Architecture and Landscape in Medieval Anatolia, 1100-1500 9781474411301

Citation preview

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Not for distribution or resale. For personal use only.

E[)JNBURGH University Press J

Not for distribution or resale. For personal use only.

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500 Edited by Patricia Blessing and Rachel Goshgarian

Not for distribution or resale. For personal use only.

Edinburgh University Press is one of the leading university presses in the UK. We publish academic books and journals in our selected subject areas across the humanities and social sciences, combining cutting-edge scholarship with high editorial and production values to produce academic works of lasting importance. For more information visit our website: edinburghuniversitypress.com © editorial matter and organisation Patricia Blessing and Rachel Goshgarian, 2017 © the chapters their several authors, 2017 Edinburgh University Press Ltd The Tun – Holyrood Road 12 (2f) Jackson’s Entry Edinburgh EH8 8PJ Typeset in 10/12 pt Trump Medieval by Servis Filmsetting Ltd, Stockport, Cheshire, and printed and bound in Great Britain by CPI Group (UK) Ltd, Croydon CR0 4YY A CIP record for this book is available from the British Library ISBN 978 1 4744 1129 5 (hardback) ISBN 978 1 4744 1130 1 (webready PDF) ISBN 978 1 4744 1131 8 (epub) The right of the contributors to be identified as authors of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988 and the Copyright and Related Rights Regulations 2003 (SI No. 2498). Published with the support of the University of Edinburgh Scholarly Publishing Initiatives Fund.

Not for distribution or resale. For personal use only.

Contents

List of Illustrations and Tablesvii Acknowledgements xi Foreword by Scott Redford xii 1 Introduction – Space and Place: Applications to Medieval Anatolia Patricia Blessing and Rachel Goshgarian

1

Part I  Building: Masons and Infrastructure 2 Craftsmen in Medieval Anatolia: Methods and Mobility 27 Richard P. McClary 3 Stones for Travellers: Notes on the Masonry of Seljuk Road Caravanserais 59 Cinzia Tavernari Part II  Social Groups: Akhis and Futuwwa 4 Suggestions on the Social Meaning, Structure and Functions of Akhi Communities and their Hospices in Medieval Anatolia · I klil Selçuk 5 Social Graces and Urban Spaces: Brotherhood and the Ambiguities of Masculinity and Religious Practice in Late Medieval Anatolia Rachel Goshgarian Part III  Exchange: Islamic and Christian Architecture 6 Transformation of the ‘Sacred’ Image of a Byzantine Cappadocian Settlement Fatma Gül Öztürk 7 The ‘Islamicness’ of Some Decorative Patterns in the Church of Tigran Honents in Ani Mattia Guidetti

95

114

135 155

Not for distribution or resale. For personal use only.

vi

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Part IV  Frameworks: Language, Geography and Identity   8 Harvesting Garden Semantics in Late Medieval Anatolia185 Nicolas Trépanier   9 All Quiet on the Eastern Frontier? The Contemporaries of Early Ottoman Architecture in Eastern Anatolia 200 Patricia Blessing 10 The ‘Dual Identity’ of Mahperi Khatun: Piety, Patronage and Marriage across Frontiers in Seljuk Anatolia224 Suzan Yalman Notes on Contributors Bibliography Index

Not for distribution or resale. For personal use only.

253 257 287

Illustrations and Tables

Maps    1 Anatolia, c. 1100    2 Anatolia, c. 1210    3 Anatolia, c. 1240    4 Anatolia, c. 1300    5 Anatolia, c. 1360–80

xiv xv xvi xvii xviii

Figures   2.1 Glazed terracotta plasterer’s trowel, Arkeoloji Müzesi, Sivas30   2.2 St Hovhannes Mkrtitch (St John the Baptist) Church, Gandzazar (1216–38) and Kilij Arslan II minbar, Konya (1155)33   2.3 Citadel Mosque, Divri©i; portal cross-section at 130 cm above top step 36   2.4 Citadel Mosque portal, Divri©i (1180–1); door jamb detail and portal lintel detail 37   2.5 Sitte Melik Tomb, Divri©i (1196–7); portal door jamb detail39   2.6 Mama Khatun Tomb, Tercan (c. 1200); upper band of muqarnas above left niche 40   2.7 Upper portion of a tombstone in the southwest corner of Ahlat graveyard, late twelfth century 40   2.8 Gök Medrese portal, Sivas (1271–2); iron cramp 42   2.9 Elbow brackets: Citadel Mosque, Konya; al-Aqsa Mosque, Jerusalem 45   3.1 The courtyard of Evdir Han 64 ·   3.2 The interior of the vaulted hall of Incir Han 69 ·   3.3 The mason’s mark on the buttress of Incir Han 71   3.4 The portal and the inscription of Çardak Han 74   3.5 The bond between the courtyard and the main façade (southeast corner) of Çardak Han 80

Not for distribution or resale. For personal use only.

viii

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

  3.6 The bond between the façade and the portal (southwest side) of Evdir Han 80 ·   3.7 An example of spolia in Incir Han 83   3.8 Finishing of the outer surface of a stone in Evdir Han 84   3.9 Finishing of the outer surface of a stone in the caravanserai of ‘Itna, built in Syria c. 1230 86   6.1 Map of Cappadocia 137   6.2 Plan of Area 5, Açıksaray 140   6.3 Site plan of Açıksaray showing the location of individual Areas and stand-alone churches 141   6.4 Vestibule in Area 5, Açıksaray 143   6.5 Supposed mi˙råb in the vestibule, in Area 5, Açıksaray 144   6.6 Stable of Area 2, Açıksaray 145   6.7 Hall of Area 1, Açıksaray 145   6.8 Independent hall (‘mosque’), Açıksaray 148   6.9 Probable mi˙råb in the independent hall (‘mosque’), Açıksaray149   7.1 View of the interior of the Church of Tigran Honents (St Gregory the Illuminator), Ani, 1215 157   7.2 Carved relief on the spandrels of the blind arcade, southern exterior wall, Church of Tigran Honents (St Gregory the Illuminator), Ani, 1215 159   7.3 Facing dragons, dihedral niche on the southern ­exterior wall, Church of Tigran Honents (St Gregory the Illuminator), Ani, 1215 160   7.4 Bevelled style, dihedral niche on the southern exterior wall, Church of Tigran Honents (St Gregory the Illuminator), Ani, 1215 161   7.5 Capital in the zhamatun, Church of Tigran Honents (St Gregory the Illuminator), Ani, early thirteenth century162   7.6 Relief from Kubatchi, Daghestan, medieval 162   7.7 Bowl decorated in the ‘bevelled style’, Samarqand, tenth century165   7.8 Naming of John the Baptist, Gospel manuscripts, British Museum and Vatican Library 166   7.9 Drawing of the muqarnas decorating an engaged pillar, Great Mosque of Mosul, c. 1172–3 168 7.10 Ceiling of one bay of the prayer hall of the so-called mosque of Manouchihr, early thirteenth century, Ani 171 7.11 Main portal, gavit, Church of St Apostles, Ani, 1217 172 7.12 Detail of a fragment of the cornice of the main portal, gavit, Church of St Apostles, Ani, 1217 173 7.13 Painted imitation of a textile decorated with simurghs, Church of Tigran Honents, Ani, 1215 176   9.1 Interior view towards qibla wall, taken from gallery, Aslanhane, Ankara 206

Not for distribution or resale. For personal use only.

illustrations and tables

  9.2 Güdük Minare in Sivas 208  9.3 Kö∞k Medrese, view 210   9.4 View of portal façade, Hatuniye Medrese 212   9.5 View of courtyard, c. 1895, Hatuniye Medrese 213 · ·   9.6 View of mausoleum, Ibrahim Bey Imaret215 · ·  9.7 Ibrahim Bey Imaret, Karaman, portal 217   9.8 Hasbey Darülhuffaz, Konya, view 219 10.1 General view of Huand (Mahperi) Hatun Complex, Kayseri225 10.2 Plan of Huand Hatun Complex, Kayseri 227 10.3 Huand Hatun Madrasa, Kayseri, view of central Èwån228 10.4 Huand Hatun kümbet, Kayseri, interior view with three cenotaphs229 10.5 General view of Shaykh Turasan zåwiya, Cappadocia, · Incesu, 1242–3 232 10.6 Tomb tower of Malika Adiliyya (‘Çifte Kümbet’), Kayseri, 1247–8 234 10.7 Tomb tower of Mahperi Khatun, Kayseri, with detail of changes in masonry 238 · 10.8 Shaykh Turasan zåwiya, Cappadocia, Incesu, foundation inscription242 Plates Between pages 174 and 175

1 Citadel Mosque, Divri©i (1180–1); portal 2 Citadel Mosque, Konya, north portal; Madrasa al-Sultaniyya, Aleppo 3 Malatya Great Mosque (1247); large structural brick muqarnas cells in maqßËra dome squinches 4 The portal of Evdir Han · 5 The courtyard and the main façade of the vaulted hall of Incir Han 6 The main façade of Çardak Han 7 Açıksaray, Areas 1 and 2 8 Façade of Area 5, Açıksaray 9 Façade of Area 1, Açıksaray 10 View of the Church of Tigran Honents (St Gregory the Illuminator), Ani, 1215 11 View of the mosque of Manouchihr, Ani 12 Detail of upper section of mi˙råb, Aslanhane, Ankara 13 Signature on portal, left side, Hatuniye Medrese, Karaman · · 14 View of Ibrahim Bey Imaret, Karaman 15 Huand Hatun Mosque, Kayseri; general view of west portal with conical dome of tomb tower to the left 16 Tomb tower (kümbet) of Mahperi Khatun, Kayseri

Not for distribution or resale. For personal use only.

ix

x

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Tables 2.1 Processes involved in stone construction 48 2.2 Processes involved in the production and use of bricks 49 2.3 Processes involved in the production and installation of glazed elements 49 2.4 Processes involved in the production and installation of mortar50 2.5 Processes involved in timber construction and woodworking50 2.6 Processes involved in the production of iron objects 51 2.7 Miscellaneous site roles 51 4.1 Comparative distribution of akhi endowments98 4.2 The distribution of initial land grants to sheikhs, fakÈhs and akhis according to the Hüdåvendigår livåsı cadastral surveys99

Not for distribution or resale. For personal use only.

Acknowledgements

This volume is the result of two conference panels organised by the editors at the Middle East Studies Association conference in 2013, and at the Society of Architectural Historians’ Annual Meeting in 2014. We would like to thank both organisations for the opportunity to present and share the projects now published here. At the same time, we would like to thank the contributors to the panels, including those who are not part of the present publication. For funding that allowed us to commission maps, we are grateful to the Academic Research Committee of Lafayette College. In turn, we thank Philip Schwartzberg of Meridian Mapping for professionally preparing the maps. We would also like to thank Noah Root for his significant editing contribution to this project. At Edinburgh University Press, we are grateful to Commissioning Editor Nicola Ramsey, Assistant Commissioning Editor Ellie Bush, and Desk Editor Eddie Clark for their support of our project and their great attention to detail in the publishing process. We would also like to thank the anonymous reviewers for their gracious and productive feedback. Finally, we are grateful to Scott Redford for agreeing to write a Foreword for this volume, as every single contributor has benefitted from his work and insights.

Not for distribution or resale. For personal use only.

Foreword Scott Redford

One year, close to the end of the semester, I overheard one of my undergraduates talking to a fellow student. ‘It’s really easy,’ he said. ‘After the Romans came the Byzantines. Then came the Seljuks and the Ottomans.’ In giving a class on the architecture of classical and medieval Anatolia, I had, of course, also mentioned Georgians, Armenians, Beyliks, Danishmendids, Mengüjekids, Trapezuntines, Crusaders, and other states and cultures. But this was cram time, and the boileddown version was being prepared for examination booklets. The readers of this volume are most definitely not be served the boiled-down and canned version of medieval Anatolia. Instead, the editors and contributors have served up a new set of dishes: crossing traditional ethnic, chronological, confessional and disciplinary borders; in a phrase, stirring things up. One of the editors is an art historian, the other an historian: the contributors are a mix of the two. And yet all the authors are engaged with the material culture of medieval Anatolia, from the rural to the urban, from the sacred to the profane, with topography and geography as well as with inscriptions and a wide variety of textual sources. This mixes things up as well, but not in a random way; in each of these essays is a sense of place, of regional specificity, and of the issues of the time. Another thing the contributors take for granted is cultural interaction and complexity. They are suspicious of master narratives and truisms, be these clichés concerning Oriental despots, gender roles, nomads, dervishes, monks, gazis, craftsmen, guilds or motivations that are somehow not based on war or religion. No one can learn equally well the languages and literatures of all the cultures of the area that is now Turkey, and so there is a lot of cross-referencing going on between the essays – all to the advantage of the reader, and to the study of medieval Anatolia, a study that is progressing beyond paradigms of nationalism, provincialism, periphery and orientalism (in both the traditional and Saidian senses of the word). In exploring

Not for distribution or resale. For personal use only.

xiii

foreword

cultural experimentation, byways, gardens, stables, the tombs of holy men, exclusive male clubs, not to mention cities and monuments off the beaten track, the reader will be moved to explore new trajectories, both chronological and geographical, and the research of a new generation of historians and art historians, itself diverse. Where will this lead? There is no saying. I do not think that the authors are moving towards one particular new paradigm. But especially at a time when Turkey itself is changing in response to perceived visions of past glories, and at a time when its historical monuments are being rebuilt to reflect not constructional truths, but perceived economic and societal aims, there is a certain solace in encountering essays that tread roads not (often) taken, that question verities and paradigms, and make us think of this complex, challenging and creative period in new ways.

Not for distribution or resale. For personal use only.

B l a c k Not for distribution or resale. For personal use only.

~~ ft. b ..qOq • .. \.~ ...; rr?_ ,..,

"

S e

4

I

, .... ...

•• •• •

•••

...

7e~ ·

(to-IJenice.

••••

Mediterranean Sea

:~Ogdod

·: : :.~w~~ ~ -- ••< •• •• •• • •••••••

0

50

100

150 miles

Ac11tl

•••••••••• -:

Principality of Antioch .

K1ngdom of Jerusalem

.

•

.•. ]. Be1rut

••

•• •

.

Ayyubid • Sultanate ••f+ ••••

Map 2 Anatolia, c. 1210 (Patricia Blessing and Philip Schwartzberg/Meridian Mapping)

•

~

•

Anatolia in 1240 Oust before the Mongol Conquest) Not for distribution or resale. For personal use only.

} B l a c k

S e

4

f) ...

Mediterranean Sea 0

50

100

150 miles

Map 3 Anatolia, c. 1240 (Patricia Blessing and Philip Schwartzberg/Meridian Mapping)

B l a c k

Sea

Not for distribution or resale. For personal use only.

Mediterranean Sea 0

50

100

150 miles

Map 4 Anatolia, c. 1300 (Patricia Blessing and Philip Schwartzberg/Meridian Mapping)

Sea

B l a c k Not for distribution or resale. For personal use only.

~

(to~VeniceJ. • •

S e

4

of St. John

•••••

Mediterranean Sea

0

50

100

150 miles

•• •••

•• •• ••••

Map 5 Anatolia, c. 1360–80 (Patricia Blessing and Philip Schwartzberg/Meridian Mapping)

CHAPTER ONE

Introduction – Space and Place: Applications to Medieval Anatolia Patricia Blessing and Rachel Goshgarian

Imagining late medieval Anatolia presents a unique challenge. The geographical region is most usually referred to as the Asian portion of the modern Republic of Turkey. During the twelfth–sixteenth centuries, it became home to increasing numbers of Turkish-speaking Muslims. Even the most nuanced of thinkers can experience difficulties when trying to see this series of centuries as anything other than a process of Turkification and Islamisation. And, to some, insisting that terms such as ‘Anatolia’ or ‘Lands of RËm’ are most appropriate when discussing the region during this time period might seem like an historiographical version of René Magritte’s La trahison des images, which in this case might be represented by an imaginary map of Anatolia with a cursive ‘This is not Turkey’ written underneath. Years after the original painting became a part of public discourse, Magritte explained: The famous pipe. How people reproached me for it! And, still, could you stuff my pipe? No, right? It’s just a representation. So, if I had written on my painting ‘This is a pipe’, I would have been lying!1 Similarly, representing this time period in Anatolian history ­ rimarily through the lens of Turkey and Turkification keeps us p from approaching a more nuanced and honest representation of the past. The overarching goal of this volume is to use both art historical and historical approaches to expand our knowledge of complex, medieval Anatolian societies by looking beyond political structures and towards: a reconsideration of the interactions between the rural and the urban; an analysis of the relationships between architecture, culture and power; and an examination of the region’s multiple geographies (perceptional, physical, political and religious). In order to widen historiographical perspectives, the contributors use a broad variety of sources (architectural, artistic, documentary and literary),

Not for distribution or resale. For personal use only.

2

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

including texts composed in several languages (Arabic, Armenian, Byzantine Greek, Persian and Turkish). Each chapter’s study actively engages with the permeable kind of boundaries that existed in late medieval Anatolian society, and suggests the possibility of (and need for) deeper and more nuanced examination that can help us to better understand how inclusions and exclusions, participation and indifference, proximity and distance might have been understood in the medieval Anatolian experience. An engagement with the conversation on space and place, based initially on Henri Lefebvre’s work, is central to this volume’s foundation and will also situate the study of this time period within a larger theoretical and intellectual discussion of the ways in which space is represented and place is experienced.2 The conceptual framework of space and place offers a useful window through which to view Anatolia during this time period, specifically due to the fractured nature of its political history and the overwhelmingly complex ‘factography’ it inspires.3 The authors in this volume look beyond the histories of transition and ethnocentric narratives and towards a more intimate collaboration between the fields of History and Art History.4 This step beyond disciplinary boundaries, coupled with a collective preoccupation with space and place, offers us an opportunity to look at the cultural landscapes of late medieval Anatolia as both conceived and physical products of the people who lived there. Historical Overview and New Approaches In this volume, we engage with a version of the nouvelle histoire within which material formerly perceived of as peripheral is placed at the centre of our canvas.5 Explicitly rejecting ethnocentrism, the chapters present in this volume paint focused, realistic images of the architectural, cultural and social landscapes of late medieval Anatolia, c. 1100–1500. During this time period, Anatolia was home to a large number of polities that ruled over various regions concomitantly and in relatively quick succession. Sometimes, the territories claimed by these polities even overlapped. Because of the seemingly consistent political fluctuation that characterised this time period, intellectual engagement with uniquely one polity, one ethnic or linguistic group, or even one kind of architectural structure ultimately limits our capacity to understand the realities of the time period. The spatial turn that has recently been embraced by historians and art historians allows us to ask questions about the relationship(s) between the representations of space and experiences of place. In the case of late medieval Anatolia, these kind of questions are particularly useful as they help us to detect and analyse continuities across regions, religions and languages. The administrations and intellectuals living in these polities

Not for distribution or resale. For personal use only.

introduction

expressed themselves using different languages; some used Arabic, others used Armenian, Byzantine Greek, Georgian, Latin, Persian or Turkish. Leaders of these realms practised various faiths; some embraced Sunni Islam and some practised Byzantine (Chalcedonian) Christianity. Still others made Shia Islam or non-Chalcedonian Christianity their official faith. And each of these political administrations ruled over heterogeneous populations that included both people who shared their language and the faith practices, and also relatively large sections of the population who spoke languages and practised religions that differed from those embraced by the administration. Nomadic peoples in Anatolia consistently challenged the organisational capacity and administrative dexterity of the polities located there. Some principalities (and cities) were able to strike a relatively happy modus vivendi with the nomads living within their boundaries, while others struggled to live peacefully with them. During this long period of complex cultural engagement and political fracturing, the twelfth-century Anatolian experiences appear similar enough to those of the sixteenth century that consideration of these four centuries of history as a unit allows for useful investigation. As Cemal Kafadar has suggested, these centuries represent a unique and long period of Anatolian history that he likened to the †awå’if (Party Kings) of medieval Iberia and should be ‘characterized in its own right’.6 Whether engaged with the study of art history or history, many scholars have focused on this region and time period of study either through the lens of the slow decline of Byzantium or through the framework of an inevitable rise of the Ottomans. This kind of teleological scholarship aims to illustrate a process and has encouraged an acceptance of ideal, exact moments (i.e., beginning and end) and, in turn, the building of exclusive narratives that obscure the multiple realities of medieval Anatolian experiences.7 At the same time, a unique combination of circumstances has led to a certain degree of ‘ghettoisation’ of the various fields associated with the study of this time period. The effects of nationalism, the political estrangement of various countries – such as Iran, Turkey, Greece, Georgia and Armenia – from one another (and, often, of intellectuals in those countries), the type and availability of source materials, and the linguistic challenges associated with those sources have left us with disconnected histories of Anatolia. A critical body of scholarship has recently emerged that considers the porous nature of the various frontiers (physical, cultural, linguistic, social, religious, philosophical) that existed in late medieval Anatolia as central to the production of local cultural, political and social fabrics. Increasingly, scholars have begun to engage with analyses across the perceived boundaries between nations and religious communities that allow for improved comprehension of a time period that can best be understood as a collection of complex

Not for distribution or resale. For personal use only.

3

4

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

moments of cultural, political and social competition, interchange and overlap. This volume takes a step further in this direction. Anatolia: Discourses on a Medieval Geographical Entity In any discussion of late medieval Anatolia, it is crucial to begin with the question of both temporal and geographical definitions and descriptions. What do we mean when we use the term ‘Anatolia’? What has this term meant and how have its meanings changed over time? What are the benefits of using this term? What might be some of its limitations and prejudices? First, however, we need to consider the problematic of geohistorical terms in and of themselves. In the introduction to his Toward a Geography of Art (2004), Thomas DaCosta Kaufmann explains the need to re-evaluate the relationship between geography and art history via the use of ‘geohistory’. According to Kaufmann, geohistory emphasises the historical aspects of geographies; individual circumstances (rather than spatial processes); and the specificity of ‘places where processes and factors are actually evinced’.8 These same kind of contextualisations and emphases are also necessary for post-nationalistic historical analysis and its treatment of geography, issues that are central for any study of medieval Anatolia. Today, the term Anatolia is most commonly used to describe the area roughly defined as mainland Turkey, extending from the Sea of Marmara in the west to the eastern borders with Armenia, Georgia, Iran, Iraq and Syria, and from the Black Sea coast in the north to the Mediterranean in the south and southwest. Using the term Anatolia to describe a medieval region presents us with a teleological approach, understanding medieval geography through the lens of the inevitable emergence of a modern nation-state. During the late medieval period, the regions that constituted what is today considered Anatolia were defined in different languages and cultural contexts by a range of terms. These include RËm, Anatolia, Trebizond, Nicaea, Armenia, Georgia, Turkey, Mesopotamia and Kurdistan, terms that were both dynamic and coexistent. The geographies of the region in question, as understood and experienced in medieval texts composed in various languages, suggest that the boundaries of these regions were not necessarily static or universally understood during this time period. It is, perhaps, because of the dynamism of these regional boundaries and the ensuing question of their significance in everyday life that it becomes even more difficult to name the series of lands in question. To further complicate our already cluttered canvas, the factors that determined the naming of places were not always consistent. For example, the medieval Armenian scholar Mxit‘ar Goš (d. 1213) considered demography to be a decisive geographical factor and referred to large swaths of land that were: (1) neither a part of a geographical area that had previously

Not for distribution or resale. For personal use only.

introduction

been considered ‘Armenia’, nor (2) under the control of an Armenian polity, as Armenia because the populations there were either significantly (or, predominantly) Armenian.9 In modern scholarship, the different terms used for the region have various implications. The Latin expression Asia Minor suggests a link with Roman heritage, while Anatolia is a Greek term that was adopted into Turkish as ‘Anadolu’ (a term whose folk etymology is ‘full of mothers’), in order to point to the old and layered civilisations there, but is often associated with the rise of late Ottoman and early republican Turkish nationalism.10 The term ‘lands of RËm’, which means the eastern Roman domains and designates both Anatolia and the Balkans, has been increasingly used in recent studies. Scholarship that uses this term engages with a critique of Orientalist and nationalist discourses, and embraces the multiple cultural geographies of the region – part Roman by way of Byzantium, part Islamic by way of the Arabic term for the Byzantine realm. A remarkably supple term, ‘RËm’ can refer at once to Rome, Byzantium, Anatolia and the Balkans, while the adjective ‘RËmÈ’ can mean Greek, Byzantine, Anatolian or Ottoman.11 Moreover, the term ‘lands of RËm’ alludes to certain concepts that are central to the study of the medieval Mediterranean: namely, a preoccupation with mobility, frontier and geography. Within the framework of a re-evaluation of this region as a frontier zone, scholarship that engages with the term ‘lands of RËm’ generally encourages broader discussions of artistic geographies as well as centres and peripheries in the medieval and early modern worlds.12 As the work at hand does not engage with research on the Balkans and because the region of Anatolia can be considered a self-contained area (albeit problematic), we embrace the term ‘Anatolia’ as a representation of an amalgam of spaces that share a geohistory in the medieval period and constitute a conceived space whose borders are grainy and fluid. In fact, we use it because it is a term that lacks clarity and consistency, similar to conceptions of this region during the late medieval period, and reflective of the complex nature of the cultural and political setup that characterised it. Even though Anatolia itself has been increasingly studied as a frontier region, comparisons with other medieval frontier zones have yet to be clearly established. Still, places with similar dynamics of religious exchange and military conquest that challenge commonly held geographical and national narratives should become part of a larger discourse on frontier regions as zones of crosscultural exchange in the Middle Ages. Within such a context, medieval Anatolia could easily be viewed in comparative perspective with other frontier zones, such as twelfth-century northern India, medieval Sicily or Iberia, a region in which the dynamics of conquest and reconquista fostered a similarly fraught and complicated political milieu.13 It is within this larger context of medieval

Not for distribution or resale. For personal use only.

5

6

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

borderlands (or frontier studies) and centre–periphery dynamics that consideration of Anatolia will become useful beyond simple reflection upon itself. Considering medieval Iberia through the lenses of diversity and multiculturalism has been a significant academic course of inquiry. In the 1940s, Spanish historian, Américo Castro, coined the term convivencia (coexistence) in order to describe the tolerant interactions between Christians, Jews and Muslims in medieval Iberia.14 He understood convivencia as a kind of medieval multiculturalism and focused much of his research on architectural, literary and religious engagements, generating an image of peaceful and productive cohabitation and exchange. About fifty years later, American historian David Nirenberg reconsidered the notion of convivencia and suggested that these religious groups coexisted in a relatively constant state of competition; the landscape of relatively peaceful coexistence was punctuated by brief, violent episodes.15 With regard to medieval Anatolia, scholars are beginning to engage with the ways in which its multiculturalism affected the interactions between various people(s) and have been interested primarily in the notion of considering this through the lens of ‘frontier studies’. While conceptually, the comparison and contrast between the two regions is useful, and can be beneficial in exploring the issue of frontier in medieval history, the limits of this approach must be acknowledged. Specifically, terms such as convivencia and Party Kings are most meaningful in a Spanish context, and are not easily directly applied to Anatolia. Nevertheless, as Peacock, de Nicola and Yıldız have suggested, certain benefits can arise from the adaptation and adoption of the abstract concepts raised in the study of medieval Iberia for the study of medieval Anatolia. General Notes on Historiography There are two central problems related to the study of medieval Anatolia in a general context: no recent overarching history of the region exists and earlier historical surveys tend to focus either on the Turkification of the region or on the decline of Byzantium. Because of this, in some cases, the dynastic divisions established by Islamic historiography dominate the telling of historical events and many nuances are lost. At the same time, a preoccupation with the fall of Byzantium has encouraged studies that consistently debate when, where and how Byzantium experienced decline. For general overviews on medieval Islamic Anatolia after the Seljuk conquest, the reader unfamiliar with Turkish must still rely on Claude Cahen’s work, first published in 1968.16 Cahen accepts an overall narrative that explains the medieval political history of Anatolia through the lens of a history of ‘Turks’ and a land yet to be known as ‘Turkey’. This narrative has been almost

Not for distribution or resale. For personal use only.

introduction

completely absorbed by political historians.17 A counterargument to the Turkish-centred narrative offered in Cahen’s work, and many subsequent Turkish works on the subject, is proposed in The Decline of Medieval Hellenism in Asia Minor and the Process of Islamization from the Eleventh through the Fifteenth Century. The title of the monumental tome written by Speros Vryonis perfectly describes the book’s framework and aims. A robustly researched work, Vryonis’ book focuses specifically on central and western Anatolia and suggests that the combination of nomadisation and Islamisation in these regions led to the decline of Byzantine civilisation.18 This ‘history of cultural transformation’19 assigns blame for the Byzantine disaster to Turko-Muslim conquerors and analyses the destruction of towns and church administration by carefully studying nomadic life and the establishment of Islamic institutions in Anatolia. Other influential presentations of the region have focused on specific polities. While often overlapping, possessing changing borders and varying degrees of direct political and economic control, these polities have traditionally been studied separately, organised into the following general categories: Byzantine, Seljuk, (Armenian) Cilician, Mongol, Beylik, Ottoman. Most medieval chronicles had polity-focused orientations, thereby making history writing along polity lines very practical, both in terms of organisation and language skills; this kind of intellectual exercise has contributed to the excavation of significant historical evidence. Still, the modern temptation to explain events through an ethnocentric (and occasionally nationalist) lens has dominated many of these polity-focused studies and, ultimately, limited their analyses. More recent scholarship has turned away from these models and towards comparative and integrative histories with general success. Nevertheless, Osman Turan’s Selçuklular zamanında Türkiye (Turkey in the Seljuk Period) remains one of the most important and complete works on the Seljuks and on Anatolia in the late medieval period. Turan’s nationalist approach sought to combine Turkish and · Islamic history (in what has come to be known as the Türk-Islam sentezi or the Turkish-Islamic synthesis) and to validate the idea of a unique kind of ‘Turkish Islam’ with a linear historical progression.20 The first edition of his book was published to coincide with the 900th anniversary of the battle of Manzikert in 1971, clearly suggesting the kind of linear narrative the author embraced.21 Beginning with the first Mongol incursions into Anatolia in the 1230s, Sara Nur Yıldız has provided a new account of the political history of Anatolia after the Mongol conquest in 1243. Her work is one of the first attempts at directly studying the Ilkhanid presence in Anatolia, and specifically looks at the ways in which the Mongol and Seljuk dynasties interacted with each other.22 Yıldız’s insights are unique and significant in that they offer a superior analysis of how two different entities vying for control over the lands and peoples of

Not for distribution or resale. For personal use only.

7

8

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Anatolia engaged with each other and represents a real shift in our understanding of medieval Anatolia. The decline of Ilkhanid power in the fourteenth century led to the rise of numerous new, local powers across Anatolia. Many of them, particularly in central and eastern Anatolia, were former Ilkhanid governors who saw the moment as an opportunity and took their chances at independence, experiencing varying levels · of success. Ismail Hakkı Uzunçar∞ılı’s work on these Anatolian Beyliks, first published in 1937, is still considered the standard work on the period.23 Still, much more attention is devoted to the Beyliks of western and central Anatolia (e.g., the GermiyanoÌulları, the Karamanids and the ÇandaroÌulları) than is offered to their contemporaries further east. The dizzying number of principalities coexisting in post-Mongol Anatolia – the Eretnids, Qadi Burhaneddin, the Dhu’l-Qadir and others – along with the major Turkmen dynasties of the Qaraquyunlu and Aqquyunlu, effectively forced Uzunçar∞ılı to provide only a summary survey. Subsequent studies fall into the dynastic model and, for the most part, are available only in Turkish.24 One exception is a survey by John Woods on the Aqquyunlu that focuses on Iran, but includes the dynasty’s activities in Anatolia beginning in the late fourteenth century.25 Byzantine histories that deal with Anatolia during the time period under consideration here predominantly consider Constantinople as imperial centre (or, for a while, as the seat of the Latin Empire, 1204–61), or at Trebizond, Epirus or Nicaea as alternate centres of Byzantine power.26 Other studies of later Byzantium look to links between the Byzantine world and European intellectual, religious and commercial currents.27 More recently, however, scholars have turned their interest towards the relations between Byzantium and the East, and to a certain extent to the relations between the Byzantine Empire and the Ottomans.28 While these recent, more comparative and boundary-breaking studies indicate a turn in the field, the tendency amongst Byzantine historians to analyse most things through an imperial lens is almost inevitable. However, Byzantine Studies is a field that has long been invested in a diverse array of historiographical approaches, including social, cultural, economic, minority, agrarian and rural histories. Armenian history has consistently been studied as the ‘history of the Armenian people’ (in Armenian, hayoc patmut‘iwn). Thus, the work of scholars writing histories in Armenian has almost always been ethnocentric in focus, albeit often less polity-focused than other historiographical traditions, as the Armenian people have lived a polycentric diasporic existence since at least the fifteenth century. At the same time, the history of Armenians – whether in Greater Armenia, Cilicia or beyond Anatolia – has frequently been equated with the study of non-Chalcedonian Christians surrounded by Muslims, Chalcedonians or Catholics. Robert Bedrosian’s

Not for distribution or resale. For personal use only.

introduction

dissertation turned a new page in the understanding of Armenian history during this time period; it specifically addresses the ways in which Armenian noble families were affected by Seljuk and, later, by Mongol rule.29 Bedrosian managed to reorient the field of late medieval Armenian history towards the greater framework of the geopolitical realities of the Seljuk and Mongol invasions. Over the course of the past nearly forty years, scholarship on Armenia both inside the Republic of Armenia and outside it has looked more and more towards connections between Armenians and the world(s) around them. While nationalist tendencies continue to inspire certain scholars in the region, some histories and art histories of Armenia and Georgia have been moving in similar directions and towards less narrowly drawn definitions of culture and identity in the medieval world, as have various treatments of Syriac histories of the region. Still, the most successful of these efforts have often been conceived in a Byzantine framework.30 Ottoman Exceptionalism One Beylik is conspicuously absent in Uzunçar∞ılı’s work: that of the Ottomans. Already in the 1960s, the pre-imperial Ottoman experience had been formally separated from competing contemporaries in intellectual discourse. Uzunçar∞ılı seemingly participated in the establishment of what has come to be understood as an Ottoman historiographical exceptionalism: the early Anatolian context of the dynasty is almost completely erased. In fact, this tendency to isolate the study of the Ottoman dynasty deeply affected not only the ways in which the Ottoman Empire itself has been examined, but also the ways in which medieval Anatolia has been considered. Much ink has been spilt over the question: how was the Ottoman Empire established? For nearly a century, scholars have tried to understand what drove the Ottoman conquests and how inclusion in the Ottoman experience was promoted and encouraged. The identity of the Ottomans, whether framed in reference to their inherited lifestyle, religious orientation or genetic code, is a central focus of this debate. In the 1930s, Austrian historian Paul Wittek proposed that the Ottomans were able to transform their early polity into a great empire because they were inspired and driven by gaza, or religious plunder.31 This framework has shaped the way in which generations of scholars have considered the rise of the Ottomans. In 1959, Mehmet Fuad Köprülü, drawing from a much richer source base, suggested that it was the Turkish tribal tradition of the Ottomans that contributed to their success in building an empire.32 Various other scholars have reconsidered the success of the Ottomans through different lenses. In 1986, Rudi Lindner used anthropological methodologies to analyse the tribal aspects revealed by fourteenthcentury Ottoman sources; he argued that Ottoman successes can

Not for distribution or resale. For personal use only.

9

10

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

best be understood via the administrative decline of Byzantium and an Ottoman desire for economic control of Bithynia.33 Similarly, Colin Imber has criticised Wittek’s understanding of the term gazi and suggested that the Ottoman Empire was anything but exclusively Islamic or Turkish.34 Cemal Kafadar re-evaluated sources and scholarship and pointed out the dangers of using essentialising terms and arguments.35 Kafadar’s work encourages a move away from the notion that identities can be defined and inspires an intellectual engagement with the fluidities and liquidities of life in the border regions. Heath Lowry’s more recent analysis of the early Ottoman experience also rejects the gazi thesis and presents his version of a re-evaluation of the early Ottoman Empire.36 In her review of said monograph, Kate Fleet suggested that: ‘as we move into the 21st century, perhaps it is time to relinquish Wittek and, for the health of early Ottoman studies, to advance bravely without him’.37 Slowly, but surely, it seems that scholarship is beginning to move away from an embracing of Ottoman exceptionalism. As this trend continues, we will begin to better understand the late medieval Anatolian – and Ottoman – experience(s) as something more complex and profound than a simple backdrop or context for the eventual establishment of the Ottoman Empire. Narratives in Art History Due, in part, to the ways in which linear national narratives developed in the context of history writing, the field of Art History has also been highly influenced by national(ist) narratives; initially, some of the main scholars of medieval Anatolian history were also involved in elaborating narratives for art history. Thus, Köprülü was involved in commissioning the French archaeologist Albert Gabriel (1883–1972) to write a two-volume history of the ‘Turkish’ monuments of Anatolia.38 Gabriel also produced further volumes on southeastern Anatolia, and on Bursa as an Ottoman capital.39 The introduction of formalist methods in art history, based largely on Köprülü’s collaboration with Austrian art historian Josef Strzygowski (1862–1941), and thanks to the presence of Strzygowski’s student Heinrich Glück in Turkey, had a far-reaching impact on the way in which medieval Anatolia was studied.40 Even though he did not visit Turkey, Strzygowski was commissioned to write an article on Turkish art that encompassed a pan-Turkic vision, from Central Asia to Anatolia.41 This narrative was a break with the nineteenthcentury discussion on the subject. Scholars in the late nineteenth and early twentieth centuries, in the context of the Ottoman and German empires, viewed Seljuk architecture in Anatolia as Persian. The German art historian Friedrich Sarre, in particular, emphasised the larger Persianate culture of the Seljuk court and highlighted connections to Iran.42 With these conflicting narratives – Persianate

Not for distribution or resale. For personal use only.

introduction

culture and Turkish nation – the historiography of medieval Anatolia is fraught with tensions and contradictions. The historiography of Armenian architecture appears separate from that of other monuments in Anatolia for much of the twentieth century. Only recently have projects begun to look at Armenian monuments in Anatolia alongside contemporary Islamic structures. In addition, scholars have also begun to look at architecture produced in the modern Republic of Armenia (and Georgia) in a larger Anatolian context, as part of a larger movement towards considering aspects of the cultural and political situations of medieval Anatolia and the Caucasus together.43 In the early twentieth century, the arts of Iran and Armenia were often discussed together.44 In addition to his focus on a consistent transmission of Persian culture over centuries, Pope discusses the possibility of an Armenian influence on the formation of the Gothic in Europe in a paper that focuses on the structure of vaults. Ultimately, Pope opts for the primacy of the older Persian tradition over Armenian influence, a decision that is understandable within the context of his own work, which is notably dominated by the study of Iranian art of all periods.45 This is one of the instances in which the historiographies of medieval Iran and Armenia intersect, reaching all the way back to Strzygowski’s studies of Armenian architecture.46 In a deconstructive work that criticises the oeuvre of Strzygowski for its deep links to the theories of ‘race’ and ‘nation’, Christina Maranci demonstrated how Islamic influence was painted as a factor that led to a decline of Armenian architecture both by Strzygowski and by those scholars who had been influenced by his book. For the most part, the early Arab incursions of the seventh and eighth centuries and the Seljuk conquest of the second half of the eleventh century are regarded as detrimental to Armenia and the work of its builders.47 Strzygowski’s book is very much a product of its own time, situating Armenian architecture within the context of Indo-Aryan studies and of nationalistic tendencies in the wake of the collapse of the AustroHungarian Empire after the First World War. Strzygowski argued that a combination of Armenian and Islamic traditions shaped an architectural vocabulary visually similar to what in Western medieval architecture is considered Gothic.48 Donald Wilber returned to the reflections on a connection between Gothic and Iranian architecture, perhaps feeling the need to address a discussion that was widespread in the 1930s and 1940s; he refers to the two directions that this trend in scholarship took, and highlights a controversy that took place between scholars focusing on the construction of vaults.49 Overall, Seljuk and Ilkhanid presence in today’s Armenia is little studied, and monuments such as the few caravanserais that have survived since the fourteenth century have often been rather poorly preserved in the recent past, a period where the restoration of monasteries and churches has likely been deemed more important

Not for distribution or resale. For personal use only.

11

12

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

in the aftermath of the collapse of the Soviet Union. In more recent studies, however, the Mongol period in Armenia (and beyond) has been portrayed as a period of relative religious freedom, which is especially the case for the time before the Ilkhanid sultans’ conversion to Islam in 1295.50 In fact, the period of Mongol rule – despite the devastations of the initial conquests – appears to have led to a renaissance in the construction of monasteries in the first quarter of the fourteenth century.51 In Turkey, the long-neglected medieval Armenian monuments most visible at Ani, but present in various regions of Anatolia, have recently begun to receive more attention primarily from academics outside and inside Turkey, but also buttressed by recent support from local and global NGOs and the Ministry of Culture.52 Social and Cultural History As shown above, historians who tend to analyse and explain the past within the framework of political actors have dominated scholarship on medieval Anatolia. More recently, however, scholars working on this time period have turned towards social and cultural history writing. By engaging with the approaches and methodologies of both cultural anthropology and literary criticism, the work of cultural and social historians is generally seen as a response to political history. Art historical approaches are often better aligned with social and cultural history, as the study of material culture and the built environment entails engagement with a sociocultural understanding of the past that stretches beyond the facts of political history. Both cultural and social histories reconstruct dynamics of expression and tell us about daily life through analysis of the structures and processes that are often seen as the background to economic and political history. Thanks both to the development of exciting contemporary historiographical approaches and also to newly available sources, much of these new social and cultural histories of medieval Anatolia aim to augment or to provide correctives to established narratives. This turn towards cultural and social histories has led to more enthusiastic collaboration between historians and the fields of archaeology, architectural and art history. For the period before the Mongol conquest, and especially the peak of centralising Seljuk rule under Ala al-Din Kayqubad (r. 1220–37), Scott Redford has provided a wide range of studies engaging architecture, archaeology and epigraphy. In her recent dissertation and article on the same patron, Suzan Yalman has carefully explored the place of the Seljuk sultan and his patronage within medieval Anatolia, but also in a broader Mediterranean context.53 Studies on the architecture of this period have focused on the role of Sufi communities and, most recently, on Anatolia’s place within the global context of the Mongol Empire.54

Not for distribution or resale. For personal use only.

introduction

The political and social history of the time period has been understudied by Armenologists both inside and outside the Republic of Armenia. However, scholars engaged with the literature of this time period have produced exciting research that suggests, once again, the importance of interdisciplinarity and the use of a range of sources and methodologies when attempting to understand this time period. At the same time, Armenologists have made particularly good use of the relatively large amount of poetry that was composed in late medieval Anatolia and have begun to analyse it in order to shed light on social and cultural attitudes and experiences.55 An important key to better understanding the social history of late medieval Anatolia is an intellectual engagement with the large number of homosocial, semi-spiritual organisations that were housed both in structures inside city walls or in relatively remote monastic complexes. Whether self-identifying as Christian or Muslim, these associations offered members a series of prescriptions both for everyday life and for a spiritual journey to some kind of illumination. Recent scholarship on these associations has moved away from analyses limited by ethnocentrism and nationalism and towards a more meaningful pondering of the complexities of identity, social performance and the political engagement they encouraged.56 At the same time, a deeper engagement with the kind of spaces (both physical and cultural) that these associations occupied has allowed us to better understand the ways in which places were experienced during this time period. Cities and their Hinterlands Scholarship has frequently focused on the study of the cities of late medieval Anatolia. This includes Konya and Bursa first and foremost, as the respective capitals of the Seljuks and Ottomans, but also other cities such as Sivas, Erzurum and Kayseri that served as trade centres in the thirteenth century and beyond.57 Moreover, an emphasis on an urban, Persianate culture of the Seljuk realm has led to an approach that may appear skewed, yet that also reflects the places of production of chronicles in medieval Anatolia.58 For similar reasons, Byzantine studies on this time period generally focus on the study of Constantinople, Trebizond and Nicaea. Scholars of medieval Armenian history have, by and large, neglected the study of ‘Greater Armenia’ in favour of the study of ‘Lesser Armenia’ (or the Kingdom of Cilicia) and the monasteries and cities there that produced more accessible documents and also provide certain scholars with the kind of state-oriented paradigm that can make political history writing easier.59 Still, historians have studied certain cities beyond the borders of Cilicia and more recently have engaged enthusiastically with study of cities like Ani and Erzincan.60 Cities were the nodes of trade, and hence also at the forefront of

Not for distribution or resale. For personal use only.

13

14

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

networks of merchants, patrons, religious scholars (and leaders) and craftsmen who moved across the region. These individuals produced historical documents (religious texts, poetry, buildings, inscriptions, miniatures) in a range of languages and cultural or religious contexts. The fact that they left behind material that is accessible today makes them easier to trace in written sources, but when we imagine how many documents have been destroyed (either naturally or intentionally) and how many individuals did not produce materials that are accessible to us today (either because they were illiterate or because they did not make art or build lasting structures), we understand the pressing need for engagement with a broader source base that involves archaeological investigation. This will be discussed in greater detail below. Only recently has more work emerged on rural areas and central places in the hinterland of Anatolia, dominated by nomads, but also home to Muslim and Christian shrines, and villages with mixed populations, often tied to endowments in urban centres.61 In his recent book on food and food production in medieval Anatolia, Nicolas Trépanier shows how using a wide range of sources makes it possible to write social histories of this time period. Specifically, he investigates agricultural production in urban and rural contexts, as well as the preparation and consumption of the foods that are made from produce.62 In her study of the shrines of Seyyid Battal Gazi and Hacı Bekta∞, two thirteenth-century complexes constructed within Seljuk and Mongol territories, respectively, and expanded due to Ottoman patronage in the fifteenth century, Zeynep Yürekli has traced the history of two important sacred sites located beyond urban centres. By using a range of source materials, including hagiographies, architecture and chronicles, she analyses the ways in which the cultural place of these structures evolved as the political contexts around them changed from fluid frontier zones to integrated regions of a centralised (Ottoman) empire.63 Effectively, many shrines that we think of today as urban, such as the Mevlana Complex in Konya, were built extra muros and were thus part of a liminal space between the city and its hinterland, accessible to both urban and rural populations.64 In this context, the dynamics of and perceived boundaries between urban and rural spaces and their populations need to be reconsidered, as well as how, when and why these dynamics changed over time. Scholars of Byzantium have engaged with rural-oriented histories for more than fifty years. In 1958, Paul Lemerle published his Agrarian History of Byzantium, and the field has engaged in several waves of activity.65 With regard to the history of late medieval Anatolia, Helen Ahrweiler’s work on thirteenth-century Smyrna (Òzmir) and Antony Eastmond’s oeuvre on Trebizond (Trabzon) offer great insights into the ways in which Byzantine urban elements engaged with the hinterlands and also with Turkish peoples and

Not for distribution or resale. For personal use only.

introduction

cultures.66 A recently published monograph by Sharon E. Gerstel, Rural Lives and Landscapes in Late Byzantium: Art, Archaeology, Ethnography, suggests that the turn in medieval Anatolian studies towards an interest in daily life and cultural production is not unique to individuals using primarily Turkish- and Persian-language sources. One would hope that as the boundaries and borders between Art History and History begin to blur, so might the borders between Byzantine Studies and other histories of late medieval Anatolia. Sources: Strengths and Weaknesses For many years, scholars were discouraged from studying medieval Anatolia because of the supposed lack of sources available.67 While it is true that the kind of sources available have produced a longstanding challenge for any art historian or historian, this particular sense was amplified by the popularity of political history writing in that there are few sources related to dynastic entities (as compared with other regions and/or time periods). The reality is that there are many sources, but of different genres and composed in a dizzying number of languages. Scholars have more recently begun to look at sources across languages, genres and fields and to consider them in creative ways in order to write different kinds of histories. The field of archaeology, for example, has proven to be instrumental to scholars in terms of providing novel kinds of material evidence related to the study of medieval Anatolia. In the 1940s, Köprülü pointed to the lack of available sources, and suggested that vast amounts of material may still be hidden in manuscript libraries.68 Even though more manuscripts have been discovered since his time,69 and others have been edited in recent years, major texts such as the full-length version of Ibn Bibi’s history of Seljuk Anatolia up to the 1280s are not yet available in critical editions or translations.70 At the same time, it has become clear that expectations about sources that grew out of the study of the central Islamic lands need to be revised when it comes to medieval Anatolia: thus, chronicles of single cities are scarce; and biographical dictionaries of regional ulema simply do not exist prior to the sixteenth century, although several recent studies have attempted to reconstruct relevant information from other sources and later compilations.71 Other sources such as waqf documents, hagiographies and even monumental inscriptions (further discussed below) have become crucial for the re-evaluation of medieval Anatolia. Moreover, the wide array of languages (Arabic, Persian, Turkish, Greek, Armenian, Georgian, Syriac and others) that can be used is a major challenge and only collaborative projects can bridge the divides imposed by linguistic boundaries. Sufi hagiographies as sources for Islamic history are relatively

Not for distribution or resale. For personal use only.

15

16

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

new even for regions outside Anatolia, and our field would greatly benefit from engagement with scholarship that takes these texts seriously, and takes into consideration the limitations and possibilities of a genre that is not interested in straightforward history writing.72 Byzantine studies, for example, has a long tradition of editing, translating and using hagiographies to analyse the ways in which Byzantine society engaged with religious practice and various aspects of morality, especially considering that, as Kazhdan pointed out, ‘the most edifying genre of literature was, at the same time, the most entertaining’.73 In fact, as early as 1968, Evelyne Patlagean (a student of Claude Lévi-Strauss) published a paradigm-shifting article that demonstrated just how useful hagiographies could be in the writing of social histories.74 More recent work has shown the ways in which this genre developed over time; in fact, for the time period that concerns us, hagiography writing changed substantially.75 A difficulty in assessing the material culture of medieval Anatolia lies both in the lack of sources that discuss workshops, craftsmen and architects, and in the absence of texts that address aesthetic and visual understanding among contemporaries. This poses a central methodological problem for art historical studies, as style becomes an essential tool in assessing art and architecture. Written sources of the period, such as chronicles and hagiographies, do not inform us about the creators of the monuments, only about their patrons. Occasional mentions of prominent buildings or of major projects, such as the reconstruction of a city’s walls appear, yet they are neither coupled with descriptions of the monuments, nor with judgement on their aesthetic merit. Still, we know little about the reception of these monuments at the time of construction, and can only infer which aspects of a structure might represent a prevalent local or regional style and which aspects we might attribute to the tastes of a specific patron. Rare signatures of architects and craftsmen, carved into the stone of portals or into the wood of mosque furniture, are the only hints at the identity of those who created these monuments. A similar case is to be made for manuscripts, where colophons are relatively rare. (Although the Armenian example paints a different picture within which colophons are much more prevalent and have been used relatively frequently in scholarship.76) Even though nisbas (such as al-Qunawi, al-Balkhi, etc.) at the end of names in signatures often refer to places, this should not necessarily be pinned to the immediate place of a specific figure’s origin.77 Thus, we do not know whether Qaluyan al-Qunawi, a possible ­master-builder whose name appears on several mid-thirteenth-­ century monuments in central Anatolia was indeed from Konya, as his nisba suggests, trained there or had a remote family connection.78 Similarly, the Sufi master Rumi (as he is most commonly known in English) is named on his cenotaph as Muhammad ibn Muhammad

Not for distribution or resale. For personal use only.

introduction

ibn al-Husayn al-Balkhi. This usage refers to the family’s origin in Balkh in today’s Afghanistan, and eschews the name Jalal al-Din Rumi, the known modern Persian and Arabic version of the name, while in Turkish, he is usually referred to as Mevlana.79 The study of monumental inscriptions is fundamental to an analysis of medieval architecture around the Mediterranean. Studies published as part of a recent volume co-edited by Òrvin Schick and Mohammad Gharipour place increased importance on epigraphic texts that are not historical inscriptions (poetry, Qur’anic passages, for instance) across the Islamic world.80 These texts were mostly disregarded in the epigraphic surveys of the early twentieth century, and still need to be documented in many cases. Irene Bierman’s groundbreaking work on the visual and propagandistic language constructed by inscriptions in Fatimid Cairo seems to have had far-reaching influence.81 In the Byzantine context, Amy Papalexandrou has shown how inscriptions function as material objects, and a recent collection of essays on inscriptions in the eastern Mediterranean edited by Antony Eastmond includes chapters on medieval Anatolia.82 Monumental inscriptions are central to all studies of the architecture of medieval Anatolia as they are the written source most closely related to each monument: written on buildings, these inscriptions are texts as much as works of art. Often, these texts are even the single source on a specific patron, or can otherwise provide details that are absent from the written record that we more conventionally regard as such, such as manuscripts. J. Michael Rogers opened up this vein of research for Anatolia in his article on epigraphy as a source on patrons and their waqfs.83 The Répertoire chronologique d’épigraphie arabe (RCEA), edited beginning in 1931 with significant sections from medieval Anatolia, has been an important source for historians of the region. Recent detailed studies of the full epigraphic programmes of city walls, first and foremost for Sinop by Redford, and for Antalya by Redford and Leiser, have shown how monumental epigraphy can be considered a particularly productive source as it can be analysed both from a textual perspective as well as from a material standpoint.84 Calligraphers would have created templates to be used on the monuments, translated from paper into different media. Carved in stone or inlaid in tile mosaic, foundation inscriptions and other monumental inscriptions in medieval Anatolia are almost exclusively in Arabic (very few exceptions exist, and are often bits of Persian poetry). Qur’anic passages are important elements in overall programmes of inscriptions, yet they have not been well studied, and are neglected in many of the epigraphic surveys. These observations also point to the problem of literacy, and the question of how widespread an understanding of the formulaic Arabic in these texts might have been in Anatolia, where Greek, Armenian and Turkish were dominant, while Persian was the literary language

Not for distribution or resale. For personal use only.

17

18

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

of the Seljuk court. As is widely accepted, Arabic was the principal language of Islam, and the converted section of the population would have known enough Arabic to complete their daily prayers and participate in other rituals, but this was unlikely to entail the literacy necessary to decipher inscriptions. Hardly any archival documents are available for pre-Ottoman Anatolia. The exception is a corpus of endowment deeds (waqfiyyas). These documents belong to Islamic foundations (waqf) that were established and served to ensure the maintenance of a building, to provide an income to caretakers and other employees (such as the imam of a mosque), and to carry out charitable tasks such as feeding the poor in a city or neighbourhood. Endowment deeds were written down in order to specifically delineate the property that belonged to such a foundation, the salaries to be paid and the charitable functions to be carried out. In medieval Anatolia, these legal documents were often drawn up a few years after the completion of a monument and may reflect a status quo at the time of composition, rather than the endowment’s ‘ideal’ state. As legally binding documents, these deeds had to be approved by a kådÈ (judge of Islamic law) in the presence of several witnesses, who signed the bottom of the document. Thanks to their detailed descriptions of property – urban real estate, gardens, fields and sometimes entire villages – these documents offer a glimpse of the wealth of a patron, but also give great insight into urban spaces that have often not been preserved. Our record of the medieval urban Anatolian fabric is far from complete, but endowment deeds can at least give us insights into the scale of properties, and sometimes mention monuments that are no longer extant.85 The few thirteenth- and fourteenth-century examples of these deeds that have been preserved in their original form are scrolls, unfurling on a length of several yards as the writing appears on their short side. These examples are very rare, and many other such documents survive only in later copies, written into the journals (defters) of the Ottoman administration. These copies date from the sixteenth into the early twentieth century and show the longevity of many of these endowments. The last ones were abolished in the 1920s as part of the secularisation movement in the newly founded Republic of Turkey. Landscape and Architecture in Medieval Anatolia Considering medieval Anatolia through the lens of landscape and architecture offers us an opportunity to intentionally remove a focus on any singular polity or collectivity and to reposition our lens on the webs between people and the spaces and places around them. While a conversation on space and place as originally conceived by Lefebvre was associated primarily with post-capitalist societies, when imagined as a bit of a wider net, it is a particularly useful

Not for distribution or resale. For personal use only.

introduction

framework for integrating fields and for cross-disciplinary activities. A central point of this volume is to move beyond identity politics and towards understanding the ways in which individuals interacted with and understood the spaces around them and the places they occupied by considering both a range of sources of analysis and different methodological approaches. An important goal of the volume is to ensure the applicability of the studies, specific to Anatolia, and informed by a thorough knowledge of its social, cultural and political history, to larger theoretical conversations across medieval studies. The chapters gathered in this volume, each focusing on a distinct problem, region or medium related to medieval Anatolia, move in this direction. The chapters in this volume cover a wide range of subjects and represent a cumulative of research composed using a diverse array of materials and languages. All these chapters, however, are linked by their insistence upon a reconsideration of various aspects of medieval Anatolian history and art history and the concern of their authors to reconfigure aspects of the ways in which the myriad regional landscapes were constructed and experienced. The authors of the chapters in this volume question the (im)permeability of boundaries (spatial, linguistic, religious), consider the kind of communications that took place between different groups of people (whether understood by the language(s) they spoke, by the association to which they belonged or by their craft or trade) and the architecture they produced, and show how architectural styles were created using an array of creative traditions and material elements. Each chapter is, in essence, a microhistory. Each chapter attempts to understand and contextualise the individual circumstances behind the construction of the particular geography under consideration. In his work on stone and brick, Richard McClary uses the monuments he analyses in a new fashion; he considers them as testaments or memorials to the craftsmen who built them. Through his skilled analysis, he is able to show that in the late twelfth and early thirteenth centuries, local stone masons (most likely Byzantine or Armenian) worked hand-in-hand with newly arrived Muslim craftsmen and that, together, they created a new architectural aesthetic in central and eastern Anatolia. In a close study of construction techniques and building phases in several thirteenth-century caravanserais, Cinzia Tavernari shows that these monuments were carefully constructed prestige projects. Thus, unlike in Ayyubid Syria, contemporary caravanserais were often quickly built brick constructions, the monuments built under Seljuk royal patronage were as much a mark of presence as an attempt to create unified infrastructure for trade. From this point of view, Tavernari argues, caravanserais in the first quarter of the thirteenth century were part of a centrally planned initiative that did not, however, extend to other types of monuments.

Not for distribution or resale. For personal use only.

19

20

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

The relationship between Sufis, akhis and the city has long been a topic of scholarly scrutiny. While many attempts to understand this · complex web of relationships have failed, Iklil Selçuk new research on land grants in Bursa – buttressed by a detailed contextual framework – sheds important light on the relationship between akhis, Sufis and the Ottoman enterprise in the fourteenth century. Selçuk posits that under the Ottomans, the akhis lost power early on, suggesting that their social and trade functions were quite different from those of the local Sufis. In the subsequent chapter, Rachel Goshgarian postulates that performing manliness was central to the medieval Anatolian experience of futuwwa, in the same way that performing Islam (as opposed to Christianity) was central to futuwwa participation. Still, the definitions of both masculinity and Islam (and, the ways in which they were performed) were filled with ambiguities and contradictions that, ultimately, made them less than exclusive and allowed both individuals and organisations a certain degree of freedom. Considering both current and art historical narratives on the Byzantine Cappadocian site at Açıksaray, Fatma Gül Öztürk suggests that the boundaries between the secular and the sacred might not have been as distinct as architectural historians have previously thought. Suggesting that Byzantine structures in Cappadocia were ‘shared spaces’, used by both Muslims and Christians (particularly in the thirteenth and fourteenth centuries), her chapter insists upon the notion that ethnic and religious differences might not have been experienced as exclusive boundaries in the landscape of Açıksaray. Mattia Guidetti’s contribution engages with the ways in which communication and exchange might have taken place between Mosul and Anatolia and between Armenians and their Muslim neighbours. His complex and new interpretation of ‘Islamic’ elements in the early thirteenth-century Armenian construction programme at Ani insists on something more than an idea of ‘influence’. In fact, Guidetti suggests that certain members of the Armenian population at Ani (and, possibly, elsewhere) were invested in engaging with what Guidetti calls ‘outerness’, and intentionally altered what was considered a traditional form of Armenian architecture in order to illustrate this liminality. His chapter suggests that Armenians in Ani actively participated in the ‘Islamisation’ of the thirteenthcentury Anatolian landscape. Working from a philological standpoint, Nicolas Trépanier paints an exciting picture of the ways in which Arabic, Persian and Turkish co-existed in the linguistic landscapes of medieval Anatolian Muslims in his analysis of various terms used to describe gardens. His research suggests that the Turkish incorporation of various Persian terms (and complete rejection of any Arabic ones) allude to a kind of bridge-building between the new literary Turkish and the

Not for distribution or resale. For personal use only.

introduction

Persian linguistic elitism thought to have permeated all areas of the Anatolian Muslim world. Patricia Blessing’s chapter shows that the practices associated with patronage and construction in post-Mongol central and eastern Anatolia must be applied to the ways in which early Ottoman architecture developed in western Anatolia. Her contribution highlights the important interactions that took place between regional architecture, local styles and fragmented patrons. Blessing illustrates the important reality that the Ottomans were a Beylik quite similar to other Beyliks in the late medieval period: a centralised architectural aesthetic did not exist and much of what was designed took place on the ground, between a single patron and a team of skilled craftsmen and local labourers. Finally, Suzan Yalman analyses the ‘complex’ identity of Mahperi Hatun, the Armenian or Georgian wife of Kayqubad I and mother of Kayqubad II. Her contribution suggests that the ‘dual identity’ paradigm (recently elaborated by Rustam Shukurov) is essential to understanding the ways in which the Seljuk elites experienced the social and religious landscapes in which they operated. In the case of Mahperi Hatun, it is also clear that the message of her patronage projects is one that reduplicates or re-emphasises this complex framework of identity. Yalman tells us, however, that Mahperi was not alone in this kind of endeavour, although she was a unique female actor of the thirteenth century.

Notes 1. Torczyner, L’ami Magritte, p. 71. 2. Lefebvre, La production de l’espace. 3. Factography (Rus., faktografiia) is a term used to describe a kind of history writing that became popular in the late Soviet period and was characterised by a focus on, and description of, facts and minutiae without analysis. A positivistic intellectual endeavour linked to expressions of nationalism, Soviet faktografiia saw intellectual inquiry into the past as part of ‘a collective search for origins and identities’. Kozlov, ‘The Historical Turn in Late Soviet Culture’. 4. See, for instance, NecipoÌlu, ‘The Concept of Islamic Art’. 5. Nouvelle histoire was developed by Jacques LeGoff and Pierre Nora in the late 1970s. This version of Annales history is most closely linked with the histoire des mentalités, history of representations and cultural history. A response to the notion of histoire événementielle, the goal of nouvelle histoire is to write ‘good’ history and to avoid the superficial. See LeGoff, La nouvelle histoire, introduction. 6. Kafadar, ‘A Rome of One’s Own’, p. 8. 7. For this aspect of historiography, see PancaroÌlu, ‘Formalism and the Academic Foundation of Turkish Art’. 8. Kaufmann, Toward a Geography of Art, p. 9. 9. Bedrosian, ‘The Turco-Mongol Invasions and the Lords of Armenia’, p. 91, n. 161.

Not for distribution or resale. For personal use only.

21

22

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

10. For a fascinating analysis of the ways in which ‘Anatolia’ has been linked to nationalist and modernist discourses in the Republic of Turkey, see Bilsel, ‘“Our Anatolia”’. 11. BozdoÌan and NecipoÌlu, ‘Entangled Discourses’; Kafadar, ‘A Rome of One’s Own’. 12. Campbell, ‘Artistic Geographies’: Castelnuovo and Ginzburg, ‘Symbolic Domination and Artistic Geography’. 13. Kafadar, ‘Rome of One’s Own’; Goshgarian, ‘Diversity in the Medieval Middle East’. The comparison with medieval Spain is suggested in Peacock, de Nicola and Yıldız, ‘Introduction’, pp. 13–16. For studies on other medieval frontier regions, see Flood, Objects of Translation; Robinson, Imagining the Passion; Dodds, Balbale and Menocal, The Arts of Intimacy. 14. Castro, España en Su Historia. 15. Nirenberg, Communities of Violence, see introduction. 16. Cahen, Pre-Ottoman Turkey. This first edition, largely published without references and a limited bibliography, was the basis for an updated French version, published posthumously with the help of Cahen’s notes. That second version was translated in 2001 and provides the most detailed account of Seljuk Anatolia available in English. 17. Cahen, The Formation of Turkey. 18. Vryonis, Decline of Medieval Hellenism, p. 354. 19. Vryonis, Decline of Medieval Hellenism, p. vii. 20. Strohmeier, Seldschukische Geschichte, pp. 152–9. 21. Turan, Selçuklular zamanında Türkiye. For controversy on the subject · with Ibrahim KafesoÌlu, see KafesoÌlu, A History of the Seljuks. 22. Yıldız, ‘Mongol Rule in Thirteenth-century Seljuk Anatolia’. 23. Uzunçar∞ılı, Anadolu beylikleri ve Akkoyunlu. 24. S¸entürk, Asya’dan Anadolu’ya rüzgar. 25. Woods, Aqquyunlu. 26. Angold, A Government in Exile; Nicol, Despotate of Epiros; Bryer, Empire of Trebizond and the Pontos. 27. Laiou, Constantinople and the Latins; Kazhdan, ‘Italian and Late Byzantine Cities’. 28. El-Cheikh, Byzantium as Viewed by the Arabs; NecipoÌlu, Byzantium between the Latins and the Ottomans; Estangui Gomez, Byzance face aux Ottomans. 29. Bedrosian, ‘The Turco-Mongol Invasions and the Lords of Armenia’. 30. Eastmond, Eastern Approaches to Byzantium. 31. Wittek, The Rise of the Ottoman Empire. 32. Köprülü, Les origines de l’empire ottoman. On the relationship between the two scholars, see Lindner, ‘Wittek and Köprülü’. 33. Lindner, Nomads and Ottomans. 34. Imber, The Ottoman Empire. 35. Kafadar, Between Two Worlds. 36. Lowry, The Nature of the Early Ottoman State. 37. Fleet, review, Heath Lowry, The Nature, p. 683. 38. Gabriel, Monuments Turcs; Redford, ‘“What Have You Done for Anatolia Today?”’, p. 244; Redford, ‘Albert Gabriel’, pp. 80–1. 39. Gabriel, Voyages archéologiques; Gabriel, Une capital turque: Brousse. 40. PancaroÌlu, ‘Formalism and the Academic Foundation of Turkish Art’. 41. Strzygowski, ‘Türkler ve Orta Asya san‘atı meselesi’; Redford, ‘“What Have You Done for Anatolia Today?”’, n. 5.

Not for distribution or resale. For personal use only.

introduction

42. PancaroÌlu, ‘A Fin-de-Siècle Reconnaissance’; Blessing, ‘Friedrich Sarre’. 43. Blessing, ‘Medieval Monuments from Empire to Nation-State’. 44. This was the product of a general scholarly notion that the cultures of Armenia and Iran were so linked that they should be considered sisters. In fact, until 1875, scholars thought that Armenian was a dialect of Persian, see Hubschmann, ‘Ueber die Stellung des Armenischen’. 45. Pope, ‘Possible Iranian Contributions’. 46. Strzygowski, Die Baukunst der Armenier und Europa. 47. Maranci, Medieval Armenian Architecture, pp. 26–37. 48. Strzygowski, Die Baukunst der Armenier, vol. II, p. 576; Strzygowski, ‘Beiträge zur Kunstgeschichte des Mittelalters von Nordmesopotamien, Hellas und dem Abendlande’, p. 343. 49. Wilber, The Architecture of Islamic Iran, p. 61; further on Wilber, see Blessing, ‘From the Survey of Persian Art to the CIA’. 50. Kouymjian, ‘L’architecture en Grande Arménie (milieu XIIIe–milieu XVe siècle)’, pp. 300–4. 51. Margarian, ‘La Grande Arménie (XIIIe–XVe siècle)’, pp. 295–9. 52. See at: http://www.virtualani.org, last accessed 24 May 2016; Kalas and Özkaya, ‘The Georgian Aspects of Medieval Architecture at Ani’; Watenpaugh, ‘Preserving the Medieval City of Ani’. 53. Yalman, ‘Building the Sultanate of Rum’; Yalman, ‘Ala al-Din Kayqubad Illuminated’. 54. Wolper, Cities and Saints; Blessing, Rebuilding Anatolia. 55. Van Lint, ‘Kostandin of Erznka’; Russell, Yovhanne¯ s T‘lkuranc‘i; Cowe, ‘Models’. 56. See especially articles collected the following edited volumes: Muqarnas 24 (2007), eds BozdoÌan and NecipoÌlu; Beyazit, At the Crossroads of Empires; Asutay-Effenberger and Daim, Der Doppeladler; Peacock, de Nicola and Yıldız, Islam and Christianity in Medieval Anatolia; Peacock and Yıldız, Islamic Literature and Intellectual Life in Fourteenth- and Fifteenth-Century Anatolia. Further relevant publications are cited in individual chapters and collected in the bibliography at the end of this volume. 57. Rogers, ‘Patronage in Seljuk Architecture’; Konyalı, Konya tarihi; Wolper, Cities and Saints; Yalman, ‘Ala al-Din Kayqubad Illuminated’; Blessing, Rebuilding Anatolia. 58. Hillenbrand, ‘RåvandÈ’; Melville, ‘The Early Persian Historiography’. 59. See Bedrosian, ‘The Turco-Mongol Invasions and the Lords of Armenia’, introduction. 60. Van Lint, ‘Kostandin of Erznka’, introduction; Goshgarian, ‘Futuwwa in Thirteenth-century RËm and Armenia’. 61. PancaroÌlu, ‘Caves, Borderlands and Configurations of Sacred Topography’. 62. Trépanier, Foodways. 63. Yürekli, Architecture and Hagiography. 64. Blessing, Rebuilding Anatolia, pp. 61–2. 65. Lemerle, The Agrarian History of Byzantium. 66. Eastmond, Art and Identity in Thirteenth-century Byzantium; Ahrweiler, ‘L’histoire et la géographie’. 67. Leiser, ‘The Seljuks of Anatolia’. 69. See, for instance, the project The Islamisation of Anatolia, c. 1100– 1500, funded by the European Research Council, 2012–16, directed by

Not for distribution or resale. For personal use only.

23

24

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Andrew C. S. Peacock at the University of St Andrews, available at: http://www.islam-anatolia.ac.uk, last accessed 10 November 2015. 70. Yıldız, ‘Mongol Rule in Thirteenth-century Seljuk Anatolia’, pp. 433–40. 71. Marlow, ‘A Thirteenth-century Scholar in the Mediterranean’; Nassiri, ‘Turco-Persian Civilization’; Atcil, ‘The Formation of the Ottoman Learned Class’. We thank Cailah Jackson for suggesting these references. 72. Karamustafa, ‘Origins of Sufism in Anatolia’. For work on Central Asia and Iran, see DeWeese, Islamization and Native Religion in the Golden Horde, and Bashir, Sufi Bodies, along with multiple articles by both authors. 73. Kazhdan, ‘Byzantine Hagiography and Sex’, p. 131. 74. Patlagean, ‘Ancienne hagiographie byzantine’. 75. A two-volume series, The Ashgate Research Companion to Byzantine Hagiography, was published in 2014. It traces the evolution of the genre from the fifth to the fifteenth centuries, and offers great insight into the ways in which historians have used texts composed in this genre (written in Arabic, Armenian, Byzantine Greek and Syriac) to uncover a great deal about various aspects of the Byzantine past: Efthymiadis, Byzantine Hagiography. 76. Sanjian, Armenian Colophons. 77. See Redford, Landscape and the State, and Trépanier, Foodways. 78. Brend, ‘The Patronage of Fa¢r al-Din ‘Ali ibn al-Husain’; Rogers, ‘Patronage in Seljuk Architecture’, pp. 166–7 and 445–7; Rogers, ‘The Çifte Minare Medrese’, pp. 80–1; Blessing, Rebuilding Anatolia, pp. 50 and 115–17. 79. Combe, Sauvaget and Wiet, Répertoire chronologique (RCEA), No. 4681; for a discussion of the inscription, see Blessing, Rebuilding Anatolia, pp. 64–5. 80. Gharipour and Schick, Calligraphy and Architecture. 81. Bierman, Writing Signs. 82. Papalexandrou, ‘Echoes of Orality’; Eastmond, Viewing Inscriptions. 83. Rogers, ‘Waqf and Patronage’. 84. Redford and Leiser, Victory Inscribed; Redford, Legends of Authority. 85. Trépanier, Foodways; Wolper, Cities and Saints; Blessing, Rebuilding Anatolia.

Not for distribution or resale. For personal use only.

CHAPTER TWO

Craftsmen in Medieval Anatolia: Methods and Mobility Richard P. McClary

The following analysis of the two primary materials, stone and brick, and the methods of combining and manipulating them, is used to delineate a methodological framework for determining the processes required to build large-scale construction projects. The specific focus here is the situation in Anatolia in the late twelfth and early thirteenth centuries, but the principle of analysing the constituent elements of buildings in order to understand the working methods of the craftsmen is applicable to much of the architecture of the premodern world. In lieu of any surviving technical treatise, the detailed study of techniques and decorative motifs can provide a wider understanding of the training, geographic origins and mobility of the craftsmen responsible for the construction of the Islamic architecture of medieval Anatolia.1 Stylistic and comparative analysis with structures in the wider Islamic world, from Iran and Central Asia to Syria and Egypt, can provide evidence for assertions with regard to the origin and training of a number of the craftsmen working in Anatolia in the twelfth and thirteenth centuries. Aside from the indigenous building traditions, two clearly discernible, external sources were integral to the development of a unique Anatolian architectural aesthetic in the first quarter of the thirteenth century. First, stereotomic stone-working methods, developed primarily for marble construction in Zengid and Ayyubid Syria in the late twelfth and early thirteenth centuries, can be found in Konya. Second, the brick and glazed-tile building style that flourished in Iran under Ildegüzid patronage in the twelfth century was used in a number of surviving buildings across Anatolia.2 This chapter presents evidence of craft practices gleaned from a close analysis of buildings in order to understand the working methods, mobility and origins of craftsmen active across Anatolia. The primary focus is on buildings erected between 1170 and 1220, and mainly involves buildings with RËm Seljuk connections because the surviving corpus of buildings is primarily associated with their patronage. However, the mobility of craftsmen, as well as the fluidity

Not for distribution or resale. For personal use only.

28

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

of boundaries and political alliances, means that the results also apply to a number of the surviving SaltËkid and Mengüjekid structures in central and northeastern Anatolia. The strong influence of, and in some cases on, the indigenous Armenian and Byzantine traditions,3 alongside the evidence of Persianate brick and glazed-tile structural and decorative techniques is clear. This chapter presents a new understanding of the processes required to extract, manufacture and combine the constituent materials into the extant corpus of buildings. This approach, not seen in previous scholarship, is achieved through the examination of a number of little-known structural and decorative details and a study of the constituent materials. The surviving structures may be viewed as records of, and memorials to, the anonymous craftsmen and labourers who built them, as much as to the elite patrons who paid for them. Craftsmen working on the most prestigious commissions, commissioned by royal patrons and involving a wide range of materials and decorations had to move from one site to the next upon completion of a project. It is possible that it was local resident builders who built vernacular structures such as residential buildings or small neighbourhood mosques and shrines in the larger urban context. The focus here is on the peripatetic master craftsmen who were responsible for the majority of the surviving corpus of elite buildings. It was this group of individuals4 who set the formal and aesthetic tone for the builders of structures across the region, including those that were less prestigious. When studying the pre-Mongol architecture of Anatolia, scholarship is inevitably drawn to the smallest proportion of structures (namely, those with elite patrons) as it is those that remain extant. It is important to keep in mind that this group of buildings did set the tone for the wider corpus. Unfortunately, most architecture from this time period has not survived; and it is impossible to know the total number of monuments that were built during the late twelfth and early thirteenth centuries. Workshop Organisation and Tools There is limited archaeological evidence upon which to form an understanding of the working methods of tile-makers, or indeed any craftsmen working in Anatolia during the medieval period. In fact, much of what we do know about the methodologies and practices of craftsmen comes from written sources. During the twelfth and thirteenth centuries, most of the raw materials necessary for building would have been transported to construction sites, often over considerable distances.5 Consequently, study of the working methods and mobility of craftsmen cannot be separated from an examination of the process of extraction and transportation of the materials required to build the extant structures. Moreover, production facilities, such as kilns and forges, would have been built anew at each

Not for distribution or resale. For personal use only.

craftsmen in medieval anatolia

major construction site as the craftsmen moved around the region. The number of different processes, and the organisation required to coordinate the various skilled craftsmen involved, attests to the complexity of the management of large-scale construction projects and the skill of the masters responsible for overseeing them. An example of an on-site production facility is the small threeroomed rectangular workshop that Arık excavated at the Bey∞ehir Palace site (c. 1235) in southwest Anatolia in 1986. Located to the south of the palace complex, this workshop was found to contain blacksmiths’ tools, wasters, ash and glaze slag, along with the remains of forges, a kiln and a lime store.6 This rare example of an industrial workshop that was excavated on-site can be used as firm evidence for the presence of itinerant masters who worked in temporary workshops on major construction projects in the early part of the thirteenth century,7 as a palace would not be the site of an ongoing production facility once completed. The on-site production of ceramic tiles was introduced in the thirteenth century, as there is clear evidence of the importation of such items in the late twelfth century. We can observe the architectural use of limited numbers of glazed bowls, at the Mengücek Gazi Tomb in Kemah (c. 1190) and the Kamereddin Tomb in Divri©i (1196–7),8 as well as imported Kashan tiles from Iran at the Kilij Arslan II Kiosk (c. 1174) in Konya. These imported examples indicate that there was no on-site production in the twelfth century, and that importation was the only way of sourcing glazed elements for use in an architectural context in the early period of the development of Islamic buildings in Anatolia. The organisation of the business of supplying and transporting materials remains poorly understood because of the lack of written sources. Occasional brief passages in traveller’s accounts give insights into the working methods of medieval craftsmen in the Islamic world. Nasir-i Khusraw, a Persian poet and traveller writing in 1047, described seeing marble being cut with a toothless saw in Syria, with ‘Mekkah sand’ used as an abrasive for the blade.9 Nearly two centuries later, the chronicler Ibn al-Athir (d. 1233) mentions a large timber yard in Baghdad, and it is likely that other major cities would have had similar stores and timber merchants.10 Such a facility would have bridged the gap between the harvesting and processing of the raw material, on the one hand, and the supply of timber for the construction trade, on the other hand. Still, however, little is known about the process of gathering, transporting, preparing and distributing such construction materials. Similarly, very few tools survive from the Seljuk period. A rare example of a plasterer’s trowel is currently held in the Arkeoloji Müzesi (Archaeology Museum) in Sivas. It is likely to have been the type of tool used to apply a smooth-finish coat of plaster to walls (Figure 2.1). The rectangular body is made of fired clay, with a roughly formed handle and a smooth turquoise glazed underside.11 The use

Not for distribution or resale. For personal use only.

29

Not for distribution or resale. For personal use only.

Figure 2.1  Glazed terracotta plasterer’s trowel, Arkeoloji Müzesi, Sivas (photographs: Richard McClary)

craftsmen in medieval anatolia

of glaze on a tool is very unusual, although it would have allowed for the consistent creation of a very smooth surface, and would also have been easy to keep clean. This plasterer’s trowel illustrates the diversity of glaze usage in late medieval Anatolia. Alongside the more common use on vessels and architectural revetment, it demonstrates an entirely practical application of the medium, which has no decorative function whatsoever. The use of a turquoise glaze in this purely functional context suggests that, alongside being the most common in the early period, it was most likely the cheapest and easiest to produce. There was a need for large numbers of tiles with tonal consistency, and turquoise glazed wares were the pottery of common use, being inexpensive and produced at a large number of sites.12 Therefore, it makes sense that turquoise was the most common glaze colour employed in architectural revetment as well. Muslim and Christian Craftsmen A craftsman’s nisba, a moniker (or epithet) that can indicate birthplace (of an individual or his family), place of residence, vocation, tribe or school of law, is rarely proof of his place of training or birth.13 A nisba can, however, offer useful secondary evidence when the stylistic elements suggest strong links with a certain region or city. As will be discussed below, close stylistic and formal connections between the architecture of Iran and Anatolia demonstrate the westward movement of individuals trained in the brick-building and glazed-tile decorating traditions of the areas to the east. In addition, stylistic evidence and the signature of Muhammad ibn Khawlan alDimashqi on the north wall Konya’s Citadel Mosque (also known as the Alaeddin Mosque) point to the introduction of Aleppine hardstone building techniques in the early thirteenth century.14 There is no evidence of Syrian masons having worked in Anatolia prior to that date.15 This would suggest a relatively early and fluid movement of artisans and styles between Anatolia, Iran and Syria. Despite a dearth of names and masons’ marks pertaining to Armenian craftsmen on the Islamic buildings of Anatolia, the units of measurement used in both Islamic and Armenian architecture are comparable.16 However, absence of evidence is not, in this case, evidence of absence: Armenians could still have been involved in the construction of stone buildings in the region since most buildings feature only one craftsman’s name, if any, and not all participating workers would have been named on any given monument. Given the numerous similarities between Armenian church architecture and the architecture commissioned by the Turco-Muslim elites of Anatolia – especially the tombs, such as the Mama Khatun Tomb (c. 1200) near Tercan – it is quite likely that Armenian-trained stonemasons were involved. In addition, it is very unlikely that during the early period the Turkic nomads would have had the requisite

Not for distribution or resale. For personal use only.

31

32

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

stone-working skills, and those with expertise in building would have come from a region with a long tradition of brick rather than stone construction. It is possible that Christian workers, both skilled and unskilled, were recruited and operated under the direction or fiduciary control of a Muslim amir, or an architect, referred to by the Persian chronicler of the RËm Seljuks, Ibn Bibi, as mimårån-i kår-dån.17 It remains to be seen whether or not the lack of Christian craftsmen’s names on buildings was a result of their dhimmÈ status, as the guild system, site management and architectural design processes in Anatolia during the twelfth and thirteenth centuries remain clouded in uncertainty.18 A process of exchange was in place between the various architectural traditions in the region. One of the clearest examples of the transfer of motifs and techniques from what is generally considered the Islamic tradition into Armenian architecture can be seen at the church of St Hovhannes Mkrtitch (St John the Baptist), in Gandzasar. The church is located 40 km north-northwest of Stepanakert, in the disputed territory of Nagorno-Karabakh. Construction of the church started in 1216, and was completed by 1238.19 The building has several identifiably Islamic elements on the interior and exterior, with the eastern façade of the church having two recesses crowned with polylobed arches featuring trefoil pendants (Figure 2.2, middle). These are almost identical to the examples employed in wood over the top of minbar doors of the twelfth century, an example being the Kilij Arslan II minbar (1155) in Konya (Figure 2.2, bottom). A similar motif, in stone, can be seen on arches in the courtyard of the Zinciriye Madrasa in Diyarbakır, built in the late twelfth century. In addition, the internal walls flanking the altar have steps running up the side, in the manner of those up to the entrance of the Quraysh Baba Tomb near Sinanpa∞a, Afyon (c. 1209–10). The underside of each step has been carved out to form a single muqarnas cell. At the church in Gandzasar, the exterior of the central drum section has intaglio patterns of split palmettes in rectilinear and curvilinear panels. The element of the building most closely associated with earlier Islamic structures is the square-plan shallow pyramid skylight inside, which is constructed entirely of muqarnas cells. Such muqarnas do not occur in Armenian architecture prior to the early thirteenth century. A similar muqarnas skylight survives in the south narthex (gavit) of the Church of the Holy Apostles in Ani, added in the early thirteenth century.20 This feature, above all, suggests the direct involvement of itinerant Armenian masons in the construction of earlier stone-built structures with Turco-Muslim patrons,21 the lack of epigraphic evidence notwithstanding.

Not for distribution or resale. For personal use only.

craftsmen in medieval anatolia

Figure 2.2  St Hovhannes Mkrtitch (St John the Baptist) Church, Gandzazar (1216–38) and Kilij Arslan II minbar, Konya (1155) (bottom) (photographs: Richard McClary)

Not for distribution or resale. For personal use only.

33

34

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Materials From the study of the craftsmen, our attention must turn to the study of the main constituent materials with which the ­surviving corpus of buildings was constructed. Detailed study of the materials can reveal information about the working methods of  the  craftsmen.  In late  medieval Anatolia, the materials used to construct ­buildings were, primarily, stone and brick. The incisions found in a number of decorative ashlars reveal valuable information about the early planning stage of geometric patterns, while the size and bond of brick used across Anatolia reveals the close connections to the architecture of Iran. Materials: Stone Stone, primarily limestone or sandstone, along with smaller amounts of basalt and marble, was the primary building medium of construction in the RËm Seljuk sultanate and the other Turko-Muslim polities of the region.22 This reflected the pre-existing tradition of lithic construction that existed in the region prior to the arrival of the Seljuks and the Türkmen nomads in the mid-eleventh century. The dominance of the local material, in contrast to the limited use of brick, suggests that, initially at least, most of the craftsmen were locals, the vast majority of whom were Christians.23 By adapting their own skills to the desires of the new patrons and adopting elements introduced by the Muslim craftsmen they would have been working alongside, a new and identifiably Anatolian architectural aesthetic emerged in the early thirteenth century. The technique of using a dressed-ashlar skin, with a rough inner surface to increase mortar adhesion, on either side of a rubble and mortar core was long established in the indigenous building traditions of Anatolia (namely, Armenian, Georgian and Byzantine).24 The same method was also employed to the south in northern Syria in both the pre-Islamic and Islamic periods. The stone buildings of the RËm Seljuks and other Turko-Muslim dynasties were built using the established techniques, with the exception of the stereotomic muqarnas hoods, and innovation was reserved for the incorporation of new modes of decoration, such as glazed tiles. The established structural elements, along with the innovative decorative motifs and forms, rapidly came together to create patterns and forms that became part of the architectural aesthetic of late medieval Anatolia. There are three primary phases in the stone-building process. The stone must be quarried and transported to the site, after which the stones are cut, before they are finally set. It may be assumed that cranes and pulleys would be needed to lift all but the smallest blocks from the ground. The cutting process required one set of tools, while the process of laying stones required another. For cutting, the tools required include axes, gavels, chisels, mallets and hand saws; while

Not for distribution or resale. For personal use only.

craftsmen in medieval anatolia

the stone-setters would have used hammers, large chisels and tools for winding up blocks. The individual masons would have had their own square.25 Ibn al-Razzaz al-Jazari (1136–1206), the inventor and scholar who worked as chief engineer under the Artukids in Mardin, attests to the use of plumb lines in the late twelfth century in his Kitåb fÈ ma‘rifat al-˙iyal al-handasiyya (Compendium of Science and Useful Practice in the Making of Mechanical Devices).26 The different colours seen in sandstone are caused by variations in the type of inclusions. Grey results from clay, while red, brown and yellow are due to the presence of ferric oxides in the stone.27 The softness of stone, due to its higher water content when freshly quarried, makes it relatively easy to work, and as the stone dries it hardens. This is particularly the case with the volcanic tuff that was commonly used, especially in Kayseri.28 The harsh winter weather in Anatolia precluded year-round building. In addition to the inclement working conditions in winter, low temperatures kept the mortar from setting, and the moisture content in the freshly quarried stones, which made them easier to work, also made them too brittle to work if frozen.29 The ashlars could be prepared with an axe or hammer, and finegrained sandstone and limestone could be cut with a toothed saw.30 Stone pattern design and execution Divri©i’s Citadel Mosque boasts the earliest surviving decorative portal attached to an Islamic building in Anatolia (Plate 1) and is securely dated, by the upper inscription, to 576/1180–1.31 The mosque is rectangular, with the short side facing towards the qibla. The portal is built in a hybrid style that reflects the eclectic nature of the early phase of Islamic architecture in Anatolia. The mosque was built for the Mengüjekid ruler Sayf al-Din Shahanshah (r. c. 1175–97),32 and the lintel features the signature of the craftsman in Kufic script, Hasan (?) ibn Piruz (?) al-Maraghi.33 It has the largest number of incised construction lines of all the extant stone portals in Anatolia, and it appears to be the case that, over time, lines were incised less deeply (or not in a permanent manner), particularly in the period after the death of Izz al-Din Kay Kawus I in 1219–20, the point at which a mature regional style of Islamic architecture can be seen to have emerged. The portal of the Citadel Mosque in Divri©i has two noteworthy examples of ­ construction lines. The rectilinear geometrical pattern in the top right corner of the door has clearly visible grid-lines (Figure 2.4). This was probably the first part of the decoration to be executed, as the top left corner, which features the same pattern, has far fewer visible construction lines than does the top right corner. It is likely that the craftsman, working on the blocks in situ, had worked out the underlying design required for the execution of the pattern by the time he got to the other side. A reduced-scale set of incised lines used for the planning of a muqarnas vault (and now lost) was found on the south façade of the

Not for distribution or resale. For personal use only.

35

CJ 0

E

Not for distribution or resale. For personal use only.

Figure 2.3  Citadel Mosque, Divri©i; portal cross-section at 130 cm above top step (drawing: Richard McClary)

CJ 0

E

ll)

craftsmen in medieval anatolia

Figure 2.4  Citadel Mosque portal, Divri©i (1180–1); door jamb detail (top) and portal lintel detail (bottom) (photographs: Richard McClary)

Not for distribution or resale. For personal use only.

37

38

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

gavit of the Asvatsankal Monastery in Armenia and has been cited as evidence for the use of architectural drawings in the thirteenth century.34 However, as it also contained construction lines and was located on the actual building, it was probably more like the examples of on-site design seen in the pattern and muqarnas cell designs discussed in this chapter, rather than evidence for the widespread use of architectural drawing in the architecture of the late medieval period. In addition to the corner patterns, the bottom of the tympanum above the door and below the square panels with geometric patterns (which also feature visible construction lines) is a band of incisions that is the underlying design for a strip of decoration that was never completed. The circle-based pattern consists of two rows of wavy lines (Figure 2.4). The lack of any finished pattern allows a rare glimpse into the underlying planning methods for the execution of patterns by the craftsmen responsible for one of the earliest decorative stone portals in Anatolia. The square stone panels with geometric patterns in the tympanum above the door are similar to panels on two tombs in Nakhchivan City, located 860 km east of Divri©i, in what was Ildegüzid territory and is now administered by Azerbaijan. The buildings are the Yusuf ibn Kutheyyir tomb (1162–3) and the Mumina Khatun tomb (1186–7), executed in brick rather than stone. The surrounding arch and spandrels in Divri©i are constructed with stones cut to look like bricks (Plate 1). In addition, the spandrels feature glazed inserts.35 The geometric patterns are much shallower than those on the rest of the corpus of surviving portals in Anatolia and are the only examples to have a flat rather than a curved or V-incised surface. It is also the only mosque portal in the region that has a lintel instead of a shallow arch over the door. It is only with close examination of the details of the surviving structures that small clues about working methods emerge. The mix of patterns, and even methods, developed in the brick-building tradition of Seljuk Iran with materials and techniques employed in the indigenous lithic tradition of Anatolia make this portal unique. It represents an intersection point of two traditions; east and west, indigenous and imported. As the cross-section shows (Figure 2.3), a number of features make it an important structure in our understanding of the development of a unique Anatolian aesthetic, including the facet at 45 degrees, the engaged columns and the use of intaglio rather than relief patterns. It is with this portal that its master builder (ustådh) laid the groundwork and created a stylistic precedent for a number of later craftsmen and the structures they would build. At the north edge of Divri©i, below the citadel, is the Sitte Melik Tomb (1196–7).36 The patterns on the portal of that tomb also feature a number of incised grid-lines. In addition, the surface of the pattern around the entrance has visible tooling marks (Figure 2.5). While on the one hand they provide the grid for geometrical strapwork interlace patterns, shallow surface incisions were also used to plan the

Not for distribution or resale. For personal use only.

craftsmen in medieval anatolia

Figure 2.5  Sitte Melik Tomb, Divri©i (1196–7); portal door jamb detail (photograph: Richard McClary)

carving of muqarnas cells into flat surfaces. Surviving examples of construction lines can be seen over the flanking niches of the outer enclosure portal of the Mama Khatun Tomb in Tercan (c. 596/1200) (Figure 2.6).37 The style of funerary stelae developed by the stonemasons in Ahlat emerged in the late twelfth century, following the demise of the ruling Sökmenid dynasty.38 A dwindling number of elite patrons led to a subsequent decline in the architectural patronage that had thrived there during the previous twenty years.39 Oya Pancaro©lu has argued convincingly that Ahlat was most likely the place where muqarnas cells were first translated into stone in Anatolia and subsequently introduced into the architecture of the region.40 A surviving tombstone, located in the southwest corner of the large graveyard outside Ahlat, features the same form of muqarnas cells and incised construction lines as seen in the later tomb portals at Tercan and Divri©i. There is a strong stylistic continuity in the form of the tombstones, of which several thousand survive, with dates ranging from the late twelfth century to the fourteenth century.41 The incised construction lines are clearly visible (Figure 2.7), and the form and technique can be related to those of the bands of muqarnas cells over the niches flanking the portal in Tercan. The nisba of the craftsman of the earliest surviving architectural stone muqarnas, at Divri©i and Aksaray, is al-KhilåtÈ, which, along with the lithic evidence, gives further support to the argument that Ahlat (Arab. Khilåt) was where the carving of muqarnas cells in stone developed in Anatolia.42

Not for distribution or resale. For personal use only.

39

40

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Figure 2.6  Mama Khatun Tomb, Tercan (c. 1200); upper band of muqarnas above left niche (photograph: Richard McClary)

Figure 2.7  Upper portion of a tombstone in the southwest corner of Ahlat graveyard, late twelfth century (photograph: Richard McClary)

Not for distribution or resale. For personal use only.

craftsmen in medieval anatolia

The late twelfth century represents a brief moment in time when the design process for patterns and forms that were to develop and proliferate in the following century was laid bare for posterity. Subsequently, the planning was hidden from the eyes of later observers. It is only through extremely close observation of the material that it is possible to discern the tool marks that reveal the working processes of the medieval craftsmen. Wood carvers, when laying out their geometric patterns, also employed the method of incising construction lines used by stone carvers.43 Little is known about the degree of trade specialisation during the late medieval period, but Rogers has suggested that the same people are not thought to have worked on different materials.44 That said, such similarities of technique, as well as the use of chisels, axes and saws to work both stone and wood, may go some way to explain why the name carved on to the stone-built Alay Han near Aksaray (late twelfth century) includes the epithet al-najjår (the woodcarver).45 It may be the case that said individual had taken on a supervisory role, but he might well have been capable of working in both wood and stone, suggesting a less media-specific division of labour than previously thought. It is not just the external details of buildings that reveal information about working methods and the transfer of techniques from one architectural tradition to another; there are occasional glimpses of internal methods as well. Iron cramps The use of large iron cramps to hold marble blocks together was a structural technique employed in the earlier Byzantine building tradition.46 As it is an internal system, in order to see the cramps a building has to have experienced extensive fabric loss or physical deterioration. The only visible example in the Islamic context, or dating from the period of Islamic rule in Anatolia, is to be found in Sivas, on the lower right-hand side of the entrance of the Gök Medrese (1271–2).47 The cramp is secured by being set into an oversized hole in the top of each block, with the iron cramp then encased in lead. The Sivas example is visible only because sections of the marble facing block have come away either due to water entering through cracks and freezing and forcing the stone away, or due to the iron rusting and the resultant expansion forcing the marble away (Figure 2.8). Like the use of wood tie-beams, particularly common in Ak∞ehir,48 techniques developed in the Roman–Byzantine tradition of building can be seen to have continued long after the external aesthetic for which they were originally developed had fallen out of use. Muqarnas hoods The muqarnas hood is the most visually striking and technically challenging aspect of the portals in Anatolia. It is a motif that, although occasionally used by other contemporaneous dynasties and carried on

Not for distribution or resale. For personal use only.

41

42

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Figure 2.8  Gök Medrese portal, Sivas (1271–2); iron cramp (photograph: Richard McClary)

under various later ones, has come to be associated primarily with the RËm Seljuk architectural aesthetic in the thirteenth century. There is evidence that the same craftsmen worked for both RËm Seljuk and Mengüjekid royal patrons in the last decade of the twelfth century,49 but by the early thirteenth century such an approach was less common. The absence of the use of muqarnas hoods for the primary portals or the mi˙råb of the monumental Mengüjekid mosque and hospital complex in Divri©i, completed in 1228–9,50 suggests that it was increasingly recognised as a more specifically RËm Seljuk, rather than a generally Turco-Muslim motif in Anatolia. Through the course of the thirteenth century the form can be seen to have proliferated across the expansive RËm Seljuk sultanate. The two earliest muqarnas hood portals to survive, at the Seljukid Alay Han near Aksaray and the Mengüjekid Sitte Melik Tomb in Divri©i, over 450 km apart, bear the signature of the same TËtbeg ibn Bahram al-Khilati,51 showing both the mobility and the diverse patronage of craftsmen at the time. The Alay Han is thought to date from late in the rule of the RËm Seljuk sultan Kilij Arslan II, probably around 1190 according to Pancaro©lu, while the tomb has a terminus ante quem of 1196–7 based on the epigraphic band around the top.52 The Divri©i tomb was built for the same patron as the Divri©i Citadel Mosque, Sayf al-Din Shahanshah (r. 1175–97).53 Unlike all the other portals in the corpus, it does not have a cavetto frame, featuring instead a stepped recess with a patterned edge. Neither of the portals have flanking niches, but the Alay Han portal has a low-relief carving of a pair of affronted lions that share a single forward-looking head at the base of the muqarnas hood. This symbol of royal authority is also seen on the portals of the Çifte Madrasa in Kayseri (1205–6), and the Izz al-Din Kay Kawus Hospital in Sivas (1217–18).

Not for distribution or resale. For personal use only.

craftsmen in medieval anatolia

Moving from the central and eastern parts of Anatolia, our attention now turns to the study of the use of imported forms and patterns on two marble portals in Konya. An Aleppine motif writ large: stereotomic strapwork The bi-chrome marble blocks that decorate the arch and spandrels of the portal (c. 1219–20), set into the north enclosure wall of the Citadel Mosque in Konya,54 are examples of complex structural stereotomy, with the stones forming the interlace pattern being fully bonded with the masonry of the building55 (Plate 2). The geometric interlace is incised with two parallel sets of three lines across polygonal stones that add a sense of relief and accentuate the dynamism of the entire composition. The closest parallel is to be found in the decoration of the coeval mi˙råb surrounds in a number of the Ayyubid madrasas in Aleppo, the Madrasa al-Sultaniyya (619/1223) being the most similar (Plate 2). Unlike many of the other Syrian examples,56 it also has small ajouré bosses in the upper corners in the manner of the Karatay Madrasa portal (c. 1220), located just to the north of the Konya citadel.57 The courtyard Èwån (a vaulted room enclosed on three sides, the fourth side being open with a monumental arch) of the Mashhad al-Husayn in Aleppo (1189) is thought to be the earliest example of strapwork stereotomy.58 Scott Redford has argued that the spandrel decoration of the marble portals in Konya represents the monumentalising and externalising of a form previously reserved for the mi˙råb.59 The Mashhad al-Husayn Èwån decoration suggests that the use of the motif in the mi˙råb context may have been the miniaturisation and internalising of a previously monumental external form.60 The lack of projection and muqarnas on the Mashhad al-Husayn Èwån is similar to the form of the Citadel Mosque portal in Konya. In contrast, the large epigraphic panel along the top and projection above the rest of the façade is in the manner of the Karatay Madrasa portal in Konya.61 Although the closest parallels for the interlace patterns are to be found in the architecture of Aleppo, there are several monumental examples of lithic interlace to be found in Mesopotamia. The pattern employed on the Konya portals is of a kind that has been described as the ‘Syrian knot’, although the motif rapidly became dispersed across a wider region.62 The city of al-Amadiyya, about 160 km northeast of Mosul, was within the domain of the Zangid-turned-independent ruler Badr al-Din Lulu (d. 1259), a man of Armenian servile origins, to whom the surviving gate is attributed.63 The Bab al-Mawsil in alAmadiyya (early thirteenth century) features marble interlace in the spandrels of the arch, but the decoration is in bas-relief and monochrome.64 The overlapping semicircle pattern on the arch voussoirs, which form the bodies of two dragons (explained in greater detail below), is the most similar element of the composition.65 Another contemporary of the Konya structures was the Bab alTilism (Talisman Gate) in Baghdad, dated to 1221.66 Although it was

Not for distribution or resale. For personal use only.

43

44

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

set into a brick tower that formed part of the city wall, the portal itself consisted of carved marble spandrels, joggled voussoirs and columns. The bas-relief interlace was more organic and free flowing than the other examples discussed thus far, and clearly formed the bodies of two dragons that flanked a seated ruler, probably Caliph alNasir (r. 1180–1225), at the apex of the arch.67 It was the only known example of its kind in Baghdad, and demonstrates the wide-ranging geographical scope, if limited number, of this broad type of portal in the first two decades of the thirteenth century. There were extensive political connections between the caliph in Baghdad and the sultan in Konya at this time, including the RËm Seljuk sultan Izz al-Din Kay Kawus I’s membership in the caliph’s futuwwa movement.68 The corners of the al-Amadiyya gate spandrels are decorated with a figural relief of a dragon-slayer with a sword. In Mesopotamia, figural elements replace the rectilinear interlace. It is tempting to think that on the religious buildings of Konya and Aleppo, the aniconic rectilinear and curvilinear motifs were employed by the builders to act as abstracted symbols for the apotropaic depiction of victory over evil, a theme so clearly displayed in figural form in the contemporary secular city gates in al-Amadiyya and Baghdad. The overlapping semicircles on the arch, while clearly representing the body of the dragon on the secular city gates, alludes to such a form in its use in a religious context, and suggests an attempt to conflate orthodoxy and apotropaic protection of religious buildings.69 The introduction of a Crusader motif: elbow brackets Following the Latin occupation of Jerusalem, elements of Crusader stone-carving motifs entered the Islamic architectural tradition after the re-conquest of the city by the Ayyubids in 1187. The marble elbow brackets on the Citadel Mosque portal in Konya are a slightly simplified copy of the ones seen on the north façade of the al-Aqsa Mosque in Jerusalem (Figure 2.9).70 The latter building had been remodelled and used as a palace by the Knights Templars in the twelfth century. Following the re-consecration of the building as a mosque in 1187, the north porch was rebuilt in 1217–18 under the patronage of Salah al-Din’s nephew al-Malik al-Muazzam.71 The brackets, referred to as angle shafts by Hamilton, are cut from single blocks of medium-hard limestone, and are incorporated into the north porch.72 The conspicuous appropriation of an identifiable aesthetic of the defeated Christians73 on the most prestigious mosque in Jerusalem and second holiest in Islam, suggests that it may have been intended as a sign of the victory of Islam and the subjugation of Christianity.74 Following this logic, it seems quite possible that the RËm Seljuk sultan Izz al-Din Kay Kawus I (r. 1211–19) was attempting to make a similar political statement through the prominent use of an array of Ayyubid decorative elements on the portal of the most prestigious mosque in the RËm Seljuk sultanate.75 This use of visual

Not for distribution or resale. For personal use only.

craftsmen in medieval anatolia

Figure 2.9  Elbow brackets: Citadel Mosque, Konya (top); al-Aqsa Mosque, Jerusalem (bottom) (photographs: Richard McClary)

culture was but one part of the process of creating and supporting an imperial identity, alongside coinage and the expansion of commercial activity. The hiring of Syrian craftsmen to work in Konya, rather than local or Iranian ones, resulted in the appropriation of what was a specifically Aleppine Ayyubid aesthetic. It was introduced into Konya, the nexus of RËm Seljuk power, and the newly synthesised aesthetic was projected across the city in what could be viewed as a symbolic appropriation of the Ayyubids and their lands. The restriction of this style to Konya resulted in the creation of a specific visual identity for the capital. The use of craftsmen building in the style of the established Iranian aesthetic for the major commission of Izz al-Din Kay Kawus I, at the hospital in Sivas, may be seen as having served a similar function, on a much grander scale,76 with regard to the former lands of the

Not for distribution or resale. For personal use only.

45

46

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Great Seljuks in Greater Iran. This use of architecture as a physical cypher for universal dominion is reflected in the sultanic titles of the period, which also tend towards claims of universal dominion.77 Materials: Brick Although the west of Anatolia had a long tradition of brick construction during the period of Byzantine rule, the brick-built structures dating from after the Turko-Muslim conquest of the region are all built in the Iranian style. There are numerous stylistic connections between the Islamic architecture of Iran and Anatolia during the period of study, in terms of form and glazed-tile decoration. However, it is the metrological similarities which reveal that it was craftsmen moving from Iran into Anatolia that were responsible for the aesthetic, rather than the often nebulous ‘influence’ of Iranian architecture that led to the erection of Persianate brick buildings in Anatolia. There is considerable variation in the size of Byzantine bricks, with the standard brick being between 32 cm and 36 cm square and 3.5–5 cm thick.78 The bricks used in the buildings of the Islamic period across Iran and Central Asia also have a wide variation in size, with the measured samples having a range of 18–25 cm per side and a thickness of 3.8–6 cm. However, the bricks employed in the late medieval Islamic architecture of Anatolia are about 20 cm square and closer to 5 cm thick on average, making the majority of the surviving examples smaller and thicker than the Byzantine bricks. Given the stylistic similarities between much of the architecture of Anatolia and that of northwestern Iran, it is to be expected that the size of brick used would also be quite similar. Bricks dated to the Great Seljuk period have been excavated at Gurgan, on the southeast coast of the Caspian Sea, which measure 20 cm square and 4 cm thick.79 In addition, the Gunbad-i Qabus in Gurgan (1007) has bricks that measure 20.8 cm square and 4.7 cm thick.80 Such close correlation between the size of bricks used in Iran and Anatolia, along with the formal and decorative similarities of the structures and the use of glaze, all point towards the conclusion that the majority of the brick workers in Anatolia were migrants from Greater Iran, trained in the Iranian architectural tradition. The literature on medieval metrology includes a number of seemingly impossibly precise definitions for some of the units used by craftsmen in the medieval period.81 However, there are a number of factors that suggest a more haphazard and site- or craftsman-specific approach was taken with regard to the units of measurement used by the craftsmen working in medieval Anatolia. These include the irregular shrinkage rate of baked bricks, the lack of standardised rule sizes, and irregularities in the plans and decorative details of the extant structures. The abovementioned Kitåb fÈ ma‘rifat al-˙iyal alhandasiyya, written in Diyarbakır (Amid) in the late twelfth century by al-Jazari,82 and described as the earliest manual of engineering

Not for distribution or resale. For personal use only.

craftsmen in medieval anatolia

practice,83 gives three measurements. It states that a shibr is half a dhirå (cubit),84 or about 25 cm.85 This distance, which roughly corresponds to the span of a human hand, is very close to the width of the average brick, plus one rising joint of mortar. It appears to be the case that the shibr was the basic unit of measurement used by the brick builders, as well as engineers such as al-Jazari, working in the region. Another unit given by al-Jazari is the ißba maftË˙ (the length of a finger), a distance equivalent to about 4 cm.86 This unit of measurement is very similar to the average thickness of the bricks used in Anatolia, and suggests that the two basic units, for the width and the height of bricks, were standardised across a wide geographic area, from Central Asia to western Anatolia. Iranian Architecture in Exile: Malatya’s Great Mosque The Great Mosque in Malatya (c. 1247) features a number of brick structural forms and decorative elements that demonstrate the northwest Iranian origin of the craftsmen who built it.87 In the typically syncretic style of RËm Seljuk architecture, the stone (but not marble) west portal has a similar, if rather more crude, version of the two bi-chrome stereotomic portals in Konya, while the maqßËra dome is almost entirely Iranian in style. It features wide, brick-built, load-bearing muqarnas cells that effect the transition from the square structure to the round dome (Plate 3). They form the lower section of the squinches of the maqßËra dome, and appear to be the last examples of structural brick muqarnas in Anatolia.88 They are very similar to those found in northwestern Iranian mosques, such as the Masjid-i Jami in Marand,89 a mosque that has been categorised by Robert Hillenbrand as forming a distant part of the Qazvini style of architecture.90 The similarities between the Marand and Malatya mosques, as well as other examples in Isfahan and Qazvin, demonstrate the mobility of craftsmen, who appear to have travelled from the former Ildegüzid and Great Seljuk lands to the east. The Malatya dome appears to be the only example of this type of brick muqarnas dome support to survive in Anatolia.91 Similar, if smaller, structural brick muqarnas cells can be seen on the balcony of the minaret attached to the Great Mosque in Sivas, 1212–13, and on an even smaller scale in the two niches in the north Èwån of the nearby hospital, founded by Izz al-Din Kay Kawus I in 1217–18. Both these structures have been associated with a craftsman named Ahmad Ibn Abi Bakr al-Marandi,92 so the structural evidence for the mobility of craftsmen is supported by the nisba of a craftsman known to have been working the region. The architectural syncretism, consisting of indigenous and imported lithic elements integrated with brick-built forms and glazed decoration developed in Greater Iran, shows the presence of a wide range of craftsmen from both within and without Anatolia that worked on the surviving corpus of buildings.

Not for distribution or resale. For personal use only.

47

48

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Role Division Hypotheses In order to achieve a deeper understanding of the working methods and mobility of craftsmen in medieval Anatolia, analysis of the different processes involved in the extraction, manufacture and combination of building materials into finished structures can been divided into seven categories. These correspond to the six main materials: stone, brick, glaze, mortar, timber and iron, along with one miscellaneous group of non-media-specific roles. This has resulted in the identification of 120 roles93 requiring different skill levels, with the only obvious overlap across materials being their transportation. Within each group, the same individual may well have performed many different roles, and the precise details of the division of roles will probably never be known. What can be hypothesised is the number of distinct, skilled roles required for the successful completion of each aspect of the construction process, from resource extraction to project completion. For stone construction there were six discernible roles (Table 2.1) and for brick, eight (Table 2.2), while glazed tiles required seven different skilled roles (Table 2.3). Mortar and plaster also required six (Table 2.4), assuming that brick workers and stone-setters would be responsible for the final combination of their respective materials with the bonding mortar. Of the nineteen basic roles involved in the architectural use of timber (Table 2.5), eight were skilled or semi-skilled roles, while the extraction and manipulation of iron required four skilled roles (Table 2.6). Regarding the miscellaneous and non-media specific Table 2.1  Processes involved in stone construction Stone construction Quarry stone Select stone Rough shaping and dressing Transport building stones and rubble Site levelling (under supervision) Run string lines and survey site Dig foundations Finish dress ashlars Cut mouldings, voussoirs and patterns Design and create stereotomic muqarnas hoods Carve figural sculptural elements Carve marble epigraphic panels Install ashlars Place rubble and mortar infill Build, install and operate lifting machines Erect scaffolding Move stone and mortar around site

Unskilled Semi-skilled Skilled x x x

x

x

x

x

x

Not for distribution or resale. For personal use only.

x

x x x x x x x x

49

craftsmen in medieval anatolia

Table 2.2  Processes involved in the production and use of bricks Brick production and construction

Unskilled Semi-skilled Skilled

Calculate the number of bricks required for the building Select clay Quarry clay Collect and sieve sand Transport raw materials to kiln site Puddle clay with water and let stand Mix sand into clay at the correct ratio Fill wood or metal brick moulds Lay moulds out in the sun to air dry Gather and transport fuel for kiln Load bricks into kiln Fuel kiln and ensure temperature reaches 800°–900°C Determine point at which the bricks are successfully fired Unload kiln Grade bricks Transport bricks to site   (if kiln is off-site) Custom-cut corner, detail and Kufic letter bricks Distribute bricks around building site and up scaffolding Lay bricks in courses between beds of mortar

x x

x x x

x x

x x

x

x

Mine ores Transport ores Smelt various metals for glaze Quarry clay Transport clay Prepare clay for moulds Fill moulds and form tiles Gather soda plants Burn soda plants Prepare glaze mixture Gather fuel for kiln Load kiln Add fuel to kiln Manage temperature of kiln and length of firing time Paint underglaze designs Apply glaze Cut tiles Transport tiles Plan overall design Prepare surface Install tiles

Unskilled x x x x

x x x

x

x

x

x

Table 2.3  Processes involved in the production and installation of glazed elements Glazed-tile production

x

Semi-skilled

Skilled

x x x x x

x

x x

x

x x x x

x x

x x

Not for distribution or resale. For personal use only.

50

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Table 2.4  Processes involved in the production and installation of mortar Mortar production Quarry limestone and gypsum Gather aggregates Sieve and grade aggregates Transport materials to site Burn lime Dig pits Prepare mortar and leave to slake Rehydrate mortar Add aggregates as binders and pozzolans Move mortar around site and deliver to craftsmen Use mortar for setting bricks into walls and vaults Incise patterns in the mortar joints Use mortar for rubble infill and setting ashlars Apply to vertical surfaces and domes as finish layer Paint external surfaces with geometric patterns Make stucco moulds Fill stucco moulds Install stucco panels Carve epigraphic inscriptions

Unskilled x x x

x

Semi-skilled

Skilled

x x x x x

x

x x x x

x

x x x x

Table 2.5  Processes involved in timber construction and woodworking Timber construction and woodworking Determine the amount of timber required Select the trees Fell the trees Limb the trees Gather poles for scaffolding Rough-cut timber for transport Transport timber Stack timber for drying Cut timbers with two-man saw Shape and notch timbers with adze and axe Precision-cut sections for doors, shutters and minbars Carve patterns and calligraphy into panels Cut and assemble kundakiri sections Move timbers around construction site Install tie beams into walls Rough-cut wood for scaffold planks, bracing and centring Erect arch vault centring and scaffolding Build and erect cranes and other lifting machines Design and install roofing systems

Unskilled

x x x x

x x

Semi-skilled

Skilled x x x

x x x x x x x x

Not for distribution or resale. For personal use only.

x x

51

craftsmen in medieval anatolia

Table 2.6  Processes involved in the production of iron objects Iron production and use Mine iron ore Transport iron ore Smelt the ore to produce iron Transport ingots Gather fuel for forge Load fuel and operate bellows Determine and maintain forge temperature Forge and shape iron objects Temper iron objects Make clay models of items Make mould of item Cast and finish item Fabricate component parts Incise patterns into the finished object Produce steel for tool edges Sharpen tools Move metal objects around site

Unskilled Semi-skilled x x x

x

Skilled

x x x

x

x x x x x x x x x

Table 2.7  Miscellaneous site roles Miscellaneous site roles Fiduciary management: pay wages and suppliers Site management: coordinating different trades Chef: food preparation Leatherworker: repair leather gloves and aprons Scribe: prepare contracts and lay out epigraphy Doctor: treat work-site injuries Deliver water for drinking and mortar production Remove waste materials from site

Unskilled Semi-skilled

Skilled x x

x x x

x

x

tasks (Table 2.7), such as site management, food preparation and treatment of injuries, seven of the eight roles would have required some considerable degree of skill. So, of the total of 120 roles, 43 (or about one-third) may be deemed to have required significant levels of training and experience. Such a ratio may suggest a division of the workforce into three fairly equal sections: labourers, apprentices with varying levels of experience and fully-trained craftsmen, of whom only a few would be masters in their field. It may be assumed that for the most part, the labourers were recruited locally from the predominantly Christian population.94 With the exception of the few named master masons who were

Not for distribution or resale. For personal use only.

52

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

working in a very similar manner to, and based on their nisbas had moved from, Ayyubid-ruled Syria, the stonemasons were largely Anatolian, and probably predominantly Armenian and Georgian. Such a view is based on the formal and decorative similarities between the surviving churches and the stone-built Islamic architecture of the period.95 The same is likely to have been the case for the blacksmiths and many of the carpenters. In contrast, it is likely that the majority of the higher-skilled individuals, especially those working with baked bricks and glazed tiles, were immigrants to Anatolia from Iran, and possibly even Central Asia, especially in the late twelfth and early thirteenth centuries. Conclusion In late medieval Anatolia a significant methodological change regarding the execution of stone patterns took place. This change demonstrates the dynamic and innovative development in techniques and styles that was underway in the region during this time period. The external decorative innovations of many of the buildings can be seen, on occasion, to mask internal structural conservatism. The structural methods exhibit a strong sense of continuity from the earlier Christian architecture of the region. In order to accommodate the different functional and aesthetic needs of the ruling Muslim patrons and the growing Muslim community a number of spatial and decorative changes were made. At the same time, new materials (such as turquoise-glazed tiles) and new motifs (such as muqarnas cells) executed in wood, brick and stone were introduced. The combination of different materials and traditions in the same building, seen across Anatolia in places such as Ak∞ehir in the west, Sivas in the centre and Erzurum in the east, demonstrates the eclectic and universal character of the construction business in Anatolia during the late twelfth and early thirteenth centuries. Stonemasons and brick workers, Christians and Muslims, indigenous and immigrant; all worked together on a wide range of structures for the various Muslim dynasties across the region. By the early decades of the thirteenth century these craftsmen had created a new and uniquely Anatolian architectural aesthetic.

Notes 1. Buildings built under Muslim patronage or for the purposes of prayer or religious study. The corpus consists of tombs, mosques and madrasas, as well as the commercial infrastructure built by the Muslim ruling elite, including caravanserais, bridges and markets. Within the broad term of Islamic architecture may also be included the hospitals, palaces and mural fortifications erected during the period of Muslim rule.

Not for distribution or resale. For personal use only.

craftsmen in medieval anatolia

2. Examples include the Great Mosque minaret (1212–13) and the Izz al-Din Kay Kawus Hospital (1217–18) in Sivas, as well as the Mengücek Gazi Tomb in Kemah (c. 1190). 3. The largely lithic medium and several elements of decoration have been drawn from Armenian architectural tradition. See McClary, ‘RËm SaljËq Architecture’, pp. 259–61. In contrast, the use of exposed timber beams and opus mixtum construction, as seen in the mosques of Ak∞ehir, shows clear links to Byzantine traditions of building. See ibid., pp. 240–2. 4. Although few of the individuals are named, one example is Ahmad ibn Abi Bakr al-Marandi, who worked in Sivas, Niksar and Tokat in the first three decades of the thirteenth century. See McClary, ‘RËm SaljËq Architecture’, pp. 130 and 356, for more details of his work. 5. Redford, ‘Just Landscape in Medieval Anatolia’, p. 83, cites a contemporary source, by Ibn Said, which refers to floating pine lumber down the river from forests located near Amasya northward to Sinop, for use in building the arsenal. 6. Arık, ‘Tiles in Anatolian Seljuk Palace Architecture’, pp. 496–7. 7. Arık, ‘Tiles in Anatolian Seljuk Palace Architecture’, p. 498. 8. See Meinecke, Fayencedekorationen, vol. 2, p. 113, fig. 21, for a drawing showing the location of the bowls. 9. Greenhalgh, Marble Past, Marble Present, p. 137, citing Guy Le Strange (trans. and ed.), Nasir-i Khusrau: Diary of a Journey through Syria and Palestine (London: Committee of the Palestine Exploration Fund, 1893), p. 132. 10. Richards, The Chronicle of Ibn al-AthÈr, p. 342. The translation of Ibn al-Athir’s al-Kamil fi l-Tarikh states that a fire in the store yards in 1187 destroyed large stocks of timber. 11. The tool measures approximately 25 cm × 10 cm on the base. 12. Tonghini, Qal’at Ja’bar Pottery, p. 70, adds that the turquoise monochrome wares were not particularly sophisticated, being cheaper and inferior to other glazed table wares. 13. For the definition of nisba, see Sourdel and Sourdel-Thomine, A Glossary of Islam, p. 127. On the problem of using nisbas for identification, see Rogers, ‘Patronage in Seljuk Architecture’, p. 446, and Snelders, Identity and Christian–Muslim Interaction, p. 91. 14. See Redford, ‘The Alaeddin Mosque’, pp. 56 and 73. 15. Abu Ali al-Halabi ibn al-Kattani, possibly from Syria, his nisba being the only evidence available, worked on the walls of Sinop in 1215, but there is nothing distinctively Syrian or Aleppine about the work he did. See Redford, ‘Sinop in the Summer of 1215’, p. 131. 16. Rogers, ‘Art and Architecture in Anatolia’, p. 966. Ghazarian and Ousterhout, ‘A Muqarnas Drawing’, p. 145, note the presence of a named Armenian master, Yovhannes, on the incised lines used to lay out a muqarnas vault on the Asvantsankal Monastery in Armenia of 1244; Blessing, Rebuilding Anatolia, p. 103, citing Laor-Sirak, ‘The Role of Armenians’, pp. 173–4. 17. Yalman, ‘Building the Sultanate of Rum’, p. 143. 18. The division of roles in the upper echelon of the construction business remains unclear. See Rogers, ‘Patronage in Seljuk Architecture’, esp. pp. 296–367 and 400–18, for a good attempt to determine what can be known from the limited sources. See Snelders, Identity and Christian– Muslim Interaction, pp. 90–2, for details of mixed Muslim–Christian workshops in the Mosul area during the period of study.

Not for distribution or resale. For personal use only.

53

54

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

19. Mkrtchyan, Treasures of Artsakh-Karabagh, p. 123. Eastmond, Art and Identity in Thirteenth-century Byzantium, p. 92, states that the church was founded by the Christian Armenian king of Khachen, Hasan Jalal Dawla/Haykaz (r. 1214–61). 20. Eastmond, Art and Identity in Thirteenth-century Byzantium, pp. 92–3, states that Ani was liberated from Muslim rule by the Mqargrdzeli family in 1199. Their court was the most mixed of all the dynasties in eastern Anatolia, as they were Kurds who married into Armenian, Georgian and Seljuk families. For details of the design and c­ onstruction of a similar vault at the Asvatsankal Monestary near Yerevan, see Ghazarian and Ousterhout, ‘A Muqarnas Drawing’, esp. pp. 146–50. 21. A similar muqarnas roof can be seen in the Great Mosque of Erzurum (late twelfth century). 22. The surviving Danishmendid structures, primarily located in in Sivas, Kayseri and Niksar, are almost exclusively stone-built. 23. There was a wide array of different Christian denominations in the region at the time, including Syrian Orthodox, Greek, Armenian and Georgian. It is likely that craftsmen from a number of different denominations would have worked for the new rulers on their increasingly large number of architectural commissions. 24. See Ousterhout, Master Builders of Byzantium, pp. 169–79 for details of the Byzantine methods of wall construction. 25. Knoop and Jones, The Mediaeval Mason, p. 54. The authors draw primarily on European sources, but the techniques and tools used for stone and brick construction remained basically unchanged from the Roman period until the nineteenth century. 26. Hill, The Book of Knowledge, p. 278. 27. Clifton-Taylor, The Pattern of English Building, pp. 58 and 110. 28. Rogers, ‘Patronage in Seljuk Architecture’, p. 439. 29. Knoop and Jones, The Mediaeval Mason, pp. 118–19. 30. Knoop and Jones, The Mediaeval Mason, pp. 77–8. 31. RCEA, vol. 8, p. 111. Located at: lat: 39° 22’ 32” N, lon: 038° 07’ 24” E. 32. Bosworth, The New Islamic Dynasties, p. 217. 33. Pancaro©lu, The House of Mengüjek, p. 32. The Kufic inscription as it appears on the building, which begins with ‘the builder is the master’ (al-åmil ustådh . . .), includes an extra letter, a låm instead of an alif at the beginning of ustådh and a number of non-standard breaks between letters, resulting in isolated medial forms appearing in the text. See McClary, ‘RËm Saljuq Architecture’, p. 39, for the Arabic text as displayed on the portal. 34. Ghazarian and Ousterhout, ‘A Muqarnas Drawing’, p. 144. 35. The nearby Kamereddin Tomb (1196), at the bottom of the citadel hill, is another rare example of stone and glazed elements being combined. In that case, there is a band of shallow circular recesses around the top of the wall into which glazed bowls, rather than architectural tiles, were set. 36. Located at lat: 39° 22’ 25” N, lon: 038° 07’ 09” E. See Önkal, Anadolu Selçuklu türbeleri, pp. 37–42, for elevation drawings of the Sitte Melik Tomb. 37. Located at: lat: 39° 46’ 40” N, lon: 040° 23’ 11” E. For a detailed description and drawings of the plan and decoration of the Mama Khatun portal, see Ünal, Les monuments islamiques, pp. 129–42. 38. See Bosworth, The New Islamic Dynasties, p. 197, for details about the Sökmenids.

Not for distribution or resale. For personal use only.

craftsmen in medieval anatolia

39. Pancaro©lu, The House of Mengüjek, p. 55. 40. Pancaro©lu, The House of Mengüjek, p. 55. 41. Rogers, ‘Calligraphy and Common Script’, p. 109, states that the earliest standing tombstone in the Ahlat cemetery is dated 1189. 42. Earlier stone muqarnas can be found in the architecture of Fatimid Cairo, an example being on the inside of the Bab al-Futuh (c. 1087). See Bloom, Arts of the City Victorious, pp. 125–6 and fig. 93. 43. A door, originally from the thirteenth-century Kuyulu Hoca Pa∞a Mescidi in Ankara, and now in the Etnografya Müzesi (Ethnographic Museum) in Ankara, Accession No. 8015, has clearly visible incised construction lines. 44. Rogers, ‘Waqf and Patronage’, p. 101. 45. Pancaro©lu, ‘The House of Mengüjek’, p. 41. 46. Ousterhout, Master Builders of Byzantium, pp. 212–14, describes the various sites where iron cramps are known to have been employed with marble. The most common use is to tie together the blocks of the dome cornice of churches, an example being the Fatih Camii in Enez. Ousterhout, Master Builders of Byzantium, p. 215, fig. 175, shows cramps set in lead that are almost identical in shape and size to the one in Sivas. 47. Located at: lat. 39° 44’ 39” N, lon: 037° 01’ 00” E. For more details about the Gök Medrese, see Blessing, Rebuilding Anatolia, pp. 104–15. 48. Surviving examples in Ak∞ehir include the Great Mosque, the Güdük Minare Mescidi and the Ferruh S¸ ah Mescidi, all built in the first quarter of the thirteenth century. 49. Pancaro©lu, ‘The House of Mengüjek’, p. 41. 50. Kuban, The Miracle of Divri©i, p. 42, cites the date as from an epigraphic inscription and suggests, based on the scale and the complexity of the carving, that the complex was probably started in the early 1220s. In contrast, Rogers, ‘Patronage in Seljuk Architecture’, p. 91, suggests that the date 1228–9 marks building’s foundation rather than its completion. Kuban, The Miracle of Divri©i, p. 148, notes that the only use of a muqarnas hood is over the later east window, which was not part of the original design, and was executed by a different craftsman. 51. Pancaro©lu, ‘The House of Mengüjek’, pp. 39–41. See Önkal, Anadolu Selçuklu türbeleri, pp. 37–42, for elevation drawings of the Sitte Melik Tomb. 52. Pancaro©lu, ‘The House of Mengüjek’, p. 39. No specific reason for the attribution to Kilij Arslan II is given, and Öney, Summary of Lion Figures, p. 52, assumes that the Alay Han is of thirteenth-century construction. 53. Pancaro©lu, ‘The House of Mengüjek’, p. 35. It is quite possible that the tomb was built during his lifetime, which, coupled with the lack of a firm date for the han, could indicate that the notion that the Alay Han having being constructed earlier is a baseless assumption. 54. The portal to the mosque, also referred to as the Alaeddin Mosque, is located at: lat: 37° 52’ 25” N, lon: 032° 29’ 34” E. 55. Tabbaa, The Transformation of Islamic Art, p. 160. 56. See Tabbaa, Constructions of Power, pp. 80–1, for numerous examples of surviving Ayyubid portals with similar strapwork decoration in Aleppo and Damascus. 57. Located at: lat: 37° 52’ 29” N, lon: 032° 29’ 35” E. See Rogers, ‘Waqf

Not for distribution or resale. For personal use only.

55

56

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

and Patronage’, pp. 77–80, and McClary, ‘RËm SaljËq Architecture’, pp. 77–9, for evidence that the portal predates the madrasa to which it is attached, which was built in 1251–2. 58. Tabbaa, Constructions of Power, pp. 112 and 118. 59. Redford, ‘The Alaeddin Mosque’, p. 71. 60. The Mashhad al-Husayn’s Èwån is enclosed within a courtyard, and thus not as external or as visible as the two Konya portals. 61. The Èwån strapwork features a line through the central upper circle in the manner of the Madrasa al-Sultaniyya mi˙råb, unlike in Konya, while the square motif in the two upper corners resolves in a slightly different manner to that of the two Konya portals. Although the upper band of epigraphy had been altered, it may be assumed that the original foundation inscription was in the same location. 62. Gierlichs, ‘Das Mosul-Tor’, p. 202. In ibid., p. 195, the author argues that the decorative and morphological grammar employed in the Byzantine churches in the south Anatolian–north Mesopotamian art region provides a number of the foundations of the visual language employed by the designers of the Islamic structures in the region and beyond. 63. Janabi, Studies in Medieval Iraqi Architecture, p. 253. Although Badr al-Din Lulu was not officially recognised by the caliph until 1234 (ibid., p. 53), it is possible that the gate was built prior to that date, as he was appointed regent to the Zangid prince from 607/ 1210 onwards. For details of his rule, see Patton, ‘A History of the Atabegs’, pp. 157–74. 64. See Janabi, Studies in Medieval Iraqi Architecture, pl. 175. 65. Janabi, Studies in Medieval Iraqi Architecture, p. 353, points out that the coiled and elongated bodies of the dragons form the arch of the portal (see ibid., pl. 175). The portal has been recently rebuilt after extensive damage and earlier images give a more accurate impression of its original appearance. 66. Bell, Amurath, p. 190 and figs 114 and 115. The gate was destroyed in 1917 by the British army. For a detailed study of the gate, see Sarre and Herzfeld, Archäologische Reise, vol. 1, pp. 34–40; Kuehn, The Dragon, pp. 124–9. 67. Janabi, Studies in Medieval Iraqi Architecture, p. 252. See Bell, Amurath, fig. 115. Pancaro©lu, in ‘The Itinerant Dragon-Slayer’, p. 160, disagrees, suggesting that the whole composition is an apotropaic device, and that the seated figure is a personification of the sun. 68. For details of the caliphal futuwwa movement, see Goshgarian, Futuwwa in Thirteenth-century RËm and Armenia’, p. 230. 69. The multivalent symbolism of the dragon has been noted by Kuehn in The Dragon, p. 124, and the abstracted symbolism in the Konya examples fits into a wider understanding of its use. 70. In both cases, the brackets are purely decorative as they have no loadbearing role, having been cut from the impost block that supports the arch. 71. Grabar, Jerusalem, p. 142. The east and west façades date from the Latin occupation, but the work on the north façade appears to be datable to the work on the north porch. That work represents Muslim patronage using Crusader motifs, or maybe even spolia, such as the elbow brackets. 72. Hamilton, The Structural History of the Aqsa Mosque, pp. 39–40. He goes on to suggest that the blocks may be twelfth century spolia. See

Not for distribution or resale. For personal use only.

craftsmen in medieval anatolia

ibid., p. 40, fig. 21, for a plan showing the location of the blocks, along with pls XXII.3, XXIII.1–6 and XXIV.1–4, for images of all the surviving blocks on the porch in 1949. 73. Hazard, A History of the Crusades, p. 80, describes the elbow bracket as a characteristic invention of the Crusaders. Surviving examples in a Christian context can be found on the western wall of the cloister of the Church of the Holy Sepulchre in Jerusalem. 74. Hillenbrand, The Crusades, p. 383. 75. See Yalman, ‘Building the Sultanate of Rum’, pp. 137 and 214, for an argument that the RËm Seljuk sultans were attempting to connect Konya with the idea of a heavenly Jerusalem. 76. As well as a glazed-tile façade for the tomb of the patron and other glazed accents in the north Èwån, the majority of the structure in Sivas are built in brick, in the manner developed in Iran in the previous two centuries. 77. The foundation inscription of the Izz al-Din Kay Kawus Hospital in Sivas describes him as ‘the pillar of Islam and Muslims, the Sul†ån of the land and the sea, the crown of the SaljËq family. . . amÈr of the believers’. 78. Ousterhout, Master Builders of Byzantium, p. 131. Bricks generally shrink by about 10 per cent during the firing process. 79. Kiani, Recent Excavations in Jurjan, p. 126. 80. Godard, Gurgån, p. 972, gives the size, but notes that there are some variations in the thickness of the bricks. For more details, and historical context of the tomb, see Michailidis, ‘The Lofty Castle’, pp. 120–38. 81. See Hinz, Islamische Masse, for Islamic metrology, and Schilbach, Byzantinische Metrologie, for a study of the units of measurement used by Byzantine craftsmen. 82. See Hill, The Book of Knowledge, for a translation of the Bodleian Library MS. Graves 27, dated 1486, a copy of a text dating from 1341. 83. Hill, The Book of Knowledge, p. 279. 84. Hinz, Islamische Masse, p. 64, states that the Arabic ad-dhirå assar‘iyya corresponds exactly with the Persian Ωar-e sar È, given as 49.875 cm, but makes no mention of the shibr. In the Byzantine context, Schilbach (1970), pp. 19 and 44–5, mentions the ‘span’ (spithami, σπιϑαμή) and gives a distance of 23.46 cm used in thirteenth-century Trebizond. The ubiquitous nature of the human hand is the reason for the common measurement. Byzantine bricks appear to be based on the foot (πούς): ibid., pp. 14–16, gives a range between 31.18 cm and 31.89 cm 85. Hill, The Book of Knowledge, p. 278. 86. Hill, The Book of Knowledge, p. 278. Hinz, Islamische Masse, p. 54, states that an ißba is 1/24 of a dhirå (c. 2 cm), but suggests that an ißba in Egypt was 3.125 cm. He makes no mention of the ißba maftË˙. 87. Arık, ‘Tiles in Anatolian Seljuk Palace Architecture’ pp. 73–4, tentatively accepts the earlier of the two dates on the building as indicating the main phase of construction, done in brick with glazed-tile decoration, but suggests that it may be as early as the reign of Izz al-Din Kay Kawus I. The later date, 1274, appears to be the date of restoration, primarily executed in stone. Stylistically, the first half of the thirteenth century is more likely. See also Meinecke, Fayencedekorationen, vol. 1, pp. 390–400. In addition RCEA, vol. 7, p. 181, gives the inscription from the west portal that refers to a Persian, ‘the work of the master Khusraw, the builder’ (amal ustådh Khusraw al-bannå).

Not for distribution or resale. For personal use only.

57

58

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

88. For a study of the full corpus of early brick muqarnas cells in Anatolia, see McClary, ‘Brick Muqarnas’, pp. 1–11. 89. See Hillenbrand, SaljËq Dome Chambers, p. 353, fig. 6. 90. Hillenbrand, SaljËq Dome Chambers, p. 360. For other examples of large-scale structural muqarnas cells in Iran, see Pope and Ackerman, A Survey of Persian Art, pl. 320, for Ardistan Masjid-i Jami (1055–8), ibid., pl. 305, for Qazvin Masjid-i Jami (1113 or 1119). 91. A more elaborate example, at the Great Mosque in Van (early twelfth century), which featured a larger number of smaller cells was photographed, in a ruinous state, by Bachmann prior to the catastrophic earthquake in the early twentieth century. See Bachmann, Kirchen und Moscheen, pp. 69–74 and pls 59–63. 92. Based on stylistic similarities, Meinecke, Fayencedekorationen, vol. 2, p. 461, suggests that A˙mad Ibn Abi Bakr al-Marandi was also responsible for the Sivas Great Mosque minaret. See ibid., pls 41.2 and 41.3, for pre-restoration images of the two niches. 93. Several of the processes, such as glaze production and metallurgy, could be sub-divided further, but to do so would add unnecessary complexity and increased conjecture without aiding the understanding of the wider topic. 94. Dadoyan, The Armenians, p. 147, states that the population of Anatolia in the early thirteenth century was overwhelmingly Greek in the west and Armenian in the east. Mecit, Kingship and Ideology, pp. 103–5, states that populations of Sinop and Antalya were predominantly Greek Christians. Yaqut al-Hamawi wrote in the early thirteenth century that the majority of the population of Erzincan were Christian Armenians. Such a view is echoed in other texts by Ibn Bibi, Ibn Battuta and Marco Polo. See Goshgarian, ‘Futuwwa in Thirteenth-century RËm and Armenia’, p. 239. There were also large numbers of Syrian Orthodox Christians living in the south of the region, around Malatya. Leiser, ‘The “History of the Patriarchs”’, p. 114, states that the Greek population appeared to enjoy justice and fair treatment, according to the History of the Patriarchs of the Egyptian Church by Abu’l-Makarim ibn Barakat, written in 1207, however, no mention is made of the significant Armenian population. Vryonis, Decline of Medieval Hellenism, p. 180, a rather dated and at times questionable source, states that the overwhelming majority of the population remained Christian. In ibid., p. 183, he argues that the Christian population was the main source of tax revenue in the latter part of the thirteenth century. 95. For further details and examples of the connections between the indigenous ecclesiastical architecture of the Georgian and Armenian communities, on the one hand, and the late medieval stone-built Islamic architecture, on the other, see McClary, ‘RËm SaljËq Architecture’, pp. 260–1 and 294–6.

Not for distribution or resale. For personal use only.

CHAPTER THREE

Stones for Travellers: Notes on the Masonry of Seljuk Road Caravanserais Cinzia Tavernari Road caravanserais represent possibly the most impressive and important construction programme that survives from medieval Anatolia.1 The majority of these structures date from the Seljuk period and some studies estimate that the Seljuks promoted the construction of approximately 300 road caravanserais,2 a remarkably high number considering the limited chronological timeframe of Seljuk rule in Anatolia.3 Most Seljuk caravanserais mark the communication axes connecting Konya and Kayseri to the Mediterranean and the Black Sea, and were built within one single century, that is, from the 1180s to the 1280s,4 when ‘Anatolia witnessed one of the most ambitious building programmes in Islamic history’ as Scott Redford has noted.5 In this chapter, I argue that studying the archaeology of buildings can provide us with new kinds of analyses of road caravanserais through stratigraphic examination of the material evidence and investigation of the masonry techniques used. Focusing in particular on the analysis of masonry characteristics, this study seeks to provide new data on the relationship between the technical environment that characterised the construction of road caravanserais and the range of political, economic and social elements whose interaction determined the creation of these edifices. The State of Research Originally recorded and contemplated primarily in travelogues, a systematic analysis of the road caravanserais of Anatolia began only towards the middle of the twentieth century when the Turkish scholar Osman Turan6 and, most notably, German art historian Kurt Erdmann produced academic work on the structures. In the mid- to late 1950s, Erdmann conducted several field surveys that led to the publication of the first – and still fundamental – a­ rchitectural study of Seljuk caravanserais; this work recorded both monuments that were extant on the ground when Erdmann conducted

Not for distribution or resale. For personal use only.

60

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

his research, and those that were known uniquely from written sources.7 Erdmann’s endeavour stimulated further interest and work on Seljuk road caravanserais, primarily by scholars in Turkey. These studies mainly focused on the architecture of road caravanserais and on the search for yet unknown similar structures.8 It is within this context that we must consider the catalogues of extant caravanserais and more detailed studies of a few individual buildings.9 Unfortunately, conceptual approaches to these monuments are still rare. The most representative efforts in this direction have come from Ay∞ıl Tükel Yavuz, Scott Redford and Mustafa Önge, who have exposed innovative ways of exploring the caravanserai phenomenon in Seljuk Anatolia. An influential article published in the 1990s by Yavuz proposed the investigation of caravanserais by relating these institutions not only to commerce, but also to the culture of the Seljuk court.10 Redford identified and highlighted the significant relationship between the landscape and the building programme promoted by the Seljuk dynasty. More recently, he has stressed the importance of elements such as patronage and the road network for a more comprehensive interpretation of the road caravanserai phenomenon.11 Önge has considered the role of road caravanserais in relation to the political power of their wealthy patrons.12 Caravanserai studies in Turkey have recently gained momentum with respect to Seljuk caravanserais and Ottoman caravanserais alike, the latter group finally receiving the long overdue scholarly attention it deserves.13 The success of the workshop held in 2014 in Kayseri in order to consider the state of research on road caravanserais in Turkey suggests that there is a general renewed interest in the field.14 Recent scholarly approaches to road caravanserais include a more comprehensive consideration of their patronage, a new interest in the study of these buildings in relation to the road networks15 and the use of GIS-based analyses.16 Notwithstanding Fernand Braudel’s famous statement that there would be no roads without stopovers,17 in fact, until recent times road caravanserais have rarely been considered as spatial–temporal tags that could provide relevant information on the development and resilience of routes.18 GIS-based analyses are particularly useful for the study of the relationship between routes and road caravanserais and, more importantly, are essential to achieving an enhanced comprehension of caravanserais as a fully connected network of buildings, provided that they are truthfully contextualised.19 A thorough comprehension of road caravanserais as a network represents an essential step in elucidating the precise nature and function of these structures. Such a diversity of research approaches is particularly promising, yet some noteworthy research methods have failed to be considered until now, such as those that engage with the archaeology of buildings. In the first part of this chapter, I will present the aims and methods of the archaeology of buildings and illustrate the material evidence.

Not for distribution or resale. For personal use only.

stones for travellers

In presenting the material evidence, I examine and describe in detail the characteristics of the masonry techniques employed in three road caravanserais built at the beginning of the thirteenth century, the heyday of road caravanserai construction. In addition to their construction dates, these buildings share other significant characteristics: they are fairly well preserved and, even more importantly, have either not been restored at all or only very partially. This is an important point because restoration operations can severely obliterate essential features of the built structure, thus impeding a thorough and correct masonry investigation and stratigraphic analysis. I successively analyse the constructional characteristics of the road caravanserais investigated during the survey, particularly focusing on the construction process, the typological characteristics of the masonry and the tools of the craftsmen. Finally, in the conclusion, I attempt a broader contextual comprehension of the Seljuk road caravanserais by considering them in relation to the landscape, their patronage and the development of equivalent structures in neighbouring Bilad al-Sham. My approach informs our understanding of the ways in which this caravanserai construction programme sheds light on the ways in which landscape, investment, patronage and power were understood within a Seljuk context. At the same time, considering the constructive techniques, technical knowledge, and the mobility and origin of the craftsmen who built these structures helps us to better understand the kind of craftsmen networks that existed in the medieval Middle East. Aims and Methods As stated above, a large part of the research on road caravanserais focuses on the architecture of these buildings and, in particular, on their layout.20 Still, it is important to note that few offer detailed information on the characteristics of the masonry used in these edifices.21 Even when information on masonry techniques is present, the way the data is organised and presented often prevents us from easily comprehending the building techniques and methods of the craftsmen who built the caravanserais. Moreover, the study of building techniques nowadays pays attention to elements that past architectural studies disregarded, such as the use of wedges or the traces left by the masons’ tools.22 Therefore, one of the main objectives of the current work on Seljuk road caravanserais is to present a new and methodologically relevant characterisation of the masonry techniques employed in these edifices. The approach to the material evidence used in this research draws on the modus operandi of the archaeology of buildings, a methodology that developed in the late 1970s and that has now acquired a well-established place among the disciplines that focus on the

Not for distribution or resale. For personal use only.

61

62

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

study of historical buildings. The archaeology of buildings primarily focuses on the direct and detailed analysis of the preserved material structure, considered ‘the repository of a substantial mass of information’23 on its constructive development. The archaeology of buildings developed from excavation archaeology and borrowed – and later adapted – the principles of stratigraphic analysis to the study of standing structures.24 Stratigraphic analysis allows archaeologists to investigate the constructional development of an edifice, an objective that, together with the characterisation of building techniques, represents the focus of archaeologists who work on historical buildings. It is important to reconstruct as precisely as possible the constructional history of an edifice because the latter is never the product of chance, but the result of very specific choices that depend on three primary factors: the nature of patronage, the availability of material and the technical knowledge of the craftsmen. These elements are closely interconnected, as well as linked to a broader context, and provide a basis for the construction of interpretative models. These models offer new elements to our understanding of the technical – and also the economic, social, cultural and political – background of an edifice. Recreating this complex background is, in fact, one of the ultimate goals of the archaeology of buildings. In striving to create a comprehensive history of a structure, the archaeology of buildings also aims at contributing to a suitable preservation plan or restoration project. At this stage in my research, I have been able to conduct only a preliminary survey. I collected new data on the masonry and building techniques of some Seljuk road caravanserais, but could not carry out a detailed stratigraphic analysis; the stratigraphic study has, therefore, been limited to some important, but partial, observations. The Survey and the Description of Material Evidence The survey that took place in May 2015 focused on three caravanserais: Evdir Han (built between 1210 and 1219) and Òncir Han (1238), both on the Antalya–Burdur road, and Çardak Han (built in 1230) on the Denizli–Dinar road. The most important factor underlying the choice of these caravanserais was that their structures remain remarkably legible, as they are all quite well preserved. They have not been profoundly modified by modern building activities, and have either been only partially affected by restoration work or not at all. As noted above, restoration activities often severely hamper the recognition of the original characteristics of the masonry and of the stratigraphic evidence. A correct analysis of the original mortar of an edifice, for instance, is usually impossible after a building has been heavily restored. This is because the original mortar is usually eliminated during the restoration process and is replaced with brand new mortar. Moreover, earlier restorations are often poorly documented

Not for distribution or resale. For personal use only.

stones for travellers

and the extremely rare historical photographs (such as Erdmann’s) are sometimes the only means of working out what is part of the original structure and what is not.25 Evdir Han, Òncir Han and Çardak Han are among the few road caravanserais that have not yet been restored – or rebuilt – in the burgeoning construction boom of contemporary Turkey. They are, however, currently undergoing cleaning and research (Evdir Han) or have recently been excavated (Òncir Han26 and Çardak Han27) as part of a plan of restoration.28 Due to these considerations, it seemed even more urgent to focus on the examination and study of these caravanserais. Rescue excavations and metric surveys are usually carried out prior to restoration. However, the documentation produced during these operations is generally quite difficult to access, especially when a private company carries out the work rather than a university or research centre (and, in the modern Republic of Turkey, private companies often carry out these surveys). While more comprehensive investigations of the structures would allow for a deeper understanding of the historical significance of each structure within its environment, such a study would require a larger time investment and that does not seem to suit the speed required by most restoration operations. Additional criteria were also considered in the choice of which caravanserais to survey, including the construction period and the availability of a reliable construction date provided either by the epigraphic evidence (Çardak Han and Òncir Han) or by textual sources (Evdir Han).29 The existence of sources that enable us to date these caravanserais is an important factor, as it allows us to better contextualise the material data. Aside from the three caravanserais chosen as the main objects of investigation here, our survey also included other road caravanserais to help place the collected data in comparative perspective. Some of these caravanserais have already been restored, such as Sultan Han in Aksaray30 or Zazadin Han,31 near Konya. Some other non-restored edifices were considered to be less significant within the framework of this preliminary study, either because they are less well preserved (i.e., Dolay Han or Do©ala Han) or because they do not seem to be threatened by an imminent restoration project, as is the case with Kargı Han. Eleven buildings were surveyed in total: Òncir Han, Evdir Han and Kargı Han in the Antalya region; Çardak Han in the Denizli region; Sultan Han and Zazadin Han in the Konya region; and Alay Han,32 Sari Han, Dolay Han, Do©ala Han, Tuzhisar Sultan Han and Karatay Han in the Kayseri region. The description of the surveyed buildings focuses on those elements that can provide the most significant data at this stage of the research. Some elements could be observed only from a distance, and it is important to note that it is impossible to conduct a detailed

Not for distribution or resale. For personal use only.

63

64

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

technical study of the apertures without a precise metric survey. As such, the level of detail of the description varies greatly for different elements (e.g., buttresses, openings, etc.). The portal, for example, is usually only briefly explained in my description and only its masonry and stratigraphic relationships are considered; the decoration of the caravanserais’ main entrances has not been taken into account in this study. Finally, the collapse of some portions of the buildings frequently made it possible to observe the characteristics of the wall core. Evdir Han The caravanserai of Evdir Han is located about 20 km north of Antalya, on the road to Burdur, and it is apparently one of the oldest extant dated caravanserais (Figure 3.1; Plate 4). An inscription, which is now lost but that Rudolf M. Riefstahl could see in 1928,33 states that the caravanserai was built by ‘Izz al-Din Kay Kawus I (r. 1210–19). The caravanserai consists of a large open courtyard surrounded on all sides by open arcades. The portal faces south and buttresses align on the outer walls of the building; on the western side one of them was used as a latrine; while the function of the buttress west of the portal, which is also connected to the vaulted gallery, is unclear. No windows open in the perimeter wall. The southern side is the best preserved part of the building; a large portion of the vaulting system is still standing and the perimeter

Figure 3.1  The courtyard of Evdir Han (photograph: Cinzia

Tavernari)

Not for distribution or resale. For personal use only.

stones for travellers

wall almost retains its original height. Part of the original flooring is also preserved in this area of the caravanserai, both the bedding and the floor surface, which is made of a layer of mortar. The gallery is also partially preserved on the eastern side and, to a lesser extent, on the western side; here the Èwån that marked the east–west axis of the caravanserai has disappeared, retaining its original characteristics only on the eastern side. The northern side has mostly collapsed, rendering it difficult to identify the original layout; similarly, in this area, the perimeter wall is mostly reduced to just a few courses (horizontal layers of stones) of masonry. Only a thorough cleaning of the area and, possibly, archaeological excavation will allow for the retrieval of precise information on the northern side of the building. The perimeter wall of the courtyard The external face of the perimeter wall is preserved on three sides of the caravanserai (south, east and west), while on the northern side only the exposed core survives. The perimeter walls of the structure comprise a core between two faces for a total width of approximately 140 cm. The core is banked in correspondence with the external courses and reflects their height. Material of various sizes is used in the core, but there is a prevalence of medium-sized roughly hewn material bound with mortar. The perimeter wall has an external face of smooth-dressed ashlars (a finely dressed stone with straight edges) and the stone is laid in horizontal, parallel courses bound with mortar. Both horizontal and vertical joints are narrow. The majority of the ashlars are probably specifically prepared; they are laid lengthways, with rare ashlars laid head on. In a few cases, joints overlap over two or three courses, or stones are partially cut to adapt to other material.34 The stones appear to be dressed on five of the faces and trimmed on the sixth. No perpend stone (a stone that extends through the width of the entire wall) could be identified, neither could any particular system of bonding the core and the outer face together. It was not possible to verify if the ashlars, laid head on, provide further bonding between the core and the wall, but they appear to be so few that it is unlikely, and at any rate this technique would not have been effective. The stones of the outer wall appear to have an average depth of about 50 cm. Some examples of masons’ marks occur on the outer faces of the stones. The outer face of the perimeter wall thus presents high-quality masonry and appears to be quite homogeneous, albeit with some exceptions. On the western side, where preserved, the upper part of the external wall consists of a different and rougher masonry than the rest of the façade. Here, the wall displays big, trimmed blocks, most of them laid head on. The joints are very thick and involve the use of numerous stone wedges. The same kind of masonry seemingly appears also on the eastern side, but in this case only a few stones are preserved. This radical change in the material employed for the masonry of the outer façade could result from

Not for distribution or resale. For personal use only.

65

66

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

different causes. It could indicate: (1) a destructive event involving a large part of the building, followed by a reconstruction; (2) a scarcity of skilled stone-cutters able to produce ashlars – possibly due to financial problems; or (3) a structural change, for example, from the perimeter wall to the vault. At this stage of our research we cannot offer more than mere hypothesis, but some evidence seems to support the latter possibility. In addition to the fact that the masonry, where it is preserved, changes at exactly the same level all along the western wall and at a similar height also on the eastern side,35 it is possible to observe that, at this same level, the core changes as well. From this level on, the core is not banked because it is no longer the core of a two-faced wall, but the conglomerate masonry that backs the vault. Nonetheless, only a more comprehensive investigation could provide conclusive data to confirm or dismiss this opinion. The second element that disrupts the homogeneity of the external perimeter wall can be detected on its southern side, east of the portal. At the bottom of the wall, near the buttress, we notice two courses of ashlars that stand due to their dimensions. These stones are bigger than those employed in the rest of the wall; they consist of a different lithotype (a specific geological unit characterised by its specific lithology, i.e., type of stone); and their external surfaces present different finishes. Some surfaces present a banding and are smooth-dressed with a flat-bladed chisel, while others are finished with a toothed tool. These stones seem to be the only example of reused material in the outer face of the perimeter wall. The internal face of the perimeter wall is very different from its external counterpart. Unlike the high-quality masonry that characterises all the outer faces of the caravanserai, the masonry of the internal face is uneven. Moreover, its treatment appears to be different depending on the side of the building, as higher quality stones were clearly used only for the entrance. In addition to ashlars, the internal face of the perimeter wall employs squared (or roughly squared) blocks laid in horizontal and parallel courses bound with mortar. The main characteristic of the internal face of the perimeter wall is the use of a large quantity of reused stones,36 eventually recut. Therefore, the shape and dimensions of the material are uneven,37 resulting in thick bed joints and perpends often involving the use of both stone and brick wedges. Several inscribed or decorated reused stones occur at the bottom of the wall. Upper courses usually employ less obvious reused material or show stones with the decorated face placed backwards. Different kinds of finish are also visible on the external surfaces, where I could recognise the use of toothed tools, flat-bladed chisels, pointed chisels and, in some instances, quarry picks. In the latter case, the stones are probably laid backwards. The presence of such different finishes further strengthens the hypothesis that the majority of the stones employed in the internal faces of the perimeter wall are, in fact, reused material that was most likely recut. The upper part of

Not for distribution or resale. For personal use only.

stones for travellers

the internal face of the perimeter wall, from the level of the springing of the vaults, mainly employs medium-sized and often only roughly squared blocks, sometimes even trimmed blocks. In the southern side, though, trimmed stones occur only rarely even in the upper part of the walls. The perimeter wall possesses ashlar masonry on both sides in only one instance: the still preserved Èwån on the eastern side. The mortar of the joints is spread beyond the joints and smoothed flush with the stone surface. It is presently difficult to establish if this finish is original or not, but it most likely corresponds to an original feature of the building. In fact, the uneven material used would have conveyed a rough appearance that the use of reflow mortar in the joints would surely have smoothed and regularised.38 The buttresses Six square-shaped buttresses are still partially preserved at Evdir Han. Their masonry presents the same characteristics as the outer face of the perimeter wall, both in the construction of the core and the outer faces. The arches The slightly pointed arches (c. 94 cm wide) that sustain the vaults of the gallery rest on piers of ashlar stones. On the side of the perimeter wall, though, the arches spring directly from the wall. The two elements probably belong to the same building phase,39 even if the construction of the arches involved wedges and superposing vertical joints.40 Both the jambs of the arches and the voussoirs (wedge-shaped stone forming the curved part of an arch) appear to have been built with specifically prepared ashlar stones. Some voussoirs bear masons’ marks. The spandrels (the area between the extrados of two adjoining arches) of the arches are usually built with medium-sized roughly squared stones, ashlar stone being used only occasionally and particularly on the southern side, near the entrance. The spandrels form the support for the vaulting system and their upper limit establishes the level of the impost of the vaults. Some arches have a keystone, while others do not. Most of the external surfaces of the voussoirs are finished with a flat-bladed chisel. A fallen voussoir made it possible to observe its dressing and finish on all the faces. The backward face of the stone possesses large drafted margins and a lower central surface, roughly finished with a point chisel or pickaxe. This treatment created a better bond between the voussoirs and the spandrels or the vault. The vaults The galleries of the caravanserai are covered with slightly pointed barrel vaults made of small rough-hewn stone, bound with abundant mortar and laid in quite regular courses parallel to the impost of the vault; some of the stone is laid lengthways. Two holes that may have

Not for distribution or resale. For personal use only.

67

68

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

been used as seating for centring are visible in the entrance corridor, but this is the only occurrence of this element that I could identify. It is still possible to see the marks left in the mortar by the use of centring (a temporary framework for supporting a masonry arch or vault), particularly between the haunches and the upper part of the vault. This element could indicate that in some cases the centring was used only in the latter stage of the construction of the vault, but not for the haunches (the side of an arch, between the crown and the pier). The haunches use the same material as the vault, except in the Èwån, where the former are ashlar stones laid in courses parallel to the impost of the vault and with narrow perpends (vertical joints) and beddings (horizontal joints). On the southern side, the vaults feature strengthening transverse ribs in ashlar stones that rest on simple moulded corbels springing from piers or embedded in the walls. The coating still covering the internal faces of some vaults may not be the original finish, but a plaster or mortar coat undoubtedly initially masked the roughness of the vault material. The openings The portal comprises an arched entrance surmounted by a stalactite half-dome. The dome comprises a pointed arch framed by a rectangular cornice, whose upper part is now destroyed. A muqarnas niche stands on each side of the entrance. The basket arch that provides access to the caravanserai has a keystone with darker colour than the rest of the voussoirs. All the elements are perfectly cut to measure in order to adapt to their position in the portal. The stones present a smooth-dressed surface finished with a flat-bladed chisel and, in some cases, they also show a draft margin executed with the same tool. I could identify the presence of a marble block, probably a reused stone that had been recut and employed in the portal as the uppermost part of one of the muqarnas niches. Òncir Han This caravanserai (Plate 5) lies on the road leading from Antalya to Burdur, approximately 50 km north of Evdir Han; it stands in a fertile green valley, some kilometres to the east of the modern road. An inscription still preserved in situ mentions the Seljuk sultan Ghiyath al-Din Kaykhusraw II (r. 1237–46) as the patron of the caravanserai and gives the date 636/1238–9. Recent cleaning and excavation have exposed the layout of the caravanserai’s courtyard and most of the façades of the vaulted hall. In comparison with the time of Erdmann’s visit,41 when several metres of soil covered most of the external façades of the caravanserai, this is a very different situation. The caravanserai faces south and comprises a fairly well-preserved covered hall (Figure 3.2) and a ruined courtyard, where almost nothing is preserved above soil level. The covered hall has a central

Not for distribution or resale. For personal use only.

stones for travellers

Figure 3.2  The interior of the vaulted hall of Òncir Han (photograph: Cinzia

Tavernari)

Not for distribution or resale. For personal use only.

69

70

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

and higher aisle flanked on each side by two smaller aisles. The vault of the central aisle runs in a north–south direction, while the lateral aisles are covered by vaults running in an east–west direction. Several buttresses of different forms align both along the wall of the covered hall and of the courtyard. The latter apparently comprised a central, open space with a series of small spaces aligning along its eastern side and two rooms flanking the entrance on the southern side. A course of springers (the first voussoir resting on the impost of an arch) jutting from the façade of the vaulted hall indicates the presence of a covered space also in the western part of the courtyard. The perimeter wall of the covered hall The perimeter wall of the covered hall is quite well preserved on all sides, particularly on the western side where it almost maintains its original height; on the other sides, the core is sometimes largely visible. The perimeter wall comprises a core between two faces for a total width of approximately 220 cm. The external face employs ashlar stone laid in horizontal, parallel courses bound with mortar; most of the material is laid lengthways, with rare ashlars laid head on. The core is banked in correspondence with the external courses and reflects their height. Material of various sizes is used in the core, but there is a prevalence of medium-sized roughly-hewn material bound with mortar. The core and the outer face of the perimeter wall do not appear to be connected through specific bondstones.42 The stone appears to be dressed on five of the faces and usually roughly squared on the sixth; the latter is finished with a pickaxe, while the others are smooth-dressed with a toothed tool. The external surface also presents a banding of approximately 3–5 cm. Both horizontal and vertical joints are narrow and there are no wedges. In the lowest portion of the wall, on the southern, northern and western sides of the building, the masonry employs some reused stones – possibly recut. The rest of the material probably consists of specifically prepared new stones. There are few occurrences of uncrossed joints, although ‘L’ joints are more common. The outer face of the perimeter wall thus presents homogeneous high-quality masonry. On the western and southern sides, though, the lowest visible course is different from the rest of the masonry. This course employs reused stones, some roughly-squared stones and ashlars whose external surface often shows a roughlyexecuted finish, probably with a pickaxe. Moreover, on the western side, the course juts from the vertical plane of the external wall face. This course could have had the function of connecting the foundation of the covered hall with the rest of the wall, but at the present stage of research this is a hypothesis that needs to be confirmed by further investigation along the other sides of the building. The internal face of the perimeter wall presents the same characteristics as the outer face, with the difference that only one example of clearly reused stone has been identified in the internal face of the wall in the covered hall.

Not for distribution or resale. For personal use only.

stones for travellers

71

The perimeter wall of the courtyard The masonry of the perimeter wall of the courtyard presents the same characteristics as the perimeter wall of the vaulted hall, both on its external and internal faces. In the northeastern corner, where the core is exposed, it was not possible to detect any seam between the perimeter wall of the vaulted hall and of the courtyard; the two elements are thus strongly connected. On the western side, the perimeter wall of the courtyard is largely in ruins and most of the time only one course of it is preserved. The latter shows a nicely decorated ashlar of large dimensions next to the destroyed buttress. The stones present a smooth-dressed toothed tool finish on five of their faces, while the sixth bears a rougher finish, executed with a pointed tool. The lower course of the internal face of the northern perimeter wall is comparable with that on the western side of the vaulted hall: it employs roughly-squared or trimmed blocks, but does not jut from the vertical plane of the rest of the wall surface. The buttresses The buttresses present the same masonry characteristics as the perimeter wall. A mason’s mark has been identified on one of the buttresses, on the western side of the vaulted hall (Figure 3.3). The horizontal joints of the buttresses and the perimeter wall do not

Figure 3.3  The mason’s mark on the buttress of Òncir Han (photograph: Cinzia

Tavernari)

Not for distribution or resale. For personal use only.

72

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

usually align, so the stones must often have been recut in order to provide for a structural bond between these two elements. Notwithstanding the recurrent presence of ‘L’ joints as the result of recut stones, I believe that the buttresses and the vaulted hall belong to the same construction phase. The buttresses comprise a square platform over which develops a full masonry body that can take different shapes. In the courtyard, only one buttress is preserved above the platform level and shows a polygonal shape. The western side of the covered hall displays three rectangular buttresses and a circular one in the northwestern side. On the northern side, there are two polygonal buttresses and again a circular one at the corner with the eastern side. The situation of the latter is more varied as it alternates between a polygonal, circular and rectangular buttress at its southeastern corner. It is interesting to note that the rectangular bases of the buttresses display different heights. The arches of the covered hall The slightly pointed arches (c. 108 cm wide) that sustain the vaults of the covered hall present a double row of voussoirs; they rest on piers of ashlar stones (c. 118 cm) or lie on ashlar springers embedded in the perimeter wall. The building of the arches involved occasional superposing of the vertical joints and ‘L’ joints, but the construction of the two elements belongs to the same constructive phase. Only specifically prepared stones were used for the voussoirs and the piers of the arches. The arches of both the central and lateral aisles have a keystone, except for three on the northern side. The voussoirs are smooth-dressed with a toothed tool, thus presenting the same finish as the ashlars employed in the perimeter wall. One occurrence of masons’ marks has been identified on a voussoir. The spandrels and the abutments exclusively employ finely-dressed ashlar stone. The vaults Vaults are preserved only in the covered hall. In fact, nothing remains of the two vaults that sprang from the southern façade of the covered hall except for a row of springers on each side of the portal. The springers are in ashlar stone, but whether the entire vault also employed ashlars is now impossible to assert. The covered hall displays slightly pointed barrel vaults exclusively made of ashlar stone laid in perfectly horizontal courses parallel to the impost of the vaults. Each vault has a keystone whose shortest side dimension is usually smaller than the rest of the material. In the southeastern corner of the covered hall, part of the vault has collapsed and it has been possible to observe that some of the ashlars, particularly the keystone, functioned as bond stones in order to provide a stronger connection between the vault and its core. In the central aisle, the vault features strengthening transverse ribs in ashlar stone that rest on simple moulded corbels embedded in the walls of the central

Not for distribution or resale. For personal use only.

stones for travellers

aisle. In the southeastern vault of the covered hall, one of the ashlars used as a keystone displays a human head, sculpted in high relief. It appears to be a unique occurrence and raises the question whether this decorative element could indicate a reused and recut stone. The openings A series of slits open in the perimeter wall of the vaulted hall; these openings develop in a niche recessed in the thickness of the perimeter wall and are perfectly bonded to it. The pointed vault of the niche has a keystone that penetrates deeper into the core of the wall than the other voussoirs in order to provide a stronger bond between the two elements. Due to their height, their only function could have been to provide light to the vaulted hall. The inclination of their base (i.e., towards the interior of the vaulted hall, rather than towards the exterior) also supports this assumption. The portal consists of a protruding niche surmounted by a vault in the shape of a scalloped shell that frames the Arabic foundation inscription. A basket arch with joggle-jointed voussoirs locked in place by a keystone gives access to the vaulted hall. All the material is prepared specifically for the portal and is shaped according to its position within this element. The external surfaces of the stones are smooth-dressed with a flatbladed chisel. Çardak Han This building lies in the small city of Çardak, approximately 50 km east of Denizli on the road that leads to Dinar, Òsparta and finally to Konya. The caravanserai is dated by an in situ inscription (Figure 3.4) that attributes its foundation to Ayaz bin Abdallah al-Shahabi (or Shihabi),43 an emir of the Seljuk sultan Ala al-Din Kayqubad I (r, 1220–37), in the year 627/1230. The caravanserai faces east and comprises a fairly well-preserved covered hall and a ruined courtyard (Plate 6). The hall has five naves, the central one being larger and higher than the lateral ones. In the southeast corner of the hall, a staircase leads to the roof and maybe helped to illuminate the hall. In fact, no windows opened in the walls of the hall, its only source of lighting being from a series of skylights that open in the vaults.44 An excavation carried out between 2006 and 2008 has exposed the layout of the courtyard and revealed that the caravanserai had a mosque and a bath located next to the entrance of the courtyard on the southern and northern sides, respectively. Some of the partition walls that defined the rooms are still preserved on the northern side of the courtyard, where the excavations have identified the presence of kitchen-related facilities. The southern side of the courtyard is less well preserved, but the traces of two arches45 springing from the main façade of the hall and from the wall of the courtyard on the opposite side point to the existence of a porticoed space, which

Not for distribution or resale. For personal use only.

73

74

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Figure 3.4  The portal and the inscription of Çardak Han (photograph: Cinzia

Tavernari)

Not for distribution or resale. For personal use only.

stones for travellers

appears to be confirmed by the archaeological evidence. Some restorations, carried out at an unknown date, appear to have been limited to the application of a new coating and the creation of a new concrete floor in the covered hall.46 The perimeter wall of the covered hall The perimeter wall of the covered hall is fairly well preserved on all its sides. It comprises a core between two faces for a  total  width  of  approximately 150 cm. The core is visible only in the southeastern and northeastern angles and behind the portal. The core is banked in correspondence with the external courses and employs mainly mediumsize, roughly hewn or unworked material bound with mortar. The entrance façade is built with ashlar stone laid in horizontal, parallel courses and bound with mortar; most of the ashlars are laid lengthways and only few are laid head on. Both horizontal and vertical joints are narrow and there are no wedges. All the material appears to be specially prepared and dimensions are fairly standardised. The ashlars show a smooth-dressed external surface with a banding of approximately 5 cm; both the banding and the central part of the surface exhibit a flat-bladed chisel finish. Some examples of masons’ marks appear on the outer face of the ashlars, mainly six-pointed stars. The southern side also employs exclusively ashlar stones laid in horizontal parallel courses and bound with mortar. The material is laid lengthways, with few occurrences of stones laid head on. Both beddings and perpends are narrow and they do not involve the use of wedges. The material employed in the southern façade is thus very similar to the eastern one but for the finish of the external faces of the stones. The ashlars of the southern side, in fact, present a wide variety of finishes, an element that could suggest the presence of several reused stones. Some have smooth-dressed faces finished with a toothed tool, while others present rougher surfaces, finished with a point chisel or, in some cases, with a pickaxe. There are also few examples of rusticated stone. Near the southwestern corner, it is possible to note a construction seam that involves the use of a few wedges and features bedding joints that do not align. In this case, wedges have been used instead of stones recut to adapt to the alignment of the joints. In the southeastern corner two stones significantly protrude from the wall, apparently in order to act as bond stones between the vaulted hall and the courtyard. The western outer façade of the perimeter wall employs squared or roughly squared stones laid in horizontal and parallel courses bound with mortar; the blocks are usually laid lengthways, with few elements laid head on. Due to the usage of several reused stones, the material is uneven, even if blocks of sizeable dimensions largely prevail. Both beddings and perpends, therefore, are quite thick and often involve the use of stone wedges of smaller dimensions. It is possible to observe that the inconsistency of the material employed also caused the doubling of

Not for distribution or resale. For personal use only.

75

76

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

a course. The external surfaces of the stones often present a rough finish, mainly executed with a point chisel or a pickaxe. Similar to the western façade, the northern façade also employs roughly-squared stones laid in parallel and horizontal joints bound with mortar, mainly of large dimensions. Both vertical and horizontal joints are thick and involve the ubiquitous use of smaller stone wedges. Most of the material is laid lengthways, with few stones laid head on. In several cases, the external face of the blocks is severely degraded and it is not possible to distinguish their finish; where the surface is better preserved, though, it presents a rough finish executed with a point chisel or a pickaxe. In the northeastern corner, a protruding stone appears to have been placed in order to act as a bond with another structure. Even though it was not possible to thoroughly examine the internal face of the perimeter wall due to a thick plaster coating covering most of the evidence, it is clear that the outer and internal faces show completely different masonry techniques. The latter employs mostly medium-size rough-hewn blocks and possibly unworked stones laid in tendentially horizontal and parallel courses bound with abundant mortar. Only on the eastern side, at the back of the portal, does the masonry employ roughly squared stones, sometimes of larger dimensions than the rest of the material. This better quality material allows for thinner joints and does not involve the use of numerous wedges to keep the horizontality and height of the courses more regular. The internal faces of the covered hall do not seem to employ reused material, as is the case with the external faces. The uneven characteristics of the material result in thick beddings and perpends, usually involving the use of wedges. A layer of whitewash most likely covered the wall surface inside the vaulted hall, or at least mortar was spread beyond the joints and smoothed flush with the stone surface. The perimeter wall of the courtyard The perimeter wall of the courtyard comprises a core between two faces. The core is mainly built of medium- to small-sized unworked stones, but it has not been possible to distinguish if the material is banked due to its poor state of preservation. The external and internal faces of the perimeter wall of the courtyard present different characteristics. All the internal faces employ roughly squared or trimmed stones, most medium sized, laid in intentionally horizontal and parallel courses. The shape and dimensions of the material are uneven, resulting in thick perpends and beddings, which involve the use of small- or medium-sized wedges. Bricks are also used either as wedges or constructive material. It is possible to observe that the mortar was spread beyond the joints and smoothed flush with the stone surface. No traces of a further coating have been identified so far. On the southern side, the wall mainly employs squared blocks of large dimensions laid in parallel and horizontal courses with narrow horizontal and vertical joints. No wedges appear on the outer face. The external

Not for distribution or resale. For personal use only.

stones for travellers

surface of the stones shows a rough aspect, but at present it is difficult to understand if this is the result of a rough finish or the consequence of the significant deterioration process that has affected the stones. As yet, no trace of a peculiar finish has been detected. The masonry of the eastern façade of the courtyard presents the same characteristics as the southern one, with the exception of the ­external surface of the stones, which present a smooth-dressed finish executed with a flat-bladed chisel. The southern façade of the vaulted hall does not present the same level of deterioration as that of the courtyard façade, so this phenomenon could be more the result of the different quality of the material than of the wall orientation. Finally, the northern side of the courtyard mostly employs roughly squared stones, as well as some medium-sized reused stones laid in horizontal and parallel courses bounded with abundant mortar. The vertical and horizontal joints are thick and involve the use of numerous small-sized wedges. The buttresses The building has buttresses on the eastern, southern and northern sides of the vaulted hall and on the southern side of the courtyard; they are all quite well preserved and clearly bonded to the rest of the wall. Apart from the two polygonal buttresses on the southern side of the vaulted hall, the others employ the same masonry as the perimeter wall to which they are bonded. The polygonal buttresses of the southern façade present the same masonry technique as the two cylindrical buttresses of the entrance side. These four elements employ smooth-dressed ashlar stone laid in horizontal and parallel courses. Both the horizontal and vertical joints are narrow and there are no wedges. The external surface of the stone presents a flat-bladed chisel finish and a banding executed with the same tool. The buttress on the southern side of the courtyard has a triangular shape like those on the northern side of the vaulted hall. The arches The round arches (c. 92 cm wide) of the covered hall run perpendicular to the entrance façade. They employ specifically prepared stone for the voussoirs and they rest on piers of ashlar stone (c. 92 cm × 113 cm). Some of the arches have a keystone, and a mason’s mark has been identified on one of the voussoirs. The spandrels of the arches of the central nave employ ashlar stone, while in the lateral naves rough-hewn stones appear to have been used. However, the plaster coating that almost entirely covers the lateral naves makes it difficult to be more precise on the characteristics of the stones used for these elements. The external surfaces of the voussoirs are smooth-dressed with a flat-bladed tool, as are the ashlars employed in the spandrels. In this case, the external surface also shows a draft margin, which also appears on the bottom surface, associated with a pickaxe or pointed chisel rough finish.47

Not for distribution or resale. For personal use only.

77

78

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

The vaults The hall is covered by north–south-oriented barrel vaults made of small rough-hewn stone bound abundantly with mortar. It is difficult to perceive the arrangement of the stones in the vaults (i.e., if they are laid in courses) because of the thick plaster coating that almost completely covers them. The haunches are visible only in the central nave, where they employ roughly squared or trimmed medium-sized stone laid in two horizontal courses parallel to the impost of the vault. Here, the haunches lay on a course of ashlar stone whose upper edge constitutes the impost of the vault. Two holes that were probably used as a seating for centring open in the second course of the haunches; their position suggests that centring was used only in the latter stage of the vault construction. The vaults feature strengthening transverse ribs in ashlar stones, which, in the central aisle, rest on piers with triangular-shaped corbels, sometimes bearing a decoration. The three decorated corbels all align on the northern side of the central aisle and, from west to east, they bear the images of two fish, a human or monkey head and a bull’s head, respectively. The images are carved in high relief, and the presence of a mason’s mark can be noticed between the two fish of the westernmost corbel. In the lateral naves, the transverse ribs rest on springers embedded in the wall and laid protruding from the vertical plane. The piers, the corbels and the transverse ribs all present an external surface smooth-dressed with a flat-bladed tool. The thick coating covering the internal faces of the vaults most probably is not original, but I believe it actually duplicates an original feature of the caravanserai that was intended to cover the rough and uneven appearance of this element. The openings The portal of the vaulted hall consists of a niche formed by a slightly pointed arch that was framed by a rectangular cornice, now largely destroyed. In the niche, a basket arch with joggle-jointed voussoirs locked in place by a keystone gives access to the vaulted hall. The foundation inscription is placed over the door, framed by two lions resting on stalactites decorated corbels. The portal employs only specially prepared material with smooth-dressed surfaces finished with a flat-bladed chisel. Constructional and Typological Characteristics The Construction Process The preliminary analysis of the selected caravanserais suggests that the current appearance of Evdir Han and Òncir Han is mainly the result of a single phase of construction. This condition implies that the general layout and main architectural elements of each caravanserai have been determined by a single major building campaign

Not for distribution or resale. For personal use only.

stones for travellers

and have not subsequently been altered so as to become unrecognisable.48 The later modifications that I could clearly identify in these buildings do not alter the original layout of the caravanserais, but consist of restorations or of modest structures abutting on the external walls of Evdir Han and Òncir Han. The parasite structures in the latter’s courtyard quite securely belong to a much later date than the Seljuk period; in fact, these structures employ exclusive reused material from the caravanserai, including decorated stones from the portal of the vaulted hall, thus revealing that Òncir Han was already in ruins at the moment of their construction. The most complex case is Çardak Han. In this building, at least the upper parts of the perimeter walls of the hall and of the courtyard, belong to two different activities because the masonry of these elements is not regularly bonded together. I could not investigate the lower parts of these walls due to vegetation.49 The vestiges of an arch springing from the eastern façade of the vaulted hall, however, appear to be in phase with the façade’s masonry, thus suggesting that a courtyard was part of the original construction plan of the caravanserai. The presence of stones protruding from the southern and northern façades of the vaulted hall to act as a bond with another structure also supports this hypothesis (Figure 3.5). Accordingly, a courtyard was most probably planned from the beginning, but at the present stage of research it is impossible to state if the existing wall of the courtyard is partly the result of a restoration or if it belongs to the same construction phase of the vaulted hall and simply indicates the sequence of the construction process. Regarding Çardak Han as the result of a single building phase, it appears that the masons first built the vaulted hall and then moved on to the courtyard. This sequence is generally considered as the usual construction process employed in the caravanserais formed by a vaulted hall and a courtyard. Òncir Han, however, does not follow this procedure as its vaulted hall and courtyard were built concomitantly. The stratigraphic analysis, in fact, shows that the masonry of these two elements is regularly bonded together.50 The investigation of the constructive characteristics of the caravanserais’ main portals also provides useful information on the construction process of these buildings. The external sides of the protruding portals present an irregular bond to the perimeter wall of the courtyard or of the hall that entails the presence of recut stones or, less frequently, construction seams or wedges (see Figure 3.6). These characteristics can be observed both on the outer and inner faces of the wall. In this case, the stratigraphic bond does not indicate a temporal sequence, but a constructive technique. The use of decorated elements of specific forms and dimensions or the use of different material than the particular status of the portal involved, in fact, made it more difficult – from a construction point of view – to physically bond the portal with the perimeter wall of the courtyard

Not for distribution or resale. For personal use only.

79

80

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Figure 3.5  The bond between the courtyard and the main façade

(southeast corner) of Çardak Han (photograph: Cinzia Tavernari)

Figure 3.6  The bond between the façade and the portal (southwest

side) of Evdir Han (photograph: Cinzia Tavernari)

Not for distribution or resale. For personal use only.

stones for travellers

or the vaulted hall. For the same reason, the niches embedded in the lower part of the projecting portal usually simply abut on the jamb of the entrance arch. The analysis of the vault that covers the portal, though, permits inference of the contemporaneity of these different elements of the access system. The majority of the portals I could explore in May 2015 presented similar characteristics, which indicates that these construction practices were common to most Seljuk caravanserais. There are other cases where elements that are not always regularly bonded together can nonetheless be quite securely ascribed to the same phase of construction and probably to the same building activity;51 such is the case of the buttresses in the caravanserai of Evdir Han. The occasional non-corresponding bed joints and the ‘L’ joints between the buttresses and the perimeter wall are related to the technical difficulties involved in connecting two different constructive elements, similar to the case of the portals. The two spaces that open in the buttresses inside the gallery of Evdir Han appear to be in phase with the rest of the perimeter wall, thus supporting the contemporaneity of the two elements. After observing this structure in person, I disagree with earlier suggestions that the buttress on the southern side, to the west of the portal, underwent significant modifications. I could not detect any construction seam or even uncrossed joint indicating that the space inside it and its access could represent a later addition. The only clear modification seems to be a partial blocking of the entrance leading to the space inside the southern buttress. The Masonry As Scott Redford explains,52 Anatolian Seljuk architecture was concerned with the appearance of buildings. This is apparent in several caravanserais, such as Evdir Han, Çardak Han, Zazadin Han and, to a certain extent, also Kargı Han and Do©ala Han, where the most visible – and important – parts of the building receive particular attention. In Evdir Han, the use of ashlar stone for the external face of the perimeter walls clearly shows the pre-eminence of the exterior façades over the interior surfaces. Internally, it was easier to conceal the uneven appearance of the wall faces using plaster or an abundance of mortar smoothed flush with the surface of the stone, both probably common solutions. The same consideration applies to Zazadin Han and Çardak Han, where the entrance façades of the courtyard and of the covered hall show masonry of higher quality when compared with that of the other façades. There are also sometimes differences among the other façades and, in Zazadin Han, the eastern outer face of the courtyard appears to be more carefully built than the other sides of the building, while in Çardak Han it is the southern outer face of the covered hall. This difference could be

Not for distribution or resale. For personal use only.

81

82

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

explained in relation to the presence of the road used by the travellers on these sides of the buildings,53 but this hypothesis should be considered with caution at this stage of the research, especially in the case of Çardak Han, whose complexity I have previously mentioned. The lower quality of the masonry that is often used in the courtyard appears to confirm that the covered hall was the most important part of the caravanserai. In fact, even if the construction of the courtyard was apparently always planned from the beginning, the execution of this element and particularly the organisation of its space could have occurred at a later time. Among the non-restored examples that I have seen, only Alay Han employs fine ashlar stone in what is still preserved of its courtyard. Even in Òncir Han, the walls that define the spaces of the courtyard appear to use rougher material than the rest of the building, although what is now visible could actually belong to the foundation and not to the wall. For the rest, Òncir Han is a very homogeneous building that employs high-quality masonry exclusively composed of ashlars. The use of ashlar stone in the vaults clearly distinguishes Òncir Han from the majority of the other caravanserais, where vaults made with rough-hewn stones largely prevail even when the walls of the vaulted hall employ ashlar stone. The use of perfectly cut and dressed material in the vault points to an important financial investment that – u ­ nsurprisingly – well suits a royal foundation. It is not possible to assert that all sultan-founded caravanserais had vaults made with ashlar stones, as Evdir Han did not, but it surely designated a high-ranking patron. The use of ashlar even for the vaults could seem incongruous with the use of reused material: the first element points to an important investment for a significant building, while the second is usually associated with reducing the costs of construction. While reuse was often a practical choice very common in the medieval world, in the case of a particularly significant building the patron could purposely decide against the use of inferior material. The widespread use of reused stones in most caravanserais and even within the framework of a high-quality royal foundation such as Òncir Han (Figure 3.7) could suggest that this was the norm for Seljuk caravanserais whenever any construction material was readily accessible. Transportation costs usually represented the highest expense within the construction site economy,54 and could significantly increase the total cost of a building depending on its proximity to a suitable quarry. It is thus possible that the Seljuk sultans were ready to invest only a limited amount of money in the construction of a road caravanserai, even if it was a royal building. This statement implies several further ­questions – on the cost and availability of specialised skilled workers, for example – that it is impossible to answer at the present stage of research. It is of paramount importance, however, to identify the presence of quarries close to the caravanserais because such knowledge will allow a better understanding of the reasons underlying a specific choice

Not for distribution or resale. For personal use only.

stones for travellers

Figure 3.7  An example of spolia in Òncir Han (photograph: Cinzia Tavernari)

of construction material. I expect to carry out this research during future surveys. In the three caravanserais I have surveyed more thoroughly, identifiable reused stones are mainly placed at the bottom of the walls, in the rear walls or with the eventual decorated face placed backwards, as if the builders wished to hold them back. On the other hand, there are caravanserais where the reused material is clearly displayed and arranged with a decorative intent, such as in Kandin Han and Zazadin Han.55 The Tools In the caravanserais I have examined, the external surfaces of the stones present several different kinds of finish. Both toothed tools and flat-bladed tools appear to be used at the same time in a single caravanserai, even if sometimes the latter are used only in the portal, as in Òncir Han. It is interesting to note that in all the selected caravanserais the stones of the portal present smooth-dressed surfaces finished with flat-bladed chisels. This regularity points to the existence of specific groups of craftsmen involved in the preparation of the material for the portal, mainly its decorative elements. The use of several different tools on the stone surfaces of Evdir Han and Çardak Han confirms the employment of reused material; it could also point to the presence

Not for distribution or resale. For personal use only.

83

84

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

of different teams of workers with different habits. The existence in Evdir Han of a finishing characteristic from the Zengid and Ayyubid period in Syria56 (Figure 3.8) seems to point to the presence of masons of Syrian origin.57 Nonetheless, there are only two stones that present this kind of finishing (pointed chisel central part with a well-defined banding), and such a small number could point to a possible reuse of these elements. Alternatively, this finishing could represent a tradition common to all the Near East, but the absence of studies in the field of masons’ tools for the Byzantine and Seljuk period do not allow us to propose more detailed considerations. Apart from the portal, as I have mentioned, Òncir Han exclusively displays stone surfaces finished with a toothed tool, identified as an oriental diamond-point dressing hammer (shauta) by J-C. Bessac. It is interesting that some decorative elements reused in the caravanserai appear to exhibit the same finish as the other stones of the caravanserai. Moreover, some stone surfaces feature several layers of finish, all executed with a toothed tool, maybe always a shauta. At this stage of research some hypothesis are possible. Stonecutters from different periods could have used the same tool58 or, conversely, nearly all the stones used for the construction of the caravanserai could have been reused. A precise identification of the tools used on the stones that present several layers of different traces is paramount

Figure 3.8  Finishing of the outer surface of a stone in Evdir Han.

This finishing is similar to Zengid and Ayyubid examples found in the Citadel of Damascus (photograph: Cinzia Tavernari)

Not for distribution or resale. For personal use only.

stones for travellers

in order to reach more definite conclusions on the ratio of reused material utilised in Òncir Han. Conclusion Notwithstanding the preliminary nature of my research, I believe it is important to offer some conclusions even at this very first stage because I hope that my remarks will help to fuel historical and archaeological research on Seljuk road caravanserais. In fact, I hope that the following lines represent temporary conclusions that will be updated or even challenged when new research on road caravanserais becomes available. The eleven road caravanserais that I have surveyed represent only a small part of the total number of road caravanserais built during the Seljuk period; they present several similarities as well as some significant differences, particularly regarding their masonry. The kind of masonry used in the caravanserais can vary greatly because it appears to be heavily influenced by their location. This element seems to act as the predominant one among the primary factors that determine the construction of an edifice.59 This is evident in Kandin Han, for example, where the exterior of the perimeter wall exclusively utilises reused material, which it is possible to surmise was readily at hand.60 Kargı Han, a quite remote caravanserai located in the wooded foothills behind Antalya, mainly uses rough-hewn stones, thus adapting the masonry to the options offered by the location: no material to reuse and no good quarry at a convenient distance. Reused material appears to be used whenever it is available, even in royal caravanserais, and sometimes it displays a decorative intent, such as in Zazadin Han. The utilisation of reused material can influence the construction process, and the ubiquitous presence of ‘L’ joints in Òncir Han is probably a consequence of this practice. Alternatively, ‘L’ joints can indicate that, in some cases, the construction process progressed vertically.61 This seems to be confirmed by the way the buttresses and the perimeter walls are bonded together in Evdir Han and also, partly, by the ways in which the portals are connected to the rest of the masonry.62 In the caravanserais I have surveyed, the use of re-employed material does not appear to be a discriminating element giving information on the date of construction or allowing inference of the rank of the patron. On the other hand, the consistent employment of ashlars in a caravanserai and their use in the vaults provides a steady link to a particularly high-ranking patron. The widespread use of ashlars as construction material for road caravanserais – as well as the preference for cut-to-measure stones instead of wedges – points to the presence of skilled stone-cutters in Anatolia during the Seljuk period. The use of rough-hewn stones in certain caravanserais or in the less significant parts of the

Not for distribution or resale. For personal use only.

85

86

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

building thus appear to be mainly related to concerns about the production costs and not to a low level of technological skills and knowledge. Comparing the masonry of Seljuk road caravanserais to that of contemporary Ayyubid examples gives the impression that highly specialised stone-cutters did not exist in Syria. Ayyubid road caravanserais, in fact, invariably employ rough-hewn stones, using ashlars only for structurally significant elements, such as quoins and arches. A broader look at Ayyubid monuments, however, reveals that the use of rough-hewn stones apparently did not depend on the level of technological skills of the craftsmen.63 Though it is possible that fewer skilled craftsmen were available in Syria than in Anatolia – thus raising the cost of such workers – it appears at any rate that the Ayyubids were not interested in paying highly specialised workers for the construction of road caravanserais. This is consistent with the historical context of the region in the first half of the twelfth century: Ayyubids most probably preferred to focus on urban structures or, outside cities, on fortified structures. A common element to Seljuk and Ayyubid caravanserais has been identified in the finishing of some stones’ outer surfaces. This finishing has been defined as ‘taille ornementale pointée fine et cernée de marges régulières ciselées grain d’orge’,64 and appears in the caravanserais of Evdir Han and ‘Itna (Figures 3.8 and 3.9), currently a small town to the northeast of Damascus where an Ayyubid emir founded a caravanserai c. 1230.65 This specific kind of finishing is also present in the Citadel of Damascus and in the castle of Beaufort66

Figure 3.9  Finishing of the outer surface of a stone in the

caravanserai of ‘Itna, built in Syria c. 1230 (photograph: Cinzia Tavernari)

Not for distribution or resale. For personal use only.

stones for travellers

in Lebanon. Ayyubid Syria and Seljuk Anatolia also appear to share a second kind of finishing of the stone surfaces. The finishing is executed with a flat-bladed tool and characterises the main façade of Çardak Han; it can also be seen in the Citadel of Damascus.67 At this stage of the research, it is unclear if these common treatments of the stones surfaces reveal a technical tradition shared by a large part of the Near East, or if they are the result of an itinerant workforce or of a migration of craftsmen. Structures represent visual elements that ensure an immediate correlation between a geographic space and a power. In Ayyubid Syria, these elements are castles and citadels. In Seljuk Anatolia, on the other hand, this role appears to have been mainly bestowed on road caravanserais. The different scale – in terms of economic investment – of the road caravanserai phenomenon in Seljuk ­ Anatolia and Ayyubid Syria clearly points to a diverse significance of these buildings for the elites of these regions. In the same chronological set, an equally high-ranking patron belonging to the ruling power produced a very different result for seemingly the same purpose: fulfilling the needs of travellers, whether merchants, pilgrims or wanderers. Following Redford and Tate, I am convinced that Seljuk road caravanserais are ‘far more than a network of way stations for travellers’ and that they represent the ‘nexus point for a variety of functions’68 related to the projection of the Seljuk state in the landscape of Anatolia, both from a political and cultural point of view.

Notes 1. I would like to thank Patricia Blessing and Rachel Goshgarian for their tireless reviews and suggestions, which have greatly helped me to improve this chapter. I greatly profited of their deep knowledge of Seljuk history. I am also grateful to Scott Redford, Mustafa Önge, Jacopo Turchetto and Robin Wimmel for the ideas, suggestions and information on road caravanserais they kindly shared with me. They enabled me to enhance my comprehension of road caravanserais. I want to thank JeanClaude Bessac for answering my questions on the stonecutters’ tools. 2. See Önge, ‘Caravanserais as Symbols of Power’, pp. 49–69; Yavuz, ‘Anadolu Selçuklu Kervansarayları’, pp. 239–59. 3. Compare this figure with the seventy-two road caravanserais in use that I could identify for the Ayyubid and Mamluk periods in Bilad al-Sham. See Tavernari, ‘Caravansérails et réseaux routiers du Bilad al-Sham’. 4. The earliest dated caravanserai, Öresin Han, bears the date of 1188, while the latest dated caravanserai, Çay Han. was built in 1278. Redford, ‘Caravanserais, Roads & Routes’; Önge, ‘Caravanserais as Symbols of Power’, p. 65. 5. Redford, ‘The Alaeddin Mosque’, pp. 54–74. 6. Turan, ‘Selçuk Kervansarayları’, pp. 471–96; Turan, ‘Celaleddin Karatay’, pp. 17–170. 7. Erdmann, Das anatolische Karavansaray. In his three-volume work,

Not for distribution or resale. For personal use only.

87

88

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Erdmann sometimes refers to his fieldwork diaries, preserved at the Staatliche Museen zu Berlin – Museum für Islamische Kunst. On Erdmann’s diaries, see Blessing, ‘Recording the Transformation of Urban Landscapes’, pp. 415–25. 8. See, e.g., Ünal, ‘A propos de la datation des Khans Seldjoukides’; Bekta∞, Selçuklu Kervansarayları, p. 160. 9. Among the studies of individual buildings, it is particularly important to mention a number of unpublished Master’s theses in architecture; see, e.g., Önge, ‘The Restoration of Zazadin Han’. 10. Yavuz, ‘The Concepts that Shape’, pp. 80–95. 11. See, e.g., Redford, Landscape and the State; Redford, ‘Caravanserais, Roads & Routes’; Redford, ‘The Inscription of the Kırkgöz Hanı’, pp. 347–59. 12. Önge, ‘Caravanserais as Symbols of Power’, pp. 49–69. 13. Robin Wimmel has recently defended a very welcome PhD dedicated to the road caravanserais of the Ottoman period. Unfortunately, I have not yet been able to access this thesis. 14. The workshop ‘Cultural Heritage on the Road: The Caravanserais of Turkey’ was held at Abdullah Gül University in Kayseri, 18–22 October 2014. The workshop was organised with the generous support of the Barakat Trust Foundation, CIERAM (International Centre for the Study of Roads and Caravanserais) and Abdullah Gül University. 15. In addition to the works of Scott Redford cited above, see, e.g., Turchetto, ‘Cappadocia from Above’, pp. 399–407. 16. Kaymakcı, ‘Geoarchaeological Investigation’; Tate, ‘The Seljuk Caravanserais. Unfortunately, the latter research proposal has never been carried out, the author having abandoned his PhD studies. I wish to thank Prof. Scott Branting for this information. 17. Braudel, La Méditerranée, vol. 1, p. 339. 18. Tavernari, ‘From the Caravanserai to the Road’, pp. 711–27. 19. During a recent conversation, Jacopo Turchetto rightly pointed out that the rare GIS analyses that have been carried out on Seljuk road caravanserais lack both a historical and archaeological background. I wish to thank Jacopo Turchetto to draw my attention to this important point. 20. Erdmann’s publication produced an extensive effort to classify and date caravanserais on the basis of their layout. See Erdmann, Das anatolische Karavansaray. 21. The restoration projects that have been prepared as Master’s theses in several faculties of architecture in Turkey usually present the most detailed and up-to-date descriptions of the masonry employed in the caravanserais. For a good example, see Önge, ‘The Restoration of Zazadin Han’. 22. Although groundbreaking, Erdmann’s descriptions of the caravanserais’ masonry do not provide information on these elements. Moreover, some edifices have since been cleaned so that I could see them in better conditions than Erdmann, and was thus able to note details that he could not have seen. 23. Tonghini, Shayzar I, p. 95. 24. For an outline of the history of the discipline, see D’Ulizia, ‘L’archeologia dell’architettura in Italia’, pp. 9–42. 25. I would like to thank Patricia Blessing for pointing this issue out to me. 26. Òncir Han was excavated in 1993. See Ünal, ‘L’apport des fouilles d’Incir khan’, pp. 695–705; Ünal, ‘Òncir Hanı 1993’, pp. 117–29.

Not for distribution or resale. For personal use only.

stones for travellers

27. Çardak Han was excavated between 2006 and 2008 by the Denizli Museum. See Kutlu, ‘Seljuk Caravanserais’. 28. There has been a surge in road caravanserais’ restoration, especially in the last few years, in order to convert them in venues for tourists. To name a few examples, the covered section of Alay Han (Kayseri– Aksaray road) was reconstructed between 2008 and 2011; Kirkgöz Han (Antalya–Burdur road) was restored in 2007. 29. Riefstahl, Turkish Architecture, p. 62. 30. This caravanserai, one of the most famous examples of Seljuk road caravanserais, was first restored in 1957. 31. This caravanserai was restored in 2007–8, on the initiative of the mayor of the town of Selçuklu. 32. Alay Han comprised a covered hall and an open courtyard, both now destroyed. A restoration project that rebuilt the entire covered hall was carried out some years ago, probably in 2008. The courtyard, whose layout is presently distinguishable, has not been reconstructed and could still provide much useful information. It probably underwent some sort of excavation, but I could not secure any precise information on this issue. The exploration of the ruins of a separate small building, 20 m away from the caravanserai, next to the southwestern corner of the courtyard, could also yield interesting data. This structure’s function is unclear, but it was probably connected to the caravanserai. 33. Riefstahl, Turkish Architecture, p. 62. 34. Henceforth, this particular solution will be termed ‘L’ joints. See Brogiolo and Cagnana, Archeologia dell’Architettura, p. 144. 35. On the western side the masonry changes above the eighth course and on the eastern side above the ninth course. Due to the uneven level of the soil around the building we can fairly estimate that the change appears at the same level on both sides. 36. This is in contrast to the observations made by Spratt and Forbes. These travellers state that the stones of the caravanserai were cut specifically for this building in spite of the proximity of some neighbouring ruins. The ongoing cleaning of the caravanserai is probably responsible for this different appreciation. See Spratt and Forbes, Travels in Lycia, vol. 1, p. 226. 37. Both the height and the length of the stones vary greatly. Examples of sample measures include height: 32 cm, 47 cm, 58 cm; length: 65 cm, 110 cm, 220 cm. 38. In his article on Kargı Han, Scott Redford states that he believes that the plaster that coats the wall of the caravanserai is an original feature of the building. I agree with him, and I believe similar solutions were adopted in caravanserais where the walls present uneven surfaces. Traces of plaster can be seen in Evdir Han and reflow mortar appears in Evdir Han and Çardak Han. These finishes have probably been renewed several times but, as in Kargı Han, they probably reflect an original feature of the edifice. See Redford, ‘The Kible Wall’, p. 352. 39. The expression ‘building phase’ indicates ‘a set containing a number of building operations within a project unit’; see Tonghini, Shayzar I, p. 98. 40. Only further investigation will allow a clearer understanding of whether all the arches are concomitant with the perimeter wall. 41. Kurt Erdmann visited the site in 1953 and 1955. See Erdmann, Das anatolische Karavansaray, p. 107. 42. The rain spouts, even if they penetrate deeper into the core than the rest

Not for distribution or resale. For personal use only.

89

90

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

of the ashlars, cannot be effectively regarded as bondstones. They are too few. 43. The recent Master’s thesis by Kutlu provides a good summary of the researches carried out on this caravanserai, and of the studies and documents dealing with this building. Kutlu states the the use of the term ‘ribat’ in the caravanserai’s inscription could be related to the origin of its founder, who was from Syria. The preserved epigraphic record, though, clearly shows that the word ‘ribat’ is never used in caravanserais foundation inscriptions in Syria in the medieval period. In the most ancient recorded inscription, the caravanserai, founded by Salah al-Din in 1181 on the road from Damascus to Aleppo, is called a ‘funduq’. From the beginning of the thirteenth century, Syrian caravanserais are invariably called ‘khans’ in their foundation inscriptions. Written sources also tend to confirm that the word ‘ribat’ was not used for caravanserais in the Syrian region as, for example, the traveller Ibn Jubayr never uses it in relation to the Syrian caravanserais where he stops during his journey. See Kutlu, ‘Seljuk Caravanserais’, p. 27; Tavernari, ‘Caravansérails et réseaux routiers du Bilad al-Sham’, pp. 135–46. 44. The presence of these elements confirm that there was no second floor and that the staircase led only to the roof. 45. Nothing remains of one of the arches apart from the ‘negative’, that is, the construction seam that indicated its presence in the southeastern corner. 46. Kutlu mentions reparations and restoration works carried out in the 1920s, but he does not identify them. I could not distinguish them during my brief visit, but they possibly focused on the covered hall, which was then used as a sheepfold or barn. I maintain that the caravanserai does not appear to have undergone intrusive restoration. See Kutlu, ‘Seljuk Caravanserais’, p. 26. 47. The finish of the bottom surface I describe may not be a regular characteristic of the ashlar stones used inside the vaulted hall of Çardak Han. In fact, it was possible to observe only one occurrence of this finish thanks to a hole in the masonry. 48. For this reason, I consider the courtyard and the vaulted hall to be the result of a single phase of construction. For a more detailed explanation of the use of the term ‘phase’ used within the framework of the archaeology of standing structures, see Tonghini, Shayzar I, p. 98. 49. It is nonetheless important to note that the situation is very dirty and a thorough cleaning would be necessary in order to clarify the stratigraphic relationship. 50. This building constantly fails to display aligned horizontal joints between its constructive elements. This characteristic is usually considered as evidence of indenting operations that aim at bonding together elements that belong to different phases of construction. I believe that this is not the case here, as the presence of ‘L’ joints, one of the main clues that the archaeologist is facing two different construction phases, is too ubiquitous here to be the result of major chronological differences. I thus argue that the reasons for the ‘L’ joints should be searched for elsewhere, either in the organisation of the work on the building site or in the material (i.e., problems in supplying homogeneous material, large employ of reused stones). The skills of the masons could also be questioned. Erdmann states that the courtyard and the vaulted hall are not bonded together even if the construction of the former started when

Not for distribution or resale. For personal use only.

stones for travellers

the latter was near completion, as the course of springers jutting from the façade of the vaulted hall clearly proves. On the basis of the stratigraphic relationship I could observe, I argue, instead, that the perimeter walls of the courtyard and of the vaulted hall are bonded together. When Erdmann saw Òncir Han, both the walls of the courtyard and a considerable part of the walls of the vaulted hall were buried under a thick layer of soil that surely prevented a clear examination of the relationship between the vaulted hall and the courtyard. The caravanserai has now undergone excavation in order to retrieve the original soil level and to clear the walls from of accumulated soil. See Erdmann, Das anatolische Karavansaray, vol. 1, p. 109. 51. A building ‘activity’ can be defined as the construction work that is ‘carried out as part of a single building operation’; see Tonghini, Shayzar I, p. 97. The architectural elements composing an ‘activity’ are bonded together thus defining a stratigraphic relationship of contemporaneity. 52. Redford, ‘The Kible Wall’, p. 352. 53. Kutlu, ‘Seljuk Caravanserais’, p. 35. 54. ‘In ogni epoca il costo del trasporto, in termini di fatica umana e di tempo, era tra i più alti di tute le operazioni del cantiere.’ Cagnana, Archeologia dei materiali, p. 54. 55. On the use of spolia, see Blessing, Reframing the Lands of Rum, pp. 128–32. 56. This finishing can be found in the Damascus’ citadel. See Bessac, ‘Problématique et méthodologie archéologiques’, pp. 59–76. 57. This kind of finishing has been identified by J. C. Bessac, who has worked extensively on the Citadel of Damascus. I would like to thank him for providing me with this information. See Bessac, ‘Problématique et méthodologie archéologiques’, pp. 59–76. 58. Traces of a diamond-point dressing hammer already appear in HalabyyaZenobia (Syria) in the fifth and sixth centuries. See Bessac, ‘Techniques classiques de construction et de décor’, pp. 9–23, fig. 14. 59. These primary factors are: the nature of patronage; the availability of material; and the technical knowledge of the craftsmen. 60. In May 2015, I could not observe the interior façade of the perimeter wall because the caravanserai was closed. 61. See Brogiolo and Cagnana, Archeologia dell’Architettura, p. 138, figs 121, 122. 62. In the case of the portals, in fact, the eventual difference of the material employed and the presence of decorative elements of specific dimensions should also be taken into account as a cause for the presence of ‘L’ joints. Nonetheless, the fact that the ‘L’ joints also appear on the inner façade of the perimeter wall appear to support a construction that proceeded vertically. 63. See, e.g., the castle of Shayzar. The building CF2 is precisely dated by an inscription to the year 1233 and it employs finely worked rusticated ashlars in its perimeter wall. See Tonghini, Shayzar I, p. 79. 64. Bessac, ‘Problématique et méthodologie archéologiques’, p. 67. 65. Tavernari, ‘Caravansérails et réseaux routiers du Bilad al-Sham’, pp. 8–11. 66. See Bessac and Yasmine, ‘Étude préliminaire des chantiers’. 67. Bessac, personal communication. I would like to thank Jean-Claude Bessac for providing me with this information. 68. Tate, ‘The Seljuk Caravanserais’, p. 5.

Not for distribution or resale. For personal use only.

91

E[)JNBURGH University Press J

Not for distribution or resale. For personal use only.

CHAPTER FOUR

Suggestions on the Social Meaning and Functions of Akhi Communities and their Hospices in Medieval Anatolia Òklil Selçuk The Islamic tradition of futuwwa (Turk. fütüvvet), or ethical codes, received a certain degree of political legitimacy when the Abbasid caliph al-Nasir li-Din Allah (r. 1180–1225) admitted it to courtly life. Soon after, the caliph formally invited the Anatolian Seljuk sultan Izz al-Din Kay Kawus (r. 1211–20) to his futuwwa by sending him a prestigious emissary and gifts imbued with futuwwa symbolism. In time, the akhi brotherhoods of Anatolia adopted and popularised futuwwa teachings during a period of political fragmentation in the region. In fact, in the post-Mongol period, akhis rose to political prominence in many urban centres. Aside from their ethical teachings and political actions, akhi brotherhoods left their mark by erecting hospices (zåwiyas), where members of mystic orders conducted communal gatherings and hosted guests. This chapter will look into the possibility of distinguishing akhi communities of medieval Anatolia from the larger category of Sufis by examining the ways in which hospices built by these communities had particular social meanings and functions. More specifically, this chapter proposes that studying the locations of akhi hospices may allow us to further understand how akhi communities functioned in the late medieval period. As they were concentrated in towns along trade routes, akhi hospices might have had significant ties to the economy of this region. Relevant narrative sources buttress this supposition, especially futuwwa manuals that underline akhi teachings, such as their focus on professional crafts and the marketplace, and that might shed further light upon them in the broader context of medieval Anatolia. As per the difficulties with distinguishing Sufi and akhi buildings and the limited scope of this study, an art historical and architectural analysis beyond that offered by recent studies is not attempted here.

Not for distribution or resale. For personal use only.

96

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Historiography: Akhis, Dervishes and the Economy Although akhi brotherhoods have attracted the attention of many scholars, much of the literature on akhis is problematic.1 Some of the difficulties with studying akhis seem to relate to the sources and the way they are utilised by scholars. Some studies exclude crucial primary sources, often because of the philological skills necessary to interpret them. On the other hand, omitting source material in Arabic also stands out as an ideological concern with detaching Anatolian akhis from the Arabic-futuwwa tradition for the purpose of underlining their ‘Turkishness’.2 A thorough analysis of akhi monuments in Anatolia necessitates documentation of their inscriptions and a systematic analysis of these inscriptions along with endowment deeds of the pious foundations of which these monuments were a part. Both inscriptions and endowment deeds pose their own problems as to their value as documentary sources. Some endowment deeds housed in the archives are copies of originals, made at a later date. Due to scribal errors, some of these copies lack significant phrases to the extent that the original meaning is lost. Yet another problem stems from dilapidated monuments and scattered inscriptions; the gradual loss of the relatively modest buildings erected by akhi brotherhoods and the tombstones of individuals with the akhi epithet is a further challenge.3 Even in the rare cases where inscriptions and endowment deeds can be matched, further questions as to the identity of holders of the akhi epithet arise. Epigraphic and archival sources often lack specific information on akhi monuments that directly distinguishes them from the general category of dervish hospices. Questions such as whether the akhi epithet was acquired or inherited; if akhis assumed economic authority in medieval Anatolia; and the extent of akhis’ relationship to making production decisions and their involvement in artisanship remain unanswered.4 Last, but not least, the status of women among akhi communities is a question to be dealt with. Fifteenth-century Ottoman chronicler Å∞ıkpa∞azåde categorises the four main social groups of medieval Anatolia as frontier warlords (gåzÈyån-ı rËm), dervishes (abdalån-ı rËm), akhis (åhiyån-ı rËm) and the women’s branch of akhi associations (båcıyån-ı rËm). While it is tempting to dismiss women akhis as a distinctive category among medieval Anatolian communities for lack of evidence, sources of this period do mention women with the akhi epithet. Still, understanding the status of these women and how they acquired their epithets is difficult. There seems to be no way of knowing to what extent a formal training and organisation, as for instance in a Sufi order, was necessary for these women to receive their titles. Moreover, as far as land grants received by women with the akhi title were concerned, the land in question could be inherited. Women might have received their titles simply as daughters

Not for distribution or resale. For personal use only.

AKHI COMMUNITIES AND THEIR HOSPICES

or wives of prominent akhis. It is difficult to determine how much these women were involved in the activities of the akhis of the hospice or in the study of futuwwa ethics in a systematic fashion.5 Given these challenges, which relate to the existing literature about akhis, the historical sources and to the ongoing destruction of unassuming historical monuments such as dervish hospices, I suggest that only an extensive use of a variety of archival, epigraphic, narrative, archaeological and architectural material including inscriptions, tombstones, genealogies, hagiographies, futuwwa manuals and anecdotal information, wherever possible, might allow the differentiation of the functions of akhi brotherhoods and uncover details of their hospices. Such a task has not been undertaken so far, and this chapter is a first step in this direction. Recent studies on Sufism and art and architectural history have provided a wider outlook by dealing with issues such as the place of futuwwa ethics in ‘cities’ as central units in the multi-linguistic and multi-religious environment of medieval Anatolia.6 They have also placed akhis in a broader historical perspective by considering the social forces in the longue durée that shaped and transformed Sufi and akhi monuments.7 In the light of these studies and my research into foundation registers, I will start by looking into the possibility of further explaining the social functions of akhi communities and their hospices by considering their locations along major trade routes in late medieval Anatolia. Documenting Akhi Presence A diachronic scrutiny of foundation registers looking for deeds of akhi endowments shows that akhi hospices were concentrated around the towns of Kütahya, Kastamonu, Amasya, Tokat and Ankara. In an effort to assess the status of akhis in the early Ottoman period, I looked for initial land grants by early Ottoman rulers to Sufi dervishes, prayer leaders and akhis in the Hüdåvendigår cadastral surveys of the greater Bursa region. I found that the land grants to akhis were comparatively few in number. Based on these observations, I suggest that even though they seem to have been significant actors in medieval Anatolia, especially along significant trade routes, akhis lost their influence in Bursa earlier than in the other aforementioned cities, perhaps due to the Ottoman rulers’ systematic efforts to establish Bursa as a capital following its conquest in the early fourteenth century.8 Hence, the process of centralisation may have been more pronounced there. As to the specific functions of akhi hospices, they may be related to the links they constituted between urban and rural areas, along trade routes through which they facilitated the safe travel of merchants. A survey of endowment registers, dating from 1220 to 1600 and housed in the foundation registry archives (Vakıflar Ar∞ivi) in

Not for distribution or resale. For personal use only.

97

98

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Table 4.1  Comparative distribution of akhi endowments Location Ankara Kütahya Kastamonu Tokat Bursa and Òznik

No. of records

No. of separate akhi endowments

26 15 14 12 5

22 11 11 10  4

Ankara, shows the distribution of akhi endowments (Arab. waqf; Turk. vakıf) among the highest-ranking centres in comparison with those located in Bursa and Òznik, as shown in Table 4.1. According to this survey, in spite of its size and significance, Bursa was home to a very limited number of akhi endowments. The important role of the towns of Tokat, Amasya and Ankara in the medieval period has been underlined by Patricia Blessing, who has studied the post-Mongol situation in medieval Anatolia from an architectural point of view.9 From the broader perspective provided by this work, these small towns lay on routes that connected the Mediterranean and central Anatolia to the Black Sea, Tabriz to Konya, and the whole area to the greater system of the Silk Road.10 The Mongol invasion led to an increase in commercial activity across Anatolia, linking the region especially to Black Sea trade.11 Kütahya, to the west of the aforementioned towns, is a passageway between the coastal and central Anatolian areas. Considering Kütahya, I also tentatively suggest that the route along which these towns lay corresponds to the Byzantine frontier both on the western (Nicean) and northern (Pontus) lines.12 Land Granted to Akhis by the Early Ottoman Rulers Cadastral records from the Ottoman period, which comprise surveys of the population, land and other sources of revenue, provide another dimension for a scrutiny of akhi presence. While endowment registers of Anatolian pious foundations reflect the locations of dervish and akhi monuments throughout centuries beginning from 1200s well into the Ottoman period, the cadastral surveys of the Bursa region, administratively known as the Hüdåvendigår subprovince, supply information on land bestowed by early Ottoman rulers (temlÈks).13 These land grants provided the pioneers of the Bursa region, such as dervishes and akhis, with the initial opportunity to establish hospices that also led to new settlements. The collection of these cadastral records, Hüdåvendigår livası tahrir defteri, includes three surveys dated 1487, 1521 and 1573. The 1487 survey, however, also includes records of land grants by the earliest Ottoman rulers beginning with those granted by Osman Gazi (r. 1290–1324), the founder of the Ottoman dynasty. A similar

Not for distribution or resale. For personal use only.

99

AKHI COMMUNITIES AND THEIR HOSPICES

inquiry into the cadastral surveys concerning the akhi presence indicates that the overall number of land grants to akhis by Ottoman rulers around Bursa was modest, thus in conformity with the endowment registers. More specifically, Ottoman cadastral surveys of the Bursa region can further be scrutinised as to which significant social groups or individuals received land grants from the Ottoman sultans.14 Several questions emerge in this context, such as in what capacity did these groups use their land grants? What was the place of akhis in this enterprise? What was the proportion of land grants to dervishes and prayer leaders (fakÈhs) compared with those of akhis? Given the lack of established distinction between akhi and Sufi buildings, an issue that will be explored further below, one might expect these actors to have similar numbers of endowments in their respective appearance among early Ottoman settlements. But this is not at all the case. As shown in Table 4.2, according to the statistics extracted from the cadastral surveys of Hüdåvendigår, the number of land plots granted to akhis was considerably fewer than those granted to dervishes and fakÈhs. As land grants to akhis are significantly fewer in number compared with those made to sheikhs and to fakÈhs (Arab. faqÈh), it appears that sheikhs received more land grants than fakÈhs (Table 4.2). Despite obscurity in quite a number of records, what is available offers a discernible pattern.15 One of these observations not reflected in Table 4.2, since only the initial land grants are accounted for, is that initial land grants to fakÈhs generally passed on to fakÈhs in the coming generations. Who exactly were the fakÈhs as a social group? It appears that they mostly functioned as prayer leaders (imams) in the community. Although the more recognised meaning of fakÈh is ‘jurisprudent’, it is unlikely that every fakÈh who received a land grant in the early Ottoman period could have been a well-established religious scholar.16 In some cases, their function as prayer leaders or those who perform the call to prayer (müezzins) might explain why original land grants to fakÈhs tended to stay in the hands of future generations of fakÈhs. Theirs were quite technical services that would require some kind of religious expertise and training, even if not a formal madrasa education. In fact, some of these land grants Table 4.2  The distribution of initial land grants to sheikhs, fakÈhs and akhis according to the Hüdåvendigår livåsı cadastral surveys Land Grants

Osman . Beg

Orhan . Beg

Süleyman Pa∞a

Murad I

Bayezid I

Sheikhs FakÈhs Akhis

4

12  9  3

46 18  6

20 16  4

12 10  1

2

Mehmed Mehmed II I 1

Not for distribution or resale. For personal use only.

2 3

100

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

were accorded to whoever held the position of prayer leader of a certain mosque.17 The Similarity or ‘Sameness’ of Sufi and Akhi Hospices The historiography of medieval Anatolian social groups has concentrated on the issues of the conversion of Christian communities to Islam, the settlement of new immigrants into Anatolia, hospices that provided communal activities to the newly settled and existing population of the lands of RËm, and Turkification and the establishment of the Ottoman Empire. While Fuad Köprülü produced the pioneering works among Turkish scholars focusing on these issues,18 Ömer Lütfi Barkan’s 1942 article has been influential largely by defining the interactions between various social groups in Anatolia.19 Barkan’s work asserts that the success of the early Ottomans depended as much on an extensive settlement and colonisation process as on military activity.20 The main actors in the settlement and colonisation movement, whose land grants are documented, were sheikhs, babas (leaders of dervish groups) and akhis. This perspective places the hospices of colonising dervishes within a group of significance for economic history because they were intentionally constructed on the main migration paths and trade routes that linked settlements. In Barkan’s framework there does not seem to be a clear-cut distinction between the hospices of akhis and sheikhs.21 In her study of dervish lodges and dervish biographies, Ethel Sara Wolper contextualises the proliferation of Sufi lodges in Sivas, Tokat and Amasya within larger political events, such as the Båbå RasËl revolt (c. 1239–41) and the Mongol presence (c. 1243–1335), and focuses on the significance of Sufi hospices in community formation. Wolper underlines the general pattern of operation in the larger category of lodges, which included not only spiritual leaders along the path to find God (those who were served), but also groups of people who took care of mundane affairs and engaged in income-producing agricultural activity (those who served). From this perspective, the economic aspects of the akhi path do not seem to be specific to akhi brotherhoods. Moreover, considering the blurred boundaries between akhi and Sufi structures, Wolper suggests that not only did akhis occupy the same buildings with other Sufis, but also that akhi was simply an epithet used among many other titles in various communities, rather than a distinct social category.22 Oya Pancaro©lu puts forward that medieval Anatolian buildings such as caravanserais, Sufi and akhi lodges, and other socio-religious complexes all served the common purpose of providing food and lodging to passers-by in the period between the thirteenth and

Not for distribution or resale. For personal use only.

AKHI COMMUNITIES AND THEIR HOSPICES

fifteenth centuries.23 Seljuk caravanserais proliferated along the Konya–Aksaray–Kayseri trade route and assumed the hospitality function in the Anatolian countryside. The madrasa, or the school for Islamic sciences, on the other hand, served urban communities under the influence of Sufism.24 Beginning in the second half of the thirteenth century, multifunctional hospice buildings initiated by Sufi and akhi groups answered to the needs of devotion and hospitality.25 Parallel organisations met similar needs of Christian communities around futuwwa-like ethics in the thirteenth and fourteenth centuries as shown by Rachel Goshgarian’s study of Armenian confraternities in Anatolia.26 Though they have different scopes and focuses in their analyses, both Wolper and Pancaro©lu converge on the likeness or equivalence of hospice buildings utilised by akhi and Sufi communities. It is indeed exceedingly difficult, based on studies made so far, to distinguish between the monuments of either group considering that few monuments whose patrons were related to akhi communities or futuwwa circles have survived.27 Blessing provides an analysis of the fourteenth-century Akhi S¸ erafeddin Mausoleum in Ankara.28 This mausoleum is connected through its patron to the physically separate Akhi S¸ erafeddin (Aslanhane) Mosque. The mausoleum, the mosque and the genealogical document of the family of Akhi S¸ erafeddin reflect the significance of akhi communities in Ankara, especially in the late thirteenth and early fourteenth centuries, even though the endowment deed has not been preserved. Conversely, many other monuments that we know of through foundation records have been destroyed. Archaeological excavations of Sufi and akhi lodges may illuminate many unanswered queries on this matter.29 The first question that comes to mind is whether or not akhi lodges in particular had sturdier connections to artisanship, production or raw-material distribution. While maintaining the position on the similarity, even the sameness, of akhi and Sufi structures, this chapter follows the recurring theme of the probable proximity of akhi communities and futuwwa ideology to economic concerns. This point slightly diverges from Wolper’s assertion, conceiving akhi as merely an epithet utilised interchangeably with other contemporary titles used by the broad category of Sufis. I agree that it is precarious to assume the akhi title to have had one and the same meaning throughout centuries. Based on my research into the survival of akhis in the Ottoman period, I put forward that akhi was probably an epithet that was rather loosely used by various individuals in Ottoman times. It looks rather far-fetched to draw institutional ties between akhis and Ottoman guilds.30 I also believe that more can be done in terms of micro studies on akhis looking into individual cases of acquired or inherited titles, and whether or not individual akhis played distinguishable roles in particular moments and contexts.

Not for distribution or resale. For personal use only.

101

102

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Hospices Linking Rural and Urban Communities Passers-by and travellers hosted at akhi hospices probably created links between rural and urban communities. While dervish lodges in the rural areas provided public services, such as safety and a place to stay for the passers-by, akhi hospices similarly opened their doors to host them in urban centres. While some akhis might have been connected to the rural economy through raw materials used in their trade, some may have been primarily engaged in various professional activities in town centres, and would have had little reason to venture into the countryside. Rural caravanserais and hospices on ancient trade routes, also utilised by the Ottomans, hosted the caravan merchants. Yusuf O©uzo©lu’s work on the economic development of Òznik and the settlements and endowments on the Òznik–Gevye–Mudurnu–Bolu road support this premise. In this framework, O©uzo©lu demonstrates the relationship between Bursa and its rural hinterland, including primarily villages with names bearing the suffix ‘kızık’, a term that refers to Turcoman tribes who settled there in the early fourteenth century (‘Bursa Kızık Köyleri’).31 The business and secure delivery of caravan merchants to the urban market through the aforesaid trade routes might have been among the concerns of urban akhis engaged in craft production and the provisioning of the town. Sufi and Akhi Attitudes towards Commercial Life based on Narrative Sources Akhi brotherhoods are thought to have functioned in various respects in medieval Anatolia, mainly as mystic, chivalric and professional figures.32 The following parts of this chapter study narrative sources such as dervish biographies and futuwwa manuals, in an effort to discern Sufi and akhi attitudes towards commercial activity in Anatolian urban centres. This analysis is based on the not-so-certain assumption that prescriptive futuwwa manuals might have impacted akhi communities’ practices. Anecdotal information from hagiographies and biographies of members of the Sufi milieu offers insight on the contrast between the attitudes of Sufis and akhis to crafts and trades.33 Anecdotes that appear in biographies of Sufi masters sometimes mention craftsmen. As we will see in the following examples from such sources, Sufis’ attitudes towards commercial activities seem to differ from that of akhis as described by futuwwa manuals.34 While the first two anecdotes reflect convivial encounters between artisans and Sufi masters, further examples illustrate the problematic aspects of having a trade or craft for mystics. The first story involves a miracle regarding a fifteenth-century former slave, Ba∞çı Òbrahim of Bursa, who achieved success in

Not for distribution or resale. For personal use only.

AKHI COMMUNITIES AND THEIR HOSPICES

business as a freed man. Once a week, Sufi master Emir Sultan (d. 1429)35 frequented the eatery of Ba∞çı Òbrahim for sheep’s head (kelle). The latter, as a follower of the master, offered him a cooked head every day. One night, having placed some heads to boil in preparation for the next business day, he saw that all of the heads had turned into gold by a miracle of the master. Upon seeing this, Òbrahim set out to build one mosque in the name of the master, and one in his own name.36 In this story, the Sufi master rewards the artisan/business owner in return for a regular offer of food, which in turn leads to a pious endowment by the artisan. In another episode, a Bursa merchant called Akka∞zåde es-Seyyid Abdurrahman Çelebi reports that once he bought rice for trade and placed it in a caravanserai cellar. When he checked on it, he saw that the rice was infested with bugs, which distressed him. On his way out of the caravanserai, he saw Sufi master Emir Sultan sitting outside. The master looked at Abdurrahman Çelebi and said: ‘Send us rice (pilåv) and turn your face towards us.’ In response, Abdurrahman Çelebi immediately said: ‘Send a bag and take as much rice as you wish.’ When he went over to look at the bags of rice, he saw that there were no bugs left, which removed all his grief.37 Both of these anecdotes feature the master of a dervish order whom market men respected. It is noteworthy that the master in question in these episodes, though away from his habitat of the lodge, was confined to his expected vocation of showing miracles. In various other stories, however, we encounter Sufi masters who are ‘forced’ to assume the roles of apprenticeship or craftsmanship, which seems to be a source of concern to them. Üftåde (d.1580) was an influential master of the Celveti order who lived in Bursa a century later.38 In his youth, his father sent him to be an apprentice under a silk-maker (kazzåz), which put him through misery. In his words, being a silk-maker was not very difficult in comparison with other occupations, yet he was not fond of the job and worked half-heartedly. One day, Üftåde ran into his silk-maker master at the mosque, and perhaps because of his disdain for the job, he could not help but curse his master. Üftåde was to regret his curse, which showed its effect immediately. His master passed away on the very same night, followed within a week by Üftåde’s father who had forced him to work in silk production.39 In a similar anecdote, Üftåde met a Bayrami master whom he served from the beginning of his apprenticeship until the master’s death. This master was Hıdır Dede who used to be a shepherd but took the Sufi path after losing his feet to the cold. During his training in mysticism, Üftåde continued to make a living as a silk-maker since his father had passed away and he had to take care of his mother and two siblings. It is said that his mother aided him by spinning silk thread at home. When his brother died, his sister got married and his mother left to live with her, so Üftåde was able to devote

Not for distribution or resale. For personal use only.

103

104

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

himself fully to the service of his master. He served Hıdır Dede for eight years until the Dede died. Subsequently, he served as a prayer leader, performing the call to prayer and reading the Qur’an.40 In his lifetime, Üftade established an endowment for a religious complex including a hospice and a mosque using his personal resources.41 A more striking story concerns Üftåde’s disciple, Aziz Mahmud HüdåyÈ (b. 1543). During his training under Üftåde, the master asked HüdåyÈ to buy liver, offal and giblets from the market, to hang them on rods and to sell them on the busiest spot in town – a rather humiliating assignment. HüdåyÈ had held the kådÈship of Bursa prior to devoting himself to the service of the master, and from this had retained a measure of pride fuelling his carnal soul (nefs).42 Although the task of selling giblets in the streets was detestable, he set out to do it out of respect for his master. Upon seeing him, the public made fun of him, shouting out that the kådÈ had lost his mind. HüdåyÈ patiently endured the challenge and managed to retain full control of himself. The rods HüdåyÈ supposedly used to sell the giblets can be seen in Üftåde’s hospice in Bursa today.43 Such anecdotes found in dervish hagiographies and biographies show occasional involvement of masters and their disciples in artisanship. They also display negative attitudes towards these commercial activities, associating them with the burden of earning a living or, in the case of HüdåyÈ, a trial imposed by one’s master. Though Sufi accounts display a complex picture, as we are about to see, they already offer a point of contrast to those attitudes that appear in the futuwwa literature. The Marketplace and futuwwa in Relationship to Professions Nasıri, who wrote his futuwwa manual in 1290 in central Anatolia, states the main difference between futuwwa and Sufi mysticism to be the obligation of the follower of futuwwa to make a living by practising a trade or craft. Similarly, Yaha ibn Halil Çoban (Fatå al-Burgazi), who wrote a futuwwa treatise in Turkish around the mid-thirteenth to mid-fourteenth century in Aleppo, views the ideal follower of futuwwa as a generous and honest man who earns his own livelihood.44 Nasıri’s futuwwa manual describes the attitude of futuwwa towards professions as follows: ‘Always have an occupation. Receive the approval of the people of skill. The unemployed always engage in vanities. Lazy people cannot achieve anything.’45 Al-Burgazi’s position is in accord as he states: ‘Akhis must have a craft to keep them occupied.’46 By this layman’s approach, akhis earned their own living, kept their personal expenditure to a minimum, and spent it at their hospices to finance the hosting of passers-by and those who needed help. This approach found in the futuwwa literature is in concert with Ibn Battuta’s accounts on Anatolian akhis who performed their profession in the day, and gathered for their communal

Not for distribution or resale. For personal use only.

AKHI COMMUNITIES AND THEIR HOSPICES

dinner and rituals in the evening.47 According to Ibn Battuta, akhis’ hospitality functions, on which they spent their earnings, included offering their guests a meal, taking them to the bath, feeding them sweetmeats and fruits, and reciting the Qur’an, followed by singing and dancing.48 Further exploring the ‘prescriptive’ features of futuwwa manuals looking for parallel themes to the aforementioned links between akhi teachings and professions of the market, it is seen that futuwwa manuals from various localities and periods convey that each craft group in the path of futuwwa had a patron saint (pÈr). Patron saints were either real or fictitious characters whose stories appeared in earlier Islamic and Biblical sources.49 Numerous futuwwa manuscripts include lists of crafts and their patron saints, whose names were mentioned while opening shops in the market, in prayers and in the girdling ceremonies that survived well into Ottoman times.50 The girdling ceremony of entering or rising within futuwwa ranks involved various ways of tying a symbolic girdle (∞edd) to the pupil. The girdle symbolised sexual chastity, as shown by Rachel Goshgarian in Chapter 5 (this volume), as well as professional affiliation.51 Different ways of girdling were associated with various professions.52 Sheikh Seyyid Hüseyin’s fifteenth-century futuwwanåme includes different girdling shapes, each of which was appropriate for a different profession.53 Even the ‘inappropriate’ professions that were not readily admitted to futuwwa nevertheless required expertise: bathhouse workers, shavers, butchers and surgeons each had a corresponding girdle shape.54 Similarly, ceremonial descriptions relating to particular professions mentioned in the Turkish manual of Fütüvvetnåme-i KebÈr include barbers (selmeniye), whose initiation ceremony features a stone and razor and mentions the name of their patron saint Salmån al-FårisÈ. In the ceremony of bread bakers (habbåz), the apprentice puts on a work shirt. Tanners (debbå©) mention the name of Ahi Evran (patron saint of Anatolian tanners) and Zayd al-Hindi (patron saint of all tanners). The apprentice to bathhouse keepers (hammåm) receives a vest or waistcoat, a turban and short breeches, and is girdled with a shawl. The apprentice to butchers, whose patron saint is Cuvanmerd-i Kassab, receives a large knife and a knife sharpener. The apprentice to carpenters (patron saint Da’ud ibn al-Rahman) receives a hand saw and a hammer.55 A manual by Molla Husein Vaiz (Husein al-Ka∞ifi, d. 1504–5), titled Fütüvvetnåme-i SultånÈ and written in Persian, provides significant details on the futuwwa–craft relationships.56 Although this manuscript mainly targeted Persian-speaking audiences, Gölpınarlı highlights the fact that professions found in the Fütüvvetnåme-i SultånÈ were not foreign to the streets and markets of contemporaneous Ottoman towns. The sixth chapter of this manuscript contains a section devoted to the marketplace, which groups the people of

Not for distribution or resale. For personal use only.

105

106

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

the bazaar under the divisions of entertainers, water servers, fortune tellers and, finally, those who tell stories and tales.57 Artisans’ utensils and their use mentioned in this work of futuwwa occasionally relate to a more religious discourse: besides the labour involved in the task, the job of basket-carriers also had a symbolic meaning of ‘remembering the weight of that which is entrusted with the people for safekeeping by God’.58 Futuwwa and Hisba Perhaps an equally tangible parallel between fütüvvet and the marketplace is the connection with the hisba literature.59 Based on the Qur’anic injunction of ‘al amr bi’l-ma‘rËf wa’n-nahy ‘an-il-munkar’, hisba refers to the principle of promoting good deeds and forbidding wrong deeds by the Islamic community, and by a state-appointed official called a muhtasib (Turk. muhtesib) in the marketplace.60 An immediately recognisable parallel between the two literatures is the fundamental ‘knowledge’ (‘ilm) mentioned in Fütüvvetnåme-i KebÈr, which is the ability to distinguish between good and evil, between the permissible and the despicable. This ability to distinguish between ‘permissible and despicable’ directly echoes the Qur’anic injunction of ‘al amr bi’l-ma‘rËf wa’n-nahy ‘an-il-munkar’. This ability is passed on to the apprentice by a master craftsman, who is at the same time a spiritual leader with knowledge of futuwwa. The appropriate training takes place with the consent of the masters and the prominent leaders of the futuwwa path.61 Aside from its engagement with the general principle of distinguishing between permissible and despicable acts, Fütüvvetnåme-i KebÈr includes some anecdotes that underline similar rules to the principles of hisba: ‘If the exact price is not known (by the buyer), one should not sell the commodity for more than its worth, nor sell it below its value.’62 In a tradition mentioned in the manual, Prophet Muhammed, after discovering that a merchant had moistened the wheat he sold in the market, supposedly said: ‘Anyone who deceives me or my community does not belong to my people.’63 Some Qur’anic verses cited in the Fütüvvetnåme also support principles on fair trade and proper weights and measures in the market. Surah al-Hud, XI, 85 states: ‘O my people, [always] give full measure and weight, with equity, and do not deprive people of what is rightfully theirs, and do not act wickedly on earth by spreading corruption.’ Surah al-Rahman, LV, 7, 8, 9 states: ‘And the skies has He raised high, and has devised [for all things] a measure, so that you [too, O men] might never transgress the measure [of what is right] weigh, therefore, [your deeds] with equity, and cut not the measure short!’ Surah al-A‘raf VII, 85 states: ‘Give, therefore, full measure and weight [in all your dealings], and do not deprive people of what is rightfully theirs; and do not spread corruption on earth after it has

Not for distribution or resale. For personal use only.

AKHI COMMUNITIES AND THEIR HOSPICES

been so well ordered.’64 Destroying false coins is recorded in heaven with good deeds. It is equivalent to giving alms worth 100 akças or fasting for 100 days.65 Other references to ‘just’ commerce are found in the ceremony of admission to futuwwa. Three ceremonies are held for the initiation to the lowest rank, admission to the second rank and promotion to the rank of the one who is bound with a girdle (sahib-i fütüvvet).66 One step among these ceremonies requires the person that performs the ritual to lift commercial scales and to turn towards the apprentice saying: ‘You shall trade with equity and always give another his due (to which he is entitled). Never give short weight nor sell a defective product. Praise God for your earnings. Whenever you pick up the scales, remember the scales in the life hereafter (which will weigh all your deeds, the good and the bad ones) and turn away from forbidden things.’ Following these words, the senior person in charge of the ritual gives the scales to the apprentice, and the apprentice kisses the hands of his master and the other elderly masters of the congregation.67 In addition to the rules that relate to fair trade, just price, fair profit and giving the right measure, Fütüvvetnåme-i KebÈr also includes the rules to do with proper religious conduct in the market: craftsmen must perform the five daily prayers, and the artisan, merchant and fatå should pray every day upon entering the bazaar and opening their shops, which is also in conformity with the Islamic hisba tradition.68 Conclusion Late medieval Anatolian futuwwa teachings emphasise ‘modesty’ and place limits and responsibilities on their followers in areas from profit-making in the market to chastity and proper conduct of conversation and hospitality. Numerous akhi buildings that are known through endowment deeds have been destroyed and are lost today, possibly due to the very humility of their structures. Many akhi monuments can only be archeologically excavated, if at all, considering that the historical centres of many cities in Turkey are now heavily built up. Scholarship on art and architecture of the period suggests that akhi hospices occupied the same structures with dervishes. Consequently, some related questions, such as which particular services akhi hospices provided, remain unanswered. A diachronic scrutiny of foundation registers looking for akhi endowment deeds shows that akhi hospices were concentrated around the towns of Kütahya, Kastamonu, Amasya, Tokat and Ankara. These towns lay alongside significant trade routes, as well as the Byzantine frontier. Akhis’ network in this region might have constituted links along these trade routes between rural and urban stops. Modest numbers of akhi endowments were found in Bursa, the first capital of the Ottomans.

Not for distribution or resale. For personal use only.

107

108

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Similarly, according to cadastral surveys of initial land grants by early Ottoman rulers, the land grants to akhis were considerably less in number compared with those made to sheikhs and prayer leaders in this region. Based on observations of endowment registers and cadastral surveys, I suggest that akhis might have lost their influence in this area earlier, perhaps due to the processes of capital-building and centralisation that might have been more pronounced in Bursa. Though it is difficult to distinguish functions of akhi hospices, it is possible to underline certain akhi practices, particularly their links to artisanship and Islamic market regulations that might have distinguished akhi communities from Sufis. Dervish hagiographies and biographies show a complex picture of attitudes towards commercial activities. Members of Sufi circles certainly do not seem to have been pleased by their involvement in artisanship in case they were obliged to make a living in the marketplace. On the other hand, futuwwa manuals feature patron saints of artisans and girdling ceremonies in detail regarding numerous craft groups. There are notable parallels between futuwwa treatises and hisba rules that helped Islamic governments regulate urban markets. Ibn Battuta’s observations on Anatolian akhis are in compliance with such prescriptive literature. All these findings suggest a deeper relationship between akhi communities and commercial life compared with that of Sufis. Was akhi merely an epithet interchangeably used by coexisting communities within the blurred boundaries of medieval Anatolia? That is more likely to have been the case during the Ottoman period. As for what came before, the question deserves further consideration.

Notes 1. Some Western Orientalists have discussed futuwwa ethics as part of a linear and static Islamic ‘tradition’. See, e.g., Hammer-Purgstall, Rückert and Thylmann, Ghaselen des Dschelal-eddin Rumi; Thorning, Beiträge zur Kenntnis des islamischen Vereinswesens; Taeschner, Gülschehris Mesnevi. For a more detailed account of this categorisation, see Ocak ‘Türkiye’de Ahilik ara∞tırmalarına ele∞tirel bir bakı∞’, p. 129. In so doing they avoid a historicisation of futuwwa’s evolution in specific regard to the akhi brotherhoods of late medieval Anatolia. Broad-themed historical studies on medieval Anatolia mention the subject of akhis only in passing. Cemal Kafadar, while providing a broad survey and critique of the historiography on the birth of the Ottoman state, underlines the importance of akhis, but focuses on the pros and cons of the ‘gaza thesis’, which has not paid much attention to them. Kafadar, Between Two Worlds, pp. 34–5. While a number of other wideranging studies likewise touch upon the subject without discussing it at length, see, e.g., Cahen, La Turquie pré-ottomane; Vryonis, Decline of Medieval Hellenism. 2. Numerous works on akhis produced by Turkish scholars display significant shortcomings by excluding important primary sources,

Not for distribution or resale. For personal use only.

AKHI COMMUNITIES AND THEIR HOSPICES

avoiding discussions of the legacy of pre-Islamic or pre-Turkic Anatolia, and by emphasising the ‘Turkishness’ of the institution to the extent of rejecting its ties with the Arabic-language futuwwa literature. An example is Ça©atay, Bir Türk kurumu olan ahilik. Yet others postulate connections between futuwwa and akhis, without historicising or analysing futuwwa ideology. That said, some works – though subject to criticism – have met high scholarly standards. See Köprülü, Türk edebiyatında ilk mutasavvıflar; Köprülü, Bizans müesseselerinin Osmanlı müesseselerine tesiri; Köprülü, Osmanlı devleti’nin kurulu∞u; Köprülü, Türk edebiyatı tarihi; Ergin, Mecelle-i umûr-ı belediyye; Gölpınarlı, Mevlana’dan sonra Mevlevilik; Muallim Cevdet, Supplément à un chapitre de la relation de voyage d’Ibni Battuta; Bayram, Ahi Evren ve Ahi te∞kilatının kurulu∞u. 3. A case in point is the so-called Ahi S¸ eker Mausoleum of Ak∞ehir, documented by Òbrahim Hakkı Konyalı in his Ak∞ehir Tarihi. In 1945, at the time of Konyalı’s publication, the only remains of the mausoleum were four tombstones squeezed in the middle of houses in the Nidir village of Ak∞ehir in central Anatolia. 4. Works on the continuity of akhis’ principles and futuwwa ethics into the Ottoman period are equally problematic. As Sabri Ülgener’s Òktisadi çözülmenin ahlak ve zihniyet dünyası (The Moral and Mentality World of Economic Disintegration) puts forward a Weberian approach to pre-capitalist ‘medieval’ tendencies in the ‘East’, it displays serious shortcomings, such as a timeless and static conceptualisation of medieval Anatolian economy that fails to account for change. Other works, on social and economic history in the strictest sense, analyse Ottoman organisation of craftsmen and markets during the later phase of the incorporation of the akhi brotherhoods into the centralised Ottoman entity from the sixteenth century onwards. See Ònalcık, The Ottoman Empire: The Classical Age. Such historical accounts of medieval Anatolian town economy and its absorption into the Ottoman imperial system offer neither a direct account of the role of akhis, nor of the functions of their hospices. Studies on Ottoman economic mentality concentrate on the Anatolian economy, but avoid the topic of akhis. See Ònalcık, ‘Ottoman Economic Mind’; Kafadar, ‘When Coins Turned into Drops of Dew and Bankers Became Robbers of Shadows’. For a discussion of Turkish scholarship on akhis’ influence on Ottoman organisation of artisans, see Kal‘a, ‘Osmanlı esnafı ve sanayisi üzerine yapılan çalı∞malarla ilgili genel bir de©erlendirme’, pp. 245–66. 5. Despite all these ambiguities, Mikâil Bayram, a historian of Seljuk history, is quite certain of the existence of the organisation (Bayram, Ahi Evren ve Ahi te∞kilatının Kurulu∞u, and Fatma Bacı ve Baciyan-ı Rum). Rachel Goshgarian underlines a similar opacity regarding women in Chapter 5, this volume, ‘Social Graces and Urban Spaces: Brotherhood and the Ambiguities of Masculinity and Religious Practice in Late Medieval Anatolia’, suggesting that Armenian futuwwa tradition in Anatolia occasionally presents women as ‘tangential and associate members’, p. 126. 6. See Goshgarian, ‘Opening and Closing’; Blessing, Rebuilding Anatolia. 7. Pancaro©lu, ‘Devotion, Hospitality, and Architecture in Medieval Anatolia’; Ça©aptay, ‘Prousa/Bursa, a City within the City’. 8. On the building of Bursa and its neighbourhoods around dynastic monuments throughout the early Ottoman period, see Pancaro©lu,

Not for distribution or resale. For personal use only.

109

110

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

‘Architecture, Landscape, and Patronage in Bursa’; Ça©aptay, ‘Prousa/ Bursa, a City within the City’, pp. 45–69; Ça©aptay, ‘Frontierscape’, pp. 156–91. 9. Blessing, Rebuilding Anatolia, pp. 165–203. 10. Blessing, Rebuilding Anatolia, p. 165. 11. Blessing, Rebuilding Anatolia, p. 171. 12. The meaning of frontier, economic or otherwise, and possible roles that akhis could have played within the frontier region is as yet a speculative idea that must be further explored. On the historical and architectural meaning of this frontier, including likely parallels between Anatolian and Iberian peninsulas of medieval times, see Kafadar, Between Two Worlds. See also Ça©aptay, ‘Frontierscape’. 13. Barkan and Meriçli, Hüdavendigâr livası tahrir defterleri. 14. See Barkan, ‘Osmanlı imparatorlu©unda bir iskân ve kolonizasyon’. 15. Many records either do not mention the name of the sultan or the initial temlÈk was granted to an Ottoman bey and then transferred to an akhi or a fakÈh. Some records also involve grants that were shared by akhis and sheikhs, or fakÈhs. Another obstacle is the question of which sultan is meant by the record. The titles of hüdåvendigår, hündkår, gåzÈ, bey and sultan, all seem to have been used by the early Ottoman rulers, although Sultan Murad I (r. 1362–89) is thought to have used the hüdåvendigår title for the first time. 16. Kafadar, Between Two Worlds, p. 181, fn. 136. 17. HatÈbs and huffåz (reciters of the Qur’an), who often held the epithet of mevlånå hatÈb, seem to have functioned in a similar fashion to fakÈhs. 18. For a survey of Turkish scholarship on Islamisation, Turkification and settlements in medieval Anatolia, and criticism – albeit with some problems – of various interpretations of the role of dervishes and their structures in these processes, see Wolper, Cities and Saints, pp. 5–6. 19. Barkan, ‘Osmanlı imparatorlu©unda bir iskân ve kolonizasyon’. 20. Barkan and Meriçli, ‘Osmanlı Òmparatorlu©unda nüfus ve vergi sayımları’, introductory essay to the Hüdavendigâr livası tahrir defterleri, p. 134. 21. On social functions of Sufi lodges, see also Faroqhi, ‘The Tekke of Hacı Bekta∞’; Faroqhi, ‘Seyyid Gazi Revisited’. 22. Wolper, Cities and Saints, pp. 77–8. 23. Pancaro©lu, ‘Devotion, Hospitality, and Architecture’, p. 49. 24. Pancaro©lu, ‘Devotion, Hospitality, and Architecture’, p. 54. 25. Pancaro©lu, ‘Devotion, Hospitality, and Architecture’, p. 62. 26. Goshgarian, ‘Beyond the Social and the Spiritual’; Goshgarian, Chapter 5, this volume. 27. One is the thirteenth-century Ahi Evran Lodge in Kır∞ehir. This monument is of significance owing to the revered position of Akhi Evran, the patron saint of tanners within the akhi tradition. An architectural analysis of this monument is not, to my knowledge, available. Foundation registers do house some related documents to this building. See S¸ ükrü Akkaya (trans.), ‘Kır∞ehir’de Ahi Evran Zaviyesinin mütevellisine ait 1238/1822–23 tarihli bir berat’. On another lodge, see Yediyıldız, ‘Niksarlı Ahi Pehlivan’in Dårü’s-Sulehåsı’. 28. Blessing, Rebuilding Anatolia, pp. 192–7. 29. However, there are specific difficulties brought upon potential archeological studies by modern Turkey’s urbanisation.

Not for distribution or resale. For personal use only.

AKHI COMMUNITIES AND THEIR HOSPICES

30. Perhaps the continuous respect paid to Akhi Evran, the master saint of tanners in the Ottoman era, has created the idea of akhis’ prominence over craft organisations of the later period, which has resonated through historiography on Anatolian guilds. See Selçuk, ‘Tracing Esnåf in Early Fifteenth-century Bursa Court Records’, pp. 62–5. 31. O©uzo©lu, ‘Orhan Bey döneminde (1324–1362) Bursa ∞ehrinde görülen de©i∞iklikler’; O©uzo©lu, ‘Kırsal ya∞am ve kentle ili∞kisi’; O©uzo©lu, ‘Tarihsel süreç içinde Òznik ve yöresinin iktisadi geli∞imi’; O©uzo©lu,‘Bursa Kızık köylerinin tarihsel kökeni’. I thank the author for providing me with his articles. 32. Goshgarian (Chapter 5, this volume) underlines yet another characteristic of brotherhoods organised around futuwwa ideals in Anatolia, namely ‘manliness’, by illustrating attitudes of Christian brotherhoods to the subject, see p. 126. 33. On various narrative sources available to the student of medieval Anatolia and ways to interpret them, see Trépanier, Foodways, appendix on the sources, pp. 131–42. 34. Two fütüvvet manuals are explored in relation to these questions: the Fütüvvetnåme of Sheikh Hüseyin ibn Seyyid Gaybï, written in southeastern Anatolia c.1450–82, and published by Gölpınarlı; and Miftåh el-DakåÈk fÈ Beyån el-Fütüvve ve’l-Hakåyik, also known as Fütüvvetnåme-i KebÈr, by Seyyid Muhammed ibn al-Seyyid al el-DÈn elHuseyni el-Razavi. For an analysis of the fütüvvetnåme, see Breebaart, ‘The Development and Structure of the Turkish FutËwah Guilds’; Breebaart, ‘Miscellanea: The Fütüvvet-nåme-i KebÈr’, p. 207. 35. An important figure among the Sufi milieu of early Ottoman Bursa, S¸ emseddin Mehmed Ali el-Hüseyin el-Buhari (Emir Sultan), born in Bukhara, was a pupil of prominent kådÈ, jurisconsult and madrasa teacher Molla Fenari of Bursa. He married the daughter of Sultan Bayezid I (r. 1389–1402) and lived as one of the most revered Sufi figures in Bursa. 36. Baldırzåde SelÈsÈ S¸ eyh Mehmed, Ravzat’il-Evliya, Bursa Velayetnameleri, vol. I, p. 87. 37. Baldırzåde SelÈsÈ S¸ eyh Mehmed, Ravzat’il-Evliya, p.109. 38. Beldiceanu-Steinherr, ‘Scheich Üftade’. 39. Beli©, Güldeste-i Riyâz-ı Òrfân ve Vefeyât-ı Dâni∞verân-ı Nâdiredân, p. 108. 40. Menåkıb-ı PÈr, 150 vd., cited in Kara, Bursa’da tarikatlar ve tekkeler, p. 297; Òsmail Hakkı BursevÈ, Silsile-i Tarik-i CelvetÈ, p. 79, cited in Kara, Bursa’da tarikatlar ve tekkeler, p. 297. 41. Erken, Be∞ba∞ and Denizli, Türkiye’de vakıf abideler ve eski eserler, p. 395. 42. Beldiceanu-Steinherr, ‘Scheich Üftade’, p. 131. 43. Vassaf, Sefine-i Evliya, vol. II, p. 394; Yılmaz, AzÈz Mahmud HüdåyÈ, p. 77. See also Kara, Bursa’da Tarikatlar ve Tekkeler, p. 318. 44. A more detailed description of futuwwa ethics based on these principles involves closing three features and opening three traits in one’s character, soul or actions: opening the face (friendliness), the door (hospitality) and the tablecloth (generosity); and closing the eyes (not to see the forbidden), the tongue (not to deceive anybody or speak ill of others), and the belt of his trousers (∞alvar) for chastity. 45. Letåif in NåsırÈ Ms. 98 b. [775]: Gölpınarlı, ‘Òslam ve Türk illerinde fütüvvet te∞kilatı ve kaynaklar’, p. 348.

Not for distribution or resale. For personal use only.

111

112

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

46. Süleymaniye Kütüphanesi, Aya Sofya MS 2049, 30a. Gölpınarlı, ‘Burgazi ve fütüvvet-namesi’, p. 125. 47. Ibn Battuta, Travels in Asia and Africa, pp. 413–68. 48. Ibn Battuta, Travels in Asia and Africa, pp. 426–7. For an analysis of the devotional and hospitality function of akhis and their hospices, see Pancaro©lu, ‘Devotion, Hospitality, and Architecture’, pp. 60–72. 49. Gölpınarlı, ‘Islam ve Türk illerinde fütüvvet te∞kilatı ve kaynaklar, section VII: Fütüvvet ve Esnaf Te∞kilatı’, p. 91. 50. Not only craft groups, but also professions or vocations such as poets (∞u‘arå) were included in the futuwwa path. Gölpınarlı, ‘Òslam ve Türk illerinde fütüvvet te∞kilatı ve kaynaklar, section VII: Fütüvvet ve Esnaf Te∞kilatı’, p. 92. 51. Goshgarian, Chapter 5, this volume, pp. 114–31. 52. Fütüvvetnåme-i KebÈr tells the history of the ∞edd (girdling) ceremony, which reveals traditional sources for the patron saints of many professions: first to be girdled by Archangel Cebrail (Gabriel) was Adem (Adam). Flour and honey came down from heaven and Adem saved some for Havva (Eve) as she was absent from the ceremony. From then on, it was a custom to send helva to those who were absent from the ceremony. (Helva was a sweet prepared from flour or semolina and honey that was used on religious occasions such as celebrations or funerals. For an analysis of helva in medieval Anatolian everyday life, see Trépanier, Foodways, pp, 116–17.) The traditions of girdling continues from Adam, on to his son Seth, then Noah, Abraham, Ismail, the nephew and the son-in-law of Ali ibn Abi Talib, and Prophet Muhammed’s companions, including the patron saint of barbers and all corporations (Salman alFarisi), messenger and footman of the prophet, the guild of footmen and couriers (‘Amr b. Umayyah al-Damri), the master of those who perform the call to prayer Bilal al-Haba∞i), etc. Ibid., pp. 152–8. Pre-Islamic patrons of many crafts originate from the life stories of prophets (kısås el-enbiyå), and are recognised and adopted by the futuwwa tradition: Adem (farmer), Suleiman (basket maker), Shith (weaver), Ibrahim (the first mason who built the Kaba), Ismail (hunter), Isa (carpenter), Nuh (built the arc), Da’ud (weaponsmith). According to this tradition, following the death of the Prophet Muhammed, four initiators of futuwwa were sent to their countries to spread it further: Salman al-Farisi to Iran, S¸ uhayb al-Rumi to Greece, Dhu an-Nun al-Mısri to Egypt, Sayfi al-Yamani to Yemen. Ibid., p. 159. 53. Different shapes: elif, kavs, mihråb, låmelif, süleymånÈ, yËsufÈ and håfÈ. Gölpınarlı, ‘S¸ eyh Seyyid Hüseyin b. GaybÈ Fütüvvetnamesi’, p. 19. 54. MS 88. 117. a. Gölpınarlı ‘S¸ eyh Seyyid Hüseyin b. GaybÈ Fütüvvetnamesi’, p. 120. 55. See Breebaart ‘The Development and Structure of the Turkish FutËwah Guilds’, pp. 182–6. 56. Molla Husein Våiz, Fütüvvet-Nåme-i SultånÈ, British Museum. A dd. No. 22705, cited in Gölpınarlı, ‘Fütüvvet-Nåme-i SultånÈ ve fütüvvet hakkında bazı notlar’, p. 128. 57. Gölpınarlı, ‘Fütüvvet-Nåme-i SultånÈ ve fütüvvet hakkında bazı notlar’. 58. Gölpınarlı, ‘Fütüvvet-Nåme-i SultånÈ ve fütüvvet hakkında bazı notlar’, p. 143. 59. For this discussion I base my analysis of the fütüvvetnåme of Sheikh Hüseyin ibn Sheikh Seyyid Gaybi on the oldest copy of the manual (c. fifteenth or sixteenth century) published by Gölpınarlı, S¸eyh Seyyid

Not for distribution or resale. For personal use only.

AKHI COMMUNITIES AND THEIR HOSPICES

Gaybî O©lu S¸ eyh Seyyid Hüseyn’in Fütüvvet-Nâmesi’, pp. 27–126. Gölpınarlı’s analysis shows that the main sources of Seyyid Hüseyin’s book were Nakka∞ Ahmed’s book in Arabic, Nasıri’s work in Persian and the manual written by Burgazi in Turkish. The latter, also known as Fütüvvetnåme-i KebÈr, is analysed based on Breebaart’s dissertation, ‘The Development and Structure of the Turkish FutËwah Guilds’. I also refer to ‘Miscellanea: The Fütüvvet-nåme-i KebÈr. A Manual on Turkish Guilds’, by the same author. 60. The existing literature, both in the form of primary and secondary material, on the principle of hisba, as well as its implementation in markets of Islamic communities, is vast and among many others includes Cook, Commanding Right and Forbidding Wrong in Islamic Thought, which focuses on the principle of hisba; Stilt, Islamic Law in Action, which covers the implementation of this principle in Mamluk times; and Kazıcı, Osmanlılarda ihtisab müessesesi, which approaches the subject from the Ottoman perspective. 61. Breebaart, ‘The Development and Structure of Turkish FutËwwa Guilds’, p. 172. 62. Breebaart likens this principle to justum precium (just price) of medieval European guilds, ‘The Development and Structure of Turkish FutËwwa Guilds’, p. 168. 63. Breebaart, ‘The Development and Structure of Turkish FutËwwa Guilds’. 64. The Message of the Qur’ån, trans. Muhammad Asad. 65. Breebaart, ‘The Development and Structure of the Turkish FutËwwa Guilds’, p. 170. 66. Breebaart, ‘The Development and Structure of the Turkish FutËwwa Guilds’, p. 194. 67. Breebaart, ‘The Development and Structure of the Turkish FutËwwa Guilds’, p. 182. 68. Breebaart, ‘The Development and Structure of the Turkish FutËwwa Guilds’, p. 172.

Not for distribution or resale. For personal use only.

113

CHAPTER FIVE

Social Graces and Urban Spaces: Brotherhood and the Ambiguities of Masculinity and Religious Practice in Late Medieval Anatolia Rachel Goshgarian The history of thirteenth-century Anatolia can probably best be visualized via contemplation of its robust construction projects. The caravanserais, madrasas, churches, mosques and dervish lodges that were built under Armenian, Byzantine, Georgian, Mongol and Seljuk rule altered cityscapes and landscapes from Ani to Konya, and from Trabzon to Antalya. Hence, these construction projects altered physical space and the ways in which people on the ground both conceived of and experienced the places in which they lived. This is to say, that the visual language constructed in the thirteenth and fourteenth centuries – particularly in cities, where construction projects were very enthusiastic – was understood according to an individual’s life experience. And there was a diversity of life experience in late medieval Anatolian cities. Trying to understand the intersectionalities and liminalities of these structures can help us to better understand the ways in which social spaces were constructed and experienced in medieval Anatolia, and can illustrate how late medieval Anatolians lived in landscapes that were home to places (both physical and social) profoundly marked by ambiguities and inconsistencies. While each of the structures present in medieval Anatolian cities is most commonly associated with a particular type of activity, scholarship has begun to uncover the myriad ways in which these medieval structures were used by a range of communities or social groups in Anatolia.1 At the same time, in a century marked by regional political competition, urban landscapes were increasingly used as stages where various groups engaged in competitive practices that sought to legitimise their approaches to organisation and power. In the fourteenth-century hagiography of Mawlana Jalal al-Din Rumi (d. 1273) and his family, Shams al-Din Aflaki (d. 1360) describes an interaction between Seljuk Sultan Ala al-Din Kayqubad (d. 1237) and Baha ad-Din Walad (d. 1231) upon the latter’s initial arrival in the city of Konya, the place he would eventually call home.

Not for distribution or resale. For personal use only.

social graces and urban spaces

Sultan Ala al-Din offered his hospitality by inviting the religious scholar to stay in his palace. Baha ad-Din (the father of Mawlana) refused, saying: A madrasa is appropriate for imams, a khanaqah for shaykhs, a palace for commanders, a caravanserai for merchants, lodges  (zaviya) for the rogues (runud) and the mastaba [bench] for foreigners.2 Of course, whether this kind of structural separation was really explained by Baha al-Din or was composed over one hundred years after his death by his hagiographer, we cannot know. Still, this passage suggests that an attempt to assign certain groups of people to specific structures was seen as a necessity by at least one late m ­ edieval Anatolian author. Ultimately, this passage tells us that certain individuals (in this case, either the Seljuk sultan or Aflaki himself) were invested in assigning spaces to certain groups and to regulating the flow of individuals within the urban landscape. Still, at other points in this very same hagiography (composed, to the best of our knowledge, by one author) we see various different members of the groups outlined above engaged in activities in spaces that are ‘not their own’, according to this particular passage. And while various kinds of text composed in Anatolia in the thirteenth and fourteenth ­centuries – including poems, chronicles, inscriptions, hagiographies and manuals – have a tendency to prescribe the usage of certain buildings as specific to certain groups of people, the same kinds of text describe these structures as being used by a wide range of social groups. The kind of contradictions present in some thirteenth- and ­fourteenth-century texts tell us a lot about the ways in which certain social groups worked, both in the context of the places they occupied in the cities and beyond them. In fact, much of what we understand as contradictory or ambiguous seems to highlight the complexities of the lived landscapes of the time as a succession of liminal physical and social spaces. These ambiguities (or contradictions) are not uniquely seen in texts with regard to the usages of structures, but also with regard to conceptions and definitions of masculinity as described (and prescribed) in contemporary texts. This chapter suggests that the ‘liminality’ or ‘in-betweenness’ that existed in late medieval Anatolian cities allowed individuals the freedom to embrace ambiguities, both with regard to religious and social practices and, also, in some cases, with regard to sexual practices. Late Medieval Anatolia: Political Complexities and Creative Social Responses During the thirteenth and fourteenth centuries, Anatolia was home to a series of polities that governed over lands having changing

Not for distribution or resale. For personal use only.

115

116

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

borders for varying lengths of time. Perhaps because of the region’s consistent geopolitical fluctuations and the presence of so many competing polities, Anatolia experienced an increased urbanity. And cities maintained a certain degree of internal stability, even if their overlords either changed or were challenged relatively frequently. While some scholars paint a picture of Anatolia that was in a constant state of war,3 the time period has also been construed as one during which patterns of political, religious and social life were altered to such an extent that a kind of rupture with the recent past was created, allowing for the creation of new institutions and practices.4 The late thirteenth-century Seljuk chancellor and chronicler Ibn Bibi, in describing the development of the city of Konya under the Seljuk sultan Ala al-Din Kayqubad (r. 1220–37), composed the following poem: Men from all lands hasten To make in that city of happiness a homeland. Not a city, you were a complete world; You were a deep sea and the city of renown.5 That the Seljuk chronicler Ibn Bibi describes the capital city of Konya as a ‘world unto itself’ suggests a medieval Anatolian reverence for cities and their development, most likely as a sign of stability, and intellectual, social and tangible wealth. The movement towards urban revival and development is generally linked with Seljuk expansion, as initiated under Sultan Ala al-Din Kayqubad.6 However, throughout the thirteenth century and even into the early fourteenth century, great construction projects were undertaken in many cities in Anatolia –both before and after the Mongol invasions in 1243 – and the result was the creation of a vivacious urban life.7 Because futuwwa associations were intimately linked with cities, their activities as performers in the urban ‘worlds’ of late medieval Anatolia shed light upon how and why futuwwa communities were formed and why there seems to have been an invigorated necessity for them in Anatolia during this time period. At the same time, the interest in texts that define and prescribe ‘masculinity’ in cities is curious. In fact, one could posit that this underlying insecurity of the time encouraged an evaluation of identities and ‘normativities’, including those related to gender expectations. Considering urban units as an alternative basis of historical analysis offers a means for comparison that allows us to see continuity in the story of the competing hierarchies of late medieval Anatolia in a more focused context. The latter is possible because the type of actors engaged in city life were generally the same, even if each city may have had a different history of political governance. These actors include akhis (members of futuwwa confraternities),

Not for distribution or resale. For personal use only.

social graces and urban spaces

dervishes, priests and bishops (Armenian, Greek, Syriac), imams, kådÈs, anti-nomians (whether Christian or Muslim) and nobles (whether Arabic-, Armenian-, Georgian-, Greek-, Mongol-, Persian-, Syriac- or Turkish-speaking). And not only were the actors in cities across Anatolia similar, but so were many of their concerns, including the establishment of exclusive, homosocial, hierarchical associations. In fact, considering late medieval sources, it becomes clear that identities were often formed around the city, rather than focused on language or faith. Thus, a shared habitus superseded identifications by shared religion or language.8 A poem by Sultan Walad (d. 1313), the son of Mawlana, illustrates the author’s reverence for the Seljuk capital: O Konya, full of cavalry soldiers You are the throne of the territories of Rum Every city is grand like an Emir You are the head of the cities like a Shah Every citadel is a luminous star You are the head of the stars like the Moon Since his highness our Shah selected you You are the Mecca and the Ka‘ba Divine.9 In the late medieval Anatolian Armenian context, as well, it would appear that urban identities were important, even though on top of urban allegiances towards local, non-Armenian rulers, there also existed relatively loose allegiances of place and power to Armenian nobles and mecatuns (nouveau riches) as well as to two Armenian hierarchies: that of the Armenian Church and that of the Armenian kingdom of Cilicia (1178–1375). The late thirteenth-century Armenian priest-poet Yovhann≠s of Erzincan10 – who composed two Armenian-language treatises on futuwwa – compares a man’s body to the structure of a city: As you see the body of a man, it is like a city with one fortification and five gates surrounding it: one gate is the eye, another is the ears, one is the nose, another is the mouth, and one is the hand and foot which are tangibles. All good and evil enters and exits through these gates.11 That Yovhann≠s constructed such an intimate relationship between the city and a man’s body suggests that this author knew that his audience would relate to such an image, and to the anthropomorphising comparison he makes. In what was perhaps a reference to the numerous attacks by various powers on the cities in which Yovhann≠s had lived, he compares Satan to invaders attempting to steal booty from the city:

Not for distribution or resale. For personal use only.

117

118

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

The soul is the inhabitant of this city. Man’s good works are the treasure of this city, its source and its jewel. The mind is the king and the watchman of the jewel. Satan and the devils are evilminded and want to take the city and the good soul, which is its treasure, and steal it.12 His message is clear: if one is meant to protect the booty in one’s city from invaders, then one must be sure to protect one’s soul from Satan. The centrality of the city itself was a fundamental part of a late medieval Anatolian conception of reality, regardless of the language one spoke or the faith one practised. It was, in fact, within the framework of the city that urban dwellers understood their immediate reality – and not necessarily within the context of a particular polity or religious hierarchy. It is within this context that urban-based, homosocial organisations grew in medieval Anatolian cities that reinforced a localised, male dominance of public life. These included (but were not limited) to: Sufi organisations, futuwwa-based associations and various Christian monastic orders. Many authors (both medieval and contemporary) have considered futuwwa to be a branch of Sufism, or a less-stringent version of the mystical path.13 In fact, in her essay · (Chapter 4, above), Iklil Selçuk attempts to tease out the meaning and functions of the futuwwa-based brotherhoods (whose members were known as akhis) in the context of Sufi brotherhoods.14 Many scholars have suggested that the definitive difference between Sufi engagement and that of the akhis is the notion that akhis prioritised work and an engagement with craftsmanship over mystical devotion, while Sufis prioritised ritual practice and piety over work. Still, the clearest difference between Sufism and futuwwa was that performing manliness was at the basis of the conceptual framework attached to futuwwa. Futuwwa: A History of Contradictions? Futuwwa essentially crystallised as a code prescribing ‘manliness’ in an Islamic context in the eighth century, but it was not until the eleventh century that the first code on futuwwa was composed in Arabic. The author of that code, Abu al-Qasim al-Qushayri (d. 1074), in a chapter dedicated to futuwwa in his Risålat al-QushayrÈ, explains the code of futuwwa as a set of morals and practices that a student could acquire through study, emulation and practice.15 Thus, from the beginning, the code of futuwwa stressed the relationship between the novice and his master as an object of emulation. Similar codes were composed in the medieval Middle East afterwards as associations claiming adherence to futuwwa fanned out throughout Islamdom.

Not for distribution or resale. For personal use only.

social graces and urban spaces

In part because of the popularity of these codes and in part due to the general weakening of the caliphate, the Abbasid caliph alNasir li-Din Allah (d. 1225) both joined the futuwwa and eventually reformed it. In the aftermath of his reform, he sent the then elderly intellectual Shihab al-Din Abu Hafs Umar al-Suhrawardi (d. 1234) to encourage his caliphate-centred reform of the institution in Anatolia. And it was in Anatolia that futuwwa-based brotherhoods flourished in the thirteenth and fourteenth centuries, and seem to have played more of a productive and mainstream role in city life than ever before. As Selçuk describes, these organisations played an important role in central and eastern Anatolian cities in the late medieval period, but less of a role in Ottoman-dominated areas further west.16 Codes on futuwwa composed in late medieval Anatolia describe the qualities and activities associated with members of the urban, homosocial confraternities who were meant to abide by them. The characteristics that are most commonly reproduced in these codes include faithfulness, generosity, work ethic and bravery. The codes also prescribe the behaviours and activities of everyday life down to how one should walk, eat, drink and laugh. Even from its very inception, however, futuwwa has consistently been expressed as a series of contradictions. Futuwwa codes have been composed with contradictions, to some extent, as a means of offering adherents a certain amount of freedom in their activities related to the subversion of local (and regional) power structures and as a kind of motor for alternative forms of urban self-governance. This is, in part, the reason these codes became prevalent in Anatolia: because their contradictions consistently allowed adherents the space to establish alternative power structures and internal relations driven by different kinds of religious and communal practice. These ambiguities were not new in the Anatolian codes of futuwwa. For example, while these codes have often been associated with a notion of manliness, the term futuwwa means ‘male youthfulness’, even though a requirement for entrance into an organisation that uses a code of futuwwa is physical maturity (Arab. bulËgh). At the same time, while the Muslim authors of these codes uniformly consider Islam to be the foundation of futuwwa, the consistently cited first example of futuwwa­-ness is Hatim al-Ta‘i (d. 578), a well-known and much loved pre-Islamic, Arabic poet. And while in the thirteenth century the caliph al-Nasir li-Din Allah (d. 1225) made himself the centralising pole of all futuwwa organisations, thus transforming the code and those associations that adhered to it into more ‘orthodox’ associations linked directly to the caliphate, veneration for Ali as the fatå (example of futuwwa) par excellence suggests a real attachment to some ideals that would have been more closely linked to Shiism.17 As mentioned in Chapter 4, above, by · Iklil Selçuk, although manliness is central to the code itself, it seems

Not for distribution or resale. For personal use only.

119

120

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

probable that in medieval Anatolia futuwwa was occasionally practised by women, as some women did use the title akhi and the famed Ottoman historian Å∞ıkpa∞azåde suggested there existed a female counterpart to the akhi organisations known as the båcıyån-ı rËm.18 In the late medieval Anatolian context, the fact that Armenian Christians in Anatolia wrote their own futuwwa (translated into Armenian as ‘manktut‘iwn’, or youthfulness) treatises that were rooted in Christianity (rather than Islam) forces us to interrogate another level in the series of ambiguities and contradictions in the code related to these associations. Futuwwa Treatises in Anatolia: the Freedom of Being ‘In-Between’ Many scholars have cautioned against using the constitutions produced by members of futuwwa brotherhoods, suggesting that they are inherently flawed because of their prescriptive nature. Even a brief engagement with the futuwwa constitutions will leave the reader with the distinct impression that she or he is reading a series of descriptions of the behaviour of an ideal member of a brotherhood. A historian could, therefore, posit that no source is either inherently descriptive or prescriptive.19 Whatever their limitations may be, treatises on futuwwa offer us a unique opportunity to engage in a diachronic analysis of how a genre of literature specifically associated with the kind of homosocial associations under consideration here defined manliness and the spaces associated with the performance of manliness; the differences between adherents of futuwwa and Sufism; and, even, the spaces these two kinds of homosocial group occupied. At the same time, a comparison between the trends observed in the contemporary texts that contribute to our understanding of the activities of members of futuwwa associations can offer another layer of analysis and consideration of the ways in which these individuals both imagine and performed masculinity and brotherhood. This, in turn, sheds light on how alliances and systems of allegiance were established in opposition to religious hierarchies as well as to regional overlords. According to most prescriptive texts, both medieval Anatolian Sufi and akhi groups specifically excluded women from participation.20 As Julie Meisami has suggested, the necessary abstinence from women (and the absence of women) that is articulated in the codes of certain homosocial associations infers a preference for the company of men and a reinforcement of the creation of a male space of dominance.21 In effect, most communal organisations in late medieval Anatolia (and elsewhere) were primarily homosocial in nature. But as noted above, the fascinating characteristic of futuwwa associations is not only that they excluded women and, therefore, tell us something about the way in which gender was experienced in Anatolian cities during this time period, but that being male or

Not for distribution or resale. For personal use only.

social graces and urban spaces

performing masculinity was central to their mission and role in society. Whereas recent scholarship has suggested that medieval Sufism rejected the idea of any superiority other than spiritual superiority, and thus voiced a criticism of ‘prevalent notions of intrinsic male superiority’,22 futuwwa explicitly prefers masculinity. Still, not all males were welcome participants in futuwwa associations. Each of the Islamic Anatolian futuwwa treatises that was composed in an Islamic context names non-believers (i.e., nonMuslims) as the first category of unacceptable association members, thus highlighting the ‘faith-based focus’ of the association. Many scholars have suggested, in fact, that futuwwa confraternities, like organisations of dervishes, represented a means by which nonMuslims were encouraged to convert to Islam in Anatolia during this time period.23 And, yet, in his fourteenth-century hagiography, Shams al-Din Aflaki makes references to Armenian akhis24 in both Konya and Sivas, while an inscription near Konya suggests that there was at least one Greek akhi living in the city.25 Vryonis intimates that this evidence suggests that Christians were also members of futuwwa-based associations, but does not determine whether they participated in Muslim versions thereof or created their own.26 Still, given the fact that in these particular examples, Armenian and Greek participants maintained their ‘Armenian’ or ‘Greek’ identity while still being called akhi, one might suppose that not all akhis in late medieval Anatolia were actually Muslim. We could even imagine that the complex space occupied by these kind of brotherhoods represented something so contradictory and ambiguous that they were able to occupy different kinds of faith spaces such that adherence to the brotherhood became more of an indication of station in city spaces than did one’s religious or linguistic practice. Famed Moroccan travel writer Ibn Battuta (d. 1368–9), who visited the region in the early fourteenth century, claimed that akhis were to be found ‘in all the lands inhabited by the Turkmens in Anatolia in every district, town and village’.27 Thanks to the extensive descriptions in his travelogue and to information gleaned from various other local sources (namely, pious endowment deeds, futuwwa treatises, chronicles and hagiographies), we know of the presence of akhis in Amasya, Ankara, Antalya, Bursa, Erzincan, Eski∞ehir, Harput, Kir∞ehir, Konya, Kütahya, Sivas and Tokat, cities that were inhabited by both Muslims and non-Muslims.28 At the same time, although many scholars have suggested that akhis and dervish associations were responsible for the conversion of local Christians, extant Armenian sources do not lament the actions of dervishes or akhis as encouraging conversion, although they consistently blame the Mongols (whom they called the ‘Nation of Archers’) for violent, forced mass conversions.29 At the same time, however, Armenian cultural trends of the late medieval period trend towards reduplicating or at least reflecting the

Not for distribution or resale. For personal use only.

121

122

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

poetic imagery and styles of Turkish- and Persian-speaking Muslims, largely propagated by homosocial groups located in cities where they would have had opportunities to engage with local Armenians. Thus, both the physical and social spaces occupied by these communal groups became sites of specific kinds of interaction between convergence, opposition and reflection. This kind of ‘in-betweenness’ was present in various aspects of social life in late medieval Anatolian cities, reinforcing a landscape that was relatively comfortable with ambiguities. Sexuality in Futuwwa Each futuwwa treatise composed in Anatolia – whether written in Arabic, Armenian, Persian or Turkish – aims to control the behaviours of its adherents by prescribing their appearance, their relationship with their master-teachers, their sexual activity (or lack thereof), their affiliation with other hierarchical institutions, their faith, their work, and the most basic and detailed aspects of their daily life centred around the dining table. Each constitution expresses concern with ‘pretenders’ (i.e., those individuals who seemed like members of futuwwa organisations but were not), while also explicitly describing those people who were to be excluded from the organisations. And all futuwwa treatises attempt to define gender roles while at the same time betraying (either explicitly or implicitly) a tension between homosociality and prescribed attitudes towards sexuality. While individuals living alongside futuwwa associations occasionally accused their members of homosexual practices,30 modern scholarship has had a tendency to avoid engaging with the blurred lines between homosocialisation and control over gender norms and sexual practices that are present in futuwwa treatises composed in the thirteenth and fourteenth centuries. The insinuations and imprecisions of these texts with regard to this particular set of relationships suggests to me that the freedom enjoyed by these associations was reinforced because they operated on the peripheries of more ‘official’ local, urban hierarchies and represented an alternative form of self-governance. This freedom – or space – from the ‘centrality’ of other organisations made futuwwa associations spaces within which ambiguities became normative. These ambiguities could have existed with regard to religious identities and practices as well as with regard to sexual activities.31 While heterosocial relationships were clearly and carefully defined within the context of futuwwa, homosocial relationships were not. The language used to prescribe the relationships between men in treatises on futuwwa, as well as in the descriptive literature related to the associations, is filled with ambiguities. And these ambiguities suggest that while the relationships between men and women were

Not for distribution or resale. For personal use only.

social graces and urban spaces

conceived in rather limited ways and clearly dictated by various codes and institutions, certain aspects of the relationships that existed between those men engaged with futuwwa were complex or, perhaps, less than normative and, therefore, intentionally left ambiguous.32 Texts on futuwwa in Anatolia do attempt to encourage chastity and even abstinence. However, the relationship between novice and master is often depicted as an intimate bond between two people who engage with one another both emotionally and physically. The dictated exclusion of women in these texts indicates a strong preference for male company, and much of the language used insinuates that hospitality between members (and even visitors) could be extended beyond the dinner table. In fact, new work on medieval North Africa has directly linked the performance of hospitality to sexuality. At the same time, chastity is prescribed even if various kinds of physical intimacy are intimated. Erzincan: a Micro-history Two Armenian-language futuwwa (Arm. manktut‘iwn) treatises, filled with allusions to Christianity, were composed in the city of Erzincan in the late thirteenth century by a celibate Armenian priest.33 The texts are remarkably similar to the Arabic-, Persian- and Turkish-language futuwwa treatises penned in Anatolia during the thirteenth and fourteenth centuries and, comparable to those texts, were reproduced for several centuries in cities all over the Ottoman Empire.34 The author in question was named Yovhann≠s (or John, d. 1293), and was the same prolific, peripatetic priest who had previously translated the Arabic-language Isma‘ili text, the Epistles of the Brethren of Purity. He went on to author a commentary on the Movements of the Celestial Bodies and a wide range of advice (Arab. naßi˙at; Arm. xratakan) literature for everyone from ‘average’ people to the Armenian nobility in the region surrounding Erzincan.35 In taking into consideration the changing political and social landscapes of the late thirteenth and early fourteenth centuries, the complex cultural situation in the city of Erzincan, and the life of Yovhann≠s, I suggest that these two Armenian futuwwa treatises represent a specific moment of religious and social interchange in the urban history of Anatolia. These two texts, in fact, reflect the socio-cultural porosity that prevailed, in particular in the aftermath of the Mongol expansion there,36 when futuwwa­-based urban organisations became more prevalent and seem to have offered opportunities for alternative forms of self-governance in cities. I have argued elsewhere that the impetus behind the composition of Armenian treatises was part of a larger Armenian response to the political and social insecurities of the time period, most generally understood through not only the lens of the arrival of the Seljuks and

Not for distribution or resale. For personal use only.

123

124

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

the Mongols, but also an Armenian awareness of material loss, the establishment of Islamic institutions in the region and a communal consciousness of Armenian conversion (both forced and voluntary) to Islam.37 It is within the context of Erzincan’s multi-valent environment that both Yovhann≠s (d. 1293) and Kostandin (d. 1314) were educated and produced many of their voluminous written works both for members of religious and lay brotherhoods and for the masses. While we know that Yovhann≠s was ordained a vardapet of the church, most scholars agree that Kostandin was most likely partially educated within the walls of a monastery, but left the protection and domination of the church to lead a more worldly life.38 The fact that these two Erzincan-based authors were contemporaries offers us rare insight into the realities of the phenomenon of brotherhoods in Anatolian cities. The prolific work completed by the travelling priest-poet Yovhann≠s suggests that he was familiar with both Arabic and Persian (based on his translations and adaptations of texts from those languages) and that he was deeply engaged with guiding people of various life stations, including average people, nobles, priests and members of the Armenian futuwwalike brotherhoods. The image that his work carves of Erzincan and Anatolia in the late thirteenth century is limited by the genres within which he writes. And while we are able to grasp bits of the type of person he might have been thanks to some textual hints, his writing is more formal and distant than that of his contemporary, Kostandin. The imagery in Kostandin’s poetry encourages us to imagine groups of familiar men, seated together, reading poetry, drinking wine, dreaming of God and love, roses and nightingales: O Lord on high, I beseech thee Have mercy on me, compassionate Father And bestow your great heavenly gifts upon me, unworthy one So that I may be drunk with the grace of the Spirit and love the heavenly life And leave this transitory world behind me, Kostandin.39 Many of his poems begin, ‘Dear brother, bend your ear’; some are dedicated to sharing specific messages with brothers (‘About brotherhood, good and bad’), and others are composed for specific individuals. Kostandin also writes about love and creates layered images of romance: When I see her/his elegance while (s)he is dancing Her/his moonlike countenance with her/his black hair Then I approach her/him with the saz40 And tell her/him I will gladly be her/his servant.

Not for distribution or resale. For personal use only.

social graces and urban spaces

That hour will be a sign for us, When I drink a glass Brimming with clear wine The colour of pomegranates Which resembles your face . . . There are some brothers who like me And want a worldly song in writing, Therefore I have recited many a love poem In a well-known melody.41 Other poems paint a picture of homosocial drunken debauchery in which the beloved (the language is ambiguous as to whether the beloved is male or female) is celebrated in gardens. The poetry inspired by the beloved (and by the alcohol) speaks to a certain loss of control, all too familiar to anyone who has ever been inebriated: I call together a gathering in the field, The nightingale is present, inebriated, Bright cupbearers in that big house, They stop and give us wine. Cupbearer,42 fill and present the goblet, My heart delights in your love, Let me drink of that goblet, Which is filled with scarlet wine. Singer, equip yourself with the six-string violin, That the baby tree may dance in our midst, O, slender and long twig, Your face is like the moon. Your eyes are like the sea, Your brows are drawn like a bow, Like sentinels of your face, I fear you will pierce me to death.43 The romantic eroticism of this particular poem and the author’s fear that the beloved might ‘pierce’ him ‘to death’ seems to me the clearest indication of the ambiguous homoeroticism of the work of Kostandin. At the same time, it would seem that relations between the brothers engaged in Kostandin’s poetic, fun-filled, spiritually-concerned confraternity were not without a competitive edge: Today my soul is joyful, I saw a deep wish fulfilled; I am drinking without lips a glass of that wine; I am drunk with that love and my thoughts are there, Where He is, I do not need the many men that wish me ill and pursue me. In this world I am like a fool because of my meditations

Not for distribution or resale. For personal use only.

125

126

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Some consider me wise, others think I am mad and out of my senses Some wish me ill and gnash their teeth against me, Some deem it fit to shed my blood, then their wish would come true . . .44 It is within this colourfully complex social and political environment that we must imagine Yovhann≠ sitting down (probably in the monastery of St Minas, situated amongst what is known today as Ke∞i∞ Da©ları (Turk. the priests’ mountains)) to compose his treatises on Armenian futuwwa. A reader of Arabic and Persian who was in the employ of the Armenian Church (based in Sis, over 600 km southwest of Erzincan) and lived in a trading hub, Yovhann≠s must have been well aware of the great political changes taking place around him in the aftermath of the Mongol takeover of the region. He must also have been conscious of the kind of activities that were a part of the quotidian in the lay confraternities in and around the city, whose members (based on analysis of Kostandin’s poetry) seem to have lived in between the marketplace and poetry parties, straddling a quest for spiritual enlightenment and an engagement with worldly pleasures. The Armenian Treatises: Exclusive Brotherhood, Inclusive Intimacy and Wine The two Armenian texts do not explicitly state that women are forbidden from participating, however, the association is called an association of manliness. Still, in the first of the two treatises, one might get the impression that women were tangential or associate members as there are explicit directions for almost doubling the number of times a female must kneel during prayer: ‘Stand in prayer three times every day . . . At every time, let each male perform 12 genuflexions and every female, 20’.45 Still, women are present in the textual space almost uniquely with reference to marriage and sex. Men in the association were meant to be (at least primarily) celibate. The first treatise reads: ‘Keep yourselves holy from sins and remain in pure virginity until you are worthy of taking the crown of blessing such that your bodies become gifts to God . . .’46 Thus, celibacy prior to marriage (and the donning of the crown associated with the celebration of that particular sacrament) is considered significant. The text continues: After taking the crown of blessing and being united in holy matrimony with your wife, who is appropriate in the eyes of God, keep yourself holy from the foreign and impure bed, such that you not lose the crown of blessing and fall into destructive malediction and deception . . .47

Not for distribution or resale. For personal use only.

social graces and urban spaces

The second text repeats this logic and suggests that desire was not only uniquely felt for women, but that feeling desire for a woman other than one’s wife could turn a man into a dog: As much as you can, remain without a woman, keep yourself holy from foreign sins. And if you want to marry, take an obedient and halal [Arab. appropriate, permissible] woman according to the blessing of the crown and do not show desire for other women, for he who mixes with foreign is called a dog and not a man.48 While relationships with women are frowned upon in futuwwa treatises – and specifically physical relationships with women other than one’s halal wife, the relationships between men are considered fundamental to the experience of futuwwa. Specifically, the rapport between master and novice is essential to the organisation and its code. Each novice was ‘adopted’ by his master who acts as a parent to the youth, initiating him into the kind of knowledge associated with the association and also behaving as an individual worthy of emulation. At the same time, Yovhann≠s considers the ‘leader of the youths’ more than just a person; he is a place where ‘lost’ individuals can be found: The leader of the youths must do good works, such that he is blessed and benevolent, for the benevolent man wins over everyone and his blessed conduct, like a father and a parent, becomes a place of meeting for those lacking a leader and for those who are foreign and blessed for he has pity and blesses the poor and the needy, the orphans and the men.49 At the same time, these associations did behave like mini hotels for travellers and visitors from other places. At the end of the day, after pooling their funds together and preparing a meal, they would often make that meal available for guests. The text forces one to question whether or not these moments of socialisation offered opportunities for more than just intellectual or spiritual intimacy. (As noted above, recent work has suggested a relationship between the hospitality of medieval confraternities and physical intimacy.) If a vardapet [Arm. teacher] should come from another country, the head brothers should go to visit him and honour him, and take from him his scholarly words and convene around him according to the preaching of the Holy Gospels, and when he leaves, they should send him off with love and according to their physical possibility and capacity, give corporeal fertility to the spiritual sower. May they do the same for bishops. And also for monastics, prayerful brothers and hermits, may they be hospitable and giving of rest.50

Not for distribution or resale. For personal use only.

127

128

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

The language of this particular passage lends itself to an erotic or sexual interpretation and also links some of the imagery from Kostandin’s ‘worldly’ poetry to the ‘religious’ practices related to instruction in the institution Yovhann≠s describes. While not explicit, the notion that a novice should ‘give corporeal fertility to the spiritual sower’ and ‘send him off with love’ seems almost to clearly engage with the idea of insemination, whether intellectual or physical. Similar to the futuwwa associations in the rest of Anatolia, the Armenian associations in Erzincan used a distinctive dress to set themselves and their brotherhoods apart from the ‘others’. This was a kind of belt that was attached to a girdle. In Yovhann≠s’ description of said belt, he specifically links the wearing of it to the preservation of celibacy, or at least to an attempt to control sexual desire: The fifth kind of belt is that of the travellers: the belt ties the girdle [Arm. drawssak]. According to that, they draw the belt along their loins to bridle desire, for desire is born in the loins and below the areas that are near the liver and the kidneys.51 Again, while sexual engagement is not necessarily prohibited in these codes, it is highly regulated by the rules set forth in the two constitutions. And abstinence prior to marriage is highly praised while sexual engagement with ‘foreign’52 women is highly criticised. Still, similar to other Anatolian codes on futuwwa, the Armenian codes tell us that there was a ceremony within which the belts were placed on the ‘novice’, thus creating a space of physical intimacy (and vulnerability?) amongst the brothers. Wine creates a problem for our analysis of these texts as they are contradictory as to whether one should or should not drink it. In the first of the two treatises – the text that seems like it was written for a larger audience associated with the association and that may have even included women, tangentially – Yovhann≠s encourages eating and drinking. He writes, ‘If you eat and drink, may you do so with all the glory of God.’53 In the second text, he criticises wine and the negative effects it can have on a man’s life: ‘Do not love wine and drunkenness, for wine mislays four main things of man: when he uses it, it brings harm to the soul, to the mind, to chastity and to duty.’54 Again, it is not necessarily wine itself that is criticised, but the drunkenness that it inspires and that makes a man less chaste, among other things. Still, again, the insistence on sexual activity seems disproportionate to the interest of the texts with other, more spiritual, topics. Conclusion The political confusion of late medieval Anatolia allowed for an important urbanisation to take place in the region. Because of the ‘in-betweenness’ of late medieval Anatolia, marginalised

Not for distribution or resale. For personal use only.

social graces and urban spaces

(and potentially subversive) organisations like futuwwa confraternities played more important roles in certain cities. At the same time, this particular organisation – whose codes were rooted in a series of seemingly intentional ambiguities and contradictions – created spaces in Anatolian cities for different kinds of cultural, faith and social practices that challenged normativities prescribed by religious hierarchies, polities and customs. While academia has known a relatively robust fetishisation of homoerotic poetry in the medieval Islamic context, scholarly engagement has come short of helping us to understand this phenomenon in the Islamic world amongst non-Muslim populations. The liminalities of the time period encouraged cross-cultural fertilisation and the blurring of boundaries between faith groups and the spaces they occupied, as diverse communities strove to engage with one another. Gender and sexuality are understudied in medieval Anatolian contexts, but the sources can and do provide us with a means to move towards better understanding the ways in which gender was constructed and sexuality expressed during this time period. Tentatively, this brief study argues that even those institutions prescribing ‘manliness’ in late medieval Anatolia are filled with ambiguities and contradictions with regard to behaviours surrounding socialisation and sexuality. Ultimately, these ambiguities were also (re)produced in an Armenian, Christian context, at the peripheries of both ‘Armenia’ and the Seljuk sultanate.

Notes 1. For example, see Pancaro©lu, ‘Hospitality’. 2. Aflaki, Feats of the Knowers of God, p. 48. 3. Vryonis, Decline of Medieval Hellenism, p. 155. 4. Melville, ‘Anatolia’, pp. 97–101. 5. As cited and translated in Redford, Landscape and the State, p. 53. 6. In her recent PhD dissertation, Suzan Yalman has termed this particular sultan’s efforts to develop construction and social services in cities a ‘commitment to urbanism’, see Yalman, ‘Building the Sultanate of Rum’, p. 233. 7. See Rogers, ‘Waqf and Patronage’; Erdmann, Das anatolische Karavansaray; Turan, ‘Selçuklu Kervansarayları’; Wolper, Cities and Saints; and Blessing, Rebuilding Anatolia. 8. While the equation of habitus with culture is not universally accepted, some social scientists have argued that habitus can operate in ways very similar to culture, occasionally giving rise to a sense of ethnicity. See Bentley, ‘Ethnicity’. Pierre Bourdieu would vehemently disagree, see Bourdieu, Outlines, pp. 24–7. But local Anatolians might not. 9. Sultan Veled, Divani Sultan Veled, p. 462. 10. There are several different transliteration systems used for the Armenian language. By and large, Armenologists employ the Hübschmann-Meillet system while the ALA-LC system is also popular. At the same time, the Armenian alphabet underwent an orthographic change during

Not for distribution or resale. For personal use only.

129

130

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

Armenian SSR. For both of these reasons, the name ‘John’ is transcribed as Yovhann≠s, Hovhann≠s, Hovhannes, etc. in various contexts. 11. Pałtasaryan, Hovhannes, p. 217. 12. Pałtasaryan, Hovhannes, p. 217. 13. Ohlander, Sufism, pp. 271–91. 14. Selçuk, Chapter 4, this volume, pp. 100–4. 15. See al-Qushayri, Risålat. 16. Selçuk, Chapter 4, this volume, p. 99. 17. See Karamustafa, Sufism: The Formative Period. 18. Selçuk, Chapter 4, this volume, p. 96. 19. This argument is carefully made by Jo Van Steenbergen in his article about ritual and the Mamluk capital, Cairo. Van Steenbergen, ‘Ritual’, pp. 227–35. 20. Ridgeon, Jawanmardi, p. 8. 21. Meisami, ‘Writing Medieval Women’, pp. 49–51. See also Toorawa, ‘Language’, pp. 251–65. 22. Shaikh, Sufi Narratives, p. 13. 23. Barkan, ‘Osmanlı imparatorlu©unda bir iskân ve kolonizasyon metodu olarak vakıflar ve temlikler I’, pp. 279–353; Vryonis, Decline of Medieval Hellenism, pp. 396–402. 24. The title akhi (Arab. ‘my brother’; Turk. a version of the term akı, ‘brave man’) was used in futuwwa-based confraternities in Anatolia for ‘master’ members. 25. Taeschner, ‘Beiträge’, p. 20; Hasluck, Christianity and Islam, p. 383; Vryonis, Decline of Medieval Hellenism, p. 401. 26. Vryonis, Decline of Medieval Hellenism, p. 401. 27. Ibn Battuta, Travels in Asia and Africa, 1325–1354, p. 125. 28. Selçuk, Chapter 4, this volume, p. 107. 29. Xa¥ikyan, Colophons, pp. 41–7 and 101–2. 30. The most critical commentary on futuwwa practices was composed in Cairo in the early fourteenth century by al-Turkumani al-Hanafi, a student of Ibn Taymiyyah (d. 1328). In particular, he criticises the older members of futuwwa associations gazing upon the naked bodies of the youthful members as they invest them with the pants of futuwwa during their initiation rites. See Irwin, ‘Futuwwa’, pp. 161–70. See also al-Hanafi, Kitåb al-luma‘, pp. 113–24 and 501–10; Ridgeon, Jawanmardi, p. 36. 31. For a larger picture of this in the general context of Islamic literary culture, see Bauer, ‘Die Kultur der Ambiguität’. 32. This interpretation is highly inspired by the work of Khaled Rouyaheb, Walter G. Andrews and Mehmet Kalpaklı. 33. The Armenian term manktut‘iwn has a meaning similar to the Arabic term futuwwa. Armenian manouk means male youth (sing.) and mankti functions as a pluralis tantum with the same meaning. Thus, the term manktut‘iwn translates as ‘the qualities of being a male youth’, similar to the Arabic term futuwwa. The Armenian-language title manktawag (or ‘leader of the youths’) is used from the eleventh to the fifteenth centuries, from the Monastery of Goš (north of Lake Sevan, in present-day Armenia), to Erzurum, to Ani to Soltaniyeh (in presentday Iran). By the sixteenth century, it would appear that this term had been completely replaced by the Turkish akhi. 34. Anasian, ‘The Turkish fütüvvet’, pp. 161–84; Grigoryan, ‘Arevmtyan’, pp. 115–27; P‘ork‘šayan, ‘Nor Naxiǰevani’, pp. 184–92.

Not for distribution or resale. For personal use only.

social graces and urban spaces

35. Pałtasaryan, Hovhannes. 36. Kafadar, Between Two Worlds, introduction. 37. Goshgarian, ‘Futuwwa in Thirteenth-century RËm and Armenia’, pp. 227–63. 38. Van Lint, ‘Kostandin of Erznka’, pp. 8–11. 39. Van Lint, ‘Kostandin of Erznka’, p. 118. 40. A saz is a stringed instrument, commonly associated with poetry gatherings. 41. Van Lint, ‘Kostandin of Erznka’, pp. 338–9. 42. He uses the term sałi (Pers., såghi), for cupbearer. In the medieval Persian poetic tradition – which was very much alive in thirteenthcentury Anatolia – the role of the såghi was central to any feast. Usually male, and of handsome countenance, he was both the server of wine and oftentimes acted as a muse. 43. Van Lint, ‘Kostandin of Erznka’, p. 355. 44. Van Lint, ‘Kostandin of Erznka’, p. 190. 45. Pałtasaryan, Hovhannes, p. 225. 46. Pałtasaryan, Hovhannes, p. 227. 47. Pałtasaryan, Hovhannes, p. 227. 48. Pałtasaryan, Hovhannes, p. 235. 49. Pałtasaryan, Hovhannes, p. 236. 50. Pałtasaryan, Hovhannes, p. 237. 51. Pałtasaryan, Hovhannes, p. 232. 52. The term used in the text is otar, meaning ‘other’ or ‘foreign’. As it does not specify another religious group (i.e., aylazg for Muslim), I have understood this term to mean a woman to whom a man is not married, or a woman foreign or other to the matrimonial bed. 53. Pałtasaryan, Hovhannes, p. 225. 54. Pałtasaryan, Hovhannes, p. 236.

Not for distribution or resale. For personal use only.

131

E[)JNBURGH University Press J

Not for distribution or resale. For personal use only.

CHAPTER SIX

Transformation of the ‘Sacred’ Image of a Byzantine Cappadocian Settlement Fatma Gül Öztürk Until recently, there were two popular misconceptions in the ­academic community related to Byzantine Cappadocia, the region in central Turkey: (1) that Cappadocia was a Byzantine monastic centre; and (2) that the history of Byzantine Cappadocia came to an end with the arrival of the Seljuks in the last quarter of the eleventh century.1 Accordingly, convinced of the inordinately generalised monastic identity of the region, scholars had initially categorised the settlement of Açıksaray in Cappadocia as a monastic site that dates back to the tenth and eleventh centuries, although literary and physical evidence is absent. At the other extreme, while advocating a secular purpose, Thomas Mathews and Anna-Christine Daskalakis-Mathews suggest that the inverted T-plan scheme and decorative elements of the rock-cut façades of those in Açıksaray and similar complexes in the region are secular Islamic in origin.2 The dual character of the settlement of Açıksaray, which apparently bears both Christian and Islamic features, is also strengthened by legends that put the settlement in connection with Hacı Bekta∞ Veli, the founder of the Bekta∞i Sufi order in the thirteenth century, who was settled in Sulucakarahöyük (today Hacıbekta∞), just 30 km north of Açıksaray. Açıksaray’s diverse architectural vocabulary and entangled identities can be directly linked with its location between the Christian and Islamic worlds. Paying particular attention to this, and to the impact of prior scholarship on the formation of these identities, this chapter will discuss the invention of a ‘sacred’ Christian image of a secular Byzantine site that emerged on a medieval border, as well as the adaptability and transmission of this image of a non-Muslim site in an Islamic land up to the present day, through Açıksaray. Cappadocia is the name commonly given to a particular area in central Anatolia with boundaries that have been drawn and redrawn on many occasions throughout its history. Cappadocia can be defined as an area of unique geological formations that have been subjected over time to intensive rock-cut architectural activity (Plate 7). The boundaries of this idiosyncratic volcanic landscape can be marked

Not for distribution or resale. For personal use only.

136

ARCHITECTURE AND LANDSCAPE IN MEDIEVAL ANATOLIA, 1100–1500

by the contemporary cities of Aksaray, Nev∞ehir, Kayseri and Ni©de in present-day Turkey (Figure 6.1). Although the majority of surviving rock-cut architecture in this area – which include ecclesiastical, monastic and secular establishments – are generally dated to the Byzantine period, especially to the ninth–eleventh centuries, the practice of carving the rock into living spaces seems to have been a continuous and prevalent tradition that is still practised today, but has origins back in antiquity and beyond. Since becoming known to the Western world in the early eighteenth century, Cappadocia has consistently been considered a Byzantine monastic centre; however, there is not a single document referring to Cappadocia in this sense, and it is unlikely that any will ever come to light.3 Moreover, recent scholarly pursuits have challenged the monastic identity of Cappadocia, redefining several settlements that were initially thought to be monastic as secular instead.4 On top of this, the notion suggesting that the Byzantine moment in Cappadocia ended with the arrival of the Seljuks ignores all the evidence pointing to a revival of Byzantine architectural and artistic production in the region during the thirteenth century. On the contrary, evidence indicates that in medieval Cappadocia, boundaries between the ‘sacred’ and ‘secular’ spheres, as well as between ‘religious’ and ‘ethnic’ groups, were less distinct than is often suggested today. Cappadocia: Monastic Interpretation Paul Lucas, appointed to travel to the eastern Mediterranean by the French court of King Louis XIV, was the first modern European to visit Cappadocia in the early eighteenth century. The engraving published alongside his travel notes shows the unique geological formations as uniform cones, topped with imaginary statues of Christian figures. Lucas, apparently unable to truly understand this peculiar landscape, claimed that these ‘strange carved spaces in the volcanic cones were the hermitages of Byzantine monks’;5 and the European travellers and explorers who followed Lucas also suggesting that the harsh volcanic wilderness was likely to have attracted a large monastic community, echoing Lucas’ initial assumption of an accumulation of monks in the region. Later, nineteenth-century travellers such as Henry Tozer would suggested a direct association between the supposedly monastic establishments in Cappadocia and the legacy of St Basil the Great (d. 379), the bishop of Caesarea (today Kayseri), one of the Cappadocian fathers, and probably the most important figure in the development of eastern monasticism.6 These claims, and the fact that for most of the twentieth century Cappadocia was studied primarily by art historians who were concerned mainly with the religious paintings found in the numerous rock-cut churches in the region, served to strengthen the exclusively

Not for distribution or resale. For personal use only.

•

I

Hactbekta~

~~~

)

\ \ \

Not for distribution or resale. For personal use only.

'

Salt Sea

"

\

/-

v At;Iksaray

Bezir Harle

.

£

c

~-y

K1ht;Iar

¢Jrgiip

NEV~EHIR. ~e ~ Hallat;

~

•

--------

AKSARAY.

~

Derinkuyu •

\.

1

1------------

~