practice and in imagination. This chapter thus “thinks through” the construction of Armenia as a frontier space during a series of medieval geopolitical projects. The chapter starts by examining literary configurations of Armenia in the tenth century, during the ‘Abbāsid Caliphate. The next episode is situated after the Armenian highlands are encompassed within the Saljūq Empire in the eleventh/twelfth centuries. Finally, the chapter examines practices of spatial production during the thirteenth through late-fourteenth centuries, when sectors of Armenia were administered in alliance with the Bagratid Georgian kingdom, which was itself quickly and heterogeneously subsumed within the Mongol Empire. As already discussed above, the aim of the chapter is thus to compare ways in which Armenia was imagined in spatial terms as an edge or a space-­between—­in other words, as a frontier—­during these political periods, rather than a discussion of where at any given time the “real frontier” lay in relation to Armenia and its cities, villages, rivers, and roads. A central question related to the archaeology and historiography of frontiers that emerges in this chapter is, “a frontier for whom?” The frontier-­ness or liminality of Armenia and its people is a shifting substance over the medieval period, and varies drastically depending on who is writing and for what audience—­whether Arab geographies and histories, Armenian dynastic histories, architectural endowment inscriptions, or legal texts. These overlapping historical perspectives create a strange “hall of mirrors” through which to contemplate Armenia as a place in the past; the addition of archaeological evidence into the analysis complicates rather than resolves our understanding of the palimpsest that is Armenia-­as-­frontier. The question of “frontier for whom” also pertains to discussions of Armenia as a frontier or even as a component of the world of Islam in the present. The question of how the cities and mountain valleys of Armenia functioned as a frontier for, for instance, the Umayyad Caliphate was antithetical to traditional Armenian historiography, which framed Arab administration in Armenia as an “alien element” that left no lasting cultural imprint.12 This is in profound 198

K AT H RY N J. F R A N K LI N

contrast to the impression provided by archaeological and art-­ historical reviews of medieval material culture, which integrate Armenia seamlessly into corpuses of (for example) ceramics centered in Iran or Baghdad.13 Meanwhile, as this chapter explores, medieval Armenian commentators maintained their peripherality to the world of Islam even as they participated within worlds of political practice that spanned the Near East and transcended the apparent boundaries of religious orthodoxy. These frontiers of the imagination form the center of my commentary at the end of the chapter; as the condition of possibility for arguments of absolute alterity between selves and others, such boundaries attain a pernicious strength that outlasts the stone-­and-­mortar of the historical thaghr. THE FRONTIER AS AN IMAGINARY APPLIED TO THE MEDIEVAL PERIOD

Theoretical engagements with the frontier in history, and in the medieval period in particular, struggle to establish or define categorical distinctions between natural and political frontiers.14 This ideal distinction rests on a persistent conviction that while political frontiers or “moral barriers” are produced by states and are therefore created by politics, natural frontiers are easily defined as the line between nature and culture, and are therefore antecedent to political argument.15 The “frontier thesis” of Frederick Jackson Turner16 has been cited as the original formulation of this concept of a natural expanse that influenced political formations; however a by-­now already established thread of research in political ecology has developed the critique that the very idea of an empty “natural expanse” to which human society could be opposed was already heavily laden with politics.17 This is true not only for early modern America but for the medieval Near East as well. A critical source on the creation of the “Islamic frontier” is the ninth-­century compilation of accounts of the Arab conquests by Aḥmad ibn Yaḥyā Balādhurī,18 a Baghdadi poet and scholar. Balādhurī’s work is important not only in its own right, but because of the enormous influence it had on later histories and geographies composed in the medieval period.19 Enumerating the provinces of Armenia and their great cities (Vaspurakan, Bałravand, Dvin), Balādhurī describes how the historic rulers of Armenia built castles, gates, and roads in the mountains.20 Foremost among these rulers was the Persian Khusraw I Anushirwan, who Balādhurī describes as building a great wall between the mountains and the sea. However, this state of affairs changes abruptly at the time of the Arab conquest: “Thus Armenia was in the hands of the Persians until the appearance of Islam, at M aking Worlds at the Edge of Everywhere

199

which time many Siyasijun [Persians settled by Chosroes I] abandoned their forts and cities, which fell into ruins. The Khazar and Greeks thus got possession of what was once in their hands.”21 Interestingly, Balādhurī deploys an almost identical formula to describe the Arab conquest of al-­Sham, describing how the conquerors arrived to seize lands once covered in mighty fortresses but recently fallen into ruins.22 What is fascinating for our purposes is not the implications this has for the “truthfulness” of Balādhurī’s account (which for some would reflect on the question of “real” frontiers versus frontiers of “propaganda”), but rather the way Balādhurī visibly crafts a limit to Islamic civilization in historical imagination. The writing of history thereby emerges as a technique by which places like Armenia are situated as being beyond the pale of civility, both spatially but also temporally: their ruination proclaims that their own societies are in the past. The politically empty space filled by the conquest—­a terra nullius for the Islamic context—­is thus already full of social meaning. Balādhurī’s frontier, like Turner’s, is first and foremost an imaginary, the “meeting point between savagery and civilization.”23 To frame the Islamic frontier as either physical or political is therefore to take a reductively historical (either true account or fabrication) approach to a social and spatial question. As social landscapes, places that are socially produced in time through action, perception, and discourse, frontiers are both-­and-­ neither material and imaginary,24 simultaneously real and “fictively political” barriers.25 Historical and archaeological attempts to categorically separate real from imaginary medieval boundaries seem a perverse operation, given that medieval scholarship is itself premised on several divisions or frontiers that are simultaneously imaginary and very real in terms of their disciplinary and political ramifications: the line between barbarism and civilization, and the frontier between modernity and “pre-­” or “non-­modernity.” The mutual misrecognition of real and imaginary results in the elision of temporal, social, and spatial divisions, such that barbarism can be a time, or premodernity a place or a culture.26 As useful though it may be to analysis (including the current one), “the medieval” as a concept is premised on a narrative break between barbarism and civilization, between the Dark Age inhabited by Europe’s “peoples without history” and the kindling of the Renaissance.27 As was observed by Manzano Moreno, the frontier that divides the temporal realm of ruins from that of ­rebuilding or Reconquista shapes the way the medieval world is imagined by academic disciplines, and the frontier manifests as separating Christian and Islamic “worlds” that contain the hidden burden of the spatiotemporal assumptions elided within them.28 Applications of the frontier concept to the medieval period29 therefore both miss the historically contingent 200

K AT H RY N J. F R A N K LI N

and imaginary nature of various ideas of frontier—­nature/culture, barbarism/ civilization, medieval/modern—­and also neglect the possibility that people in the eighth to fifteenth centuries were active in producing their own spatial imaginaries that had for them very real and localized meanings. This applies as well to strict categorical distinctions such as Brauer’s division between “internal” and “external” frontiers for the medieval period, which hangs on the easy and reliable differentiation between frontiers separating merely political adversaries (internal), and frontiers lying between selves and ontological others (external).30 As observed above, fundamental to the mechanics of frontiers is a division (in the space of perception and imagination) between the familiar and the alien, the orthodox and the heterodox; this dissolves the utility of contrasting “merely political” and flexible internal frontiers versus rigid external frontiers that reflect stark ontological divisions. As I explore here, these distinctions not only shift but are shaped by argument, perception, and practice over time. MAKING GEOGRAPHICAL FRONTIERS IN THE MEDIEVAL PERIOD

The example of Balādhurī provided a glimpse of the way Armenia was imagined as a particular kind of acutely social and political place through the medium of written history. This task of world-­making was also undertaken in a very deliberate and straightforward way by medieval geographers, who catalogued the material culture of the Islamic world and schematized it in lists and represented it in maps. These geographies were as much evocations of a landscape of taste as they were descriptions of mountains or routes: medieval Arab authors recounted in lavish detail the commodities and substances brought to the Near East from India and China, including perfumes, spices, rare woods, silk, and jewels.31 In such geographies the region of the Caucasus is delimited in terms of itineraries for travel, but also through a “cartography of taste,” a world imagined in terms of the desirable, delectable things found and produced there.32 Arabic historians and geographers such as Ibn Khurradādhbih and Ibn Rusta enumerated the furs and swords that originated in Rus and were carried by caravan along the roads of the Caucasus to Khwarezmia, Transoxiana, and Iran. A major element of the north–­south trade in the medieval period was slaves: human commodities that originated in Slavic Eastern Europe and Turkic Central Asia, and that were transported to the Near East, where they became the mamlūk or ghulām armies of the Ayyubid dynasty.33 Within Arab geography the Caucasus was constructed first as distant, then as M aking Worlds at the Edge of Everywhere

201

delectable. The situated perspective of tenth-­century traveler and historian al-­ Muqaddasī has been encapsulated by Wheatley, who imagined travel through the Caucasus from the point of view of the Islamic center in Baghdad.34 Wheatley’s discussion of transportation foci developed the impression of the Caucasus as a mountainous matrix of movement, through which passed the caravans connecting the great empires of Byzantium, Khazaria, and India to that of Islam. The Caucasus was known to al-­Muqaddasī and his contemporaries as al-­Riḥāb, “the spacious,” perpetuating the practice of Balādhurī to conceptualize the Armenian highlands and surrounding territories as vacant of people (though somehow full of commodities). As routes and cities were enumerated, so were the tasteful things to be found in them. Exports along these trade routes from Armīniya, a province within the spacious Riḥāb, included apples and melons, carpets, horses and cattle, saffron, mules, slaves, beaver pelts, sturgeon and herring, and borax from Lake Van and naphtha from the wells of Baku.35 Tenth-­century authors agreed on the importance of textile imports from the region, including the raising of silkworms at Partaw for ibrism silk36 exported to Fars and Khūzistān. Riḥāb’s contribution to the Islamic textile trade was remarkable for two dyes: first, red dye made from madder roots was used in Arrān and Shirvan. Second, cochineal dye made from the qirmiz insect was local to Dvin as well as other cities; the geographer Ibn Ḥawqal references the production of this dye at Dvin, attributing it to a “certain worm.”37 Al-­Muqaddasī’s listing of products and manufactures of the south Caucasus is significantly inextricable from his sense of Riḥāb’s location at the rim of the dār al-­Islām, the Muslim ecumene: “Sturdy wools are produced here, superb surcingles [girdles] usually dyed crimson . . . the fruits cannot be counted nor weighed. Along with that it is an important frontier (thughūr), an exalted region. Here were those who controlled the Aras river (al-­Rass) below the mountains Ḥuwayrith and Ḥarith [Ararat] . . . It is a glory to Islam, an area that invites invaders.”38 Al-­Muqaddasī’s description of Dvin (Dabīl in Arabic) further exhibits that author’s attempts to categorize the space of the Caucasus as simultaneously within the known world and located on the margin of a Muslim cosmology centered in urban Baghdad: Dabīl is an important town with a fortress, well-­fortified. It has many products; its name is famous. Its wool is important, its river fluent; gardens surround it, and it has ancient suburbs, a strong fortress. Its markets are cruciform, its fertility amazing. The mosque is on a large hill, beside which is a church which the Kurds control; in it is a fortress. The buildings are of clay and stone. It has a 202

K AT H RY N J. F R A N K LI N

number of gates, among which are the Bab Kidar (Kiran), Bab Tiflis, Bab Ani. However, despite its distinction, the majority of its inhabitants are Christian, for its population has declined and its fortress gone to ruin.39

Several things are immediately apparent from al-­Muqaddasī’s brief description of Dvin in the tenth century: the city is presented as a rich center, old, and connected by highways to the foremost cities of the region, but at the same time (according to the geographer) ruled by foreigners and inhabited by infidels. Al-­Muqaddasī again repeats the characterization of Armenia as a society once great but now in decline and “gone to ruin:” the civilization of the Armenians is thus situated in the territory of the past. Elsewhere in this section of Aḥsan al-­taqāsīm f ī ma‘rifat al-­aqālīm (The Best Divisions for Knowledge of the Regions), al-­Muqaddasī complains of the difficulties he has in engaging local jurists in questions of philosophy, owing to the difference of their sectarian positions; even the faithful within Armenia are separated from him by distances of understanding. This legalistic frustration compounds the author’s impression of the region as intractable, difficult to parse, and verging on barbarism. The “frontier-­ness” of al-­Riḥāb is thus for the worldly traveler al-­ Muqaddasī primarily and critically a question of ecumenical edges—­that is, the perceptible limits of his coherent and comfortable cultural and religious world. The effect of al-­Muqaddasī’s project—­marking distance in the strangeness of taste and of local traditions—­is the production of an imagined world. A construction of the world in terms of goods to be exploited and exported, of invaders to be kept at bay, and of Christians who dwell among ruins, al-­ Muqaddasī’s geography is inherently political. As in the geographies of value produced and perpetuated by the world s­ ystems of global capital, these imagined worlds of taste support a politics also centered on the medieval cities of Arabia and Europe. Al-­Muqaddasī’s written world of values (religious orthodoxy, intelligibility of language, wealth in goods) itself circulated as a medieval text, and was negotiated not only by urban Baghdadis but also by world travelers. Such worlds-­in-­text—­and their accompanying maps—­created an imagined world with the Islamic cities in the center and the tastes and customs of the empire at distances befitting their exoticness, at the edges and limits of civilization. MEDIEVAL ARMENIAN WORLD-­M AKING

The geopolitical effect of imagined worlds like al-­Muqaddasī’s is to situate places like Armenia on the frontier of the civilized landscape, both to increase the allure of its products and to (literally) marginalize the actions M aking Worlds at the Edge of Everywhere

203

and interests of the people living there. Postcolonial histories and archaeologies that would pick apart or add nuance to systems approaches to the past often state their aim as “focusing on the periphery.”40 But as mentioned in chapter 1 of this volume, people living within the peripheries of imagined worlds centered elsewhere undertook their own projects of world­and boundary-­making—­and the medieval Armenians are a stellar example. Throughout the fifth to seventh centuries, Armenian naxarars and the clerics under their patronage distinguished themselves ecumenically from the Byzantine Christian world through a series of theological councils. At successive summits held at Dvin, the capital city in the Ararat plain, the Armenian nobility and clergy (a class united by kinship as well as patronage) rejected the Christological referenda of the 451 Council of Chalcedon.41 This long-­ upheld doctrinal distinction is regarded by historians as a strategic political move by the Armenians to distance themselves both from Byzantine orthodoxy as well as Persian affinity, since the dual nature of Christ demanded by Chalcedon resonated strongly with tenets of Zoroastrianism. Medieval Armenian Christianity retained a peripheral status despite the territorial envelopment of the highlands themselves within the Byzantine or Sasanian Persian Empires. The construction and maintenance of a distinct Armenian Church through the medieval period was therefore a labor of literal ecumenical production, creating a world of religious orthodoxy with frontiers that corresponded to no political territory but rather to the rhetorical performance of Armenians’ identity as congregants as well as subjects within empires. The labor of world-­making undertaken by medieval Armenians is demonstrated as well by attempts to situate the Armenian kingdoms and their subjects within geographical depictions of the world—­specifically that of the seventh-­ century scholar Ananias of Shirak (called Širakac’i). One of several works on science produced by the medieval scholar, the Ašxarhac’oyc’ (“Geography,” literally “Show of the world”) presents an authoritative account of the world at that time. Interestingly, in his arguments about the shape of the world and delineation of its centers and edges, Ananias himself crosses the frontiers of standing traditions of imagining the world in the medieval Near East, as he works between scriptural and Hellenistic (Ptolemaic) cosmological models.42 For example, while following a Ptolemaic model in describing the shape of the global earth and its various zones, Ananias departs from Ptolemy and follows the scriptures in declaring the city of Jerusalem to be the center—­a “place equidistant from all extremities”—­of this inhabited world; the project of Širakac’i is thus an ecumenical geography, aligning the world according to moral valences, with “extremity” understood in civilizational terms as well 204

K AT H RY N J. F R A N K LI N

as in terms of “real” distance.43 A tension emerges however, as Širakac’i situates his Ašxarhac’oyc’ from a perspective particular to Armenia—­a place definitely on the edge of a map centered on Jerusalem. Širakac’i describes Upper Armenia as the region highest in elevation, “not only of Armenia but of the whole world,” citing the fact that the waters of the Armenian highlands feed the great rivers that flow “to the four corners of the earth.”44 The geographer thus evokes the region as a gravitational center, an axis mundi, and a point of origin. In a similar mode, Širakac’i draws on literary precedent, arguing that “Constantine of Antioch . . . says that the Ark came from the east to us in the middle of the earth”45 and thus rhetorically situates Armenia (the Ark’s final destination) and the highlands squarely at the center of the inhabited, civilized regions described in classical geography and scripture. The literary world-­making of Širakac’i further demonstrates how, in the practice of locating Armenia in a broader political world, Armenians tacked back and forth between worldviews of their own, in which Armenia was central, and the cosmologies of others, which relegated the Armenians to the edge of the map. These doctrinal and literary works of frontier-­making were paralleled and reinforced by the production of ontological boundaries, or demarcations between political selves and others in terms of the conception of proper rule and law. This practice is especially well illustrated in the medieval period by the law code or Datastanagirk’ compiled by the Armenian cleric Mxit’ar of Goš (Gošec’i).46 A major part of the context of the Datastanagirk’ is the administration of Armenia through the eleventh and twelfth centuries under the Saljūq Sultanate47 when, as historian Robert Thomson observed, Armenians like Gošec’i were acutely aware that they were not “masters of their own destiny,”48 a phenomenon felt particularly acutely in the sense of Armenian daily life being regulated by institutions external to the world of Armenian Christian practice. Mxit’ar states in his own introduction that the intention of the law code is to juridically distinguish legally regulated Armenian daily and public life from the traditions upheld in Muslim Saljūq courts, given that, due to the lack of a written code, the Armenians had been exposed to “blame not only from foreigners but also from Christians.”49 Written at the end of the twelfth century, the proscriptive and prescriptive statutes of the Datastanagirk’ encompass all aspects of the Armenian social world, from adultery to spoils of war to the minting of money to the theft of eggs from nesting birds. The majority of the law code’s precepts is drawn from scripture and Biblical commentary; however, the code clearly pronounces its jurisdiction over the bodies of Armenians as political subjects as opposed to merely their spiritual state as congregants. Through the medium of the written law—­and presumably, its M aking Worlds at the Edge of Everywhere

205

continued enforcement—­Mxit’ar and his contemporaries produced a world of definitively and legitimately Armenian practice with margins that could be upheld in court, thus enabling a practical maintenance of the “limits of the civilized”—­a legal frontier. The written document of the Datastanagirk’, like the written ecumenical geography of Ananias Širakac’i, was more important in maintaining a cultural boundary between Armenians and “others” during the latter twelfth century than a physical frontier. However, like the liturgical implications of the ecumenical world-­making of the Armenian clergy, the law code also produced a world in practice centered on the Armenian highlands, with ramifications for space, landscape, and the movement of Armenians and “foreigners.” By the time of Mxit’ar Gošec’i’s death in 1213, highland Armenia had been encompassed as vassal principalities within the Bagratid Kingdom of Georgia. During the following period of rule by Armenian princes in the name of the Georgian Crown, the Armenian landscape was further characterized by projects of world-­making that appear more concerned with the expansiveness of their scope than with the hard lines of their edges. For the rest of this chapter I consider the medieval Armenian road inn, or caravanserai, as a project of world-­making that was contextualized by Armenia’s status as an “edge” but was also part of active practices of interiority rather than liminality, drawing us back to the question of how people on edges produce centers through material and discursive labors. In this I build on other discussions in this volume,50 which have already elaborated on the role of architecture in producing coherent centers on apparent edges. CARAVAN INNS IN MEDIEVAL ARMENIA: CENTERING ON THE FRONTIER

Halls designated for traveling humans and their goods and animals had a long tradition by the medieval period, dating to infrastructural institutions of the Achaemenid and Roman Empires.51 Within medieval geographies like that of al-­Muqaddasī, inns for travelers were designated using either the Persian-­derived term caravanserai (karwan, “group of travelers”; saray, “house, hall, palace”) or a number of other terms, including khan, funduq (derived from the Greek pandocheion), or ribāṭ, depending on variations in the structure and function of the building.52 The institution of the caravanserai manifested differently and attained varying social meanings in different periods and regions of the Near East, from Karakhanid Iran to Mamlūk al-­Sham to Saljūq Anatolia.53 206

K AT H RY N J. F R A N K LI N

From dedication inscriptions and historical sources, we know that the išxan’ner (princes) of medieval Armenia constructed road inns, which were designated in those sources as karavanatn’ner (caravan houses), pandoks (likewise derived from the original pandocheion), and xanaparhs (urban hostels, sometimes with associated xalpaks, or storefronts/stalls). Within his law code, Mxit’ar of Goš stipulated not only that the endowment of “dismounting houses” (often called ijavanactn’ner) was part of the prerogative and duty of princes, but also that the revenue from inns was a lawful and proper form of princely income.54 Such profits from karavanatn’ner frequently manifested as donations made by princes to churches in the highlands: for instance, the famous 1215 donation inscription of Prince Tigran Honenc’ on St. Gregory’s Church in Ani mentions the incomes from multiple inns that the prince presented to the monastery.55 By thus converting revenues into the currency of pious charity, the Armenian princes operated in a manner equivalent to the Muslim waqf system, which functioned through the Near East and Rum Sultanate during the medieval period.56 Due to this shared importance within both Christian and Muslim institutions of charity, hospitality, and piety on the part of upwardly mobile or princely donors, the institution of the road inn in Armenia was therefore a political space that crossed apparent political and religious borders. This capacity of the road inn as an “in-­between” institution that encompassed multiple political ecumenes round a roadside center was built into the architectural space of the inns themselves. The function of the karavanatun as an architectural space that enclosed travelers within a far-­reaching world of politicized hospitality emerges from analysis of the extant road inns within the Republic of Armenia as well as from intensive research at the Arai-­ Bazarjuł karavanatun, excavated in 2011 (figure 9.2). Historical data suggest that the now-­ruined caravan inn south of the contemporary village of Arai-­ Bazarjuł, located in the Kasakh River valley, Aragatsotn, was constructed in 1213 by the Armenian merchant-­išxan Vače Vačut’yan.57 Excavations within the still-­standing northern wall uncovered the interior of a rectangular structure approximately 30 m in length, divided by two lines of low arches into three parallel longitudinal galleries (see figures 9.2–­9.4). Within the central barrel-­vaulted gallery between the arches the floor was constructed of hard-­ packed clay (marked occasionally by burning), whereas the two lateral galleries, lined with flagstones, were an average of 50 cm lower and they sloped slightly downward from the central gallery to the outer edge of the building to facilitate cleaning and draining. The upper edge of the lateral galleries also featured stone troughs running lengthwise just outside of the arch plinths: at one point M aking Worlds at the Edge of Everywhere

207

Figure 9.2. Plan view of excavations at the Arai-­ Bazarjuł caravanserai in the Aragatsotn region, Armenia. The excavations are located in the structure’s northeast corner. Note the standing north wall of the building as well as the arch base in the west baulk (with associated clay flooring), the stone trough, and the paving-­ stone flooring.

the rim of the stone trough had been pecked through to create a tethering hole, corroborating the impression that the troughs were meant to contain fodder for caravan pack animals such as horses, donkeys, or even perhaps camels (all are attested from medieval Armenia). This structure unifies a number of road inns from the thirteenth to fourteenth centuries in Armenia, including inns at Jrapi, Aruč, Kotrac’, Yaiji (Harjis), and the Selim Pass. The form of these buildings—­long, vaulted halls built of a stone facing over a rubble core, supported by columns on square bases—­also formally unifies the road inns with preserved monastic architecture from the medieval period, specifically the narthex, or gavit, of medieval churches. Significantly, the role of the gavit was parallel to the caravanserai, as a relatively open, hospitable liturgical space available to lay worshippers outside the restricted zone of the church itself: an “in-­between” space mediating the passage of grace between the sacred sanctuary and the secular world outside. These buildings were constructed of volcanic stone—­often in red-­and-­black-­checkered patterns—­and 208

K AT H RY N J. F R A N K LI N

Figure 9.3. Reconstructed plan of the Arai-­Bazarjuł caravanserai showing excavated areas (shaded) as well as locations of (a) square gallery supports, (b) long troughs, and (c) central drainage channels.

featured barrel-­vaulted ceilings built from intersecting arches. The southern section of the Arai-­Bazarjuł karavanatun had collapsed in places, such that it was possible to recreate the dimensions of the volcanic tufa facings from which the arched ceiling had been constructed. This barrel vault was structurally nearly identical to the ceiling inside the nave of the cathedral of Astvac’nkal, located 3 km to the east in the Kasakh Valley and constructed by Vače Vačut’yan’s son, Kurd. It was at this church, Astvac’nkal, that a sketched design for the now-collapsed narthex’s muqarnas ceiling was found inscribed on a spare piece of building stone during renovations.58 The effect of the muqarnas carving in medieval architecture is to create structural form (dome, arch, vault) that is simultaneously delicately organic and harmoniously orderly. In their deployment of muqarnas in skylights, entrances, and ceilings, the gavit and caravan inn in Armenia are also, un-­accidentally, very structurally similar M aking Worlds at the Edge of Everywhere

209

Figure 9.4. Reconstructed scale elevation of the Arai-­Bazarjuł caravanserai galleried hall. Note raised central clay floor for human patrons and the lower, outer (lateral) galleries for caravan animals.

to Saljūq civic architecture, including khans.59 The architectural form suggests that the Armenian išxan’ner were eager in their endowed buildings to center their political subjects (congregations and travelers) within an orderly model of the world with their benevolent power at its apex. At the very least, it is significant that the princely technique of situating subjects in built space was shared across the putative Saljūq-­Armenian frontier. The similarities of form across seemingly religious and political boundaries resonates in the institutional function of the road inn not only as a rest stop on the highway but also as a locus for pious authority in the Armenian tradition. As mentioned above, caravanserais were implicated in the performance of pious charitable donations undertaken by Armenian nobles through their inclusion in lists of assets donated to monasteries and churches. Caravanserais were also in other ways drawn into the careful construction of worlds of political authority in medieval Armenia through the techniques of architectural inscription. That is, the public, hospitable space of the caravan inn was one of the repertoire of architectural places decorated with stone inscriptions by princes in an effort to perform the political identity they aspired to embody in the medieval world. So the same spatial tradition of politics that encouraged Tigran Honenc’ to donate inns to his church at Ani, in return for prayers for his immortal soul and worldly power, also manifested in the prince Česar Orbelyan in 1332, who inscribed directly on his caravan inn at Selim a solicitation for the prayers of 210

K AT H RY N J. F R A N K LI N

Figure 9.5. The entrance to the caravanserai located in the Selim Pass, Vayoc Dzor region, Armenia. This caravanserai was built in 1332 CE by Česar Orbelyan. Note the muqarnas and zoomorphic decoration on the entrance above the door; the dedication inscription is on an interior wall.

passersby (see figure 9.5). In both cases (Honenc’ and Orbelyan) the inscriptions start with a performative location of the prince in relation to a worldly hierarchy surmounted by higher powers—­the Mxargreceli/Zakaryan princess for Honenc’, and Abū Sa‘id Bahadur Khan of the Ilkhanids for Orbelyan.60 This style of referential donation can also be found on Saljūq khans from the thirteenth century; however, rather than merely an architectural or epigraphic “trend,” it might be more compelling to consider this as a practice of expansive political world-­making shared across the Islamic and Christian traditions. While the situatedness of Armenian politics between and within Byzantine and Arab worlds has garnered extensive and eloquent historical discussion, such analyses have foregrounded the polysemy of Armenian rulership at the expense of the significant role played by material culture—­such as crowns and silken garments—­in the production of that shared ecumene.61 Following on the discussion of world-­making through the description of exotic places and M aking Worlds at the Edge of Everywhere

211

things in Arab geographies in the sections above, I am interested in the ways that material culture, including architecture, inscriptions, and also everyday objects as well as food, co-­constructed a medieval world for Armenians. The question of epigraphic performance speaks primarily to the significance of the caravanserai institution for the people who built them: what did the institution consist of in terms of practices located within the space of the inn? The material-culture assemblage recovered at Arai-­Bazarjuł comprised a very small sample of glazed and sgraffiato-­decorated pottery wares and a much larger assortment of slipped red wares. These two general ceramic datasets (glazed wares and “plain” or red wares) support different kinds of discusions about practice and material imaginaries situated within the galleries of the caravanserai. The glazed assemblage indicates in a programmatic way the integration of the Kasakh Valley highway nodes with medieval urban production and consumption trends.62 From the tenth through the fifteenth centuries, sgraffiato-­glazed ceramics were iconographically and stylistically integrated within reperotoires that unified material forms (architecture, textiles, jewelry) and spaces of consumption (understood as generally “elite” if not explicitly royal) across the Near East and Mediterranean.63 Red-­slipped red wares and plain wares have not previously supported analysis of social action in medieval Armenia, outside of oppositional models of rural-­vs.-­urban and low-­vs.-­high-­class consumption.64 Yet these objects, when approached as vessels for harissa, or stew, as opposed to the bearers of heraldic or religious signs, can inform on the relationships between the local and the global as mediated by shared meals. As I have explored in recent and ongoing research, the assemblage of red-ware cooking pots and footed bowls at the caravanserai strongly suggests the provisioning of travelers with food raised and prepared locally, which indicates that medieval Armenian society participated in an institution of hospitality that emphasized feeding and housing strangers, and that tied together the Armenian polities with their Muslim neighbors and fellow travelers.65 Scholarly discussion on the material politics of hospitality has argued that at the core of “commensal politics” is the power of food to tether memory, power, place, and time together into a co-­created world of shared power relations.66 The commensal politics of hospitality were provided at inns built and provisioned by the medieval Armenian princes; travelers were literally roofed and walled by local power. Such hospitality focused on collapsing space and time, as enabled by commensality—­providing meals for travelers to share (as is indicated by the forms in the ceramic assemblage) and inviting them and their particular projects (trade, pilgrimage, exploration) into a shared world centered in Armenia.The caravan inn as an institution—­both a built space and 212

K AT H RY N J. F R A N K LI N

a locus of practices—­was a form of place-­making that produced a clear sense of political ecumenism centered in Armenia but overlapping and encompassing wider realms. DISCUSSION: FRONTIER FOR WHOM?

The intent of this chapter has been to engage with the case of medieval Armenia as a way of thinking past the concept of frontiers in the medieval period as a question of material versus imaginary. By conceiving of the frontier as part of the imaginary space produced in the larger project of making social worlds for political action, the question to ask about the medieval period becomes not, “How was it a frontier?” but, “Frontier for whom?” Within this chapter I have examined a few of the ways in which Armenia was constituted as a frontier by and for the Islamic world through the writing of history and geography. I argued that the production of imagined worlds—­whether as narratives of conquest or geographies of foreign tastes—­was an inherently political project that situated Armenia as a periphery in relation to urban Islamic centers. I then shifted the field of vision to the ways in which social actors situated in Armenia produced their own imagined world. These Armenian-­ created worlds were sometimes very concerned with edges, as in the case of Christian orthodoxy and native Armenian law. Sometimes, however—­and in the medieval period in particular—­Armenians strove for the production of centrality. The Vačut’yans of the Kasakh Valley and other išxan’ner worked hard to locate themselves at the center of lineages in close political communion with their patron lords: the Georgians, the Ilkhanids.67 At the same time, they performed authority for their subjects in a way that was simultaneously very particularly Armenian but that also shared a spatial syntax with broader Anatolian traditions. What the discussion of the caravan inn in Armenia brings to the broader discussion of the Islamic frontier in the medieval period is a sense that in Armenia, at least during this period, actors like Vače Vačut’yan, Tigran Honenc’, or Česar Orbelyan endeavored to build social worlds that centered on their own projects but that also contained the world of the highway and the tastes and traders of distant cities. The road inn as a medieval institution (a structure and a repeated practice of hospitality and patronage), and especially as located in medieval Armenia, collapsed distinctions between frontier and metropolis, center and periphery, into a microcosm of worldly and pious authority.68 Further research on the social history of medieval Armenia will continue to explore the degree to which the regions of Islam, western M aking Worlds at the Edge of Everywhere

213

Christianity, and East Asia were constituted as peripheries to the Armenian political worldview, as was certainly the case. The case study of the medieval Armenian frontier constitutes one of this volume’s several case studies concerning the ways that frontiers must be imagined before they can be constructed—­both in the past and in modern analysis. As crucial though it is to emphasize that the categories of center and frontier were both imaginary and politically real for actors in medieval Armenia and the Islamic world,69 there are also ongoing ramifications for considering (or failing to consider) frontiers as simultaneously imagined and political in the contemporary context. “The world of Islam” continues to represent an ecumene whose frontiers are simultaneously hard drawn and ephemeral, depending on whether the scale and locale of analysis is the limit of juridical practice, border enforcement of sovereign states, or recruitment in online forums. In this sense, scholarship of the medieval Near East can bring its insights to bear on persistent discourses that have been shaped by the same “sociospatiotemporal collapsing” that, as I discussed above, has run through medieval scholarship from its foundations, and that continues to manifest in the way that Islamic culture and the medieval period (itself cribbed as “Islamic”) are discussed. To give an extreme example, after the January 7, 2015 shooting of staff from the French satire magazine Charlie Hebdo, the Western media responded in a number of ways, but a few comments illustrate the persisting power of boundaries that are both imagined and real. In a published public response to the shootings, author Salman Rushdie dismissed fundamentalist terror as “a mediaeval form of unreason,”70 thus situating the values and projects of the people he discussed on the other side of a temporal frontier, in a past time of barbarism and ruins. While in no way diminishing the atrocity of the terrorist attack, I argue that the terminology used to cope with it reveals that the same real/imagined frontiers that cut through medieval histories—­between barbarism and civilization, ancient and modern, past and present, foreign and familiar—­haunt the political present. The comments following the Charlie Hebdo attack are part of a pattern of commentary reacting to violence undertaken by Islamic agents, further exemplified by the discourses surrounding the actions of the Islamic State of Iraq and the Levant (ISIL), the unrecognized extremist Islamist state occupying shifting territories in Iraq, Syria, and Lebanon at the time of writing. In October, 2015, one of the US Republican presidential candidates, Carly Fiorina (soon to be chosen as Ted Cruz’s vice-presidential running mate), declared that “what ISIL wants to do is drive us back to the Middle Ages, literally.”71 Similar comments have followed not only the violent acts on humans carried out by ISIL but 214

K AT H RY N J. F R A N K LI N

also their acts of “performative destruction” directed at archaeological objects, sites, and built heritage.72 Responses to such commentary point out that to call the tech-­savy and profoundly modern ISIL medieval is a political move, casting them across a temporal frontier into the medieval and by so doing defining them as primitive and thus ontologically incapable of state sovereignty.73 But the dangers of rhetorically fighting ISIL on these grounds are obvious, as drawing links between Islam, the temporal space of the medieval, and a condition of primitive savagery enables a culture of dehumanization in the West that cannot be limited only to a single extremist group but that produces frontiers of the imagination as effective as any that separated the medieval actors discussed in this chapter. Switching the question asked—­not, “What kind of frontier?” but, “A frontier for whose world?”—­has relevance not only for Near Eastern history and archaeology as disciplines, but also for the world those disciplines continue to shape. NOTES

I would like to thank Asa Eger for the initial invitation to share my work at the 2012 ASOR Meetings in Chicago, as well as for the opportunity to publish my developing ideas on medieval Armenia as frontier(s). The original research, which provided data for this chapter, was supported by a Dissertation Fieldwork grant from the Wenner Gren Foundation, and was carried out in 2011 in collaboration with Project ArAGATS. In the final stages, I owe a debt of thanks to Elizabeth Fagan and Hannah Chazin for reading and edits, and to Josh Cannon and Gwendolyn Kristy for help with Arabic translation. My profound thanks to the anonymous reviewers for refinements and critical redirections, as well as for interventions in my thinking that I will continue to work into future discussions. All missteps are my own. 1. A. Eger (chapter 1), this volume. 2. Yildiz 2010, 91–­122. 3. Al-­Muqaddasī 1994. 4. Marr 1934; Hovannisian 1997. 5. Jackson and Morgan 2009; Polo 1904. 6. Indeed, an artifact of contemporary geopolitics altogether? 7. Lightfoot and Martinez 1993, 487. 8. Franklin, 2014a, 20–­39. 9. Shapira 1999, 143. 10. Toumanoff 1963; Ter-­Ghevondyan 1976. 11. Franklin 2014a, chapter 4; Bedrosian 1997, 241–­272.

M aking Worlds at the Edge of Everywhere

215

12. Ter-­Ghevondyan 1976; Garsoian 1984–­1985, 16–­17. 13. See for instance, Watson 2006; Lane 1947; Kalantaryan et al. 2009. But contrast this with the centrist perspective presented in F. Babayan 1981. 14. Power 1999, 1–­13; Eger 2014; Lightfoot and Martinez 1993. 15. Curta 2010, 1–­12. 16. Turner 1893. 17. A few sources: Morrison 2007, 81–­99; Bauer 2011, 83–­113; Gosden 2001, 242–­261. Rosenzweig 2014; Schama 1996; Scott 1998; Smith 2003. 18. Died 279/892. In Hitti 1916. 19. Ibid., 10. 20. For a more extensive discussion of these physical walls and the ways they may have mapped onto imagined political and social boundaries, see Alizadeh (chapter 7), this volume. 21. Hitti 1916, 309. 22. Eger 2014, 2–­3; Hitti 1916, 210. 23. Turner 1893, 1. 24. In contrast to some applications of the term in archaeology, and specifically following the work of Benedict Anderson, I define imaginary not in the sense of a figment or “the opposite of real” but as existing in the realm of affect, image, discourse, dream, and aspiration—­in other words, as the socially real. 25. Smith 2003; Monroe 2012, 16. 26. Biddick 1998, 58–­82. 27. Davis 2008; Wells 2008. 28. Manzano-­Moreno 2004, 34; see also the increasing elision of medieval for barbaric in popular discourse, giving even greater pause to this and other academics when Islamic and medieval are increasingly used interchangeably within Western scholarly circles. 29. Lewis 1958, 475–­483; Thompson 1913, 490–­504. 30. Brauer 1995, 9–­11. 31. Ibn Ḥawqal 1800, 132–­134; al-­Muqaddasī 1994. 32. Stahl 2002, 827–­845; Golden 1991, 69. 33. Golden 1991, 70; Crone 1980. 34. Wheatley 2000. 35. For a further discussion of the landscape of luxuries by which Arab writers imagined the Caucasus during the medieval period, see Vorderstrasse (chapter 8), this volume. 36. Long-­fibered silk produced by killing the silkworm before it hatches, thus preserving the cocoon unbroken. See Serjeant 1972, 198b. 37. Ibn Ḥawqal 1800, 160–­161. 216

K AT H RY N J. F R A N K LI N

38. Translation from al-­Muqaddasi 1906, 373, in al-­Muqaddasī 1994, 332. 39. al-­Muqaddasī 1994, 333. 40. Stein 1999; A. Eger 2014, 410; Kohl 2009, 13–­24; Kardulias 2010, 53–­80. 41. Redgate 2007, 281–­306; Adalian 2010, 286. Also: Andrews 2013, 29–­42. 42. Glick et al. 2014, 189. 43. Hewsen 1992, 46. 44. Ibid., 59. 45. Ibid., 43. 46. Gošec’i 2000. 47. Manandian 1979; Babayan 1976, 540–­557. 48. Thomson 1997, 119–­127. 49. Gošec’i 2000, 71. 50. Lauricella (chapter 2) and Holod and Kahlaoui (chapter 3), this volume. 51. Kleiss 2001. 52. Hillenbrand 1994, 332–­333. 53. Franklin 2014a, chapter 3; Cytryn-­Silverman 2010; Yavuz 1997, 80–­95. 54. Gošec’i 2000, 114, 295. 55. Mahe 2001, 1319–­1342; Arakelyan 1958; Franklin 2016, 131–­146. 56. Önge 2007, 49–­69. 57. Šahkatunyan 1842, 168–­ 169; Harut’yunyan 1960; Petrosyanc 1988; Franklin 2014b 1–­21. 58. Ghazarian and Ousterhout 2001, 141–­154. 59. Yavuz 1997; Ertuğ and Bragner 1991; Erdmann 1961. 60. Franklin 2014a, chapters 3–­4; Kiesling 2001, 73. 61. Jones 2001–­2002, 341–­398. 62. These connections are currently under further investigation but the glazed wares found at Arai-­Bazarjuł were most likely produced at a city or town in the relatively nearby highlands, such as Dvin or Tbilisi: see Franklin 2014a; F. Babayan 1981. 63. Redford 2005; 283–­310; Hoffman 2001, 17–­50; Grabar 2006. 64. See the model formulated in Łafadaryan 1952. 65. For a discussion of food and hospitality at endowed institutions in medieval Anatolia, see Trépanier 2014; for further discussion of provisioning, ceramics, and hospitality at the Arai-­Bazarjuł caravanserai see Franklin 2014a; Franklin et al. 2017. 66. Michael 2001, 65–­114; Munn 1992. 67. Ališan 1881, 164–­165. 68. Goshgarian 2007. 69. See also Straughn 2006. 70. Rushdie 2015. 71. Holsinger 2015; Wood 2015. M aking Worlds at the Edge of Everywhere

217

72. Shaheen 2015. Performative destruction is a term developed by the ASOR Cultural Heritage Initiative; see ASOR CHI reports on the destruction of sites and monuments in Syria and Iraq: e.g., Danti et al. 2015. 73. Terry 2015; Tharoor 2015; Perry 2016. REFERENCES

Adalian, Rouben. 2010. Historical Dictionary of Armenia. Lanham, MD: Scarecrow Press. Ališan, Łevond. 1881. Širak: tełagrut’yun patkeracoy’c. Venice: I Mxit’aray Vans I.S. Łazar. Andrews, Tara. 2013. “Identity, Philosophy and the Problem of Armenian History in the Sixth Century.” In History and Identity in the Late Antique Near East, edited by Philip Wood, 29–­42. Oxford Studies in Late Antiquity. New York: Oxford University Press. Arakelyan, Babken. 1958. Kałak’nerə yev arhestnerə Hayastanum 6–­13-­yerrord darerum. Hator I. Yerevan: Haykakan SSR GA Hratarakč’ut’yun. Babayan, Frina. 1981. Mijnadaryan Hayastani gełarvestakan xetsełeni zardadzeverə. Haykakan SSR Gitut’yunneri Akademiya Hnagitutyan yev Azgagrutyan Institut. Yerevan: Haykakan SSH GA Hratarakčut’yun. Babayan, L. H. 1976. “The Development of Zakarid Power and New Armenian Feudal Houses.” In Hay Žołovrdi Patmut’yun. Prak VIII, edited by Aghayan et al., 540–­557. Yerevan: Haykakan SSH Gitut’yunneri Akademiayi Hratarakč’ut’yun. Bauer, Andrew. 2011. “Producing the Political Landscape: Monuments, Labor, Water and Place in Iron Age Karnataka.” In The Archaeology of Politics: The Materiality of Political Practice and Action in the Past, edited by P. Johansen and A. Bauer, 83–­113. Newcastle upon Tyne: Cambridge Scholars Press. Bedrosian, Robert. 1997. “Armenia during the Seljuk and Mongol Periods.” In Armenian People from Ancient to Modern Times, Volume 1, edited by Richard Hovannisian, 241–­272. New York: St. Martin’s Press. Biddick, Kathleen. 1998. “English America: Worth Dying For?” In The Shock of Medievalism. 58–­82. Durham, NC: Duke University Press. Brauer, Ralph W. 1995. “Boundaries and Frontiers in Medieval Muslim Geography.” Transactions of the American Philosophical Society, New Series 85.6: 1–­73. Crone, Patricia. 1980. Slaves on Horses: The Evolution of the Islamic Polity. Cambridge, UK: Cambridge University Press. Curta, Florin. 2010. “Introduction.” In Borders, Barriers and Ethnogenesis: Frontiers in Late Antiquity and the Middle Ages, edited by F. Curta, 1–­12. Studies in the Early Middle Ages Series. Turnhout, Belgium: Brepols Publishers. 218

K AT H RY N J. F R A N K LI N

Cytryn-­Silverman, Katia. 2010. The Road Inns (Khans) in Bilad al-­Sham. BAR International Series 2130. Oxford, UK: British Archaeological Reports. Danti, Michael, Jesse Casana, Tate Paulette, Kathryn Franklin, and Cheikhmous Ali. Cultural Heritage Intiatives (CHI): Planning for Safeguarding Heritage Sites in Syria and Iraq, Weekly Report 25 ( January 26, 2015). Last Accessed November 6, 2016. http://www.asor-syrianheritage.org/syrian-heritage-initiative-weekly-report-25 -january-26-2015/. Davis, Kathleen. 2008. Periodization and Sovereignty: How Ideas of Feudalization and Secularization Govern the Politics of Time. The Middle Ages Series. Philadelphia: University of Pennsylvania Press. Dietler, Michael. 2001. “Theorizing the Feast: Rituals of Consumption, Commensal Politics, and Power in African Contexts.” In Feasts: Archaeological and Ethnographic Perspectives on Food, Politics, and Power, edited by M. Dietler and B. Hayden, 65–­114. Washington, DC: Smithsonian Institution Press. Eger, Asa. 2014. The Spaces between the Teeth: Interaction and Exchange among Muslim and Christian Communities. London: I.B.Tauris. Erdmann, Kurt. 1961. Das Anatolische karavansaray des 13. Jahrhunderts. Teil 1, 2. Berlin: Verlag Gebr. Mann. Ertuğ, Ahmet, and Mine Timur Bragner.1991. The Seljuks: A Journey through Anatolian Architecture. Istanbul: Ahmet Ertug. Franklin, Kathryn. 2014a. “‘This World Is an Inn’: Cosmopolitanism and Caravan Trade in Late Medieval Armenia.” PhD diss., Department of Anthropology, University of Chicago. Franklin, Kathryn. 2014b. “A House for Trade, a Space for Politics: Excavations at the Arai-­Bazarjugh Caravanatun; Aragatsotn, Armenia.” Anatolica XL: 1–­21. Franklin, Kathryn. 2016. “Assembling Subjects: Cosmopolitanism in Late Medieval Armenia.” In Incomplete Archaeologies: Assembling Knowledge in the Past and Present, edited by Emily Miller Bonney, K. Franklin, and J. Johnson, 131–­146. Oxford, UK: Oxbow. Franklin, Kathryn, Tasha Vorderstrasse, and Frina Babayan. 2017. “Examining the Late Medieval Village from the Case at Ambroyi, Armenia.” Journal of Near Eastern Studies 76.1. Garsoian, Nina G. 1984–­1985. “The Early-­Mediaeval Armenian City: An Alien Element?” JANES 16–­17: 67–­83. Ghazarian, Armen, and Robert Ousterhout. 2001. “A Muqarnas Drawing from Thirteenth-­Century Armenia and the Use of Architectural Drawings during the Middle Ages.” Muqarnas: An Annual on the Visual Culture of the Islamic World, XVIII: 141–­154.

M aking Worlds at the Edge of Everywhere

219

Glick, Thomas, S. Livesey, and F. Wallis, eds. 2014. Medieval Science, Technology and Medicine: An Encyclopedia. London: Routledge. Goshgarian, Rachel. 2007. “Beyond the Social and the Spiritual: Redefining the Urban Confraternities of Late Medieval Anatolia.” PhD diss., Department of History, Harvard University. Golden, Peter. 1991. “Aspects of the Nomadic Factor in the Economic Development of Kievan Rus.” In Ukrainian Economic History: Interpretive Essays, edited by I. S. Koropeskyj. Edmonton, AB: Canadian Institute of Ukrainian Studies. Gosden, Chris. 2001. “Postcolonial Archaeology: Issues of Culture, Identity and Knowledge.” In Archaeological Theory Today, edited by Ian Hodder, 242–­261. Cambridge, UK: Polity Press. Gošec’i, Mxitar. 2000. The Lawcode (Datastanagirk’) of Mxit’ar Goš. Translated with commentary and indices by Robert W. Thomson. Amsterdam: Rodopi. Grabar, Oleg. 2006. “The Shared Culture of Objects.” In Islamic Visual Culture, 1100–­1800, volume II, Constructing the Study of Islamic Art, 51–­68. Hampshire, UK: Ashgate Publishing Limited. Harut’yunyan, Vrazdat. 1960. Karavanat’nerə yev kamurjnerə mijnadar Hayastanum. Yerevan: HayPetHrat. Hewsen, Robert. 1992. The Geography of Ananias of Šhirak (Ašxarhac’oyc’): The Long and Short Recensions. Wiesbaden: Dr. Ludwig Reichert Verlag. Hillenbrand, Robert. 1994. Islamic Architecture: Form, Function, and Meaning. New York: Columbia University Press. Hitti, Phillip K. 1916. The Origins of the Islamic State: Being a Translation from the Arabic, Accompanied with Annotations Geographic and Historic Notes of the Kitab Futuh al-­Buldan of al-­Imam abu-­l ‘Abbas Ahmad ibn Jabir al-­Balādhurī. London: Longmans, Green. Hoffman, Eva R. 2001. “Pathways of Portability: Islamic and Christian Interchange from the Tenth to the Twelfth Century.” Art History 24.1: 17–­50. Holsinger, Bruce. 2015. “Carly Fiorina Goes Medieval,” New York Times, October 8, 2015. Last accessed November 6, 2016. https://www.nytimes.com/2015/10/08 /opinion/carly-fiorina-goes-medieval.html. Hovannisian, Richard. 1997. The Armenian People from Ancient to Modern Times, Volume II: The Dynastic Periods; from Antiquity to the Fourteenth Century. New York: St. Martin’s Press. Ibn Ḥawqal, Muḥammad. [1800]. Ṣūrat al-­’Arḍ, or the Oriental Geography. Translated by W. Ouseley. London: Wilson and Co. Oriental Press. Jackson, Peter, and David Morgan. 2009. The Mission of Friar William of Rubruck: His Journey to the Court of the Great Khan Mongke, 1252–­1255. Indianapolis, IN: Hacket Publishing Ltd. 220

K AT H RY N J. F R A N K LI N

Jones, Lynn. 2001–­2002. “The Visual Expression of Bagratuni Rulership: Ceremonial and Portraiture.” Revue des Études Arméniennes 28: 341–­398. Kalantaryan, A, G. Karakhanyan, H. Melkonyan, H. Petrosyan, N. Hakobyan, F. Babayan, A. Zhamkochyan, K. Nawasardyah, and A. Hayrapetyan. 2009. Armenia in the Cultural Context of East and West: Ceramics and Glass (4th–­14th Centuries). Joint Research Project of the Swiss National Science Foundation. Republic of Armenia National Academy of Sciences Institute of Archaeology and Ethnography. Yerevan, Armenia: NAS RA Gitutyun Hratarakchutyun. Kardulias, P. Nick. 2010. “World-­Systems Applications for Understanding the Bronze Age in the Eastern Mediterranean.” In Archaic State Interaction: The Eastern Mediterranean in the Bronze Age, edited by W. Parkinson and M. Galaty, 53–­80. Santa Fe, NM: SAR Press. Kiesling, Brady. 2001. Rediscovering Armenia: An Archaeological/Touristic Gazetteer and Map Set for the Historical Monuments of Armenia, edited by Raffi Kojian. Yerevan: Tigran Mets. Kleiss, Wolfram. 2001. Karawanenbauten in Iran. Vols. 1–­6. Deutsches Archaologisches Institut Eurasien-­Abteilung Aussenstelle. Berlin: Reimer. Kohl, Philip. 2009. “The Ancient Economy, Transferable Technologies and the Bronze Age World-­System: A View from the North-­Eastern Frontier of the Ancient Near East.” In Centre and Periphery in the Ancient World, New Directions in Archaeology, edited by M. Rowlands, M. Larsen, and K. Kristiansen, 13–­24. Cambridge, UK: Cambridge University Press. Łafadaryan, K. G. 1952. Dvin kałakə ev nra pełumnerə. Volume I. Archaeological Expeditions 1937–­1950. Yerevan: Haykakan SSH Academy of Sciences. Lane, Arthur. 1947. Early Islamic Pottery: Mesopotamia, Egypt and Persia. London: Faber and Faber. Lewis, Archibald. 1958. “The Closing of the Medieval Frontier 1250–­1350.” Speculum 33.4: 475–­483. Lightfoot, Kent, and Antoinette Martinez. 1993. “Frontiers and Boundaries in Archaeological Perspective.” Annual Review of Anthropology 24: 471–­492. Mahe, Jean-­Pierre. 2001. “ ‘Le testament de Tigran Honenc:’ la fortune d’un marchand armenien d’Ani aux XIIe–XIIIe siècles.” Comptes rendus des séances de l’Academie des Inscriptions et Belles-­Lettres, 145e année, N.3.1319–­1342. Manandian, Hakob. 1979. Yerker III Seljukyan šrjanic minjev sefyanneri hastatumə Iranum. Haykakan SSH Gitut’yunneri Akademiya, Patmutyan Institut. Yerevan: Haykakan SSH Hratarakč’ut’yun. Manzano-­Moreno, Eduardo. 2004. “The Creation of a Medieval Frontier: Islam and Christianity in the Iberian Peninsula, Eighth to Eleventh Centuries.” In Frontiers

M aking Worlds at the Edge of Everywhere

221

in Question: Eurasian Borderlands 700–­1700, edited by D. Power and N. Standen, 32–­54. New York: St. Martin’s Press. Marr, Nicolai. 1934. Ani: knizhnaya istoriya goroda i raskopki na meste gorodischa. Gosudarstvennaya Akademiya Istorii Materiyalnoi Kulturye. Moscow: Gosudarstvennoye Sotsialno-­Aekonomicheskoye Izdatelstvo. Monroe, J. Cameron. 2012 “Power and Landscape in Atlantic West Africa.” In Power and Landscape in Atlantic West Africa: Archaeological Perspectives, edited by J. C. Monroe and A. Ogundiran. Cambridge, UK: Cambridge University Press. Morrison, Kathleen. 2007. “Making Places and Making States: Agriculture, Metallurgy, and the Wealth of Nature in South India.” In The Wealth of Nature: How Natural Resources Have Shaped Asian History, 1600–­2000, edited by P. Boomgaard and G. Bankoff, 81–­99. Basingstoke, UK: Palgrave MacMillan. Munn, Nancy. 1990. Constructing regional worlds in experience: Kula exchange, witchcraft and Gawan local events. Man New Series, Vol. 25 (1). 1–17. Munn, Nancy D. 1992. The Fame of Gawa: A Symbolic Study of Value Transformation in a Massim (Papua New Guinea) Society. Durham, NC: Duke University Press. al-­Muqaddasi, Muḥammad. 1906. Kitāb aḥsan al-­taqāsīm f ī ma‘rifat al-­āqālīm. Transcribed and annotated by M. J. de Goeje. Leiden: Brill. al-­Muqaddasī, Muḥammad. 1994. The Best Divisions for Knowledge of the Regions. Translated by Basil Anthony Collins. Centre for Muslim Contribution to Civilization. Reading, UK: Garnet Publishing. Mönge, Mustafa. 2007. “Caravanserais as Symbols of Power in Seljuk Anatolia.” In Power and Culture: Identity, Ideology, Representation, edited by J. Osmond and A. Cimdiņa, 49–­69. Pisa: Pisa University Press. Perry, David. 2016. “This Is Not the Crusades: There Is Nothing Medieval about ISIS,” CNN​.com, October 26, 2016. Last accessed November 6, 2016, http://​www​ .cnn​.com/​2016/​10/​16/​opinions/​nothing​-medieval​-about​-isis​-perry/. Petrosyanc, V. M. 1988.Nig-­Aparani Patmačartarapetakan Hušardzannere. Haykakan SSR Patmut’yun yev kulturayi hušardzanneri pahpanut’yan enkerut’yun. Yerevan: “Hayastan.” Polo, Marco. 1904. The Most Noble and Famous Travels of Marco Polo, One of the Nobility of the State of Venice, into the East Parts of the World, as Armenia, Persia, Arabia, Tartary, with Many Other Kingdoms and Provinces. Marsden Translation, edited by Thomas Wright. New York: C. Scribner. Power, Daniel. 1999. “Frontiers: Terms, Concepts, and the Historians of Medieval and Early Modern Europe.” In Frontiers in Question: Eurasian Borderlands 700–­1700, edited by D. Power and N. Standen. New York: St. Martin’s Press. Redford, S. 2005. “A Grammar of Rum Seljuk Ornament.” Mesogeios 25–­26: 283–­310.

222

K AT H RY N J. F R A N K LI N

Redgate, Elizabeth. 2007. “Myth and Reality: Armenian Identity in the Early Middle Ages.” National Identities 9.4: 281–­306. Rosenzweig, Melissa. 2014. “Imperial Environments: The Politics of Agricultural Practice at Ziyaret Tepe, Turkey in the First Millenium BCE.” PhD diss., Department of Anthropology, University of Chicago. Rushdie, Salman. 2015. “Salman Rushdie: ‘I Stand With Charlie Hebdo, as We All Must’” Wall Street Journal, January 7, 2015. Last accessed August 17, 2016. http://​on​ .wsj​.com/​1xRFMt0. Šahkatunyan, Yovhannu Episciposi. 1842. Storagrut’yun katołike Ejmiadzni yev hing gavarac Araratay. Hator yerkrord. Yerevan: Ejmiadzin. Schama, Simon. 1996. Landscape and Memory. New York: Vintage. Scott, James. 1998. Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed. New Haven, CT: Yale University Press. Serjeant, R. B. 1972. Islamic Textiles. Beirut: Librairie du Liban. Shaheen, Kareem. 2015. “ISIS Attacks on Ancient Sites Erasing History of Humanity, Says Iraq.” The Guardian, March 9, 2015. Last accessed November 6, 2016. https://​www​.theguardian​.com/​world/​2015/​mar/​09/​iraq​-condemns​-isis​-destruction​ -ancient​-sites. Shapira, Dan. 1999. “Pahlavi References to Armenia.” Iran and the Caucasus 3/4 (1999/2000): 143–­146. Smith, Adam T. 2003. The Political Landscape: Constellations of Authority in Early Complex Polities. Berkeley: University of California Press. Stahl, Ann B. 2002. “Colonial Entanglements and the Practices of Taste: An Alternative to Logocentric Approaches.” American Anthropologist 104.3: 827–­845. Stein, Gil. 1999. Rethinking World Systems. Tuscon: University of Arizona Press. Straughn, Ian. 2006. “Materializing Islam: An Archaeology of Landscape in Early Islamic Period Syria (c. 600–­1000 C.E.).” Unpublished PhD diss., Department of Anthropology, University of Chicago. Ter-­Ghevondyan, Aram. 1976. The Arab Emirates in Bagratid Armenia. Translated by Nina S. Garsoian. Lisbon: Calouste Gulbenkian Foundation. Terry, John. 2015. “Why ISIS Isn’t Medieval.” Slate, February 9, 2015. https://slate.com /news-and-politics/2015/02/isis-isnt-medieval-its-revisionist-history-only-claims -to-be-rooted-in-early-arab-conquests.html. Tharoor, Ishaan. 2015. “The Misguided Debate about How Islamic the Islamic State Is.” The Washington Post, February 20, 2015. Last accessed Nov. 6, 2016. https://​www​ .washingtonpost​.com/​news/​worldviews/​wp/​2015/​02/​20/​the​-misguided​-debate​ -about​-how​-islamic​-the​-islamic​-state​-is/​?tid​=​a​_inl.

M aking Worlds at the Edge of Everywhere

223

Thompson, James Westfal. 1913. “Profitable Fields of Investigation in Medieval History.” The American Historical Review 18.3: 490–­504. Thomson, Robert. 1997. “Mxit’ar Goš and His Law Code.” In Armenian Perspectives: 10th Anniversary Conference of the Association Internationale des Études Arméniennes, School of Oriental and African Studies, 119–­127. London: Psychology Press. Toumanoff, Cyril. 1963. Studies in Christian Caucasian History. Georgetown, DC: Georgetown University Press. Trépanier, Nicholas. 2014. Foodways and Daily Life in Medieval Anatolia. Austin: University of Texas Press. Turner, Frederick J. 1893. “The Significance of the Frontier in American History.” Last accessed August 17, 2016. https://​en​.wikisource​.org/​wiki/​The​_Frontier​_in​ _American​_History. Watson, Oliver. 2006. Ceramics from Islamic Lands. New York: Thames and Hudson. Wells, Peter. 2008. Barbarians to Angels: The Dark Ages Reconsidered. New York: Norton. Wheatley, Paul. 2000. The Places Where Men Pray Together: Cities in Islamic Lands, Seventh through the Tenth Centuries. Chicago: University of Chicago Press. Wood, Graeme. 2015. “What ISIS Really Wants.” The Atlantic. Last accessed November 6, 2016. http://​www​.theatlantic​.com/​magazine/​archive/​2015/​03/​what​-­­isis​-­­really​ -­­wants/​384980/. Yavuz, Aysil Tükel. 1997. “The Concepts that Shape Anatolian Seljuk Caravanserais.” Muqarnas 14: 80–­95. Yildiz, Sara N. 2010. “Reconceptualizing the Seljuk-­Cilician Frontier: Armenians, Latins and Turks in Conflict and Alliance during the Early Thirteenth Century.” In Borders, Barriers and Ethnogenesis: Frontiers in Late Antiquity and the Middle Ages, edited by Florin Curta, Studies in the Early Middle Ages Series, 91–­122. Turnhout, Belgium: Brepols Publishers.

224

K AT H RY N J. F R A N K LI N

About the Authors

Karim Alizadeh is Visiting Professor of Anthropologyat Grand Valley State University, Michigan. His research focuses on two transitioning periods in the history of the Caucasus region and Iran: (1) the development of social complexity in late prehistory; and (2) the Sasanian imperial era to medieval Islamic society. His research includes social complexity, social boundaries and border studies, early urbanism, and societal collapse. His excavations and surveys in Iranian Azerbaijan trace the development of complex societies in the Caucasus and the collapse of the Sasanian Empire in Iran. Karim Alizadeh received his BA and MA from the University of Tehran (Archaeology) and PhD from Harvard University (Anthropology). A. Asa Eger is Associate Professor at the University of North Carolina at Greensboro in the Department of History, and faculty member in the Archaeology Program. He has surveyed and excavated in Turkey, Israel, Cyprus, and Greece. Two recent books focus on the Islamic-­Byzantine frontier from the sixth to the twelfth centuries: The Islamic-­ Byzantine Frontier: Interaction and Exchange Among Muslim and Christian Communities (2015) and The Spaces between the Teeth: A Gazetteer of Towns on the Islamic-­Byzantine Frontier (second edition, 2016). Currently he is publishing three final reports: (1) the excavation that he directed at Tüpraş Field/ Ḥiṣn al-­Tīnāt, on coastal Turkey; (2) Antioch’s sector 17-­O, an unpublished area of the city excavated by Princeton in the 1930s, and reanalyzed by a small team of scholars; and (3) the Nahr Qoueiq Survey, a reanalysis of the Islamic settlement

and material culture of the 1970s survey in collaboration with Tasha Vorderstrasse. In addition, he is coauthoring a volume on the complete history of Antioch with Andrea de Giorgi. He is co-­editor of the new series Monographs in Islamic Archaeology (Equinox Publishing) and serves on the editorial board of the Journal of Islamic Archaeology. Jana Eger studied Egyptology and prehistoric archaeology in Heidelberg and Berlin, and graduated in 2013 with a thesis on remote sensing and reconstruction of ancient trade routes in the Sudanese Western Desert. After having worked as archaeologist and GIS analyst at several excavations and survey projects in Sudan, Egypt, Turkey, Mongolia, and Germany, she worked from 2013 until 2018 as a research employee at the “Forschungsstelle Alter Sudan” at Münster University within the field project “Wadi Abu Dom Itinerary” in the Bayuda Desert (Sudan). Since 2018, she has been Field Director of Münster University’s field project “Interregional Linkage Investigations in Northern Kordofan (InterLINK),” and investigates the function of the historical landscape of Kordofan (western Sudan) as a frontier and contact zone between the ancient and medieval Nile cultures and central Africa. Kathryn J. Franklin is an archaeologist of medieval Armenia and the Caucasus, where she has been working on collaborative projects in the Republic of Armenia for a decade, exploring the ways that local politics and Silk Road culture were tangled together in landscape and space-­time. She is currently the co-­PI of the Vayots Dzor Silk Road Survey, a multiscalar project exploring Silk Road spaces in southwestern Armenia. Her work as a medievalist centers on experiences of medieval travel, intimacies of medieval embodiment, and the profound and mundane practices of medieval and early modern hospitality and cosmopolitanism. She is currently appointed as a lecturer in medieval history at Birkbeck, University of London, where her research is concerned with world-­ making as a locus of politics, material culture as a mediator of spatiotemporal distances, and the interpenetration of literary and “real” landscapes in archaeological work. Renata Holod is College of Women Class of 1963 Term Professor in the Humanities, History of Art Department, and curator, Museum of Archaeology and Anthropology, both at the University of Pennsylvania. She has carried out archaeological and architectural fieldwork in Syria, Iran, Morocco, Turkey, Central Asia, Tunisia, and Ukraine. She was convenor, steering-committee member, and chair, Master Jury for the Aga Khan Award for Architecture; and was Director, Assessment Board, Aga Khan Trust for Culture, 2012–­2013. She is coauthor or coeditor of City in the Desert (Harvard 1978), Architecture and Community: Building in the Islamic World Today (Aperture 1983), The Mosque and the Modern World (Thames and Hudson 1997), The City in the Islamic World (Brill 2008), and An Island through Time: Jerba Studies ( Journal of Roman Archaeology 226

A bout the Authors

2009). On several advisory boards, she has also served as President, Board of Trustees, at The Ukrainian Museum (2013–­2017). Tarek Kahlaoui, is Assistant Professor of History in the Mediterranean School of Business (Tunisia). He focuses on Islamic cartography, archaeology, and the premodern history of North Africa. He took part in the Jerba Project and edited its second volume with Renata Holod. He recently published Creating the Mediterranean (Brill 2017) on the history of Islamic perceptions of the Mediterranean between the tenth and sixteenth centuries. Anthony J. Lauricella is a PhD candidate in Islamic archaeology at the University of Chicago. He has conducted fieldwork in Sicily, Palestine, Azerbaijan, and Turkey. Ian Randall received his PhD in archaeology from Brown University in 2018. An island archaeologist, he has worked in the Baltic, South Pacific, and Caribbean as well as the Mediterranean. His work focuses primarily on Cyprus and Crete in the transition from late antiquity to the early Middle Ages, with an emphasis on issues of connectivity, identity negotiation, and pottery consumption. His latest project examines the cultural impact of early Muslim seafaring and ports as heterotopic space. Giovanni R. Ruffini is Professor of Classical Studies in the Department of History at Fairfield University, Connecticut. He specializes in documentary papyrology and the social history of Greco-­Roman Egypt. His most recent publication in that field is Life in an Egyptian Village in Late Antiquity: Aphrodito Before and After the Islamic Conquest (Cambridge 2018). His publications in Nubian Studies include Medieval Nubia: A Social and Economic History (Oxford 2012) and The Bishop, The Eparch, and The King: Old Nubian Texts from Qasr Ibrim ( Journal of Juristic Papyrology Supplement 22, 2014). He is the founding coeditor of Dotawo: A Journal of Nubian Studies. Tasha Vorderstrasse is the University and Continuing Education Program Coordinator and Research Associate at the Oriental Institute of the University of Chicago. She received her PhD in Near Eastern Archaeology at the University of Chicago in 2004. Her work concentrates on the material culture of the Near East, Caucasus, and Central Asia, and the relationship between these regions and China.

A bout the Authors

227

Index

A Abbāsid, 7, 8, 11, 15, 16, 18, 86, 92, 94, 151, 152, 168, 172, 176, 198; Byzantine frontier, 47 Abū Sulaymān al-Tlātlī, 53 Acrocorinth, 90 Adamo de Citella, 55 Aegean, 17, 23, 81, 84–89, 92, 96 Aegina, 86, 88 Aghlabid, 42, 48, 65, 68, 70, 72, 74 Agios Epiphanios, 88 Agios Georgios, 84, 88 Ajīm, 49 al-Andalūs, 4, 7, 17–18, 22, 47, 82 Alamut, 18 Alan Gate, 148, 151, 152 Alans, 18, 168, 171–174, 180–182 Albania (Caucasian), 151, 174 Alborz, 18 Alexander (the Great)’s Gate/ Wall (Sadd-i Iskandar), 154, 159, 172 Alexandria, 89, 180 Alfonso V, 55 Algeria, 31, 32, 35, 42, 52 ‘Alī (caliph), 15, 16 Amazigh (dialect), 35 Amīnāj, 34 Amorgos, 23 Amu Darya River (Oxus), 157 Antioch, 18, 205 Arabia, 16, 203 Aragon, 55, 66 Arrān, 159, 172, 202 Aras River, 202 Araxes River, 151

Aruč, 208 Ashoka (Mauryan emperor), 180 Assyrian Empire, 142, 148, 166 Athens (thalassocracy), 83 ‘awāṣim, 8, 11 Ayyubid dynasty, 201 Azerbaijan, 148, 150, 151, 156, 172, 197 B Bacrtia, 180; Greco-Bactrian, 175 Baghdad, 7, 86, 91–93, 113, 202, 203 Bagratid, 197, 198, 206 Balanjar, 148, 151, 153, 159, 160 Banganarti, 109 Banī Maghzil, 60, 62 Banī Maʿqil, 53, 71 Banī Smūmin, 56 Banū Kanz, 13, 18, 114 Banū Yahrāsin, 64, 69 Baqt (Treaty), 112 Baṣra, 152, 154 Battle of Ṣiffīn, 15–16 Battle of the Masts, 89 Baylaqān, 151 Bayuda, 122, 129 Berber, 14, 31, 43, 52, 69 Besh-Barmak, 145, 148 Bir al-‘Ayn, 127, 129 Black Sea, 171, 173–175 Bolshaya Laba River, 176 Bulgars, 174 Buqṭūra, 34 Burj, 61, 65 Burj al-Kabīr, 55

Burj Ghazī Muṣṭafa, 55, 65, 67 Islamic-Byzantine/Byzantine-Islamic frontier, 3, 5, 7, 8, 11, 15, 32, 47, 70, 83, 84, 91 Byzantine, 8, 9, 12, 15, 17, 18, 31, 47, 48, 52, 64, 68, 72, 74, 81, 88–97, 106, 142, 150, 155, 159, 168, 169, 172–175, 177, 180, 181, 195, 197, 204, 211 C Cabezo de la Cobertera, 18 Cairo (al-Qāhirah), 13, 61, 72, 109, 113, 130, 131 Cappadocia, 17 caravanserai, 9, 10, 13, 19, 196, 197, 206, 208–212, 217 Caspian Gates. See Darband Caspian Sea, 144, 146, 160, 173; pass, 148; plain, 150 Catalan, 55, 67 Central Asia, 6 Česar Orbelyan, 210 Chalcedon (Council), 204 Chandax, 95 China, 170–179 Chios, 86, 88 Cilicia, 91, 198; Armenian, 195 Çirax-kala, 145 Constans II, 89 Constantia, 86, 88 Constantine of Antioch, 205 Constantine IV, 88 Constantine VII Porphyrogenitus, 89 Constantinople, 84, 86, 89, 91–93, 109, 176 Crete, 17, 18, 81, 82, 84–89, 91, 92, 94 Cyclades Islands, 23 Cyprus, 17, 72, 81, 82, 84–88, 91, 92, 94 D al-Dabba, 129; Nile segment, 131 Dabīl, 202 Dag-Bary. See Darband Dagestan, 146, 147 Danube River, 142 dār al-ḥarb, 80 dār al-Islām, 80, 94 Darband: settlement and wall, 8, 12, 144–155, 158–160; site, 139 Darfur, 120 Dariali, 148, 152, 153, 174 Dehibat, 35, 36 Delhi, 21

230

I ndex

Donatists, 17, 31, 32 Dongola, 107, 108, 114, 120, 130 Dongolawi-Kenzi, 114 Dargūth Bāshā, 54 Dvin, 199, 202–204, 217 E Egypt, 10–13, 35, 55, 84, 86–89, 91, 95, 105, 106, 107, 109–114, 119, 120, 128, 133, 177; Fusṭāṭ, 35 Egyptian-Nubian border, 9, 10, 20, 105, 106, 109 Ethiopia, 23 Euoptios, 87 Euphrates River, 142, 157 F Faṭimids, 12, 16, 54, 74, 110, 111, 133 Flavian Period. See Roman Frederic I, 55 France (French), 67, 109, 120, 214 Fursaṭā, 35, 38, 39 G Ganja, 148, 151, 159 Ganja-Gazakh Plain, 148 gasr, 36, 37, 41 Geniza (Cairo), 55, 71, 110, 174 Genoa, 55 Georgian, 158, 172 al-Ghazāl, 130, 131 Ghazī Muṣṭafā, 67 Gog, 20, 150, 154, 172 Great Wall of China, 13 Great Wall of Gurgān, 143, 159 Gulf of Bu Ghara, 48, 54 Gurgān Plain, 8, 143, 167 H Hadrian’s Wall, 143 al-Ḥakim, 61, 74, 110, 112 al-Hara, 41, 60, 61 Hafsid, 48, 55–57, 65–68 Hagia Sophia, 89 Hamuda Bey, 67 Hashāshiyūn, 18 Hellenistic, 180, 204; period, 128 Hepththalites, 6 Hindustān, 21

ḥiṣn, 65 Ḥumat Ghardāya, 52 Ḥumat al-Sūq, 52, 56, 67 Huns, 145, 147, 150 I Iberia, 14, 173 Iberian-Andalusī frontier, 3, 7 Idunat, 43 Ifrīqiya, 47, 48, 51, 54, 64, 65, 68, 70, 71 Imazighen, 31; Ibāḍīs, 35 India, 7, 16, 21, 169, 177, 178, 180, 201, 202 Iraklion, 95 Iran, 4, 6, 22, 150, 152, 157–159, 161, 169–172, 174, 197, 199, 201, 206 Īrānshahr (Iranzamin), 144, 157 Iraq, 18, 152, 157, 161, 214, 218 Islamic-Christian frontier. See IberianAndalusī frontier Israel, 157 Istanbul, 56, 88 Italy, 12, 55 J Jādū, 34–36, 39 Japan, 44, 171 Jrapi, 196, 208 Jerusalem, 14, 92, 93, 109, 204, 205 Jews/Jewish (also Judaism), 31, 32, 34, 43, 51, 55, 71, 159, 169; Karaite, 17 Jordan, 19 Justinian: fortifications, 88; plague, 51 K Kababīsh, 127 Kabāw, 35, 39 Kasakh Valley, 207, 209, 212, 213 Kanembu (language), 35 Karakhanid, 206 Kerma period, 125 Khārijī, 48, 52, 68, 70 Khazar, 9, 12, 18, 148, 150–154, 159, 160, 168, 173–175, 180, 181, 200; Khazaria, 202 Khusraw I Anushirwan, 146, 150, 199 Khurāsān, 6–7, 15 Kopetra, 94 Koyrkos, 91 Kotrac’, 208 Kourion, 86, 88, 94

Kuban: Bulgars, 174; river, 176 Kurdufān, 120, 121, 127, 129–131 L Lake Chad, 35 Late Roman, 91 Lazic War, 150 Lebanon, 91, 214 Levant, 84, 86, 87, 214 Limes, 22, 35, 40, 41 M Magog, 20, 150, 154, 172 Maghrib, 3, 9, 31, 36, 44, 47, 55, 61–62, 68, 74, 91 Makuria, 11, 119, 120, 121, 125, 128, 129, 131, 132, 133 Majorca, 47 Malta, 47 Mamlūk, 176, 206 Mamlūk-Mongol frontier, 5, 19 Manichaeans, 6 Masjid Zayd, 57 Mecca, 14, 21, 92, 93 Medina, 14 Meninx, 51 Mesopotamia, 142, 148, 150, 172–174 Mil Plain, 8 Minorca, 47 Miscona, 66 Mistawa, 66 Moabia, 66 Morocco, 16, 18, 31 Mzab, 32, 35, 41–42 N Nafūsans, 34, 42 Nahāvand, 150 Nālūt, 34–36, 39 Napatan period, 125–126 Naqsh-e Rustam, 150 Naxos, 23 Negev, 19 Nile, 13, 17, 107, 111–112, 119–120, 122, 125–133 Nobadia/Nobatia, 112, 119, 120, 128 Normans, 14, 48, 55, 68, 71 North Africa, 15, 16, 17, 31, 32, 40, 48, 52, 55, 56, 61, 74, 84 Nukkār, 16, 52–53, 54, 56, 61, 66, 68, 70

I ndex

231

O Orbelyan, 211 Oman, 16, 32 Ören Qala, 8, 148, 159 Orthodox Christianity, 81, 92, 213 Ottoman, 5, 19, 48, 50, 54–56, 65, 67, 68, 74, 142 P Palermo, 55, 72 Palestine, 43, 94, 157 Pantelleria, 47 Paphos, 94 Paros, 23 Peace of Westphalia, 139 Persia, 6, 15, 20, 88, 120, 130, 133, 150, 160, 174, 182, 196, 199, 200, 204 Piri Reis, 56 Pitsunda, 176 Q al-Qashtīl, 57, 65, 74 Qashtīl al-Wādī, 54 Qaṣr Ibrim, 10, 12, 107, 109, 110, 113 Qayrawān, 31, 32, 40, 61 Qishi’in, 57, 59 R Rhobia, 107 Rhodes, 82 ribāṭ, 16, 42, 47, 51, 61, 65, 70, 72, 206 Roger of Loria, 66, 74 Roman, 15, 17, 22, 31, 35, 40, 41, 49, 51, 56, 86, 128, 142, 143, 145, 147, 150, 157, 181, 196, 206 Rustamid dynasty, 52 S Safavid, 16, 52, 142 Sahara/Sahel, 22, 34, 40, 133 Salaḥ al-Dīn, 109 Salīma, 130 Saljūq, 195, 197, 198, 205, 206, 210, 211 Sanctus Nicolaus, 55 Sanchur Pass, 176 Santorini, 23 Sardinia, 47 Sarmatian, 171 Sasanian, 8, 10, 12, 13, 20, 89, 140, 142, 143–155, 173, 174, 175, 179, 181, 197, 204

232

I ndex

Scandinavia, 180 Scythians, 171 Selim, 210; Pass, 208, 211 Sengim, 179 Serra East, 23 Shabiran, 148 al-Sham, 200, 206 Sharūs, 34, 35, 37, 42, 44 Shengjinkou, 179 Shirvan, 202 Sicily, 12, 47, 48, 54–55, 65–66, 72, 84, 109 Sīdī Zikrī, 64 Ṣiffīn, 16, 32 Sijilmasa, 16 Silk Road, 13, 168, 170–171, 175, 179, 181 Slav, 106, 160 Sogdian, 6, 175–177, 181, 182 Sousse, 61 Spanish, 48, 54, 67, 68 Sub-Sahara, 35, 120, 133, 140 Sudan, 9, 10, 17, 35, 81, 106, 114, 120, 122, 125, 130, 131 Sūq al-Qiblī, 56 Syria, 32, 43, 47, 84, 88, 89, 107, 109, 151, 154, 173, 177, 214, 218 Syro-Anatolian frontier. See IslamicByzantine frontier T Tafialt Oasis, 16 Tāhart (Tiaret), 52, 53, 71, 74 Tala, 52 Tamtam, 131 Tantura, 88 Ṭarāblus. See Tripoli Ṭarī’. 13, 17, 19, 120, 130–131 Tbilisi, 217 thaghr, 47, 51, 152, 168, 199, 202 thughūr, 5, 8, 15, 63, 64, 68, 80, 82, 202 Tiaret. See Tāhart Tigharmīn, 35 Tigris River, 142 Tirmisa, 36 Tmizda, 38–40 Tripoli (Ṭarāblus), 47 Tripolitania, 18, 33, 35, 36 Tunis, 55, 56, 67, 72 Tunisia, 9, 18, 31–33, 41, 48, 67, 70 Tuquq, 179

Turfan, 179, 180 Turgut Reis, 54 Turkey, 197 U Ultan Qalasi, 8, 148, 151 Umayyad, 15, 31, 86, 89, 94, 175–176, 198 Uswān (Aswan), 18, 107 ‘Uthmān, 150, 151 V Vandals, 31, 40 Venice, 55 Vizari, 88 Volga River, 172, 174 W Wādī Abū Dūm, 130, 131 Wādī Huwār, 121, 125, 131 Wādī Malik, 120, 121, 125, 127, 129–132 Wādī Qa‘ab, 125, 129 Wādī Zabīb, 64 Wahbī, 16, 52–54, 56, 60, 64, 66, 68–70, 74 Warthān, 151

Westphalia, 139 West Africa (Gold Road, Slave Road), 13, 22 Wīghū, 34–35 Wuryūrī, 37 Y Yaiji (Harjis), 208 Yassi Ada, 88 Yemen, 20 Yenikapı, 88 Z Zacharias (Nubian king), 113 Zakaryan, 211 Zankur, 120, 121, 127, 130 Zanzibar, 32 zāwīya, 56, 60, 61 al-Zāwiya, 69 al-Zāwīya al-Jimniya, 56 Zāwīyat Sidi Būsa‘īd, 70 Zintān, 35, 36 Zirids, 48, 54, 68 Zoroastrianism, 204 Zwāgha Berbers, 52

I ndex

233