The Biographical Tradition in Sufism: The Tabaqat Genre from Al-Sulami to Jami 9780700713592, 9781315028477

Citation preview

THE BIOGRAPHICAL TRADITION IN SUFISM

C u rz o n S tu d ies in A sian R e lig io n Series Editor: Sue Ham ilton, K ing’s College, London Editorial Advisory Board:

N ick Allen, University o f O xford C atherine Despeux, IN A LCO , Paris Chris M inkowski, C ornell University Fabio Ram belli, Williams College, Massachusetts Andrew R ippin, University o f Victoria C urzon Press publishes a Series specifically devoted to Asian R eligion, considered from a variety o f perspectives: those o f theology, philosophy, anthropology, sociology, history, politics and literature. T he prim ary objects o f study will be all the religious traditions o f the Indian sub-continent, Tibet, China, Japan, SouthEast Asia, Central Asia, and the N ear and M iddle East. T he m ethodology used in the works published in the Series is either comparative or one focused on (a feature of) a specific tradition. T he level o f readership ranges from undergraduates to specialist scholars. T he type o f book varies from the introductory textbook to the scholarly m onograph.

Tradition and Liberation T he H indu Tradition in the Indian W om en’s M ovem ent Catherine A . Robinson

Shinto in History Ways o f the Kami John Breen and Mark Teeuwen

Beyond Personal Identity Dogen, Nishida, and a Phenom enology o f N o-S elf Gere on K opf

Krsna: Lord or Avatdra? T he relationship betw een Krsna and Visnu Freda Matchett Proposals or scripts for the Series will be welcomed by the Series Editor or by Jonathan Price, C hief Editor; Curzon Press.

THE BIOGRAPHICAL TRADITION IN SUFISM The tabaqat genre from al-Sulami to Jämi

Jawid A. Mojaddedi

First Published in 2001 by Curzon Press Published 2013 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN 711 Third Avenue, New York, NY 10017 USA Routledge is an imprint of the Taylor & Francis Group, an informa business © 2001 Jawid A. Mojaddedi

All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record of this book is available from the British Library Library of Congress Cataloguing in Publication Data A catalogue record for this book has been requested ISBN 978-0-700-71359-2 ISBN 978-1-315-02847-7 (eISBN)

In m em ory o f N o rm an Calder

This page intentionally left blank

Contents

viii

Abbreviations

ix

N ote on presentation

Introduction 1

Part One C hapter 1

Sulam fs Tabaqät al-Süfiyya

C hapter 2

T h e H ilyat al-awliyä’

41

C hapter 3

T he Persian Tabaqät al-Süfiyya

69

9

Part Two C hapter 4

Q u sh ay n ’s Risala

C hapter 5

H u jw m ’s K a sh f al-mahjüb

99

125

Part Three C hapter 6

Jäm fs Nafahät al-uns

151

C onclusion

177

Notes

183

A ppendix

213

Bibliography

219

Index

224

vii

Abbreviations

T he following abbreviations are used: E l1 E l2 H KM L N Q R TABA TABS

T h e E ncyclopaedia o f Islam T h e E ncyclopaedia o f Islam

(First edition) (Second edition)

Abü N u caym al-Isfahänl, H ily a t a l-a w liy ä ’ cAll HujwTrl, K a s h f a l-m a h jü b (Ed. V. Zhukovski) Abü Nasr al-Sarräj, K itä b a l- lu m a c (Ed. R .A . Nicholson) cA bd al-R ahm än JämT, N a fa h ä t a l-u n s (Ed. M. Abedi) T h e Q u r’an (Cairo edition) Abu ’l-Q äsim al-Q ushayri, a l-R isä la (Eds A .H . M ahm ud & M. al-Sharif) (Tehran reprint) "Abdullah al-Ansärl, Tabaqät a l- S ü fiy y a (Ed. A .H . Habibi) Abü cAbd al-R ahm än al-Sulaml, Tabaqät a l-S ü fiy y a (Ed. J. Pedersen)

(Further details about these works are included in the bibliography.)

vili

Note on presentation

All Arabic and Persian terms have been transliterated, unless they are included in the Concise O xford Dictionary (N inth Edition), such as Q u r’an, hadith, shariah and fatwa. M y transliteration follows the system o f the Encyclopaedia o f Islam , w ith the modifications customary in works in English. Formulaic prayers are om itted. T he full titles o f the works referred to in the notes o f the text, together w ith their publication details, are provided in the bibliography

IX

This page intentionally left blank

Introduction

he tabaqât genre has been one o f the most productive genres o f the Islamic literary tradition. Com piled on the basis o f chronological principles o f organisation and using generic criteria for selection, these works take the form o f collections o f biographies.1 Tabaqât writings invariably depict the past o f a particular tradition o f religious affiliation or scholarship, the chronological parameters o f w hich conventionally stretch from an authoritative starting-point to the generation (tabaqa) im mediately preceding the assumed author. W hat is perhaps most remarkable about this genre is that generation after generation o f successors in each particular tradition have com piled their own tabaqât works, thereby producing a constant supply o f versions o f the same period o f history, only extended by a generation each time. T he sheer vastness o f this corpus o f literature has m eant that its use is widespread. In particular, they tend to be used as sources o f ‘ready-m ade’ biographies o f medieval scholars. M any historians have treated such works, for the sake o f convenience, as m o re-o rless accurate ‘databanks’ w ith w hich to reconstruct the history o f the periods they depict.2 However, in view o f recent develop­ ments in literary criticism and its increasing influence on the study o f Islamic historiography it seems untenable to assume ‘the im m ed iate equivalence o f w o rd and w o rld ’.3 D espite its undoubted im portance in the Islamic literary tradition, there has so far been no m onograph-length study devoted to the tabaqât genre w hich serves to identify its essential characteristics and functions. In this study I have explored these issues by means o f the literary study o f a single, neatly-defm ed and representative group o f texts, namely the tradition o f tabaqât w riting about the past o f Sufism.

T

1

Introduction

T he six tabaqat writings considered in this study constitute the main tradition o f Sufi historiography They thus occupy a position o f im portance comparable to the tabaqat writings o f each o f the main religious traditions o f Islam, such as the legal schools. M oreover, they share the same basic literary characteristics, in that they consist o f the biographies o f past representatives arranged in a predom inantly chronological pattern. W ithin such a context, each biography serves to convey inform ation both about its individual subject as well as about the com m unity to w hich he or she belongs, whilst at the same time constituting an integral part o f the linear narrative about the past o f that com m unity w hich is expressed by this arrangem ent. By focusing in detail on the structural characteristics o f each o f the six major tabaqat writings o f the Sufi tradition and the inter-relationships betw een them , it is anticipated that conclusions may be reached w hich are also applicable to the w ider use o f the genre. This book is divided into three parts. Part O ne consists o f three chapters, each devoted to a w ork traditionally ascribed to an eleventh century author. C hapter O ne is a study o f Abü cAbd alR ahm án al-Sulaml’s (d.412/1021) Tabaqat al-Süfiyya , the earliest example o f the Sufi tabaqat genre. C hapter Two examines the Hilyat al-awliya\ w hich is traditionally ascribed to Abü N u'aym alIsfahanl (d.430/1038), focusing on the portion o f the (ten-volume) Hilya w hich corresponds to the scope o f the whole o f Sulamfs work. C hapter Three considers the Persian Tabaqat al-Süfiyya, w hich is traditionally ascribed to "Abdullah al-Ansarl (d.481/1089). This is the first Persian w ork o f the genre, and is often considered to be essentially a Persian version o f its earlier namesake. T he three works discussed in Part O ne, by virtue o f being the earliest examples o f Sufi tabaqat works, tend to be used as authoritative sources o f inform ation about the early history o f Sufism. This study highlights the organisational structures o f these works and the characteristics o f their basic units o f material, in order to evaluate their literary functions and re-assess their historical significance. M oreover, a significant issue w hich is afforded considerable attention in Part O ne is the degree o f authorial control apparent in the surviving forms o f the works concerned. Part Two consists o f two chapters, devoted to tabaqat writings, each o f w hich constitute one com ponent section o f an eleventh century Sufi manual. C hapter Four is a study o f the tabaqat section o f Abü ’l-Qasim al-Q ushayri’s (d.465/1072) Risala , and C hapter 2

Introduction

Five that o f the tabaqat section o f cAlT ’1-HujwirI’s (d.ca.467/1074) K ashf al-mahjub.4 These two works have proven through the centuries to be the most popular works o f their kind in Arabic and Persian respectively. T he relationship betw een the tabaqat section and the rem ainder o f each o f these two works is examined in order to dem onstrate the functions o f this genre o f writing, highlighted w ithin such a literary context. Part Three consists o f one chapter, C hapter Six, w hich deals w ith cAbd al-R ahm an Jam fs (d.898/1492) Nafahat al-uns. This Persian w ork represents a fifteenth century revival o f the Sufi tabaqat genre. M ore specifically, Jam i introduces it as the direct successor o f the aforem entioned Persian Tabaqat al-Sufiyya , w hich is traditionally ascribed to Ansarl. Its biographies o f Sufis living before the twelfth century constitute a redaction o f that earlier work, whereas the rem ainder o f the Nafahat covers the period betw een the twelfth and the late fifteenth century. C hapter Six demonstrates the way in w hich the earlier w ork has been recontextualised, by identifying the priorities o f the redaction and highlighting the way in w hich the earlier w ork was actually utilised as a source, in contrast to the way in w hich it was classified. T he organisational fram ework o f the rem ainder o f Jam fs Nafahat is also considered, in order to observe how the biographies o f the later Sufis that have been added are linked back to those o f Sufis living before the twelfth century in a m anner w hich is characteristic o f the tabaqat genre. T he six works examined here represent a cohesive tradition o f writing. M ost o f them include explicit cross-references to each other. The Hilyat al-awliya’ includes a specific reference to Sulamfs Tabaqat al-Sufiyya , which implies that this ten-volum e work reached its final form after the completion o f its considerably shorter counterpart.5 T he title o f the Persian Tabaqat al-Sufiyya in itself is an indication o f the influence upon it from its Arabic precursor, which is also evident in the predominant principle o f organisation o f its biographies. Although QushayrT prefers to use the expression o f oral transmission, the tabaqat section o f his Risala , in both form and content, is based closely on Sulamfs w ork.6 H ujw irfs K ashf almahjub includes a specific cross-reference to the arrangement o f biographies in both Sulaml’s Tabaqat and Qushayri’s Risala;7 and, as m entioned above, Jami identifies his own work as the successor o f the Persian Tabaqat al-Sufiyya , which, in turn, he classifies as the successor o f Sulaml’s foundational work o f the genre.8 3

Introduction

T he analysis in each chapter o f this study follows the same principle o f focusing in ever-increasing detail on the individual text concerned; at first, the organisational framework is examined, followed by the overall structure o f a representative biography, and culm inating in the scrutiny o f the individual ‘building-block’ segments o f the text concerned. T he decision to num ber the com ponent sections o f each chapter, rather than to label them w ith headings, is because each chapter’s analysis proceeds according to a corresponding sequence o f stages. T he biographies o f Abu Yazld al-Bastaml9 (d.261/874) and Abu ’1-Qasim al-Junayd (d.297/910) have been selected as the representative biographies for each o f the works considered. This serves to facilitate a direct com parison o f the form and content o f corresponding material in each work, as well as to m onitor the developm ent o f these two specific biographical traditions. In the three chapters o f Part O ne additional attention is given to the re-assessment o f the historical value o f the texts concerned. This involves the observation o f the recurrence o f variants and com m on motifs, as well as the S itz im Leben suggested by the structures o f the text and the purported m ethods (predominantly oral) by means o f w hich its material was generated and transmitted. T he two chapters o f Part Two include additional discussions o f the implications o f the juxtaposition o f the tabaqat genre next to further com ponent sections o f the same w ork w hich belong to alternative genres. This includes comparisons o f the ways in w hich material about Abu Yazld, Junayd and others has been selected and arranged in the different sections o f the w ork concerned. In my analyses I deconstruct each tabaqat w ork to distinguish its organisational structures and biographical co n ten t, and to demarcate the divisions betw een the discrete segments w hich constitute the latter. In addition to shedding light on the process o f com pilation o f each w ork and highlighting the overall context in w hich discrete items are em bedded, this approach facilitates the com parison betw een different works ‘o f like w ith like’ in an appropriate and relevant manner. That is to say, the organisational framework, the overall structure and com ponent sections o f an individual biography as well as its discrete segments o f material are com pared w ith their respective counterparts in the other works. Although the significance o f precise distinctions betw een the different levels o f structure in a particular text may not be im mediately clear in isolation, it should becom e increasingly 4

Introduction

apparent as the inter-relationships w ith other texts o f the genre and the developm ent o f specific biographical traditions are explored in further detail. W hilst this book is a study o f a historiographical genre, it also unavoidably contains a certain am ount o f the narration o f history in the form o f the biographies o f the authors o f the individual works that are under scrutiny I have attem pted to restrict these biographies to the bare m inim um , and to make explicit the sources for individual items o f inform ation as well as to point out the growth and developm ent o f such details, where applicable. At the cost o f the frequent use o f expressions o f qualification, such as ‘attributed’ and ‘alleged’, I have given priority to m aintaining consistency w ith my observations in this study o f the historical value in general o f biographical genres. This approach has the benefit o f guarding against the influence o f later biographies o f the authors on the interpretation o f their surviving works. O n the same principle, in the exploration o f the processes o f their com pilation the actual structures o f the individual texts are given precedence over the accounts provided ostensibly by their assumed authors, as well as those offered in later biographical sources; the latter, and any additional sources, are taken into consideration only in so far as they correspond to the structures o f the surviving texts, w hich serve as the most authoritative points o f reference regarding their ow n history and com position. As H ans-G eorg G adam er sums up such an approach: [I]t is a basic principle for the historian that tradition is to be interpreted in another sense than the texts, o f themselves, call for. H e will always go back behind them and the m eaning they express to enquire into the reality o f w hich they are the involuntary expression.10 An earlier version o f parts o f C hapter Three was presented at Societas Iranologica Europaea : third European conference o f Iranian Studies (University o f Cam bridge, 1995), and was subsequently published in its proceedings (Ed. C.Melville; Weisbaden, 1999). An earlier version o f parts o f C hapter Four was presented at the 209th m eeting o f the Am erican O riental Society (Baltimore, 1999), and has been published in Studia Islamica.]] I should like to express my gratitude to all those w ho have helped to make it possible for me to w rite this book. T he British 5

Introduction

Academy are to be thanked for making the original research financially viable w ith the award o f a m ajor postgraduate studentship, and the British Institute o f Persian Studies for contributing towards a period o f study at the m anuscript library o f Tehran University. Patricia Crone made it possible for me to devote a year to the preparation o f the final draft at the Institute for Advanced Study in Princeton, w here I benefited also from the com pany and collegial encouragem ent o f M aroun Aouad, M ichael C ook, O leg Grabar and W olfhart Heinrichs. G .R . Hawting, E dm und Herzig and Alex Samely each read the original thesis in preparation for my viva voce exam ination (University o f M anchester). Julie Scott Meisami and Andrew R ippin both read a recent draft o f the revised and updated book version. I am very grateful for their time and attention. W hilst this w ork has benefited considerably from their comments, criticisms, reassurance and encouragem ent, I alone am responsible for its contents. I w ould also like to thank Jonathan Price o f C urzon Press and David M cC arthy o f Laserscript for their assistance and patience throughout the publication process. I remain indebted to my m other for the efforts she has made throughout my life to ensure single-handedly that I always had the opportunity to concentrate on my studies as m uch as I wished. N egin Nabavi also deserves a special m ention for reading the earliest and the final draft o f this book, as well as a num ber o f interm ediate versions, and consistently offering insightful sugges­ tions and criticisms. T he single most im portant influence on my academic w ork has been my doktorvater, the late N orm an Calder. I had the unique privilege o f studying under his guidance from my first year as an undergraduate to the com pletion o f my doctoral thesis. M y debt to him is immense, far beyond the scope o f a m onograph, and it is to his m em ory that I hum bly dedicate my work.

6

Pa rt O n e

A ll these memories, superimposed upon one another, now formed a single mass, but had not so far coalesced that I could not discern between them — between my oldest, my instinctive memories, and those others, inspired more recently by a taste or ‘p erfume’, and finally those which were actually the memories o f another person from whom I had acquired them at second hand — if not real fissures, real geological faults, at least that veining, that variegation o f colouring, which in certain rocks, in certain blocks o f marble, points to differences o f origin, age and formation. (M arcel Proust, Remembrance of things past, Vol. I, p.203)

This page intentionally left blank

Chapt er One

Sulam i’s Tabaqat al-Sufiyya

I

Ab

bu cAbd al-R ahm an M uham m ad b. al-Husayn al-Sulami 1 1 (d.412/1021) was an eleventh century Sufi scholar from Nishapur. O ver twenty works that are ascribed to him have survived.1 W hilst little is know n about the events o f his life,2 it w ould appear that he was greatly revered by his fellow-citizens; his biography in al-Khatib al-Baghdadfs (d.463/1071) Ta’rikh Baghdad m entions that his grave in Nishapur had already becom e a pilgrimage destination.3 It also specifies that the grave was located w ithin a small monastery (duwayra )4 w hich was nam ed after Sulami. T he monastery may have been established only after his death as an extension o f the grave, or it may even have been the place w here he had taught and com piled his works. T he inclusion o f a biography o f Sulami in the Ta’rikh Baghdad is apparently on account o f his visits to that city, w hen he would transmit reports about the Sufi leaders o f Khurasan to the scholars o f Baghdad (wa-haddatha biha can sh u yu kh K hurasan).5 Such occasions could also have provided an opportunity for him to collect reports about the Sufis o f Baghdad, w hich account for a substantial proportion o f the contents o f his own works. Sulami took his nisba (title denoting descent), by w hich he is com m only know n, from his m aternal grandfather, Isma'll b. Nujayd al-Sulami (d.365/976), w ho is usually referred to as Ibn N ujayd.6 H e describes the latter as one o f the em inent followers o f Abu cU thm an al-H irl (d.298/910), ‘w ho spread the Sufi path in N ishapur’.7 Ibn Nujayd is often classified in later tradition as a m em ber o f the M alam atiyya (the people o f blame),8 and a treatise about this group is in fact counted amongst Sulaml’s ow n surviving 9

Sulam l’s Tabaqat al-Sufiyya

works. It confirms that he himself was at least familiar w ith the term and its connotations.9 Ibn Nujayd is likely then to have been an im portant early influence on SulamI, perhaps even as his first teacher in Sufism. cAbd al-Karlm b. M uham m ad al-Samcani (d. 562/1162), over a century later, informs us that Ibn Nujayd was also very wealthy, and that Sulaml’s m other was the sole heir to his fortune.10 In this way, he may have provided the resources for his grandson to pursue his interest in Sufi scholarship, as well as the original inspiration. A lthough SulamI is claimed as one o f their ow n as m uch by Shafi’ite as by Sufi biographers, his surviving works indicate that he was prim arily a Sufi. In fact, the aforem entioned Risalat alMalamatiyya , as well as the Kitdb al-samac,u w hich is the earliest surviving m onograph on Sufi musical audition, suggest that he was im mersed in Sufism to an extent that he took an interest even in the m ore contentious aspects o f the tradition.12 T he two works that are most often m entioned by name in his medieval biographies are the Haqa’iq al-tafsTr, a w ork o f mystical exegesis,13 and the Tabaqat al-Sufiyya , a collection o f biographies o f Sufis.14 T he latter, as the earliest example o f the Sufi tabaqat genre, is arguably SulamI’s m ost influential work.

II T he Tabaqat al-Sufiyya is a collection o f 103 biographies o f Sufis.15 T he biographies are grouped into five consecutively num bered sections called tabaqat (generations) (See Fig.l), framed by an introduction at the beginning o f the w ork and a postscript at the end. T he first four o f Sulaml’s generation sections are each made up o f 20 biographies.16 T he fifth generation section consists o f 23 biographies.17 This final section contains an excess o f biographies perhaps because, since it represents the generation closest to the time in w hich the w ork was compiled, it proved too difficult to restrict to only twenty biographies. That is, SulamI may have intended to include only twenty biographies in each section, but decided to make the final generation the exception. T he introduction o f the Tabaqat al-Sufiyya describes the w ork as consisting o f 100 biographies divided into five ‘generations’, each made up o f 20 biographies (afaluh cala khams tabaqat . . . wa-adhkur f t kull tabaqat cashrTn shaykhan).]H M oreover, in the postscript it is 10

Sulann’s Tabaqat al-Sufiyya

stated that each biography contains ‘about tw enty segments (hikayat)\ despite the fact that this is clearly not the case for most o f th em .19 T he evident discrepancies are perhaps due to the wish to offer a neat and balanced account o f the m ethodology applied in the com pilation o f the work. M ost o f the biographies in the Tabaqat al-Sufiyya offer the date o f their subject’s death.20 An examination o f these dates shows that Sulaml’s five generations cover a period extending back from the late 4 th /1 0 th century to the late 2 n d /8 th century. T hat is to say, the earliest o f the first generation lived in the eighth century, whilst the latest o f the fifth generation lived in the late tenth century. It is perhaps to be expected that the latest ones to be included should be the im mediate predecessors o f SulamI, but the reason why the earliest o f them should be from the eighth century is not selfevident. An explanation is provided in the introduction o f the work, w here it is stated that M uham m ad, the last o f the prophets, was succeeded by saints (wa-atbaca ’l-anbiya’ calayhim al-salam bi-rlawliya’),2] and that the first three generations o f these ‘saints’ were the sahaba (the P rophet’s companions), the tabi'un (the successors o f the sahaba) and the tabicu ’l-tabicm (the successors o f the successors) respectively. T he latter constitute the same three ‘generations’ w ho are know n collectively as the ‘pious predeces­ sors’ (al-salaf al-salih ), and are considered to be the religious successors to M uham m ad in the Sunni tradition.22 T he continua­ tion o f the introduction concerns the position o f the individuals included in the Tabaqat al-Sufiyya in relation to the salafi in the image o f the past that is being structured: I already m entioned, in the Kitab al-Zuhd , the sahaba, the tabicun and the tabicu ’l-tabicfn , century by century and generation by generation, until the turn o f those endow ed w ith mystical states (ahwat), w ho speak about unicity (tafirid), the truths o f unity (tawhid) and the application o f the methods o f detachm ent (tajnd). I therefore wished to compile a book about the lives o f the later awliya\ w hich I name Tabaqat al-Sufiyya. (TABS, 5 .7 -1 0 )

T he Sufis whose biographies are contained in the Tabaqat alSufiyya are thus presented as the successors o f the salafi w ho are said to have been the subject o f an earlier w ork by SulamI, entitled the Kitab al-Zuhd.23 It is implied that the Tabaqat al-Sufiyya is its sequel, 11

Sulam i’s Tabaqat al-Sufiyya

and that the Sufis it includes are therefore successors ultimately o f the Prophet himself, through those three interm ediary salaf generations. This serves to explain the fact that the earliest individuals to be included in the Tabaqat are from the eighth century, and it takes a form that one w ould expect from the representative o f any Sunni M uslim tradition; SulamI is depicting the past o f his own specific tradition as extending back to the time o f the prophet o f Islam, by a m ethod, or route, that serves to attribute to Sufism the same foundations as Sunni Islam. T he Sufis included in the Tabaqat are the ‘later aw liya\ w ho are characterised as being ‘those endow ed w ith mystical states, w ho speak about unicity, the truths o f unity and the application o f the m ethods o f detachm ent’, w ith the implication that these are their distinctive qualities in relation to those w ho preceded them. It is perhaps best to return to the actual structure o f the w ork itself, in order to gain an insight into the m ethodology that has been applied (see F ig .l). T h e criteria used for grouping biographies into the five generation sections is particularly instructive. T he members o f Sulami’s fifth generation section, for w hom dates are supplied in the Tabaqat itself, are said to have died betw een the years 378/988 and 341 /9 5 2 ;24 his fourth generation section betw een 340/951 and 328/940;23 his third generation section betw een 330/941 and 291/903;26 and his second generation section betw een 319/931 and 283 /8 9 6 .27 Thus a sequential pattern through time emerges in the last four sections, each o f w hich is made up o f the biographies o f individuals w ho could actually have been contem poraries.28 It is at this point that Sulaml’s first generation stands out as being especially problematic. This is because it includes not only the biographies o f individuals from the th ird /n in th century (e.g. H am dun al-Qassar (d.271/884)) as one would expect from the sequential pattern in the other generation sections o f the w ork (which represent a gradual recession in time), but also those o f individuals from as early as the second/eighth century (e.g. alFudayl b. cIyad, d. 178/803). T he pattern is broken by the inclusion o f the biographies o f these earliest figures, w ho could not possibly have belonged to the same actual generation as Qassar and his contem poraries, together w ith w hom they have been grouped into the same ‘generation’. O ne can therefore see that the first generation is anomalous for the opposite reasons to the final generation; whilst the latter is overlong in terms o f the num ber o f 12

Sulam i’s Tabaqat al-Sufiyya

biographies it contains for the same (relatively short) time span, the form er is overlong in terms o f the time-span covered by its selection o f tw enty individuals. R ath er than adding extra sections, SulamI has decided to cover a p eriod o f over a century in a single generation section. There is no inherent reason why he should have restricted him self to five generation sections. It may be then that brevity in general was a priority, or there may have been a scarcity o f Sufis from the earliest period (he needed twenty). T he fact that it is the first generation section w hich is anomalous in this way suggests that it was the most difficult to assemble, w ith regard to structuring a credible ‘generation’. T he inclusion o f the biographies o f individuals from the eighth century causes the break up o f the established pattern o f gradual recession, so it suggests that they were considered necessary by SulamI in spite o f this. This was probably because representatives o f the Sufis from this period w ould be required in order to bridge a continuity w ith the salaf. In theory, it w ould be possible for individuals living in the eighth century to have m et and obtained authority from the tabicu ’l-tab\c\n , the final generation o f the salaf. T he irregularity o f the first generation section indicates also that the Tabaqat al-Sufiyya was not in fact a straight-forward sequel to an earlier work, w ritten ‘generation by generation’ as its introduction suggests. R ath er it appears to be a backreading o f the past o f Sufism, from the standpoint o f the early fifth/eleventh century, w ith the ultimate goal o f reaching as far back as the authoritative salaf.29 A lthough the generation sections o f Sulami’s Tabaqat al-Sufiyya are presented in a chronological sequence, the individual biographies w ithin them are plainly not. T heir order may appear at first to be arbitrary, but w hen one compares their length, one can observe that the longest ones are positioned either right at the head, or at least towards the beginning o f each generation, thus suggesting a hierarchical pattern. T he m ore interesting biographies - the ones that have attracted the most material - are given first in each generation. In addition to this organisational principle, and often modifying it, the biographies o f individuals w ho are said to have had som ething in com m on, or to have had dealings w ith each other, are often juxtaposed (e.g. M uham m ad b. cAll ’1-KattanI (d.320/932), Ishaq b. M uham m ad al-N ahrajurl (d.330/941) and cAli b. M uham m ad al-M uzayyin (d.328/939), each o f w hom died in M ecca).30 13

Sulam i’s Tabaqat al-Sufiyya

T he individuals whose biographies are at the head o f the third, fourth and fifth generations have another significant factor in com m on.31 T hey are each said to have associated with, and thus obtained authority from (sahiba)32 the head o f the second generation, Abu ’1-Qasim al-Junayd (d.297/910). Junayd is the first person in the list o f the ‘authoritative associates’ o f Abu Sa'ld b. al-AcrabI (d.341/952), the head o f the fifth generation. H e is the only one nam ed for Abu Bakr al-Shibll (d.334/945), the head o f the fourth generation. Al-Jurayrl (d.311/924), the head o f the third generation, is said to have been the immediate successor to Junayd.33 In his ow n biography, w hich is positioned at the head o f the second generation, Junayd is said to have been approved by everyone (wa-maqbul ca ld ja m f al-alsina). It is also stated that he was a jurist w ho practised according to the school o f Abu Thaw r (d. 240/854).34 W hilst there may be no corroboration for them , in this context, such assertions serve as m ethods o f legitimization. Regardless o f w hether or not they are based on historical facts, they make it clear why association w ith Junayd could be especially valuable for members o f later generations —in order that they too m ight be considered authoritative. T he overall effect is that Junayd is presented in this schema as the most im portant Sufi o f his own time, from w hom the pre-em inent tradition o f later generations acquired authority. Sulami’s choice o f ‘head biographies’ most likely indicates a personal preference. It is significant, in this regard, that the individuals whose biographies are at the head o f the last four sections are mostly from Baghdad. Although the Tabaqat includes nearly twice as many Khurasanians to Baghdadians, the biographies o f the latter are consistently placed in positions o f greater prom inence in the hierarchical structure o f each generation.35 Junayd, JurayrI and Shibll are each said to be from Baghdad, as well as the individuals w ho follow immediately after them (see Fig.l). Ibn al-Acrabi is described as a student o f Junayd, but he is said to have settled in Mecca. His biography is followed by that o f som eone else w ho is said to have settled in Mecca, namely Abu cAm r al-Zajjajl. (This com m on factor probably accounts for the position o f prom inence o f the latter’s biography).The third biography, w hich is also the longest in this generation, is that o f Jacfar al-K huldl from Baghdad, a m ajor transm itter o f biographical material about Junayd.36 O n the whole, therefore, the order o f 14

Sulam i’s Tabaqat al-Sufiyya

biographies in the Tabaqat suggests a preference for the Sufis o f Baghdad, particularly those w ho can be associated w ith Junayd. This is not altogether surprising in view o f the aforem entioned report o f Sulami’s visits to Baghdad, w here he was considered a notew orthy scholar. T he Tabaqat also reveals a special interest in ShiblT, for, not only is his biography ‘the head’ o f the fourth generation, but it is also the longest biography in the w hole w ork by a considerable margin. It w ould appear to be in keeping w ith his authorship o f works on contentious topics, that Sulami should show a special interest in one o f the most controversial Sufis o f Baghdad.37 As already m entioned, the biography w hich is placed at the head o f the final generation is that o f Ibn al-AcrabI. It is significant that he should be chosen in preference to Sulami’s grandfather, Ibn Nujayd, and NasrabadhI, not to m ention the aforem entioned Khuldi. In his biography, Ibn A'rabI is described as a scholar w ho com piled m any works about the Sufis (wa-sannafa li-’l-qawm kutuban kathira).3H M oreover, in another source from this period, the Hilyat al-awliya\ he is attributed w ith a (no longer extant) w ork called the Tabaqat al-Nussak.39 T he title suggests that it could have belonged to the same genre as the Tabaqat al-Sufiyya. Therefore, an explanation for the position o f his biography is that Sulami may have wished to position a scholar amongst the Sufis (like himself) at the head o f this final generation. T hat is to say, the choice o f Ibn A 'rabI as the head o f the final generation section is an indication o f the im portance w hich Sulami accorded to Sufi scholarship, the notion o f ‘scholarship’ here being history and continuity.40 T he anomalous first generation section also deserves a m ention, w ith regard to the hierarchical order, for despite the inclusion o f the biographies o f Sari ’1-SaqatI (d.251/865) and al-H arith al-Muhasibl (d.243/857),w ho are both rem em bered as teachers o f Junayd, they are not positioned at the head, as one m ight expect from the pattern in the remainder o f the work. In fact, they are only the fifth and sixth biographies, respectively, being preceded, in order, by the biographies o f al-Fudayl b. cIyad, Ibrahim b. Adham, D hu ’1-Nun al-Misrl and Bishr b. al-H arith.41 T he precedence o f three o f these biographies (all apart from D hu ’1-Nun) is probably due to the anomalous nature o f this section, in that it includes figures from the 2 n d /8 th century as well as those from the 3 rd /9 th century, and consequently is not determ ined simply by the application o f a hierarchical principle to a selection o f contemporaries. T he date 15

Sulam fs Tabaqat al-Sufiyya

given for Fudayl’s death is the earliest provided in the work at 178/ 803; although a date is not supplied in this work for Ibn Adham, his death has already been dated as 161/777—8 in an earlier source.42 M oreover, SulamI mentions that, whilst in Mecca, he was an associate o f Sufyan al-Thawri (d. 161/777—8)43 and Fudayl (kharaja ila M akka wa-sahiba biha Sufyan al-Thawri wa-’l-Fudayl b. cIyad);44

Bishr b. al-H arith (d.227/841) is also linked back to Fudayl, and this association is probably responsible for the position o f his biography in the section.45 However, the position o f D hu ’1-Nun’s (d.245/860) biography, which is also the longest in the first generation, cannot be accounted for simply on the basis o f chronology. It w ould appear that D hu ’1-Nun was held in particularly high esteem, even in relation to his contemporaries Saqati and MuhasibI, the teachers o f Junayd. It is im portant to point out at this stage the lack o f corroboration for the implication that the eighth century figures included in the Tabaqat were actually Sufis at all. For instance, Abu M uham m ad b. Qutayba (d.276/889) includes, in his w ork entitled al-M acarif a biography for only one o f these figures, namely Fudayl b. cIyad.46 W riting in the ninth century w ith a specific interest in diverse groups o f Muslims, Ibn Qutayba shows no knowledge o f the existence o f Sufism. In fact, he classifies Fudayl as a traditionalist ( cabid, calim).94

VI T he body o f the biography o f Junayd, like that o f Abu Yazld, is made up almost entirely o f utterances. They include examples o f each o f the types encountered in the biography o f Abu Yazld (isolated statements, question-answer constructs, short prayers, advice). O ne utterance attributed to Junayd was perceived in a dream by the first transm itter,95 according to its isnad:

26

Sulami’s Tabaqat al-Sufiyya

I heard Abu ’1-Hasan al-Q azwInl say that he heard Jacfar alKhuldT say, I saw Junayd in a dream (ra’aytu ’l-Junayd f t ’l-manam ), and I asked him, Are not the words o f the prophets ‘allusions’ (isharat) on the authority o f ‘witnessings’ (;mushahadat)?96 H e smiled and said, T he words o f the prophets are reports (naba’) on the authority o f presence (.hudur), and the words o f the sincere ones (.siddTqunf7 are ‘allusions’ on the authority o f ‘witnessings’. (TABS, 147.6-10)

This utterance is presented w ith an isnad as a significant segment o f material for the biography o f Junayd, regardless o f the fact that it is said to have been perceived in a dream. It is treated as being just as authoritative as an utterance witnessed in the (physical) presence o f Junayd. T he reason why this particular utterance is said to have been perceived in a dream, is certainly not, as is often the case, on account o f any tim e-gap betw een the lifetime o f the first transm itter (Khuldl) and that o f Junayd.98 It may be due then to the content o f Junayd’s utterance; the segment depicts him as answering a question about the degree o f authenticity o f the speech o f prophets, by pointing out that it is actually a degree higher than at first assumed by Khuldl. It implies that Junayd him self possesses the knowledge to confirm the relation betw een the prophets’ speech and w hat they describe. (It perhaps also serves to underline a conviction in the superiority o f prophets over Sufis through attribution o f such a com m ent to the highest ranking Sufi o f all).99 T he provision o f a dream context implies that such knowledge was attained by Junayd only after death, thereby avoiding its attribution to him while he was still alive, w hich could perhaps have been less acceptable. Finally, it should be m entioned that this utterance exemplifies com m on features amongst those attributed to Junayd, by virtue o f its function o f definition and its use o f technical Sufi terms. T he biography o f Junayd includes examples o f a couple o f segment belonging to categories that have not been encountered in the biography o f Abu Yazid. T he first o f these can be described as ‘epistolary’. T he length and careful com position o f this category o f segments suggest literary origins, as they are typically made up o f a series o f balanced, rhym ing clauses, and are often introduced as representing w ritten correspondence to ‘one o f his brethren’ (.kataba ’l-Junayd ila bacd ikhwanih ) .10() 27

Sulami’s Tabaqat al-Sufiyya

I heard Abu ’1-Hasan say that he heard Jacfar say, Junayd wrote to one o f his brethren [the following]: W hoever invites [people] to God whilst depending upon other than him, God puts him to the test and veils (hajaba) his remembrance from his heart whilst he performs it with his tongue. If he takes heed and separates from the one on w hom he depends, and returns to the one to w hom he invites, God will remove his trials and tribulation. But if he persists in his dependance (.sukun ), God will strip away {nazad) mercy for him from the hearts o f the people, and clothe him (albasah) in ‘the dress o f greed’. Then his claim upon them will increase despite the loss o f mercy from their hearts, and his life will become a failure, his death miserable, and his afterlife filled with regret. We take refuge in God from dependence on other than God. (TABS, 147.10—148.5)101

This segment suggests that written correspondence between Sufis was already a familiar activity. Its attribution to Junayd is in keeping with his image as a Sufi who had widespread authority.102 The text o f the segment reads like a sermon, warning o f the consequences of hypocrisy in one’s teaching; if one invites people to God, but depends oneself on someone other than God, then one will face tribulations. Junayd himself is thereby implicitly attributed with sincerity in his teaching, as well as the authority to advise other religious teachers.103 T he final category o f segment to be m entioned is that o f verse. T he Tabaqat as a w hole actually contains many segments belonging to this category, although there are no examples in the biography o f Abu Yazid, nor strictly speaking in the biography o f Junayd. The latter, however, does include a distich o f verse, w hich serves as the conclusion o f the following segment: I heard Mansur b. Abdullah say that he heard Abu cU m ar alAnmatl say that a man asked Junayd, O n account o f what does the lover feel sorrowful, with regards to his mystical moments (awqat). He replied, O n account o f the time o f an expansion (bast) which brings on a contraction (qabd), or the time o f an intimacy (uns) which brings on an estrangement (wahsha). T hen he composed (ansha’a) [the following], reciting: I had a drink which was becoming pure by the vision of you, The power of time spoilt it once it had become pure. (TABS, 149.8-150.2)

28

Sulamfs Tabaqat al-Sujiyya

T he distich o f verse serves to reinforce the them e o f the preceding utterance, namely the transience o f blissful experiences. Again one can observe the repeated use o f technical Sufi terms in this definitive response attributed to Junayd. It also implies that Junayd com posed the two hemistiches o f verse, enhancing his portrayal as an erudite Sufi authority, w ho produced w ritten compositions.

V II There are a num ber o f ways in w hich one can attem pt to gain an insight into the social context, or S itz im Leben , o f Sulamfs collection and com pilation o f the material o f his Tabaqat. R ather than following the m ethod o f accepting at face value the accounts o f later biographers o f SulamI, w hich are unlikely to be free from (inevitable) retrojection, a certain am ount o f clues about this process can be picked up from the Tabaqat itself. In this regard, both the way in w hich segments are introduced by their isnads, and the social settings that are described in their matns can be instructive. As explained above, most o f the segments are not provided w ith their ow n isnad, but w hen isnads are given, the vast m ajority o f them indicate oral transmission, usually by means o f the formula, samxtu fulan yaqill. A few segments in other parts o f the w ork specify that the context is that o f a study circle (halqa), such as the following, w hich is repeated three times in the biography o f Shibll: kuntu yawmanfThalqatih wa-samictuh yaqul . . . (O ne day I was in his circle w hen I heard him (Shibll) say . . .).104 T here are also a num ber o f m ore elaborate descriptions o f such situations, w hich often function to boost the standing o f their subjects in relation to the other participants o f a study circle. For example, the following com m ent is found in the introduction to the biography o f Abu cA m r al-Zajjajl: I heard my grandfather say, I was in M ecca while al-Kattam, al-N ahrajurl and al-M urtacish and other shaykhs (masha’ikh) were there. They were sitting in a circle (yaqcudun halqatan), the focus (sadr al-halqa) o f w hich was Abu cAmr. W henever they spoke about som ething they all consulted w hat Abu cA m r would say (raja'a ja m fu h u m ila ma yaqul Abu cA mr). (TABS, 44 9 .6 -1 0 )

29

Sulam i’s Tabaqat al-Sufiyya

This anecdote depicts Abu cA m r (ZajjajI) as a Sufi whose superior status to the others present, including the em inent authorities Kattani, Nahrajuri and M u rta'ish,105 is indicated by their deference to him in consulting his opinion. Abu cAm r is seated at the focus o f the circle, and this is evidently w here the leading authority o f the gathering would sit.106 It is also w orthw hile to underline that the above anecdote suggests that, whilst the m ost authoritative participant may have been treated w ith deference, the teaching situation was w hat we w ould term (approvingly) today as ‘interactive’ — m ore o f a ‘seminar discussion’ than a m onologue. Whilst the exact situation described in the above anecdote is most probably fictional, it offers a setting that was at least judged by SulamI to be credible enough for the purpose o f boosting Abu cA m r’s authority. It could therefore reflect the halqa context about which SulamI himself would have had direct knowledge as a participant, and from which he obtained his own collection o f segments. References in the Tabaqat to teaching circles or sessions (majlis, halqa)nn usually do not specify their actual location. In the few exceptions, a mosque, or even the main mosque o f a town, is said to have been the location.108 Shibll’s circle, for example, is said, on one occasion, to have been located in the main mosque o f M edina (kuntu waqifan cala halqat AbTBakr al-ShiblTftjam ic al-Madma . . .).109 Abu Hamza ’1-BaghdadI is also said to have spoken (takallama) in the main mosque o f M edina, where a change in his mystical state caused him to ‘fall from his seat’! ( takallama yawman f t ja m ic alM adina ja-taghayyara calayh haluh wa-saqata can kursiyyih).U{)

Although the suggestions that these individuals held sessions in the same im portant m osque prim arily serve the purpose o f boosting their authority, nonetheless they also indicate that the holding o f such sessions in mosques, sometimes perhaps even in the most prestigious mosques, was credible in Sulami’s lifetime. In the Abu cAm r anecdote m entioned above, the individuals w ho form the circle are themselves Sufi authorities (masha’ikh ); this is necessary for the sake o f fulfilling the particular functions o f that anecdote (i.e. boosting the status o f Abu cAmr). In other examples, however, the presence o f Sufis w ho were not all em inent shaykhs is implied. There are also indications that such sessions were not closed to outsiders. For instance, the transm itter o f the segment about Shibll’s session in the main mosque o f M edina describes him self as ‘an onlooker’ (kuntu waqifan cala halqat AbT Bakr al-Shiblt). This is perhaps to be expected in the (public) main 30

Sulann’s Tabaqat al-Sufiyya

mosque. O th er indications o f the openness o f meetings can be found, as for example in the statement that Shibll him self started out on the Sufi path at one o f Khayr al-Nassaj’s sessions (taba f t majlis Khayr al-Nassaj).]U Sulaml’s Tabaqat also contains a relatively small proportion o f segments w ith isnads indicating w ritten transmission. T he most com m on formula for this purpose is ra’aytu bi-khatt al-fulan (I saw

in the handw riting o f X ).112 For example, w ritten transmission from both Sulaml’s father and grandfather are indicated in this way, ra’aytu bi-khatt abt/ jaddt.U3 T he use o f khatt in such formulae specifies that the w ritten source is in the handw riting o f its author, thus adding m ore weight to its credibility, as the following introduction to a segment emphasises: wajadtu f t kitab jaddt AbT cUthman, bi-khatt yadih . . . (I found, in ‘the b o o k ’ o f my grandfather, Abu cU thm an, in his handw riting .. .).114 T he following link in an isnad is also instructive: sam ttu Ahm ad b. Sulayman al-Kafarshila’T qala wajadtu f t kitabi can Hatim al-Asamm

. . . (I heard Ahm ad b. Sulayman al-Kafarshlla’I say, I found [it] in my kitab , on the authority o f Hatim al-Asamm .. .).115 It seems to imply that Kafarshlla’I was unfamiliar w ith the contents o f his own kitab , so that it is unlikely that it could possibly be referring to a literary w ork that he had compiled. Since he simply came across the segment in his kitab , it is perhaps m ore likely that the kitab is instead a reference to a collection o f such segments, possibly in the form o f a notebook that could have been used to record segments as an aid to m em ory.116 T here are also a few isnads that indicate transmission by means o f licence to transmit (ijazatan). For instance, the first and second segments o f the body o f the biography o f M uham m ad b. Khafif are introduced using the formula akhbarana A bu Abdullah Muhammad b. K haftf ijazatan,u 7 whilst the following remark is found at the end o f this biography: kull hadhihi ’l-hikayat akhbaramha A bu Abdullah M uhammad b. K haftf rahimah Allah ijazatan li bi-khattih (Abu 'Abdullah M uham m ad b. Khafif, may G od have mercy on him, inform ed me about all o f these [preceding] segments by licence, in his ow n handw riting).118 This final statem ent may be a gloss, w ith the purpose o f attributing authority to all o f those segments in the biography w hich precede it, and are introduced merely by the simple conjunction wa-; that is, as far back as the aforem entioned initial two segments o f the biography w hich possess their own isnads. T he remark may nonetheless be significant as an indication 31

Sulami’s Tabaqat al-Sufiyya

that the w ritten transmission o f a group o f segments by ijaza was a current practice amongst Sufi scholars. T he relatively few indications o f w ritten transmission in the isnads o f segments in this w ork may be an indication that the expression o f oral transmission was preferred. T he specific references to ijdzas and ‘the handw riting’ o f those ascribed w ith w ritten sources are then perhaps a further indication o f the dubiousness w ith w hich w ritten transmissions may have been perceived. If the isnad claims that the source was in the handw riting o f the ‘author’, or transmitted w ith his blessing, then it could offer reassurance that it is not a corrupted copy made by som ebody else, or a second-hand transmission prone to textual adulteration and growth. SulamI him self suggests that the use o f a w ritten source w ithout licence is less credible than the other forms o f transmission (both oral and w ritten through licence) which involve some form o f direct encounter w ith the transm itter.119

VIII Anyone w ho reads closely an early w ork o f Islamic historiography, such as the Tabaqat al-Sufiyya , will soon notice that it contains num erous segments that are variants o f each other. It should perhaps be no surprise that variants should have been generated and transmitted, but the fact that they are included in the same literary w ork demands further consideration. Variants in the Tabaqat may be found in the same, as well as in different biographies. That is to say, segments that are variants o f each other may be attributed to the same, or to different individuals. T he biography o f Junayd itself contains the following two variant segments: 1 I heard Abu ’1-Hasan cAll b. M uham m ad al-Q azwInl say that he heard Abu ’1-Tayyib al-cAkkl say that he heard Jacfar al-Khuldl say that he heard Junayd say, T he opening o f every honourable door is [achieved by] the exertion o f effort (fath kull bdb sh a rf badhl al-majhud). (TABS, 147.4-6)

2 I heard cAbd al-W ahid b. Bakr say that he heard H am m am b. al-H arith say that he heard Junayd say, T he door to every pure and glorious knowledge is the exertion o f effort (bdb kull cilm nafis jalxl badhl al-majhud ), and the one 32

Sulami’s Tabaqat al-Sufiyya

w ho seeks G od by the exertion o f effort is not like the one w ho seeks him by the path o f liberality (tariq al-jud). (TABS,142.10-143.1)

Each o f the above segments is given on the authority o f its own specific isnad. Having ostensibly been obtained separately they have both been included as utterances by Junayd. If one assumes that the inclusion o f variants in the same biography would have been undesirable, then one might see this case as an oversight. This would also support the aforementioned suggestion that the entry o f segments into biographies may have been a collaborative process, rather than checked by a single individual. However, attention should be drawn to the fact that the variants have significant differences, which could have made each o f them worthy o f inclusion. Variant 2 is the more elaborate version; it specifies that ‘the door’ is that o f knowledge, and also includes an extra rhyming clause which serves to contrast the m ethod o f exerting effort (badhl almajhiid) against that o f ‘liberality’ (jud). Whilst the term ju d does not usually have negative connotations,120 in this context it is probably being employed polemically, and is presumably directed at a com peting path that attaches less importance to personal effort.121 T he existence o f these two variants highlights the fact that segments have had their ow n histories before being entered into the Tabaqat, during w hich they may already have been trans­ formed. In this specific example there are indications that the variants may have been generated as a result o f factors, such as the preference for conciseness, greater precision (bab kull cilm / kull bab), and polemical interests (badhl al-majhud/tariq al-jud). T he biography o f Abu Yazld, w hich is located in the first generation section, contains a segment that is a close variant o f segments found in two other biographies, namely those o f Abu Bakr al-Kattanl (fourth generation) and Abu ’l-cAbbas al-Dlnawarl (fifth generation). They thus place variants o f the same utterance in the m ouths o f individuals from three different generations: 1 And Abu Yazld said, God observed the hearts o f his saints (.awliya’ih), and amongst them were those w ho were not fit to bear gnosis in a pure form , so he preoccupied them w ith worship (man lam yakun yasluh li-haml al-macrifa sirfan fa-shaghalahum bi- }l- cibada). (TABS, 63 .1 1 -1 3 )

33

Sulanu’s Tabaqat al-Sufiyya

2 KattanT said, G od looked at some o f his worshippers ( cubaydih) and did not consider them fit for his gnosis (ahlan li-macrifatih), so he preoccupied them w ith his service (fa-shaghalahum bi-khidmatih) (TABS, 39 1 .3 -4 )

3 And I saw in the handw riting o f my father122 that Abu ’1cAbbas al-Dlnawarl said, God has worshippers ( cubbad) w hom he does not consider fit for his gnosis (lam yastaslihhum li-macrifatih), so he preoccupies them w ith his service (fa-shaghalahum bi-khidmatih ), and he also has worshippers w hom he does not consider fit for his service, so he disregards them (lam yastaslihhum li-khidmatih faahmalahum). (TABS, 50 3 .7 -9 )

Each o f the above three variant utterances informs o f the existence o f certain devotees o f G od ( cubbad, cubayd, awliya’) w ho are not deem ed fit for gnosis, and are therefore kept busy (shaghala) by G od w ith alternative preoccupations ( cibada, khidma). T h e im plication is that gnosis is reserved for an elite, whilst other devotees are engaged in inferior activities because o f their inadequacy. T he third example, w hich also happens to be provided w ith its ow n isnad, is the most elaborate version; it includes an extra statem ent at the end, w hich introduces a new category, even further dow n the hierarchy — those w ho are not even deem ed fit for G o d ’s service, and are therefore disregarded altogether. T he inclusion o f variant segments in the same biography, and, perhaps even m ore so in different biographies as utterances attributed to different individuals, highlights the hazards o f treating a w ork like the Tabaqat al-Sufiyya as a source o f historical facts. Since it is unlikely that the compilers themselves were unaware o f the existence o f such variants in biographical traditions, that it did not seem to cause misgivings implies that the value o f biographical material did not lie (merely) in its historicity, but rather the continued relevance o f its message. It has already been noted that many o f the utterances attributed to the subject o f a biography are presented in a question-answer construct. T he questions that are used for this purpose tend to be formulaic, and they can serve to highlight the topics o f interest

34

Sulam i’s Tabaqat al-Sufiyya

w hich were considered w orthy o f inclusion. For example, the biography o f Junayd contains two segments, in w hich the utterance attributed to him is presented as a response to the question ‘W h o is the gnostic?’ (man al-carif). 1 Junayd was asked, W h o is the gnostic (man al-carif )? Fie replied, O ne w ho is not distracted, neither by a glance nor an utterance. (TABS, 145.6-8)

2 O n being asked, W ho is the gnostic (man al- carif) , Junayd replied, T he one w ho speaks about your secret while you are silent. (TABS, 143.9-11)

These two segments offer different, although not m utually exclusive, responses to the same question: T he gnostic is not distracted (from God) by what he may see or hear, and he can also read other people’s minds. T he biography o f Abu Yazid also includes, amongst several utterances about the carif and ma'rifa , those that are presented as responses to questions, such as ‘W hat is the sign ( calama) o f the carifV ;u 3 ‘W hat are the degrees (darajat) o f the ca rifT ;124 ‘For w hat did they receive ma'rifaV ;125 and ‘By means o f what did you acquire m acrifa? u 6 R etu rn in g to the biography o f Junayd, it should be m entioned that it contains another instance w here two utterances are each presented as responses to the same formulaic question, namely ‘W h o m should I follow?’ (man ashabu?). O n this occasion the segments are actually juxtaposed: 1 Junayd said to a man w ho asked him, W h o m should I follow? T he one to w hom you are able to reveal w hat God knows about you (ma yaclamuh Allah mink). 2 H e was asked another time, W h o m should I follow? The one w ho is able to forget his belongings and carry out w hat is incum bent on him. (TABS, 148.9-12)

T he first o f the above variants attributed to Junayd is particularly significant in relation to the two biographical traditions being considered in this study, because the following variant o f it, employing the same conspicuously rare calima min verbal construct, is attributed to Abu Yazid in the Hilyat al-awliya}:

35

Sulam i’s Tabaqat al-Sujiyya

H. A man asked him (Abu Yazld), W h o m should I follow? H e said, The one from w hom you do not need to hide what G od knows about you (ma ya'lamuh Allah mink). (H, X , 3 8 .7 -8 )

W hilst this is admittedly a unique case w ith regard to the specific biographical traditions o f Abu Yazld and Junayd, nonetheless it demonstrates that close variant utterances have been attributed in contem porary works even to the two figures w ho in later tradition came to represent opposite poles o f Sufism.127 It serves as a further rem inder o f the pitfalls in treating works o f this genre as repositories o f facts about individuals belonging to earlier generations, and that they are m ore likely to represent a ‘m irror o f the environm ent’ in w hich they were com piled.128

IX T he most recurrent topic in the biographies o f both Abu Yazld and Junayd is gnosis (m acrifa).]29 T he same can also be said for most o f the biographies in Sulami’s Tabaqat. It suggests that m acrifa was the most popular topic w hen the w ork was compiled. Segments about this topic attributed to Sufis o f the past were preserved. Nonetheless, the biographical traditions o f Abu Yazld and Junayd appear to have already acquired certain distinctive characteristics. T he Abu YazTd tradition can be distinguished from that o f Junayd by the recurrence o f comparisons o f gnostics w ith other (implicitly inferior) types o f religious devotees, but, as the above variants demonstrate, this kind o f utterance is by no means exclusively attributed to Abu Yazld. T he Junayd tradition already includes epistolary segments, which, in com m on w ith the utterances attributed to him, frequently employ technical Sufi terminology. W hilst this distinction is perhaps fitting for som eone rem em bered as an erudite Sufi w ho had widespread authority, these types o f segments are neither attributed exclusively to Junayd. Thus, whilst individual segments from each o f these biographies would not look out o f place if they were filed into an alternative one, the accumulated assemblages as a whole, o f w hich they are com ponent parts, already begin to depict Abu Yazld and Junayd as examples o f distinct types o f mystics — the form er bold and provocative, the latter erudite and cautious. 36

Sularm’s Tabaqat al-Sufiyya

C hristopher M elchert has recently pointed out that the contents o f the biographies o f the earliest individuals to be included in the Tabaqat are distinct from the others, since they contain a greater emphasis on asceticism and piety than on mysticism and gnosis.130 M oreover, Johanes Pedersen had noticed that these biographies are also distinct by virtue o f their possession o f a higher proportion o f isnads that imply w ritten transmission, as well as lengthy narrative segments.131 All o f these characteristics can be interpreted as indications that the biographical traditions o f the earliest members — those living the furthest back from Sulaml’s ow n time — had already reached a relatively mature form. T he possibility o f recreating these established biographical portraits, such as by means o f backprojections, was restricted considerably by their familiarity. It is therefore perhaps to be expected that, o f all the biographies in the Tabaqat, they should be the ones that are distinct from the rest in this regard. M oreover, this distinction hardly represents sufficient grounds for assuming that, if SulamI did not rework the early biographies significantly himself, then they must be historically accurate. In fact, recent studies by B onner and C ooperson have already dem onstrated that, prior to the eleventh century, the biographies o f Ibrahim b. Adham and Bishr al-Hafi (respectively) offer contrastingly different portraits from those in Sulaml’s Tabaqat. Evidently, by the time o f Sulaml’s work, these biographical traditions had already reached a relatively mature stage, but only after having previously undergone processes o f reworking and developm ent.132 T he Tabaqat al-Sufiyya holds obvious attractions for those w ho are interested in reconstructing the early history o f Sufism. A lthough this is not the aim o f the present study, the examples provided above can serve as a note o f caution w ith regard to treating this w ork as a repository o f facts about the individuals it describes. W hilst one should not deny that there is likely to be a certain am ount o f historically accurate inform ation amongst the biographical segments that make up Sulamfs portraits, it should be pointed out that it was precisely on account o f their continued relevance in the judgem ent o f later generations that they had been preserved and transm itted (whilst presumably an indeterm inable am ount o f other segments would conversely have been lost in the process). M oreover, it is difficult to identify precisely the ‘authentic’ segments w ith any degree o f certainty, and even if some o f them m ight have been preserved in their original form 37

Sulam i’s Tabaqat al-Sufiyya

they are likely now to convey different messages, as shaped by the new contexts in w hich they have been embedded. Given that the authors o f tabaqat works did not necessarily share the same aims as m odern historians, it is perhaps safer to appreciate their functions in their present contexts than to speculate about their accuracy in describing earlier historical periods. T here are a num ber o f ways in w hich biographical segments about past Sufis could have been relevant for a contem porary audience or readership. For instance, they could serve as models o f ideal belief and conduct, for the purpose o f emulation. In this way, they could shape the definition o f Sufism for the contem porary com m unity by w hom they were used. Polemical segments, such as the ikhtilaf al-culama’ utterance attributed to Abu Yazld and the w ritten correspondence attributed to Junayd, w hich is aimed against hypocritical teachers, lend themselves most obviously to this function.133 Furtherm ore, biographical segments also serve to delight the listener/reader. For instance, the aforem entioned verse attributed to Junayd serves to reinforce, in an enjoyable and m emorable m anner, the message o f the m atter-of-fact utterance w hich precedes it.134 T he following pithy utterances attributed to Abu Yazld, w hich represent a com m on type o f segment found in Sulami’s biographies, also seem to be designed to give pleasure and to stimulate reflection on lofty issues: kufir ahl al-himma aslam min Tman ahl al-minna (The unbelief o f those w ith aspiration is sounder than the faith o f those w ho seek rewards!);135 ab'ad al-khalq min Allah aktharuhum isharatan ilayh (The furthest from God amongst m ankind are the ones w ho point to him the most!).136 In the preceding sections o f this chapter, reference has been made to the occasional incongruities in the overall framework o f the Tabaqat, as well as the ‘unsystematic’ order o f its biographical segments. In spite o f this, however, and especially in view o f the fact that its overall fram ework is predom inantly unified and systematic, it would seem justified to classify the Tabaqat al-Sufiyya as a w ork w hich bears the stamp o f the authorial control o f SulamT. (This should becom e all the m ore clear w hen its occasional incongruities are com pared w ith the large num ber evident in the other eleventh century Sufi tabaqat works, w hich constitute the subjects o f the next two chapters o f this study). In fact, the Tabaqat al-Sufiyya merits being counted as Sulami’s most im portant work, since, by virtue o f its position as the earliest 38

Sulam i’s Tabaqat al-Sufiyya

example o f its genre, it came to acquire a symbolic authority amongst later Sufis.137 Thus, its im portance lies in how it was perceived, regardless o f its historical accuracy Its unrivalled influence on later works o f the Sufi tabaqat genre is two-fold: in the first place, its clearly defined organisational framework has influenced the main examples o f the genre; and secondly, the form and content o f its biographies have determ ined the parameters for the later developm ent o f individual biographical traditions, w hich continued to accumulate material to supplem ent that w hich was derived from Sulami’s own portraits. In these ways, Sulami’s eleventh century re-reading o f the past o f Sufism, in the form o f the first w ork o f the Sufi tabaqat genre, shaped the corresponding re-readings o f his successors in Sufi historiography. T he influence that it has exerted on later tabaqat writings is illustrated especially in Chapters T hree and Four o f the present study. However, attention will be drawn first o f all to a w ork o f the same genre w hich is traditionally ascribed to a contem porary o f SulamI, namely Abü N u'aym al-Isfahànï.

39

This page intentionally left blank

Chapter

Two

The Hilyat al-awliya’

I

A

bu N u 'ay m Ahm ad b. 'Abdullah al-Isfahanl (d. 430/1038), the . assumed author o f the Hilyat al-awliya\ is rem em bered prim arily as a Shafi’ite hadith transmitter. Subkl m entions that about eighty people transm itted hadiths from h im .1 Later tradition reinforces Abu N u 'ay m ’s Shafi’ite credentials by associating him w ith Shafi’ite/H anbalite rivalries in his hom e town; he is even attributed w ith the miracle o f causing the mosque o f Isfahan to collapse and crush its congregation, in response to being expelled on account o f his loyalty to Shafi’ism.2 W hilst Abu N u'aym him self is not rem em bered as an im portant past authority in the later Sufi biographical tradition, his grand­ father (Abu 'Abdullah M uham m ad b. Yusuf) Ibn M a'dan alB anna’, is identified as the teacher o f 'All b. Sahl al-Isfahanl (d.307/920), w ho was the most celebrated Sufi from Isfahan.3 In the introduction to the Hilyat al-awliya’, Ibn M a'dan is m entioned as Abu N u 'ay m ’s forerunner in Sufi scholarship.4 M oreover the Hilya closes w ith the biographies o f recent generations o f Ibn M a'd an ’s students.5 This w ould suggest that the main influence on Abu N u'aym in the direction o f Sufism was the m em ory o f his grandfather, and association w ith his successors in Isfahan. T he ten-volum e tabaqat work, the Hilyat al-awliya’ wa-tabaqat alasfiya\ is the longest w ork ascribed to Abu N u'aym by a considerable margin. T he other m ajor works ascribed to him, namely the D hikr akhbar Isbahan and the D ala’il al-nubuwwa , also display an interest in biographies and hadith transmissions.6

41

T he Hilyat al-awliyâ*

II T he Hilyat al-awliyâ\ w hich is introduced as a Sufi tabaqât w ork,7 consists o f approximately 650 biographies, am ounting to ten volumes in the printed edition (approx. 4, 000 pages). A lthough it does not have a clearly-defined organisational fram ework com par­ able w ith Sulamî’s Tabaqât, a general chronological principle is evident in the order o f biographies, since they begin w ith the four ‘rightly-guided’ Caliphs (al-khulafâ’ al-râshidün) and culminate w ith Abü N u 'a y m ’s co ntem poraries.8 T h e tim e-span is covered predom inantly by individuals w ho are not usually identified as Sufis; these include the generations o f the salaf, the first six Imams o f Shi’ism, scholastic theologians, the founders o f major law schools and other celebrated pietists. These religious authorities are attributed w ith some Sufi utterances (amongst other material) in their respective biographies, even though they may not usually be rem em bered in this way.9 T heir inclusion can be attributed to Abü N u 'ay m ’s own interests in scholasticism and traditionalism. Whilst the overall arrangement o f the biographies in the Hilya is chronological, the individual biographies do not follow in a strict chronological order. Even though the dates o f the deaths o f the subjects o f biographies are seldom offered, there are a num ber o f glaring inconsistencies, such as w hen a student’s biography is presented many volumes before that o f his own teacher.10 The reason for such discrepancies appears to be the application o f a num ber o f competing organisational principles w ithout sufficient coordination. In addition to the chronological principle (and often in conflict with it), biographies can be found juxtaposed because o f the geographical origins o f their subjects,11 the relationships between their subjects,12 and the associations between their contents (e.g. com m on theme, motif, key-word).13 The conflict between these competing principles o f order suggests that they may not have been applied by a single author. This is further corroborated by the num ber o f instances where two separate biographies are provided about the same individual, w hich even overlap in co n ten t.14 Therefore, although the overall chronological progression o f the biographies is clear and comparable with other works o f the tabaqât genre, in that it structures a continuity from an authoritative point in the past to the generation immediately preceding the assumed author, a closer inspection reveals a relatively large am ount o f inconsistencies, suggesting the work o f ‘many hands’.15 42

T he H ilyat al-awliya}

T he Hilya includes, amongst its approximately 650 biographies, 76 out o f the 103 that make up Sulamfs Tabaqàt al-Sufiyya. M ost o f them (68) are found in the tenth volum e.16 In view o f the fact that the biographies w hich correspond to Sulamfs collection constitute only a small proportion (approximately a ninth) o f the total num ber in the Hilya, its omissions may be m ore significant. T he longest biography in the Tabaqàt for w hich there is no counterpart in the Hilya is that o f al-Husayn b. M ansur al-Hallaj. It seems remarkable that Hallàj should not be included in a w ork o f this scope and volume. It may be that he was considered too controversial, especially in the light o f the com m ents in the introduction o f the H ilya.17 Later tradition, perhaps by way o f explanation for this conspicuous omission, reports a dispute betw een Hallàj and the people o f Isfahan, w ho apparently sought to kill him after he quarrelled w ith cAlI b. Sahl al-Isfahànl.18 For the purpose o f the present study, the tenth volum e o f the Hilya, w hich contains most o f the biographies that correspond to those found in Sulamfs Tabaqàt, will be the focus o f attention. O n initial investigation, this volume appears to be made up o f distinct groups o f juxtaposed biographies, each preceded by its ow n ‘group introduction’.19 T he ‘group introduction’ ostensibly explains the reason why biographies belonging to the same particular group are juxtaposed. Each is given under the rubric qàla ’l-shaykh Abu N u caym (or variants o f this), suggesting ostensibly that they represent the com m ents o f Abu N u'aym himself, as recorded by a student. For example, at the end o f the biography o f Abu Yazld there is an introduction to ‘the suns o f the East and their em inent m em bers’ (qàla l-shaykh Abu N u caym rahimah Allàh ammà shumUs ahi al-mashriq wa-aclàmuhum fa - . . .);20 the subsequent six biogra­ phies are appropriately about Khurasanian Sufis. However, it should also be pointed out that they are in turn followed by biographies o f Sufis from Baghdad, such as H àrith al-M uhàsibi and Sari ’1-SaqatI, w ithout a further introduction to signal the start o f a group o f Baghdadians. This suggests that the group introductions do not represent the original framework o f the Hilya, for they are not provided consistently throughout the tenth volume, not to m ention the rem ainder o f the w ork.21 A nother group introduction in the tenth volume names Abu Sacld al-Khazzàz (usually Kharràz), specifically as a m em ber o f the group o f gnostics from Iraq whose biographies will follow (qàla ’l-shaykh dhikr ja m à ’a min jamahfr al-càrifm min al-cIràqiyyin . . . ka43

T he Hilyat al-awliya’

A b l S a id al-K hazzaz) 22 This introduction is followed, appro­

priately enough, by the biographies o f five Sufis from Iraq. However the biography o f Khazzaz him self is found only m uch later (introduced under his ’ism Ahm ad b. cIsa), after several intervening biographies o f individuals from Khurasan.23 T he first o f these Khurasanian biographies begins w ith a group introduction to ‘a group o f recent sages from the East’ (qala ’l-shaykh wa-min liukama’ al-mashriq min al-muta’akhkhinn ja m a ’a minhum . . .).24 Thus the following sequence is observed: Introduction to ‘gnostics from Iraq’, including Abu Sacld al-Khazzaz I Individuals from Iraq 1 Introduction to < ‘recent Eastern sages’ I Individuals from Khurasan (followed by others from Iraq) I Ahm ad b. cIsa (Abu Sacid al-Khazzaz) T he separation betw een the biography o f Khazzaz and the preceding reference to it in a group introduction is probably due to interpolation and growth in the text. Biographies about ‘recent Eastern sages’ may have been interpolated betw een his biography and those o f the other ‘gnostics from Iraq’, eventually generating their own group introduction. This w ould imply that the group introductions represent later redactional efforts to explain the order o f biographies in the work, rather than definitive signposts belonging to an original plan. They cannot therefore be relied upon at face value to account for the position o f each biography in the tenth volume, since they represent layers o f redaction rather than a static com pilation under the control o f an individual author.25

Ill T he biographies o f Abu Yazld and Junayd are both found in the tenth volume o f the Hilya. Each o f them is considerably longer than its counterpart in Sulaml’s Tabaqat. In the printed edition, the 44

T he Hilyat al-awliya’

biography o f Junayd is approximately 800 lines in total length, m aking it the second-longest in the tenth volume (after that o f MuhasibI), whilst the biography o f Abu Yazid is approximately 200 lines long. Despite the difference in length betw een these two biographies, both o f them , in com m on w ith the vast m ajority o f biographies in the Hilya , consist o f three distinct parts, w hich correspond to those o f the biographies in Sulam fs w o rk .26 However, these parts are invariably found in the alternative order of: the introduction, followed by the body, and finally the hadith transmission.27 Similar to the bodies o f the biographies in Sulamfs work, the body o f each biography in the Hilya is made up o f juxtaposed discrete segments, which are introduced by one o f a num ber o f different types o f introductions. M ost segments in the biographies o f both Abu Yazid and Junayd are introduced by their own isnad.28 However, out o f those segments that do not possess an isnad, more than twice as many are introduced by a simple conjunction (wa-/fa-), rather than a phrase that re-introduces a narrator (qala wa-, the subject being possibly Abu N u'aym or a transmitter in a preceding isnad). Therefore, as in the case o f Sulamfs Tabaqat, a significant proportion o f the segments are presented w ithout being linked unambiguously to a preceding isnad. T he m odern edition o f the Hilya employs a m ethod o f paragraphing w hich implies that all o f the segments w ithout an isnad, regardless o f w hich alternative type o f introduction they possess, are given on the authority o f the nearest isnad that precedes them . This is achieved by paragraphing at the point o f each isnad. However, just as it has been observed in Sulamfs Tabaqat, there is no secure basis in the text itself for this convenient interpretation.29 A passage from the body o f the biography o f Abu Yazid is presented below to illustrate the segmental form o f the text: 1 I heard al-Fadl b. Ja'far say that he heard M uham m ad b. M ansur say that cUbayd b. cAbd al-Q ahir said, A group o f people sat around Abu Yazid (jalasa qawm ila Abx Yazid). H e hung his head down for a while, then he raised it to them and said, W hile you have been sitting dow n before me, here I have been, roaming my thoughts, looking for a rotten grain w hich you can bear, to extract it for you, but I did not find [anything].30

45

T he H ilyat al-awliya’

2 H e said, Abu Yazld said, I was absent from G od for thirty

years. M y absence from him was [as a result of] my m entioning him (dhikri ), for w hen I refrained from it I found him in every state.31 3 A man said to m e,32 W hy do you not travel ? [Abu Yazld] said, Because my com panion does not travel and I am staying w ith him. T he questioner opposed him by means o f an analogy, saying, Ablution w ith still water is disliked! Abu Yazld responded, They did not see any fault w ith sea w ater; its w ater is clean and its dead things are permissible!33 T hen [Abu Yazld] said,You may see the rivers flowing w ith droning and m urm uring until they approach the sea. W hen they mix w ith it their m urm uring and turbulence subsides, and the sea water does not notice them , neither does an increase appear in it , nor would a decrease appear in it if they were to leave it. 4 cU m ar b. Ahm ad related to me that "Abdullah b. Ahmad related to him that Ahm ad b. M uham m ad related to him that cU thm an related to him on the authority o f Abu Musa, saying that Abu Yazld said, I did not cease (lam azal) for thirty years, w henever I wished to m ention (adhkur) God, to rinse and wash my tongue, deem ing him too high (ijlalan li-’llah) to m ention him. 5a "Uthman b. M uham m ad al-'U thm anl related to me that Abu ’1-Hasan al-RazI said that he heard Yusuf b. alHusayn say that he heard Yahya b. M u'adh say that he heard Abu Yazld al-Bastaml say, I did not cease (lam azal) to w ander in the field o f unity (maydan al-tawhid) until I departed to the enclosure o f unicity (dar al-tafnd). T h en I did not cease (lam azal) to wander inside the realm o f unicity until I departed to everlastingness (daymumiyya ), and I drank from his cup such a drink that, thanks to its m em ory (min dhikrih ), I will certainly not thirst ever again. 5b Yusuf said, I used to hear this speech in different words from D hu ’1-Nun, there being an addition to it; D hu ’1N u n w ould not utter it except w hen excited (nashat) and overwhelm ed by his state (ghalabat haluh calayh). H e used to say that, and follow it with, Yours is the glory and beauty, yours is the perfection. Glory be to you, glory be

46

T he H ilyat al-awliya’

to you. May the tongues o f praises and the m ouths o f glorification venerate you. You, you, eternal, eternal. His love for me is eternal. 6. Abu ’1-Fadl Ahm ad b. Abi cIm ran related to m e that Musa related to him that M ansur b. "Abdullah related to him, saying that he heard Abu cImran Musa b. cIsa say that he heard his father say that Abu Yazld said, I was absent from G od for thirty years. M y absence from him was [a result of] my m entioning him, for w hen I refrained from it I found him in every state, until it was as if he was me (hatta ka-annah ana).34 (H, X , 3 4 .18-35.17)

T he above passage can be divided into six discrete segments, as shown. Segments 2 and 3 are the only ones w ithout an isnad , being introduced by qala wa- and wa-, respectively. These juxtaposed segments also appear to have a them atic association w ith each other; the utterance attributed to Abu Yazld in Segment 2 implies that refraining from the activity o f m entioning G od (dhikr) had overcome absence from him, whilst the utterance in Segment 3 asserts that refraining from the activity o f travel is preferable w ith regards to achieving proxim ity to God. T he them atic association betw een the two utterances seems to have caused the use o f an inappropriate pronoun at the beginning o f Segment 3 (IT instead o f lah); it is most likely a scribal error due to the force o f the utterance in Segm ent 2, w hich im mediately precedes it.35 T hat is to say, Segm ent 3 may in fact have been interpreted as a continuation o f the direct speech o f Abu Yazld in Segm ent 2, and, as a consequence, the pronoun may have been changed (short­ sightedly) to the first person in order to facilitate assimilation. Segments 2 and 6 in the above passage are close variants. T heir wording is almost identical, the only significant difference being the extra phrase at the end o f Segment 6 (ka-annah and , ‘as if he was m e’), describing the perception o f identity betw een the speaker (Abu Yazld) and God. There is an allusion to such an experience also in the allegory w hich forms Abu Yazld’s final response in Segment 3 (.. . and the sea water does not notice them , nor does there appear in it an increase, nor w ould there appear a decrease in it if they were to leave it). O ne m ight therefore have anticipated the juxtaposition o f Segments 3 and 6, on account o f this particular association. 47

T he Hilyat al-awliya’

T he segments that separate Segments 3 and 6 do not appear to do so merely by coincidence, or as a result o f haphazard ordering, since patterns o f association are also discernible betw een them. Segment 4 depicts Abu Yazld as declaring his persistence (for thirty years again) in the pious habit o f washing his tongue out w henever he w anted to m ention G od’s name. It is perhaps associated w ith Segment 2 through the key-w ord dhikr. Segment 5 can be divided into two parts, the second part being introduced w ith a reference back to the first part by its introduction (kuntu asm au hadha ’l-kalam). T he first part (5a) consists o f an utterance attributed to Abu Yazld, describing his progression towards ever closer proxim ity to God (maydan, dar), and implying the eventual attainm ent o f some form o f union and perm anence (tawhid, tafrid, daymumiyya). H e reaches a climax w here he drinks such a drink (shurba) from ‘G od’s cup’ (ka’suh) that he will never thirst again, simply because o f the m em ory (dhikr) he retains o f it. This is clearly related to the other segments in the passage w hich tell o f a climactic experience o f union. Segment 5a is also associated w ith Segment 4 by the repeated use o f the phrase lam yazal , w ith w hich both o f these segments begin. This is unlikely to be merely a coincidence, for no other segments in the entire biography begin w ith the same construct.36 The above analysis demonstrates thematic and key-word/phrase associations between juxtaposed segments, with or without isnads. Each segment qualifies by association, in some measure, to be included in its position, suggesting that their juxtaposition is not coincidental. However, in comparison, Segments 2, 3 and 6 are the most closely related. The first two o f these are juxtaposed, and seem to have been considered as a continuous segment by at least one scribe. However, Segment 6 is separated from them by two segments, which themselves are not unrelated to their neighbours. The separation may be due to the interpolation o f these segments (4 and 5), which are also, albeit to a lesser degree, associated with Segments 2 and 3, as well as with each other. It may also be the case that the two variants (Segments 2 and 6) were added to the text at different times, the interpolator o f the later one to be added, being either unaware, or unconcerned, that a variant had already been included. In any case, this representative passage seems to have been produced by layers o f entries o f segments, rather than a static compilation prepared at one time by a single individual. In this way, it corresponds to the ‘multi-layered’ organisational framework o f the Hilya. 48

T h e H ilyat al-awliya’

T he inclusion o f Segment 5b, w hich informs o f the existence o f a variant o f Segm ent 5a, but attributed to D h ü ’l-N ü n , is w orthy o f consideration. It seems to function not merely to inform o f a variant, but also thereby to provide support for the validity o f the utterance attributed to Abü Yazld, by attributing a variant o f it to a m ore authoritative individual. T he provision o f this kind o f corroboration occurs once again in his biography (in the hadith transmission).37 However, other segments that appear equally, if not more, in need o f this kind o f corroboration also are included in the biography w ithout any such corroboration.38 This is perhaps yet another indication o f inconsistency in the com pilation, reinforcing the impression o f it as a product o f ‘many hands’.

IV T he passage from the biography o f Abü Yazld w hich is presented above, at the beginning o f Section III, includes two variants o f the same utterance (Segments 2 and 6), both o f w hich are attributed to Abü Yazld, as well as a a variant o f a different utterance w hich is attributed to D hü ’1-Nun (Segment 5b).39 T he biography o f Junayd also contains a num ber o f such variants, and they can be identified amongst its narrative segments as well as its utterance segments.40 T he passage presented above in Section III contains examples o f the two main categories into w hich all the segments in the body o f the biography o f Abü Yazld can be classified: utterances (Segments 2, 4, 5, 6) and narratives (Segments 1, 3).41 These are the same two categories o f segments that are offered in the corresponding biography o f Sulaml’s Tabaqât. T he biography o f Junayd, on the other hand, contains additional categories o f segments, such as epistolary and verse segments.42 Amongst the utterances attributed to Junayd in the Hilya are three examples o f a form o f utterance that is not represented at all in its earlier counterpart, namely ‘lengthy prayers’.43 W hilst the latter differ from the short prayers found in Sulaml’s Tabaqât because o f their length (20-71 lines), they consist o f a series o f the same forms o f exclamations in praise o f God and appeals to G od for help (e.g. Ilâhï laysa f ï ufuq samawâtik wa-lâ f ï qarâr ardik fïfu sh â t aqâlïmihâ man yuhibb an yahmada ghayrak . . . ‘God, there is not in the horizon o f your heavens, nor on the firmness o f your earth, in 49

T he H ilyat al-awliya’

the vastness o f its climes, anyone w ho wants to praise som ething other than you . . .).44 T he lengthy prayers in this biography are each introduced as having been recited aloud by Junayd (sam itu 7Junayd b. Muhammad yaqul wa-huwa yadcu bi-hadha ’l-d u a ’).45 In one case, a man comes to him to complain about depression, as a response to w hich Junayd teaches him the prayer (fa-ja’ah rajul fa shaka ilayh al-dayq fa - allamah wa-qala qul..).46 This com m ent suggests an explanation for the recording and preservation o f prayers such as these; the prayers attributed to an illustrious Sufi, like Junayd, may have been recorded in w riting, m em orised and used in o n e’s own personal worship. This factor could also have influenced the developm ent o f lengthy, elaborate and poetic prayers. Fifteen o f the segments in the body o f the biography o f Junayd are epistolary.47 They possess the same characteristics as the sole example o f this category presented in Sulaml’s biography o f Junayd, which, in fact, is also amongst those included here in the Hilya;4H they are made up o f rhym ing balanced clauses, and they offer advice and instruction. Since their average length is tw entyfive lines, w hen com bined they account for almost half o f the total length o f the biography.49 T hree o f them are juxtaposed, covering a total o f approxim ately 100 lines b etw een th e m .50 This conspicuous series o f segments suggests that juxtaposition accord­ ing to segm ent-category may have been a further organising principle that was applied during the developm ent o f the text. (They may also have been stored or transmitted together before compilation, despite their conflicting isnads). Fourteen o f the segments in this biography are narratives. Included amongst them are autobiographical narratives about encounters betw een Junayd and each o f his nam ed teachers, nam ely H arith al-M uhasibi and Sari ’1-SaqatI.51 T hey offer descriptions o f contrasting student-teacher relationships. In the first o f the two juxtaposed narratives, w hich form the following continuous passage, Junayd is depicted favourably in relation to Saqatl, w ho was also his m aternal uncle: 1 Ja'far b. M uham m ad inform ed us in som ething that he w rote52 and M uham m ad b. Ibrahim related to me, saying that he heard Abu ’1-Qasim al-Junayd b. M uham m ad say, O ne day I entered before Sari ’1-SaqatI and saw that he was unsettled. I said, Shaykh, I see you are unsettled. Fie said,

50

T he H ilyat al-awliya’

Just now there was a knock on my door and I said, Enter. A youth (shabb), w ho looked like he was about the right age for discipleship, entered before me and asked me about the m eaning o f tawba (repentance), so I inform ed him. H e then asked m e about the condition o f tawba so I told him. T h en he said, That is the m eaning o f tawba and this is its condition, so w hat is its essence (haqTqa)? I said, The essence o f repentance for you is that you do not forget the reason for repentance. H e said, It is not like that for us. So I asked him, W hat is the essence o f repentance for you then? H e said, T he essence o f repentance is that you do not rem em ber the reason for repentance. I am now contem plating his words. Junayd said, I then responded, H ow beautiful are his words! T h en [Saqati] said to me, Junayd, w hat is the m eaning o f this? I said, Teacher (ustadh), if I am w ith you in the state o f harshness (hal alja fa ’) and you transfer me from that state to the state o f purity (hal al-safa’) then my rem em brance o f harshness whilst being in this state o f purity w ould be negligence. 2 I entered before him on another day (yawman akhir) and saw him unsettled, so I said, Shaykh, I notice that you are anxious. H e said, Yesterday I was in the main mosque (ja m i ) w hen a youth (shabb) stopped before me and said to me, Shaykh the slave does know that G od has accepted him! I said, H e does not know. H e said, O h yes he does. And he said once again, O h yes he does. I said to him, H ow does he know? H e said, If I see that G od has held me back from every act o f disobedience (ma siya) and made m e conform to every act o f obedience (ta a), then I know that G od has accepted me. (H, X , 274, 9 -1 5 )

T he two similar narratives, w hich form the above passage, are linked together by the phrase ‘on another day’ (yawman akhir). They are therefore not presented as discrete segments, although they appear to be variants o f each other, but rather as a continuous narration by Junayd. T he first narrative (1) serves to depict Junayd as a precocious student w ho appreciates and explains cleverly the wisdom o f the youth (shabb)53 to his ow n teacher (shaykh, ustadh), Saqati. T he reader may anticipate a similar evaluation o f the answer o f the youth by Junayd in the second narrative, but none is given. 51

T he H ilya t al-aw liya’

T he latter does not therefore function to attribute directly any positive qualities to Junayd himself. However, it again describes Saqatl as being outsm arted by a youth, and the juxtaposition influences the reader to associate Junayd once again w ith the latter, in preference to his own teacher. These narratives therefore depict Junayd as being wiser than Saqatl by means o f the third character o f the shabb , thereby conveniently avoiding any suggestion o f conflict betw een the pair themselves. It may be the case that the narratives in the above passage are polemical in origin, w ith the aim o f denigrating Saqatl, for he does not him self recognise the wisdom o f the youth in either o f them. In view o f that, it is perhaps no coincidence that M uhasibi is portrayed positively in all o f the segments about him in this biography. T he following pair o f autobiographical narratives about Junayd’s encounters w ith H arith al-M uhasibl parallel in a num ber o f ways the above pair involving Saqatl: 1 I heard Abu ’1-Hasan Ahmad b. M uham m ad b. Miqsam say that he heard Abu Bakr al-Khawwas say that he heard al-Junayd b. M uham m ad say, A l-H arith b. Asad alM uhasibl used to come to our house and say, C om e out for a stroll w ith me. I would say to him, You [want to] take me away from my seclusion and security against my carnal soul (nafs) to the streets, the realm o f wretchedness and lustful sights! H e w ould say, C om e out w ith me and d o n ’t be scared! So I w ould go out w ith him, and it w ould be as if the street were em pty o f everything —we w ould not see anything to dislike! (H, X, 255.18-22)

2 I used to say often to H arith, [W hat about] my seclusion and my intimacy? You [want to] take me out to the vileness o f the vision o f the people and the streets! H e would say to me, H ow often will you say, [W hat about] my intimacy and my seclusion ? If half o f m ankind were to approach me I w ould not find any intimacy w ith them, and if the other half kept away from me I w ould not feel lonely because o f their distance! (H, X, 255.25-256.3)

52

T h e H ilya t al-aw liya’

T he above variants are both narrated autobiographically, and in the past continuous tense; these encounters are being related as habitual occurrences, rather than specific events. T hey are separated from each other by only one intervening segment, w hich itself is also an autobiographical narrative in the past continuous tense about Junayd’s experiences w ith M uhasibl.54 This suggests that they were juxtaposed because all three are segments o f the same type. In both o f the above variants Junayd describes his complaints to M uhasibl for urging him out o f his seclusion. In Variant 1, Junayd discovers the correctness o f M uhasibi’s approach by experiencing (perhaps m iraculously) that his fear o f distraction by the ‘wretchedness and lustful sights’ (afat, ru’yat al-shahawat) was unwarranted, whilst in Variant 2, M uhasibl rebukes him for his fear, declaring that he him self is not affected by exposure to m ankind. They both function to dem onstrate the latter’s superior wisdom, and are therefore perhaps unlikely selections for the biography o f Junayd, w ho is portrayed as som eone w ith still a lot to learn.55 T heir portrayal o f M uhasibl in relation to Junayd contrasts sharply w ith the aforem entioned autobiographical narratives about Junayd’s encounters w ith Saqatl, his other nam ed teacher. B oth pairs o f segments may be part o f the same polemical effort to prom ote the authority o f Muhasibl. T he other narrative segments in the biography o f Junayd include the following two variant segments, w hich each narrate an encounter w ith Abu Bakr al-Shibll, w ho is often listed as one o f Junayd’s students:56 1 Ja'far inform ed us and cU thm an also told me about it, saying, I was walking w ith Junayd w hen Shibll m et him and said to him, Abu ’1-Qasim, what do you have to say about som eone for w hom G od (al-haqq) suffices as attribute (na t), knowledge (cilm) and existence (w ujud )? H e responded, Abu Bakr, divinity is lofty and lordship is mighty; there are a thousand levels betw een the em inent o f the generation and you, at the first level o f w hich o n e’s identity disappears. (H, X, 267.4-7)

2 H e said, Shibll stopped before [Junayd] and said, W hat do you have to say, Abu ’1-Qasim, about som eone whose existence (wujud) is a truth (haqTqa), and not [merely] in 53

T h e H ilya t al-aw liya’

theory (Ulman ) ? H e responded, Abu Bakr, betw een the em inent people and you are seventy steps, the lowest o f w hich is that you forget yourself. (H, X, 270.19-21)

Each o f these variants describes Shibll confronting Junayd w ith a question w hich is implicitly boastful. H e is rebuffed by the latter in both cases. T he questions posed by Shibll in the two variants are form ed using similar vocabulary (yd A b a ’l-Q asim ma taqulfT man . . . , w ujud.. haqq/haqTqa.. cilm). T he rebuttals are even m ore alike. This is probably not merely a coincidence, but rather on account o f a shared history. Narratives o f this type describing encounters betw een two Sufis, one o f w hom is boastful whilst the other rebuffs his claims, are far from uncom m on. Even the biography o f Abu Yazid contains an exam ple (presented below), despite the fact that his ow n biographical tradition itself contains num erous examples o f his own boasts. I heard M uham m ad b. al-Husayn say that he heard M ansur say that he heard Abu Yacqub al-N ahrajurl say that he heard cAli b. "Ubayd al-Sahmdanl say that Yahya b. M u'adh wrote to Abu Yazid, I became intoxicated (sakartu) by the volume that I drank from the cup o f his love (ka ’s mahabbatih). Abu Yazid wrote to him in his reply, You became intoxicated and w hat you drank were mere drops! [Meanwhile] som eone else (ghayri)57 has drunk the oceans o f the heavens and the earth and his thirst has still not been quenched; his tongue is hanging dow n from thirst and he is asking, Is there any more? (H, X, 40.10-14)

A lthough this narrative may differ ostensibly from the variants describing encounters betw een Junayd and Shibll, it is actually parallel to them; they each describe the subject o f the biography (Abu Yazid / Junayd) as responding dismissively to the boasts o f another Sufi (Ibn M u'adh / Shibll), by suggesting that the latter has not reached the high level that he claims.They appear to be anecdotal topoi, w ith the function o f affirming the discerning authority o f their subjects.58 In other words, our examples function to attribute to Abu Yazid and Junayd respectively, the knowledge that is required for discrimination betw een genuine experiences and the claims o f pretenders. 54

T he H ilya t al-aw liya’

There are two narratives in the biography o f Junayd, each o f w hich describes the m anner o f his death. They give alternative accounts, even though both are presented as eye-witness reports. 1 I heard M uham m ad b. al-Husayn say that he heard Abu "Abdullah al-D arim l say that he heard Abu Bakr al-"Atawi say, I was w ith Junayd w hen he died; he recited the w hole o f the Q u r’an, then, starting from Surat al-Baqara , he recited seventy verses and died. (H, X, 264.1-3)

2 I heard "Abd al-M un'im b. "Umar say that he heard Abu Sa"ld b. al-A"rabI say that he heard Abu Bakr al-"Attar say, I was present w ith Junayd, Abu ’1-Qasim, on his death, amongst a group o f our companions. H e said, [Junayd] was praying whilst sitting and he w ould bend his legs w hen he w anted to prostrate himself. H e continued like this until the life w ent out o f them . It became difficult for him to move his legs, so he stretched them out . O ne o f his friends amongst those w ho were present at that time he was called ‘al-Bassaml’ - saw him [in that condition]. Abu ’1-Qasim’s legs had swollen, so he asked, Abu ’1Qasim, w hat is this? H e replied, These are G od’s blessings, God is the greatest. W hen he com pleted his prayer, Abu M uham m ad Jurayri said to him, Abu ’1-Qasim, if only you w ould lie down. So he responded, Abu M uham m ad, this is the time o f G od’s kindness. G od is the greatest.59 That rem ained his state until he died. (H, X, 281.11-18)

These different accounts o f the same event underline the hazards in using biographical segments as sources for historical events. A lthough it is know n in this case that the event in question (the death o f Junayd) must have happened in some fashion, and both o f the segments that report it seem plausible, there is no reason to accept one over the other. However, at the same time, it is apparent that both versions have the same function, namely the attribution o f a pious death to Junayd. This is perhaps where their significance lies: their function, rather than their historicity. N either o f them can be accepted as a historical account, but each o f them serves to attribute a pious death to Junayd, w ho is portrayed as persevering w ith acts o f worship until the final mom ents o f his life. 55

T h e H ilyat al-aw liya*

Version 2, quoted above, w hich is appropriately the final segment o f the body before the hadith transmission, is more com plex than its rival account. W hereas Version 1 is a b rief report o f a pious death, after reciting the Q u r’an, Version 2, by offering details about Junayd’s swollen legs, emphasises his persistence in pious worship, despite the obstacles one m ight expect on the brink o f death. It also names three people present w hen Junayd died: Abu Bakr al-cAttar (the narrator), ‘al-Bassaml’ and JurayrT. Only the latter, JurayrT, is rem em bered in other sources as a close associate o f Junayd; he is in fact described as Junayd’s successor.60 To my knowledge, the only individual w ith a name similar to ‘alBassaml’, w ho is associated w ith Junayd in tradition, is actually Abu Yazld. C ould it be a misspelling o f his nisba ? Abu Yazld, however, is associated w ith Junayd only by virtue o f the latter’s com m entaries on his utterances,61 and, in any case, he is believed to have died at least twenty-six years before him. C ould the provision o f this name perhaps be an error then, made in the attem pt to name individuals associated w ith Junayd as being witnesses to his death? T he closest I have found to an ‘Abu Bakr alcA ttar’ w ho is associated w ith Junayd is the narrator o f our Version 1 [of his death], ‘Abu Bakr al-cAtawi’. C ould it be that som eone w ith such a name was com m only rem em bered as having been present at Junayd’s death, thus generating the inclusion o f these names in each narrative’s account? T he H ilya has long been regarded by historians as a rich source o f narratives, since it usually contains a far greater quantity for each biography than other contem porary works, not to m ention the fact that it contains many biographies that are simply not found elsewhere before the eleventh century However, rather than assuming that it can serve as a source o f neutral data for historical reconstruction, it is w orthw hile to take advantage o f the length o f the w ork to observe such factors as the frequency o f variants and anecdotal topoi, as well as the logistical problems posed by the existence o f com peting accounts o f the same event. It w ould be short-sighted to assume that the contents o f these biographies constitute an accurate reflection o f historical reality, and, more significantly, it w ould encourage a superficial interpretation o f biographical works, one w hich w ould fail to do justice to the dynamic traditions that produced them.

56

T he H ilya t al-aw liya}

V T he introductions o f the biographies o f Abu Yazld and Junayd are typical o f such introductions in the Hilya. T he form er is presented below as a representative example: T he shaykh and hafiz, Abu N u'aym , may G od have mercy upon him, said (qala ’l-shaykh al-hafiz . . .), Amongst them is the lone roamer, the solitary w anderer (ta’ih wahTd, h a ’im farid ), al-Bastaml Abu Yazld. H e roamed and withdrew, he w andered and then returned. H e w ithdrew beyond the limits (;mahdudat ) to the originator o f perceptibles (mahsusat) and non-existents (ma dumat). H e separated him self from crea­ tion. H e was consistent w ith the truth. H e was helped by secret retreats, and reinforced by the mastering o f piety. His allusions are plain and his expressions hidden; to those w ho understand them they give security, to those w ho reject them they are a tem ptation (isharatuh h a ’ina wa- ibaratuh kam ina licdrifThim damina w a-li-m unkirihim fa tin a ) . (H, X, 33.20-34.1)

This introduction is presented under the rubric qala ’l-shaykh , in com m on w ith the other statements that have a demarcating fu n ctio n th ro u g h o u t the ten -v o lu m e w ork. It provides a characterisation o f Abu Yazld in rhym ing prose (sa f), and it contrasts w ith its counterpart in Sulaml’s Tabaqat a l-Sufiyya , by offering no ‘factual’ details, such as the full name, origins and date o f death o f the subject.62 T he introduction corresponds w ith m uch o f the contents o f the body o f the biography w hich follows it, by acknow ledging that utterances attributed to Abu Yazld are controversial, and it defends these utterances implicitly, by contrasting the judgem ent o f ‘those w ho reject th em ’ (munkiruha) w ith ‘those w ho understand th em ’ (carifuha ), thus im plying that the rejectors fail to understand them. T he introduction o f the biography o f Junayd similarly refers to its subject in s a j\ and in a way that corresponds to impressions given by the segments in the body o f the biography. For instance, it begins w ith the following balanced rhym ing clauses: al-murabba bifu n u n a l-ilm , a l-m u ’ayyad bi-cuyun al-hilm (trained in the range o f sciences, supported by the gems o f gentle persuasion).63 T he epistles, verse and lengthy prayers, in particular, offer an image o f Junayd w hich conforms to these characteristics.64 57

T he H ilya t al-aw liyay

As already m entioned, in the H ilya the hadith transmission is normally presented at the end o f a biography, that is, after its main body T he hadith transmission o f the biography o f Abu Yazld is representative o f this convention.6^ However, an alternative transmission o f the same hadith (without the inclusion o f Abu Yazld in its isnad) is also presented here. Perhaps Abu Yazld’s controversial reputation, and especially the view o f ‘those w ho oppose h im ’, m eant that the hadith transm ission required corroboration o f this kind. The unusual presentation o f an alternative isnad , one w ithout Abu Yazld, w ould therefore function in the same way as the aforem entioned variant o f a controversial utterance w hich is attributed to the m ore authoritative D hu ’1N u n .66 T he hadith transmission in the biography o f Abu Yazld is preceded by the following statement attributed to Abu N u caym, w hich serves as a dem arcator betw een it and the preceding body o f the biography: T he shaykh said (qala ’l-shaykh . . .), We restricted ourselves to this am ount o f his words because o f the deep allusions (isharat camTqa) they contain, whose depths no-one reaches except one w ho has dived into his sea, has drunk from the pure waves o f his breast, and understood the utterances o f his soul, w hich are produced and spread from his intoxication (.sukr ). As for hadith transmission (riwdya ) from him . . . (H, X, 41.6-8)

M ain body/hadith transmission demarcators such as the one presented above are seldom given, but w hen they are, they often suggest that the am ount o f material included in the body o f the biography has been deliberately restricted; this is not peculiar to the biography o f Abu Yazld.67 W hereas this example consists o f com m ents solely about the (preceding) body o f the biography, other such demarcators remark also about the subsequent hadith transmission. For instance, the one in the biography o f Junayd describes him as having mastered jurisprudence (ahkama cala 7sharfca), whilst also adding that his engagem ent in fulfilling ‘the truths’ o f transmissions prevented him from transm itting many him self (;wa-kana ’l-qiyam bi-haqa’iq al-athar yadfacuh can al-riwaya w a -’l-athar).68 T he above dem arcator in the biography o f Abu Yazld is particularly significant because o f its claim that his utterances are 58

T he H ilya t al-aw liya1

generated ultimately by his intoxication (.sukr ). A similar explana­ tion presented in the body o f this very same biography has been discussed already In that instance it is offered about D hu ’1-Nun, w ho utters a variant o f the same (controversial) claim about a mystical experience as Abu Yazld, but only w hen overwhelm ed by ecstasy.69 T he only other instance w here intoxication (sukr) is m entioned in this biography is in the segment about w ritten correspondence betw een Abu Yazld and Yahya b. M u 'a d h .70 Again, it is not Abu Yazld w ho is described as being intoxicated there, but, on the contrary, he rebukes Ibn M u'adh dismissively for becom ing intoxicated too easily, and im agining that he has reached lofty heights. Since the attribution o f intoxication to Abu Yazld in the main body/hadith transmission dem arcator conflicts w ith the references to intoxication that are found in the actual body o f his biography, its role is perhaps not so m uch to summarise accurately his depiction provided there, but rather to offer a general excuse for the controversial utterances attributed to him — one w hich reflects the association o f his name w ith mystical intoxication at the time o f the later redaction o f the H ilya w hen such demarcators were applied.

VI In Section II o f this chapter, it was observed that the qala ’l-shaykh demarcatory statements in the form o f ‘group introductions’ represent a late redaction o f the text, rather than its original plan. It was also suggested that the text appears to have grown with interpolation even after this particular redaction. Further support for this interpretation can be found in the overall structure o f the biography o f Junayd, for its hadith transmission (which has already been noted above as being introduced by a qala ’l-shaykh statement) does not mark the end as one would expect. Instead, it is followed by further segments o f material, which form a resumption o f the main body.71 Six segments in total are included in this resumption, each o f which is provided with an isnad. T heir isnads name, in total, three different immediate sources, and they include three different categories o f segments, in the following order: 1) lengthy prayer - 2) lengthy prayer - 3) epistle - 4) verse 5) verse - 6) lengthy prayer

59

T he H ilya t al-aw liya’

Lengthy prayers represent the only category w hich is found exclusively here in the resum ption o f the main body o f the biography o f Junayd, for several further examples o f epistolic and verse segments are included before the hadith transmission. W hilst it may be possible that the lengthy prayers were grouped intentionally after the hadith transmission, as a form o f appendix, this w ould be inconsistent in the w ider context o f the H ilya , and it would still fall short o f explaining why the remaining three segments should have been placed apart from the other examples o f their categories present in the main body o f the biography. T he fact that the final segment o f the main body before the hadith transmission is an account o f Junayd’s death only underlines that it once signalled its conclusion.72 T he provision o f a resum ption o f the body o f the biography, consisting o f segments w hich are wholly appropriate for inclusion in the main body o f the biography, therefore suggests that they were added after an initial redaction. T he body o f the biography may have becom e too familiar, and therefore closed to further interpolation, so that new segments could only be added after the hadith transmission, rather than at an appropriate point w ithin the body. Alternatively, they may have been tagged on at the end simply due to a lack o f consideration for the regular format o f biographies in the Hilya. A similar observation is made by David Daube in the study o f R om an and Biblical codes o f law.73 H e points out that the order o f rules in such codes often appears to be ‘illogical’. Whereas one would anticipate new rules to be inserted into a code in accordance with their contents, between the old provisions, it is often the case that they are added on at the end. He argues that this is probably due to the fact that the code to which they were added had become too well established. It could no longer be altered internally so there was no option but to add the new rules at the end. Whereas a biography is admittedly a very different context to a code o f law, it may still be the case that the segments found after the hadith transmission o f the biography o f Junayd (including the three examples o f lengthy prayers) were positioned there for similar reasons. Although one would expect to find these segments in the main body o f the biography, before the hadith transmission, if they are not included there this may be because it had already reached a fixed form, concluding appropriately with a narrative about the subject’s death. It had (eventually) become too well recognised to be tampered with; further segments thus accumulated as an appendix. 60

T he H ilya t al-aw liya’

VII All o f the isnads that are provided in the biography o f Abu Yazid and nearly three quarters o f those in the biography o f Junayd, suggest that segments were obtained through oral transmission (e.g. haddathana, sam i'tu . . . yaqul). Such isnads in themselves offer few clues regarding the S itz im leben in w hich this oral transmission took place. However, the narratives found in the two biographies offer some descriptions o f the contexts in w hich oral transmission could have taken place. They describe situations that were at least familiar and credible to the producers o f the material, and were perhaps even projected back to the lifetimes o f Abu Yazid and Junayd respectively. For instance, the passage presented above in Section III o f this chapter describes a group o f people (qawm) sitting around (jalasa ila) Abu Yazid, while he hangs his head down for some time, before raising it to them to say, ‘W hile you have been sitting down before me, here I have been, roaming my thoughts, looking for a rotten grain w hich you can bear, to extract for you, but I did not find [anything]’.74 The context which is described (albeit as a means o f portraying Abu Yazld’s superiority and disdain) is that o f a teaching circle, in which segments (even ‘rotten grains’) are transmitted. M ost o f the isnads w hich suggest w ritten modes o f transmission identify the source as ‘so m eth in g w ritte n ’75 by Ja cfar b. M uham m ad al-K huldl (d.348/959).76 T here are three formulae in total by w hich isnads indicate this: 1 akhbarand fT-ma kataba ilayya ‘H e inform ed me in som ething that he w rote to m e’77 2 kataba ilayya . . . qala ‘H e w rote to me, saying that . .. ’78 3 akhbaram f t kitabih ‘H e inform ed me in som ething that he w ro te’79 It is not clear w hether these formulae are being employed to indicate alternative modes o f w ritten transmission, or a single m ode. T h e first two form ulae appear to refer to w ritten correspondence from K huldl (presumably to Abu N u'aym ), but, according to the traditional dates, Abu N u caym would have been only twelve years old w hen K huldl died.80 In that case, w hat would be the purpose in claiming this? O ne possible explanation is that the status o f w ritten correspondence w ould have had the effect o f 61

T he H ilya t al-aw liya’

boosting the authority o f the material, since it w ould indicate a direct and personal form o f w ritten transmission. There are several segments in the biographies that have been examined w hich describe, w ithin their rnatm , w ritten correspon­ dence betw een Sufis. For instance, in one o f the narratives presented above, Yahya b. M u'adh writes to Abu Yazld to boast about his intoxication, only to receive a dismissive w ritten reply.81 Similarly, Ahm ad b. Harb sends, along w ith a prayer-mat, a w ritten request for Abu Yazld to pray on it at night (wajjaha ilayh hasTran wa-kataba ma ah ilayh, Sil calayh b i-’l-layl ); Abu Yazld returns a dismissive w ritten response: fa-kataba A b u YazTd ilayh, In m ja m a ctu cibadat ahl al-samawat w a -’l-ardin al-sabca fa -ja caltuha f t mikhadda wawadactuha tahta khaddT (I have gathered the acts o f worship o f the

people o f the heavens and the seven climes, and I have put them in a pillow, and placed it under my cheek (i.e. to sleep on)!).82 W ritten correspondence is thus another recurrent m o tif in the matns o f segments. A lthough the descriptions are perhaps retrojections, they probably reflect the actual practice, at least by the time o f the compilation o f the text. T he last o f the three formulae presented above (akhbaram f t kitabih) is the most problematic because, taken at face value, kitab may refer, as well as to w ritten correspondence, to a literary w ork o f some form. However, there is no extant book ascribed to Khuldl, and even the Sufi biographical tradition itself has not ascribed books to him .83 T he biography o f Khuldl in the tenth volume o f the H ilya itself states that he w rote down segments (,kataba ’l-athar).84 Sulami’s Tabaqat also contains a biography o f Khuldl, w hich states that he was ‘the authority to refer to regarding the sciences o f the Sufis, their books, the segments about them and their biographies’ (wa-kana >l-marja ilayh f t culum al-qaum w a -kutubihim w a-hikayatihim w a-siyarihim ) .8S Significantly, this com m ent m entions the books o f the (other) Sufis w ithout suggesting that Khuldl w rote any himself. If it is unlikely then that the formula a k h b a ra n d ft kitabih is a reference to a w ritten w ork com posed by Khuldl, it may instead be a reference to w ritten segments, perhaps in the form o f a notebook. (O ur evidence for the existence o f such a source is, after all, precisely w ritten segments). Two further isnads in the biography o f Junayd appear to refer to sources o f the latter kind, although w ithout reference to K huldl.86 T he m ore elaborate o f these is presented below: 62

T he H ilya t a l-a id iy a ’

I recited to (qara’tu cala) Abu ’l-Husayn M uham m ad b. cAli b. Habish al-N aqid al-Sufi, a com panion o f Abu ’l-cAbbas b. cA ta’, in Baghdad in the year 359 from som ething w ritten by him (kitab ), and he acknowledged it [as being his own] (faaqarra bih). I said that I heard Abu ’l-Q asim al-Junayd b. M uham m ad say . . . (H, X, 256.4-6)

T he above account o f an elaborate form o f transmission concurs w ith C alder’s observation o f qira’a in third century legal texts.87 This particular example suggests that Abu N u caym recited to Ibn Hablsh from a copy o f som ething the latter had w ritten, and, in response, he confirm ed that it was indeed his work. T he specific details concerning the time and location o f this transmission function to dispel any doubts over its authenticity. T he kitabs , or ‘w ritings’, ascribed to renow ned transmitters such as K huldl may have consisted o f w ritten segments (hikayat ) transm itted by means o f qira’a in a similar fashion. T he following narrative segment in the biography o f Junayd offers a vivid description o f the process o f the production o f w ritten segments from a teaching session: A nd w henever I came across [Muhasibi] in the place where he would sit (al-m akan alladhi yajlis f l h ), he would say to me, Ask me something! I w ould tell him, I do not have a single question to ask you. H e w ould then say, Ask me about anything that comes to your mind! Questions w ould then rain dow n upon me and I w ould ask him about them . H e w ould answer them for me on the spot, and then go to his house and put them into writing. (H, X, 255.18-256.3)

This narrative alludes to a situation that can be inferred from the literary structures o f the H ilya , namely the production o f w ritten segments by the recording o f questions and answers from a teaching session. This particular description is remarkable in that it describes M uhasibi, the teacher, as the one w ho compiles the segments after him self having answered Junayd’s questions. This appears to be in order to emphasise the degree to w hich he valued the sessions w ith his precocious student.88 T he depiction o f M uhasibi may also function to attribute to him the authorship o f books based on the inspiration provided by Junayd, w ho eventually 63

T h e H ilya t al-audiya’

became m ore authoritative in tradition than his own teacher. O nce again, in contrast to the narratives involving Saqatl, M uhasibl is depicted as a wise teacher, who, in this particular narrative, has the perception to foresee that Junayd w ould be inspired w ith w orthw hile questions for him to answer.89 To recapitulate, the isnads in our representative biographies o f the H ilya mostly indicate oral transmission. They suggest that the segments were produced in a teaching milieu, similar to the segments that make up the bodies o f the biographies in Sulaml’s Tabaqat. There also exists a small proportion o f segments that appear ostensibly to have been obtained from w ritten sources, nam ely w ritten correspondence, and the w ritten segments belonging to the collections o f other scholars. M ost o f these sources also w ould appear to have originated in an oral teaching context. T he transmitters and compilers o f the H ilya seem to have been participants themselves, and their ow n collections o f segments probably form ed the bulk o f the com ponent materials o f the work. T he isnads o f all the segments in the H ilya are normally understood to represent transmission to Abu N u'aym , the assumed author o f the work. However, there are indications that this cannot always be taken for granted, for Abu N u'aym him self is m entioned in the isnads o f a few segments; he is nam ed as the immediate source by oral transmission o f at least three segments,90 and in one instance that I have found, he is as far removed as the second transm itter (by qira’a) before the com piler.91 There is little chance that a namesake is being referred to in these examples because his full name is given, accom panied by the titles and epithets that are usually attributed to him (al-shaykh, al-hafiz etc.).92 T he inclusion o f a qird’a transmission from Abu N u'aym suggests that he him self possessed a collection o f w ritten segments that was used as a source, perhaps even the main source, for the H ilya. However, these examples prim arily testify that at least some o f the segments in the H ilya were com piled by his students (and their successors). Abu N u caym, the individual to w hom the H ilya is traditionally ascribed, may have started the compilation, and even made the most significant input towards the final product, but nonetheless, the H ilya , in the form in w hich it has survived today, was not the product o f his authorial control from beginning to eventual com pletion.

64

T h e H ilya t al-aw liya’

VIII T he final statem ent in the H ilya , before the concluding ham d and salat in praise o f G od and the Prophet respectively, offers a date for the com pletion o f its compilation: T he author (m u ’a llif) said, This is the last o f what I dictated (amlaytu), on the last Friday o f the m onth o f D h u ’l-hijja in the year 422. (H, X, 408.18-19)

T he above statem ent gives the impression that the w ork was put into w riting through a formal process o f dictation by the author. T he com pletion o f the dictation is dated w ithin the lifetime o f Abu N u'aym , so he is presumably the one w ho is being referred to by the title m u ’allif, even though, to my knowledge, this title is not used elsewhere in the H ilya . (O f course, it may have been used uniquely at this point because o f the function o f this closing statem ent w ith regards to the attribution o f authorship). The inconsistencies and incoherencies in the form o f this w ork suggest that it was not in fact under the control o f a single 'au th o r’ in the m anner o f a dictation, neither Abu N u'aym nor any other individual. It w ould appear that a nu m b er o f individuals participated in the w ork’s compilation, w ith little co-ordination betw een their efforts, and that it was com pleted after Abu N u caym ’s death, incorporating material that had been obtained from him via intermediaries. T he fact that the biographies o f Sufis from Isfahan are at the culm ination o f the work, would suggest that the actual compilers probably shared an allegiance to the tradition o f Sufism w hich was associated w ith Abu N u caym ’s grandfather Ibn M a'dan. T he attribution o f the H ilya to Abu N u caym is probably due to the fact that he was the illustrious collector o f biographical segments par excellence associated w ith this school tradition o f Isfahan (as testified by his surviving works). M oreover, his own collection o f segments perhaps constituted the starting point for the H ilya , as well as the main source for its final version. As already m entioned, the biographies o f Abu Yazld and Junayd are amongst sixty-eight biographies in the tenth volume o f the H ilya w hich correspond to those found in Sulaml’s Tabaqat alSufiyya. They are typical o f such biographies in that they contain many segments o f material that are also found in that earlier work. However, the isnads, on the authority o f w hich they are presented, 65

T he H ilya t al-aw liya’

do not acknowledge it as their source.93 M ost o f these isnads do not identify SulamI as the source at all, whilst those that do invariably indicate oral transmission (e.g. haddathana M uham m ad b. al-H usayn qala / samx tu M uham m ad b. al-H usayn ya q u l).94 T he order o f presentation in the H ilya , in its present form, o f the material that is also found in the Tabaqat, does not suggest that the earlier work was used systematically as a source; segments that are found juxtaposed in the earlier work, are often presented apart in the H ilya (and vice versa), w ith o u t any obvious ex p lan atio n .9d However, the likelihood is that this is simply a further effect o f the general inconsistency in the ordering o f segments, and it is probably attributable to layers o f grow th and interpolation in the biographies. Perhaps the most obvious indications o f the use o f Sulamfs Tabaqat are the sequences o f biographies that are juxtaposed in both works. For instance, the biography o f M uham m ad b. al-Fadl al-Balkhl is the first m em ber o f a sequence o f nine juxtaposed biographies that have counterparts in the earlier work, the first three o f w hich are juxtaposed in exactly the same order in both works. M oreover, the content o f these biographies reveals a heavy reliance on Sulamfs Tabaqat. T he biography o f Balkhl, for instance, possesses an introduction derived from that o f its counterpart in the earlier w ork instead o f an introduction in s a f w hich one would normally expect in the H ilya.96 Furtherm ore, its hadith transmis­ sion is the same one that is given in its earlier counterpart (just as in the examples o f Abu Yazld and Junayd),97 and its body consists entirely o f segments derived from that w ork.98 SulamI is m entioned twice in the H ilya outside o f the context o f an isnad. In the second volume, amongst the biographies o f ‘the people o f the canopy’ (ahl al-suffa) 99 there is a statem ent under the rubric qala il-shaykh , in w hich SulamI is nam ed as som eone w ho had m entioned (dhakara) those w ho ‘occupied the canopy’.100 This is followed by a biographical introduction to SulamI himself, describing him as som eone w hom the com m entator (presumably Abu N u'aym ) had m et (wa-huwa ahad man laqmah). However, it does not ascribe any w ritten works to SulamI. This is all the more surprising since, immediately after this introduction to SulamI, Abu SacId b. al-Acrabi is introduced as the second person w ho had m entioned the ahl al-suffa , but, in sharp contrast, he is described as the author o f famous works. O ne o f them , the Tabaqat a l-N ussak , is even nam ed specifically as the m odel for the biographies o f ‘the 66

T he H ilya t al-aw liya’

successors’ in the H ilya (lah al-tasanxjal-mashhura . . . w a -’qtafxfTbaqT ’l-kitab m in dhikr a l-ta b im hadhwah idh huwa sh a ra a fT ta ’llf Tabaqat al-N u ssa k).un In this context, the failure to ascribe w ritten works

to SulamI suggests that, at the tim e this biographical introduction was entered into the text, his works were not yet known. SulamI seems to have been recognised simply as som eone w ho m entioned 'the people o f the canopy’ in an oral teaching context. T he second m ention o f SulamI outside o f the context o f an isnad occurs in the following qala ’l-shaykh demarcatory group introduction, w hich is found at the end o f the biography o f Abu Yazld, and immediately before a sequence o f six biographies o f fellow Khurasanians: T he Shaykh Abu N u caym said, As for ‘the suns o f the East’ (.shumus al-mashriq) and their em inent ones, Shaykh Abu cAbd al-R ahm an al-Sulaml o f N ishapur has taken care to m ention them in his book, entitled Tabaqat al-Sufiyya. I wished to put down, in concise form, the names o f a group o f their famous members in my ow n book. (H, X, 41.24-42.2)

This group introduction testifies to familiarity w ith Sulaml’s Tabaqat a l-Sufiyya , and so it has presumably been entered into the text after the aforem entioned biographical introduction to SulamI w hich fails to m ention that he was an author at all. T he discrepancy betw een allusions to SulamI in the above passages suggests that the redaction o f the text by means o f such qala ’lshaykh statements (attributed to Abu N u'aym ) was itself a gradual process, during w hich Sulaml’s Tabaqat al-Sufiyya became familiar. It is likely that at this stage it was used by the redactors as a source o f material. T he H ilya appears to have grown even after the addition o f material from Sulaml’s Tabaqat, such that, in the form in w hich it eventually reached closure, it does not im mediately suggest that the Tabaqat was used systematically as a source, whilst at the same time it leaves little doubt that it was relied upon heavily at one stage as a source for both segmental material (the buildingblocks) and (architectural) dem arcatory introductions.

67

This page intentionally left blank

C h a p t e r Three

The Persian Tabaqdt al-Sufiyya

I he surviving works ascribed to Abu Isma'il "Abdullah b. M uham m ad al-Ansarl (d.481/1089) suggest that he was a Sufi author from Herat, w ho was also a Hanbalite. His nisba implies that his ancestors may have been amongst the M edinan ‘helpers’ (ansar) o f the Prophet M uham m ad’s early followers, w ho are said to have accom m odated the M eccan m igrants.1 T he earliest biography o f Ansarl is found in M uham m ad b. Abl Yacla’s (d.527/1133) Tabaqdt al-Hanabila. Ibn Abl Yacla introduces Ansarl, in his b rief (15-line) biography, as ‘the leader o f the Sunnites (ahl al-sunna) in H erat’, w ho is know n by the title Shaykh al-Islam ,2 and is called KhatTb a l-cajam (the orator o f the Persians) ‘on account o f the depth o f his knowledge, his eloquence and his em inence’.3 H e gives Ansari Hanbalite credentials by describing him as a firm opponent o f Ash’arism (kdna shadfdan cald 7A sh cariyya) and a student o f Hanbalite authorities on hadith and tafsir. Ibn Abl Ya'la finally attributes two distiches o f verse to Ansarl, w hich begin w ith the statem ent and hanbalT ma hayaytu . . . (I’ll be a Hanbalite as long as I live .. .).4 N o suggestion is made in this collection o f biographies o f Hanbalites that Ansarl m ight have also been a mystic. Ibn Yacla also informs us that Ansarl had two sons, cAbd al-H adl and Jabir, the form er o f w hom was allegedly killed by Ismailis (faqatalathu ’l-B atiniyya ). DhahabI (d.748/1348) relates that Ansarl him self was involved in religious disputes on account o f his Hanbalite allegiance, being put on trial many times.5 T he bulk o f

T

69

T he Persian Tabaqat a l-Sufiyya

D hahabl’s 145-line biography o f Ansar! is in fact taken up w ith presenting him as a staunch H anbalite w ho w ould never compromise his belief, even at the cost o f being expelled from his native Herat, and putting his life in danger;6 Ansarl is protected and supported by G od in his encounters w ith opponents because o f his adherence to the ‘correct’ school. These are com m on motifs in biographies o f Hanbalites, as well as religious scholars o f other affiliations, containing echoes o f the biographical tradition o f Ibn Hanbal himself.7 It should finally be noted that DhahabI depicts Ansarl also as a Sufi, and ascribes both Sufi and Hanbalite works to him .8 W hilst Ansarl is a celebrated figure from the past for the Hanbalite tradition, the surviving works attributed to him suggest that he was prim arily a Sufi. Eleven o f the sixteen extant works ascribed to him are specifically mystical in subject matter, while only two can be classified as Hanbalite.9 T he Tabaqat al-Sufiyya , w hich is traditionally ascribed to Ansarl, is the earliest w ork o f the Sufi tabaqat genre to have been w ritten in Persian, and it is extant in five manuscripts, the oldest o f w hich states in its colophon that it was copied in 771/1335, nearly 250 years after the death o f the assumed author. M uch o f the text is w ritten in an archaic form o f Persian that W. Ivanow has identified as the local dialect in 11th— 12th century South Eastern Khurasan.10

II T he Persian Tabaqat a l-S ü fiyya 1^ contains over two hundred biographies in an overall chronological arrangement. O n first impression, the order o f biographies may seem to follow the generation system o f Sulaml’s work. However, a closer examina­ tion reveals the influence o f com peting m ethods o f arrangement, to an even greater degree than has been observed already in the H ilyat al-aw liya\ M ost o f the biographies appear to be arranged according to two different principles, namely, according to generation, and according to the name o f the subject. T he former o f these predominant methods is comparable to what is observed in Sulamfs Tabaqat al-Süfiyya , where biographies are grouped together into generation sections.12 T he Persian Tabaqat alSüfiyya includes six generation sections, the first five o f which correspond approximately (in time-span) to those structured by 70

T he Persian Tabaqat a l-Sufiyya

Sulaml. Each generation section, apart from the first, is introduced by means o f a heading,13 and most biographies begin by classifying the subject as a m em ber o f a particular generation.14 These factors make the generation system the most influential one in the final form o f the work, and it is presumably on account o f this system that it has been given the same title as Sulamfs earlier work. For almost all o f the biographies found in Sulamfs Tabaqät alSüfiyya there are corresponding ones in this w o rk .15 These biographies are also usually found in the corresponding generation section.16 T he Persian w ork also contains additional biographies in each o f the five corresponding generation sections, as well as an extra ‘sixth generation’ (made up o f the biographies o f individuals w ho lived too late to be included by Sulaml). It has already been observed that the order o f biographies w ithin the generation sections o f Sulamfs w ork is significant, in that the head biographies tend to be the most im portant.17 A comparison betw een the corresponding generation sections o f the two works shows a considerable discrepancy w ith regard to the head biographies. For example, in the Persian work, the second generation section begins w ith the biography o f Sahl b. "Abdullah al-Tustarl (d.2 8 3 /8 9 6 ),18 rather than Junayd, whose biography is the fourteenth to be listed.19 However, in spite o f this, Junayd’s biography is once again the longest in the second generation, and he also features prom inently in the introduction and the appendix o f the w ork. T he order o f biographies w ithin generation sections therefore does not seem to be managed according to the same principles, or at least not as carefully, as in Sulamfs w ork.20 It has been noted that the first generation section in Sulamfs Tabaqät is designed to forge continuity w ith the time o f the Prophet, through the three generations o f ‘the pious predecessors’ (ial-salaf al-sälih). Therefore it begins w ith the biographies o f Sufis from the second century, and they are linked back to ‘the pious predecessors’ by a specific statement in the introduction o f the w o rk .21 T h e m ost pro m in en t difference betw een the first generation sections o f the two works is the inclusion, at the start o f the Persian Tabaqät, o f the biography o f Abü Häshim al-Süfl, w ho is not m entioned at all by Sulaml. Abü Häshim is said to have been the first person to be called ‘Suff (awwal kassT ki ’ü-rä S ü ß gufiand B ü H äshim S ü ß Tdh).22 This is the implicit reason why his biography is the first to be listed. In addition to a few segments about Abü Häshim, it includes a m yth about ‘the first khäniqäW 71

T he Persian Tabaqat a l-Sufiyya

(Sufi lodge) to have been built, w hich is not related at all to the life-story o f Abu H ashim .23 Segments about ‘firsts’ (aw a’il) therefore appear to have been attracted to the beginning o f this work. T he second most prom inent principle for the arrangem ent o f biographies in the Persian Tabaqat is name classification. For example, the following groups o f juxtaposed biographies w hen com bined account for m ore than a fifth o f the whole work: • ten juxtaposed biographies about subjects who have in common the kunya Abu ’l-cAbbas/ Bu’l-cAbbas. (TABA, 294-312)

• nine juxtaposed biographies about subjects w ho have in common the kunya Abu Bakr/ Bu Bakr. (TABA, 429-437)

• seven juxtaposed biographies about subjects w ho have in common the kunya Abu Ya'qub/ Ba Ya'qub. (TABA, 275-283)

• seven juxtaposed biographies about subjects who have in common the kunya Abu Ja cfar/ Bu Jacfar. (TABA, 345-351)

• six juxtaposed biographies about subjects who have in common the kunya Abu Bakr. (TABA, 390-393)

• five juxtaposed biographies about subjects whose names rhyme (Samnun, Z ahrun, cArun, M aym un and Sacdun M ajnun). (TABA, 225-231)

This m ethod o f arrangem ent appears, for the most part, in harm ony w ith the generation principle. T he two groups o f juxtaposed biographies about subjects w ith the kunya Abu Bakr, for instance, occur in separate generation sections, suggesting that they were deliberately separated for this specific reason. However, there are a few instances w here it appears, at first glance, as if the name classification principle conflicts w ith the generation principle. For instance, the aforem entioned group o f ten juxtaposed biographies w hose subjects are called ‘Abu ’1-"Abbas’ is found in the third generation section, despite the fact that two o f its members are classified at the beginning o f their 72

T he Persian Tabaqat a l-Sufiyya

biographies as belonging to the fifth generation. T hat is to say, the subjects o f two o f these biographies are classified as members o f the later generation, but nonetheless, they are included amongst the third generation because o f their names.24 In m any o f the examples o f this kind, name association is not the only factor. For instance, the biography o f Abü Ya'qüb al-N ahrajürï is am ongst the aforem entioned group o f seven juxtaposed biographies whose subjects are called ‘Abü Yacqüb’, located in the third generation section, despite the fact that he is classified as a m em ber o f the fourth generation. His biography follows immediately after that o f Abü Y a'qüb al-Süsï w ho is said to have been his teacher.25 Similarly, the biography o f Bü cU thm àn al-M aghribï, w ho is classified as belonging to ‘the fifth or the sixth generation’, immediately follows the biography o f Abü cU thm ân al-Hïrî, w ho is classified as a m em ber o f the second generation. In the introduction to M aghribl’s biography, it is reported that his grave is next to that o f H lrl.26 These juxtapositions that conflict w ith the generation system, p o in t to the existence o f a third, less immediately apparent factor, namely the juxtaposition o f biogra­ phies because o f associations betw een their contents.27 This is not only a factor in causing the juxtaposition o f biographies w hich are classified as belonging to different generation sections, but also in creating inconsistencies in the application o f the name classification principle. For instance, a pair o f juxtaposed biographies o f subjects w ith the kunya Abü "Abdullah is separated from the biography o f Abü "Abdullah al-Jalla’ by the biography o f Abü cUbayd al-Busrl, w hich begins w ith a statement attributed to Jallâ’.28 Similarly, the biography o f Abü Jacfar M ajdhüm is separated from the aforem entioned group o f seven Abü Jacfar biographies by that o f Abü "Abdullah b. al-Barql. Ibn al-Barqï is m entioned in the single segment that makes up the biography that precedes his own (the biography o f Abü Jacfar M u cadh al-Misrl).29 To recapitulate, three different factors have been identified in the arrangem ent o f biographies in the Persian Tabaqat al-S u fiyya : the generation system established by Sulaml’s earlier work, a name classification system and the juxtaposition o f biographies whose contents are associated. T he generation system is likely to have been applied last o f all; it is the most prom inent in the final form o f the work, and, as already m entioned, generation section headings are provided, as well as allusions to this system at the beginning o f most biographies (not to m ention in the title o f the work). It is 73

T he Persian Tabaqat a l-Sufiyya

perhaps no w onder then that the later Sufi tradition considered the w ork to be based on Sulaml’s Tabaqat.30 T he juxtaposition o f biographies according to th eir subjects’ names is a m ore elem entary m ethod o f arranging biographies, w hich was perhaps derived from the way segments o f material were stored.31 W hilst it may facilitate finding a particular biography, its arbitrariness limits the ability o f a com piler to control the structuring o f the past. The juxtaposition o f biographies due to association between their contents is comparable with, and often an extension of, the juxtaposition o f com ponent segments o f material w ith such associations, w hich is especially prom inent in this w ork.32 T he conflict and inconsistencies in the application o f each o f these principles may have been caused by growth o f the text and the direct im port o f sources w ith contrasting formats. T he end-result is comparable w ith w hat has been observed in the H ilya , in that there is evidence o f layers o f redaction activity, rather than a static com pilation by a single editor. T h e Persian Tabaqat a l-S ü jiyya is com parable w ith Abü N u 'a y m ’s H ilya , not only due to the inconsistent application o f com peting organisational principles, but also because its redaction devices, such as its introductions to constituent parts o f the work, do not always correspond to the context in w hich they are found. For example, it contains a few cases o f (ostensibly) repeated biographies. That is, introductions are given to separate biogra­ phies about the same individual, as in the case o f Fath b. Shakhraf al-MarwazI, but only one o f them is followed by the body o f an actual biography o f M arwazl.33 T he Persian Tabaqat also contains, uniquely for the Sufi tabaqat genre up to this point, non-biographical chapters interm ingled amongst its biographies. An example o f such a chapter appears in the middle o f the biography o f Junayd, and will therefore be considered at length in the fourth section o f this chapter.34 These features further emphasise the lack o f a single author in control o f the final form o f the work, after its layers o f redactions.

Ill M ost o f the biographies in the Persian Tabaqat al-Süfiyya consist o f two parts: the introduction and the body. They therefore differ from the biographies o f earlier tabaqat works, in that they do not 74

T he Persian Tabaqat a l-Sufiyya

contain a hadith transmission. T he latter seems to have been no longer a m atter o f significance for individual biographies, although hadiths still play a significant role in the w ork outside o f them .35 T he body o f a biography in this w ork consists o f discrete segments juxtaposed. However, a characteristic innovation is that these segments can be about individuals other than their subjects, as well as those about the subject him self (and often in greater proportion). For instance, the body o f the biography o f Abu Yazld is made up o f 37 segments, only 12 o f w hich are about Abu Yazld himself. T he rem ainder are accounted for by 13 segments about other individuals, and 12 segments o f com m entary presented under the rubric S haykh al-Islam guft. T he latter are traditionally understood to be com m ents made by Ansari, and they usually relate to immediately preceding segments, w hether they are about Abu Yazld or another individual. Consider, for example, the following passage from the biography o f Abu Yazld: 1 Abu Yazld was seen after his death in a dream, and was asked, H ow are you? H e replied (Abu YazTd-ra pas-i marg bikhwab didand guftand hal-i tu guft), I was asked, O ld man what have you brought? I said, W hen a poor man reaches the court o f a king he is asked w hat he wants not w hat he has brought! 2 Shaykh al-Islam said that cAlI SharTfi told him, Amongst us in N ishapur there was a poor old wom an called ‘cIraqiyya\ She used to beg door-to-door. She died. She was seen in a dream, and was asked, H ow are you? She replied, I was asked, W hat have you brought? I said, O all my life I’ve received assurances that G od will provide, and now I am being asked what I have brought? It was said, She is correct, leave her alone! 3 cA li Sharifi died in the m onth o f Muharram in the year

436 and was buried next to Im am Yahya b. 'Ammar.

4 Fudayl cIyad was seen after his death in a dream, and was asked, H ow are you? H e replied, I did not see anything

for the slave better than his lord (guft lam ara li- yl - cabd

khayran min rabbih).36 5 Khayr-i Nassaj was seen after his death in a dream, and was asked, H ow are you? H e replied, W hat has it got to do w ith you (tu-ra a zm chi kdr)? For once I have escaped from this futile world o f yours.

75

T h e Persian Tabaqat a l-Sufiyya

6 Sari Saqatl says, I passed by a monastery (dayr) and called, O m onk (rahib). T he m onk answered, W hat is it pagan? I asked, H ow long have you been here? H e answered, T hirty-three years. I asked, W hat have you learnt in these thirty-three years? H e replied, W hich servant have you seen com ing out o f the house o f a king and talking about it? W hat does the secret o f the kings have to do w ith you (.tu-ra ba raz-i m uluk chi kar )? 7 Shaykh al-Islam said, Conversation w ith strangers is not correct. First be acquainted, then start to converse. (TABA, 92.6-93.9)

O nly the initial segment o f this passage taken from the biography o f Abu Yazld is about the subject himself. T he subsequent segments, the addenda, consist o f material about other individuals, and com m entary attributed to Ansarl. Segments 1, 2, 4 and 5 each appear to have developed from the same narrative motif. T he subject is seen in a dream after he/she has died and is asked about his/h er condition. This is expressed identically in Segments 1,4 and 5: fulan-ra pas-i marg bi-khwab dfdandguftand, H a l-i tu? G uft . . . Segment 2 differs because the lesser-known subject requires an introduction. After the statement that she died, this segment continues : bi-khwab dfdand guftand, H al-i tu? G u ft . . . It is the closest variant o f Segment 1, because both subjects relate their reaction to being asked the same question, chi awurdf (W hat have you brought)? Segment 6 is a narrative attributed to Saqatl about a personal encounter w ith a m onk. His account is parallel in structure to Segments 1, 2, 4 and 5, in that an outsider (Saqatl/living person) asks an insider (m onk/dead person) about w hat he has experienced there (m onastery/afterlife). Segments 3 and 7 are related only to the segment which im mediately precedes them . Segment 3 offers details about the death o f the nam ed source in segment 2, cAlI Sharlfl. Segment 7 contains an instruction not to enter into conversation w ith strangers, w hich is attributed to Ansarl. It is related to the m o n k ’s response in Segment 6, by w hich he refuses to disclose anything to Saqatl, the M uslim outsider. O nly Segment 2 in the above passage m entions its source; the remaining segments, about Abu Yazld and other individuals, do not m ention their source at all, even though the compiler could 76

T he Persian Tabaqat a l-Sufiyya

not have obtained them directly. This is typical for the w ork as a whole, w here sources are seldom m entioned. W h en they happen to be m entioned, a full isnad is sometimes provided, or, as above, simply the identity o f the first transmitter. A substantial proportion o f this ‘Persian’ work is in Arabic. Segment 3 is entirely in Arabic, as is conventional for such formulaic death notices. T he utterance contained in Segment 4 is in Arabic but the context is described in Persian, Fudayl cIyad-ra pas-i marg bi-khwab didand guftand, H a l-i tu? G uft, lam ara li-’l - cabd khayran m in rabbih. T here is also an example in the same biography

o f an Arabic utterance followed by its translation into Persian.37 Segment 3 stands apart from the rest, not only because it is the only one that is entirely in Arabic, but also because it is a ‘factual’ statem ent rather than a report o f som eone’s speech. It states w hen the transm itter o f the preceding segment died and w here he was buried. Since it only offers inform ation about the latter’s death, it is probable that it was interpolated here for the specific purpose o f guaranteeing that transmission could have taken place. It was perhaps originally a marginal gloss that became incorporated into the main body o f the text. T he final sentence o f Segment 2 fits awkwardly,38 and even w ithout it the story about the beggar-wom an w ould still be complete: she dies after a lifetime o f begging and is subsequently seen in a dream, reporting that upon being asked w hat she had brought w ith her, she had responded w ith the ironic lament, ‘O h, all my life I’ve received assurances that G od will provide and now I am being asked what I have brought?’ Segment 6 w ould be com plete even w ithout its final sentence, the second question in succession posed by the m onk, ‘W hat does the secret o f the kings have to do w ith you?’ (tu-ra ba raz-i m uluk chi kar?). T he same structure (tu-ra . . . chi kar) is used in the first part o f Khayr-i Nassaj’s retort in the preceding segment, ‘W hat has it got to do w ith you?’ It is possible that the final sentence in Segm ent 6 was generated parallel to this, and the occurrence o f a variant o f Segment 5 w ithout being juxtaposed to Segment 6, in the biography o f Khayr-i Nassaj himself, later in this work, suggests that the two segments were not necessarily transmitted together.39 T he above passage can be best described as a ‘bundle’ o f segments. It is typical o f such bundles in the Persian Tabaqat alS u fiy y a , in that it dem onstrates the hetero g en eity o f the 77

T he Persian Tabaqat al-Sufiyya

com ponent segments o f this work, as well as some o f the patterns o f association betw een those that have been juxtaposed. It also contains some examples o f w hat appear to be gloss and interpolation, and signs o f possible organic growth. Even if the segments about Abu Yazld should be considered in isolation from the other segments in his biography, then the h eterogeneity o f the segm ents, and the lack o f a single, predom inant organising principle w ould still be evident. For instance, three o f such segments about Abu Yazld him self possess a full isnad ,4() four o f them name only the first transm itter,41 and the remaining five segments offer no source details. T he latter group includes variants o f segments found in earlier works provided w ith their own isnads .42 Different m ethods o f presentation are also used for the texts o f the Abu Yazld segments; out o f the eleven utterances attributed to him, five are presented in Arabic and six in Persian. O ne o f the Arabic utterances is followed im mediately by a Persian translation,43 whilst another is preceded by an isnad and contextual fram ew ork in Persian.44 T here are therefore no indications that the segments about Abu Yazld may have once represented a ‘unified core’, in relation to w hich other segments could have been interpolated later. T hat is to say, the segments about Abu Yazld, just like the other segments in his biography, appear to have been com piled as part o f an extended process by a num ber o f individuals w ho had different preferences w ith regard to presentation. It seems strange for a w ork traditionally believed to have been produced on the basis o f SulamT’s Tabaqat al-Sufiyya that only two o f the segments in the biography o f Abu Yazld should correspond to segments in that precursor. These tw o apophthegm atic utterances are in fact juxtaposed and identical to their earlier counterparts: kufr ahl al-him m a aslam min m a n ahl al-minna (the unbelief o f those w ith aspiration is sounder than the belief o f those w ho seek reward), ab'adu ’l-khalq m in A llah aktharuhum isharatan ilayh (the furthest people from God are those w ho point to him the m ost).45 To put these two cases into perspective, it should be noted that there are just as many in this biography that are variants o f those found in the corresponding biography o f Abu N u 'ay m ’s H ilya. Nonetheless, the fact that the two aforem entioned segments are both juxtaposed and identical to their earlier counterparts suggests that Sulami’s w ork was used, at least as one o f the sources for the Persian work. It was probably used during the same 78

T h e Persian Tabaqat a l-Sufiyya

(relatively late) redaction o f the work, in w hich the generation system was applied, along w ith all o f its trappings, as the predom inant organising principle for the order o f presentation o f biographies.46 This biography o f Abu Yazld also curiously includes variants o f a couple o f short utterances w hich are found attributed to him in earlier works, only here they are attributed to other religious authorities, namely the first R ightly-guided Caliph, Abu Bakr, and the mystic authority, D hu ’1-Nun al-M isri.47 It has already been suggested that the existence o f such variants attributed to different individuals highlights the transferability o f this kind o f biographical material. However, it is unlikely to be a mere coincidence that these utterances are attributed to individuals other than Abu Yazld here in his own biography, nor that both o f the individuals concerned are m ore authoritative figures in tradition than Abu Yazld himself. It suggests that out o f a desire to give m ore weight to the respective utterances they had been attributed to m ore authoritative individuals, but had still rem ained associated w ith Abu YazTd, and thus appear in his ow n biography in this very loosely-controlled com pilation.48 The inclusion o f a high p ro p o rtio n o f segments about oth er individuals in the biography o f Abu Yazld also has the effect o f em bedding material attributed to him w ithin a w ider tradition o f mysticism and piety. In contrast to the biography o f Abu Yazld, the body o f the biography o f Junayd contains nearly twice as many segments about the subject himself, than about other individuals. This discrepancy is probably a sign o f the greater acceptance o f him in tradition, such that material attributed to him is w ithout the need for any corroboration.

IV W hilst the body o f the biography o f Junayd contains a higher proportion o f segments about the subject himself, it nonetheless contains m ore indications o f interpolation and organic growth than the body o f the biography o f Abu YazTd. This can be interpreted as another indication o f the greater im portance attached to Junayd by those w ho com piled this work. A prom inent example o f such grow th and interpolation is the appearance o f the chapter heading M a s ’ala f t ’l-tawhfd (An issue concerning divine 79

T he Persian Tabaqat a l-Sufiyya

unity) only a few pages after the heading for the biography o f Junayd itself49 This heading implies the start o f a new, thematic chapter about tawlud , and the close o f the biography o f Junayd (and it has been interpreted thus by the m odern editors). T he next heading after M a s’alafT ’1-tawhTd is that o f the biography o f Ibn alKaranbl, a Sufi associated w ith Junayd.50 These headings suggest the following sequence o f chapters: 1 T he biography o f Junayd 2 A thematic chapter about tawlud 3 T he biography o f Ibn Karanbi A close inspection o f the contents o f the so-called chapter on tawhid suggests that this sequence is not so clearly defined. Particularly conspicuous in this regard are the sequences o f segments that form its beginning and its closing sections, respectively, which are both related closely to the preceding biography o f Junayd whilst being unrelated to the topic o f tawhid. For instance, towards the end o f the chapter on tawlud , the following segment triggers the start o f a continuous sequence o f segments unrelated to tawhid , which reaches as far as the demarcating heading that signals the start o f the biography o f Ibn Karanbi: Shaykh al-Islam said that Junayd says, People think that I am the student o f Sari Saqatl, [but] I am the student o f M uham m ad 'All Qassab, w hom I asked, W hat is Sufism? H e replied, I do not know, but it is a noble (ka n m ) character w hich G od (the Noble) manifests in a noble time, through a noble man, amongst a noble people (qawm -i kiram). (TABA, 182.3-6)

T he above segment consists o f an utterance attributed to Qassab introduced by Junayd, w ho declares that Qassab was his teacher rather than Saqatl. It is unrelated to tawhid , even though the last chapter heading to precede it is M a s’ala f t ’l-tawlud , implying that it belongs to that chapter. However, it is relevant for the biography o f Junayd, especially since it begins with a statement attributed to him about his alleged teacher. In fact, it may even be associated with the polemic regarding the relative superiority o f Junayd’s teachers, which has already been suggested as a factor behind the curious denigration o f Saqatl in the corresponding biography o f the H ilya.5] T he above segment marks the start o f a final section o f forty-three lines o f text, still under the heading o f M a s’ala f t ’l-tawlud , which is 80

T he Persian Tabaqat a l-Sufiyya

not related to tawhid at all, but is related to Junayd. This final section is made up o f the above segment, an extensive com m entary to it, as well as another utterance attributed to Junayd, and commentary to that utterance. The second utterance is presented below: Junayd said, Lack o f ecstasy (wajd) is not harmful w hen it is accom panied w ith excess o f knowledge ( cilm ), because excess o f knowledge is m ore com plete than excess o f ecstasy. (TABA, 184.2-3)

T he above utterance is also unrelated to tawhid , but relevant for a biography o f Junayd. In fact, the same utterance is repeated in the main body o f the biography before the heading M a s’ala f t ’1-tawhTd w hich appears to mark its closure.52 T he repetition o f this segment in the above location, ostensibly towards the end o f the chapter M a s’ala f t ’1-tawhTd, alongside other material that is relevant to a biography o f Junayd, but not relevant at all to a thematic chapter on tawhid , confirms that the boundaries betw een chapters were not always defined as they now appear. It suggests a resum ption o f the body o f the biography o f Junayd after a diversion into the topic o f tawhid. This diversion dissects the body o f the biography o f Junayd into an initial part and a resumption, and the distance thus created probably contributed to the repetition o f the same segment. T he beginning o f the so-called chapter on tawhid confirms these indications. T he first segment, w hich is presented below, is appropriately about tawhid , but it may also be considered relevant to the biography o f Junayd. 1 Shaykh al-Islam said that Junayd said, For thirty years I have not said anything about tawhid (Divine Unity), but have spoken [only] about matters peripheral to it. 2 S haykh al-Islam said, Two individuals have said two different things. T he first, Junayd, w ho said that it is knowledge, the carpet o f w hich has not been woven for thirty years, during w hich people have spoken [only] about matters peripheral to it, that is knowledge o f tawhid. I do not know w hat he is talking about at all, because knowledge o f tawhid has nothing to do w ith tongues (i.e. is beyond words). T he second, Bu Bakr Kattanl, [who] says that the least part o f the knowledge o f Sufism is som ething w hich you cannot discover. T he latter spoke well. (TABA, 169.7-13)

81

T he Persian Tabaqat a l-Sufiyya

T he above passage can be divided into two discrete segments, as shown. Segment 1 offers an utterance attributed to Junayd, stating that he has only talked about matters peripheral to tawhld for thirty years, and not directly about tawhld itself. Segment 2, w hich may appear at first to be a com m entary on Segment 1, in fact offers a variant o f it (in w hich ‘people’ (marduman ) are said to have failed to speak about knowledge o f tawhld), followed by an utterance attributed to Abu Bakr al-Kattanl. These two utterances have apparently been juxtaposed for comparison, in spite o f the fact that the latter utterance is about 77m-/ tasaw w uf (knowledge o f Sufism), rather than 7/m-/ tawhld. Thus, the heading M a s’a l a f i }l-tawhld is followed immediately by the above two segments, both o f w hich involve Junayd, as well as relating to the topic o f tawhld. Nonetheless, w ithout the chapter heading, the above passage may have been considered merely an extension o f the preceding body o f the biography o f Junayd. A factor w hich may have made Segment 2 unusual for the biography o f Junayd is its criticism o f him and expressed preference for the opinion attributed to KattanT. T he biography o f Junayd contains no other instance o f criticism —a feature seldom found in any o f the biographies in works o f the tabaqat genre. T he inclusion o f a new chapter heading, at this particular location in the text, may therefore have been determ ined by this criticism, as an attem pt to exclude it from the biography o f the subject. This interpretation is further supported by the content o f the six segments w hich follow im mediately after the above passage, since none o f them are related to tawhld, and yet all but one are about Junayd (the exception being a segment o f com m entary attributed to Ansarl). This sequence o f six segments culminates in the following narrative: Shaykh al-Islam said that "Abdullah Sacid Kilab had becom e famous for asceticism (zu h d ). According to him, he was asked, W hy do you not refute anything about the Sufis? H e replied, I am not acquainted w ith their science (7/m). H e was told, There is a Sufi authority here, their teacher (ustadh) and leader (sara), w ho has knowledge and teachings beyond the knowledge o f a scholar. H e came and listened to Junayd’s speech, and asked him questions w hich [Junayd] answered, and then he stood up and said to his student, If there should be on earth a group over w hom our words do not have

82

T he Persian Tabaqat a l-Sujiyya

precedence, then it is them , and a science w hich refutes our science and holds it in little value, then it is their science. (TABA, 170.4-11)

T he above segment is also not related to tawhfd , the supposed topic o f the chapter in w hich it is found. R ather, it narrates an encounter involving Junayd. Although it is slightly ambiguous in places,53 it appears to function to confirm the soundness o f Junayd’s ‘science’ (of Sufism). A famous ascetic listens to Junayd and, as a result, holds the Sufis and their science in the highest estimation, rather than refuting them . This segm ent w ould therefore be most appropriate for the biography o f Junayd. T he above segment marks the end o f a sequence o f segments covering tw enty one lines at the beginning o f the so-called chapter on tawhfd , w hich consists entirely o f such material about Junayd that, apart from the first two segments, is totally unrelated to taw hfd . It is, however, followed by a continuous sequence o f segments that are related to tawhfd , and form the bulk o f the chapter. T hat sequence o f segments ends at the aforem entioned utterance attributed to Junayd about his teachers, w hich marks the beginning o f the closing section o f the chapter (see above). T he closing section consists mostly o f segments about Junayd, and contains no material at all w hich is related to tawhfd. To recapitulate, the chapter entitled M a s’ala f i ’l-tawhfd , can be divided into three sections. T he middle section, w hich is made up o f segments about tawhfd , forms the bulk o f the chapter, thus accounting for its title. T he initial section and the closing section, however, are m ore relevant for the biography o f Junayd, w hich immediately precedes the chapter. These characteristics suggest that it has in fact grown out o f that biography. It begins w ith an utterance attributed to Junayd about tawhfd , followed im mediately by other biographical segments about Junayd, prior to material about tawhfd. This suggests a gradual em ergence o f the chapter on tawhfd , by the attraction o f thematically associated segments, comparable to the creation o f a ‘bundle’ o f juxtaposed segments (see Section III above). M oreover, the chapter ends w ith a return to biographical material about Junayd, indicating a resum ption o f his biography. It reinforces the impression that the material about tawhfd dissects the biography o f Junayd, splitting its main body into two parts. T he analysis o f this part o f the w ork suggests the following sequence, presented opposite the given headings: 83

T he Persian Tabaqat a l-Sufiyya

According to headings:

According to content:

A biography ofjunayd A B thematic chapter about tawhid 1 biography ofjunayd C biography of Ibn Karanbi. B 2 3 4 5

utterance byjunayd about tawhid biographical segments about Junayd segments about tawhid biographical segments about Junayd C 6 biography of Ibn Karanbi.

T he existence o f them atic chapters is a distinctive feature o f the Persian Tabaqat al-Sujiyya. A nother example is found m id-way through the biography o f Abu Sa'ld al-Kharraz, the so-called chapter on yajt (‘finding’).54 It is not im mediately obvious why it should be located here. However, the final utterance attributed to Kharraz before this chapter also m entions yajt , ‘T he beginning o f this m atter is acceptance (qabul) . .. and the end o f it is ‘finding’ (y a fiy .55 It is therefore likely that this chapter was created by the attraction to this utterance o f segments related to the topic yajt . It ends (after twelve pages) at a segment about Kharraz w hich is not related to yajt , signalling a resum ption o f the la tier’s biography.56 A similar them atic chapter is also found w ithin the biography o f D hu ’1-Nun. This consists largely o f an expanded typology o f 77m (knowledge) that is attributed to Ansari. T he same pattern occurs as in the aforem entioned examples; that is, the chapter is preceded by an utterance attributed to D hu ’1-Nun on different types o f 77m, and it ends w ith a resum ption o f his biography.57 Them atic chapters in the Persian Tabaqat al-Sufiyya thus appear to have grown from biographies; that is, they are form ed w hen a long sequence o f thematically associated segments o f addenda is generated by a segment attributed to the subject o f the biography. D ue to the considerable length o f the ‘bundle’ that is created in this way, eventually it has managed to puzzle redactors, w ho have decided to classify it as a them atic chapter. In this way, it appears to mark the close o f the preceding biography, out o f w hich it has grown, only for that to be resumed after the bundle. It is perhaps no coincidence that such them atic chapters tend to be ‘grow ths’ from the longer, and m ore im portant, biographies in the work. Biographies that attracted the most attention from compilers and redactors during the stages o f the process o f compilation, and 84

T he Persian Tabaqat a l-Sufiyya

w hich w ould carry m ore weight, were m ore prone to interpola­ tion o f material in this way T he body o f the biography o f Junayd in the Persian Tabaqat does not contain any segments at all that are variants o f those found in the body o f Sulami’s corresponding biography However, it does contain variants o f two segments that are found in the H ilyat alaw liya \ Consider the following example: T A Junayd says, the drow ning o f ecstasy (wajd) in knowledge (7/m) is better than the drowning o f knowledge in ecstasy (TABA, 169.1-2)58

H Junayd was asked, W hich o f these is m ore complete: the drow ning o f knowledge (77m) in ecstasies (wujud) or the drow ning o f ecstasies in knowledge? H e said, T he drowning o f knowledge in ecstasies. Those w ho have knowledge o f God are not like his ecstatics (wajidun ). (H, X, 275.18-20)

As a result o f the reversal o f the respective positions o f 77m and w ajd/w ujud, the above utterances convey opposite messages: the variant in the H ilya (H) suggests that the experience o f ecstasy is superior to knowledge, whilst the variant in the Persian Tabaqat (TA) suggests that the exact opposite is the case. T he existence o f such variants, attributed to the same authority, but w ith opposite meanings, is an indication o f the dynamic nature o f the process o f transmitting, recalling and re-form ing biographical material. It demonstrates the extent to w hich biographical material can serve the purposes o f later transmitters and compilers.

V As previously m entioned, most biographies in the Persian Tabaqat can be divided into two main parts: the introduction and the body. T he introductions o f those biographies for w hich there is a corresponding one in Sulami’s Tabaqat tend to be similar to their counterparts. To illustrate this point, the introduction o f the biography o f Abu Yazld is presented below, followed by its counterpart in Sulam i’s work. For ease o f com parison the introductions are divided into num bered constituent elements.

85

T he Persian Tabaqat al-S u fiyya

Amongst the first generation is Bayazid Bastami TA 1 Shaykh al-lslam said that his name is Tayfur b. cIsa b. Adam b. Surushan. TA 2 His great-grandfather, Surushan, was a Magian w ho became a Muslim. TA3 Bayazid was a com panion o f Ahm ad Khadrawayhi and he had seen Bu Hafs, Yahya M u'adh and Shaqlq Balkhl. TA 4 Shaykh al-lslam said, His teacher (ustadh) was a Kurd. H e is next to him in the grave in Bastam. Bayazid, out o f reverence for the teacher, requested, Dig my grave deeper than that o f my teacher! (i.e. lay me to rest at a lower level than my teacher). TA 5 Bayazid was a Hanafite (sahib-i ra’y buda dar madhhab ), but a [station of] sainthood (wilayat) opened up to him in w hich school allegiance (madhhab ) was not obvious. TA 6 His death was in the year 261. It is also said that it was in the year 234, but 261 is m ore correct. (TABA, 87.10-88.8)

Amongst them (the first generation) is Abu Yazld Tayfur b. cIsa b. Surushan T S 1 His grandfather, Surushan, was a M agian w ho converted to Islam. T S 2 They were three brothers: Adam, Tayfur and cAlI, and all o f them were ascetics and pietists, and were endowed w ith mystical states. T S 3 H e was from Bastam. T S 4 H e died in the year 261 according to what I heard "Abdullah b. cAlI say, w ho heard Tayfur b. cIsa ’l-Saghlr say that he heard cAmmI ’1-BastamI say that he heard his father say, Abu Yazld died in the year 261. I heard al-Husayn b. Yahya say, Abu Yazld died in the year 234. G od knows best concerning it. (TABS, 60.1-10)

The introduction o f the biography o f Abu Yazld in the Persian Tabaqat al-Sufiyya can be divided into six elements, as shown 86

T he Persian Tabaqat a l-Sufiyya

(TA1—TA6). The second and the sixth elements (TA2, TA6) appear to be almost exact Persian translations o f the corresponding elements (TS1 and TS4) in Sulami’s introduction to Abu Yazld, whilst the first element offers information about Abu Yazld’s full name, which is provided already in the heading o f the earlier biography.59 Elements TA3, TA4 and TA5 offer new inform ation. It has already been observed that the introduction to the biography o f Abu Yazld in Sulami’s Tabaqat is particularly brief, lacking any inform ation about the subject’s associates.60 These gaps are filled by TA3 and TA4, the form er providing a list o f Sufis w ith w hom he associated (aqran) and w hom he had seen (dfda bud). TA4 specifies further that Abu Yazld’s teacher (ustadti) was the Kurd w ho is buried next to him. This point is illustrated by an anecdote emphasizing Abu Yazld’s reverence for his teacher. TA5 affirms that Abu Yazld was an affiliate o f the Hanafite school (sahib-i ra’y), and implies that it may not always have been apparent that he had any such affiliation because o f the lofty station w hich he had attained. This is perhaps an attem pt to affirm that Abu Yazld, in spite o f being a controversial figure in tradition, actually had allegiance to a form o f scholastic Islam. In this case, it would perform the equivalent function to com m ents found in the H ily a ’ s introduction to Abu Yazld, w hich asserts that Abu Yazld was essentially sound, despite any appearances to the contrary.61 A com parison betw een the introductions o f the biographies o f Junayd in the two works offers similar impressions. Am ongst the second generation is Junayd b. M uham m ad b. alQ awarlrl Abu ’1-Qasim al-Zajjaj al-Khazzaz, the C h ief o f the gnostics (sayyid a l-carifin)

TA1 It is said that his father sold glassware, and on account o f that he is called qawarir (glass trader).

TA2 It is said that his origins were from Nihavand and the place where he settled was Iraq. TA3 H e was a jurist according to [the m ethod of] Bu Thawr, the best student o f ShaficI. H e used to give his (Bu Thawr) fatwas.

TA4 H e had been a com panion o f Sari Saqatl, H arith M uhasibI and M uham m ad Qassab, and he was their student (.shagivd). 87

T he Persian Tabaqat a l-Sufiyya

TA5 H e is one o f the leaders (a ’imma ) and chiefs (sadat) o f this group, and is acceptable on all tongues (maqbul bar hama zabanha).

TA6 It is said, In the world there are three individuals belonging to this generation w ho are w ithout equals: Junayd in Baghdad, Bu 'Abdullah Jalla in Syria and Bu cU thm an H lrl in Nishapur.

TA7 Shaykh Bu Jacfar Haddad says, If intellect were a man, it would be in the form o f Junayd.

TA8 He died in the year 297, and his son prayed over him on the C aliph’s N ew Year’s Day (n a yru z-i khalifa ), a Saturday. It is said that he died at the end o f Friday and was buried on Saturday. That night, som eone saw Mustafa (the Prophet) in a hurry. That person asked him, O M essenger o f God, w here are you going? H e replied, I am hurrying to the funeral o f the Caliph (khalifa), and w ent away. T he next day, that person walked around Baghdad, asking about the Caliph. H e was told that [the Caliph] was fine. Finally, he heard that Junayd had died — the Prophet had called him the ‘C aliph’ ! (TABA, 161.8-162.12)

Amongst them (the second generation) is al-Junayd b. M uham m ad b. al-Junayd Abu ’1-Qasim al-Khazzaz

TS1 His father used to sell glass and so he was called alqawarm (glass trader).

TS2 His origins are from Nihavand, but he was b orn and brought up in Iraq. I heard Abu ’1-Qasim al-Nasrabadhl say that. TS3 H e was a jurist and used to practise according to [the m ethod of] Abu Thawr. H e used to give fatwas in his circle. TS4 H e followed Sari ’1-SaqatI, H arith al-M uhasibl and M uham m ad b. cAlI ’1-Qassab al-Baghdadl. TS5 He was one o f the leaders (a’imma) and chiefs (sadat) o f the group and is acceptable ‘on all tongues’ (maqbul cala j a m f al-alsina).

88

T he Persian Tabaqat a l-Sufiyya

T S 6 H e died in 297 on the C aliph’s N ew Year’s Day, w hich was a Saturday It is said that he died in the last hour o f Friday and was buried on Saturday I heard Abü ’1-Hasan b. Miqsam m ention that. (TABS, 141.1-142.2)

T he introduction o f the biography o f Junayd in the Persian Tabaqat includes corresponding elements for each o f the six elements that make up Sulaml’s introduction (TS1- TS6), usually in the form o f Persian translations o f the original Arabic.62 It also offers two further elements (TA6, TA7), w hich are included im mediately after a remark about the authority and universal acceptance o f Junayd. These two elements are essentially discrete biographical segments that have been inserted into their respective positions by virtue o f them atic association. T hey expand on the function o f TA5, to affirm the high estimation o f Junayd in the opinion o f other Sufis. T he final elem ent in the introduction o f the biography o f Junayd in the Persian Tabaqat offers, in addition to the dates o f his death found in its earlier counterpart, a narrative related to this event. This narrative serves to underline further Junayd’s authority, by claiming that the Prophet himself, in a dream on the night o f Junayd’s death, implied that Junayd was ‘the C aliph’ to whose funeral he was hurrying. This narrative seems to have been generated by the m ention o f Junayd’s death falling on the C aliph’s N ew Year’s Day (n a yn iz-i khalifa).63 To recapitulate, the introductions o f the biographies o f Abü YazTd and Junayd in the Persian Tabaqat al-Süfiyya appear to be based closely on the corresponding introductions in Sulaml’s work. T hey are largely made up o f Persian translations o f the constituent elements o f their earlier counterparts. In this way, they are representative o f the introductions o f biographies in the later Persian w ork for w hich there are counterparts in Sulaml’s Tabaqat. This close correlation betw een the two works w ith regard to the introductions o f their biographies, contrasts sharply w ith the lack o f correlation betw een the bodies o f th eir corresponding biographies, for it has already been observed that the latter bear little relation to each other. T he Persian Tabaqat as a whole contains a few segments w ithin the bodies o f its biographies that are introduced as transmissions obtained directly from SulamI, all but one o f which are introduced 89

T he Persian Tabaqat a l-Sufiyya

as oral transmissions.64 Two o f them, in fact, are exact Persian translations o f segments found in Sulaml’s earlier work. Significantly, although they are found in the bodies o f biographies in the later Persian work, they are both originally found in the introductions o f biographies in Sulami’s w ork.65 These indications corroborate the impression that the organisational framework o f the earlier w ork was used systematically as a model; the introductions to biographies as well as their arrangement in generation sections are thus prom inent in the Persian Tabaqat, whereas there are relatively very few indications that material may have also been obtained from the bodies o f the biographies in the precursor. Thus Sulami’s Tabaqat al-Sufiyya provided the ‘architecture’ w ithout supplying a significant proportion o f the ‘buildingblocks’. As already m entioned, a num ber o f com peting m ethods o f organising material are evident in the final form o f the work, the most predom inant o f w hich is Sulaml’s generation m ethod, w hich appears to have been the last one to have been applied. Furtherm ore, the introductions o f biographies seem to have been entered at the same stage o f redaction as the generation system, on the basis o f Sulami’s model, yet they have generated further growth o f material (e.g. khalifa dream narrative) before the eventual closure o f the text.

VI In order to gain an insight into the processes by means o f w hich the Persian Tabaqat was put together, it is w orthw hile to focus on its com ponent segments. M ost o f the segments in this w ork are introduced under the rubric Shaykh al-Islam guft (i.e. ‘Ansar! said’). Examples already encountered include both com m entary on preceding material and segments about other individuals (intro­ duced by Ansarl). Such introductions suggest that segments were transm itted orally, but do not reveal m uch else about the context o f transmission. Clues to the S itz im Leben o f this material are found in the few introductions o f segments that provide further detail, such as the following examples. 1 Shaykh al-Islam dictated to us (imla kard) on the authority o f M uham m ad b. al-Junayd, w ho said . . . (TABA, 9.12)

90

T h e Persian Tabaqat a l-Sufiyya 2 Shaykh al-Islam advised, W rite dow n this segment (hikayat)

and learn it, for you will not receive anything about Shibll, on his authority, better than this segment . . . (TABA, 380.4-5)

T h e above examples refer to the transmission o f segments by Ansar! to an audience o f students. Segments were apparently read out for the students to w rite dow n (presumably in some form o f notebook). In the second example, the students are instructed to record a specific segm ent that is considered im portant by AnsarL T hus it is im plied that segments were evaluated and collected, and th at they w ere w ritte n dow n as an aid to m em ory.66 Regardless o f w h eth er they were w ritten by Ansari’s students themselves and describe w hat actually happened or not, these descriptions rem ain significant. T hey point to a context in w hich segments were dictated and entered into notes that was familiar and credible to the compilers, w ho may have been Ansari’s students, or perhaps later generations o f his successors (since the text shows m any indications o f layers o f grow th over an extended period). There is also perhaps an indication o f the S itz im Leben o f the material in the following: 3 Shaykh al-lslam said, In my box (jacba) there is this segment from Shaykh Ahm ad KufanI w hich is o f value .. . (TABA, 522.1-2)

Assuming that ja 'b a is not being used metaphorically here, it appears to be a reference to a store o f segments. It is to be expected that w ritten segments w ould have been stored in some fashion, and in an order that w ould facilitate retrieval. T he ja 'b a may be a reference to such a ‘filing box’. Segments may have been filed according to the names o f their transmitters, as the above example suggests. T he following introduction to a segment o f com m entary (about the execution o f Hallaj) is also instructive: 4 Amongst the fragments belonging to S haykh al-lslam was, w ritten in his handw riting, his diary writings, this detail . . . (dar j u z ’ha-yi Shaykh al-lslam bud bi-khatt-i way niwishta ruznam aha-yi way in mufassal . . .) (TABA, 321.1-2)

91

T he Persian Tabaqat a l-Sufiyya

This introduction to a segment has posed problems to both scribes and editors.67 It is probably best understood as a sequence o f glosses, each o f w hich is intended to give authority to the text that follows. It begins by stating that the segment was found in Ansari’s ‘fragments’ (A arjuzw ha-yi Shaykh al-Islam bud), possibly referring to scattered notes o f w ritten segments. A phrase follows immediately w hich specifies that the segment was w ritten in Ansarl’s own handw riting (bi-khatt-i way niwishta ), presumably to reinforce its authority. (It may be taken to imply that some o f Ansarl’s collection o f ‘fragments’ may have been w ritten by others). Finally the m ention o f ‘his diary-w ritings’ (ruznam aha-yi way) seems to have been tagged on for good measure. It is w orth noting that the above sentence does not claim that the material was heard from Ansarl directly. R ath er it suggests a process o f text-creation whereby segments o f material in various w ritten sources were used to create a new text. T he sentence also implies that writings attributed to Ansarl (‘fragments’ and ‘diaryw ritings’) had survived, and were perhaps even in the possession o f the glossators themselves. O ne could speculate that the writings may have been preserved by later generations o f successors, as part o f the ‘collected w ritings’ belonging to Ansarl and his tradition (perhaps even in the ‘file-box’), from w hich segments could later be extracted and incorporated into the text. In summary, the four introductions presented above describe the transmission o f segments during teaching sessions. Ansarl dictates from his own collection o f w ritten segments. T he students do not just listen and memorise; they also take notes and thus presumably build up their own collections o f segments. Notes, w hich are taken dow n by Ansarl himself, and his students, both becom e incorporated into the final work. Introduction 2 implies, w ith regard to note-taking, that students did n o t w rite dow n everything they heard. Like students o f any age they norm ally selected for themselves w hat to record, although occasionally their teacher may have instructed them to note dow n a specific segm ent that he considered im portant. This underlines that it is norm ally the student w ho judges w hat he hears, and decides w h eth er to record a segm ent or not. Fie is therefore the author o f his ow n notes, even if they happen to be based on the teachings o f others. In both form and content these notes indicate the predilections o f the one w ho has taken them . 92

T h e Persian Tabaqat a l-Sufiyya

Ansari is quoted in Example 2 as saying to his students, ‘You will not receive (nayarand shuma-ra) anything about ShiblT, on his authority,68 better than this segment . ..’ It perhaps implies that students obtained material for their notebooks from different sources in the pursuit o f as many w orthw hile segments as possible, and highlights further hazards involved in regarding a student’s notes as the w ork o f his teacher. They can only be regarded so, w ith any degree o f confidence, if the teacher dictates a w ork from beginning to end to his student, and then checks over and redacts the notes him self into a canonical form .69 This is clearly not the case w ith the Persian Tabaqat al-Sufiyya. T he inconsistencies, incongruous juxtapositions and frequent breaks in the text all indicate that it was not subject to these kinds o f controls. N orm an Calder makes similar observations, regarding the use o f notes as an aid to oral activity. His references to their use in the fields o f adab and fiq h parallel what is found here, including the following comments from Jahiz’s Kitab al-Bayan : ‘A word preserved in your heart is better than ten in your notebook’, ‘A man writes down the best that he hears, and he memorises the best that he writes down.’70 Such segments point to the similar use o f notebooks at a (much earlier) time w hen cultural activity was still predominantly oral. Calder has argued that the transition to a literary milieu was effected largely ‘through the m ediation o f notebooks’.71 T he Persian Tabaqat al-Sufiyya would appear to be an example o f a later process o f textcreation from teaching session notes.

V II To recapitulate, taking into consideration the observations made in the previous sections o f this chapter regarding this highly problematic text, one can say w ith confidence that the Persian w ork traditionally know n as Ansarl’s Tabaqat al-Sufiyya is an ‘unauthored’ work. It is essentially a com pilation o f heterogenous segments o f material, w hich have been extracted from the notes o f a num ber o f individuals, and at different times. T he text has also grown over an extended period after the initial com pilation. It shows m ore signs o f interpolation and interference than the H ilyat al-awliya \ w hich itself appears to have grown for up to two generations after its nam ed ‘author’.72 T he Tabaqat al-Sufiyya was probably com piled in one location, namely Herat, since the use o f 93

T he Persian Tabaqat a l-Sufiyya

an archaic form o f Persian, w hich is associated w ith that region in the 11th and 12th centuries, predominates throughout. In view o f the im portance o f the teaching context for the collection o f the material, it w ould not be too far-fetched to suggest that this is a ‘school-text’ —a text that became the repository for the expression o f the views o f Ansarl’s successors in Herat. If it came to be rem em bered as his work, it is probably because he was looked back upon as the founder o f the school, on whose teachings the work may originally have been based. T he process by w hich this w ork eventually reached its final form appears to have involved at least two major redactions. This is evident from the arrangem ent o f biographies, for w hich the main organising principles appear to have been applied in succession. As already m entioned, it seems that the biographies were originally ordered according to the names o f their subjects. Later, they were rearranged (overlooking a few cases w hich now create incoher­ ency) according to a generation system derived from Sulami’s Tabaqat al-Sufiyya. Indeed, it is presumably on account o f this generation (tabaqa) system that the Persian work was also nam ed Tabaqat al-Sufiyya.

T he introductions o f biographies in Sulaml’s earlier w ork also appear to have been used as the basis for the corresponding introductions in this work. This was probably carried out at the same stage as w hen the generation system was used as a model, since introductions serve as demarcators for the overall organisa­ tional framework. Occasionally they appear to have been inserted incorrectly; this occurs w here the redactor has presumably mistaken the mere m ention o f som eone’s name in a segment for the start o f a new biography, and has consequently inserted an introduction. It confirms that they were added in a later redaction, rather than constituting the demarcations o f the original frame­ work, and they corroborate the impression that Sulaml’s Tabaqat was relied upon only at a later stage, and not as the starting-point o f the Persian work. As already m entioned, one effect o f the misapplication o f introduction demarcators is the creation (at least ostensibly) o f more than one biography about the same individual. For example, there are two introductions to biographies o f Abu ’l-cAbbas al-Surayj (d.306/918), the Shafi’ite jurist o f Baghdad.73 T he first o f these marks ostensibly the start o f the second biography to follow that o f Junayd. It continues with the narrative which is presented below. 94

T he Persian Tabaqat a l-Sufiyya

O ne day he (Surayj) passed by the edge o f Junayd’s session, listened and w ent away Som eone asked him, H ow did you find Junayd? H e replied, I do not know anything about

the secrets o f this group except that this Shaykh has an awe-inspiring presence (satvla ) w hich is not like that o f the vain.74 (TABA, 184.15—185.3)

T he above narrative appears to function to confirm the authority o f Junayd, by attributing a positive opinion o f him to the jurist Surayj. It consists o f the same m o tif as the aforem entioned narrative in w hich cAbdullah Sacid Kilab, a renow ned ascetic, listened to Junayd and then gave a positive estimation o f him .75 In this way, it is as equally relevant as that narrative for the biography o f Junayd, w hich comes to a close only a few lines o f text earlier. It was probably indeed intended for it, but eventually became mistaken for an independent biography, thus acquiring its own dem arcatory in tro d u ctio n in a later redaction. T h e o th er introduction to a biography o f Surayj is followed by a narrative w hich functions to give authority to Surayj himself, by describing him as putting a questioner into a trance (way bang bi-kard w a ’z hush bi-shud).76

W hilst it appears that Sulami’s generation system for arranging biographies, as well as his introductions to biographies, were used as a m odel in a late redaction, it is probable that the text continued to be subject to organic growth even after this stage. O ne o f the indications o f this is found at the start o f the body o f Junayd’s biography, w hich includes segments that appear to have been generated by the preceding (Sulaml-based) introduction, such as the dream narrative about the Prophet, generated by the preceding m ention o f ‘the C aliph’s N ew Year’ (see above). A nother segment that has been inserted to a position towards the start o f the biography, presumably on account o f this remark, is the following: Shaykh al-Islam said that the Caliph o f Baghdad said to R uw aym , H ey ill-m annered one! [Ruwaym] responded, I am not ill-mannered, I have kept company w ith Junayd!77 (TABA, 163.9-11)

T he above narrative describes R uw aym b. Ahm ad (d.303/915—16) as confidently contradicting the Caliph because o f his high estimation o f the virtue o f com panionship w ith Junayd. It is thus 95

T he Persian Tabaqat a l-Sufiyya

implied that Junayd’s authority is greater than that o f the Caliph o f Baghdad, at least in the eyes o f the Sufi R uw aym b. Ahmad. In this way the narrative is similar to the aforem entioned one in w hich the Prophet calls Junayd khalifa. It is likely then, that this too was generated ultimately by the introduction o f the biography. T he Persian Tabaqat al-Sufiyya ends w ith a series o f segments related to ma'rifa (gnosis). These segments concern different individuals, most o f whose biographies are included earlier in the w ork.78 This curious section, w hich makes up nearly a tw entieth o f the whole work, appears to have prom pted the scribes o f later manuscripts to give it a separate heading Fasl ft ’l-m a crifa w a -’ltawhfd (Section on gnosis and divine unity).79 It is not immediately clear why these segments have been positioned here w hen they could have been filed into the appropriate biographies in the main body o f the work. It may be the case that a special collection o f segments related to the topic o f ma'rifa had been m aintained, and that it was decided to keep these segments separate from the biographies. It is however also possible that this section was added after the whole text had reached a closed form; it w ould therefore have been positioned at the end as an appendix, even though its com ponent segments are actually appropriate for the biographies in the main body o f the work. In this way, it would parallel the appendices at the end o f certain biographies in the H ilyat al-awliya and perhaps represent a final stage o f entry o f segments into the organic school text o f Herat, w hich became know n as ‘Ansari’s’ Tabaqat al-Sufiyya.

96

Pa r t Two

“You tell me that battles are reproduced. I do fin d something aesthetic, ju s t as you said, in seeing beneath a modern battle the plan o f an older one; I can’t tell you how attractive the idea sounds. B u t then, does the genius o f the commander count fo r nothing? Does he really do no more than apply the rules?” (Marcel Proust, Remembrance of things past, Vol II, p. 113)

This page intentionally left blank

C h a p t e r Four

Qushayri’s Risàia

I

A

bu ’1-Qasim cAbd al-Karlm b. Hawazin al-Qushayri (d. 465/ . 1072) was born in the year 376, according to al-Khatib alBaghdadl, w ho claims that Qushayrl him self inform ed him directly o f this date. It w ould imply that he lived for nearly ninety years.1 T he Khatib also refers to the following visit by QushayrT to Baghdad: H e came to us in the year 448. H e transm itted hadiths in Baghdad and we w rote them down on his authority —he was trustw orthy (thiqa ). H e was good at preaching and eloquent in instructing. H e was educated in theology (usuI) according to the Ash’arite school, and in jurisprudence (furuc) according to the Shafi’ite school. (al-Baghdadl,

T a’rikh,

XI, 83.5-8)

It is perhaps a little surprising that, despite inform ing that QushayrT was a trustw orthy hadith transmitter, and a religious preacher w ho followed Ash’arite theology and Shafi’ite law, he does not indicate that QushayrT was also a Sufi.2 However, his contem porary the Sufi scholar, cAlT b. cU thm an al-H ujw m , includes a biography o f QushayrT in his K a s h f al-mahjub w hich presents him as a Sufi authority, whilst also acknowledging that he excelled in a wide range o f intellectual disciplines.3 T h e earliest narrative o f QushayrT’s life is found in al-Siyaq liTa’rTkh N aysabur o f cAbd al-G hafir b. IsmacTl al-FarisT (d.529/ 1134).4 T h e m ain events in QushayrT’s life, according to this 99

Q u sh ay n ’s R isala

biography, are sum m arised as follows: he was b o rn am ongst the Arabs w ho had settled in U stuw w a in K hurasan.5 O n visiting N ishapur as a young m an he attended a m eeting led by Abu "All ’1-Daqqaq (d. 406/1016 or 412/1021) and becam e his disciple. T he latter instructed him to study the religious sciences; he proceeded to study jurisprudence under the Shafi’ite M uham m ad b. Bakr al-Tusi (d.420/1029), and theology under the Ash’arite Abu Bakr b. Furak (d .4 0 6 /1 0 1 5). Later he perform ed hajj w ith A bu M uham m ad al-Juw aynl (d.438/1047) and A hm ad alBayhaql (d.458/1066), and together they obtained the authority to transm it hadiths by sa m a c (sam ica . . . al-hadtth) from the leading authorities in the Hijaz and Baghdad (presumably en route).6 QushayrI com piled al-TafsTr al-K ablr and o ther works w hich were widely circulated. All the later biographies o f QushayrI appear to be based on this ‘skeleton’ narrative. M ore details accumulated over time, especially regarding his death, his teachers and his w ritten compositions. T hey include motifs w hich have already been encountered in the biographical traditions o f Abu Yazld and Junayd, including perseverance in worship in the final m om ents o f life, and burial deeper in the ground than on e’s teacher.7 Twenty extant works have been attributed to QushayrI, including both Sufi and (Ash’arite) theological works.8 It is mainly due to his scholarship that he is rem em bered as an im portant figure in the Sufi biographical tradition, and his most famous work, the R isala , has proven to be one o f the most widely read works in the history o f Sufism.9

II T he Risala is a com bination o f the two main genres o f Sufi literature, namely the tabaqat and manual genres.10 It can be classified as ‘dual-generic’, since it incorporates these two genres in the form o f its main com ponent sections.11 In total, the Risala consists o f four sections w hich are demarcated by the author himself, using headings and introductions.12 Biographies are included in the second section, w hich represents the tabaqat genre (see Section III o f this chapter), whilst the third and fourth sections, in com bination, constitute the equivalent o f a Sufi manual (see below). As A rthur A rberry has already observed, the 100

Q u sh ayn ’s Risala Risala draws heavily on both Sulami’s Tabaqat al-Sufiyya and Abu Nasr al-Sarraj’s Sufi manual, the L u m a ', and yet it has managed to

surpass them in popularity. It w ould not be an exaggeration to suggest that most readers are familar w ith Sulami’s Tabaqat prim arily through Q ushayrl’s use o f it as a source for the tabaqat section o f his ow n work. This makes it all the m ore im portant to exam ine closely the ways in w hich QushayrI selected and reworked material from that precursor for his ow n purposes. Before turning attention to the tabaqat section, it is necessary to appreciate the context in w hich it is em bedded, by considering the nature o f the other sections o f the Risala. T he shortest section by far is the first section,13 on ‘the doctrine (itiqad) o f this sect (ta’ifa) in theological issues (masa’il a l-u su t)\ w hich immediately precedes the tabaqat section. It reveals Qushayrfs allegiance to Ash’arism, since it serves to attribute to Sufism the theological doctrines o f that school.14 T he introduction o f the term inology section o f the R isala , w hich follows immediately after the tabaqat section, indicates a change in orientation towards a readership o f Sufi practitioners (min salikf turuqihim w a-m uttabicT sunnatihim ) .15 It even suggests that the use o f an exclusive term inology serves the purpose o f obscuring the m eaning o f Sufi discourse from outsiders (li-takun m a'am alfazihim mustabhama cala }l-ajanib ).16 Such com m ents at this particular point in the Risala suggest that the preceding sections (consisting o f the tabaqat and the short Ash’arite theology section) are designed for a w ider audience including the uninitiated, in contrast to the subsequent Sufi manual, com prising the term inol­ ogy and ‘systematic’ sections (see below). A total o f forty-eight Sufi term s are introduced in the tw entyseven chapters that make up the term inology section.17 Q ushayrl’s m ethod is to com bine his ow n explanations w ith utterances attributed to past Sufi authorities, as well as occasional citations o f verses o f the Q u r’an and hadiths. T he term inology section leads up to the systematic section w hich is introduced by means o f the following short introduction: We will now m ention chapters (abwab) explaining the stations (;maqamat) that are the routes o f the wayfarers, then, after that, chapters detailing the states (ahwat) to the extent that G od facilitates by his grace. (R, 167.18-20)

101

Q ushayri’s Risala

T he chapters (abwâb )18 that make up this section are here introduced as belonging to two consecutive groups, namely chapters on stations (maqâmât) followed by chapters on states (ahwâl). However, the actual topics covered by these chapters are m ore wide-ranging; they include topics, such as sam ac (musical audition) and ahkâm uhum f t ’l-safar (their (the Sufis’) rules during travel), w hich can hardly be classified under either category He appears to have chosen to divide his discussions o f topics that were conventionally covered in the Sufi manual genre into two broad groups —the first constitutes his section on terminology, whilst the second constitutes the ‘systematic’ section, w hich encompasses, in addition to the ‘stations’ and ‘states’ o f the mystic itinerary, discussions o f selected Sufi customs and practices. Although the dem arcation betw een these two sections consists o f no m ore than the short introduction presented above, a transition can be seen in the form at o f the (fifty-one) chapters that follow it, from that o f the preceding term inology chapters. Qushayrl now invariably begins w ith a citation from the Q u r’an, followed in turn by a prophetic hadith, his own com m ents and utterances attributed to Sufi authorities o f the past.19 T he final chapter o f the R isala , ‘[The C hapter of] advice to disciples (al-wasiyya li-’l-m u n d ïn )\ is not demarcated at all from the section w hich precedes it. In view o f the fact that the systematic section includes a few ‘miscellaneous’ chapters in addition to those on stations and states, it should perhaps be no surprise that this final chapter o f advice has usually been classified as its final member. However, since it fails to follow the same structural pattern as the preceding fifty-one chapters w hich make up the systematic section, it is perhaps better understood to be separate from them, as a form o f appendix.20 T he introduction to the Risala offers the date o f the com pletion o f its com position (kataba ) as 437,21 whilst the postscript gives the date o f the com pletion o f its dictation (im lâ’) as 438.22 These dates therefore suggest that the Risala was eventually dictated from a master copy that had been com posed a year earlier. T he overall consistency in style and m ethod evident in this w ork indicate a relatively high degree o f authorial control, and it seems reasonable to accept these dates as representing the com pletion o f the work by Qushayri, and its dictation a year later, respectively.

102

Q ushayri’s Risala

III T he tabaqàt section o f the Risàia consists o f an introduction, 83 biographies and a conclusion. The biographies are presented in a continuous sequence, beginning with Ibrahim b. Adham (d.162/ 778-9) and ending with Ahmad b. cAta ’l-R udhbari (d.369/980). Whilst there is an obvious chronological progression from the start o f the sequence to the end, individual biographies are not listed in strict chronological order. 80 o f the (83) biographies correspond to those found in Sulami’s Tabaqàt al-Sufiyya , the (103) biographies o f which are grouped into ‘generation’ sections.23 Moreover, in spite o f the fact that the biographies in the Risàia are presented as a single sequence, rather than divided into generation sections, their order o f presentation indicates the influence o f the precursor. For example, all the biographies in the Risàia which correspond to Sulaml’s first generation are grouped together at the start o f the sequence, before the biography o f Junayd. In the earlier work, his biography signals the start o f the second generation section. In the Risàia too, it marks the point in the sequence, immediately after which follow those biographies w hich correspond to Sulami’s second generation section.24 Similarly, the biographies o f JurayrT, Shibll and Ibn A'rabI signal, within the continuous sequence, the ‘starting-points’ o f biographies corresponding to Sulaml’s third, fourth and fifth generations, respectively. In that work, their biographies are the first to be listed in the third, fourth and fifth generations. It should also be m entioned that many o f the biographies in the Risàia are not only positioned in the sequence in accordance with their classification by SulamI, but they are even juxtaposed in exactly the same order.25 The degree o f influence from Sulaml’s generation system is so high that, w ithout prior familiarity with that work, the criteria for ordering biographies in the Risàia would be indecipherable. O nly three biographies that are included in QushayrT’s Risàia do not have counterparts in Sulaml’s Tabaqàt. T hey are the biographies o f D àw ud al-Ta’I (d. 165/781), Ahm ad al-Zaqqaq al-Kablr (d.290— 1/902—4) and Abu cUbayd al-Busrl (d.238/852). Despite the fact that biographies devoted to them are not found in the earlier work, each o f them is m entioned there.26 Therefore, it appears that none o f these individuals were necessarily unknow n to SulamI, yet they did not m erit a biography according to his criteria. T heir im portance was evidently greater to Qushayrl, whose w ork was com piled at least a quarter o f a century later. 103

Q u sh ay n ’s Risala

All three o f the additional biographies have a factor in co m m o n : th e ir first isnad nam es SulamT, referred to as M uham m ad b. al-Husayn, as the im m ediate source.27 T he presence o f such isnads at the start o f these additional biographies has the effect o f m aking them seem to be based on the authority o f SulamT, despite the fact that they are not found in his Tabaqat. It facilitates their integration w ith the rest o f the biographical section, the organisational fram ework o f w hich has evidently been m odelled on that work. T he biography o f BusrI (d.238/852—3) is juxtaposed betw een those o f Samnun b. Hamza (d.ca 297/910) and Shah al-K irm anl (d.ca 300/912), Sufis o f at least one generation later. This position is conspicuous since usually biographies o f contem poraries tend to be juxtaposed in this w ork, rather than those o f different generations. T he reason for its position can be found in the introduction to the biography o f KirmanI, w here it is stated that he followed Abu cUbayd al-Busri.28 O f the 103 biographies found in Sulaml’s Tabaqat al-Sufiyya , only 25 are om itted in the Risala. M ost o f these are accounted for by the final members o f each o f Sulaml’s five generations: the final four biographies from the second generation, the final two from the third, the final four from the fourth and the final seven from the fifth are all om itted.29 This suggests that the earlier w ork was used systematically, in such a way that once enough biographies had been included from each o f Sulami’s generations in turn, then the rem ainder were simply om itted. O u t o f the remaining biographies to be om itted, that o f Hallaj is the most significant. T he omission o f Hallaj is particularly curious since many segments about him are found in the other sections o f this w ork.30 This suggests that his omission was not merely for the sake o f brevity. Since a decision was made to om it Hallaj from the tabaqat section, whilst leaving segments about him in other sections o f the R isala , it implies that his inclusion w ould have conflicted w ith the aims particular to the tabaqat section o f the work. Since the biographies in Q ushayn’s Risala correspond to those o f Sulami’s w ork they cover the same period, from the eighth century to the late tenth century. As already observed, SulamT provides an explanation for the fact that his biographies do not extend back as far as the time o f the Prophet.31 Similarly, QushayrI in the introduction to his biographical section links Sufis from the eighth century back to the time o f the Prophet. 104

Q ushayri’s Risala

Be aware that the best o f the Muslims after the apostle o f God were not called, in their own time, by a title o f distinction other than companionship o f the apostle o f G od (.suhba), since there was no virtue higher than that. Thus they were called the sahaba and w hen those o f the second period took over from them, those w ho had associated w ith the sahaba were called the tabicun , this being considered the most noble title. T hen those w ho came after them were called the atba al-tabicm. T hen the people were at variance and different ranks became discernible. T he elite o f the people amongst those w ho were preoccupied w ith religion were called the zu h h a d (ascetics) and cubbad (pietists).Then innovations emerged and challenges were made between the groups, each o f them claiming that the zu h h a d were amongst their number. T he elite o f the traditionalists (khawass ahl al-sunna ), w ho maintained their souls w ith God and safeguarded their hearts from the paths o f heedlessness, alone possessed the name tasaw wuf (Sufism). This name became well-known for these great individuals by 200 A.H. In this section we will m ention the names o f a group o f the shaykhs o f this sect from the first generation until the time o f the later ones amongst them, and briefly m ention their behaviour and sayings which contain an indication o f their principles and their customs, God willing. (R, 34.1-8)

Qushayri links Sufism back to the time o f the Prophet by asserting that the representatives o f this tradition were ultimately his successors. They are identified as the elite religious devotees from amongst the traditionalists, w ho succeeded the three generations o f the ‘pious predecessors’ (the sahaba, ta b iu n and atbac al-tabicTn), as well as the ‘original’ zu h h a d and cubbdd. This m odel is an elaboration o f that provided by Sulami in the introduction to his Tabaqat a l-Sufiyya , w here he states that the first Sufis were the successors o f the pious predecessors. If Q ushayri’s version differs, it is because o f his concern to justify the use o f different titles in different periods. T he term tasaw w uf is said to have been wellknow n only by the third century. Before this period, the elite religious devotees, that is, the predecessors o f the Sufis, had apparently been called zu h h a d and cubbad , but eventually each o f the com peting groups that em erged claimed to represent the zu h h a d , creating the need for the new name, tasaw w uf 105

Q u sh ay n ’s R isala

Qushayrl, by referring to the differences in opinion and challenges betw een com peting groups, as well as the beginning o f innovations, shows greater concern than his predecessor to distinguish Sufism from com peting traditions. H e defines the first Sufis as an elite am ongst traditionalists w ho were focused com pletely on God. This differs significantly from Sulami’s definition o f Sufis as ‘those w ho have mystical states (ahwal) and speak about unicity (tafrid), the truths o f unity (tawlnd) and the application o f the m ethods o f detachm ent (ta jrid )\2>2 It is a sign that Qushayrl wished to re-orientate the definition o f the identity o f Sufis in the tabaqat section o f his R isala , classifying them as pious traditionalists rather than mystics. It is stated in the above passage that the tabaqat section will include Sufis ‘from the first generation (tabaqa) until the tim e o f the later ones’. It actually begins w ith those w ho were classified as members o f Sulaml’s first generation o f Sufis, whilst it culminates w ith the latter’s fifth generation. Qushayrl, w ho lived a generation later than SulamI, also includes, after the final biography, the following list o f Sufi contemporaries: T he shaykhs w hom we have reached and w ith w hom we are contem poraries, even if we could not m eet them , such as, the master and witness, the spokesman o f his time w ho was unique in his era Abu cAli ’1-Hasan b. al-Daqqaq, Shaykh Abu cAbd al-R ahm an al-Sulaml, w ho was unique in his time, Abu ’1-Hasan cAlI b. Jahdam w ho lived in Mecca, Shaykh Abu ’l-cAbbas al-Qassab33 in Tabaristan, and Ahm ad alAswad in Dlnawar, Abu ’1-Qasim al-Sayrafl in Nishapur, Abu Sahl al-Khashshab al-Kablr also in Nishapur, M ansur b. Khalaf al-M aghribi, Abu Sacd al-Mallnl, Abu Tahir alK huzandl and others. If we had occupied ourselves w ith m entioning them and detailing their characteristics we would have departed from the aim o f conciseness. (R, 120.1-8)

T h e above list o f ten Sufis serves as an extension o f the preceding sequence o f biographies. It is explained that biographies had not been included for them in that sequence only due to a concern for conciseness. N onetheless, the provision o f this list at this p o in t implies that they are QushayrT’s selected successors o f the earlier generations o f Sufis, from am ongst his ow n im m ediate predecessors. 106

Q u sh ay n ’s R isala

Sulaml’s generation system culminates w ith his ow n teachers and predecessors in Sufi scholarship in the form o f his fifth generation section. Similarly the aforem entioned extension o f the sequence o f biographies in Qushayri’s tabaqàt section indicates his ow n preferences and allegiances. T he predom inance o f K hur­ asanians is striking. M oreover, just as Sulaml’s fifth generation is hierarchical, in that m ore space is devoted to the biographies placed at the head, so too is Q ushayrl’s final list, in that the first two individuals to be named, Daqqàq and SulamI, are the only ones afforded glorious epithets. Q ushayri’s ow n teacher in Sufism and his predecessor in Sufi scholarship are thus presented as the most im portant members o f the generation immediately preceding his own. Daqqàq is the most frequently quoted Sufi in the R isàia , whilst SulamI is the most frequently nam ed im mediate source o f segments in the w ork (as well as the provider o f the m odel for the tabaqàt section as a whole). Thus, in the context o f the Risàia itself, this tabaqàt section serves as an isnàd for the matn o f the Sufi manual, w hich makes up the larger part o f the work.

IV T he organisational framework o f the tabaqàt section o f Q ushayri’s Risàia provides the most conspicuous evidence o f his heavy reliance on Sulaml’s Tabaqàt al-Sufiyya. This is found not only in the selection and order o f presentation o f biographies, but also in the (demarcatory) introductions o f the individual biographies. To illustrate this point, the introduction o f the biography o f Abu Yazld in the Risàia is presented below: Abu Yazld Tayfur b. cIsà ’l-Bastaml. His grandfather was a M agian w ho converted to Islam. They were three brothers: Adam, Tayfur and cAlI. All o f them were pious ascetics and Abu Yazld was the greatest o f them w ith regards to mystical state. It is said that he died in the year 261. It is also said that it was 234. (R, 55.1-5)

T he introduction o f the biography o f Abu Yazld in the Risàia presents, in the same order, most o f the inform ation found in the corresponding introduction o f Sulaml’s Tabaqàt.34 However, it is m ore concise, in keeping w ith the tabaqàt section as a whole; it 107

Q ushayrl’s R isala

omits the isnàds on the authority o f w hich alternative dates for Abu Yazld’s death are originally given, and it offers the nisba 'alBastaml’, facilitating the omission o f a statem ent about the subject’s place o f origin. Qushayrl has also om itted, apparently for the sake o f brevity, the Persian nam e o f Abu Yazid’s grandfather, w ho is said to have converted from Zoroastrianism. T he statem ent characterising Abu Yazld and his brothers together in the introduction o f Sulami’s Tabaqàt (w a-kulluhum kànu zuhhàdan wa- cubbàdan wa-arbàb al-ahwàl) has been reworked in the R isala , where it reads: wa-kàna A b u Yazxd ajallahum hàlan. This is an im provem ent on the earlier version w hich is anomalous for characterising its subject only alongside his brothers. T he introduction o f the biography o f Junayd in Qushayrl’s R isàia , w hich is presented below, shows further signs o f reworking. Abu ’l-Q asim al-Junayd b. M uham m ad is the chief o f this sect and their leader (sayyid hàdhihi ’l-tà ’ifa w a-im am uhum ). His origins are from Nihavand and the place o f his birth and upbringing is Iraq. His father used to sell glassware and so he was called al-qawàrm ‘the glass trader’. H e was a jurist according to the school o f Abu Thawr, and he used to give fatwas in his circle, w ith him present, w hen he was twenty years old. H e associated w ith his m aternal uncle al-Sari and al-H àrith al-M uhàsibl and M uham m ad b. cAlI ’l-Qassab. He died in the year 297. (R, 70.15-71.3)

Similar to the introduction o f the biography o f Abù Yazld, the introduction o f the biography o f Junayd in the Risàia is a more concise version o f its earlier counterpart.35 Each o f the items o f inform ation that it contains appears to be derived from the earlier version,36 but some o f them have been reworked significantly. First o f all, the introduction in the Risàia begins w ith the statement that Junayd is the leader o f the Sufis. This is based on the m ore m odest statem ent towards the end o f the earlier version w hich informs that Junayd was one o f their leaders. B oth the reworking o f this statem ent and the change o f its position to the start suggest that Junayd’s supreme position amongst Sufis had becom e firmly established by Q ushayrl’s time. (Sulamfs Tabaqàt al-SUfiyya itself w ould have played a part in this process). T he statement in the Tabaqàt w hich serves to attribute authority in jurisprudence to Junayd also appears to have been reworked in 108

Q u sh ay n ’s Risala

the R isala : it has been extended by the addition o f two further details about Junayd’s delivery offatwas. T he addition o f bi-hadratih clarifies that Junayd gave fatwas w ith the authority o f Abu Thawr, since it affirms that the latter was present at the time; the addition o f wa-huwa }bn cashnn sana emphasises the precocious nature o f Junayd’s acquisition o f authority in jurisprudence — apparently, he was only tw enty years old at the time. T he introductions o f the biographies o f Abu Yazld and Junayd are representative o f the vast m ajority o f the introductions o f biographies in the R isala , in that they dem onstrate that QushayrT relied heavily on Sulaml’s work; he produced m ore concise versions, w hich were reworked according to his ow n preferences. T he re workings in these introductions serve to embellish and emphasise the credentials o f their respective subjects; Abu Yazld is now distinguished from his brothers as being the greatest mystic out o f the three, whilst Junayd is no longer introduced as merely one o f the Sufi leaders, but rather the leader o f this sect, sayyid hadhihi ’l-ta ’ifa . M oreover, he was apparently so precocious as a jurist that he m erited the authority to give fatwas in the presence o f Abu T haw r at a relatively young age. In the next section o f this chapter, attention will be turned from the organisational framework to the com ponent segments o f material o f the biographies, in order to assess accurately the nature and extent o f the influence o f SulamT’s Tabaqat (and other earlier works) on the R isala . Especially in view o f the deceptive nature o f the influence o f that w ork on the Persian Tabaqat, it is necessary to appreciate the relationship betw een the ‘building-block’ segments o f the two works, as well as their ‘architecture’.

V O u t o f the nine discrete segments w hich make up the body o f the biography o f Abu Yazld, five are demarcated by an isnad ,37 one by the formula wa-bi-hadha ’l-isnad (‘also on the authority o f the (preceding) isnad’) ,38 one by the m ention o f the first two transmitters only,39 and the remaining segments by the simple conjunction wa-.40 N o n e o f them indicate that a w ritten source has been used, even though many o f them are found in earlier works. For instance, the first two segments are virtually identical to corresponding segments in Sulami’s biography o f Abu Yazld.41 In 109

Q u sh ayn ’s Risala

fact, it is only their isnads that differ, and merely by the m ention in both o f Qushayrl’s isnads o f ‘M uham m ad b. al-H usayn’ (SulamI) as the im mediate source (added to the end o f the same isnads that are found in the latter’s own w ork).42 Oral transmission from SulamI is thus implied, even though the segments are found in both works w ithout any discrepancy. T he first segm ent describes Abu Yazld’s m ethod o f Sufism as one o f self denial, whilst the second attributes to him a concern for scholastic know ledge and its application.43 In the latter, Abu Yazld expresses his appreciation for the accepted difference o f o p in io n am ongst the scholars (ik h tila f a l- cu la m a ’) .44 T hese segments have been selected from the m any available in the earlier w ork, w hich includes pithy apophthegm s and w itty dismissals o f com peting types o f religious devotees. T h eir selection signals that Q ushayri prefers to depict Abu Yazld as a pietist w ho appreciated the activity o f religious scholars, and they have even attracted a gloss to underline this message — it states that ‘Abu Yazld did n ot depart from this w orld until he had m em orised the entire Q u r’an’.45 T here are no further segments in the biography o f Abu Yazld that are also found in Sulamfs Tabaqat, suggesting that the latter w ork was used as the starting basis for the com pilation o f the biography; that is, it was drawn upon for the introduction and the first two segments, after w hich further sources were used.46 Sulaml’s Tabaqat is only the first o f three major Sufi works that appear to have been used as sources for the biography o f Abu Yazld. T he third and fourth segments, for instance, are variants o f two juxtaposed segments in the chapter on ‘w hat has been m entioned about the (Sufi) shaykhs regarding their following (ittiba ) o f the apostle o f G od’ o f Abu Nasr al-Sarraj’s (d.378/988) Kitab a l-L u m a c. These segments are presented below: 3 Abu H atim al-Sijistani related to us, saying that Abu Nasr al-Sarraj related to him, saying that he heard Tayfur alBastaml say that he heard the one know n as cAmmI ’1Bastami say that he heard his father say that Abu Yazld said to him, Let us go to see this man w ho has made him self famous for proxim ity to God. H e was a sought after man, famous for asceticism (z u h d ), so we w ent to him. W h en he came out o f his house and entered the m osque he spat towards the qibla (direction o f prayer), and so Abu Yazld

110

Q ushayri’s Risala

turned away and did not greet him, saying, This is unfaithful to one o f the manners (adab) o f the apostle o f God, so how can he be faithful about w hat he is claiming? 4 O n the authority o f this isnad Abu Yazld said, I intended to ask G od to spare me from the need for the provision o f food and w om en, then I said, H ow can it be perm itted for me to ask G od for this w hen the apostle o f G od did not ask him for it? So I did not ask him. T h en G od spared me from the need for the provision o f w om en, such that I do not care w hether a w om an meets me or a wall ! (R, 55.15-24)47

Q ushayrl’s variants are offered on the authority o f the same isnad , just like the juxtaposed pair in the L u m a \4S T hey appear to have been obtained from Sarraj’s Kitab a l-L u m a \ even though only oral transmission (via an interm ediary) is implied by their isnads. T he failure to acknowledge the w ritten source is typical for the R isala , w hich has already been observed to owe a considerable debt to bo th Sulami’s Tabaqat and Sarraj’s L u m a c.49 In Segment 3, the claim to sainthood o f one w ho fails to follow an aspect o f the P rophet’s behaviour is rejected. In Segment 4, Abu Yazld admits that he had wished to go beyond the norm ative example o f the Prophet in asceticism, but then thought better o f it, and was consequently rewarded by the granting o f his original wish. This pair are com plem entary segments; the form er implies that one must not violate any aspect o f the P rophet’s behaviour, whilst the latter implies that it is better to confine oneself to following the P rophet’s example than to undergo extra self-denial. B oth segments function to attribute to Abu Yazld the opinion that the correct religious path is following the sunna (normative example) o f the Prophet, in favour over both falling short and striving beyond it. T he biography o f Abu Yazld also includes three segments that are variants o f corresponding segments in the H ilyat al-awliya’. Two o f them are juxtaposed in both works (in reverse order).50 T he first o f these segments is presented below, followed by its variant in the earlier work: R Abu Yazld was asked, W hat was the most difficult (ashadd) thing you encountered on G o d ’s path? H e said, It is not possible to describe it. T hen he was asked, W hat was the

111

Q ushayrl’s R isala

easiest (ahwan)5] thing that your carnal soul (nafs) encoun­ tered from you? H e said, As for this, yes [it can be described]: I sum m oned it to perform an act o f obedience (taa) and it did not accede to my request so I denied it water for a year. (R, 56.10-12)

H I heard Abu ’l-Hasan say that he heard Abu ’l-Hasan alMarwazI say that he heard Abu Yazld’s wife say that she heard Abu Yazld say, I urged my carnal soul (nafs) to go to God, but it refused me and was stubborn, so I left it behind and proceeded to God (fa-taraktuha wa-m adaytu ila ’Halt). (H, X, 36.21-3)

In both variants, Abu Yazld relates that he had asked som ething o f his carnal soul (nafs), but, since it disobeyed him, he punished it. However, they specify different tasks asked o f the carnal soul as well as the eventual punishm ents for disobedience. These details are surely not insignificant. In the H ilya Abu Yazld is portrayed as a mystic w ho left behind his carnal soul for ‘refusing to go to G o d ’, and proceeded w ithout it (fa-taraktuha wa-m adaytu ila ’llah). In the R isala , however, Abu Yazld is portrayed as a self-denying pietist w ho ‘deprived his carnal soul o f water for a year’ because it had not fulfilled the requested act o f obedience (ta ca). This transformation o f a segment from one w hich portrays Abu Yazld as a mystic to one that portrays him as a pietist is paralleled in another o f the Risala s variants o f H ilya material, w hich is presented below: R Abu Yazld said, For thirty years I prayed w ith the belief in myself, at every ritual prayer that I prayed, that I was like a M agian w anting to cut off my girdle (ka-anm majusT uridu an aqti'a zunndrf). (R, 56.13-14)

H [Abu Yazld] said, If you stand before God, M ake yourself like a Magian wanting to cut off the girdle in front o f H im (ka-annak majusT turidu an taqtica }l-zunnar). (H, X, 40.18-19)

These variants both employ the same imagery w hich is expressed in their respective clauses ka-anm majusT uridu an aqtica zunnarT/kaannak majusT turidu an taqti'a al-zunnar. In the H ilya this imagery is presented as part o f an instruction on how to behave in G od’s presence, whilst in the Risala it represents Abu Yazid’s own 112

Q ushayri’s Risala

thoughts during ritual prayer. Thus, the variant in the H ilya is an instruction to fellow mystics seeking proxim ity to God, whilst the use o f the context o f ritual prayer in the Risàia depicts Abu Yazld as a pietist, w ho wishes to perfect his attitude during ritual worship. T he remaining two segments in the biography o f Abu Yazid do not appear to have been obtained from any o f the m ajor Sufi works. O ne o f them is placed at the very end o f the biography, immediately after the variants o f material in the Hilya. In fact, it seems to be derived from the same narrative m o tif as a segment in the biography o f Abu Yazld in that w ork (feeling unw orthy to m ention G o d ’s nam e: f a - ’htashamtu an a d h ku ra h /ijlàlan li-}llàh an adhkurah).52

T he other ‘new ly-introduced’ segment has been juxtaposed immediately after the aforem entioned segment w hich suggests that to follow the prophetic sunna is better than excessive self-denial. It is thematically related to that segment, in that Abu Yazld belittles asceticism (laysa li-’l-zu h d m anzila ), and proceeds to relate his own progress towards proxim ity to God, w hich was only made possible through his grace.53 T he body o f the biography o f Junayd in Qushayn’s Risàia consists o f fourteen juxtaposed discrete segments. Six o f them are demarcated by isnàds54 four by the naming o f the immediate source only,55 and four by the simple conjunction wa-.S6 The first six segments o f the body o f this biography are also found in the corresponding biography o f Sulaml’s Tabaqàt al-Sufiyya ,57 confirming the impression given by the biography o f Abu YazTd, that the precursor was used as the starting-point for the biographies in the Risàia. T he biography o f Junayd also includes variants o f segments that are found in the H ilya and the L u m a \ beginning im mediately w ith its seventh segment, w hich is presented below alongside its earlier variant:

R Junayd said, W hoever does not m em orise the Q u r’an and w rite down hadiths should not be followed as a m odel in this m atter (Sufism) because our knowledge is restricted by the book and the sunna ( cilm unà hàdhà muqayyad b i-’l-kitàb w a -ylsunna). (R, 72.5-6)

H I heard Abu ’1-Hasan cAlI b. H arnn b. M uham m ad and Abu Bakr M uham m ad b. Ahm ad al-M ufid both say that they 113

Q u sh ay n ’s R isala

heard Abu ’1-Qasim Junayd b. M uham m ad say, m ore than once, O u r knowledge is restricted by the book and the sunna ( cilmuna madbut al-kitab w a -’l-sunna). W hoever does not m em orise the Q u r’an, w rite dow n hadith and practise jurisprudence should not be followed as a model. (H, X, 255.12-15)

T he two statements that make up the variant in the H ilya are conjoined in reverse order in the Risala. O therw ise they are virtually identical, the main discrepancy being the use o f the participle muqayyad instead o f madbut , which, in this context, is synonymous. W hat is especially interesting about the inclusion o f this segment in the Risala is the fact that it is followed im mediately by further variants, the first o f w hich is presented below: R I heard M uham m ad b. al-Husayn say that he heard Abu Nasr al-Isbahanl say that he heard Abu cAli ’1-Rudhbari 58 quote Junayd, saying, This way o f ours is bound by the principles o f the book and the sunna (madhhabuna hadha m uqayyad bi-usul al-kitab w a -’l-sunna). (R, 72.7-8)

This variant o f the im mediately preceding segment differs by specifmg that the ‘way’ w hich Junayd represents, rather than ‘the know ledge’, is bound by the principles o f M uslim revelation, but nonetheless both presumably refer to Sufism.59 T he next segment in the biography, w hich is the final m em ber o f this sequence o f juxtaposed variants, is presented below. It is a close variant o f a segment provided in the aforem entioned chapter on the Sufis’ em ulation o f the Prophet in Sarraj’s Kitab a l-L u m a \ w hich is also presented below. R Junayd said, This knowledge o f ours is built out o f the hadith o f the apostle o f God ( cilmuna hadha mushayyad bihadfth rasul A llah). (R, 72.9)

L T he shaykh [Sarraj] said that he heard cAbd al-W ahid b. cUlwan say that he heard Junayd say, This knowledge o f ours is interwoven w ith the hadith o f the apostle o f G od ( cilmuna hadha m ushtabik bi-hadith rasul A llah). (L, 103.9-11)

114

Q ushayri’s R isala

Qushayrfs variant differs from that o f Sarraj only by its use o f m ushayyad bi- (built out of) rather than mushtabik hi- (interwoven with). T he form er may imply that prophetic hadith is the substance o f Junayd’s knowledge, whilst the latter suggests that his knowledge is interwoven in a close relationship w ith hadith. T he difference is perhaps negligible. W hat is m ore significant is the fact that this is the only segment in the biography o f Junayd w hich appears to have been derived from Sarraj’s L u m a , and, just as in the case o f the biography o f BastamI, it is found originally in the chapter o f the latter w ork w hich is concerned specifically w ith em ulation o f the Prophet. T he remaining segments in the biography o f Junayd do not appear in any o f the m ajor precursors to the Risala. Am ongst them is a further example o f an anecdotal topos that has been encountered tw ice already in the biographical tradition o f Junayd.60 In this particular example, Ibn Surayj, the jurist o f Baghdad, indicates his approval and appreciation o f Junayd by stating that his ow n fine speech (kaldm hasan ) about jurisprudence was acquired thanks to discussions w ith Junayd (hadha bi-barakat mujdlasat AbT ’l-Q asim al-Junayd) .61 There are also two segments w hich depict Junayd as som eone w ho engaged in simple acts o f piety and devotion, even after having attained a lofty status. In the first o f these, Junayd responds to an enquiry about how he acquired gnosis (ma'rifa), by explaining that he sat ‘before G o d ’ for thirty years on a step in his house (min julusT thalathm sana tahta tilka ’l-daraja w a -a u m a ’a ila darajafTdarih).62

T he second example presents Junayd as responding to an enquirer w ho is baffled by the fact that he still uses prayer-beads, even after having reached a lofty status (anta m a ca sharajik ta ’khudh biyadik subha?), by declaring that he will not abandon a m ethod that has helped him to reach proxim ity to G od (tanq bih wasaltu ila rabbi la ufariquh).63 These two segments serve to reassure that Junayd was not merely a scholar o f Sufism, but also som eone w ho undertook the practical disciplines o f the tradition and never regarded him self above m aintaining the simplest acts o f devotion, despite being know n as ‘the chief o f the gnostics’. T he penultim ate segment in this biography also functions to attribute pious qualities to Junayd. Daqqaq relates that every day Junayd w ould go to his stall, perform 400 units o f ritual prayer, and then return hom e.64 This narrative functions to distance Junayd from the activity o f trade and its potential corrupting effect. It is 115

Q ushayrfs R isala

not denied that Junayd was a m erchant (his nisba indicates this plainly enough), but the reader is given the impression that he spent all o f his time at the stall in worship. A segment w ith the same m o tif is found in the biography o f Junayd in the Persian Tabaqat al-Sufiyya , w here it is even alleged that he never managed to sell anything.65 T he final segment in this biography is about Junayd’s death. It is also found in Abu N u 'a y m ’s H ilya w ith exactly the same text, but a different isnad.66 It was presumably positioned here at the end o f the biography by intention, because o f the event that it describes, in preference to a position next to this biography’s other variant o f a segment in the H ilya.67 To recapitulate, the fact that the introductions o f Q ushayrl’s biographies o f Abu Yazld and Junayd and the first segments o f their bodies correspond closely to m aterial in Sulam l’s Tabaqat alSu fiyya suggests that the latter w ork has been used systematically as their foundation; Q ushayrl begins his biographies by selecting from corresponding m aterial in Sulam l’s w ork, and then proceeds to add material, in turn, from other w ritten and oral sources. B oth o f these biographies also contain segments that are found in the corresponding biographies o f the H ily a , as well as segments that are variants o f those found in Abu Nasr al-Sarraj’s K itab alL u m a c.

T he biographies o f Abu Yazld and Junayd in the Risala also contain segments that do not appear to be variants o f material found in earlier w ritten works. T hey are positioned either im mediately after a thematically associated variant segment, or at the end o f the biography This overall pattern confirms that the first stage o f the com pilation o f the biographies consisted o f obtaining material from the aforem entioned w ritten sources, beginning w ith Sulaml’s Tabaqat al-Sufiyya. This was ‘topped u p ’ by the addition o f further segments that are not found in any o f the w ritten sources. As it is evident that Qushayrl based his biographies on material already provided in earlier works, the nature o f the material he has selected from what was available in those works becomes all the m ore significant. Firstly, Qushayrl has not included here any o f the provocative apophthegms, boastful autobiographical reports and witty dismissals o f ascetics and pietists, w hich are attributed to Abu Yazld in Sulaml’s Tabaqat and Abu N u caym’s Hilya. M oreover none o f the shathiyyat (ecstatic utterances) attributed to Abu Yazld are 116

Q ushayrfs R isala

included, despite the fact that Sarraj’s Kitab a l-L u m a c contains a section specifically devoted to them .68 T he differences betw een Qushayri’s segments about Abu Yazld and their variants in the earlier works is also instructive. T he form er are consistently less mystical and have been reworked w ith the effect o f emphasising Abu Yazld’s piety and self-denial. T he body o f the biography o fju n ay d includes three juxtaposed variants o f an utterance about the im portance o f the Q u r’an and the sunna for Sufism. T here are also two segments w hich contribute to the image o f Junayd as som eone w ho m aintained the practice o f conventional acts o f worship and good deeds, regardless o f his ow n status in Sufism.69 T he emphasis on these issues in the biography o fjunayd confirms the indications observed in that o f Abu Yazld, namely that Qushayri’s aim is to emphasise that Sufis are pious, traditionalist Muslims. H e achieves this by om itting the m ore mystical segments about these individuals that are available to him, and toning down the ones that he uses. QushayrT does not re-create his subjects as individuals w ithout any mystical characteristics. R ath er he blends together segments that describe them as mystics w ith a high concentration o f those that confirm their piety and traditionalism.

VI T he tabaqat section is not the only section o f the Risala w here one can find segments about Abu Yazld and Junayd, not to m ention the other Sufis included there. This is because all o f the sections o f the R isala , to varying degrees, are made up o f segments about past Sufi authorities. T he systematic section is o f particular interest in this regard, for the basic unit o f its chapters is the discrete segment, conveying the opinions and actions o f past Sufis w hich are relevant to the them e in question. In the light o f this, the m ere provision o f chapters on topics such as sam ac (musical audition)70 and karamat (miracles)71 indicates im mediately that the parameters o f the definition o f Sufism are significantly m ore accom m odating in the systematic section than in the tabaqat section; both sections are made up o f discrete segments about virtually the same set o f authorities,72 but contentious topics are accom m odated only w ithin the systematic section. A com parison o f the material provided about Abu Yazld and Junayd w ithin their biographies 117

Q u sh ay n ’s Risala

w ith that provided in the systematic section may begin to elucidate the nature o f the relationship betw een these two sections. T he systematic section contains the following variant o f a segment already encountered in the biography o f Abu Yazld,73 w hich employs the ‘severance o f the girdle’ motif: I heard M uham m ad b. al-Husayn say that he heard Ahm ad b. cAlI b. Ja'far say that he heard al-Husayn b. cAlluya say that Abu Yazld said, For twelve years I was the blacksmith o f my carnal soul, and then for five years I was the m irror o f my heart. For a year I gazed at what was betw een the two, and there appeared a girdle (zunnar) around my waist. T hen for twelve years I w orked at severing it (qat*), and I gazed at it again, and there appeared a girdle around my inner being (batiri). For five years I w orked at severing it, w ondering how I could cut it off, and it was disclosed to me — I gazed at mankind and saw them as dead, so I pronounced four times over them , A lla h u akbarl74 (R, 177.19-178.4)

It has already been noted that the variant provided in the Risala s biography o f Abu Yazld has a variant in the H ilyat al-awliya’ which is m ore mystical in content.75 T he provision o f the above variant in the systematic section corroborates the impression that, for the specific purposes o f his biography Qushayri preferred the variant w hich emphasises his pious attitude during ritual prayer, for it constitutes in itself evidence that he had alternative variants. In fact, the above segment belonging to the systematic section is provided in the chapter on mujahada (religious striving), for w hich the variant provided in the biography o f Abu Yazld w ould have been equally appropriate.76 T he systematic section also includes variants o f segments w hich are frequently provided in earlier works that have been used as sources for the biography o f Abu Yazld in the R isala , but are not included in that biography in any guise. For instance, consider the following example, w hich is a variant o f a segment that is provided in the H ilya ’s biography o f Abu Yazld: R Abu Yazld said, G od has certain worshippers ( cubbad) w hom , if he were to veil them in heaven from vision o f him, w ould appeal for deliverance from heaven just as the inhabitants o f hell appeal for deliverance from hell (law

118

Q ushayri’s Risala h a ja b a h u m ft ’l-janna can ru’yatih la -’staghathu m in al-janna kama yastaghith ahl al-nar m in al-nar). (R, 460.16-17)

H [Abu Yazid] said, G od has an elite group o f worshippers ( cubbad) w hom , if he were to veil them in heaven from vision o f him, w ould appeal to leave heaven just as the inhabitants o f hell appeal to leave hell (law hajabahum f t l-ja n n a can ru’yatih la -’staghathu b i-’l-khuruj m in al-janna kama yastaghith ahl al-nar b i-’l-khuruj m in al-nar). (H, X, 34.14-17)

T he above segment in the chapter on shawq (yearning) is a close variant o f the segment provided in the biography o f Abu Yazid in the H ilya. These variants both imply that Abu Yazid identifies w ith the uncom prom ising elite, whose goal is vision o f God, and w ho lack any interest in the rewards o f heaven w hich ordinary Muslims anticipate. In fact, two further variants o f this segment are provided in the biography o f Abu Yazid in the H ilya ,77 but none o f them has been selected for his biography in the Risala; a variant, however, is reserved exclusively for the systematic section o f this work, perhaps because it was considered too controversial to be included in the tabaqdt section. A nother type o f utterance that is found in earlier biographies o f Abu Yazid, but not in that given by Qushayrl, is the witty assertion o f the inferiority o f ascetics in relation to gnostics. Such a segment is found in the systematic section o f the Risala. Abu Yazid said, T he gnostic flies whilst the ascetic walks (alcarif tayyar w a -’l-za h id sayyar). (R, 441.22)

This utterance parallels those attributed to Abu Yazid in earlier sources, such as the following example from Sulaml’s Tabaqdt al­ ts fifty y a.

Abu Yazid said, T he gnostic’s concern is w hat he hopes for, while the ascetic’s concern is w hat he eats. (TABS, 66.11-67.1)

A lthough this kind o f utterance is conspicuously absent from the R isala's biography o f Abu Yazid w hen it is com pared w ith the earlier biographies that have been used as a source, the provision o f 119

Q u sh ay n ’s Risala

the above variant in the systematic section indicates that Qushayrl had examples available, if he had wished to include them. Its arrogant tone may have been considered inappropriate for the biography . As already m entioned, the segments contained in Qushayrl’s biographies o f Abu Yazld and Junayd w hich are variants o f segments found in Sarraj’s Kitdb a l-L u m a c are all found in a specific (36-line) chapter o f the earlier work, on ‘w hat has been m entioned about the Sufi shaykhs regarding their emulation o f the apostle o f G od and their expertise in that’. However the systematic section o f the R isala , w hich is influenced m ore by the L u m a c than any other previous work, contains many variants o f segments that are found in other chapters o f that work. These include two juxtaposed segments about Junayd for the chapter on sam a c,78 T he L u m a ' w ould have been an obvious source for the com pilation o f a chapter on sa m a \ and indeed it appears to serve as its foundation,79 in m uch the same way as Sulaml’s Tabaqat serves as the foundation for most o f the biographies o f the tabaqat section. Qushayri’s decision to include in his biographies o f Abu Yazld and Junayd segments from the L u m a c selected only from the chapter specifically about the Sufis’ emulation o f the Prophet suggests that this chapter may similarly have been considered the obvious source for segments that are appropriate for the particular function o f his biographies. T h e biography o f Junayd in the R isala depicts him as som eone w ho advocated that Sufi practice should be in harm ony w ith theoretical know ledge, and that adept Sufis should not abandon the discipline they u n d erto o k as novices. This makes it all the m ore conspicuous that the systematic section o f the R isala should contain m aterial that qualifies (perhaps even contradicts) this image, including the first o f the follow ing tw o juxtaposed segments at the culm ination o f the chapter on irada (desire/ discipleship). 1 Junayd was asked about the novice (mund) and the adept Sufi (murad),m and he replied, T he novice is controlled by the rule o f scholastic knowledge (siyasat a l-c\lm ), while the adept is controlled by the protection o f the Truth (n d y a t al-haqq ), because the novice walks (yasfr) while the adept flies (yat.Tr). W hen will the one w ho walks catch up with the one w ho flies?

120

Q ushayri’s R isala 2 It is said that D hü ’1-Nün sent a m an to Abü Yazïd,

instructing him , Ask him until w hen will the sleep and repose continue — the caravan has already departed ! Abü Yazïd replied, Tell my brother D hü ’1-Nün, the real man is the one w ho sleeps all night, then wakes up at the destination before the caravan [has reached it] ! D hü ’1N ü n responded, Bravo ! We, in our present state, cannot match this.81 (R, 310.17-19)

This utterance attributed to Junayd implies that adept Sufis are no longer governed by the same restrictions as novices; they are protected directly by God. Similarly, the subsequent narrative about Abü Yazïd implies that he can afford to rest w ith the reassurance that som eone in his state can reach the furthest goals regardless. T he above examples show that the segments included in the R isâ la s biographies o f Abü Yazïd and Junayd represent a restricted selection from the w ider range available to QushayrL Further segments about the same individuals are provided in other parts o f the work, displaying tangible differences to those that have been included in their respective biographies. T he biographies tend to contain the less mystical variants and those that are less controversial. T h e discrepancy in the selection criteria used for the com pilation o f the different sections o f the Risâla is not restricted to the level o f the segment alone. It has already been pointed out that the biography o f Hallâj is the longest o f those found in Sulamï’s Tabaqât, not to be included in Qushayrï’s Risala. T he significance o f this omission is further emphasised by the fact that many segments about Hallâj are found in the other sections o f the Risala. In fact, there are seventeen in the systematic section alone.82 This seems to confirm that Hallâj was considered inappropriate for inclusion in the tabaqât section o f the R isâla , whilst at the same tim e he was considered im portant enough to be m entioned in its systematic section. It is w orth considering w hat the differences are betw een the context o f a biography and the context o f a them atic chapter, in order to appreciate the factors behind Qushayrï’s selection criteria. Biographies by definition put individuals ‘under the spotlight’. T he subject is the focus o f attention because all o f the material in 121

Q ushayrfs Risala

the biography, w hich takes the form o f an assemblage o f discrete segments, is about him. In contrast, while each chapter in the systematic section is also made up o f discrete segments o f material, a variety o f individuals are quoted w ithin them; the focus is the them e o f the chapter, and so the opinions o f a variety o f authorities on that particular topic are em bedded amongst each other. Moreover, thematic chapters in the Risala begin w ith citations from the texts o f M uslim revelation (Q u r’an and Hadith), followed by Qushayrfs own introduction to the topic. T he segments that follow are then arranged so as to develop the elaboration o f the topic under discussion, facilitated by further com m ents by Qushayrl at appropriate points. If the above passage is considered as an example, one observes that it forms the culm ination o f the chapter on irada (desire), by w hich point the discussion had been directed towards the consideration o f the terms murid (novice Sufi) and murad (adept Sufi). T he way to these culm inating segments had been paved effectively, so as to accom m odate the contentious climax. T he functions o f the tabaqat section entail the w idening o f the acceptability o f Sufis and the structuring o f a harm ony between traditionalist Islam and Sufism. This is achieved by characterising Sufism in the tabaqat section by piety, hum ility and religious discipline. T he omission o f the biography ofH allaj appears to have been part o f the effort to fulfil this objective. T he inclusion in the systematic section o f several segments about Hallaj highlights the discrepancies betw een the two sections. Those segments do not make him the focus o f discussion, as they w ould have done if they had been used to form a biography. M oreover, w hen em bedded amongst the opinions o f other Sufi authorities, as well as citations from M uslim revelation, this material is m ore readily acceptable.

V II T he agenda that has been observed by the analysis o f the biographies in the Risala is not actually a hidden one. T he aims o f his tabaqat section are elaborated in its conclusion. T he purpose o f m entioning them in this place is to indicate that they are unanim ous about the veneration o f the shariah (.ta'zTm al-shan a), [that they] are characterised by travelling

122

Q u sh ay n ’s R isala

the paths o f religious discipline (turuq al-riyada), and persist in following the sunna w ith o u t abandoning any religious customs (adab al-diyana ); [that they] agree that w hoever abandons pious deeds and striving and does not build his affair on the foundation o f piety (warn0) and righteousness (taqwa ) is being insincere to God about w hat he is claiming. (R, 119.12-16)

In the above passage, Qushayrl acknowledges that he has provided his biographies for a particular end. T he final two statements suggest further that he is defining the identity o f Sufis polemically against false claimants w ho do not possess the preceding list o f required qualities. Similar sentiments are also expressed in the introduction to the Risala where he states that Sufism in his time had becom e virtually extinct;83 apparently only the ‘pretenders’ to Sufism had remained. T he Risala presents itself as an attem pt to remedy this situation by highlighting the characteristics o f the ‘real’ Sufis o f the past. A lthough such com m ents are conventional m ethods o f exhortation in the introductions o f religious works, the analysis o f the biographies o f the Risala certainly confirms that an effort has been made to redefine the identity o f Sufism. M oreover, this re-orien tatio n is new and relevant to Q u sh ay n ’s ow n experience as a traditionalist Ash’arite/Shafi’ite. T he discrepancy betw een the images presented o f Sufis in the different sections o f the Risala suggests that Q ushayn’s endeavour was not straightforward. It has been observed that Section O ne, the theology section o f the R isala , attributes Ash’arite opinions to Sufis. Similarly, Section Two, the tabaqat section, presents Sufis in a m ore acceptable light to traditionalists and jurists by attributing to them the aforem entioned characteristics. At the beginning o f the term inology section, however, Qushayrl states that it is w ritten for the practising Sufi,84 thus signalling a departure from the reorientation o f the definition o f the identity o f the Sufi com m unity to the exposition o f Sufi theory w ithin significantly w ider parameters. T he first two sections may be seen as a preparation for the main body o f the text, w hich follows. In relation to the subsequent Sufi manual, the tabaqat section can be considered to function like the isnad in relation to its matn. Qushayrl has com piled a w ork in w hich the tabaqat and manual genres function in tandem to em bed Sufism further w ithin the mainstream o f Sunni Islam, whilst m aintaining generous limits o f 123

Q ushayrï’s Risala

acceptability. His m ethod is derived from a mastery o f the conventions o f the two genres and the functions to w hich each o f them lends itself. H e manipulates carefully the building-blocks o f material at his disposal, by selecting, modifying and re-arranging them in new configurations in order to fulfil the overall aims o f his work.

124

Chapter

Five

H ujw lri’s Kashf al-mahjub

I :>u ’l-Hasan cAli b. cU thm an al-Jullabi al-H ujw m was a Sufi scholar o f the eleventh century w ho was born in Ghazna and settled eventually in Lahore.1 T he m ost reliable source o f inform ation about his life is his sole surviving work, the K a sh f almahjub.2 This w ork suggests that, whilst he was first and foremost a Sufi, he had also received a traditional scholastic education, and was particularly interested in kalam (scholastic theology); it contains repeated criticism o f anthropomorphists (mutashahbiha, hashw iyan)3 Qadarites (qadariyan)4 and M u ’tazilites (m u ctazila / ahl-i i ctizal).5 M oreover H ujw lrl frequently uses the m ethod o f dialectical reasoning in his exposition o f Sufi doctrine, as indicated by the frequent occurrence in the text o f the construct agar guyand . . . m fgu’Tm . . . (‘if they say. . . we say. .. ’).6 A comparison o f H ujw lri’s own biographies o f the eponymous founders o f the major lawschools suggest that he was most likely affiliated to the Hanafite school, since Abu Hanlfa is afforded the most attention and praise.7 H ujw lrl’s teacher in Sufism was probably the relatively littleknow n Abu ’1-Fadl al-Khuttali, w hom he describes as his role m odel on the Sufi path (iqtida’-i man dar-Tn tariqat bid-ust ).8 The K a s h f al-mahjub also contains many references to meetings with other Sufi teachers, during travels in an area extending from Syria (where Khuttali was based) to the Punjab. H ujw lrl died in Lahore w here his shrine is today the m ost celebrated pilgrim age destination. T he dates given traditionally for his death are 45 6 / 1063-4 and 4 6 4 /1 0 7 1 -2 .9 125

H ujw iri’s K a s h f al-m ahjub

T he K a sh f al-mahjub is the oldest surviving w ork o f its kind w ritten in Persian. Similar to Q ushayri’s R isala , it is a dual-generic volume covering both the tabaqat and manual genres. It was allegedly w ritten in response to questions asked by an enquirer into Sufism nam ed Abu SacTd H ujw iri,10 and the following passage suggests that it was com pleted in Lahore, and therefore probably towards the end o f the author’s life.11 M y shaykh had many transmissions (riwayat) about him, but, at the present time, [relating] m ore than this is not possible because the w ritten sources are still in Ghazna, whilst I am preoccupied (giriftar) in the region o f H ind amongst illnatured people (ndjinsan ). (KM, 110.13-16)

R eynold Nicholson has inferred from the above passage that H ujw iri was ‘taken there (Lahore) as a prisoner against his will’.12 This interpretation depends on the translation o f giriftar as ‘prisoner’, but this is not the only possible m eaning in the above context. H ujw iri’s description o f the people o f H ind as ndjinsan (ill-natured), rather than confirm ing that he was trapped by enemies, may be seen as simply part o f the description o f his current situation in negative terms to contrast w ith the past in his hom etow n and the positive m em ories he retains o f it (possibly including those o f m ore extensive literary resources). T he above passage alludes to the use o f w ritten sources (kutub) for the K a sh f al-mahjub. O n the basis o f a com parison w ith the m ajor Sufi works that had already been w ritten, it appears that, rather than suffering from a lack o f resources, H ujw iri had most o f them at his disposal.13

II T he K a s h f al-mahjub can be divided into three main sections. W hile Section Two consists mostly o f biographies (see section III below), Sections O ne and T hree are made up o f them atic chapters. Section O ne is structured as follows: 1 O n the affirmation o f knowledge (77m) 2 O n poverty (faqr) 3 O n [the name] ‘Sufism’ (tasawumf)

126

H u jw in ’s K a s h f al-m ahjub

4 O n the wearing o f patched cloaks (m uraqqaat) 5 O n their differences o f opinion regarding poverty and purity (safwa) 6 In explanation o f blame ( f t bayan al-malama). W ith the exception o f the fifth chapter, w hich is best described as an appendix relating to the second and third chapters,14 all the chapters o f Section O n e share the same structural p attern ;15 H u jw lrl briefly introduces the subject o f discussion before presenting citations from the Q u r’an and hadith literature; he then elaborates. T he most distinctive characteristic o f H u jw m ’s approach is his provision o f extensive elaboration, in w hich segments about Sufis from the past are em bedded for support and illustration. As a result, he may appear to be expressing his opinions m ore directly to the reader than contem poraries, like Qushayri, whose works tend to be made up o f juxtaposed segments w ith minim al com m entary.16 W hat stands out most in Section O ne is the sixth chapter, on ‘blam e’ (malama ). T he m ere inclusion o f a chapter on malama alongside chapters about what are usually regarded as m ore central aspects o f Sufism is in itself innovative, indicating the significance w hich H ujw lrl attaches to this subject.17 Section T hree o f the K a s h f al-mahjub consists largely o f a systematic analysis o f Sufi practice. It consists o f eleven chapters, w hich are presented as a sequence o f ‘uncoverings o f veils’ (ka sh f al-hijab).

1 T he U ncovering o f the first veil regarding gnosis (m a crifa) 2 T he U ncovering o f the second veil regarding divine unity (tawhfd)

3 T he U ncovering o f the third veil regarding faith (Tman) 4 T he U ncovering o f the fourth veil regarding purification (tahara) 5 T he U ncovering o f the fifth veil regarding ritual prayer (salat)

6 T he U ncovering o f the sixth veil regarding almsgiving (zakat)

7 T he U ncovering o f the seventh veil regarding fasting (sawm)

8 T he U ncovering o f the eighth veil regarding pilgrimage (bajf)

127

H u jw in ’s K a s h f al-m ahjub

9 T he U ncovering o f the ninth veil regarding com panion­ ship (suhba ) together w ith its manners and rules 10 T he U ncovering o f the tenth veil explaining their m anner o f speaking (mantiq ), the definitions o f their expressions (alfdz) and the realities o f th eir m eanings (haqa’iq ma canihim ) 11 T he U ncovering o f the eleventh veil regarding the practice o f audition (sama°) and the explanation o f its various forms W hilst the first three chapters are concerned w ith theological aspects o f Sufism and Chapters Four to Eight are (ostensibly) concerned w ith the rituals w hich have becom e know n as ‘the pillars o f Islam’, Chapters N ine to Eleven focus on topics related specifically to the customs and practice o f Sufism.18 T he chapters about the customs and practices peculiar to Sufism, w hich actually account for the m uch greater portion in length o f Section T h ree,19 are presented after those about the four main M uslim rituals, creating the impression that they are based on the foundation o f the ‘pillars o f Islam’. In fact, Chapters Four to Eight are only ostensibly about the M uslim rituals specified in their headings, because they each consist o f two sections, only the first o f w hich relates directly to that ritual; the second section in each o f these chapters offers a mystical in terp retatio n and elaboration. Thus, the chapter ostensibly about purification leads to a section, twice as long, about repentance (tawba); that o f prayer leads similarly to a discussion o f love (mahabba, cishq); that o f almsgiving to a section on liberality (jud) and generosity (sakhawa ); that o f fasting to the mystical significance o f ‘hunger’ (ju c); and that o f pilgrimage to a section on ‘mystical witnessings’ (mushahadat). In this way, the overall scheme o f Section Three, grounding Sufi customs and practices on the apparent foundation o f the conventional ‘pillars o f Islam’, is repeated in the m icrocosm o f the com ponent chapters. T he culm ination o f Section Three is the chapter on musical audition (sama ). It is to a large extent based on the corresponding chapter o f Sarraj’s Kitab al-lum ac.2{) However, after following closely Sarraj’s m ethod o f legitimizing the practice o f musical audition (.sama*), H ujw lrl adds further sections at the end, on dance (raqs)2] gazing at youths (a l-nazarfi ’l-ahdatti)22 and the rending o f garments (,kharq ), respectively.23 H ujw lrl endorses Sufi dance, w hich he 128

H u jw in ’s K a slif al-m ahjub

term s harakat-i wajdi (ecstatic movements) to distinguish it from other forms o f dance, as well as the tearing o f garments in the ecstasy o f sa m a \ but he draws the line at the involvement o f youths.

Ill Section Two o f the K a s h f al-mahjub consists mainly o f biographies. T hey are grouped into the first seven chapters o f this section, before the final chapter on contem porary Sufi groups {giruhha). T he first four chapters consist o f the biographies o f ‘the leaders o f the Sufis (a ’im m a tu h u m Y am ongst, in tu rn , the follow ing generations and classes o f religious authorities: 1 the four rightly-guided caliphs amongst the companions o f the Prophet (sahaba) 2 the second to the sixth Imams from amongst the family o f the Prophet (ahl al-bayt)24 3 the ‘people o f the canopy’ (ahl al-suffa) 4 ‘the followers’ o f the companions (tabxun ) and their ‘helpers’ (ansar) These four chapters encompass, in an overall chronological arrangem ent, the first two generations o f the successors to the Prophet according to the Sunni tradition’s image o f its ow n past, as well as the first six Imams o f Shi’ism, even though the individuals concerned are not rem em bered usually w ith ascetic and mystical characteristics. T he inclusion o f the ahl al-bayt, lined up alongside the salaj.‘ indicates H u jw in ’s predilection for the accom m odation o f plurality. T he fifth chapter o f Section Two accounts for m ore than three times as many pages as the first four chapters com bined. It consists o f sixty-four biographies presented under the heading, ‘C hapter m entioning their leaders from amongst the followers o f the followers [of the companions] (atbac al-tabicm) until our present day’. T he selection o f biographies for this chapter, as well as their arrangement, indicate the direct influence o f Sulami’s Tabaqat alSufiyya. T here are only two significant areas o f discrepancy: firstly, H ujw lrl includes seven biographies before that o f Fudayl b. cIyad, w hich is the first biography in Sulami’s arrangement; secondly, he does not group the biographies into ‘generations’. Nonetheless, their order o f appearance is usually just the same as in that earlier 129

H u jw in ’s K a s h f al-m ahjub

work, in a m anner comparable to the relationship betw een Sulaml’s arrangem ent and the tabaqat section o f Q ushayrl’s Risála.25 T he K a sh f al-mahjüb provides a corresponding biography for each o f those included in Sulaml’s first generation. M oreover, they are all listed before the biography o f Junayd which, in the earlier work, marks the beginning o f the second generation. A similar pattern can be observed in the biographies w hich correspond to those found in the later generations o f Sulaml’s work. For instance, the biographies w hich correspond to those found in Sulaml’s second generation are all found betw een the aforem entioned biography o f Junayd and that o f Jurayri which, in the earlier work, marks the beginning o f the third generation.26 T here are corresponding biographies also for eight o f the twenty biographies found in Sulaml’s third generation,27 whilst the biography o f Abü Bakr al-Shibll, w hich heads Sulaml’s fourth generation, is the only representative o f that generation,28 and only seven o f the tw entythree biographies in Sulaml’s fifth generation are included in the K a s h f al-mahjüb. To recapitulate, the K a sh f al-mahjüb includes, positioned betw een the biographies o f Fudayl b. cIyad and Abü ’1Hasan al-Husri, corresponding biographies for over half (54) o f the 103 biographies included in Sulaml’s Tabaqat a l-S ü fy y a , and most o f them are presented in the same order in relation to each other. T here are three biographies positioned betw een those o f Fudayl b. cIyad and HusrI for w hich there are no corresponding biographies in Sulaml’s earlier work. These ‘new ly-introduced’ biographies are devoted to D áw üd al-Ta’I, M uham m ad b. Idris alSháfi'l and Ahmad b. Hanbal, respectively. D áw üd al-Ta’I is introduced as a disciple o f Abü Hanlfa, whose biography is included already in C hapter Four o f this section.29 H e has also been included in Qushayrl’s Risála and in the H ilyat al-awliya>. T he inclusion o f the jurists ShaficI3(> and Ibn Hanbal31 is also not entirely w ithout precedent in the genre, for the H ilyat al-awliya} includes both o f them as well. However, it needs to be taken into consideration that the ten-volum e H ilya contains over 650 biographies in total. T heir inclusion in H ujw lri’s relatively small collection o f biographies therefore suggests a m ore emphatic effort on his part to em bed Sufism as an integral part o f a wide and authoritative tradition o f Muslim scholarship. It has already been n o ted that, despite containing the biographies o f num erous individuals w ho are not norm ally associated w ith Sufism, the H ilya fails to provide a biography o f 130

H ujw iri’s K a s h f al-m ahjub

Hallaj. In the K a s h f al-mahjub, however, the third longest biography in the w hole w ork is that o f Hallaj (third after those o f Abu Hanifa and Fudayl b. cIyad).32 T he mere inclusion o f Hallaj, ‘rubbing shoulders’ w ith jurists and R ightly-guided Caliphs, highlights further the fact that H ujw iri has chosen to accom m odate a diverse range o f individuals, rather than narrow down his selection to the respectable few. H ujw iri is by no means the first to include Hallaj in the Sufi tabaqat genre. SulamI after all includes him, m entioning at the outset, in the introduction o f his biography, the controversy about his legitimacy;33 SulamI informs that most shaykhs reject him (:raddah akthar al-m asha’ikh ), but identifies only those w ho defend and praise him .34 H ujw lrl’s biography o f Hallaj does not merely refer to the controversy, but it is in fact entirely taken up w ith his defence. It is w orth considering in further detail, for it is comparable w ith his discussions o f controversial issues in other sections o f the work, such as the aforem entioned chapters on malama and sama H ujw iri begins, like SulamI, by inform ing the reader o f the controversy over Hallaj’s legitimacy, and then nam ing those authorities w ho support him .36 Innovatively he adds a further list o f authorities - those w ho suspend judgem ent (andar amr-i way taw aqquf karda-and), w hich includes Junayd.37 H e finally m entions contem porary shaykhs w ho are in favour o f Hallaj, namely Shaykh Abu Sa'ld (al-M ayhanl), Abu ’1-Qasim al-G urganl and Abu ’1cAbbas al-Shaqqanl.38 H e then him self proceeds to defend Hallaj against accusations o f heresy; at first, he suggests that m any critics have confused him w ith a namesake w ho really was a heretic. T hen he states the reasons for w hich the correctly identified Hallaj has been rejected, explaining, in each instance, w hy there is insufficient reason to condem n him .39 H e also claims that Hallaj’s utterances have been m isunderstood, both by opponents and ‘false followers’ w hom he calls the H allajiyan .40 At the end o f the biography, immediately after declaring that he personally admires Hallaj, and that he has w ritten further works about him , he reproaches Hallaj’s m ethod because it requires painstaking corroboration. H e then presents w hat he judges to be a dangerous utterance attributed to Hallaj, which, w hen com pared w ith some o f the utterances attributed to him by this time in Sufi literature, is hardly the most deserving to be condem ned as ‘a bane’.41 Regardless o f H ujw lrl’s choice o f a dangerous utterance 131

H ujw iri’s K a s h f al-m ahjub

attributed to Halláj, it strikes the reader that w ith this criticisms he is m aking an unanticipated departure from his efforts up until that point in support o f Halláj. Perhaps it is intended to reassure the reader that, in spite o f his defence o f Halláj, he has been strict in his assessment. As m entioned above, HujwTrT innovatively includes Halláj together w ith religious authorities from a wide tradition o f scholarship and piety, in the tabaqat section o f his work. This context can only give further strength to the case for approval o f Halláj. T he fact that his biography ends on a critical note is comparable to what can be observed in the chapters on malama and sam ac; in both o f these chapters, after a process o f legitimizing the respective practices, H ujw lri makes a sudden departure into criticism. This suggests that it is a deliberate technique to facilitate the acceptance o f controversial issues, rather than evidence o f indecision and confusion on the part o f the author; in order to redefine the boundaries o f acceptability m ore accommodatingly, H ujw lri must reassure the reader that he is not merely a permissivist, but rather that he maintains a firm line between legitimate and illegitimate aspects o f the issue concerned.42 T he sixth chapter o f Section Two contains the biographies o f the leaders o f recent Sufis (a ’im m atuhum min al-m uta’akhkhirm ), those w ho lived too late to be included in the earlier tabaqat works.43 It includes the aforem entioned shaykhs w ho allegedly supported Halláj, Abü Sa'ld b. Abl ’l-Khayr al-Mayhánl, Abü ’1Qásim al-G urgánl and Abü ’l-cAbbás al-Shaqqánl, as well as Qushayrí, the author o f the Risála. Also included is the biography o f the relatively little know n Khuttali, w hom H ujw lri apparently took as his own role m odel.44 In the seventh chapter HujwTrT provides single-sentence references to his contem porary Sufis, w ho are listed according to the geographical region in w hich they lived; the list begins w ith the Sufi contem poraries living in Syria (where KhuttalT was based), and ends w ith those in Ghazna, from where HujwTrT himself originated. To recapitulate, the first seven chapters o f Section Two o f the K a s h f al-mahjüb , in com m on w ith other writings belonging to the Sufi tabaqat genre, contain biographies arranged in a chronological pattern. They cover the span extending from the time o f the Prophet to the time o f the author. In this way a continuity is structured betw een later Sufis, including HujwTri’s own ‘role 132

H ujw iri’s K a s h f al-m ahjub

m odel’ Khuttall, and the com panions o f the Prophet, through the intervening generations o f religious authorities. This cohesive schema serves the function o f legitimizing the authority o f the selected Sufis o f H ujw iri’s own generation, as well as the other individuals included in the diachronic com m unity he has structured, notably accom m odating Halláj. H ujw lrl, in contrast to his contem porary Qushayri, whose Risala is a dual-generic w ork like the K a s h f al-mahjüb , accom ­ modates a diverse range o f religious authorities in his choice o f Sufi leaders o f the past, from either end o f the spectrum. T here is no doubt that the inclusion o f Halláj is a deliberate decision, for H ujw lrl reveals that he is already aware o f the strategy o f om itting him w hen he com m ents that ‘not to register a m ention o f him w ould be insincere’ (ithbat na-kardan-i dhikr-i way bT-imanatTbüd-í).4S

IV T he final chapter o f Section Two consists o f descriptions o f Sufi ‘groups’ (girühha ), rather than individual biographies.46 A total o f twelve ‘groups’ are m entioned, the first ten being classified as ‘accepted’ (maqbül) whilst the final two are ‘rejected’ (mardüd). Taken at face value, this chapter m ight be understood to represent an account o f the actual schools o f Sufism that existed in H ujw lrl’s day. However, the lack o f corroboration in contem porary sources would suggest otherwise. It is in this context that his m ethod o f describing the ‘groups’ gains greater significance. T he ten ‘accepted groups’ represent a diversity o f opinions, and they are each nam ed after a specific individual whose biography is already included in the preceding chapters (e.g. the Junaydiyya after Junayd; the Qassariyya after H am dün al-Qassár).47 H ujw iri’s characterisation o f each o f the ‘groups’ consists o f an exposition o f the doctrine w hich he attributes to them . In most cases this leads to considerable elaboration. M ost o f the issues covered are contentious, such as the question o f superiority betw een intoxication (.sukr ) and sobriety (.sahw ), and annihilation ffana’) and subsistence (baqa’).4H This chapter on Sufi groups therefore constitutes a m erging o f the biographical and thematic principles o f order. H u jw iri’s objective in this chapter therefore appears to be to expound the various doctrines and debates that are current, rather 133

H u jw iri’s K a s h f al-m ahjub

than to chronicle the em ergence o f actual Sufi schools. By presenting contentious subjects as the doctrines o f particular ‘groups’ o f Sufis, each o f whose authority is derived from an illustrious early Sufi leader, he manages to give legitimacy to a wide range o f approaches and opinions. It would appear that his reference to these ‘founding fathers’ may therefore be prim arily to give authority, and this is corroborated by the fact that most o f their biographies do not themselves contain even a single reference to the doctrines w hich are attributed to them in this chapter. For example, Abü Yazld (Tayfür) al-Bastâml w ho is described as a proponent o f intoxication over sobriety in the context o f the Tayfüriyya , is not associated w ith intoxication at all in his biography. Hujwlri, in the specific context o f his account o f the Tayfüriyya , appears to be the first to attribute to him this doctrine, and thus he shapes his depiction in the later Sufi tradition.49 T he two ‘rejected’ groups are the H u lü liyya , w ho are allegedly the believers in the heresy o f incarnation, and the H allâjiyya , the so-called ‘false followers’ o f Hallâj (already referred to in the latter’s biography) w ho are characterised by allowing transgression o f the law. H ujw lri considers these two ‘rejected groups’ under a single heading,50 since their inclusion in the scheme further serves the single purpose o f re-defm ing the limits o f acceptability. In this way, the chapter on Sufi groups follows the same pattern as the biography o f Hallâj, as well as the aforem entioned chapters on malâma and sam ac; a diversity o f opinions about controversial issues is accom m odated by giving a qualified approval. This is achieved by drawing the line at illegitimate belief and practice, w hich are attributed, at the end o f the chapter, to the H ulüliyya and the H allâjiyya respectively. H ujw lri thus manages to accom m odate the doctrines o f the preceding ten groups, even if some o f them should be diametrically opposed to each other.51 T he position o f this chapter, immediately after the biographies o f Sufis, suggests that the ten ‘accepted’ groups represent the culm ination o f the Sufi tradition. T he arrangem ent also serves to present the doctrines attributed to each o f the various groups included as having the authority o f the entire diachronic com m unity that has been structured, stretching back to the time o f the Prophet. It is perhaps no coincidence then that this is the chapter for w hich H ujw lri reserves his exposition o f most o f the contentious doctrines and theories associated w ith Sufism.52

134

H ujw irfs K a s h f al-m ahjub

V Each biography in the K a sh f al-mahjub consists o f two parts: the introduction and the body T he introductions o f those biographies for w hich there is a corresponding one in Sulami’s Tabaqat alSufiyya appear to be related relatively loosely to their earlier counterparts. Consider, for example, the introduction o f the biography o f Junayd, w hich is presented below Amongst them is the shaykh o f shaykhs in the mystical path (,tariqat), and the leader o f leaders in the religious law (sharfat), Abu ’1-Qasim al-Junayd b. M uhammad b. al-Junayd al-Qawarlri. 1 H e was accepted by the literalists (ahl-i zahir) as well as those endow ed w ith hearts. 2 H e was perfect in the [various] branches o f knowledge, and the ‘m ufti’ o f theology (usul ), jurisprudence (furu ), mystical union (wusul) and social conduct (m u'am alat ). 3 H e was the leader o f the companions o f Abu Thawr. 4 H e has lofty sayings and perfect mystical states, such that all o f the mystics agree upon his leadership and no claimant or free-thinker has the ability to oppose him. 5 H e was the nephew o f Sari Saqatl as well as his disciple. (KM, 161.5-11)

T he introduction to Junayd in the K a s h f al-mahjub includes inform ation about his names, his authority and his association w ith Abu Thaw r and Sari ’1-SaqatI. All o f these points are also m entioned in the corresponding introduction in Sulami’s Tabaqat al-S u fiyya .53 However, the K a s h f al-mahjub omits any m ention o f Junayd’s origins, place o f upbringing and circumstances o f death, for such details are usually not included in his work. H ujw irfs introduction is m ore evaluative than ‘factual’, and especially places a greater emphasis on the authority o f Junayd, w hich it describes as covering both the Sufi tradition and the scholastic traditions o f Islam. Nonetheless, the elements w hich stress this point appear to be related loosely to inform ation provided originally by Sulaml.54 For instance, Elem ent 1 about the acceptance o f Junayd (maqbul-i ahl-i za h ir wa-arbab al-qulub) appears to be derived from a clause in its earlier counterpart (maqbul cala j a m f al-alsina). This re-created version has the effect o f attributing unambiguously to Junayd acceptance by both mystics and non-mystics alike.

135

H u jw iri’s K a s h f al-m ahjub

T he K a sh f al-mahjub only names Saqati from amongst the teachers o fju n ay d in Sufism w ho are listed in the Tabaqat. This is probably due to the fact that the body o f the biography begins (immediately after this m ention o f him) w ith a narrative w hich illustrates their relationship, whilst it contains no material about any other teachers.55 T he introduction o f the biography o f Abu Yazld in the K a s h f almahjub is presented below: Amongst them is the heaven o f gnosis (falak-i ma'rifat ), the angel o f love (malak-i mahabbat), Abu Yazld Tayfur alB a sta m l,

1 H e was one o f the shaykhs, his state (hat) greater than the rest and his rank (sha’n) the most magnificent o f them all, to the extent that Junayd said, Abu Yazld amongst us has the status (m anzila ) o f Gabriel amongst the angels. 2 His grandfather had been a Magian, and his father was one o f the greats o f Bastam. 3 H e has excellent transmissions o f hadiths o f the Prophet. 4 H e was one o f these ten famous leaders o f Sufism.56 5 N o -o n e before him made as many discoveries as him concerning the truths o f this science. 6 In all states he was an admirer o f scholastic knowledge (7/m) and a venerator o f the religious law (.sharfat ), not withstanding that a group, for the sake o f support for their own heresy (ilhad ), attribute [falsely] som ething to him. In the beginning, his life was based on religious striving (mujahadat) and putting into practice good conduct (barzish-i m uam alai). (KM, 132.9-19)

T h e in troduction to Abu Yazld in the K a s h f al-m ahjub is considerably longer than its counterpart in SulamT’s Tabaqat alS u fiyya ,57 as well as being relatively long by the standards o f the K a s h f al-mahjub itself, thereby im mediately suggesting that his im portance had increased in the eyes o f HujwTri. In addition to providing most o f the inform ation found already in the Tabaqat, H ujw iri’s introduction innovatively places an emphasis on Abu Yazld’s high rank amongst Sufis. Elements 1, 4 and 5 each function to emphasise this point, and it is none other than Junayd, usually rem em bered him self as the most authoritative Sufi, w ho is quoted

136

H u jw ln ’s K a s h f al-m ahjub

in confirm ation o f Abü Yazld’s pre-em inence. T he final elem ent o f H ujw lrfs introduction functions to distance Abü Yazld from any heretical doctrine that m ight be associated w ith him. In a similar m anner to his defence o f Halláj, HujwirT explains that a group have falsely attributed their ow n heresy to Abü Yazld. B oth o f the above introductions are representative o f those provided in the K a sh f al-mahjüb for individuals also included by Sulaml, in that they are loosely based on inform ation provided originally by their respective counterparts half a century before. However, they have been re-created by HujwirT relatively liberally, in order to stress the points that are im portant for the context o f his ow n work. This has the effect o f blurring the relationship betw een the two works to a considerable extent. HujwirT makes full use o f his principal source, but his over-riding concern is to shape the inherited material to his own needs, rather than to act as a faithful transm itter o f the inform ation provided in the foundational work o f the genre.

VI T he bodies o f the biographies in the K a sh f al-mahjüb typically consist o f segments about their respective subjects, followed by com m entary on their significance. In most cases w here the segments about the subject are initially presented in Arabic, a translation into Persian is also provided. A segment about the subject, followed (when appropriate) by its translation and com m entary together form an integrated ‘u n it’. T he body o f the biography o f Junayd consists o f a sequence o f five such units. As already m entioned, the first unit begins w ith a narrative segment about his relationship w ith Sari Saqati, to w hich reference is made immediately before, in the final elem ent o f the introduction o f the biography.58 O ne day Sari was asked, C an a disciple reach a higher level (daraja) than [his own] master? H e replied, Yes, the p ro o f o f this is manifest: Junayd’s rank is higher than mine. This utterance is modesty (taw aduc) from that great master (p it ). H e said w hat he said by means o f insight (basTrat), but no -o n e can see above oneself, for vision (dfdar) involves [looking] below. His utterance [therefore] is clear evidence

137

H u jw lri’s K a s h f al-m ahjub

that w hen he saw Junayd above his own level (martabat), although he perceived [Junayd as being] above, he was [really] below (KM, 161.11-15)

T he utterance attributed to Saqatl exalts Junayd by suggesting that he had surpassed his ow n master. H ujw iri’s commentary, however, is concerned w ith removing any am biguity about w hat the utterance may imply regarding Saqatl’s own status; H ujw lri argues that Saqati must have actually been at a higher level than Junayd w hen he made the com m ent. T he aim appears to be to depict Saqatl positively, and this is perhaps not unrelated to earlier polemical material about him .59 T he com m entary manages to fulfil this aim w ithout reducing the significance o f the original utterance w ith regard to the depiction o f Junayd, as an exceptional student w ho is destined to reach the highest rank. T he second unit in the biography o f Junayd reinforces the message o f the first one.60 It begins w ith another segment about the relationship betw een Saqatl and Junayd. T h e narrative describes Junayd’s elation on receiving directly from the Prophet in a dream the instruction to teach, w hich leads him to imagine that he has surpassed his own teacher. However, Saqatl, w ho had already instructed Junayd to start teaching for himself, shows that he is already aware o f Junayd’s dream w hen he asks why Junayd requires to receive the same instruction from the Prophet before having the confidence to act. This causes Junayd to recognise that Saqatl must indeed have a higher rank than himself. H ujw iri’s com m entary on this narrative consists o f the following remark ‘T here is clear evidence in this segment that the masters, regardless o f their characteristics, are aquainted w ith the inner state o f their disciples (wa andann hikdyat dalil-i wadih ast kay pTran bi-har sifat kay bashand mushrif-i hal-i m undan bashand)\61

A lthough the preceding narrative is introduced as one that is well know n (mashhur ), the only segment in this biography that can actually be traced back to an earlier w ritten source is that o f the third unit. In fact, this has been presented already in Chapter 1 above, for it is included in Sulaml’s Tabaqat,62 In that w ork it takes the form o f a dream narrative, in w hich K huldl asks Junayd about the relative authority o f the words o f the prophets and the words o f the ‘sincere ones’.63 In the K a s h f al-mahjub , only Junayd’s response is presented, as an isolated statement w ithout either being part o f a 138

H u jw iri’s K a s h f al-m ahjub

question/answ er construct or any allusions to a dream narrative context. H ujw lri’s com m entary on the segment refers to the controversy over the question o f superiority betw een saints and prophets. H e suggests that Junayd’s utterance illustrates that Prophets are superior, w hich is the view that he him self supports. H ujw iri also refers to ‘two groups o f heretics’ (mulhida ), in crossreference to a m ore detailed discussion o f the same controversy later in the w ork (in the chapter on Sufi ‘groups’).64 T he segment about Junayd in the fourth unit is a narrative describing an encounter w ith Satan (IblTs), w here the latter puts Junayd to the test unsuccessfully, for he is protected by God. H ujw iri completes the unit w ith the moral o f the story:‘God always protects his saints from Satan’s tricks’ (khudaw and ta a la aw liya’-i khud-ra andar hernia ahwal a z kaydha-yi shaytan nigah darad).65

T he final narrative, w hich concludes the biography o f Junayd, is presented below: It is related about [Junayd] that one day one o f his disciples suffered an inner affliction (ranjT bi-dit) and im agined that he had reached a [lofty] rank. H e became opposed to [Junayd]. O n e day [the disciple] came to test him. [Junayd] saw his intention because o f his ow n pre-em inence. [The disciple] asked him a question. Junayd said, D o you want a clear-cut ( cibaratt) answer or a spiritual one (m a'naw i )? H e said, Both. [Junayd] replied, If you w ant the former, it is that if you had tested yourself you w ould not have needed to test me, and you w ould not have com e here [with the intention of] testing. If you want a spiritual one, it is that I depose you o f your sainthood (wilayat). Immediately that disciple’s face turned black and he screamed, T he com fort o f certainty (yaqin) has been vanquished from my heart! H e im mersed him self in seeking forgiveness and gave up meddling. T hen Junayd said to him, You did not know that G o d ’s saints are the custodians o f the secrets (waliyan-i asrar) — you cannot bear their blows. H e cast a breath on [the disciple] and he returned to his [former] purpose, repenting from ever having acted independently in relation to the shaykhs. (KM, 163.14-164.5)

This narrative describes the relationship betw een Junayd and one o f his ow n disciples, w ho tries to test him. It attributes to Junayd 139

H u jw in ’s K a s h f al-m ahjub

the ability to read the intention o f his disciple’s question, and to strip him o f his spiritual attainments. T he latter is verified by the description o f the disciple’s im mediate feeling o f loss. Junayd explains to the disciple how he could read his intentions, ‘G o d ’s saints (awliya’) are the custodians o f the secrets’. T he narrative appears to be a culm ination o f the preceding material in the biography. T hat begins w ith a description o f the relationship betw een Junayd and his own (superior) master, Saqatl, and continues in the next unit, w hich possesses extensive com m entary on the superiority o f prophets over saints. T he m o tif o f being tested is introduced by the narrative describing Junayd’s trials w ith Satan. Junayd manages to succeed against him thanks to G od’s protection, just as in the final narrative, he sees through the questions o f his disciple to the hidden agenda. All the com ponent units in the body o f this biography, including both their segments about Junayd and any com m entary on them that is provided, are therefore linked together cohesively, producing a climax where Junayd is depicted as a Sufi saint w ith miraculous powers o f insight. It is perhaps no coincidence that the recurrent them e in this biography is the miraculous insight (firasa) o f Sufis, for the biographies o f Junayd in Sulaml’s Tabaqat and the H ilyat al-awliya’ both offer the same hadith transmission by him about precisely this topic. W hilst H ujw lri is no longer interested in including hadith transmissions in his biographies, his climactic narrative appears to be derived from the hadith transmission that had already becom e part o f the inherited biographical tradition o f Junayd.66 T he body o f the biography o f Abu Yazid contains only three units. Similar to the biography o f Junayd, the first unit appears to be an expansion on the introduction o f the biography, w hich attests to Abu Yazld’s com m itm ent to religious striving and the fulfillment o f correct conduct in spite o f the heresy that is falsely attributed to him (see above). This segment is in fact the utterance included already by SulamI w hich attests to A bu Yazld’s appreciation o f the differences o f opinion o f the scholars (ikh tila f a l-cu lam a’) and his concern to put into practice scholastic knowledge (7/m).67 H ujw lri’s com m entary explains why this task is both difficult and necessary (as a kind o f safety net).68 T he segment about Abu Yazid o f the next unit is the following utterance, w hich in Sulaml’s corresponding biography is juxta­ posed im mediately after its variant o f the aforem entioned ikh tila f utterance: ‘Heaven has no im portance in the opinion o f the lovers 140

H u jw iri’s K a s h f al-m ahjub

[of God], and the lovers are veiled by their love’ (al-janna la khatar laha cinda ahl al-mahabba wa-ahl al-mahabba mahjiibun bi-mahabbatihim ).69 T he fact that these two segments form the foci o f juxtaposed units in the K a s h f al-mahjub suggests that they are likely to have been obtained together directly from that source. H ujwlrl, in his com m entary to this utterance, explains that heaven is w ithout im portance to the lovers o f G od because it is (merely) created (makhliiq ),70 whilst love, being an attribute o f God, is uncreated (wa-mahabbat-i way sifat-i way-ast nam akhluq).71 T hen he explains that the lovers o f God are veiled by their love because love requires the duality o f lover and beloved, w hich is incom patible w ith divine unity (wa-dustan bi-dusti m ahjub-and a z anchi w ujud-i d usti d u ’T iqtida’ kunad wa-andar asl-i tawhid d u ’T surat nagirad).72

T he biography o f Abu Yazld culminates w ith the following unit, the initial segment o f w hich (about Abu Yazld) is not found in any earlier work. It is related that [Abu Yazld] said, O nce I w ent to M ecca and saw the house (the Kaaba) on its own. I said [to myself], T he pilgrimage is not acceptable for I have seen many stones o f this kind. I w ent again and saw the house and the lord o f the house. I said [to myself], This is still not the essence o f unity (haqiqat-i tawhid ). I w ent a third time and saw the lord o f the house but not the house itself. Inside me a voice exclaimed, Bayazld, if you were not to see yourself whilst seeing the w hole world, you w ould then no longer be a polytheist (.m ushrik ), but since you do not see the w hole world but you do see yourself, you remain a polytheist. T h en I repented, whilst at the same time repenting from perceiving my own existence. This is a subtle story about the soundness o f his inner state (hat) and a good example for those endow ed w ith such states. G od knows best. (KM, 134.4-11)

This autobiographical narrative parallels the preceding utterance attributed to Abu Yazld about G od’s lovers, according to the interpretation provided in the latter’s commentary. Just as the lovers place no im portance on heaven in relation to G o d ’s love because it is ‘created’, Abu Yazld, in his search for ‘the essence o f divine unity’ (haqiqat-i tawhid ), places no im portance on the perform ance o f a pilgrimage merely to ‘G o d ’s house’. B oth the 141

H u jw in ’s K a s h f al-m ahjub

lovers in the preceding u tterance and A bu Yazld in this autobiographical narrative are by implication attributed w ith an advanced status in relation to ordinary Muslims w ho are content w ith heaven and pilgrimage to the Kaaba. M oreover, in spite o f their heightened sensitivity, both the lovers and Abu Yazld ultimately remain short o f divine unity (tawhfd). In the form er case, it is because love requires the duality (d u ’t) o f lover and beloved, whilst in the latter Abu Yazld is advised that he still sees himself despite no longer seeing the Kaaba, and thus remains a polytheist. To recapitulate, whilst the segments about Abu Yazld in the last two units (out o f three) are related to each other thematically, neither o f them are related in the same way to the initial unit o f the body o f the biography. T hat unit (including both segment and comm entary) performs the function o f assuring the reader that Abu Yazid’s religious practice was sound, thereby serving as a preparation for the subsequent material. T hat is to say, it offers a secure foundation at the start o f the body o f the biography, on w hich m ore controversial opinions regarding the value o f heaven and pilgrimage can be presented. T he biographies o f Abu Yazld and Junayd exemplify the most distinctive feature o f H ujw lrl’s biographies in relation to other writings o f the Sufi tabaqat genre. For each o f them he has selected and arranged segments for inclusion in order to develop a central theme. The biographies in Qushayri’s Risala have also been observed to contain a num ber o f sequences o f thematically related segments, although, w ithout the aid o f commentary they are far from cohesive. T he dual-generic nature o f each o f these two works, which requires the arrangement o f com ponent segments according to topic for one section, whilst according to individual subject for the other, would have facilitated the merging o f the genres.

V II T he creativity in H ujw lri’s m ethod is most striking in his innovative inclusion o f topics that had previously not been brought to prom inence, such as the chapter on malama and the sub-sections on dancing, tearing o n e’s garments and gazing at youths during sama A nother conspicuous innovation is the provision o f the chapter on Sufi groups {giruhha) at the culm ination o f Section Two o f the K a s h f al-mahjub. 142

H u jw irï’s K a s h f al-m ahjub

U p until this final chapter, Section Two follows the familiar structure o f the tabaqat genre, by presenting in an overall chronological pattern the biographies o f religious authorities from the im m ediate successors o f the Prophet to the Sufi contem por­ aries o f the author. T he chapter on Sufi groups consists o f sub­ sections covering ten ‘accepted’ groups and two ‘rejected’ ones.73 Included amongst the ten accepted groups are the Tayfüriyya (the Tayfürians) and the Junaydiyya (the Junaydians), w hich are nam ed after (Tayfür) Abü Yazld al-Bastàmï and Junayd respectively. T he descriptions o f these two groups are taken up mostly w ith the account o f a debate over the question o f superiority between sobriety (sahw ) and intoxication (sukr). This debate is treated extensively in the description o f the Tayfürians, w hich im m edi­ ately precedes the description o f the Junaydians. T he Tayfürians are introduced as follows: T he Tayfürians: this group regard as their leader Abü Yazld Tayfür b. Tsâ ’1-Bastàmî. H e is one o f the Sufi chiefs and one o f their greats. His path is that o f rapture (ghalba) and intoxication (sukr). (KM, 228.18-229.1).

This introduction attributes to Abü Yazld, w hom the Tayfürians regard as their leader, a m ethod characterised by intoxication (sukr) and rapture (ghalba). However, in the rem ainder o f the description o f the Tayfürians, Abü Yazld is m entioned again only twice. In the first o f these instances it is reported that he considered intoxication (sukr) to be superior to sobriety (sahw).74 T h e second instance is the following narrative about an exchange w ith Yahyà b. M u'àdh, w hich is presented below together w ith H u jw lri’s com m entary: T he following story has been passed dow n about Abü Yazïd, [which is interpreted] the w rong way around (maqlüb): Yahyà b. M u'àdh w rote a letter to him, asking, W hat do you say regarding som eone w ho becomes drunk w ith one drop o f the sea o f love? Bàyazïd w rote in reply, W hat do you say regarding som eone who, if all the oceans o f the world were to becom e the wine o f love, would drink them all and still scream out about being thirsty? People assume that Yahyà has alluded to intoxication (sukr) and Bàyazïd to sobriety (sahw). T he opposite is the case for

143

H u jw iri’s K a s h f al-m ahjub

the possessor o f sobriety is the one w ho cannot bear a drop, while the possessor o f intoxication is the one who, in drunkenness, drinks everything and still needs more, since drinking is the means o f intoxication. It is m ore fitting for like to pair w ith like (Jins bi-jins). Sobriety is opposed to it; it does not rest comfortably w ith drink. (KM, 233.8-15)

T he above narrative is a variant in Persian o f one provided in the biography o f Abü Yazld in the H ilya t al-aw liya \ 75 It depicts Abü Yazld as rebuking Yahyâ b. M u 'ad h for boasting about becom ing d runk w ith a m ere drop o f the ‘sea o f love’ (bahr-i mahabba). Abü Yazld suggests in his response that it is better to have a greater tolerance, enabling one to thirst for even m ore o f it. This anecdote lends itself to the interpretation that Yahyâ boasted about his propensity for intoxication, whilst Abü Yazld held that rem aining sober despite drinking ‘oceans’ o f w ine is superior. HujwIrT him self admits that this is the usual interpretation, but argues that it should actually be interpreted the other way around (;maqlüb). This is in order to use the anecdote as evidence for his attribution to Abü Yazld o f the opinion that intoxication is superior to sobriety. HujwIrT suggests that a sober person cannot bear even a drop and therefore becomes drunk easily, whilst an intoxicated person can drink vast quantities because he is drunk already T he fact that HujwIrT resorts to using this particular narrative as the only one about Abü Yazïd w ith the message that intoxication is superior to sobriety, despite the fact that it can only serve his purposes if the m ore obvious interpretation is ignored in favour o f his innovative one, is understandable, since it is the only segment about Abü Yazld to be found in his earlier biographies w hich refers explicitly to intoxication.76 T he description o f the Tayfurians is taken up completely w ith H ujw ïrï’s account o f the debate about the question o f superiority betw een sobriety and intoxication. T he Tayfurians are said to prefer intoxication because it removes an individual’s self-control (.tadbfr) and free will (ikhtiyâr), which are considered to be veils over perfection. The proponents o f sobriety retort that intoxication is merely a distortion o f o n e’s perception, and so is itself such a veil.77 H ujw lrl states his preference for the latter view which he attributes to Junayd and his followers. 144

H u jw in ’s K a s h f al-m ahjub

M y ow n shaykh, w ho was o f the Junaydian school (JunaydT madhhab), used to say that intoxication is the playground o f children (bâzïgâh-i küdakân) whilst sobriety is the place o f the annihilation o f m en (fanâgâh-i mardân). I, cAlî b. cU thm an alJullàbï, say in agreem ent w ith my shaykh, that the perfection o f the state o f the possessor o f intoxication is sobriety. (KM, 232.4-6)

By this statem ent H ujw lri associates himself, through his own shaykh, w ith the ‘Junaydian school’. Even though he has indicated his ow n allegiance here, H ujw lri proceeds to offer a reconciliation betw een the two viewpoints in the debate. W h en the foundation is firm , sobriety (.sahw ) is like intoxication (.sukr ), and intoxication like sobriety, but w hen there is no foundation they are [distinct] as they are usually understood. In short, sobriety and intoxication are in the path o f men. They are caused by a difference o f opinion. But w hen the sultan o f the truth shows his beauty sobriety and in to x icatio n b o th appear to be in tru d ers ( tufayl) . . . R egarding this m atter [the poet] says, W h en the m orning appears w ith the star o f wine T he drunk and the sober will be the same. (KM, 234.6-234.11)

T he above passage shows H ujw ïri’s aim to reconcile the difference o f opinion in such a way that both sides can be considered legitimate. H e diverts attention from the debate by suggesting that such differences o f o p in io n seem like relatively w orthless ‘intruders’ w hen considered from the perspective o f union (‘w hen the sultan o f the truth shows his beauty’). T he section on the Junaydians begins w ith the following introduction: T he Junaydians. T he Junaydians regard as their leader Abü ’1Qàsim al-Junayd b. M uham m ad. In his time he used to be called ‘the peacock o f the scholars’ (tâwüs a l-culam â’), and he was the chief o f this sect and their leader o f leaders. His path is based on sobriety (sahw) contrary to the Tayfürians, and his difference o f opinion has [already] been m entioned. His is the most famous m ethod (madhhab) , and my shaykhs have all been Junaydian. Apart from this [mention] in the discussions

145

H u jw irl’s K a s h f al-m ahjub

o f the difference o f opinion, there is m uch attributed to him regarding conduct on the path. However, for fear o f prolixity, I have restricted it to this am ount, but if som eone should need to know m ore than this he must read from som ewhere else until it becomes clear, because my m ethod in this book is conciseness and desisting from prolixity. (KM, 235.1-8)

Despite attributing to Junayd glorious epithets and describing his m ethod as the most famous one in Sufism, H ujw rn surprisingly does not actually provide a detailed account o f that m ethod. Instead he refers the reader back to the previous section, about the Tayfurians, w here Junayd’s difference o f opinion (ikhtilàf) con­ cerning the relative status o f sobriety and intoxication is m entioned, as well as to other parts o f the book for his advice about conduct.78 T he fact that H ujw lrl does not offer here a detailed account o f ‘the most famous m eth o d ’ amongst the Sufis is itself an indication that this is not the aim o f his chapter on Sufi groups. R ath er the aim appears to be to outline acceptable differences o f opinion am ongst Sufis. H e achieves this by attributing opinions to various Sufi groups w ho are each nam ed after an illustrious early Sufi. H ence the Tayfurians and the Junaydians are only m entioned in the context o f the debate over the question o f superiority betw een sobriety and intoxication. T he sections on the Tayfurians and the Junaydians are two amongst ten sections in this chapter w hich describe ‘accepted Sufi groups’. They are typical o f such sections in that their prim ary concern is to explore a debate rather than offer a systematic description o f an actual Sufi group. T he fact that there is no corroboration for the existence o f most o f these Sufi groups outside o f H ujw lrl’s work supports the impression given by his m ethod o f describing them , namely that these groups are mostly his own constructs, designed to facilitate the accom m odation o f multifariousness amongst Sufis. To recapitulate, H ujw lrl’s K a sh f al-mahjub is a dual-generic Sufi manual that has been skilfully composed on the foundation o f the m ajor Sufi works that had already been w ritten. T he latter include Sulaml’s Tabaqàt a l-S xifyya , Q ushayrl’s Risàia and Sarràj’s Kitàb alL u m a \ all three o f w hich are referred to specifically by him. H ujw lrl’s K a s h f al-mahjub is an innovative work, in that it includes discussions o f topics that had not yet been included in the same 146

H u jw in ’s K a s h f al-m ahjub

level o f detail in any major w ork o f Sufism (e.g. malama, al-kharq , the legitimacy o f Hallaj, the diversity o f opinions amongst Sufi groups). In all these cases his aim is to accom m odate a contentious subject, and thus extend the boundaries o f acceptability in Sufism. A lthough H ujw lri may appear on occasion to be indecisive and incoherent, nonetheless the agenda behind these occasions remains consistent. In a num ber o f ways, Huj wiri is also the most overtly creative o f the authors w ho have contributed to the tabaqdt genre. This is highlighted not only in his deliberate inclusion o f the chapter on contem porary Sufi groups at the culm ination o f his tabaqdt section — a conspicuous m erging o f the tabaqdt and manual genres — but also in the form and content o f his actual biographies themselves; for each particular biography he has reworked and juxtaposed an exclusive selection o f material for the purposes o f developing a central them e, whilst he also provides extensive com m entary to facilitate the cem enting together o f the com ponent segments into a cohesive and unified whole.

147

This page intentionally left blank

Pa rt T h r e e

We are members o f the fie ld o f historicity, as storytellers, as novelists, as historians. We belong to history before telling stories or writing history. The game o f telling is included in the reality told. (Paul Ricoeur, ‘The narrative function’, p.294)

This page intentionally left blank

Chapter Six

Jam i’s Nafahat al-uns

I TV Tur al-Dm cAbd al-Rahman Jami (d.898/1492) is said to have X spent most o f his life in Herat, the main town near the place o f his birth. He received training in the religious sciences with a particular emphasis on scholastic theology (kalam) and the study of hadiths.1 At a young age he became the disciple o f the Naqshbandi Sufi khwaja,2 Sacd al-Dln Muhammad Kashgharl (d.860/1455). He is also linked to the person whom he describes as the most influential master o f this tradition during his own lifetime, namely Khwaja cUbaydullah Ahrar.3 Over forty works are ascribed to Jam!, most o f which have survived.4 They reveal that he was an accomplished author in a diverse range o f genres. Jam! is most renowned at a popular level for his poetry. His seven mathnawTs, known collectively as the Haft awrang (the seven thrones), include, in addition to reworkings of popular stories (e.g. LaylT wa M ajnun), a work written in honour o f the aforementioned "Ubaydullah Ahrar, called appropriately Tuhfat al-Ahrar.

The prose works ascribed to Jami include commentaries on parts o f the Qur’an and the hadith corpus as well as mystical works. The latter include a commentary on Ibn al-cArabfs Fusus alhikam .5 Jami also composed his own mystical writings, including the treatise called Lawa’ih (flashes) which is heavily influenced by the ideas o f Ibn al-cArabi.6 The most famous o f his prose works, however, is the Nafahat al-uns min hadarat al-quds, a collection of approximately six hundred biographies, covering the period from 151

Jan n ’s N afahat al-uns

the second/eighth century until the ninth/fifteenth century In the introduction o f this work, JamI informs that he was asked to compile it in the year 881/1476 by Amir Nizam al-D ln cAlI Shir N aw a’I (d.906/1500), the celebrated Chaghatay m inister and scholar.7 A num ber o f editions o f the Nafahat have been prepared, and it has been translated into Arabic and Turkish.8 R ecently M ahm ud Abedi prepared the first annotated edition, which is the one that has been used for this study. It is based on the six most im portant manuscripts o f the text, the oldest o f which is believed to date from 883/1478, that is, only two years after the reported request for JamI to compile the w ork.9

II T he N afahat al-uns consists o f an introduction10 and approximately 600 b io g rap h ies.11 T h e biographies are divided into tw o d isproportionate sequences: the m ain sequence consists o f approximately 570 biographies o f Sufi men, and it is followed by a comparatively short sequence o f only 34 biographies o f Sufi w om en, demarcated from the main sequence by a separate heading and introduction, thus constituting a form o f appendix.12 T he m ost im m ediately recognisable feature o f the main sequence o f biographies is that it begins with material obtained directly from the Persian Tabaqat al-Silfiyya. To be m ore precise, the influence o f that work is apparent in the first 322 biographies, from the first biography, that o f Abu Hashim al-Sufi, until that o f Shaykh Abu cAlI D aqqaq.13 O ver 250 (nearly 4/5) o f the biographies w hich are listed between the aforem entioned pair correspond to those found in the earlier work. M oreover the bulk o f their actual contents seems to have been obtained directly from it.14 T he first 322 biographies o f the N afahat thus constitute the first o f the three m ajor sections into w hich the main sequence o f biographies can be divided - one w hich is based closely on the Persian Tabaqat al-Sufiyya. T he biographies in this section are presented in approximately the same order as in the Persian Tabaqat itself. This means that Sulaml’s original ‘generation’ arrangem ent is also evident to a limited extent in the N afahat , filtered through the influence o f the interm ediary Persian Tabaqat. There are, however, a num ber o f 152

Jam fs N afahat al-tm s

discrepancies in the specific order o f presentation betw een the N afahat and the latter work, w hich may be instructive. T he most conspicuous change in position o f a biography concerns that o f Walld b. "Abdullah al-Saqqa’. W hilst it has been listed in the second generation section (out o f six) in the Persian Tabaqat alSu fiyya , it is found towards the very beginning o f the N afahat , following after the biography o f D hu ’1-N un.15 T he content o f the biography in both works is virtually the same - it is made up o f segments which are about D hu ’1-Nun and are transm itted by Saqqa’. This biography therefore appears to have been transferred in the later work to a position w hich is m ore appropriate in view o f its actual content. It has already been dem onstrated that a name classification system was used at one stage for ordering biographies in the Persian Tabaqat al-Siifiyya , and remains prom inent.16 Biographies in that w ork are often juxtaposed according to the names (usually the kunya) o f their subjects, creating, for example, groups o f juxtaposed biographies about Sufis w ith the kunya Abu Bakr in each o f the last four generation sections (3rd—6th) o f that w o rk .17 In the N afahat these four separate groups, as well as smaller groups and previously isolated biographies o f Sufis called Abu Bakr, have been m erged to create one continuous group o f (thirty-five) juxtaposed biographies. Since this group includes m em bers classified as belonging to a total o f four different generations, it testifies that the classification o f early Sufis into separate generations had lost its relevance by the time o f Jam fs late fifteenth century re-structuring o f the past.18 T he Persian Tabaqat is exceptional for a w ork o f its genre by virtue o f the fact that it neither begins w ith the biographies o f the generations o f the salaf nor links its first members back to them. This has not been missed in the process o f the redaction o f the w ork to fulfil the needs o f the Nafahat. W hilst the main sequence o f biographies begins in a similar fashion to the Persian Tabaqat, w ith Abu Hashim (‘the first person to be called Sufi’), it also includes, in the preceding introduction, the citation o f Q ushayrfs effort to explain how the first people know n as Sufis are the successors o f the salaf I19 Seventy o f the biographies listed betw een those o f Abu Hashim and Daqqaq (the overlapping section) in the N afahat are not found in the Persian Tabaqat at all. A num ber o f different criteria appear to have been used in deciding w here exactly to insert these new 153

Jam i’s N afahat al-uns

biographies w ithin the inherited framework. M ost frequently the new members have been positioned either next to a Sufi w ith the same name, or one w ith w hom the new m em ber is associated in some way; nineteen new biographies are found juxtaposed next to that o f an individual w ith the same k u n y a 2{) while five new biographies are similarly found juxtaposed next to that o f an individual w ith the same first nam e21 and three new biographies are found juxtaposed next to that o f an individual w ith the same nisba.22 Tw enty-nine o f the new biographies are found juxtaposed next to that o f an individual with w hom the subject o f the new biography was associated in some way.23 A few o f the remaining fourteen new ly-introduced biographies, each usually consisting o f just one narrative about their subject, appear to have been positioned im mediately after biographies containing narratives w ith the same m otif,24 whilst others are inserted next to those o f individuals from the same region.25 It w ould therefore seem that additional biographies were inserted into the inherited framework, according to whichever type o f association w ith an existing one was judged to be the most convenient.26 T he new additions include the only three biographies which had been listed originally in the first generation o f Sulaml’s Tabaqdt, but were om itted altogether from the Persian Tabaqdt alSufiyya. Furtherm ore, both the introductions and bodies o f these three biographies, w hich are devoted to Hatim b. cU nw an alAsamm, A hm ad b. Abl ’1-Hawari and 'Abdullah b. Khubayq alAntakl respectively, suggest that they have been obtained directly from Sulaml’s w ork.27 T heir inclusion indicates that Jam! was acquainted w ith both Sulaml’s Tabaqdt and the Persian Tabaqdt in sufficient detail to take note o f this discrepancy; he makes amends by including them after all o f the members w ho originally belonged to the first generation o f the Persian Tabaqdt. It has already been pointed out that, according to the headings o f the biographies in the Persian Tabaqdt m ore than one biography is provided for certain individuals,28 but in such cases usually only one o f them actually contains material about the subject himself. These are the only ones that are retained in the N afahat , as if the other ‘false’ biographies were identified as such, and rem oved.29 T he way in w hich the framework o f the Persian Tabaqdt has been transform ed in Jam i’s Nafahat therefore indicates that it has been redacted w ith the aims o f both tidying up the inconsistencies and incoherencies o f that organic text and adding new material 154

Ja n n ’s N afahat al-uns

considered relevant from Jam l’s later standpoint. O ne o f the most conspicuous effects is that the generation system, originally im ported into the Persian Tabaqat from Sulaml’s w ork, is obscured considerably. Associations by virtue o f the names or relationships o f their subjects determ ine the order o f biographies in the N afahat m ore often than their generation classification. It is perhaps understandable that, for a fifteenth century author such as Jami, the classification o f Sufis living betw een the eighth and tenth centuries into separate generations would have been less relevant than for his eleventh century predecessors in Sufi historiography. In fact, all the generations encompassed by the Persian Tabaqat are referred to collectively in the introduction o f the N afahat , simply as the mutaqaddiman (early m em bers).30

Ill T he biographies in the Nafahat w hich correspond to those found in the Persian Tabaqat usually consist o f two parts (the introduction and the body).T he biography o f Abu Yazld in the N afahat in fact is made up entirely o f material that has been provided already in the Persian Tabaqat. Its body consists o f a total o f ten segments o f material, all o f w hich are found, in the same order,31 amongst the first 23 segments (constituting the first 9 ‘bundles’)32 o f the corresponding biography in the earlier work. This would suggest that the Tabaqat was used systematically from the beginning until a sufficient am ount o f material had been extracted from it. T he segments o f material provided in that overlapping portion o f the earlier biography that were not selected from it include several segments o f addenda, that is, segments that do not refer directly to Abu Yazld himself.33 T he inappropriateness o f such segments for a biography o f Abu Yazld is probably the reason for their omission from the N afahat.34 A com parison o f the following representative passage in the N afahat w ith that o f corresponding material in the Persian Tabaqat may serve to elucidate the precise m ethods o f redaction applied: N 1 Shaykh al-Islam said, M any lies (durugh) have been told about Bayazid. O ne o f them is that he said, I arrived and pitched my tent opposite the throne o f G od (bardbar-i carsh).

155

Jam i’s N afahat al-uns

N 2 Shaykh al-Islam said, This utterance is infidelity (kufr ) in relation to the shariah and [an indication of] remoteness (bucd) in relation to the truth (haqfqat). Can you achieve the truth by projecting yourself to view? W hat is the truth? Escaping from yourself. Achieve the truth by your own n o n ­ existence (nabudjl Saying, Opposite [G od’s throne] about oneself is infidelity. Can you achieve divine unity (tawhfd) by duality (duganagi)? Descent is necessary not ascent! N 3 Husrl said, If I see the throne o f God then I am an infidel (kajir). (N, 54.22-55.6)

TA1 Shaykh al-Isldm said, M any lies (durugh) are said about Bayazld. O ne o f those that have been made up about him is that he said, I arrived and pitched my tent opposite the throne o f God (bardbar-i carsh). TA2 Shaykh al-Islam said, This utterance is infidelity (kufr) in relation to the shariah and [an indication of] remoteness (bucd) in relation to the truth (haqiqat). Can you achieve the truth by projecting yourself to view? W hat is the truth? Escaping from yourself. Achieve the truth by your own n o n ­ existence (nabudjl Saying, O pposite [G od’s throne], is infidelity. TA3 Husrl says, If I see the throne o f God then I am an infidel (m ulhid ), and if I were to arrive and pitch my tent, where exactly have you arrived? T A 4 Can you achieve divine unity (tawhfd) by duality (duganagtj? Descent is required, not ascent! (TABA, 88.14-89.6)

T he most obvious discrepancy between the two corresponding passages is that the version in the Persian Tabaqdt consists o f four segments, whilst that o f the N afahat consists o f only three segments. In the Tabaqdt, the first two segments (TA1, TA2) consist o f a report about Bayazld and com m ents on that report respectively, both o f w hich are attributed to Ansarl under the rubric Shaykh al-Islam guft. T he third segment (TA3) is a com m ent attributed to Abu ’1-Hasan al-Husri (d. 371/982) about the first segment (TA1). T he fourth segment (TA4) is a continuation o f the

156

Jam i’s N afahat al-uns

preceding segment o f com m ents attributed to Ansari (TA2), w hich seems to have been separated from it by the interpolation o f the intervening com m ent attributed to Husri (TA3). It is also probably due to that interpolation, em bedded betw een hortatory com m ents directed at an audience using the second-person, that it changes clumsily from the first to the second-person (w a’r shudam khaym a zadam kuja shudt).

T he above passage in Jam i’s Nafahat, w hich is clearly based on the Tabaqat version, differs by virtue o f its presentation o f the com m ent attributed to H usri (N3) after all the com m entary attributed to Ansari (N2) rather than in the middle, thus avoiding its separation into two parts. A nother discrepancy lies in the omission o f the problematic final clause o f the com m ent attributed to Husri. Furtherm ore, the version o f the N afahat is m ore concise, and it employs alternative vocabulary (e.g. ka fir/ mulhid) probably due to the changing connotations o f these terms in the different historical contexts. T he biography ofju n ay d in the Persian Tabaqat has been shown to be highly problematic as a result o f interpolation and organic grow th over an extended period.35 T he way in w hich it has been used as a source for the corresponding biography in the Nafahat may clarify how the effects o f growth and interpolation had been interpreted. T he body o f the biography o fju n ay d in the Nafahat consists o f twenty-five segments, nineteen o f w hich are provided already in the Persian Tabaqat, in the same order. However, only six o f these latter are based on material located strictly w ithin the parameters o f the biography o f Junayd in the earlier w ork;36 three juxtaposed segments correspond to similarly juxtaposed segments found in the chapter, ‘An issue about divine unity’,37 while the rem ainder o f the overlapping material corresponds to material located originally in its biography o f Abu ’l-cAbbas al-Surayj, w hich follows the aforem entioned chapter about divine unity, and similarly appears to be a growth from the biography o fju n ay d .38 T he fact that the biography o fju n a y d in the Nafahat contains several segments about him w hich are found originally outside o f his biography in the Persian Tabaqat, amongst sections that had grown out o f it, indicates that the problematic nature o f the earlier w ork was taken into consideration. T hat is to say, the redaction took into account the fact that material about Junayd is also found after the close o f his biography has been demarcated in the earlier work. Furtherm ore, the most conspicuous omission from the 157

Ja n n ’s N afahat al-iw s

material already provided w ithin the parameters o f the actual biography o f Junayd in the precursor is the most problematic passage o f all, namely the incoherent sequence o f overlapping segments related to the them e o f knowledge (7/m), w hich exemplifies the effects o f an extended process o f organic growth and interpolation in the Persian Tabaqat,39 T he first three segments in the body o f the biography o f Junayd in the N afahat are amongst those that are derived from sources additional to the Persian Tabaqat; in fact, they are Persian versions o f segments provided already in the corresponding biography o f 'Abdullah al-Yaficfs (d.768/1367) celebrated w ork o f the ta ’rikh genre, the M ir’at al-janan. A lthough Jam! does not name this work as th eir source w ith in these segm ents them selves, in his introduction to this biography he acknowledges the inclusion o f inform ation ‘in YafiYs ta ’r ik h \M) T he first o f the three segments that have been obtained from that source is presented below, followed by the original version. N O ne day in his childhood (aydm-i sighar) Junayd was playing w ith children. Sari Saqati said, H ey lad (ghulam ), what do you have to say about thankfulness (shukr )? H e said, Thankfulness is that you do not make use o f his favours (ni'am) in acts o f disobedience (m a cast) against him. Sari said, I fear very m uch that your share o f fortune lies only in your tongue. Junayd said, I was always frightened o f that remark until I went before him one day, having brought som ething that he needed — he said, R ejoice (bisharat bad), for I had requested from God that he send this to me in the hands o f som eone w ho w ould be successful (muflihT yd muwaffaqt)\ (N, 80.3-7)

M From his childhood (sighar) [Junayd] was made to speak about gnosis and legal regulations, such that w hen his m aternal uncle Sari was asked about thankfulness (shukr), while Junayd was playing w ith children, he asked him, Hey lad (ghulam), w hat do you say? [Junayd] replied, Thankfulness is that you do not make use o f his favours (n icam) in acts o f disobedience (m a cdsi) against him. Sari said, H ow m uch I fear for you that your fortune lies only in your tongue! Junayd said, I did not cease from fearing these words until I entered before him one day, having brought som ething that he needed — he said to me, R ejoice (abshir) for I appealed to 158

Jam i’s N afahat al-uns

G od to send that to me in the hands o f som eone w ho w ould be successful (m uflih aw qala muwaffaq)\ (al-YaficI, M ir’at, II, 234.3-8)

T hat the body o f the biography o f Junayd in the N afahat should begin w ith the above narrative (N) seems appropriate, since it describes events set in his childhood. It functions to confirm that Junayd was not only a clever speaker at a very early age, but also that he was set to continue to receive divine favour. An intriguing twist regarding the creation and developm ent o f the above narrative is revealed on com parison w ith the following, originally discrete segments o f material. T A Junayd says, As a seven-year old I was playing w ith my friend in front o f Sari Saqati’s teaching-circle (halqa). He called me back. I greeted him. H e was talking whilst standing stiffly. H e asked me, Hey boy, what is thankfulness (.shukr )? W hat is trust (taw akkul )? I replied, That you do not use his gift in disobedience (m a'siyat ). Sari said, It is clear that your share o f fortune from God is your tongue. (TABA, 168.9-13)

H I I heard Abu ’l-Hasan b. M iqsam say that he heard M uham m ad b. Sa'ld say that he heard Junayd b. M uham m ad, w hen he was asked about the essence o f thankfulness (haqTqat al-shukr ), say, T hat you do not make use o f any o f his favours (n i‘am) for disobedience (m a cast) against him. (H, X, 268.21-3)

H 2 Al-JahdamI related to me that M uham m ad b. al-Hasan related to him that Abu ’1-Qasim Bardan al-H awandl had related to him, saying that he heard Junayd say, I came to Abu ’l-Hasan al-Sari one day and knocked on his door. H e answered, W h o is that? I said, Junayd. H e said, C om e in! I entered, and there he was sitting alert. I had four dirhams w ith me, so I gave them to him. H e said, R ejoice (abshir) for you will succeed, because I needed these four dirhams and so I had prayed, O God, send it to me in the hands o f a man w ho will succeed (yufliti) before you! (H, X, 270.22-271.2)

B oth the segment in the Persian Tabaqat al-Sufiyya (TA) and the first o f the pair o f segments from the H ilya (H I) offer early variants 159

Jam i’s N afahat al-im s

o f the utterance about thankfulness (shukr), w hich eventually forms the first part o f the narrative provided by both YaficI and Jaml. The variant in the Persian Tabaqat (TA) is m ore elaborate than that o f the H ilya (H I), as it offers details about Junayd’s age and his actions at the time o f the encounter, as well as a final com m ent by Saqatl. M ost o f these additional details are preserved and elaborated in the later narratives. T he second segment o f the pair from the H ilya (H2) forms the basis o f the second part o f the eventual narrative. W ithout the continuation that this provides, it would have ended on a negative note (Junayd’s misgivings). W hilst in its earliest w ritten version this second part is an isolated narrative form ed around the m o tif o f giving m oney to a devotee o f God, once it is com bined w ith the narrative relating Saqatfs misgivings about Junayd, it provides reassurance that Junayd received m ore favours from God than merely his way with words.41 T he com plete (two-part) narrative had been com piled for the first time, to my knowledge, by Yafi'l, and was eventually obtained from that w ork by Jaml, in spite o f his detailed knowledge o f the earliest textual sources o f its com ponent parts. It suggests that he preferred complex narratives, and that their historicity was not his overriding concern. T he next two segments in this biography are also based on material in the corresponding biography o f YafiYs M ir’at. These segments, w hich are presented below followed by their earlier versions, consist o f a narrative and com m entary attributed to Yafi'i.

N Junayd said, Sari told me, H old a session (majlis) and speak to the people! I doubted my self (nafs) and did not consider it w orthy enough until I saw the Messenger in a dream on a Thursday night. H e said, Speak to the people! I woke up and w ent to the door o f Sari’s house before dawn. I knocked on the door. H e said, You did not judge me to be correct until you were told [by the Prophet]. T hen, at the break o f day, I held a session and started to speak. News spread that ‘Junayd is speaking’. A Christian youth, w ho was not wearing the usual clothing o f Christians, stood at the periphery o f the m eeting. H e asked, Shaykh, what is the m eaning o f the saying o f the M essenger o f God, Beware o f the insight (firasa) o f the believer for he sees by the light o f God? Junayd said, I hung my head down for a while, then I lifted it up and said,

160

Jam i’s N afahat al-uns

Embrace Islam for the time for you to becom e a M uslim has arrived! Imam YaficI says, People think that there is one miracle by Junayd in this, but I say that there are two: firstly, his knowledge o f the young m an’s infidelity; secondly, his knowledge that he would becom e a M uslim at that time. (N, 80.8-18)

M [Junayd] said, M y uncle Sari told me, Speak to the people! T here was reticence in my heart from speaking to the people and I used to doubt myself o f being w orthy o f that. I saw the M essenger o f G od in a dream on a Thursday night. H e said to me, Speak to the people! I went to [Sari] before dawn, and I knocked on the door. H e said to me, You would not believe until you were told [by the Prophet]! T he next day I held a session in the main m osque (jam ic) and started to speak. T he news spread amongst the people that ‘Junayd is holding a session to speak to the people’. A Christian youth came before me in disguise (m utanakkiran) . H e asked, Shaykh, what is the m eaning o f the saying o f the Messenger o f God, Beware o f the insight (firasa) o f the believer for he sees by the light o f God? Junayd said, I hung my head down for a while, then I lifted my head up and said to him, Embrace Islam for the time for you to becom e a Muslim has arrived, and the youth embraced Islam. People believe that there is one miracle o f Junayd in this, but I say that there are two: firstly, his knowledge o f the young m an’s infidelity; secondly, his knowledge that he w ould becom e a M uslim at that time. (al-YaficI, M ir’at, II, 231.18-232.7)

T he above narrative in the N afahat , w hich has been obtained from Yafi'l’s M ir’at , describes Junayd’s first teaching session. It is positioned appropriately towards the beginning o f the body o f the biography, preceded only by the narrative about Junayd’s childhood. To begin with, the differences betw een the two versions o f this narrative are w orth considering, because they exemplify how segments can be transformed, even in a close translation. For example, Jam i’s description o f the Christian youth (jawam tarsd na dar libds-i tarsayan bar kindr-i majlis bi-Tstad) is not merely a translation o f YafiYs description (fa-waqafa ca\ayya ghulam

161

Jan u ’s N afahat al-uns

nasranl m utanakkiran), but it is an interpretation o f the m eaning o f the relatively ambiguous mutanakkiran (in disguise), whilst also emphasising that the Christian is an outsider by situating him at the periphery o f the teaching circle.42 The omission o f the specific location o fju n ay d ’s first meeting, in the main mosque (b i-’l-ja m ic), may perhaps also be significant, especially since the same detail is similarly om itted on other occasions in the N afahat.43 T he above narrative can also be divided into two halves: at first, the diffident Junayd receives confirm ation from the Prophet that he should establish his own teaching sessions; subsequently, in his first teaching session, he receives a question from a Christian youth concerning a hadith w hich is specifically about miraculous insight (firasa), to w hich he responds by showing that he posseses that attribute himself. It is surely no coincidence that the hadith in question is the same one that is provided in the biographies o f Junayd in both Sulamfs Tabaqat al-Sufiyya and the H ilyat alawliya’.44 It had been presented in each o f those works simply as the example o f a hadith transmission by Junayd, but now it constitutes the focus o f a miracle story about him. H ujw lri’s K a s h f al-mahjilb also includes a narrative w hich describes Junayd as dem onstrating his possession o f firasa in a teaching session, albeit w ithout actually incorporating the hadith in question.4:> In that variant the questioner is one o fju n ay d ’s own disciples, w ho tries to test him. T he same biography also includes a variant o f the first part o f the above narrative, w hich describes Junayd’s excitem ent at receiving the instruction from the Prophet and his subsequent discovery that Saqatl, his teacher, knew about the dream all along. In that particular context, however, it functions to underline Saqatl’s pre-em inence over Junayd, as an example o f the superiority o f teachers over students.46 Self-doubt followed by a sign o f reassurance at the beginning o f o n e’s mission and the outsm arting o f doubters and infidels are o f course co m m o n anecdotal motifs in the biographies o f religious authorities in general (including that o f the Prophet). In the biographical tradition o f Junayd specifically, such motifs appear to have been already consolidated, and continued to be preserved, transm itted and reworked until eventually reaching the extended form o f the narrative in the Nafahat.

162

Jam i’s N afahat al-uns

IV T he introductions o f the biographies o f Abu Yazld and Junayd are representative o f those provided in the N afahat for the biographies that correspond to the ones already included in the Persian Tabaqat, in that they are both based closely on their precursors. For instance, the introduction to Junayd is presented below: T he C h ief o f the sect (.sayyid al-ta’ifa), Junayd al-Baghdadl

N1 Fie belongs to the second generation (az tabaqa-yi thaniya ).

N2 His kunya is Abu ’l-Qasim , while his laqabs are Qawarlri, Zajjaj, and Khazzaz. H e was called Qawarlri and Zajjaj because his father used to sell glass. According to YafiYs Ta’rikh , Khazzaz is w ritten w ith a dotted k h a ’ and a doubled and repeated z d \ H e was called Khazzaz because he traded in silk.

N3 His family origins (ast) are from Nihavand and the place w here he was born (mawlid) and brought up (m ansha’) is Baghdad. N4 H e followed the law school (madhhab ) o f Abu Thawr, ShafiYs best student. It has also been said that he followed the law school o f Sufyan Thawri.

N5 H e was a com panion (suhbat dasht) and student (shagird) o f Sari Saqati, H arith M uhasibl and M uham m ad Qassab.

N6 H e is one o f the leaders (ayimm a ) and chiefs (sadat) o f this group (the Sufis). Everyone links him self back to [Junayd] (,nisbat bi-way durust kunand), like Kharraz, R uw aym , Shibll and others.

N7 Abu ’l-cAbbas cAta says, O u r leader (imam) in this science ( cilm ), and the authority to w hom we refer (marjac) and w hom we imitate (muqtada ), is Junayd.

N8 T he Caliph o f Baghdad said to R uw aym , H ey illm annered one! H e said, Can I be ill-m annered if I have spent half a day in the com pany o f Junayd? This means that anyone w ho has spent half a day in his com pany cannot be illmannered, how then for longer?47

163

Jam i’s N afahat al-utis

N9 Shaykh Abu Jacfar Haddad says, If intellect ( caql) were a man it would be in the form o f Junayd.

N10 T he following has been said: There have been three peerless ones from this generation (tabaqa): Junayd in Baghdad, Abu "Abdullah Jalla in Syria and Abu "Uthman H lri in Nishapur.

N i l He passed away in the year 297, according to the Kitab al-Tabaqat and al-Risala al-Q ushayriyya , but according to YafiYs Ta’rikh he died in the year [2]98. It has also been said that fit was] in the year 299. God knows best. (N, 79.7-80.2)48

T he introduction o f the biography o f Junayd in the N afahat repeats most o f the material in the corresponding introduction provided in the Persian Tabaqat (much o f w hich was derived originally from Sulami’s ‘Kitab Tabaqat a l-S u fiyya '),49 whilst also including addi­ tional material from ‘Yafi'is Ta’rikh ’, namely the M ir’at al-jandn.S[) Elements N 7 and N 8 are originally found in the body o f the biography o f Junayd in the Persian Tabaqat, rather than in the introduction. They appear in that work amongst interpolations o f general statements about Junayd, w hich had accumulated towards the beginning o f the body o f the biography,and seem to have been intended for its introduction instead, since they w ould be more appropriate there.'"1 (They were perhaps interpolated sloppily, or became displaced during growth). T heir relocation in the N afahat to the introduction o f the biography has the effect o f turning them into ‘factual’ details,52 like his name and origins. In this way, Junayd’s intelligence ( caql) and his authority as a source o f reference (;marjac) and a role m odel (muqtada) are presented in a more appropriate context as established credentials. T he introduction to Abu Yazld, w hich is presented below, contains only a couple o f discrepancies w ith its precursor in the Persian Tabaqat: BayazTd BastamT

N1 H e belongs to the first generation (tabaqa-yi ’iild). N2 His name is Tayfur b. isa b. Adam b. Surushan. N3 His grandfather was a Magian w ho became a Muslim.

164

Ja n n ’s N afahat al-uns

N 4 Bayazld was a com p an io n (a z aqrdri) o f A hm ad Khadrawayh, Abu Hafs and Yahya M u'adh, and he had seen Shaqlq Balkhi. N 5 His death was in the year 261. It has also been said [that it was] in the year 234. T he form er is m ore correct. N 6 His teacher (ustadh) was a Kurd. O u t o f reverence for the teacher, [Bayazld] instructed in his will, Dig my grave lower than that o f my teacher! N 7 H e was a Hanafite (az ashab-i ra’y), but a [station of] sainthood opened up before him in w hich school allegiance (madhhab) was not apparent. (N, 54.14-21)

T he most conspicuous difference betw een the introductions in the two works is that two elements o f the version in the Persian Tabaqat are relocated in the N afahat (N6 and N7) to a position after the m ention o f Abu Yazid’s death, rather than before it.53 T he first o f these elem ents, N 6, relates to Abu Yazld’s burial, so its juxtaposition after the m ention o f the date o f his death makes it an expansion o f that detail. T he second o f these elements, N7, w hich has been relocated to the very end o f the introduction, alludes to the question o f Abu Yazld’s apparent failure to conform to a law-school (madhhab), suggesting that if this may have appeared to be the case it was only because o f his lofty mystical station, thereby implicitly excusing him. This is immediately followed by the previously cited passage from the beginning o f the body o f the biography, w hich also alludes to a controversial aspect o f Abu Yazid’s biographical tradition and absolves him, in that instance by asserting that illegitimate claims have been falsely attributed to him. In view o f the im mediately subsequent material, the position o f N 7 therefore also seems appropriate. T he com parison that has been made betw een the the Persian Tabaqat and the corresponding portion o f Jam l’s N afahat has highlighted the way in w hich the earlier w ork was redacted for Jam l’s purposes. In summary, the incoherencies and points o f disjuncture in that organic text have usually been tidied up and sm oothened over - the most problematic material is usually left out while m uch o f the material that is retained is altered in order to give greater coherency and cohesion. An effort to be concise is also

165

Jam i’s N afahat al-uns

apparent, both in the selection o f material to be included in a particular biography and in its adaptation. For instance, most o f the addenda in the earlier work, that is those segments in a biography w hich are actually about som eone other than the subject himself, are om itted altogether. M oreover, as one m ight expect, JamI, w ho was w riting in the fifteenth century, often employs m ore contem porary expressions in his reworkings o f the material at his disposal, and he also incorporates additional material from more recent sources.

V In the introduction to the N afahat al-uns , JamI presents the w ork as a continuation o f a literary tradition w hich originated with Sulaml’s Tabaqdt al-Sufiyya ,54 and was introduced into the Persian tradition in the form o f its namesake w hich is traditionally ascribed to Ansari. In the following passage, he offers an account o f the relationship betw een those eleventh century precursors: His Holiness Shaykh al-Islam , the shelter for m ankind, the protector o f the surma , the preventer o f innovation, Abu Isma'll "Abdullah b. M uham m ad al-Ansarl ’1-HarawI dictated (;imla) that (Sulaml’s Tabaqdt al-Siifiyya) in sessions (majalis) w ith companions and assemblies (m ajam ic) in w hich he w ould advise and adm onish. H e added to it further utterances belonging to some shaykhs (m ashdyikh ) w ho had not been m entioned in that book, and also some o f his own mystical experiences and inspirations (adhwaq wa mawajid). O ne o f his devotees and disciples collected them (the dictations) and com piled them into a book (ya kf a z muhibbdn wa muridan dn-ra j a m c mikarda wa dar qayd-i kitdbat miyawurda ). (N, 2.1-5)

M y ow n study o f the relationship between Sulaml’s Tabaqdt alSufiyya and its Persian namesake shows that, whilst the account o f their relationship in the above passage may be a fair inference from a superficial reading o f the texts, it is in fact inaccurate.55 O ne can safely assume that an author w ho had examined the two works as closely as JamI evidently has would also have realised that the relationship betw een them is m ore complex than he has suggested here.56 It should therefore be rem em bered that JamI was not a 166

Jam i’s N afa h a t al-uns

disinterested com m entator, w ithout any motive for emphasising that Sulaml’s w ork was the basis for that ascribed to Ansârï. In fact, after m aking the above assertion, Jam! subsequently points out that the N afahat is based directly on the Persian Tabaqât, and thereby implies that his ow n w ork is ultimately based on Sulaml’s oldest example o f the genre and must consequently be authoritative.57 In spite o f this, Jam i’s account o f the relationship betw een the two precursors to his ow n N afahat , rather than the implications o f the way in w hich he actually uses them both, has determ ined the opinion o f most m odern scholars. In the continuation o f the above passage taken from the main introduction o f the N afahat , Jam! offers his ow n explanation o f the way in w hich he has redacted the Persian Tabaqât: Truly that (the Persian Tabaqât al-Süfiyya) is a fine book and a noble collection, containing the truths o f Sufi gnosis and the subtle intricacies o f this sect. However, since it is w ritten in the old language o f H erat (zabân-i Haraun-yi qadxm), w hich was used at that time, and has now reached the state, where, on account o f the errors and the alterations o f scribes (tashxf wa tahrxf-i niwïsandagân) many sections are difficult to understand, and since it is also lim ited (muqtasar ) in its m ention o f some early members (mutaqaddimân ) and fails to m ention other [early] members, His Holiness S haykh al-Islam himself, his contem poraries (mu ‘âsirân) and the later members (m u ta ’akhkhirân), I have many times thought about trying to redact it (tahrxr wa taqrxr), in accordance w ith my ability and endurance, by: 1 conveying whatever is comprehensible w ith a w ell-know n contem porary expression; 2 leaving whatever is incom prehensible behind the veil o f concealment; 3 em broidering, on the ‘tablet o f clarification’ (lawh-x tabyxn), selected utterances and proven pieces o f gnosis w hich have been attached to them , from other esteemed books; 4 adding to the above the explanation o f the states (ahwât), stations (maqâmât), gnosis (ma'ârif), miracles (karâmât), and dates o f birth and death o f a group w ho are not m entioned in that book. (N, 2.5-14) 167

Jam i’s N afahat al-uns

JamI justifies his com position o f the N afahat by asserting that a redaction o f the Tabaqat was needed for the following reasons: because it is com posed in an archaic form o f Persian; because it contains errors and alterations that have been made by scribes (.tasluf wa tahnf); and because he considers it both incom plete (lacking the biographies o f some o f the earlier Sufis, as well as adequate material about those that are included) and outdated (lacking the biographies o f the later Sufis, including Ansari himself). This account o f his m ethod o f redaction corresponds largely to the conclusions reached by a com parison o f the two works. However, whilst JamI implies that he only leaves out the ‘incom prehensible’ material, that is not always the case. T he use o f material from Abu Yazid’s biography in the Tabaqat, for instance, shows that material was extracted in order from the beginning o f the biography, and that its final segments were all (indiscriminately) om itted once a sufficient total had been reached. JamI perhaps did not wish to declare that conciseness was also a concern for him, lest his own work be considered deficient. In the above passage JamI divides the Sufis o f the past into three broad chronological groups: i) the early members (mutaqaddiman ) ii) Ansari and his contem poraries (mu'asiran ) iii) the later members (m u ta ’akhkhiran) T he Sufis included in the Persian Tabaqat represent the mutaqaddim dn , and thus all o f them belong to a single chronological group from Jam i’s (fifteenth century) perspective. T he subsequent biographies in the N afahat must therefore represent the two latter chronological groups —Ansari and his contem poraries (mu'asiran), and the later m em bers (m u ta ’a k h k h ira n ).T h e three o f them represent the three sections into w hich the main sequence o f the N afahat is divided for the purpose o f the present analysis. T he first section, represented by the redaction o f the Persian Tabaqat, has already been identified as beginning w ith Abu Hashim (the first biography in the Nafahat) and ending w ith Daqqaq. However, the point o f demarcation betw een the second and third sections, representing Ansari and his contem poraries and the later Sufis respectively, is less easily defined. T he most obvious point is at the start o f the first o f a series o f coeval clusters o f juxtaposed biographies, each o f which represents Sufis living betw een the 168

Jam i’s N afahat al-uns

twelfth and fifteenth centuries (m u ta ’akhkhirân ). T he first cluster represents the Naqshbandl Sufi order from the twelfth century until Jâm ï’s ow n time. The biographies that intervene betw een this cluster and the preceding biography o f Daqqàq should therefore represent AnsarT and his contem poraries (m xiâsirân ). Appropriately, they consist o f the biographies o f Sufis living betw een the late tenth and the late eleventh centuries, including that o f Ansàrï himself.58

VI It seems w orthw hile to present at this point a sketch o f the organisational framework o f the groups representing the later Sufis. A lthough they represent periods that are not depicted by any o f the earlier works o f the same genre, nonetheless they may serve to illustrate the organisational principles applied in the Nafahat ‘independently’ o f its main precursors. W hilst every effort has been made to include only the necessary details here, it will be considerably easier to follow this broad outline by referring to the text itself, or at least to the diagrams provided in the appendix o f this book. An allusion has already been made to the existence w ithin the m u ta ’akhkhiran section o f clusters representing distinct sub-groups, such as the members o f a Naqshbandi line o f Sufism from the twelfth to the fifteenth century. In fact, the m u ta ’akhkhiran group includes the following eight clusters w hich account for the m ajority o f its constituent biographies: 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8

T he T he T he T he T he T he T he T he

Naqshbandi cluster cAyn al-Q udat al-H am adanl cluster cluster beginning w ith Abu Najlb al-Suhrawardl Jalal al-D ln al-Balkhl (Rum i) cluster cluster beginning w ith Shihab al-D ln al-Suhrawardl cAbd al-Q adir al-jill (GllanI) cluster M uhyl ’1-Din M uham m ad b. al-cArabI cluster Persian poets’ cluster

T he biographies in clusters 3.1, 3.3 and 3.5 each depict a Sufi line o f transmission extending as far as a contem porary o f Jam! w ho lived in the vicinity o f Herat, and they account for the m ajority o f the biographies in the m u ta ’akhkhiran section. Clusters 3.2, 3.4, 3.6 169

Jan n ’s N afahat al-uns

and 3.7 consist, in turn, o f the biographies o f cAyn al-Q udat alHamadanl, Jalal al-Dln al-Balkhl, cAbd al-Q adir al-jlll (al-Gilanl) and M uhyl ’1-Din M uham m ad b. al-cArabT, together w ith their respective associates — one specific individual represents the focus o f each o f these clusters. T he final cluster (3.8) contains the biographies o f poets w ho w rote in Persian. R etu rn in g to the Naqshbandi cluster, the sequence o f nineteen biographies beginning w ith that o f Yusuf HamadanI (d.535/1140) and culm inating with that o f "Ubaydullah Ahrar (d.896/1490) represents a line o f succesion o f the Naqshbandi order, to which JamI him self belonged (see Fig.2). Two o f the biographies in this cluster provide nothing m ore than a statem ent declaring that the subject was the successor o f the person w hose biography immediately precedes, im plying that the principal objective was to structure a cohesive and com plete Sufi lineage.59 A lthough most o f the Sufis listed in this cluster are nam ed as successors o f the im mediately preceding ones, the diachronic succession is disturbed at three points, such as w hen biographies have been incorporated that are entirely made up o f material about the subject o f a juxtaposed biography.60 The linear succession is also disturbed where the line branches into two, after the death o f N aqshband.61 Khwaja "Ubaydullah Ahrar is introduced as the representative o f the so-called Khwajagan tradition o f Sufism in Jam i’s tim e.62 His biography is followed by the unique provision o f a conclusion to the whole cluster, in w hich the tradition is glorified.63 For instance, the following four lines o f verse close this cluster, in the first hemistich o f w hich they are referred to by the name N aqshbandiyya : T he Naqshbandiyya are such amazing caravan leaders T hat they take the caravan through a hidden route to the havam. (N, 417.5-8)

T he position in w hich the Naqshbandi cluster has been included, as the first o f the coeval clusters amongst the m u ta ’akhkhiran, not to m ention its unique possession o f glorification o f the lineage, testifies to the allegiance o f JamI him self to this school o f Sufism, even though he does not explicitly acknowledge his own links to the order. W hilst he fails to m ention that he was Kashgharl’s student and son-in-law, towards the end o f the biography o f 170

Jam i’s N afahat al-im s

"Ubaydullah Ahrar in one m anuscript it is asserted that Jam! was a student o f the latter’s deputy in Bukhara, and was thereby linked to the individual w h o m he him self describes as the suprem e representative o f the Naqshbandi order in his ow n day.64 W hilst the Naqshbandi cluster may have been given special attention by JamI, the other line o f transmission clusters n one­ theless possess m any similar features. For instance, the sequence o f thirty biographies betw een those o f Abu Najlb al-Suhrawardl (d.563/1167) and Shams al-D ln Asad (d.864/1459) similarly structure a line o f transmission from the twelfth until the fifteenth century, whilst also including biographies that present material relevant to the subject o f a juxtaposed biography (see Fig.3).65 This cluster also depicts an eventual split into alternative lines o f transmission. For instance, whilst the last four members depict transmission as far as the aforem entioned Asad, they are preceded by an alternative line o f transmission extending as far as Fakhr alD ln Luristani (d.820/1417), w hom JamI claims to have m et in his childhood.66 T he fourteen biographies w hich make up the cluster that begins w ith Shihab al-D ln al-Suhrawardl (d.632/1234) and culminates w ith Shams al-D ln al-Kusu’I al-Jaml (d.863/1459) follows similar patterns in the process o f structuring a complete and cohesive lineage (see Fig. 4). T he clusters that are focused on one im portant individual are best exemplified by the relatively uncluttered example w hich is focused on Jalal al-Dln al-Balkhl, better know n today as R um i. It consists o f seven members, beginning w ith R u m i’s father Baha’ alD ln Walad and culm inating w ith R u m i’s son, Sultan Walad. His ow n biography, w hich is the longest in this cluster,67 is preceded not only by that o f his father, but also by that o f his teacher after the latter’s death, namely Burhan al-D ln M uhaqqiq.68 R u m i him self is followed immediately by his celebrated soul-m ate and the inspiration behind m uch o f his literary output, namely Shams al-D ln Tabriz!.69 Tabriz! is followed by two disciples o f R u m i,70 and finally the latter’s son Sultan Walad (d.712/1312), w ho is said to have studied at the feet o f all o f the previously listed members apart from his own grandfather (presumably because that w ould have been impossible).71 T he clusters centred around the biographies o f cAyn al-Q udat al-Hamadani, cAbd al-Q adir al-Jil! (al-Gllan!) and M uhyl ’1-Din M uham m ad b. al-cArabI are each structured on the basis o f a similar pattern to that o f the R u m i cluster, nam ing the teachers as 171

Jam i’s N afahat al-uns

well as a couple o f generations o f the students o f their respective ‘key m em bers’. They also include biographies that serve simply to provide further inform ation about the key member. For instance, the final two biographies in the cAyn al-Q udat cluster are made up almost entirely o f quotations o f the latter.72 T he final eleven members o f the main sequence o f biographies in the N afahat constitute the cluster o f Persian poets. They are listed in approximately chronological order,73 but some o f them are juxtaposed only because their subjects are associated w ith each other.74 T he subject o f the first biography o f this cluster, Hakim Sana’! (d.ca.525/1131) is introduced as a disciple o f Yusuf Hamadanl (d.535/1140), the first m em ber o f the Naqshbandi cluster.75 In this way, the whole cluster o f Sufi poets is implicitly linked back to the same tradition o f Sufism as JamI, w ho himself com bined his Sufi affiliation with the com position o f highlyesteemed Persian poetry. T he very last biography in the main sequence o f the Nafahat is that o f Shams al-D in M uham m ad al-Hafiz, arguably the last great Persian poet before JamI himself.76 It is remarkably brief, but this is explained as being only on account o f the fact that he is already very famous.77 Jam! concedes that it is unknow n w hether Hafiz was him self a Sufi or not, but he explains that his poetry is always used by Sufis, and has even been formally approved by a m em ber o f the Naqshbandi line.78 T he inclusion o f this cluster at all in a work o f the Sufi tabaqat genre is an innovation influenced by Jam i’s own participation in the tradition o f Persian poetry. H e appears to have wished to include his precursors in this field, but found it impossible to integrate them into an earlier part o f the framework o f the Nafahat. They have therefore been tagged on at the end as a separate cluster, whilst every effort is made to attribute to them collectively a link back to his own Naqshbandi tradition o f Sufism.

V II In summary, the biography o f Daqqaq represents the end o f the first section o f biographies in the N a fahat , that o f the mutaqaddim an , w hich covers the eighth until the eleventh century. T he biography o f Yusuf Ham adanl represents the beginning o f the third section, that o f the m u ta ’akhkhirdn , w hich covers the twelfth until the 172

Jam i’s N afahat al-uns

fifteenth century. T he portion o f the main sequence o f biographies betw een these two points is accounted for by the biographies o f Ansari and his contem poraries (m u casiran), living betw een the eleventh and twelfth centuries. This interm ediary section bridges the gap betw een the later members (m u ta ’akhkhiran ) and the authoritative earlier members (m utaqaddiman ), whilst at the same time it includes the most im portant contem poraries o f Ansari. The framework thereby depicts a continuous transmission o f the tradition, from the first generations to the contem poraries o f Jami. A diversity o f medieval Sufi orders and outstanding personalities are thus presented as the heirs o f the Prophet, by transmission through the contem poraries o f Ansari, the earlier members (;mutaqaddiman) and the salaf. T he cohesiveness o f the pattern o f succession that is structured is further enhanced by means o f the attribution o f specific links betw een individual members o f the three sections. For instance, Yusuf HamadanI, the first m em ber o f the Naqshbandi cluster, is said to have a ‘w ell-know n’ Sufi link w ith Abu cAlI FarmadI (d.477/1084), w ho is one o f the final members o f the m u a sira n J 9 Similarly Abu Najlb al-Suhrawardl, and through him his nephew Shihab al-Dln, are linked back to Ahm ad Ghazali (d.520/1126).80 These individual associations have the effect o f linking back the entire Sufi lineage clusters headed by each o f these three representatives o f the m u ta ’akhkhiran (and extending as far as a contem porary o f Jami) back to the m uasiran. T he biographies o f FarmadI and Ghazali play key roles in the enhancem ent o f cohesion betw een the three broad chronological groups, because they themselves are linked back in a similar fashion ultimately to representatives amongst the mutaqaddiman. FarmadI derives his authority in Sufism (intisab-i way dar tasawwuj) from two individuals, namely, Abu ’1-Qasim GurganI and Abu ’1-Hasan KharaqanI, w ho are both amongst the first members o f the m uasiran. Ahm ad Ghazali is said to have been a disciple o f the subject o f the biography w hich immediately precedes his own, namely A bu Bakr al-Nassaj (d.487/1094), who, in turn, is described as an associate o f GurganI.81 It is particularly significant that KharaqanI and GurganI should have been chosen as the ones to w hom the later generations are linked back, for each o f them is attributed w ith a Sufi lineage in the introduction o f his own biography.

173

Jam i’s N afahat al-uns

Shaykh Abu ’l-H asan’s (KharaqanT) Sufi lineage (intisab ) goes back to ‘the sultan o f the gnostics’ (sultan a l-carifm) Shaykh Abu Yazld BastamT. His training in Sufi conduct (tarbiyat-i Tshan dar suluk) was from the spirituality (ruhaniyyat) o f Abu Yazld. Shaykh Abu ’l-H asan’s birth was a while after the death o f Shaykh Abu Yazld. (N, 303.18-20)

KharaqanT (d.425/1034) was born after Abu Yazld (d. 261/875) had died, yet he is linked directly back to this illustrious m em ber o f the mutaqaddiman through the latter’s ‘spirituality’ (ruhaniyyat), presumably referring to his influence beyond the grave. G urganfs (d.469/1076) Sufi lineage is traced, through a sequence o f three interm ediaries, whose lives appropriately overlapped, back as far as Junayd: His lineage (nisbaf) reaches, by three interm ediaries (wasita), namely Shaykh Abu cU thm an M aghribI, Shaykh Abu cAlI Katib and Shaykh Abu cAlI RudbarT, to ‘the chief o f the sect’ (sayyid al-tdyifa ), Junayd. (N, 312.14-16)

As m en tioned above, the link betw een the m u ta ’akhkhiran biographies and the mu'asiran w hich leads to KharaqanT does so through Abu cAlI FarmadT, w ho is associated at the same time with GurganT. This ‘double affiliation’ is perhaps due to the nature o f KharaqanT’s relationship w ith Abu YazTd; GurganT’s com plete chain o f transmission back to Junayd reinforces the value o f association w ith FarmadT, as shown in the following diagrammatic summary o f the aforem entioned links.

Kharaqani

Abu Yazid

Gurgam

Junayd

Farmadi

Nassaj -A hm ad Ghazalim u ta ’akhkhiran

174

mu asiran

mutaqaddiman

Jam i’s Nafaha t al-uns

T he mu asirán group thus fulfils its bridging role by providing, at its culm ination, links to the m u ta ’akhkhirán, and at its start, links back to the mutaqaddim án , w ho are represented by our two heroes Abü Yazld and Junayd. W hilst Junayd had already been depicted, as far back as the first work o f the tabaqat genre, as the most authoritative representative o f the early generations, the same cannot be said for Abü Yazld. In the framework o f the N afahat , however, his biography plays a similar role to that o f Junayd, as an authoritative ‘com m on-link’ amongst the mutaqaddimán on w hom the chains o f transmission o f later generations converge.82 This testifies not only to the fact that his status had been consolidated as authoritative by the late fifteenth century, but also that his reputation had even surpassed that o f his rivals amongst his contem poraries, w ho had been considered m ore em inent by most o f Jam i’s precursors.83 It has already been observed in the study o f earlier writings o f the tabaqat genre that the allegiances and predilections o f the author (or the tradition that produced the works) are indicated by their organisational framework. T he biographies representing that allegiance most directly are usually those o f the latest members o f the tradition and are included at the end, as the culm ination o f the past o f Sufism. For instance, in Sulaml’s Tabaqat al-Silfiyya, the culm ination takes the form o f the final generation o f biographies. Since such generations are hierarchical, the first biographies to be listed w ithin them are the most instructive w ith regard to his allegiances and predilections. T he fram ework o f the N afahat is comparable to that o f Sulami’s work, since Jam i’s im mediate predecessors are included as members o f the series o f coeval clusters representing the m u ta ’akhkhiran. T he provision o f these coeval clusters in the N afahat parallels the provision o f the biographies o f im mediate predecessors in Sulami’s final generation. Just as the first members o f those generations are the most im portant, so too the first o f the coeval clusters to be presented is the most significant, namely the N aqshbandi cluster w hich represents Jam i’s own tradition o f Sufism. W hilst in the Tabaqat the length o f the head biographies is also an indication o f their pre­ eminence, in the N afahat this is indicated by the Naqshbandi cluster’s unique possession o f praise and glorification o f their tradition. Jam! can therefore be seen to have applied the same principles for the fram ework o f the N afahat as had been applied in the works 175

Jam i’s Nafahat a l-u m

o f his predecessors. Unlike them , he lived at a time after Sufi orders had becom e established, and each had produced biogra­ phical literature devoted exclusively to the predecessors o f their own specific order, many examples o f w hich have been used by Jàm ï for his m ore comprehensive project.84 Nonetheless, Jam! manages to integrate his selection o f Sufi lineages, as well as further em inent individuals o f later times, by representing them in the form o f clusters o f biographies, each o f w hich is linked back to the earlier Sufi authorities. T he authority o f those Sufis living before the eleventh century had been established by the earlier tabaqât works. Therefore, they were depicted by Jam! in the late fifteenth century, as the solid foundation w hich bolsters the authority o f the diversity o f later traditions that he decided to include in his relatively broad definition o f Sufism.

176

C onclusion

Tim e is no longer primarily a g u lf to be bridged, because it separates, but it is actually the supportive ground o f process in which the present is rooted. (H-G. Gadamer, Truth and method, p.264)

he tabaqat writings w hich have been examined in this study display close inter-relationships w ith each other at every structural level, from their overall organisational frameworks to the specific segments o f biographical material that have been passed down through the tradition in differing guises. For instance, vestiges o f the organisational framework o f SulamT’s early eleventh century Tabaqât al-Sufiyya are still discernible in Jam l’s late fifteenth century N afahât al-uns. Furtherm ore, the latter’s m ethods o f arrangem ent for the biographies o f Sufis living after the twelfth century follow the same principles as Sulamfs foundational generation system, and his biographical segments fulfil similar functions, albeit in a m anner appropriate for the new historical context. T he arrangem ent o f biographies in each o f the six tabaqat writings depicts an overall linear time-span, culm inating w ith the biographies o f the immediate predecessors o f the author (or the person to w hom the text is traditionally ascribed). A lthough they depict the time-span in a chronological order, all o f them are backreadings. T hat is, from the standpoint o f the time w hen the works were produced, their organisational frameworks serve to trace a route from as far back as a generation o f predecessors whose authority is already established in tradition. Those works w hich A.

177

Conclusion

appear to be under the control o f a single author offer the most cohesive and ordered frameworks for this task o f asserting the unbroken transmission o f the tradition from the beginnings o f Islam. T he functions o f the tabaqat genre are highlighted most clearly w hen they are juxtaposed next to contrasting genres as part o f the same literary work. In both Qushayrfs Risala and H ujw iri’s K a sh f al-mahjüb , the tabaqat sections serve to buttress the authority o f subsequent accounts o f Sufi theory and practice. In the form er work, it is followed im mediately by a section w hich is introduced as having been designed specifically for the benefit o f the initiated. In the latter work, it is closely related to an account o f the differences o f opinion amongst Sufis regarding contentious aspects o f their doctrine. T he relationship betw een the respective tabaqat sections and the sections w hich are juxtaposed after them is comparable to that between an isnad (chain o f authority) and the main (text) w hich it supports. T he other tabaqat writings also have a matn to w hich they give support; the culm ination o f their frameworks at the im mediate predecessors o f the assumed authors would suggest that they serve to buttress the authority o f the tradition o f Sufism that produced them. T he selection o f biographies to be included in a tabaqat work is a way o f indicating a definition o f the tradition whose past is being represented, in that it structures a diachronic community, the parameters o f whose identity are demarcated by the characteristics o f its individual members. B oth the established reputations o f these members as well as the actual way in which they are now portrayed shape this definition. Amongst the Sufi tabaqat writings under the control o f a single author, the tabaqat section o f H ujw iri’s K a s h f almahjilb , by its deliberate inclusion o f jurists and controversial mystics alike, offers the most accom m odating parameters. This is in keeping w ith the overall character o f that work. O n the other hand, the tabaqat section o f Q ushayri’s Risdla offers a contrastingly narrow definition o f Sufism; it not Ónly excludes the controversial Halláj, but it portrays the Sufis w ho are included as little m ore than traditionalist pietists, even at the cost o f om itting many o f the characteristics w ith w hich they had previously been described in source works. This is in keeping w ith its expressed agenda. T he order in w hich the members o f the diachronic com m unity depicted in a tabaqat w ork are listed may also be instructive with regards to the priorities o f the author. For instance, Sulami’s 178

Conclusion

arrangem ent o f biographies in successive generations allows a hierarchical m ethod o f ordering the mem bers o f the same generation. In this way, he manages to prom ote a tradition centred on Junayd as the pre-em inent tradition o f Sufism. Sulaml’s generation system forms the basis o f the representations o f the past o f Sufism as far as the eleventh century in the frameworks o f all the later Sufi tabaqat writings. JamT’s depiction o f Sufis living after the twelfth century is similar to this, in that it includes clusters o f biographies w hich depict Sufi lineages, each spanning the same period. T he first coeval cluster o f this kind to be presented is that o f his own tradition, the Naqshbandi order, w hich is thereby given precedence over all the com peting traditions. T he m ajority o f the com ponent segments o f the biographies seem to have been produced originally in an oral teaching situation. Furtherm ore their earliest presentation in literary works suggests that they already had an extended history, and may even have been preserved in the form o f notes prior to compilation. Such a context for the generation and transmission o f segments facilitated the frequent inclusion o f variant utterances and recurrence o f anecdotal topoi, as well as confusion in many instances over the identity o f the individual to w hom the dicta and deeds should be attributed. O ne need look no further than the biographical traditions o f Abu Yazld and Junayd to observe such features, including the recurrence o f the same anecdotal topoi in both biographies and the attribution o f variants o f the same utterance to each o f them . And this is in spite o f the fact that in later depictions they represent opposite types o f mysticism. A com parison o f the presentation o f biographical segments in the tabaqat works considered indicates that they are creatively re­ w orked through time. As we have seen, the modifications vary from the omission o f seemingly insignificant detail to the transform ation o f whole segments, resulting in entirely different messages being conveyed from their earlier variants. W hilst some o f the modifications may seem negligible in isolation, the new context in w hich they are presented is what needs to be taken into consideration. For instance, Qushayrfs Risala is a w ork w hich follows a particular agenda that is evident at every level o f its structure. Segments may also be transform ed significantly in form only in order to convey the same message m ore effectively for a new generation o f readers. This is best illustrated by the com bination o f 179

C onclusion

originally discrete segments to form the extended narratives that are presented in Jam i’s Nafahat. In one case a hadith transmission attributed to Junayd in earlier biographies becomes an integral part o f a miracle story — hadith transmissions had been surpassed by miracle stories by this time as evidence for the authenticity o f Sufis. Biographical segments becom e transform ed not only through successive works, but also w ithin the same work w hen that has undergone an extended period o f growth. W hilst on the one hand the organic nature o f such works raises questions about the dating and authenticity o f their contents, it also provides an invaluable insight into the dynamic process o f re-form ing the past by the ongoing school traditions. T he biographies o f the most im portant figures o f the past (from the view point o f the tradition that produced them) are particularly valuable in this regard, for they have been reworked to the greatest degree precisely because o f their continued influence. T he effects o f growth, interpolation and successive redactions in organic works, such as the Persian Tabaqdt al-Sufiyya and the H ilyat al-aw liyd\ can be witnessed at every level o f their structure, from the extended developm ent o f com ponent segments to the com peting organisational principles in their overall frameworks. Perhaps the most conspicuous indication lies in the inappropriate position o f many introductions o f biographies in the surviving forms o f the works —after a process o f growth and interpolation, as well as successive redactions, they are sometimes found strangely out o f context. T he Persian Tabaqdt has already generated an explanation in the tradition itself for its incoherent and organic form, but the H ilya shares several o f its characteristics, albeit to a less emphatic degree. It also merits classification as an organic text that underw ent an extended period o f growth beyond the death o f the individual to w hom it is attributed. Each o f the m ajor tabaqdt writings o f the Sufi tradition may seem at first sight to be little m ore than a repository o f facts about the lives o f its subjects, but on closer inspection none o f them suggests a concern for the preservation and transmission o f material in its original form. T hat is to say, rather than being concerned about historical facts, they display a perpetual creativity in their individual re-workings o f the past, w hich is evident in the selection and arrangement, as well as the modification o f the material obtained from precursors. T he perception o f works o f this genre by positivist historians as repositories o f factual inform ation 180

Conclusion

therefore appears to be short-sighted and inappropriate. Tabaqat works are too sophisticated for such a superficial reading, w hich is deceptively convenient for those w ho seek instant ‘facts’. T he form and content o f the works examined in this analysis support the suggestions in recent theories o f historiography that any depiction o f the past is inseparable from the present life o f its author, and that the actual occurrence o f the author’s re-collection o f the past is o f greater historical significance than the content o f his description. As products o f the ‘labour and savoir-faire’1 o f the individuals and school traditions w hich com piled them , tabaqat works remain invaluable as sources for the time in w hich they were produced. A degree o f scepticism towards their historical value for the past w hich they describe therefore, far from negating the value o f one o f the most prolific genres o f medieval Islamic literature, opens up the possibility o f a re-orientation towards their value as sources for the contexts o f their own form ation. T he tabaqat works considered here may serve to demonstrate emphatically that Sufism shared w ith the other main traditions o f M uslim scholarship in the pre-m odern period ‘a concept o f authority based upon precedent’.2 That is not to say that past m odels were necessarily im itated slavishly generation after generation, but rather that the developing traditions, including Sufism, sought support and legitim ation through creative appeals to precedent, in that the present context often provided ‘[both] the impetus and the desired results’.3 T he Sufi tabaqat genre therefore represents a dynamic herm eneutical process, w hich can serve as an indicator o f the vitality o f Sufism in different historical periods;4 it informs o f the tradition’s ability to continue to find significance in the past, repeatedly recalling and re-form ing it for its present needs. As Paul R ico eu r has com m ented: ‘T he past is not passé for our future is guaranteed precisely by our ability to possess a narrative identity, to recollect the past in historical or fictive form ’.5

181

This page intentionally left blank

N otes

Introduction 1 See Auchterlonie, Arabic biographical dictionaries; Gilliot, ‘tabakat’ in E l2; Hafsi, ‘R echerches’; Heffening, ‘tabakat’ in E l1. 2 The works o f Richard Bulliet exemplify this approach (e.g. see his Patricians o f Nishapur).

3 Wansbrough, R es ipsa loquitur, 10. M ore specifically, see R oy M ottahedeh’s review o f Bulliet’s Patricians, and Humphreys, Islamic History, 198—9. See further, Meisami, Persian Historiography, esp. 1-14; Waldman, Historical Narrative, 3 -2 5 . 4 The Sufi manuals that were written prior to these two works, despite containing a certain amount o f biographical material, do not contain tabaqdt sections as such. See al-Sarraj, Kitab a l-L u m a c; al-Makkl, Q u t al-qulub; alKalabadhl, Kitdb a l-Tacarruf. 5 See pp 6 5 -7 below. 6 See C h .4, especially Sections III and IV. 7 See pp 204, n.13 below. 8 There are two major collections o f Sufi biographies ascribed to authors living before the fifteenth century w hich are not included in this study, namely Ibn al-Mulaqqin’s Tabaqdt a l-A w liyd ’ and Farid al-Dln "Attar’s Tadhkirat a l-A w liya ’ (see bibliography for details). They both differ significantly in structure from the works w hich are taken into account in this study, in that their biographies are not arranged in a predominantly chronological pattern. Ibn al-Mulaqqin’s Tabaqdt, w h ich lists biographies mostly according to an alphabetical arrangement, represents a western (North African) tradition o f Sufi literature, with w hich Jam! does not appear to be familiar. 'Attar’s Tadhkira also does not possess a predominantly chronological organisational framework, and it appears to have been written for literary purposes other than those o f the tabaqdt genre. This perhaps explains why Jam! lists it amongst the works ascribed to the Persian poet ‘Attar, in his biography o f the latter, but chooses neither to name it as a precursor, nor to use it as a source o f material for his Nafahat.

183

N otes 9 W hilst I have vocalized this nisba here in accordance with recent scholarship (especially in German), in the English tradition the same name has conventionally been vocalized as BistamT. M ost references to this personality in the present work instead employ the kunya ‘Abu Yazld’, corresponding to the references provided in source works. See further the entry o f the nisba in al-Sam'anl, al-Ansab. 10 Gadamer, Truth and method, 301. 11 J.A. Mojaddedi, ‘Legitimizing Sufism in al-Qushayrfs R i s a l a S t u d i a Islamica 90 (2000), 37-50. See also the list o f errata due to be published in the next issue.

Chapter One: Sulam i’s Tabaqat al-Sufiyya 1 For a list o f the extant works ascribed to SulamI, see Sezgin, Geschichte, I, 6 7 1 -4 . 2 See al-Sulami, A dah al-suhba, 1 -2 . 3 al-Baghdadl, Ta’nkh , II, 2 4 8 .1 5 -1 6 . 4 See M inuvi’s remarks on the connotations o f duwayra in Furuzanfar, Tarjumayi Risala, 19, note 1. This term is used for a small monastery which has residents. It may be located within the grounds o f a school (madrasa). Later tradition attributes to SulamI a school in w hich he trained students. 5 al-Baghdadl, Ta’nkh, II, 248.4. 6 This nisba refers to the Arab tribe Banti Sulaym. See al-Sain'anl, al-Ansab, VII, 1 11-3. 7 TABS, 4 7 6.1-5; 1 59.1-7. 8 See al-Subkl, Tabaqat, III, 223.21. For information about the M aldmatiyya, see D e Jong, Algar & Imber, ‘al-Malamatiyya’ in EL. 9 See Hartmann, ‘As-Sulaml’s Risdlat al-M alam atiyya’. 10 See al-Sarnanl, al-Ansab, VII, 112.18-113.15. 11 Nasrollah Pourjavady has prepared an edition o f this text, and included it in his recently published collection o f Sulami’s works, M a jm u ca-yi athdr-i A b u cA b d al-Rahm dn Sulann.

12 See further R , 348.3—20, w hich consists o f a narrative in which Abu ’1Qasim al-Qushayrl is instructed by his teacher, Abu cAli al-Daqqaq, to visit SulamI and take, without his knowledge, his copy o f a work containing poems by al-Husayn b. Mansur al-Hallaj. SulamI displays his miraculous insight (firdsa), convincing Qushayri to abandon his plan to fulfil the instruction. The same narrative also refers to criticism about Sulami’s participation in sam dc. It seems to serve the function o f depicting SulamI as an upholder o f controversial views, w ho is nonetheless proved to be correct by his miraculous powers. 13 See the studies by Gerhard Bowering: The minor Q u r’an commentary, and ‘The Q ur’an commentary o f al-Sulami’. 14 Another named work com m only attributed to SulamI is the so-called Ta’nkh al-Sufiyya. This work is not extant, and the references to it offer conflicting impressions; the title is used in many later works (perhaps generically) to refer to the Tabaqat, but it is also used to refer to historiographical writings o f SulamI covering periods beyond the scope o f that work. In the excellent

184

N otes

15

16

17

18 19 20 21

22 23

24

25 26 27 28

introduction to his edition o f the Tabaqat, Pedersen discusses these conflicting references in considerable detail (see TABS, Ed.’s Intro., 5 0 -6 2 ). His observations seem to suggest that it may be more fruitful to concentrate on the text that has survived, rather than to speculate about the possible form and content o f a work that is neither extant, nor referred to in a consistently recognisable fashion. This total is reached by counting each o f the two instances where two brothers are treated together as constituting a single biography. This is because structurally they each constitute only one biography, with a single heading etc, despite possessing two subjects. Johannes Pedersen, and others follow ing him, reach a total o f 105 biographies by counting each o f these cases o f ‘com bined biographies’ as constituting two separate biographies. See further the next two notes below. The second generation section includes one ‘com bined biography’, under a single heading, w hich is devoted to the brothers Muhammad and Ahmad abna’ Abl ’1-Ward (TABS, 24 6 -2 5 0 ). Similarly to that observed in the second generation section (see above note), the fifth generation section includes, under one heading, one ‘combined biography’, w hich is devoted to the brothers Muhammad and Jacfar abna’ Ahmad b. Muhammad al-Muqrl. (TABS, 54 1 -5 4 6 ). TABS, 5 .1 1 -1 2 . e.g. TABS, 5 5 4 .1 -4 . See Rosenthal, M uslim historiography, 111. TABS, 4 .3 -5 . The problematic nature o f the term ‘saint’ in an Islamic context has already been noted elsewhere: see G.M. Smith & C. Ernst (Eds), Sainthood in Islam, especially xi—xxviii. See Chaumont, ‘al-salaf w a-’l-khalaf’ in E l- . There is no extant copy o f a work that fits the description o f the Kitab alZ u h d . Rather, there are only a couple o f ambiguous references, in other works o f the eleventh century, to historiographical material attributed to SulamI w hich may have been contained in such a work. See H, II, 25.8—13; KM, 9 8 .2 0 -9 9 .2 . See also TABS, Editor’s Intro., 5 0 -6 2 . The earliest died in 3 4 1 /9 5 2 (Abu SacTd b. al-AcrabI) and the latest in 3 7 8 / 988 (Abu ’1-Qasim al-Muqri’). For the purposes o f illuminating Sulaml’s m ethodology only those dates given by him are taken into account here (and in the follow ing references to time-spans covered by generation sections). The remaining members for w hom dates can be obtained from other sources confirm the interpretation being proposed. The earliest died in 3 2 8 /9 4 0 (al-Murtacish) and the latest after 340/951 (Abu ’1-Khayr al-Aqtac). The earliest died in 291 /9 0 3 (Ibrahim al-Khawass) and the latest in 330/941 (Abu ’1-Hasan al-Dinawari). The earliest died in 2 8 3 /8 9 6 (Sahl b. 'Abdullah al-Tustari) and the latest in 319/931 (Muhammad b. al-Fadl al-Balkhl). There is also a considerable overlap between the second and third generations (29 1 /9 0 3 —3 19/931), suggesting that there was an abundance o f Sufis belonging to this period w ho proved difficult to classify according to the generation system. It is perhaps the hierarchical facility o f Sulami’s generation system which has influenced most the shaping o f these two generation

185

N otes

29 30 31

32

33 34

35

36 37 38 39 40 41 42 43 44 45 46 47 48 49

50 51 52

sections (i.e. crucially maintaining Junayd and Jurayri as the heads o f the respective sections). See below pp. 13—15. See Dabashi, ‘Historical conditions’, 1 7 1 -2 . See TABS, 3 8 6 -4 0 0 . See pp. 1 5 -1 6 below concerning the order o f biographies in the first generation section. It is an exceptional case because, unlike the other generation sections, all o f its members could not possibly have been contemporaries o f each other. The verb sahiba, w hich in other contexts can mean simply ‘to accompany’, is used consistently in this work to refer to a relationship, through w hich one acquires authority (as in ‘the sahdba). See TABS, 253.11. N o works ascribed to Abu Thawr have survived, but he is often cited, in literature o f the eleventh century, as an authoritative jurist w ho was associated with Shafi'i (see Schacht, ‘Abu Thawr’ in E l“; Melchert, Formation , 71-7 6 ). Later he seems to have been confused with Sufyan al-Thawri (d. 161/778) (e.g. see N , 79.13). See TABS, Editor’s Intro., 21. Pedersen has identified, out o f the 105 individuals included (in the 103 biographies - see note 15 above), 59 from Khurasan, 31 from Iraq, 8 from Greater Syria, 5 from Egypt, and one each from M ecca and Qayrawan. See p p .61-4 below The biography o f Shibll is more than one and a half times as long as any other biography in the work (See TABS, 340—355). TABS, 443.6. H, II, 2 5 .1 4 -1 8 . See the citation o f Norm an Calder on p. 17 below O ne manuscript tradition presents the biography o f Dhu ’1-Nun before that o f Ibn Adham (see TABS, Editor’s Intro., p.5) See Bonner, Aristocratic violence, 126. See Melchert, Formation, 3 -4 . TABS, 13 .6 -7 . See Cooperson, ‘Ibn Hanbal and Bishr al-Hafi’, 71—101. Ibn Qutayba, a l-M a carif, 5 7 1 .1 0 -1 3 . Ibn Qutayba, a l-M a carif, 543.6. Bonner, Aristocratic violence, 125-134. Christopher M elchert has recently made a similar observation, albeit w ith an alternative interpretation; he has identified D hu ’1-Nun as the figure representing the historical turning-point from asceticism to mysticism. This attests to an accurate reading o f the Tabaqat, but his interpretation o f Sulami’s representation o f the past as an accurate account o f actual historical circumstances, especially with regard to the earliest Sufis, seems to me optimistic. See Melchert, ‘Transition’, 5 1 -7 0 . See also note 132 below. For an account o f the portraits o f the other early members o f Sulami’s first generation section, see M elchert, ‘Transition’, 52—63. See especially pp. 1 3 0 -1 3 2 below. An English translation o f this biography is already available (albeit derived from M assignon’s original French translation), in Massignon, Origins (Trans. B. Clark), 2 1 0 -2 1 3 .

186

N otes 53 Calder, ‘History and nostalgia’, 60. The fulfilment o f the process in this specific context can be witnessed in later works. See especially pp. 106—7 below, regarding Qushayrl’s inclusion o f Sulam! in the tabaqat section o f his Risala. 54 There are forty biographies in this work that do not correspond to this pattern, and this is by virtue o f the fact that they lack a hadith transmission. Presumably the subjects o f these biographies were not known for transmitting hadiths (see pp.20—21 below concerning the hadith transmission o f a biography). 55 See note 34 above 56 See Makdisi, Colleges, pp. 148—52, regarding authorisation to teach law and give fatwas. 57 ‘The Caliph’s N e w Year’s D ay’ is identified by Abu Rayhan Muhammad b. Ahmad al-BTrunl (d. 4 4 0 / 1048) as the eleventh day o f the m onth o f Hazlran (Greek calendar), in w hich ‘people in Baghdad splash in the water, strew about dust, and play other games’ (see al-BTrunl, Athar, 266.10; citation in English from Sachau’s translation, A l-B Trum ’s chronology o f ancient nations, 258). I owe this reference to Julie Scott Meisami. 58 The introduction to Abu Yazld is actually one o f the few that lack this type o f information. See note 32 above concerning the use o f the verb sahiha. 59 See TABS, 41.3. 60 The isnad o f the first segment in the body o f the biography o f Muhasibi names Junayd as the first transmitter (TABS, 50.2—7). 61 The Tabaqat does not contain a biography o f Qassab, although he is mentioned as an authority with w hom two other Sufis associated. In both these cases he is mentioned alongside Saqatl, suggesting a special relationship between them in tradition. See TABS, 151.10 (Abu ’1-Husayn al-Nuri); 186.4-5 (Samnun b. Hamza). There is a biography o f Qassab in al-Baghdadi, Ta’nkh, III, 62. 62 The introduction to Junayd also associates him with the jurist Abu Thawr, the significance o f which has already been discussed on p. 14 above. 63 See especially Chs 3 and 4 below. 64 However, see TABS, Editor’s Intro., 40, where he brings to attention the provision o f a hadith transmission at the end o f one o f the biographies, in addition to one located between its introduction and body 65 Eight biographies each contain two hadiths (e.g. the biography o f ‘Abdullah b. Khubayq al-Antaki, 131.7-132.6). See further TABS, Editor’s Intro., 40. 66 A translation o f the hadith transmission in the biography o f BastamI is offered in Sells (Ed.) Early Islamic mysticism, 235. 67 Q , 15.5. 68 Although asnada often has connotations o f tracing the ascription o f hadiths, in this context it is used to mean simply ‘to transmit’. See further TABS, Editor’s Intro. 4 0 -4 1 . 69 For information about the transmitters, see al-Sulaml, Tabaqat (Ed. N. Shariba), 156 (notes). 70 e.g. See R , 3 4 3 -5 3 . 71 e.g. The hadiths offered in the biography of'Abdullah b. Khubayq al-Antaki, w hich are about the creation o f a human in the womb, and the Prophet’s m eth o d o f p erform ing ablutions after sexual intercourse (TABS, 131.7-132.6). 72 See Rosenthal, M uslim historiography, 9 3 -5 ; Robinson, ‘A l-M ucafa b. cImran’. 73 Corroboration for this assertion is often difficult to find outside o f the Sufi

187

N otes

74

75 76

77 78 79 80

81 82 83

84 85

86 87 88 89 90

91 92 93

tradition. In fact, even the reliability o f SulamI him self as a hadith transmitter has been questioned (see al-Baghdadl, Ta’rikh, II, 248). There are significant discrepancies between the texts o f Pedersen’s and Shariba’s editions, especially with regard to the introductions o f segments, and they are highlighted in the comparison o f the different editors’ presentations o f this passage. See also Pedersen, ‘Quelques remarques’, 193-4. e.g. TABS, 62.7. See Pedersen’s quantitative analysis o f introductory formulae in the Tabaqat as a whole, as part o f the extensive introduction to his edition o f the text: TABS, Editor’s Intro., 3 3 -3 8 . See TABS, 132.6, 371.4. TABS, 150.3-8. See TABS, Editor’s Intro., 42. See TABS, Editor’s Intro., 4 3 -4 4 , 4 5 -4 7 . I have not had the opportunity to examine the manuscript concerned for myself. In spite o f advocating the method o f attributing every segment w ithout its ow n isnad back to the preceding isnad, Pedersen him self admits that one cannot verify that this interpretation applies in each instance (TABS, Editor’s Intro., 44). See TABS, 147.6-148.5; 150.3-8. e.g. TABS, 142.9-10; 1 43.5-9. e.g. See TABS, 3 4 2 .4 -9 (two segments juxtaposed), but also 3 4 5 .4 -6 (two further segments with the same final tradent). In each o f these pairs, the first segments actually possess identical isnads. See p p .29-32 below concerning the possible sources o f segments. The compilation o f an author’s books by his students is a com mon phenom enon in medieval Islamic scholarship, and is even the way in which the Sufi tradition itself accounts for the compilation o f the (Persian) Tabaqat alSufiyya, ascribed to "Abdullah Ansari (see C h.3 below). For a recent study o f this phenom enon in early fiq h , see Calder, Studies, Chs 1—5 and Ch.7. See p.49 below, concerning ‘lengthy prayers’. TABS, 6 4 .6 -7 . e.g. TABS, 145.9 -1 1 , 1 4 6 .5 -7 in the biography o f Junayd. See Ja w a m ic adab al-Sufiyya (Ed. E. Kohlberg) and Kitab A dab al-suhba (Ed. M.J. Kister). Consider the follow ing example: ‘I sat in my prayer-niche at night and stretched out my leg, and a voice (hdtif) said to me, H e w ho keeps the company o f the kings should do so with fine manners’ (TABS, 61.12—62.2). See R , 1 7 7-183 concerning the use o f the term mujahada in Sufi literature o f the eleventh century. The utterance, ikh tila f a l-culam a’ rahma had already been projected back as an utterance o f the Prophet. See Goldziher, ‘ikhtilaf’ in E l1. The isolated use o f this expression leaves it somewhat open to interpretation. In this specific context, where the focus is scholastic knowledge, it could refer to the theological debate about the nature o f divine unity, possibly as a synonym for ta'til (stripping God o f his attributes o f unity). In this case, Abu Yazld would be portrayed as drawing the line for ‘acceptable ik h tila f at the doctrine attributed to the M u’tazilites, and thus as som eone w ith ‘correct’ belief according to Ash’arite scholars. This would also not be without relevance for som eone remembered as a Hanafite (see p.86 below).

188

N otes

94 95 96 97 98 99 100 101 102

103 104 105

106 107 108 109 110 111 112 113 114

115 116

117

118 119 120

H ow ever, tajrid al-taw htd could also be referring to a mystical understanding o f divine unity, and therefore be interpreted as excluding either its questionability or its acceptability, depending on whether the ‘orthodoxizing’ function o f the remainder o f the utterance is being extended or uncompromisingly qualified by this final clause. e.g. See TABS, 6 2 .2 -7 ; 64 .1 2 -1 4 ; 6 6 .1 1 -6 7 .3 . See Kinberg, ‘Literal dreams’, 2 7 9 -3 0 0 (I owe this reference to Andrew Rippin); TABS, Editor’s Intro., 38. See L, 6 8 -6 9 . The ‘witnessings’ are described as ‘perceptions o f the heart (.qalb) from the unseen (ghayb)’. See Baldick, M ystical Islam, 39. See TABS, 147.4—6, as an example o f a segment about Junayd, the first tradent o f which is Khuldl. See further KM, 3 0 3 -3 0 7 . See further p.50 below. For the texts and translations into English o f further examples o f epistolary material attributed to Junayd, see Abdel-Kader, Life, personality and writings. Regarding the implications o f attributing letters to authoritative individuals in historiographical works, see N oth/C onrad, Early Arabic historiographical tradition, 76—87. See further p p .29-32 below TABS, 350.11. See also 342.11, 348.7. Each o f these named individuals is considered worthy o f a biography in Sulamfs Tabaqat: see TABS, 3 5 6 -3 6 1 , 3 8 6 -3 9 1 , 3 9 2 -3 9 5 . M oreover they are all included in the generation section preceding that o f ZajrjajT. This anecdote therefore serves to emphasise the latter’s authority, by claiming that he was given deference by illustrious and senior ‘colleagues’. See Makdisi, Colleges, 9 1 -3 . Ibid, 10-11; 17 -1 9 . Ibid, 12-13; 2 1 -3 . TABS, 3 4 8 .6 -8 . TABS, 2 9 4 .6 -7 . TABS, 340.8. See further TABS, Editor’s Intro., 33—35. e.g. TABS, 1 0 5 .4 -5 , 183.9, 500.10, 503.7. Sulami’s father is neither named, nor referred to outside o f the context o f these isnáds. TABS, 1 6 0 .2 -3 . This phrase is found in the biography o f Abü ‘Uthmán Sacid al-Hlri, as an explanation o f the nature o f a transmission by his grandson, Sacid b. 'Abdullah b. Sacid b. Ismail. TABS, 8 1 .5 -6 . W ith regards to the use o f notebooks in the transition between oral and literary culture, see Calder, Studies, 171 -1 8 1 . C oncerning the implications o f the term kitdb see pp.6 1 - 4 below. TABS, 4 8 6 .5 -6 . W ith regards to the implication o f ‘licence’ to transmit (ijaza), see Vajda, ‘idjáza’ in El , Makdisi, Colleges, 143—52 and Calder, Studies, 171—4. TABS, 4 9 0 .3 -4 . TABS, 170, cited in TABS, Editor’s Intro., 34. e.g. See R , 3 6 0 -3 6 7 .

189

N otes 121 See p. 115 below. 122 See p.31 above concerning this kind o f transmission. 123 TABS, 6 4 .8 -1 1 . Two variants o f this segment are also found in the corresponding biography in the H ilyat al-awliyd>: X , 37.14—15; 39.6—8. 124 TABS, 6 2 .2 -3 . See also 1 5 0 .3 -5 for a response, attributed to Junayd, to the same question. 125 TABS, 6 4 .1 -3 . 126 TABS, 6 6 .8 -1 0 . 127 See pp. 143—6 below. 128 See Kinberg, ‘Literal dreams’, 283—292. 129 The mystical notion o f m a crifa is com m on in ninth century works, but it is traditionally attributed further back to D hu ’1-Nun al-Misri (see Baldick, M ystical Islam, 35; Massignon, Essay, 143). 130 M elchert, ‘Transition’, 5 2 -5 4 . 131 TABS, Editor’s Intro., 3 9 -4 2 . 132 See Bonner, Aristocratic violence, 125—134 (I owe this reference to G.R. Hawting); Cooperson, ‘Ibn Hanbal and Bishr al-Hafi’, 71—101. Moreover, M elchert’s observations, with regards to the biographies o f the earliest individuals to be included, does not represent sufficient grounds for the assumption that the Tabaqat as a w hole ‘seldom ’ suffers from backprojection, thereby implying that it can be used as a reliable source o f facts for historical reconstruction (Melchert, ‘Transition’, 53). 133 See p.26, 28 above. See also p.203, n.69 below regarding an explicitly polemical segment included in Sulami’s biography o f Junayd which is repeated in QushayrT’s Risdla. 134 See p.28 above. 135 TABS, 64.1. 136 TABS, 66.5. 137 e.g. see pp. 166—7 below regarding Jami’s attempt to link back his Nafahdt to Sulami’s Tabaqat.

Chapter Two: The H ily a t a l-a w liy a ’ 1 al-Subkl, Tabaqat, IV, 18—22. 2 al-Nabhani, Karamat, I, 293. 3 e.g. See TABS, 229.1—8 (the introduction o f the biography o f cAlT b. Sahl alIsfahani). 4 See H, I, 4 .1 4 -1 5 . 5 See H, X , 4 0 2 -8 . 6 See the bibliography for details o f the printed editions o f these two works (Geschichte Isbahans and D a la ’il al-nubuwwa). For the minor works ascribed to Abu N u caym, see Khoury, ‘Importance’, 78. 7 H, I, 41.1. 8 See H, I, 28—87; FI, X , 402—408. The final person listed in the Hilya is Abu ’1-Hasan cAlI b. Mashadha (d.414/1023), a pious Shafi’ite scholar from Isfahan. Although he is not explicitly linked to the group o f successors o f Ibn Ma'dan whose biographies immediately precede his own, the position o f his biography implies that he was a product o f their tradition. His deliberate

190

N otes

9 10

11 12

13

14

15

inclusion at the end o f the work (wa-khatama ’l-tahqïq bi-tarïqat al-mutasawwifa b i-A b ï ’l-Hasan cA l ï b. Mâshâdha) may have been due to his importance in Isfahan at the time the work was completed there. Little information is offered about him in other biographical writings. In fact, he is known almost exclusively for his association with Abü N u caym by virtue o f the position o f his biography in the H ilya (See Ibn al-cImad (d.481/1089), Shadharàt , III, 201; al-Subkl, Tabaqât, IV, 181.4-10). e.g. See Frank, ‘Tasawwuf is..’, 74. e.g. The biography o f Fadïm Abü Hàshim, ‘before w hom Sufyân al-Thawrï sat [as a student] ’ (jalasa ilayh Sufyân al- Thawrï) , is found in the tenth volume, whilst that o f Sufyân al-Thawrï overlaps the sixth and seventh volumes (see H, VI, 356-V II, 144; H, X , 112). See also p p .71-2 below concerning the Abü Hâshim biographical tradition. The aforementioned Fadïm Abü Hâshim’s biography is juxtaposed to biographies o f fellow Sufis from Baghdad e.g. The biography o f Abü Yazïd is followed by those o f Ahmad b. alKhadrawayhi and Ibrâhïm al-Harawï. Ibn al-Khadrawayhi’s biography begins with a narrative involving Abü Yazïd, whilst Harawï is introduced as a companion o f Abü Yazïd (min aqrân A b ï Yazïd ) (See H, X , 4 2 -4 3 ). e.g. The biographies o f Abü Yacqüb al-Zayyàt and Abü Jacfar al-Küfï both contain the same anecdotal topos, concerning the handing over o f ‘dirhams’ to or from Junayd (See H, X , 223.24—224.20). This topos is also used in the biography o f Junayd itself, where he is described as handing over ‘dirhams’ to Sarï ’1-Saqatï (See H, X , 270.23-271.2). egs. Two biographies devoted to Abü Turàb al-Nakhshabï (H, X, 45-51 & 21922); two biographies devoted to Abü Ahmad al-Qalànisï (H, X, 306 & 341); two biographies devoted to Abü Yacqüb al-Zayyàt (H, X, 223 & 342-3) for an example o f overlapping material, see H, X, 223.24—224.6 cf. 342.18—25). R a if Khoury, whilst pointing out the value o f the H ilya as a work o f historiography, has also noticed the problematic nature o f its form: B ref on y sent la main de plusieurs transmetteurs qui ont copié le texte plus d'une fois, dans ces séances d'étude et de transmission, comme c'était le cas dans les cercles des lettrés islamiques à travers les siècles (Khoury, ‘L’importance’, 97). Despite these observations, Khoury continues to classify the H ilya as Abü N u caym’s work.

16

17

18 19

It remains worthwhile to avoid the attribution o f the same kind o f ‘authorship’ to works such as the Hilya, as, for example, its contemporary Sulamï’s Tabaqât, even if organic school texts are hardly uncom m on in medieval Islamic scholarship (e.g. see Calder, Studies, especially C h.7). The remaining eight are each members o f Sulamï’s first generation, and their biographies are found in Volumes VII—IX. (Ibrâhïm b. Adham, Shaqïq alBalkhï, Hàtim al-Asamm, Fudayl b. cIyàd, Bishr al-Hàfî, M a'rüf al-Karkhï, al-Dârànï and Ahmad al-Antâkï). See H , I, 3 -5 , where can be found the stated intention o f the Hilya, to distinguish ‘authentic’ Sufism from the ways o f false pretenders, with specific reference to the ‘permissivists’ (mubâhiyya) and the ‘incarnationists’ (huhlliyya). See al-Daylamï, Sïra, 98—99, where Hallâj’s visit to Isfahan is described. See also Meier, E in Wichtiger, 6 0 -1 0 6 . There are in total nine group introductions in the tenth volum e (see H, X , 41; 222; 233; 305; 311; 375; 387; 407; 408).

191

N otes 20 See H, X , 4 1 .2 4 -4 2 .2 . 21 Group introductions are seldom provided in the remaining nine volumes. For examples, see H, II, 25 (aid al-suffa ‘mentioned by SulamI and A'rabI’; see further pp.66—7 below), and 61 (tabicfm ); VI, 148 (nussdk, cubbad). 22 See H, X , 2 2 2 .2 2 -2 5 . O utside o f the H ilya, K h a z z d z is normally remembered by the alternative nisba Kharrdz. H e is m entioned here as one o f the gnostics from Iraq to w hom books are ascribed (m inhum man yunsab ilayh al-kutub al-musannafa ka-AbT S a cTd a l-K h a zzd z).

23 See H, X , 246-249.18.' 24 See H, X, 232.22. In the printed edition, it forms the first clause o f the introduction to the biography o f Muhammad b. al-Fadl. 25 See p.42; 6 3 - 4 below concerning the occurrence o f repeated biographies, usually with overlapping material, and the existence o f isnads in which Abu N u caym him self is mentioned as a tradent. 26 See p.21 above. 27 This format is shared by the vast majority o f the biographies in the Hilya, which makes it all the more conspicuous that the biography o f Junayd possesses an anomalous resumption o f its body after the hadith transmission. See pp.59—60 below. 28 O ut o f the 63 segments that make up the body o f the biography o f Abu Yazld, 32 segments are introduced by means o f an isnad, 8 segments by means o f a phrase that re-introduces a narrator (qala wa-), and 23 segments by means o f a simple conjunction (wa-/fa~). O ut o f the 107 segments that make up the body o f the biography o f Junayd, 78 segments are introduced by an isnad, 9 segments by qala wa-, and 20 segments by means o f w a -/ fa-. See also pp.21—4 above. 29 N or indeed in the manuscript tradition, including the one on which the tenth volum e o f the printed edition is based (Azhariyya). W hilst the isndds tend to be highlighted, the same often applies for the alternative types o f introduction. I am indebted to Shaykh Ahmad Khalifa o f the al-Azhar library for allowing me to view the Azhariyya manuscript briefly under his supervision. 30 The manuscript on which the printed edition is based reveals two misprints here. The text should read jalastum (not ujlistum) and fa -la m (not fa-m a ). 31 The omission o f ghaybatT after kanat is a misprint according to the Azhariyya manuscript itself. This can also be deduced from Segment 6, w hich is a variant. 32 The printed edition faithfully repeats the reading o f the Azhariyya manuscript here. However, see note 35 below. 33 The manuscript on which the printed edition is based includes the clause fa qala A b u Yazld, which is om itted in the printed text, immediately after what would then be read as a statement, inna ’l-m a ’ al-qd’im qad kuriha ’l-w udtl* minh. Thus the latter clause would represent the questioner’s allegory, whilst Abu YazTd’s response would begin with lam yaraw bi-m a1 al-bahr ba’san. 34 I should like to thank Maroun Aouad, Michael Cook, O leg Grabar and Andras Hamori for a number o f helpful suggestions regarding the translation o f this problematic passage, during the Arabic Texts Seminar at the Institute for Advanced Study in May 1999. 35 The Leiden manuscript gives instead lah ‘to him ’.

192

N otes 36 See also H, X , 35.18—19; 22—3, where two juxtaposed segments each begin with the same formula: fa - ja ’a rajul ila A b \ YazTd fa-qala. 37 See p.58 below. 38 e.g. See H, X , 35.24—36.1, which is also presented below p.62. 39 See p p .4 5 -7 above. 40 For examples o f variant utterances in the biography o f Junayd, see H, X, 2 6 3 .1 8 -2 0 cf. H, X , 2 6 7 .2 0 -2 2 . 41 See the classification o f segments in pp.24—6 above. 42 The five segments containing fragments o f verse (attributed to Junayd) range in length from one to three distiches. Although it is possible to interpret them mystically, they consist mostly o f love-poetry topoi, such as ‘the lover’s oppression’ and ‘the gossip o f the people about the lovers’ (e.g. see H, X , 2 6 9 .3 -7 ). 43 H, X , 2 8 2 .5-25; 2 8 3 .1-23; 284.2 2 -2 8 7 .1 7 . 44 H, X , 2 8 3 .7 -8 . 45 H, X , 2 8 3 .2 -3 ; 2 8 4 .2 3 -2 4 . 46 H, X , 2 8 2 .5 -6 . 47 H , X , 2 5 6 .4 -1 6 ; 2 5 7 .2 0 -2 5 9 .1 6 ; 2 5 9 .1 7 -2 6 0 .1 8 ; 2 6 0 .1 9 - 2 6 1 .2 1 ; 2 6 2 .8 -2 6 3 .1 0 ; 2 6 4 .4 -2 6 5 .1 ; 2 6 5 .2 -1 3 ; 2 6 5 .1 9 -2 6 6 .2 5 ; 2 7 1 .1 0 -2 7 3 .5 ; 276.4-277.5; 276.6-23; 278.16-22; 279.9-280.2; 280.11-25; 283.24-284.12. 48 See p.28 above, for a discussion o f this epistle. 49 M ost o f the epistolary segments are introduced simply as som ething ‘said’ by Junayd (sam i'tu . . . yaqul). Five o f them are introduced as written correspondence com posed by Junayd; four simply as something written ‘to one o f his [Sufi] brethren’ (kataba ’l-Junayd ila bacd ikhw dnih/kataba ila bacd ikhwanih kitdban yaqiil fill), whilst the remaining example is introduced as written correspondence and names the recipient as Abu Ishaq al-Maristam; this epistle can be divided into two sections, the first section o f which is also found in the biography o f Maristani itself, but as the final part o f an entirely different epistle (see H, X , 2 7 6 .4 -2 7 7 .5 cf. H, X , 33 3 .4 -1 2 ). The five examples which are presented as written correspondence by Junayd are also included in a manuscript w hich is known as ‘The R a sa ’il o f Junayd’, along with other material similar in form, style and use o f vocabulary (see Abdel-Kader, Life, personality and writings, Appendix, 1 -6 2 (Arabic text). The editor and translator o f this manuscript has classified these segments as ‘letters’, in accordance with the classification offered by their transmitters. 50 See H, X , 257.2 0 -2 6 1 .2 1 . 51 See H, X , 2 5 5 .1 5 -1 7 . See pp.5 2 -3 below, regarding the narratives which describe encounters with MuhasibT. 52 See p p .61-3. 53 The shabb m otif is com m only associated w ith wisdom and miraculous powers (e.g. see H, X , 2 7 5 .5 -1 7 ). 54 See p.63 below, where this narrative is presented. 55 In fact, both o f them are also found in the biography o f MuhasibT in the same volum e o f the H ilya (see H, X , 74.6—18). The variant o f Segment 1 is offered w ith a further isnad, in addition to transmission from Ahmad b. Muhammad, and the variant o f Segment 2 is presented with its ow n isnad, identical to that o f Segment 1.

193

N otes 56 e.g. See TABS, 340.9. 57 T he manuscript in Leiden gives ghayruk, w h ich w ould seem more appropriate. The similarity between the letters k a f and y d ’ in the manuscript may account for this discrepancy. An alternative explanation, although perhaps less likely, is that ghayruk may have been employed deliberately in order to avoid the attribution o f a ‘counter-boast’ to Abu Yazld. 58 See Fahndrich, ‘The Wdfaydt al-acy d n , 444. See also N oth/C onrad, Early Arabic historiography; 109—204. 59 The second A llah is missing in the printed edition, but it is required both to make grammatical sense and to parallel Junayd’s previous utterance. 60 See TABS, 2 5 3 .9 -2 5 4 .1 . 61 See L, 3 8 0 -3 9 5 . 62 See pp. 1 8 -1 9 above. 63 See H, X , 2 5 5 .7 -1 1 . 64 See p.50 above. 65 In fact, it presents the very same hadith as its counterpart in Sulaml’s Tabaqat (see H, X , 4 1 .6 -1 7 , cf. TABS, 6 0 .10-61.11). 66 See p.49 above. 67 e.g. H, X, 6 9 .12-13; 109.22-25; 281.18-20; 2 9 6 .1 1 -1 2 . 68 H, X , 281.2 2 -2 8 2 .4 . 69 See p p .46-7 above. 70 See p.54 above. 71 See H, X , 28 2 .5 -2 8 7 .1 7 . 72 See H , X , 2 8 1 .11-281.18. 73 Daube, Studies , 7 4 -1 0 1 . 74 H, X , 3 4 .1 8 -2 1 . 75 See below for possible interpretations o f kitdb. 76 H, X , 3 8 1 .1 0 -1 2 . See also TABS, 4 5 4 -5 . 77 e.g. H, X, 257.11. 78 e.g. H, X , 279.2. 79 e.g. H, X , 257.20. 80 O ne o f the segments introduced as part o f written correspondence offers a date akhbarana Ja Ja r b. M uham m ad b. N usayrfT -m d kataba ilayya sanat thaldtlnn wa-arbacm (H, X , 381.15). This date is presumably [3]43, when Abu N u caym would have been only seven years old. 81 See p.54 above. 82 H, X , 3 5 .2 4 -3 6 .1 . 83 M odern scholars have sometimes referred to a ‘lost b ook ’ by Khuldl, entitled H ikaydt al-aw liya\ but this title, which is derived from medieval sources, probably represents no more than a generic reference to the segments collected by Khuldl, and not to a literary work (see Abdel-Kader, Life , Personality and writings , 12). 84 H, X , 381.10. The introduction also remarks that he followed the best o f people, including Junayd sahiba ’l-akhyar al-funayd w a-’l-Thaum wa- R uwaym an. The first segment after the introduction is a hadith. 85 TABS, 4 5 4 .5 -7 . 86 See below, and H, X , 257.21. The latter example has been overlooked by the editor in his m ethod o f paragraphing clusters. 87 See Calder, Studies, 173.

194

N otes 88 See C h .3 on the Persian Tabaqat al-Sufiyya, a work which, according to tradition, was com piled by a student on the basis o f records o f ‘the author’s’ teaching sessions. 89 The above narrative also suggests that teaching sessions may have been held in a specific place away from the hom e o f the teacher. Muhasibi leaves the location o f his teaching session (al-m akan alladhf yajlis fih ) to return hom e (;thum ma yamdi ild m anzilih). See pp .29-32 above. 90 See H , VIII, 306.4; 336.25; 369.19 (in each case the isnad begins thus, haddathana ’l-shaykh al-hdfiz A b u N u 'a y m A h m a d b. ‘Abdullah rahimah A llah).

91 See H, IX, 279.20 (akhbarana l-sh a ykh A b u }l-Fadl A h m a d b. A h m a d b. alHasan b. al-H addad qird’atan calayh wa-ana asm ac qala akhbarana A b u N u 'a y m A h m a d b. A bdullah al-hafiz qird’atan calayh hadha ’1-hadTth bi-isnadih . . .). See pp.87—8 above concerning the qira’a m ode o f transmission. 92 Pedersen has directed attention to a number o f further isnads, but they

93 94

95 96 97 98 99 100 101

include an ‘Abu N u caym’ mentioned at stages, where an individual from the second century would be required. See TABS, Editor’s Intro., 67 and H , I, 72; 322; 328. eg. H, X , 3 4 .6 -1 0 cf. TABS, 6 4 .1 4 -6 5 .3 . However, consider H, X , 4 0 .7 -9 , w hich cannot be found in the Tabaqat al-Sufiyya. e.gs. H, X , 34.6; 40.1. There are also two isnads that name Sulami as the source by means o f the formula, akhbarana M uham m ad b. al-H usayn qala . . . W hilst the use o f akhbarana in this work tends to imply written transmission, this is not made explicit here by, for instance, reference to a kitab. See TABS, 6 2 .3 -6 cf. H, X , 3 7 .23-24; 39.13, and also H , X , 3 7 .2 1 -2 3 cf. TABS, 6 2 .2 -3 ; 6 4 .7 -8 . H, X , 2 3 2 .2 1 -2 4 cf. TABS, 2 0 6 .1 -8 . H, X , 2 3 3 .1 3 -1 9 cf. TABS, 206.1 0 -2 0 7 .7 . H , X , 2 3 2 . 2 5 - 2 3 3 . 1 2 cf. T A B S , 2 0 7 . 9 - 2 0 8 . 6 ; 2 0 8 .1 1 - 2 0 9 .3 ; 2 0 9 .1 3 -2 1 0 .1 . See Watt, ‘ahl al-suffa’ in E l2. See H, II, 2 5 .8 -2 2 . The Tabaqat al-N ussak is not extant. However, in contrast to the case o f Jacfar Khuldi (see pp.61—2 above), written works are consistently ascribed to A'rabI in the Sufi biographical tradition itself (egs. H, X , 375.15—16, lah altasam f al-mashhura; TABS, 443.6 (sannafa li- ’l-qawm kutuban kathTra).

Chapter Three: T he Persian Tabaqat al-Sufiyya 1 This is made explicit by what appears to be an interpolation in the introduction o f the Tabaqat al-Sufiyya, a work traditionally ascribed to Ansari. It is claimed there that his ancestor amongst the ansar was none other than the one w h o accommodated the Prophet him self after his migration from Mecca, called Abu Ayyub al-Ansari. It is also claimed here that Abu Ayyub was one o f ‘the people o f the canopy’ (ashab al-suffa), and that he died in Constantinople on a military expedition (see TABA, 1 .9 -2 .4 ). In this way, Ansari is given impeccable religious credentials w ith regards to his descent. It is therefore surprising (perhaps an oversight) that this interpolation also states that his first ancestor to move to Herat was called M a tt al-Ansari.

195

N otes 2 3 4 5 6 7 8

9

10

11

12 13

14 15

16

See Kramers, ‘Shaykh al-Islam’ in EI°. Ibn Abi Yaclä, Tabaqat, II, 2 4 7 -8 (no.684). Ibid., 248. al-Dhahabl, Ta’rTkh al-Isläm, Year 481, 53—63. See Ibid., 5 9 .1 -1 8 . e.g. see p.41 above, concerning the biography o f Abü N u caym al-Isfahänl in the Shafi’ite tradition). For more information on the biographical tradition o f Ansari, see the pioneering work o f D e Beaurecueil, Khwädja ‘A bdulläh (especially 15—149), and also that o f his translator into Dari, Ravan Farhadi, Abdullah Ansari, 3 -1 5 . See Ritter, ‘Philologika’, 89—100, and Ravan Farhadi, A bdullah Ansari, 19—23. R itter’s list (of fifteen works) omits D ham m al-Kaläm, a work which has been counted as Hanbalite and is included by both Ravan Farhadi, w ho has translated selected passages into English (pp.35-40), and Brockelmann (Geschichte, Supplementband I, 773—4). O ne should bear in mind Ravan Farhadi’s introductory comment: A bdullah Ansari o f Herat is considered a ‘great writer’, and yet he almost never wrote! (p. 19). See further below, especially Section VI. See Ivanow, ‘Tabaqat o f Ansari’, 1-34; 3 3 7 -3 8 2 . This two-part article contains extensive notes about the peculiarities o f the dialect. I would like to thank Leonard Lewisohn for lending me his copy. The work is referred to here as ‘the Persian Tabaqat a l-Süfiyyä’ to distinguish it from its earlier (Arabic) namesake. For the purposes o f this study the edition prepared by Abd al-Hayy Habibi is used, in preference to the more recent edition by Mawlayi. Whilst, as Ravan Farhadi has pointed out (pp.43-44), M awlayi’s edition provides many helpful indices, its main text does not follow the oldest manuscript as closely. Moreover, it incorporates the use o f ‘endchapter’ notes, which are hardly appropriate for a work where such divisions are often blurred (e.g. see pp.79—85 below). I have also had the opportunity to examine the oldest surviving manuscript (Nafiz Pasha 426 - photographed copy in Tehran University). I would like to thank Florian Sobieroj, Nasrollah Pourjavady and all the staff at the manuscript library o f Tehran University for their assistance. See Ch. 1, Section II above. e.g. al-Tabaqa al-thäniya min a ’immat al-Süfiyya (The second generation o f the leaders o f the Sufis), located immediately before the second generation section (TABA, 112.10). e.g. wa-min al-tabaqa a l-’ülä Bä Hafs Haddäd NTshäbürT (TABA, 90 .1 2 -1 3 ). There is no corresponding biography in the Persian Tabaqat for the following which are provided by Sulami: the biography o f "Abdullah b. Khubayq b. Säbiq al-Antäk! (TABS, 131-135), the com bined biographies o f Muhammad and Ahmad the sons o f Abl ’1-Ward (TABS, 24 6 -2 5 0 ). For the exceptions, see TAB A, 123 -1 2 6 (cf. TABS, 3 2 8 -3 3 1 ), the biography o f Abü Hamza ’l-Khurasani; TAB A, 126-131 (cf. TABS, 2 9 4 -7 ), the biography o f Abu Hamza ’l-Baghdadi. There are also four instances where a biography is classified as a member o f a particular generation section in Sulami’s work, whilst here their classification is left ambiguous (see TABA, 209 (cf. TABS, 318) concerning Mamshäd al-Dlnawarl; TABA, 270 (cf.

196

N otes

17 18 19 20 21 22 23

24 25

26 27 28 29 30 31 32 33

34 35

36 37 38

TABS, 271) concerning Tahir al-MaqdisI; TAB A, 341 (cf. TABS, 321) concerning Ibrahim al-Raqqi; TABA, 367 (cf. TABS, 386) concerning Abu Bakr al-Kattani. See pp. 1 0 -1 8 above. See TABA, 112-118. See TABA, 161-169. Nonetheless, this discrepancy suggests a prom otion o f Tustari from his mediocre position in Sulamfs Tabaqat. See p. 11 above. TABA, 7 .3 -4 . TABA, 9 .1 4 -1 0 .1 1 . The Hilya includes in its tenth volum e the biographies o f ‘Abu Hashim Fadim’ and ‘Abu Hashim al-Zahid’, w hich both contain variants o f the same anecdotes: see H, X , 112.9-11 cf. H, X , 2 2 5 .1 6 -1 9 cf. TABA, 9 .7 -8 ; H, X , 2 2 5 .1 9 -2 0 cf. TABA, 9 .9 -1 0 . An English translation o f the biography o f Abu Hashim is provided in Ravan Farhadi, Abdullah Ansari, 4 7 -9 . See TABA, 3 0 0 .5 -3 0 1 .9 (the biography o f Abu ’l- cAbbas al-Dlnawari) and 3 0 2 .5 -3 0 4 .1 0 (the biography o f Abu T-cAbbas al-Sayyari). See TABA, 279.8. This relationship is also mentioned in Sulamfs Tabaqat, so that its m ention here is unlikely to have been merely generated by the juxtaposition o f their biographies. TABA, 2 0 0 .3 -4 . See further pp.75—8 below, concerning the juxtaposition o f the ‘building-block’ segments o f biographies on account o f association between their contents. See TABA, 2 4 4 .1 1 -1 3 . See TABA, 35 2 .1 1 -3 5 3 .2 . e.g. see pp. 1 6 6 -7 below. See p. 91 below. See p p .75-8 below. See TABA, 37.7—70.1. See Section V below concerning the provision o f introductions as demarcators o f biographies. The first introduction to Ibn Shakhraf is not followed by the body o f a biography about him. Rather, it appears to have been provided because o f an anecdote relating an encounter between Ibn Shakhraf and Israfil, the subject o f the immediately preceding biography. The second introduction is followed by the body o f an actual biography. See pp.79—85 below. An interesting example is the clumsy repetition o f the same hadith in the introduction o f the work, the multi-layered structure o f w hich shows indications o f being the product o f ‘many hands’ (see TABA, 3.11; 5 .1 1 -1 2 ). Passages that are given in Arabic in the text (as opposed to the predominant Persian) are presented in bold type in the translation. TABA, 9 0 .6 -9 . In the oldest manuscript it reads, guftand rast mfguyad a z way bdz shaund. The problem is that two parties, in addition to the old beggar-woman, are referred to: the speaker (active or implied passively) and those to w hom the command bdz shawTd is directed (see TABA, 5 8 6 .1 -3 ). In the preceding part o f the segment the only speakers m entioned are the beggar-woman and the subject o f the previous guftand w ho ask(s) her what she has brought. This final

197

N otes

39 40 41 42 43 44 45 46 47 48 49 50 51 52 53

54 55 56 57 58 59

60 61 62

63 64

sentence is probably a clumsy gloss that functions to defend the beggarwom an’s ironic lament, even though that may have been originally the ironic punchline o f a joke. It would perhaps have been more appropriate as a gloss with the same function for the preceding variant, w hich contains Abu Yazld’s bold retort (to the same question asked o f the beggar-woman), ‘W hen a poor man reaches the court o f a king he is asked what he wants not what he has brought!’ O ne might speculate that the gloss was perhaps intended to influence (reassuringly) the reception o f this segment about the subject o f the biography. TABA, 2 8 5 .3 -5 . e.g. TABA, 9 0 .2 -4 . e.g. TABA, 90.6. e.g. TABA, 9 0 .9 -1 0 cf. TABS, 64.1; TABA, 88 .9 -1 1 cf. H, X , 4 0 .2 -6 . TABA, 9 0 .6 -9 . TABA, 9 4 .6 -1 3 . TABA, 9 0 .9 -1 0 cf. TABS, 63.13-64.1; 66.5. See TABA, 8 8 .9 -1 2 cf. H, X , 4 0 .2 -6 ; TABA, 93.1 6 -9 4 .1 cf. H, X , 3 8 .1 7 -1 8 . TABA, 9 5 .4 -5 cf. TABS, 6 6 .6 -8 . See pp.46—7; 49 above concerning the attribution o f a variant utterance to Dhu ’1-Nun in the biography o f Abu Yazld in the H ilyat al-awliya’. The heading M a s’ala f t ’l-tawlud is found at TABA, 169.6. The biography o f Ibn al-Karanbi begins at TABA, 184.9. See pp.50-51 above. See TABA, 1 6 7.6-7. There is considerable discrepancy between manuscript variants o f this introduction. It suggests that the copyists themselves found it problematic. The difficulties are probably due to one or more layers o f interpolation (see TABA, 170, notes 4 -5 ). TABA, 136.12-149.10. TABA, 1 3 6.3-4. TABA, 149.12. The biography o f Kharraz continues until p. 154.9. TABA, 2 0 .2 -2 5 .9 . See also the translation into French prepared by D e Beaurecueil in his Khwadja 'Abdullah , 258—264. See TABA, 1 6 8 .9 -1 3 (cf. H, X , 268.2 1 -2 3 ) for the other variants. In the Persian Tabaqdt Surushan was his great-grandfather, rather than his grandfather. This is perhaps part o f a process o f distancing him in time from his Magian ancestor. The name offered for Abu YazTd’s grandfather is A d a m , the name o f one o f his brothers according to SulamT. It may be more than a coincidence, since there is little doubt that the earlier version has been used as a source. See p. 18 above. See p.57 above. A conspicuous gloss is the com m ent informing that Abu Thawr was ShafiYs best student. It may be that later readers were less familiar with his identity, and that a need was felt to highlight his traditionalist credentials. See C h .l note 57. The exception is introduced as being found in a work called Ta’rikh al-Sufiyya ascribed to SulamT (see C h .l note 14 above). It is not found in his Tabaqdt —it

198

N otes

65 66 67 68 69 70 71 72 73 74 75 76 77 78 79

is about a Sufi belonging to the sixth generation section, namely Abu ’1Hasan Slrwani (see TAB A, 482.11). TABA, 4 4 7 .1 -4 cf. TABS, 5 1 6 .4 -7 ; TABA, 4 7 7 .3 - 6 cf. TABS, 54 1 .1 1 -5 4 2 .4 . W ith regard to memorization and notebooks, see Makdisi, Colleges, 9 9 -1 0 5 . See the manuscript variants: TABA, 321, notes 2 and 3. This is probably a gloss with the aim o f boosting the authority on which the segment is given. See pp.31—2; 63—4 above concerning qird’a and ijdza; Makdisi, Colleges, 104—5. See Calder, Studies, 176. See ibid., 1 7 1-181. See pp.64—5 above. See note 33 above for another example o f an (apparently) repeated biography. The parts in bold are originally in Arabic. See pp.8 2 -3 above This biography is found amongst the group o f ten juxtaposed biographies about individuals with the kunya Abu ’1-‘Abbas (TABA, 311.4—312.7). See also TABA, 163, notes 11 & 12 for the editor’s discussion o f the variants in extant manuscripts which suggest that kard should be understood as kardam. e.g. an utterance attributed to Abu YazTd, TABA, 561.14—562.3 . See TABA, 543, note 1.

Chapter Four: Qushayri’s R isala 1 al-Baghdadl, Ta’rlkh, XI, 8 3 .1 3 -1 4 . 2 SulamI is introduced in the Ta'rlkh Baghdad in similar terms, but it also indicates that he was a Sufi (see p.9 above). 3 KM, 2 0 9 .1 0 -2 1 0 .6 . In fact, many (Sufi) utterances attributed to Qushayri are also included in other parts o f this work (see KM, 28; 141; 189; 205; 287; 401; 439). 4 al-FarisT, al-Siyaq, BII, ff.49a—51a. This purports to be a continuation o f the ‘lost’ Ta’rikh Naysabur ascribed to “Abdullah b. al-Bayyic al-Hakim alNaysaburi (d.405/1014). 5 Ustuww a is thought to have been located about 150km N orth o f Nishapur (near Quchan in N orth Eastern Iran), before the 13th century M ongol invasions. See Gramlich, Das Sendschreiben, Einleitung, 11. 6 Regarding the connotations o f transmission by samdc, see Calder, Studies, 171—6. 7 According to cAbd al-Rahman b. cAll b. al-JawzT (d.597/1192), Qushayri died in the m onth o f Rajab and was buried next to his teacher Abu 'All ’l-Daqqaq (Ibn al-Jawzi, a l-M u n ta za m , VIII, 280). Ibn Khalllkan (d.681/1282) reports that he died on Sunday 16th R a b f al-Akhir and was buried beneath Daqqaq (Ibn Khalllkan, Wafayat, III, 205—7). Subkl relates a tradition that Qushayri died in the early m orning o f Sunday 16th R a b f al-Akhir after an illness, during w hich he continued to stand up for prayer; he was then buried next to his teacher Daqqaq (al-Subki, Tabaqat, V, 159.16—18). These versions serve to emphasise the relationship between Qushayri and Daqqaq, as well as the piety and humility o f the former (see p.55, 86 above). 8 See Brockelmann, Geschichte, I, 556—7.

199

N otes 9 The Risala is available in several manuscripts and has been translated into Persian, (O ttom an and M odern) Turkish and Urdu. A number o f commentaries to the Risala have also been produced from the thirteenth until the nineteenth century (see Von Schlegell, Principles, xii—xvii (preface by Hamid Algar). It is used today as a standard textbook in al-Azhar. 10 There are two main editions o f al-Risala al-Q ushayriyya. This study is based on the single volume Tehran reprint o f the edition prepare d by cAbd al-Hallm Mahmud and Mahmud al-Sharif (originally as a tw o-volum e work — hence discrepancies in reference details, though not in the actual text itself). Manuscript variants provided in Gramlich’s indispensable German edition have also been taken into consideration. For further details o f these editions, see the Bibliography. 11 Gerhard Bowering, w ho also recognises the dual-generic nature o f the R isala, refers to it as a ‘manual’, in order to distinguish it from works such as Sarràj’s Kitdb a l-L u m a c, which he classifies as a ‘treatise’ (see Bowering, M ystical vision o f existence, 19—21). Whilst his appreciation o f the significance o f this distinction is laudable, his choice o f terms is unhelpful. In view o f the fact that the works which Bowering labels ‘treatises’ are conventionally classified as ‘manuals’, I have decided to conform to the latter convention rather than adopt Bòw ering’s terminology; I therefore distinguish the Risala (as well as Hujwlri’s K a sh f al-mahjilb) as a ‘dual-generic’ work, thereby highlighting the actual reason for its distinction. 12 See below concerning the final chapter o f the work, which can be regarded as an appendix. 13 In fact, Gramlich classifies this section as a continuation o f the introduction o f the Risala. 14 R , 19—33. See also R , 3 1 .1 8 -1 9 , where Qushayri concludes that the segments (hikdyat) which he has presented show that ‘the doctrines ( caqd’id) o f the Sufis are in accordance with the sayings o f ‘the followers o f the truth’ (evidently in this context this title is being applied to the Ash’arites)’. 15 R , 121.12-13. 16 R , 121.1-11. 17 Thirteen chapters in this section consider Sufi terms in pairs (e.g. R , 126—9, qabd / bast); ten chapters consider them individually (e.g. R , 121—3, al-waqt). Four o f them consider terms in groups o f three (e.g. R. 162, 7/m al-yaqm / cayn al-yaqm / haqq al-yaqm ). 18 Qushayri uses the term bdb here to mean ‘chapter’, although it is used for ‘section’ in previous contexts ( tabaqdt and terminology sections). It is presumably out o f a desire to translate this term consistently that Richard Gramlich has divided his German edition o f the Risala into a total o f 54 ‘Kapitels’, 52 o f which are accounted for by the systematic section and the appendix (according to my classification), whilst the tabaqdt and terminology sections make up only one Kapitel each. I have prefered an alternative classification, despite the fact that it necessitates alternating translations o f the term bdb, in order to give a more representative overview ot the work (the systematic section does not account for almost the entire Risàia, as the 52:54 ratio would imply). 19 The different types o f material are usually presented in this order when included. Prophetic hadith are occasionally found later in a chapter as well (e.g. R , 223.22—224.3). Citations troni the Q ur’an can also be found later in

200

N otes

20

21 22 23 24 25

26

27 28

29 30 31

a chapter, w h en they are accom panied by com m entary (e.g. R , 212 .1 9 -2 1 3 .2 ). The most frequently quoted Sufi authorities in this section are Abu cAlI T-Daqqaq and Abu T-Qasim al-Junayd. See R , 5 3 9 -5 5 4 . Margaret Malamud has also referred to this final chapter as being distinct from the remainder o f the work (see Malamud, ‘Sufi organisations’, 429). See also Hartmann, Darstellung , 175—207. R , 1 5 .8 -9 . R , 554.7. See C h .l, Section II above. R , 3 5 -7 0 cf. TABS, 7 -1 4 1 . e.g. R , 1 0 7 -1 1 7 (15 biographies) cf. TABS, 4 4 3 -5 5 2 . The biography o f Sahl al-Tustarl (R, 57—8) stands out as the only major discrepancy. It is the eleventh biography in the biographical section o f the Risala, placing it amongst those corresponding to the first generation o f Sulamfs Tabaqdt, whilst it is actually found in the second generation o f that work (TABS, 199—205). This makes it the only biography in Qushayri’s tabaqdt section to be positioned in a portion o f the series o f biographies w hich does not correspond to the generation in w hich it had been classified by Sulaml. This biography is made up largely o f a narrative describing his uncle’s practical instruction to him about the performance o f dhikr (remembrance o f God). The final segment o f the immediately preceding biography, that o f Abu YazTd, is a narrative which also refers to the performance o f dhikr (R, 56.19—22). The thematic link between these narratives is perhaps a reason for the exact position o f Tustari’s biography in the Risala; they may have becom e juxtaposed in the processes o f transmission and compilation, and eventually influenced the exact position o f their respective biographies. Nonetheless, it is surely no coincidence that Tustari’s biography should once again be relocated, just as in the Persian Tabaqdt al-Sufiyya, the organisational framework o f w hich is also influenced by Sulami’s arrangement to a significant degree. Dawud al-Ta’I is named as the Sufi authority with w h om M a'ruf al-Karkhl associated (TABS, 74.1 2 -7 5 .1 ). Al-Zaqqaq al-Kablr is mentioned three times: he is one o f the Sufi authorities with w hom Abu Bakr al-Duqqi is said to have associated, and similarly one with w hom cAlT b. Bundar al-Sayrafi associated; he is also m entioned in an isnad as the transmitter o f an anecdote about Abu SacId al-Kharraz (TABS, 469; 534.2; 2 2 5 .4 -1 0 ). Abu cUbayd alBusri is m entioned four times: Abu ‘Abdullah al-Jalla’, Shah al-Kirmani and Abu SacTd al-Kharraz are each said to have associated with him; he is also m entioned positively in an anecdote about Abu Turab al-Nakhshabl (TABS, 166.4-6; 183.4-5; 2 2 3 .3 -4 ; 136.7-11). R , 79.11,15; 82.3, 7. In fact, the biographies o f Zaqqaq and Busri contain only two isnads in total, both o f which name Sulaml as the immediate source. R , 82.17. Similarly, the biography o f M a'ruf al-Karkhl w hich is positioned tenth in Sulami’s first generation is the fourth biography o f Qushayri’s series, immediately before that o f Sari ’1-SaqatI w ho is said to have been his student (R, 41.5; 43.10). See TABS, 2 4 2-252; 332-339; 427-442; 5 3 3 -5 5 3 . e.g. R , 217.9—14. See also p. 184, n.12 above (refering to R , 348.3—20). See TABS, 4 - 5 and Ch. 1, Section II above.

201

N otes 32 33 34 35 36

37 38 39 40 41 42 43 44 45 46

47 48

49 50 51

52 53

54 55 56 57

TABS, 5 .7 -1 0 . See Grämlich, Das Sendschreiben , 105, regarding manuscript variants. TABS, 6 0 .1 -1 0 . TABS, 141.2-142.2. QushayrT omits the names o f sources, an alternative precise dating o f Junayd’s death, the name o f his grandfather, and his ow n professional title, a l-K h a zz ä z (the silk trader) R , 55 .6 -9 ; 5 5.10-13; 55.15-20; 5 6 .1 -9 ; 5 6 .1 5 -1 8 . R , 5 5 .2 1 -4 . R , 56.19—22. The first two transmitters mentioned are cIsä ’1-BastämI, Abü Yazid’s brother, and his son Müsä b. cIsa ('Ammi ’l-Bastami). R , 56.10-12; 5 6 .1 3 -1 4 . See TABS, 6 6 .8-10; 6 3 .2 -5 . See TABS, 6 2 .9 -1 0 for the isnäd o f Sulami’s version o f the second segment. See R , 5 5 .6 -1 3 . This segment is presented above, pp .25-6. R , 55.14. The biography ofjun ayd corroborates this interpretation, since it similarly begins with (six) segments that are also found in SulamT’s work, whilst no further segments from the same source are included at a later point (see below). R , 5 5 .1 5 -2 4 cf L, 103.14-104.6. This is indicated by the formula wa-bi-hädhä ’l-isnäd before the second o f the pair o f segments. In Sarräj’s L u m a ' it is indicated by the repetition o f the isnäd. The isnäd in the Risäla differs only by the addition o f Sarräj and the immediate source, SijistanI, at the end o f the chain. See Arberry, Sufism , 74; Hartmann, Al-Kuschairis Darstellung, 2—4. R , 5 6 .1 3 -1 8 cf. H, X , 4 0 .1 5 -1 9 . W hile the two questions posed in the variant o f the Risäla are not logically connected, they both share the construct mä ashadd/ahwan mä . . . Similarly, the segment immediately preceding the variant in the Hilya contains the follow ing pithy utterance about the carnal soul, w hich includes the use o f ahwan and a s‘ab (a synonymn o f asliadd): 1 treated everything but I did not treat anything more difficult (ascab) than the disease o f my carnal soul (nafs), although there is nothing more worthless (ahwän) to me than that’ (H, X, 36.18—20). It is perhaps not too far-fetched to suggest that this may be more than a coincidence, and that Qushayrl’s variant was derived from the combination o f these two juxtaposed segments in the Hilya, which would then indicate that the text o f the Hilya was the (direct) source for the Risäla. R , 5 6 .1 9 -2 2 . The variant in the H ilya (H, X , 3 5 .3 -5 ) is presented above in translation, pp.44—6. R , 5 5 .1 -9 . The ‘voice’ tells Abü Yazld at first, ‘You are not capable o f being with us’ (yä A b ä YazTd lä taqwä m a cana), but then reassures him with, ‘You’ve found [it]!’ (wajadta). R , 7 1 .4 -6 ; 71 .7 -1 0 ; 7 1 .11-20; 7 2 .1 -4 ; 7 2 .7 -8 ; 7 2 .1 0 -1 6 . R , 72.17-19; 73 .2 -4 ; 7 3 .5 -7 . R , 71.21; 7 1 .22-3; 7 2 .5 -6 ; 72.9. The texts o f the first and sixth segments are identical to those o f the corresponding segments in the earlier work, whilst their isnäds differ only due

202

N otes

58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76

77 78 79 80

81

to the addition o f Sulami (Muhammad b. al-Husayn) as the immediate source (R, 7 2 .1 -4 , 7 2 .4 -6 cf. T A B S , 1 4 3.9-11, 148.5-7). The second and third segments similarly are given on the authority o f isndds w hich differ from those o f their earlier variants only by the inclusion o f Sulami, as the immediate source, whilst the texts o f both o f these segments are abbreviated versions o f their variants in Sulami’s work (R, 71.7—10, 7 1 .1 1 -2 0 cf. TABS, 1 4 3 .14-144.2, 145.5-6; see also Gramlich, Das Sendschreiben, 66 for significant manuscript variants o f the isnad). The fourth and fifth segments, are also abbreviated variants o f two juxtaposed segments in Sulaml’s work. However, they are not provided here with any isndds, whilst in the Tabaqat the variant o f the fourth segment is provided with one. al-Rudhbari (d.320/934) is remembered as one ofjun ayd’s followers. See R , 99.3. Despite m entioning Sulami in its isnad as the immediate source, this segment is not found in his work, nor in any o f the earlier biographies o f Junayd. See pp.82-3; 95 above. See R , 7 2 .1 0 -1 6 . R , 7 2 .1 7 -1 9 . R , 7 2 .1 9 -7 3 .1 . R , 7 3 .2 -4 cf. H, X , 2 6 4 .1 -3 . TABA, 1 6 2.13-15. See p.55 above concerning the two com peting accounts o f Junayd’s death, w hich have both been included in the Hilya. See above. See L, 3 8 0 -3 9 0 , and Ernst, Words o f ecstasy. See especially R , 7 1 .1 1 -2 0 , a polemical attack on the doctrine o f tark alharakdt (abandonment o f pious deeds after attaining proximity to God). R , 4 6 6 -4 8 4 . R , 4 8 5 -5 2 5 . See p. 121 below concerning the omission o f Hallaj from the tabaqat section, despite the provision o f segments about him in the systematic section. See pp. 1 1 2 -1 3 above. The final com m ent is a reference to the Muslim funeral prayer which begins with four repetitions o f A llahu akbar. See pp. 112—13 above. O ne might suggest that the variant provided here has been generated by the combination o f the ‘severance o f the girdle’ m otif and the autobiographical narrative form, which is relatively frequent in the biographical tradition o f Abu Yazld (and especially associated with spiritual progression). See Ch. 2, Section IV above. R , 4 7 7 .2 1 -4 7 8 .5 cf. L, 28 2 .9 -1 4 ; 2 8 5 .1 2 -1 8 . This can be observed by noting Gramlich’s references to parallel passages in the L u m a c (Gramlich, Sendschreiben, 4 6 1 -7 9 ). Although the term murad can have various uses, this particular segment is preceded by the follow ing definition: al-m und huwa *l-m ubtadi’ wa- }l-murad huwa ’1-muntahT (R, 310.1) The second segment parallels a segment included in the Hilya, in w hich Abu Yazld is sent a prayer-rug with a written request that he pray on it at night, to

203

N otes w hich he replies that he has already collected all acts o f worship ( cibadat), ‘put them in a pillow, and placed it under his cheek (i.e. to sleep)! (ja -ja caltuhdfT mikhadda wa-u>adactuhd tahta k h a d d iy . See H, X, 35.22—36.1. 82 See Gramlich, Das Sendschreiben, 608. These include a variant o f a segment in Sulaml’s biography o f al-Hallaj, R , 4 4 1 .3 -4 cf. TABS, 3 0 9 .6 -8 . 83 See R , 16.6-17.13. 84 See Section II above.

Chapter Five: H u jw in ’s K a sh f al-m ahjub 1 Jullab and Hujwlr are neighbouring suburbs o f Ghazna. 2 In the K a sh f al-mahjub HujwTrl refers to at least nine further works o f his own, none o f which have survived (see HujwTrl, The K a sh f al-Mahjtlb, x i-x ii). 3 e.g. KM, 314.13. 4 e.g. KM, 6.18. 5 e.g. KM, 6.18; 86.5. 6 e.g. KM, 3 3 7 .1 0 -1 2 . 7 KM, 11 2 -7 cf. 143-5; 145-6. 8 KM, 208.4. 9 See Ethe, Catalogue , I, 969 (No. 1773). R eynold N icholson has observed the discrepancy between these dates and those given for the deaths o f individuals w hom Hujwlrl appears to have outlived according to the text o f the K a sh f almahjub. He suggests a revised estimate between 4 6 5 /1 0 7 2 -3 and 4 6 9 /1 0 7 6 -7 (See Hujwlrl, The K a s h f x -x i). 10 See KM, 7.6. 11 It is extant in several manuscripts, the earliest o f which is thought to have been written in the early ninth/fifteenth century (see HujwTrl, The Kashf, 53). 12 Hujwlrl, T he Kashf, x. 13 e.g. See pp. 128—9 below, concerning the chapter on sam dc. Furthermore, although specific sources are seldom acknowledged, at one point Hujwlrl describes accurately the arrangement o f biographies in SulamT’s Tabaqdt alS u fy y a and Qushayri’s Risdla (see KM, 141.4—8). 14 Poverty is discussed in Chapter Two, whilst Chapter Three is largely taken up with a discussion o f purity (see KM, 34.18—40.6). 15 Chapter Five stands apart from the others in this first Section o f the work because it is concerned with the specific debate about the relative significance o f two different aspects o f Sufism (poverty and purity). Furthermore it fails to follow the same structural pattern as the other chapters o f Section One. and it is also more than three times shorter than each o f them. 16 Taken at face value, these opinions may appear to be incoherent and contradictory, suggesting that Hujwlrl is a confused and indecisive author. For example, see Baldick, Mystical Islam, 63—4, where he suggests that Hujwlrl ‘is unable to provide a coherent verdict on whether poverty should be preferred to wealth’. The discussion which he refers to is parallel to Hujwlrl’s discussion about intoxication (sukr) and sobriety (sahw) (see pp. 142—6 below); similarly Hujwlrl declares a preference for poverty (faqr) as an attribute applicable to a Sufi and justifies this view point, but, at the end o f

204

N otes

17

18 19 20 21 22 23 24 25 26 27

28 29 30 31 32 33 34 35

36 37 38 39

40 41

the discussion, he makes an allowance for those who prefer wealth (ghand ’). By this technique, he manages to accommodate seemingly contradictory opinions (see KM, 24.4 -2 8 .1 1 ). See further Mojaddedi, ‘Extending the boundaries’. Hujwlri also appears to be particularly interested in the customs relating to wearing cloaks. H e refers to a separate work, Asrdr al-kkiraq w a -’l-m ulaum w idt (The secrets o f patched and m ulti-coloured cloaks), w hich he claims to have also written about this topic (See KM, 63.9). Chapter Four, on tahdra, is about the purification required as preparation for the performance o f the four main Muslim rituals. Chapters N ine to Eleven account for 175 pages (pp.432-607), whilst the first eight chapters make up only 91 pages (pp.3 4 1-432). KM, 5 0 8 -5 4 6 cf. L, 2 6 7 -3 0 0 . KM, 5 4 1 .17-542.13. KM, 5 4 2 .1 3 -1 8 . KM, 5 4 2 .18-544.15. The biography o f Ali, the first Imam, is included in the preceding chapter on the ‘rightly-guided Caliphs’. See pp. 103—7 above. However, there are corresponding biographies for only eighteen o f the tw enty-one biographies in that w ork’s second generation. Two o f these biographies (Khayr al-Nassaj and Abu Hamza ’1-KhurasanT) have actually been included amongst biographies corresponding to the second generation: KM, 182-185 cf. TABS, 3 2 4 -3 3 1 . They are also juxtaposed in the same order in both works. KM, 19 5 -7 . KM, 135-137. KM, 143-145. KM, 14 5 -6 . KM, 189-193. TABS, 3 0 8 .6 -1 0 . The individuals m entioned are Abu ’1-‘Abbas b. ‘Ata’, Muhammad b. Khafif and Abu ’1-Qasim al-Nasrabadhl. H ujwiri’s list o f those w ho accept Hallaj includes the names o f the three m entioned by SulamI in addition to cAmr b. ‘Uthm an (al-Makki), Abu Ya'qub al-Nahrajuri and cAlI b. Sahl al-Isfaham (see further p? above). Hujwlri also adds that the recent Sufis accept him (wa jum la-yi m u ta ’akhkhiran qabul kardand-ish) (KM, 189.11-15). KM, 189.15-16. KM, 189.16-20. KM, 189.17-20. For example, Hujwlri accepts criticism o f Hallaj’s behaviour, giving by way o f example his departure from the company o f teachers without permission, but he argues that his faith remained sound throughout. KM, "l9 2 .8 -1 0 . The utterance is presented at first in Arabic as, ‘The tongues are seeking to speak, under their speech they are seeking to perish (al-alsina mustantiqdt tahta nutqihd m u stahlikdty . It is followed by a Persian translation, w hich conveys, ‘Talkative tongues are the destruction o f silent hearts (zabdnhd-yi guya haldk-i

205

N otes

42 43 44 45 46 47

48 49 50 51 52 53 54

55 56

57 58 59 60 61 62 63 64

dilhâ-yi khdm üsh-asty (KM, 193.2). This utterance may be interpreted as relating directly to the reason given in the later Sufi tradition for Hallaj’s execution. See further Mojaddedi, ‘Extending the boundaries’. KM, 2 0 2 -2 1 4 . See p. 125 above. KM, 1 9 0.5-6. KM, 21 8 -3 1 4 . In fact, the ten groups are listed previously in the introduction o f the biography o f Abü ’1-Hasan al-Nürï, after w hom one o f these groups, the N ilriyya, is named (See KM, 164.7—17; 236—244). See pp. 143—6 below. See pp. 1 4 3 -4 below for further details. See KM, 3 3 3 -3 4 1 . See p. 133 above. See further pp. 1 4 3 -6 below concerning al-Tayfüriyya and al-Junaydiyya. See pp. 1 8 -1 9 above. For instance, whilst in the Tabaqdt it is stated that Junayd practised jurisprudence according to the principles o f Abü Thawr (tafaqqaha cald A b\ Thawr), here in the K a sh f al-mahjiib it is stated that he was [no less than] ‘the leader’ o f the latter’s companions (im dm -i ashab-i A b ü Thawr), thus elevating his status in this context even further. Moreover, Junayd is introduced in the K a sh f al-mahjiib as a ‘mufti’ (implying som eone with authoritative opinion) in the disciplines with which he is associated. This relatively uncom m on usage is probably derived from the report in the Tabaqdt’s corresponding introduction, w hich describes him as giving fatwas (kdna yufti) in the circle o f Abü Thawr (see pp. 18—19 above). See pp. 1 3 7 -8 below This is a reference to the ten accepted Sufi groups, each o f w hom is linked back to a Sufi leader. The Tayfiiriyya are associated with Abü Yazïd (see below ). See p. 18 above. See p. 136 above. See p p .50-52 above. KM, 161.15-162.14. KM, 162.13-14. See p p .26-7 above. KM, 162.14-163.1 cf. TABS, 147.6-10. See KM, 303.8—307.4. The two groups o f heretics are both referred to as anthropomorphists (giriilu a z hashwiydn ki mujassima-yi ahl-i Khurdsdn-and . . . ginlliT digar a z mushabbiha).

65 KM, 163.1-14. 66 See pp. 160—162 below, concerning a similar narrative in Jaml’s Nafalidt, w hich even incorporates the hadith about firdsa. 67 See p.26 above. 68 KM, 132.19-133.12. 69 KM, 133.1 2 -1 3 cf. TABS, 6 3 .6 -7 . Although the conjunction wa- is ambiguous, being translatable as ‘w h ile’, as well as ‘and’, H ujw M ’s commentary shows that he prefers the latter possibility.

206

N otes 70 71 72 73 74 75 76

KM, 133.15-17. KM, 133.15-16. KM, 133.1 8 -1 9 . See p.133 above. KM, 2 3 0 .4 -5 . H, X , 4 0 .1 0 -1 4 . See p.54 above. Intoxication (sukr) is mentioned only in one other instance in the earlier biographies o f Abü Yazld, namely in the introduction o f the hadith transmission o f Abü Yazid’s biography in the H ilyat al-awliyä\ where it is suggested that his deep utterances were produced and issued from his intoxication (näfathät sirrih al-mutaivallada ’l-muntashara min sukrih ) (H, X, 41.8). 77 See KM, 2 3 0 .4 -9 . 78 This is confirmed by the text, since virtually every chapter related to Sufi conduct contains a few utterances attributed to him.

Chapter Six: Jäm l’s Nafahät al-uns 1 See N , Editor’s Introduction, 7. 2 The Naqshbandi Sufi leaders are referred to traditionally using the title khumja, meaning ‘master’, and are also referred to collectively as the Khwäjagän, especially before the time o f Bahä’ al-Dln Naqshband (e.g. see N , 4 1 0 .7 -9 ). 3 See the biographies o f Käshghari and Ahrär in Jäml’s Nafahät, 408.5—410.7; 410.8—416.14. Jämi, writing after Käshgharl’s death, does not actually m ention that he was his disciple. Ahrär’s biography is three times the length o f the biography o f Käshghari, and its position suggests that he is considered the contemporary representative o f the Naqshbandi tradition. This is even declared unambiguously at the beginning o f his biography. 4 See N , Editor’s Intro., 12—20. 5 See Jämi, N a q d al-nusils. 6 This work has been translated into English by E.H. W hinfield and M .M . Kazvlni (see bibliography). 7 See N , 1—2.21. For more information on N aw ä’I, see Barthold, Herat, and N , Editor’s Intro., 11-14; 49; 53. Jämi explains his choice o f title, by remarking that, ‘[the work] which, by virtue o f containing wafts (nafahät) o f the sweet breaths o f the shaykhs, which have arrived from paradise and have blown onto the smelling faculty o f the souls o f those yearning for instances o f presence in intimacy (uns) [with God], will be called ‘The breaths o f intimacy from the occasions o f presence in paradise’ (N, 2.24—25). 8 For details, refer to N , Editors Intro., 47—50. 9 This manuscript (Morad Mulla Library, no. 2349) shows indications that it had been checked and corrected, perhaps even by Jämi himself. 10 The introduction contains sections on wiläya (sainthood), ma'rifa (gnosis), types o f mystics (e.g. süß, malämati, qalandar), tawlud (divine unity) and types o f miracles (karämat, m u jiz a ) (see N , 1-25). 11 The main text o f the oldest manuscript contains 585 biographies. There is considerable variation between manuscripts with a tendency for the number o f biographies included to increase with time.

207

N otes 12 Although there are thirty-four biographies o f Sufi women, they amount to a total o f only twenty pages in the printed edition. The longest o f these biographies (approx. 100 lines in the printed edition) is devoted to ‘Tuhfa’, who is associated with Sari ’l-Saqatl (N, 623.6-626.3). However, most o f the biographies consist o f only one brief narrative, and many o f these are variants o f each other (e.g. see N, 630.21—631.9). Far from providing a significant resource on the contribution o f wom en to Sufism, this appendix reveals Jam! stretching his limited resources, as if under pressure to include the biographies o f women. 13 N , 2 7 -2 9 9 . The introduction o f the Persian Tabaqat also forms the basis for the final section o f the introduction o f the Nafahât (see N , 2 3 -2 5 cf. TABA,

1_6)-

14 For example, see pp. 1 5 5 -6 6 below, concerning the biographies o f Abu Yazid and Junayd. 15 N , 3 3 .8 -1 5 . 16 See p p .72-3 above. 17 See TABA, 362-372; 390-397; 429-436; 4 8 6 -4 9 2 . 18 See N , 179-204. 19 See N , 23.6—13. See further pp. 104—5 above. 20 e.g. The biography o f Abü Abdullah al-Hadramï has been inserted after that o f Abü Abdullah ‘Abbâdânï (see N , 121). 21 e.g. The biography o f Ibrâhïm al-Sayyâd al-Baghdâdï has been inserted after that o f Ibrâhïm Utrüsh (see N , 43). 22 e.g. The biography o f Muhammad Khâlid al-Ajurrl has been inserted after that o f Ibrâhïm Ajurrï Kabïr (see N , 43). 23 Eighteen o f them follow the biography o f Ibn Khafïf al-Shïrâzï (see N , 24 2 -2 5 9 ). 24 e.g. The biography o f Ahnaf al-Hamadânï consists o f a single narrative in which the subject is described as appealing and receiving G od’s help, whilst travelling alone through a desert. The final segment o f the immediately preceding biography, w hich is devoted to Abü Sacïd Kharrâz, is a narrative containing the same m otif (see N , 7 5 .7-18; 7 5 .20-76.5). 25 e.g. The biography o f Rüzbihân Baqlï (d .6 0 6 /1209) is found juxtaposed next to that o f Abü Ishâq Kâzarünï (d.426/1035), despite the fact that they lived centuries apart, apparently because they were both from Fârs (see N , 2 60.5-263.17). 26 This may be interpreted as a sign that the compilation was a collaborative effort under the supervision o f the author. 27 N , 6 3 .8 -6 5 .2 2 cf. TABS, 8 0 -92; 131-135. 28 See pp .94-5 above. 29 e.g. O nly that biography o f Surayj which contains biographical material about him is retained. The other ‘false’ biography o f Surayj, w hich actually contains biographical material about Junayd and is found shortly after the latter’s biography, is omitted. However, some o f the contents o f the omitted biography are included in the biography o f Junayd for which they are more appropriate (see p. 157 below). See also the biographies o f Sharif Hamza-yi cAqïlï (TABA, 132; 525 cf. N , 284). However, a new repetition o f a biography has been made in the compilation o f the Nafahât (N, 61 c f N , 91 - two biographies for Ahmad b.'Asim Abü ‘Abdullâh al-Antâkï, headed using his ’ism and kunya respectively.

208

N otes 30 31 32 33

34 35 36 37 38 39 40 41 42

43 44 45 46 47

48 49 50 51 52 53 54 55 56 57 58 59 60

61

See N , 2.5—13, and pp. 167—8 below. N , 5 4 .2 2 -5 5 .2 5 cf. TABA, 8 8 .1 3 -9 2 .9 . For an explanation o f the term ‘bundle’ in this context, see p.75 above. For example, a segment about Abu Musa ’1-DabllI was included in the earlier biography, simply on account o f its association with the preceding segment about Abu Yazld w hich names Dablli as its source. However, not all the segments o f addenda have been omitted (see TABA, 90.10, 91.1—2). See N , 55 .7 -1 0 ; 2 2 -2 5 . See pp .79-85 above. N , 80.20 cf. TABA, 168.2-3; N , 80.21 cf. TABA, 168.14-169.1; N , 80.24 cf. TABA, 169.2; N , 8 1 .3 -4 cf. TABA, 1 6 9.3-5. N , 8 1 .5 -2 2 cf. TABA, 182.3-183.12. N , 8 1 .2 3 -2 4 cf. TABA, 185.12-186.3; N , 8 2 .2 -1 5 cf. TABA, 188.12-189.15. TABA, 163.14-167.5. See further p. 163 below. See p. 191, n.13 above. The display o f miraculous knowledge about an outsider is a com on m otif in Sufi narratives, and indeed has been encountered already in the form o f a narrative about Abu Yazld in Sulami’s Tabaqat al-Sufiyya (see p.25 above). N , 5 5 .1 1 -1 2 cf. TABA, 9 0 .2 -6 . TABS,142.2-8; H, X , 28 1 .2 2 -2 8 2 .4 . See also pp.20-1; 58 above. See p. 139 above. See p. 138 above. Element N 8 belongs to a stock o f segments that serve to illustrate the superiority o f a religious authority over a political authority. See further TABA, 1 6 3.9-11. See TABA, 161.8-162.12. See p. 1 0 8 -9 above. See Yafi‘1, M ir'at, II, 231.1; 5 -6 . See TABA, 16 3 .3 -8 . As opposed to ‘illustrative’ details (see Fahndrich, ‘The Wafaydt al-acyan , 438). See TABA, 8 7 .1 0 -8 8 .8 . See N , 1 .7 -1 6 . His description o f Sulami’s Tabaqat al-Sufiyya is based on that w ork’s ow n introduction. See Chapter Three above. e.g. See above concerning the biographies found in Sulami’s Tabaqat and also in Jaml’s Nafiahat despite having been omitted by Ansari. See below and N , 2 .5 -1 4 . See further J.A. Mojaddedi, ‘Jaml’s re-contextualisation o f biographical traditions’. See N , 385.1-4; 5-6. For instance, the biographies o f Qutham Shaykh and Khalil Ata (N, 388.3—389.23), w hich intervene between those o f Sayyid Amir Kulal and Baha’al-D ln Naqshband are both made up almost entirely o f material attributed to Naqshband (see Fig.2). Although cAla’ al-Din cAttar is said to have been put in charge o f many o f Naqshband’s students, within the latter’s lifetime (N, 394.7-12), it is Khwaja Muhammad Parsa w ho is identified as Naqshband’s successor. Muhammad

209

N otes

62 63 64 65 66 67 68 69 70 71 72 73

74

75 76 77 78 79 80 81 82 83 84

Parsa’s biography is followed by that o f his son and successor Khwaja Nasr Parsa (N, 401.2), which precedes the biography o f "Ala’ al-D ln’s son and successor, Khwaja Hasan "Attar (N, 401.15). The next three members, Ya'qub Charkhi, Nizam al-Dln Khamush and'Abdullah ImamI Isfahan!, are all described as associates o f "Ala’ al-Dln "Attar. Sa"d al-Dln Kashgharl, w ho is traditionally remembered as Jami’s own teacher, fits into the framework by virtue o f his association with Nizam al-Dln Khamush (see Fig.2). See N ,406.13-24; 4 0 3 .7 -2 1 . N , 416.1 9 -4 1 7 .8 . See N , 415.18-416.18; 4 1 0 .7 -1 1 . e.g. See N , 4 3 9 -4 4 1 . The bulk o f the material in the two biographies which precede that o f "Ala’ al-Dawla al-Simnanl is attributed to the latter. N , 455.8-14. N , 46 1 .9 -4 6 6 .1 . N , 460.1 7 -4 6 1 .8 . N , 4 6 6 .1-469.13. N , 469.1 4 -4 7 1 .8 . N , 47 1 .1 0 -1 4 . See N , 4 1 9 -4 2 0 . The five biographies w hich follow after that o f "Abd alQadir al-Jill similarly continue to present material about him (N, 510—524). The most prominent exception is the biography o f Afdal al-Dln al-Khaqani which appears after that o f Shaykh AwhadI Isfahan! (d.738/1337). It is stated that al-Khaqani lived during the caliphate o f al-Mustadl’ (d. 595/1198) [N, 506.13-14; N , 6 0 3 .21-22]. Fakhr al-Dln "Iraqi and Amir Husaynl are both described as disciples o f Baha’ al-Dln Zakarlya’ (see N , 599.11—602.21; 602.22—603.20); Khusraw and Hasan DihlawTI are similarly both students o f Nizam al-Dln Awliya’ (N, 607.11-608.11). N , 593.13. N , 611.20. N , 6 1 2 .6 -7 . N , 6 1 2 .4 -5 . N , 3 8 0 .2 0 -2 1 . N , 4 2 0 .1 6 -1 9 . N , 3 7 5 .6 -7 . See further Calder, Studies, 2 3 6 -4 1 . Hujwiri has already been observed to grant Abu Yazld a pre-em inent status (see pp. 13 6 -7 above). For instance see N , 328—334, which represents the origins o f the Chishti tradition. The six biographies which make up this section display a striking uniformity in form and content, which suggests that they have been imported from the same work. See further J.A. Mojaddedi, ‘JamT’s recontextualisation’.

C onclusion 1 R icoeur, ‘The narrative function’, 293. 2 Wansbrough, Sectarian milieu, 130. See also C ook, Dogma, C h .ll.

210

N otes 3 Rippin, 1Quranic Studies’, 44. 4 For a contemporary example, see Nurbakhsh, M ardän-i Süß. 5 Kearney (Ed.), Dialogues, 28.

211

This page intentionally left blank

A ppendix

T h e biographies represented in these charts are num bered according to their order o f appearance in the text concerned. T he associations betw een biographies, indicated in Figs. 2 to 5, are not necessarily historical associations betw een their subjects (see further C h .6 above).

213

2. al-Murta‘ish

19b. Ahmad b. Abi ’l-Ward

19. ‘Abdullah al-Antäki 20. al-Nakhshabi

A pp en d ix 1 The framework of Sulami’s Tabaqat al-Snfiyya

20. Abü ‘Abdullah al-Sijzi

18. al-Juzjänl 19a. Muhammad b. Abi ’l-Ward

18. Ahmad al-Anüäki

19. al-Shibhi 20. al-Farrä’

20. Ibn Abi Sacdän 20. Ibn Hamdän

23. Abü ‘Abdulläh al-Dinawari

22. al-Räshinl

21b. Ja‘far al-Muqri’

21a. Muhammad al-Muqri’

18. al-Sayrafi

18. Ibn Sälim 19. al-Nasawi

18. Abü Hamza 1-Khurasäni

17. Abü “Abdullah al-Rüdhbäri

16. al-Turüghbadhi

15. al-Husrl

14. al-Nasräbädhi

10. Bundär b. al-Husayn

9. Ibn Khafif

19. al-Subayhi

17. al-Muw7allad

17. Khayr al-Nassäj 17. Abü ‘Abdullah al-Maghribi 17. Mansür b. ‘Ammär

14. Abü Sacld al-Kharräz

14. Yahyä b. M ucädh

15. al-Qirmisini

14. Abü ’l-Hasan al-Dinawari

13. al-Warräq

13. Ibn Khadrawayhi

16. Ibn Yazdäniyär

14. Ibn Hind al-Färisi

13. al-Halläj

12. Ibn ‘All ’1-Tirnndhi

15. Mumshädh al-Dinawari

12. al-Abhari 13. al-Qirmisini

12. Abü Bakr al-Wäsiti

11. Muhammad al-Balkhi

11. Hätiin al-Asamm 12. al-Hawäri

16. Ibrahim al-Qassär al-Raqqi

11. Ibn Bunän

11. Abü ’l-Husayn al-Warräq

10. al-Tustarl

10. Macrüf al-Karkhl

15. ‘All b. Sahl al-Isbahäni

13. Abü ‘Uthmän al-Maghribi

10. Ibn al-Kätib

10. Abü Hamza al-Baghdädi

9. 'Amr al-Makkl

9. al-Däränl

16. Abü ’l-‘Abbäs b. Masrüq

12. al-Dinawari

9. al-Muzayyin

9. Bunan al-Hammäl

8. Samnün b. Hamza

8. al-Bastämi

15. Abü Hafs al-Nisäbüri

11. al-Tamastäni al-Färisi

7. al-Kattäni 8. al-Nahrajüri

7. Ibrahim al-Khawwäs 8. ‘Abdullah al-Kharräz

7. Shäh al-Kirmäni

7. Shaqlq al-Balkhi

16. Hamdün al-Qassär

7. Ibn Nujayd al-Sulaml 8. al-Büshanji

6. al-Aqta‘

6. Ibn Hamid al-Tirmidhi

6. Yüsuf al-Räzi

6. al-Muhäsibi

5. al-Duqqi 6. ‘Abdulläh al-Räzi

5. Ibn Munäzil

5. Abü ‘Umar al-Dimashqi

5. Ruwaym b. Ahmad

4. al-Sayyäri

3. Abü ‘Ali ’l-Rüdhbäri 3. Ja‘far al-Khuldi

1. Abü Sa‘id b. al-A‘räbI 2. Abü cAmr al-Zajjäji

1. al-Shibll

4. al-Thaqafi

5. al-Saqati

3. Mahfuz b. Mahmüd

5th Generation

4th Generation

4. Tahir al-Maqdisi

3. al-Hiri 4. al-Jallä’

4. Bishr b. al-Harith

2. Abü ’l-‘Abbäs b. ‘Ata’

2. al-Nürl

2. Ibrahim b. Adham 3. Dhü ’1-Nün

1. al-Jurayri

1. al-Junayd

3rd Generation

2nd Generation

1st Generation

1. al-Fudayl b. ‘Iyäd

1 7 .'A bdullah Im am i Isfahan! 18. Sacd al-DIn Käshghari (d.860/1455)

16. Nizam al-D m K ham ush

19. cUbaydulläh Ahrär (d.896/1490)

15.Ya'qub Charkhi

10. Baha’ al-DIn N aqshband

7. Sayyid A m ir Kulal

A p p en d ix 2 The Naqshbandi cluster of biographies (Cluster 3.1) in the N a fa h ä t al-uns

14. Hasan cA ttar

1 1 .cAla’ al-Dm cAttar

8. Q utham Shaykh

6. M uham m ad Baba Samasiyi

5. cAlI R äm itäni

4. M ahm üd A njir Faghnawl

3. cÄ rif R iw giraw l

2. cAbd al-Khäliq G hujdaw änl

1. Y üsufH am adänl

Abü cAlI Färmadi (See Fig.5)

13. Abu Nasr Parsa

I

12. M uham m ad Parsa

9. Khalll Ätä

12. N ajm al-D in R äzi 11. Abü ’1-W afa’ K hw arazm i

10. K im ä lja n d i

28. M uham m ad Shah Farahi 30. Shams al-D in M uham m ad Asad (d.864/1459))

29. B aha’ al-D in ‘U m ar

27. Shah cAli Farahi

r

26. Fakhr al-D in Lüristam

I

25. B aha’ al-D in A bardihi

24. A khi cAli Q u tlu q Shäh

■22. ‘A bdullah G harjistani

A p p en d ix 3 The cluster of biographies beginning with Abu Najib al-Suhrawardi (Cluster 3.3) in the Nafahat al-uns

21. Sayyid cAli H am adäni

I

23. Bäbä M ah m ü d T üsi

9. Jam al al-D in Gill

20. Taqi ’1-Din Dusti

14. Jamal al-D In Jürfam

13. R adi '1-Din al-G haznaw i

; 3. R uzbihan Kabir M isn

19. AkhI M uham m ad D ihistanl

8. Sayf al-D in Bakharzi

^5. N ajm al-D in al-K ubrä .

I'

4. Isma'il Q asri

I

1. Abü N ajib al-S uhraw ardi-

17. cAli M isri —_____ 18. N aim al-D In a l- A d k ä n i____— 16. ‘Alä’ al-D aw la ’l-S im nänl :

15. N u r al-D in Isfarayini

7. Sacd al-D in H um m uw ayi

6. M ajd al-D in Baghdadi -

2. ‘A m m är Yäsir ;

Ahmad Ghazäli (See Fig.5)

(N iz ä m a l-D in H a ra w l)----------

17. A bu Y azld P uranI (d.8 6 2 /1 4 5 7 )

18. Z a h lr a l-D in K halw atl

- 16. M a h m u d Z a h id M u rg h ab I

.^15. A b ü B akr T äyibäbädl

______(A hm ad al Jam I)

10. A b d al R az z ä q K ashi

14. M u h a m m a d a l-K u sü ’I al-Jäm i (d.8 6 3 /1 4 5 8 )

, 12. A b ü B ak r al-K haw afl

I

I I . ‘A bd al R a h m a n al-M isri

7. Sham s al D in Safi

A p p en d ix 4 T h e c lu s te r o f b io g ra p h ie s b e g in n in g w ith S h ih a b a l-D ln a l-S u h r a w a rd l (C lu s te r 3 .5 ) in th e Nafahat al-uns

13. Q a w ä m a l-D In Sinjam '

9. M u h a m m a d K a s h i'

a l-D m L ü r

. 8. cA bd al-S am ad N atan zI

4. M u h a m m a d YamanT

5. Ib rah im Majdhub__

________

-_2. CA1I b. B u zg h u sh al-ShlräzI

I

20. N iz ä m a l-D m A w liya’

19. B ah a ’ a l-D In Z a k riy y a ’ M u lta n l

3. cA b d a l-R a h n ia n b. ‘A ll b. B u zg h u sh -

I. Shihab al-D In a l-S u h raw a rd l ^

1

A bu N a jib a l-S u h raw a rd l (See Fig. 3)

r

(cA y n a l-Q u d à t H a m a d à n l C luster)

B ah ä ’ al-D In Z akarlyä

(See Fig. 4)

(A m ir H usaym )

(F akhr a l-D In "Iraqi)

(P ersian P oets C luster)

3.8. H a k im S ana‘i

3.1 Y u s u f H a m a d à n l (See Fig. 2)

3.2. CA1I H asan B usti

-3 .5 . S hihab a l-D m a l-S u h ra w a rd i

- (N a jm a l-D in al-K ubra)

(see Fig. 3)

, I

2.5. A b ü A li F ärm adi

2.1. A b ü ’1-H asan a l-K h a ra q än l

I

1. A b ü Y azld

A ppendix 5 Associations between the clusters in Section 3 (m uta’akhkhimn) of the Nafahat al-nns , including their links back to Section 1 ( mutaqaddiman )

(Ibn a l-cA rabI C luster)

3.7. Ib n a l-cA rabi

I

(cA b d a l-Q ä d ir al-JIli C lu ster)

3.6. cA b d a l-Q ä d ir al-JTli ---------

(Jaläl a l-D In al-B a lk h l C lu ster)

3.4. B ah ä3 a l-D In W a la d ----------

3 .3. A b u N ajib al-S u h raw ard i

I

2.5. A h m a d al-G hazali

I

2 .5. A b ü B akr al-N assaj

I

2.1. A b ü ’1-Q äsim a l-G u rg ä n l

1. al-Junayd

Bibliography

cA l l Hassan, T h e life, personality and writings o f al-Junayd, (London, 1976) A l - A n s ä r i , ‘A b du llah, Tabaqät a l-S ü fiyy a (Ed. C A .H . H a b ib l), (Kabul, 1967) ------- , Tabaqät a l-S ü fiyy a , (Ed. M .S. M awläyi), (Tehran, 1983) ------- , Tabaqät a l-S ü fiy y a , M S. N afiz Pasha 426 (photographed copy in the m anuscript library o f Tehran University) I b n A l - cA r a b I , M u h yi T-D in, a l-F u tü h ä t a l-M a k k iy y a (4 Vols; Cairo, 1911) A r b e r r y , A . J . , S u fism , (London, 1950) ------- , ‘JämT’s B iography o f Ansari’, Islamic Q uarterly 7 (1963) ‘Attar, Farid al-D ln , Tadhkirat a l-a w liyä ’, (Ed. R .A . N ich o lso n ), (2 Vols; L ondon, 1 9 0 5 -7 ) A u c h t e r l o n i e , P., A rabic biographical dictionaries, (Durham , 1987) A l - b a g h d ä d I , A .B . al-Khatlb, T a ’n k h Baghdäd, (14 Vols; Cairo, 1 9 3 Iff.) B a l d i c k , J . , M ystical Islam , (London, 1989) B a r t h o l d , V.V., H erat z u r z e it H u sa in Baiqaras, (Trans. W. H inz), (Leipzig, 1968) A l - b i r ü n i , A . R ., a l-A th ä r al-bäqiya can al-qurün al-khäliya, (Ed. C .E. Sachau), (Leipzig, 1923) B o n n e r , M ., Aristocratic violence and holy war, (N e w H aven, 1996) B ö w e r i n g , G., T h e m ystical vision o f existence in classical Islam : the Q u r ’anic hermeneutics o f the S u fi S a h l al-T u sta n ( d .2 8 3 / 8 9 6 ) , (B e r lin /N e w York, 1980) ------- ,T h e m inor Q u r ’an com m entary o f A b ü cA b d a r-R a h m ä n M u h a m m a d b. alH u sa yn a s-S u la m i (d. 4 1 2 / 1 0 2 1 ) , (Beirut, 1995) ------- , ‘T h e Q u r’an com m entary o f al-Sulam l’ in W. Hallaq & D. Little (Eds) Islamic studies presented to Charles J. A d a m s, (Leiden, 1991) B r o c k e l m a n n , C ., Geschichte der arabischen Litteratur, (5 Vols; Leiden, 1 9 3 7 -4 9 ) B u l l i e t , R .W ., T h e patricians o f N ishapur, (Cam bridge Mass., 1972) C a l d e r , N ., Studies in early M u slim jurisprudence, (O xford, 1993) ------- , ‘H istory and nostalgia: reflections on John W ansbrough’s T h e Sectarian M ilie u *, in M eth o d and theory in the stu d y o f religion 9.1 (1997) A b d e l-k a d er,

219

Bibliography C h a b b i, J., ‘R em arques sur le d évelop pem en t historique des m ouvem ents ascétiques et m ystiques au Khurasan’, Stu d ia Islamica 46 (1977) C h a u m o n t , E., ‘al-salaf w a -’l-khalaf’, Encyclopaedia o f Islam (2nd Edn) C o o k , M ., E arly M u slim dogm a , (Cam bridge, 1981) C o o p e r s o n , M ., ‘Ibn Hanbal and Bishr al-H áfí’, S tu d ia Islamica 8 6 (1997) D a b a s h i , H ., ‘H istorical conditions o f Persian Sufism during the Seljuk period ’, in L. L ew isohn (Ed.), Classical Persian S u fism : fro m its origins to R u m i, (London, 1994) D a u b e , D . , Studies in Biblical Law, (Cam bridge, 1947) A l-D a y la m I , A bü ’1-Hasan, STrat-i Ibn K h a ftf al-ShirazT, (Ed. A. Schim m el), (Ankara, 1955) D e B e a u r e c u e i l , Serge de Laugier, K hw á d ja 'A bdullah al-Ansari: M ystiq u e H anbalite, (R echerches d ’Institut de Lettres O rientales de B eyrouth 26), (Beirut, 1965) ------- , ‘al-Ansari al-H araw f, Encyclopaedia o f Islam (2nd Edn) ------- , ‘Abdullah al-Ansari al-H aravi’, Encyclopaedia Iranica D e G o e j e , E .M . et al., Catalogus codicum arabicorum , (Leiden, 1888—1907) D e J o n g , E et al., ‘al-M alam atiyya’, Encyclopaedia o f Islam (2nd Edn) A l - d h a h a b I , Shams al-D in , Siyar a'lam a l-n u b a W (Ed. Arnaut S. et al.), (23 Vols; Beirut, 1 9 8 1 -5 ) ------- , Ta’rlkh al-Islam , (Beirut, 1994ff) E l i o t , T.S., Collected poem s Í 9 0 9 - Í 9 6 2 , (London, 1963) E r n s t , C.W ., Words o f ecstasy in S u fism , (N e w York, 1985) E th é,

H .,

Catalogue o f the Persian M S S . in the library o f the India Office,

(O xford, 1903) F a h n d r i c h , E . , ‘T h e W afayat a l-a cyan o f Ibn Khallikán: a n ew approach’, Jo u rn a l o f the A m erican O riental Society 93.4 (1973) A l- F a r is I , cA bd al-Ghafir, ‘al-Siyaq li-T a ’n k h NTsábür’, in R . Frye (Ed.), T h e Histories o f N ishapur, (London, 1965) F r a n k , T., ‘ T a sa w w u f is . . on a type o f mystical aphorism ’, fo u r n a l o f the A m erican O riental Society 104.1 (1984) F u r ü z á n f a r , B., Tarjum a-yi R isa la -y i Q ush a yriyya , (Tehran, 1967) G a d a m e r , H .G ., Truth and m ethod (Trans. G . Barden & J. C um m ing) (London, 1975 [1965]) G o l d z i h e r , I., M u slim Studies (Ed. S.M . Stern), (2 Vols; N e w York, 1 9 6 7 -7 1 ) ------- , ‘ikhtiláf’, Encyclopaedia o f Islam (1st Edn) G r a m l i c h , R ., D as sendschreiben al-Q u sa yñ s uber das S u fitu m , (Stuttgart, 1989) ------- , Schlaglichter uber das S u fitu m , (Stuttgart, 1990) ------- , D ie N a h ru n g der H erzen : A b ü Talib M a k k is Q ü t al-qulüb (4 Vols; Stuttgart, 1995) G r o s s , J.A., ‘A uthority and m iraculous behaviour: reflections on kardm at stories o f Khwája "Ubaydullah Ahrár’, in L .L ew isohn (Ed.), T h e legacy o f medieval Persian S u fism , (London, 1992) H a f s i , I., ‘R ech erches sur le genre ‘tabaqát’ dans la littérature arabe’ (Parts I, II and III), Arabica 23 (1976), 24 (1977) H a r a w I , B., T a ka m m u la -yi L a n , (Kabul, 1964) I T a r t m a n n , R ., al-Kuschairis darstellung des S u fitu m s, ( B e r l i n , 1 9 1 4 ) ------- , ‘A s-Sulam l’s R isa la t a l-M a la m a tiyya , D er Islam 8 (1918)

220

Bibliography G .R ., ‘T h e d evelop m en t o f the biography o f al-H arith ibn Kalada and the relationship b etw een M ed icin e and Islam ’, in C .E . B osw orth et al. (Eds), T h e Islam ic world from classical to m odern times, (P rinceton , 1988) H e f f e n i n g , W., ‘tabakat’, Encyclopaedia o f Islam (1st Edn) H e r m a n s e n , M .K ., ‘Interdisciplinary approaches to Islamic biographical materials’, R eligion 18.4 (1988) ------- , ‘B iography and hagiography’, O xfo rd encyclopaedia o f the modern Islamic world, (N e w York and O xford, 1994) H o m e r i n , Th. Em il, From A ra b Poet to M u slim Sa in t, (South Carolina, 1 9 9 4 ) A l - H u j w I r I , cA 1ï, K a s h f al-m ahjüb (Ed. V. Zhukovski), (Tehran, 1 9 5 7 [St Petersburg, 1 8 9 9 ] ) ------- , K a s h f al-m ahjüb o f a l-H u jw m , (Trans. R .A . N ich olson ), (London, 1911) H u m p h r e y s , R .S ., Islamic history: a fra m ew o rk fo r enquiry, (Princeton, 1991) I b n A b I y a ' l â , Tabaqàt al-H anâbila, (2 Vols; Cairo, 1952) I b n A l - ci m à d , A b ü T-Falàh, Shadharât a l-d h a h a b fïa k h b â r m an dhahab, (8 Vols; Cairo, 1931) I b n A l - j a w z ï , A b ü ’1-Faraj, a l-M u n ta z a m f ï ta ’rïkh a l-m u lü k w a -’l-u m a m , (10 Vols; Beirut, 1940) Ib n A l - m u l a q q i n , A b ü Hafs, Tabaqàt a l-A w liy â ’, (Ed. N . Shariba), (C airo, 1973) I b n M u n a w w a r , M uham m ad, A srâr a l-taw hïd f ï m aqâm ât a l-shaykh A b ü S a ï d , (Ed. M .S. KadkanT), (2 Vols; Tehran, 1987) I b n Q u t a y b a , Abü M uham m ad, K itâ b a l- M a câ r if (Ed. M . al-Sâwï), (Beirut, 1970) A L - I s f a h à n î , A b ü N u 'a y m , Geschichte Isbahans, (Ed. S. D e d e rin g ), (L eiden, 1 9 3 1 -1 9 3 4 ) ------- , H ily a t a l-a w liyâ ’, (10 vols; Cairo, 1932—38) ------- , H ily a t a l-a w liyâ ’, M S. U niversity o f Leiden O r 311a,b ------- , H ily a t a l-a w liyâ ’, M S. U niversity o f al-Azhar ------- , D a lâ ’il al-nubuw w a, (Hyderabad, 1950) I v a n o w , V., ‘Tabaqat o f Ansari in the old language o f H erat \ Jo u rn a l o f the R o ya l A sia tic Society (1923) J à m ï , cA bd al-R ahm ân, L a w â ’ih, (Trans. E .H . W h in field (Ed. I. Afshâr), (Tehran, 1979) ------- , N a fa h â t al-uns m in hadarât al-quds, (Ed. M . AbidI), (Tehran, 1992) ------- , N a q d a l- n u s ü s fïsharh naqsh al-Fusüs, (Ed. W .C. Chittick), (Tehran, 1977) A l - K a l à b à d h ï , A b ü Bakr, K itâb a l-ta ca r m f li-m adhhab ahl a l-ta sa w w u f (Ed. A.J. Arberry), (Cairo, 1934) K e a r n e y , R . (Ed.), D ialogues w ith contem porary continental thinkers:the phenom enological heritage, (M anchester, 1984) K h a l i f a , Hâjjî, K a s h f a l-Z u n ü n , (Istanbul, n.d.) I b n K h a l l i k à n , A bü T-cAbbàs, W afayât a l-a cyân w a -a n b â ’ a b n â ’ a l-za m â n , (Ed. I. cAbbâs), (6 Vols; Beirut, 1977) K h o u r y , R .G ., ‘Im portance et authenticité de textes de H ily a t a l-a w liyâ r’, S tu d ia Islamica 46 (1977) K i n b e r g , L .,‘Literal dreams and prophetic hadïths in classical Islam: a com parison o f tw o ways o f legitim ation ’, in D er Islam 70.1 (1993) M a k d i s i , G., T h e R ise o f Colleges, (Edinburgh, 1981). H a w tin g ,

221

Bibliography A l- M a k k I , A b u Talib, Q iit al-qulub, (4 Vols; Cairo, 1932) M a l a m u d , M . , ‘Sufi organisations and structures o f authority in m edieval N ishapur’, International Jou rn a l o f M id d le E astern Studies 26 (1994) M a l t i - D o u g l a s , F., ‘Controversy and its effects in the biographical tradition o f al-Khatib al-Baghdadl’, S tu d ia Islamica 46 (1977) M a s s i g n o n , L., E ssay on the origins o f the technical language o f Islamic m ysticism , (Trans. B. Clark), (N otre D am e, 1997) M e i e r , E, ‘Ein w ichtiger handschriftenfund zur sufik’, O riens 20 (1967) M e i s a m i , J.S., Persian H istoriography to the end o f the twelfth century ; (Edinburgh, 1999) M e l c h e r t , C., ‘T h e transition from asceticism to m ysticism at the m iddle o f the ninth century C .E.’, S tu d ia Islamica 83 (1996) ------- , T h e fo rm a tio n o f the S u n n i schools in the 9 th to 10th centuries, (Leiden, 1998) M o j a d d e d i , J.A., ‘E xtendin g the boundaries o f Sufism: H ujw iri’s K a s h f alm a h ju l), Sufi 35 (1997). ------- , ‘L egitim izing Sufism in al-Q ushayri’s R isala, S tudia Islamica 90 (2000). See also the list o f errata due to be published in the next issue. ------- , ‘Jam l’s recontextualisation o f biographical traditions: ‘the biography o f Ansarf in the framework o f the N a fa h a t al-uns, in G .R . H aw ting, J.A. M ojaddedi & A. Sam ely (Eds), Studies in Islamic and M id d le Eastern texts and traditions in m em ory o f N o rm a n Calder ; (O xford, in press) M o t t a h e d e h , R ., R e v ie w o f R . Bulliet, Patricians o f N ishapur, in Jo u rn a l o f the A m erican O riental Society 95 (1975) A l - N a b h a n I , J a m i c karam dt al-a w liyd ’, (2 Vols; Cairo, 1962) N e u s n e r , J., E liezer ben H yrcanus: the tradition and the m a n , ( 2 Vols; Leiden, 1973) N o t h , A . / C o n r a d , L., T h e early A rabic historiographical tradition: a sourcecritical study, (Princeton, 1994) N u r b a k h s h , J., M a rd a n -i S u fi, (London, 1996) P e d e r s e n , J., ‘Q uelques remarques au sujet du texte des Tabaqat a l-su fiy y a \ M elanges Louis M assignon I I I (Damascus, 1957) P o u r j a v a d y , N . (Ed.), M a jm u ca -yi athar-i A b u cA b d a l-R a h m a n Su la n u (2 Vols; Tehran, 1990) P r o p p , V., M orphology o f the fo lkta le, (Trans. L. Scott), (Austin, 1968 [1927]) P r o u s t , M ., Rem em brance o f things p a st , (Trans. C .K .S. M o n c rieff & T. K ilm artin), (N e w Y o r k /l9 8 1 ) A l - Q u s h a y r T , A bu ’l-Q a sim , a l- M icraj, (Ed. CA. cA b d a l-Q ad ir), (C airo, 1964) ------- , a l-R a sa }il a l-Q u sh a yriyya , (Ed. M . Hasan), (Karachi, 1964) ------- , A r b a c rasa’il fi ’l-ta sa w w u f (Ed. Q. al-Samarra’I), (Baghdad, 1969) ------- , L a ta ’i f al-isharat, (Ed. I. Basyuni), (3 Vols; Cairo, 1981) ------- , a l-R isa la f (Eds C A .H . M ahm ud & M . Ibn al-Sharif), (Tehran, 1995 [2 Vols; Cairo, 1 9 7 2 -4 ]) ------- , al-Tahbir w a-'l-tadhkfr, (Ed. I. Basyuni), (Cairo, 1968) ------- , Sharh astna’ A lla h al-husna, (Ed. A. al-H ilw anl), (Cairo, 1969) R a d t k e , B. & O ’Kane, J., T h e concept o f sainthood in early Islamic m ysticism, (London, 1996) R a v a n F a r h a d i , A .G ., ‘A b d u lla h A n sa rf o f H erat, (London, 1995)

222

Bibliography R i c o e u r , P., ‘T h e narrative fu n ction ’, in J.B. T h om p son (Ed.), H erm eneutics

and

the

hum an

sciences:

essays

on

language,

action

and

interpretation,

(Cam bridge, 1981) ------- , T im e and narrative (Trans. K. M cLaughlin et al.), (3 Vols; C hicago, 1 9 8 4 -8 ) R i p p i n , A ., ‘Q uranic Studies, part IV: som e m eth odological n otes’, in M eth o d and theory in the stu d y o f religion 9.1 (1997) R i t t e r , H ., ‘P h ilologik a’, D er Islam 22 (1935) R o b i n s o n , C.F., ‘A l-M u cáfá b. cImran and the B egin nings o f the Tabaqat Literature’, J o u m a l o f the A m erican O riental Society 116.1 (1996) R o s e n t h a l , E, A history o f M u slim historiography, (Leiden, 1968) S a c h a u , C .E. (Trans.), T h e chronology o f ancient nations, (L ondon, 1879; reprinted Lahore 1983) A l - S a m cá n I , A bü Sacd, al-Ansab, (Ed. M . cAwana), (Beirut, 1976) A l - S a r r á j , A bü Nasr, K ita b a l-lu m a cf t 'l-ta sa w w u f (Ed. R .A . N ich olson ), (London, 1914) ' S c h a c h t , J., O rigins o f M u h a m m a d a n jurisprudence, (O xford, 1959 [1950]) S c h a c h t , J., ‘A bü T haw r’, Encyclopaedia o f Islam (2nd Edn) S c h i m m e l , A ., M ystical dim ensions o f Islam , (Chapel H ill, 1975) S e l l s , M . (Ed.), E arly Islamic mysticism: sufi, Q u r ’an, m i craj, poetic and theological writings, (N e w York, 1996) S e z g i n , E, Geschichte des arabischen Schrifttum s, (Leiden, 1967ff.) S m ith , G .M . & E r n s t , C. (Eds), M anifestations o f sainthood in Islam , (Istanbul, 1993) A l - S u b k I , cA bd al-Wahhab, Tabaqat a l-S h a ficiyya al-kubra, (Eds M .Tanahl & A. a l-H u lw ), (10 Vols; Cairo, 1 9 6 4 -1 9 7 6 ) A l - S u l a m i , A b ü cA b d al-R ahm an, Tabaqat a l-S ü fiyy a , (Ed. N . Shariba), (Cairo, 1953) ------- , K ita b A d a b al-suhba, (Ed. M.J. Kister), (Jerusalem, 1954) ------- , K ita b Tabaqat a l-S ü fiyy a , (Ed. J. Pedersen), (Leiden, 1960) ------- , J a w a m ic A d a b a l-S ü fiyy a , (Ed. E. K ohlberg), (Jerusalem, 1976) T r i m i n g h a m , J.S., T h e S u fi orders in Islam , (O xford, 1971) V ajd a, G., ‘idjaza’, Encyclopaedia o f Islam (2nd Edn) V o n S c h l e g e l l , B., Principles o f S u fism , (Berkeley, 1990) W a l d m a n , M .R ., Toward a theory o f historical narrative: a case stu d y in PersoIslamicate historiography, (C olum bus O H , 1980) W a n s b r o u g h , J., T h e sectarian m ilieu: content and composition o f Islamic salvation history, (O xford, 1978) ------- , R e s Ipsa L o q u itu r : H isto ry and M im esis, (Jerusalem, 1987) Y afT I, "Abdullah, M ir ’át al-janan (4 Vols; Haidarabad, 1920)

223

Index

‘Abd al-Qädir al-Jllänl QTlT) 169, 171, 210 n. 72 Abdel-Kader, ‘A.H . 189 n.101, 193 n. 49, 194 n. 83 Abedi, M ahmud 152 ‘Ä b id (pl. ‘ubbäd, pietists) 26, 105, 1 1 8-119, 192 n. 21 Abü Bakr, Caliph 79 Abü Hamza al-Baghdädl 30 Abü Hanlfa 125, 130-131 Abü Häshim al-Süfi 7 1 -7 2 , 152-153, 168, 191191 n. 1 0 -1 1 , 197 n. 23 Abü N u ‘aym Ahmad b. ‘Abdullah al-Isfahäni 2, 39, 4 1 -4 3 , 45, 5 7 -5 8 , 61, 6 3 -6 7 , 74, 78, 116, 190 n. 6, 191 n. 8, 15, 192 n. 25, 194 n. 80, 195 n. 92, 196 n. 7 Abü Thawr 14, 18, 88, 108 -1 0 9 , 135, 163, 186 n. 34, 187 n. 62, 198 n. 62, 206 n. 54 Abü ‘Uthmän al-Maghribl 73, 174 Abü Yazld Tayfür, al-Bastämi 4, 1 8 -2 2 , 2 4 -2 8 , 33, 3 5 -3 6 , 38, 4 3 -4 9 , 54, 5 6 -5 9 , 6 1 -6 2 , 6 5 -6 7 , 7 5 -7 6 , 7 8 -7 9 , 85, 87, 89, 100, 1 0 7-113, 1 1 5-121, 134, 136-137, 140 -1 4 4 , 155, 1 6 3-165, 168, 1 7 4-175, 179, 184 n. 9 (Intro.), 187 n. 58, 65, 188 n. 93, 191 n. 12, 192 n. 28, 33, 193 n. 36, 194 n. 57, 198 n. 38, 48, 59, 199 n. 78, 201 n. 25, 202 n. 39, 53, 203 n. 76, 81, 206

224

n. 56, 207 n. 76, 208 n. 14, 209 n. 33, 42, 210 n. 83 A h l al-bayt (the family o f the Prophet) 129 A h l al-suffa (the ‘people o f the canopy’) 129, 192 n. 21, 195 n. 1 Ahrär, ‘Ubaydulläh 151, 170—171, 207 n. 3 A ltw äl (mystical states) 11, 101-102, 106, 167 Algar, H. 200 n. 9 ‘Ä lim 26 ‘Amr b. ‘Uthm än al-Makkl 205 n. 34 al-Ansäri, ‘Abdullah 2 -3 , 6 9 -7 0 , 7 5 -7 6 , 82, 84, 9 0 -9 4 , 96, 156 -1 5 7 , 166-169, 173, 188 n. 85, 195 n. 1, 196 n. 8 -9 , 209 n. 56 al-Ansäri, Abü Ayyüb 195 n. 1 al-Antäkl, ‘Abdulläh b. Khubayq 154, 187 n. 65, 71, 196 n. 15 al-Antäkl, Ahmad 191 n. 16 Aouad, M. 191 n. 34 Arberry, A.J. 100, 202 n. 49 Ä r i f 26, 35, 119 Asad, Shams al-Dln 171 al-Asamm, Hätim b. ‘Unwän 154, 191 n. 16 Ash’arite school 69, 99—101, 123, 188 n. 93, 200 n. 14 al-Aswad, Ahmad 106 ‘Atä al-Rüdhbäri, Ahmad b. 103 ‘Attär, ‘Alä’ al-Din 209—210 n. 61

Index A tb ä ‘ al-täbi‘Tn. See l a b i ‘ü ’l-tä b i‘Tn.

al-‘Attär, Abu Bakr 5 5 -5 6 ‘Attär, Fand al-Dln 183 n. 8 Auchterlonie, P. 183 n. 1 A w liy ä ’ (saints) 140

Baghdad 9, 1 4 -1 5 , 17, 43, 9 4 -9 6 , 9 9 -1 0 0 , 163 -1 6 4 , 187 n. 57, 191 n. 11 al-Baghdädl, Abü Hamza 196 n. 16 al-Baghdädl, al-Khatib 9, 99, 184 n. 3, 187 n. 61, 188 ni 73, 199 n. 1 Baidick, J. 189 n. 97, 190 n. 129, 204 n. 16 al-Balkhl, Jaläl al-Dln. See R üm l, Jaläl al-Din. al-Balkhl, M uhammad b. al-Fadl 66, 185 n. 27, 192 n. 24 al-Balkhl, Shaqlq 191 n. 16 B aqä’ (subsistence) 133 Barthold, V.V. 207 n. 7 al-Bassäml 55—56 al-Bayhaql, Ahmad 100 Biblical law 60 al-Blrünl, Abu Rayhän 187 n. 57 Bishr b. al-Härith 1 5 -1 6 Bonner, M. 16, 37, 186 n. 42, 48, 190 n. 132 Böwering, G. 184 n. 13, 200 n. 11 Brockelmann, C. 196 n. 9, 199 n. 8 Bukhara 171 Bulliet, R.W . 183 n. 2 ‘Bundles’ in the Persian Tabaqät al-Süfiyya 7 5 -7 9 , 8 4 -8 5 , 155 al-Busri, Abü ‘Ubayd 73, 103 -1 0 4 , 201 n. 2 6 -2 7 Calder, N . 17, 63, 93, 186 n. 40, 187 n. 53, 188 n. 85, 189 n. 1 1 6-117, 191 n. 15, 194 n. 87, 199 n. 70, 6, 210 n. 82 Chaumont, E. 185 n. 22 Conrad, L. 189 n. 102, 194 n. 58 Constantinople 195 n. 1 C ook, M. 192 n. 34 Cooperson, M. 37, 186 n. 45, 190 n. 132 Correspondence between Sufis 6 1 -6 2

Dabashi, H. 186 n. 29 al-Dablli, Abü Müsä 209 n. 33 al-Daqqäq, Abü ‘All 100, 106 -1 0 7 , 115, 152 -1 5 3 , 168 -1 6 9 , 172, 184 n. 12, 199 n. 7, 201 n. 19 al-Däränl 191 n. 16 Daube, D. 60, 194 n. 73 al-Daylaml, Abü ’1-Hasan 191 n. 18 D aym üm iyya (permanence) 48 D e Beaurecueil, S. de L. 196 n. 8, 198 n. 57 D e Jong, F. 184 n.8 al-Dhahabl, Shams al-Dln 6 9 -7 0 , 196 n. 5 D h ikr (remembrance o f God) 201 n. 25 D hü ’l-N ü n al-Misri 1 5 -1 6 , 46, 49, 5 8 -5 9 , 79, 84, 121, 153, 186 n. 41, 49, 190 n. 129, 198 n. 48 Dihlawl, Hasan 210 n. 74 Dlnawar 106 al-Dinawari, Abü ’l- ‘Abbäs 33—34, 197 n. 24 al-Dinawari, Mamshäd 196 n. 16 al-Duqql, Abü Bakr 201 n. 26 Egypt 186 n. 35 Epistles 2 7 -2 8 , 50, 193 n. 49 Ernst, C. 185 n. 21, 203 n. 68 Ethe, H. 204 n. 9 Fahndrich, E. 194 n. 58, 209 n. 52

F anä’ (annihilation) 133, 145 Faqr (poverty) 126, 204 n. 16

al-FärisI, ‘Abd al-Ghäfir 99, 199 n. 4 Färmadi, Abü ‘All 1 7 3 -1 7 4 Firäsa (miraculous insight) 20, 25, 140, 1 6 0-162, 184 n. 12, 206 n. 66 Frank, T. 191 n. 9 al-Fudaylb. ‘Iyäd 12, 1 5 -1 6 , 1 2 9-131, 191 n. 16 Furüzänfar, B. 184 n. 4 Gadamer, H .-G . 5, 177, 184 n. 10 (Intro.) Ghalba (rapture) 143 Ghazäll, Ahmad 1 7 3 -1 7 4 Ghazna 1 2 5 -1 2 6 , 132, 204 n. 1 Gilliot, C. 183 n. 1 Goldziher, I. 188 n. 92

225

Index Grabar, O. 192 n. 34 Grämlich, R . 199 n. 5, 200 n. 1 0 -1 1 , 13, 18, 202 n. 33, 203 n. 57, 79, 204 n. 82 al-Gurgänl, Abü T-Qäsim 131—132, 1 7 3 -1 7 4

Ibn Abi T-Hawärl, Ahmad 154 Ibn Abi YaTä, M uhammad 69, 196 n. 3 Ibn Adham, Ibrahim 15—16, 37, 103, 186 n. 41, 191 n. 16 Ibn al-A‘räbI, Abü Sa‘Id 14—15, 17, 66, 103, 185 n. 24, 192 n. 21, 195 n. 101 Ibn al-‘Arabl, M uhyl al-Dln 151, Hadith sections o f biographies 2 0 -2 1 , 58 169-171 al-Hafl, Bishr 37, 191 n. 16 Ibn al-Barql, Abü ‘Abdullah 73 al-Häfiz, Shams al-Dln Muhammad Ibn al-‘Imäd, Abü T-Faläh 191 n. 8 172 ’ Ibn al-jawzl, ‘Abd al-Rahmän 199 n. 7 Hafsi, I. 183 n. 1 Ibn al-Karanbl 80, 84, 198 n. 50 H äl (inner state) 136, 141 Ibn al-Khadrawayhi, Ahmad 191 n. 12 al-Halläj, al-Husayn b. Mansür 17, 43, Ibn al-Mulaqqin, Abü Hafs 183 n. 8 91, 104, 121 -1 2 2 , 13 1 -1 3 4 , 137, Ibn ‘Ammär, Imäm Yahyä 75 147, 178, 184 n. 12, 191 n. 18, 203 Ibn ‘Atä’, Abü T-‘Abbäs 205 n. 34 Ibn Fürak, Abü Bakr 100 n. 72, 204 n. 82, 205 n. 34, 39, 206 n. 41 Ibn Hablsh, Abü T-Husayn Halläjiyya 134 Muhammad b. ‘All 63 Ibn Hamza, Samnün 104 Halqa (study circle) 29—30, 159 al-Hamadänl, Ahnaf 208 n. 24 Ibn Hanbal, Ahmad 70, 130 al-Hamadani, £Ayn al-Qudat 169—172 Ibn Harb, Ahmad 62 Ibn Jahdam, Abü ’1-Hasan ‘All 106 Hamadänl, Yüsuf 170, 1 7 2 -1 7 3 Hamori, A. 192 n. 34 Ibn Khafif, Muhammad 31, 205 n. 33 Hanafite school 87, 125, 165, 188 n. 93 Ibn Khalllkän, Abü ’l - ‘Abbäs 199 n. 7 Ibn Kunäsa 16 Hanbalite school 41, 6 9 -7 0 Ibn M a‘dän al-Bannä’ 41, 65, 190-191 Harakät-i wajdi (ecstatic movements). See Raqs. n. 8 al-HarawI, Ibrahim 191 n. 12 Ibn Mäshädha, Abü ’1-Hasan ‘All 190-191 n. 8 Hartmann, R . 184 n. 9 (Ch. 1), 201 n. 20, 202 n. 49 Ibn M u‘ädh, Yahyä 54, 59, 62, 1 4 3 -1 4 4 , 165 Hashwiyän (anthropomorphists) 125 Hawting, G .R . 190 n. 132 Ibn Nujayd, Ismä‘Il, al-Sulaml 9 -1 0 , 15 Heffening, W. 183 n. 1 Ibn Qutayba, Abü Muhammad 16, Herat 6 9 -7 0 , 9 3 -9 4 , 96, 151, 167, 186 n. 4 6 -4 7 169, 195195 n. 1 Hijaz 100 Ibn Surayj 115 al-Hlri, Abu ‘Uthm än 9, 73, 88, 164, Ijäza (licence to transmit) 31—32, 189 n. 117, 199 n. 69 189 n. 114 Hujwlr 204 n. 1 Introductions o f biographies 1 8 -1 9 , al-HujwIri, ‘All 3, 99, 125 -1 4 7 , 162, 57, 8 5 -9 0 , 9 4 -9 6 , 1 0 7-109, 178, 200 200 n. 11, 204 n. 2, 9, 1 1 -1 3 , 135 -1 3 7 , 1 6 3 -1 6 6 Introductions o f segments 2 2 -2 4 , 16, 205 n. 17, 34, 39, 206 n. 69, 210 n. 83 2 9 -3 2 , 4 4 -4 5 , 6 1 -6 4 , 7 4 -7 5 , 9 0 - 9 3 ,1 0 9 - 1 1 0 H u M iy y a 134, 191 n. 17 Humphreys, R.S. 183 n. 3 Iräda (desire/discipleship) 120, 122 Husaynl, Amir 210 n. 74 Iraq 1 7 -1 8 , 4 3 -4 4 , 108, 186 n. 35 al-Husrl, Abü ’1-Hasan 130, 156 ‘Iraqi, Fakhr al-Dln 210 n. 74

226

Index Isfahan 41, 43, 65, 190 η. 8, 191 η. 18 al-lsfahäm, ‘All b. Sahl 41, 43, 190 n. 3, 205 n. 35 (Ishq (divine love) 128 Ismailism 69 Isräfil 197 n. 33 Ivanow, W. 70, 196 n. 10 ‘Iyäd, Fudayl 75, 77 Jähiz, N ur al-Dm 93 al-Jallä’, Abü ‘Abdullah 73, 88, 164, 201 n. 26 JämI, ‘Abd al-Rahmän 3, 1 5 1-177, 1 7 9 -1 8 0 , 183 n. 8, 190 n. 137, 206 n. 66, 207 n. 3, 5, 7, 9, 208 n. 12, 209 n n. 56, 210 n. 61 al-Jämi, Shams al-Dln al-Kusü’I 171 J ü d (liberality) 128 Julläb 204 η. 1 al-Junayd, Abü ’l-Q äsim 4, 1 4 -2 0 , 23, 2 6 -2 9 , 3 2 -3 3 , 3 5 -3 6 , 38, 4 4 -4 5 , 4 9 -5 7 , 5 9 -6 6 , 71, 74, 7 9 -8 5 , 8 7 -9 0 , 9 4 -9 6 , 100, 103, 108-109, 1 1 3 -1 1 7 , 1 2 0 -1 2 1 , 1 3 0 -1 3 1 , 133, 1 3 5 -1 4 0 , 1 4 2 -1 4 6 , 157 -1 6 4 , 1 7 4 -1 7 5 , 1 7 9 -1 8 0 , 186 n.28, 187 n. 60, 62, 188 n. 88, 189 n. 98, 101, 190 n. 124, 133, 191 n. 13, 192 n. 27, 193 n. 40, 42, 49, 194 n. 59, 84, 201 n. 19, 202 n. 36, 46, 203 n. 5 8 -5 9 , 66, 206 n. 54, 208 n. 14, 29 Junaydiyya 133, 143, 1 4 5 -1 4 6 al-Jurayri, Abü Muhammad 14, 5 5 -5 6 , 103, 130 al-Juwaynl, Abü M uhammad 100 Kaaba 1 4 1 -1 4 2 Kafarshilä’I, Ahmad b. Sulaymän 31 al-Kaläbädhl, M uhammad b. Ishäq 183 n. 4 Karämät (miracles) 117, 167, 207 n. 10 al-Karkhl, M a£rüf 191 n. 16, 201 n. 26, 28 K a s h f al-hijäb (‘uncovering o f the veil’) 127 Käshghari, Sa‘d al-Dln Muhammad 151, 170, 207 n. 3, 210 n. 61 Kätib, Abü ‘AIT 174

al-Kattani, Abu Bakr M uhammad b. ‘All, 13, 2 9 -3 0 , 3 3 -34,’ 8 1 -8 2 , 197 n. 16 Kearney, R . 211 n. 5 Khäniqäh (Sufi lodge) 71 al-Khäqänl, Afdal al-Dln 210 n. 73 Kharaqänl, Abü T-Hasan 1 7 3 -1 7 4 Kharq (rending o f garments) 128, 147 al-Kharräz, Abü Sa‘Id 4 3 -4 4 , 84, 163, 192 n. 22, 198 n. 56, 201 n. 26, 208 n. 24 al-Khashshäb al-Kablr, Abü Sahl 106 Khayr al-Nassaj 31, 75, 77, 205 n. 27 al-Khazzäz, Abü Sa‘Td. See al-Kharräz, Abü Sa‘ld. Khoury, R . 191 n. 15 al-Khuldl, Ja‘far b. Muhammad 1 4 -15, 27, 61 6 1 -6 3 , 138, 189 n. 98, 194 n. 83, 195 n. 101 Khurasan 9, 14, 17, 4 3 -4 4 , 67, 70, 100, 107107, 186 n. 35 al-Khuräsänl, Abü Hamza 196 n. 16, 205 n. 27 al-Khuttall, Abü ’1-Fadl 125, 132 -1 3 3 al-Khüzandl, Abü Tahir 106 Khwäjagän 170 Kiläb, ‘Abdullah Sa‘ld 82, 95 Kinberg, L. 189 n. 95, 190 n. 128 al-Kirmänl, Shäh 104, 201 n. 26 Kramers, J.H. 196 n. 2 al-Küfi, A b ü ja ‘far 191 n. 13 Lahore 1 2 5 -1 2 6 Lüristänl, Fakhr al-Dln 171 al-Maghribl, Mansür b. Khalaf 106

Mahabba (divine love) 128, 141—142,

144 Mahmüd, ‘Abd al-Halim 200 n. 10 Majdhüm, Abü Ja‘far 73 M ajlis (teaching session) 3 0 -3 1 , 160, 166 Makdisi, G. 187 n. 56, 189 n. 106, 117, 199 n. 66, 70 al-Makkl, Abü Talib al-Wa‘iz 183 n. 4 M aläma (blame) 127, 131, 134, 142, 147 M aläm atiyya 9, 184 n. 8, 207 n. 10 Malamud, M. 201 n. 20

227

Index al-Malim, Abu Sa‘d 106 Manual genre 2—3, 100—103, 11 7-124, 1 2 6-129, 178, 183 n. 4, 200 n. n. 11 M aqam at (spiritual stations) 1 0 1-102, 167 al-MaqdisI, Tahir 197 n. 16 M a ‘rifa (mystical knowledge) 26, 3 5 -3 6 , 96, 115, 127, 167, 190 n. 129, 207 n. 10 al-Maristanl, Abu Ishaq 193 n. 49 al-MarwazI, Fath b. Shakhraf 74, 197 n. 33 M a sh a ’ikh 2 9 -3 0 , 131, 166 Massignon, L. 186 n. 52, 190 n. 129 al-MayhanT, Abu Sa‘id 131—132 M ecca 1 3 -1 4 , 16, 29, 69, 106, 141, 186 n. 35, 195 n. 1 Medina 30, 69 Meier, F. 191 n. 18 Meisami, J.S. 183 n. 3 Melchert, C. 37, 186 n. 34, 43, 4 9 -5 0 , 190 n. n. 130, 132 M inuvi, M. 184 n. 4 al-Misri, Abu Ja‘far M u‘adh 73 Mojaddedi, J.A. 184 n. 11 (Intro.), 205 n. 16, 35, 206 n. 42, 209 n. 58, 210 n. 84 M ottahedeh, R . 183 n. 3 M ubdhiyya 191 n. 17 Muhammad, the Prophet 1 1 -1 2 , 17, 2 0 -2 1 , 65, 69, 8 8 -8 9 , 9 5 -9 6 , 10 4-105, 111, 114 -1 1 5 , 120, 129, 132, 134, 136, 138, 143, 160 -1 6 2 , 187 n. 71, 188 n. 92, 195 n. 1 Muhaqqiq, Burhan al-Dln 171 al-Muhasibl, al-Harith 1 5 -1 6 , 18 -1 9 , 43, 45, 50, 5 2 -5 3 , 6 3 -6 4 , 8 7 -8 8 , 108, 163, 187 n. 60, 193 n. 51, 55, 195 n. 89 M ujahada (religious striving) 26, 118, 136, 188 n. 91 al-Multani, Baha’ al-Dln Zakariya’, 210 n. 74 al-Muqri’, Abu T-Qasim 185 n. 24 M urad (adept Sufi) 120, 122, 203 n. 80 M uraqqa‘dt (patched cloaks) 127 M u n d (novice Sufi) 120, 122, 203 n. 80

228

al-Murta‘ish 2 9 -3 0

M ushähadät (‘mystical witnessings’) 128 M utashabbiha (anthropomorphists) 125

M u ’tazilites 125, 188 n. 93 al-Muzayyin, ‘All b. Muhammad 13

al-Nabhänl 190 n. 2 N afs (carnal soul) 112 al-Nahrajüri, Abü Ya‘qüb Ishäq b. Muhammad 13, 29—30, 73, 205 n. 35 al-Nakhshabl, Abü Turäb 191 n. 14, 201 n. n. 26 Naqshband, Muhammad b. Muhammad Bahä’ al-Dln 170, 207 n. 2, 209 n. 6 0 -6 1 Naqshbandiyya 151, 1 6 9 -1 7 3 , 175, 179, 207 n. 2 - 3 Narrative motifs 5 2 -5 6 , 7 5 -7 9 , 8 2 -8 3 , 95, 100, 1 1 5-116, 13 9 -1 4 0 , 143 -1 4 4 , 1 5 8 -1 6 2 , 179, 191 n. 13, 193 n. 53, 199 n. 7, 201 n. 25, 203 . 76 209 n. 42, 47 Nasräbädhi, Abü ’l-Qäsim 15, 18, 205 n. 33 al-Nassäj, Abü Bakr 173—174 Naw ä’i, ‘All Shlr 152, 207 n. 7 a l-N a za r f l ’l-ahdäth (gazing at youths) 128 N icholson, R eynold 126, 204 n. 9 Nihavand 18, 8 7 -8 8 , 108, 163 Nishapur 9, 75, 100, 106, 164, 199 n. 5 N otebooks 31, 6 1 -6 3 , 9 0 -9 3 , 179, 194 n. 83 N oth , A. 189 n. 102, 194 n. 58 Nurbakhsh, J. 211 n. 4 al-Nürf, Abü ’1-Hasan 206 n. 47 N üriyya 206 n. 47 Order o f biographies 1 0 -1 5 , 17 -1 8 , 4 2 -4 4 , 7 0 -7 4 , 1 0 3 -1 0 7 , 1 2 9-133, 15 2-155, 168 -1 6 9 , 1 7 2-179, 197 n. 35 Organic growth 5 9 -6 0 , 7 9 -8 5 , 9 5 -9 6 , 180, 191 n. 15 Origins o f Sufism (traditional accounts of) 1 1 -1 2 , 1 5 -1 6 , 3 7 -3 8 , 7 1 -7 2 , 105

Index Parsa, Khwaia M uhammad 2 0 9 -2 1 0 n. 61 Pärsä, Khwäja Nasr 210 n. 61 Pedersen, J. 37, 185 n. 1 4 -1 5 , 186 n. 35, 188 n. 74, 76, 80, 195 n. 92 Pilgrimage 141—142 Pouijavady, N . 184 n. 11 Prayers 24, 4 9 -5 0 Punjab 125 Qabül (acceptance) 84 Qadarites 125 al-QalänisI, Abü Ahmad 191 n. 14 al-Qassäb, Abü T-‘Abbäs 106 al-Qassäb, M uhammad b. ‘Ali 19, 80, 8 7 -8 8 , 108, 163, 187 n. 61 al-Qassär, Hamdün 12, 133 Qassäriyya 133 Qayrawan 186 n. 35 Q irä’a (recitation) 6 3 -6 4 , 199 n. 69 Quchan 199 n. 5 Q ur’an 21, 5 5 -5 6 , 101 -1 0 2 , 110, 1 13 -1 1 4 , 117, 122, 127, 151, 200 n. 19 al-Qushayri, Abü T-Qäsim 2 -3 , 9 9 -1 1 1 , 113, 1 1 5 -1 2 3 , 126, 130, 1 3 2 -1 3 3 , 142, 146, 153, 1 7 8-179, 184 n. 12, 187 n. 53, 190 n. 133, 199 n. 3, 7, 200 n. 14, 18, 201 n. 25, 28, 202 n. 36, 51, 204 n. 13

n.

al-Raqqi, Ibrahim 197 n. 16 R aqs (Sufi dance) 1 2 8 -1 2 9 Ravan Farhadi, A.G. 196 n. 8 -9 , 11, 197 n. n. 23 R epetition o f biographies 42, 74, 9 4 -9 5 , 154, 197 n. 33 R icoeur, P 149, 181, 210 n. 1 R ippin, A. 189 n. 95, 211 n. 3 Ritter, H. 196 n. 9 Robinson, C.F. 187 n. 72 R om an law 60 Rosenthal, F. 185, n. 20, 187 n. 72 al-Rüdhbäri, Abü ‘Ali 114, 174, 203 n. 58 R üm i, Jaläl al-Din 169-171 Ruwaym b. Ahmad 9 5 -9 6 , 163 Safwa (purity) 127

Sahaba (the Prophet’s companions) 11,

105, 129, 133

Sahw (sobriety) 133, 14 3 -1 4 5 , 204

η. 16

a l-Salaf al-sälih (‘the pious

predecessors’) 11—13, 16—17, 21, 42, 71, 129, 153, 173 S a m ä ‘ (musical audition) 102, 117, 120, 1 2 8 -1 2 9 , 131, 134, 142, 199 n. 6, 204 n. 13 al-Sam‘äm, Abü Sa‘d 184 n. 9 (Intro.), 10 (Ch. 1) al-Sam‘äni, ‘Abd al-Karim b. Muhammad 10 Sanä’i, Hakim 172 al-Saqati, Sari 1 5 -1 6 , 1 8 -1 9 , 43, 5 0 -5 3 , 64, 76, 80, 8 7 -8 8 , 1 35 -1 3 8 , 140, 158 -1 6 3 , 187 n. 61, 191 n. n. 13, 201 n. 28, 208 n. 12 al-Saqqä’, Walid b. ‘Abdullah 153 al-Sarräj, Abü Nasr 101, 1 1 0-111, 1 14 -1 1 7 , 120, 128, 146, 183 n. 4, 200 n. n. 11, 202 n. 48 al-Sayrafi, Abü Ί-Qäsim 106 al-Sayrafi, ‘Ali b. Bundär 201 η. 26 al-Sayyäri, Abü ’l - ‘Abbäs 197 n. 24 Schacht, J. 186 n. 34 Sells, M. 187 n. 66 Sezgin, F. 184 η. 1 al-Shäfi‘i, Muhammad b. Idris 130, 163, 186 n. 34 Shafi’ite school 10, 41, 94, 99—100, 123, 190 n. 8, 196 n. 7 al-Shaqqäni, Abü ’l- ‘Abbäs 131 -1 3 2 al-Sharif, Mahmüd 200 n. 10 Sharifl, ‘Ali 75—76 Shathiyyät (ecstatic utterances) 116 Shawq (religious yearning) 119 al-Shibli, Abü Bakr 1 4 -1 5 , 23, 2 9 -3 1 , 5 3 -5 4 , 91, 103, 130, 163, 186 n. 37 Shi’i Islam 42, 129 al-Sijistäni, Abü Hätim 110, 202 n. 48 Sirwäni, Abü ’l-Hasan 199 n. 64 Smith, G.M. 185 n. 21 al-Subki, ‘Abd al-Wahhäb 41, 184 n. 8, 190 η. 1, 191 n. 8, 199 n. 7 Sufyän al-Thawri 16, 163, 186 n. 34, 191 n. 10 al-Suhrawardi, Abü Najib 169, 171

229

Index al-Suhrawardi, Shihab al-Din 169, 171 S u kr (intoxication) 58—59, 133, 1 4 3-145, 204 n. 16, 207 n. 76 al-Sulaml, Abu ‘Abd al-Rahman 2 -3 , 9 -2 5 , 2 9 -3 2 , 3 6 -3 9 , 4 2 -4 5 , 4 9 -5 0 , 57, 62, 6 4 -6 7 , 7 0 -7 1 , 7 3 -7 4 , 78, 85, 87, 8 9 -9 0 , 9 4 -9 5 , 101, 103 -1 0 4 , 106 -1 1 1 , 113, 116, 119, 121, 129 -1 3 1 , 135 -1 3 8 , 140, 146, 152, 1 5 4-155, 162, 164, 1 6 6-167, 175, 177 -1 7 9 , 184 n. 1 -2 , 4, 1 1 -1 2 , 14, 185 n. 2 3 -2 4 , 186, n. 4 9 -5 0 , 187 n. 53, 69, 189 n. 105, 113, 190 n. 133, 137, 191 n. 1 5 -1 6 , 192 n. 21, 194 n. 65, 195 n. 94, 196 n. 1 5 -1 6 , 197 n. 20, 25, 198 n. 59, 64, 199 n. 2, 201 n. 25, 2 7 -2 8 , 202 n. 42, 46, 203 n. 57, 59, 204 n. 82, 13, 205 n. 34, 209 n. 42, 54, 56 al-Sulaml, Isma‘ll b. Nujayd. See Ibn Nujayd, Isma‘ll, al-Sulami. Sunni Islam 11—12, 69, 123, 129 al-Surayj, Abu ’l- ‘Abbas 9 4 -9 5 , 157, 208 n. 29 al-Susi, Abu Ya‘qub 73 Syria 125, 132, 164, 186 n. 35 7a ‘a (act o f obedience) 112 Tabaristan 106 7 abi ‘un (the successors o f the sahaba) 11, 105105, 129, 192 n. 21 7a b i‘u '1-tabi‘m (the successors o f the successors) 11, 13, 105, 129 Tabriz!, Shams al-Din 171 TafrTd (unicity) 11, 46, 48, 106 al-Ta Ϊ, D aw ud 103, 130, 201 n. 26 Tajnd (detachment) 11, 106 Tajnd al-tawhid 26, 189 n. 93 Tawba (repentance) 51, 128 Tawhid (unity) 11, 46, 48, 7 9 -8 4 , 106, 127, 141 -1 4 2 , 156, 207 n. 10

230

Tayfüriyya 134, 143-144, 146, 206

n. 56 Teaching contexts 2 4 -2 5 , 2 9 -3 1 , 46, 6 1 -6 4 , 9 0 -9 4 , 139, 1 6 0 -1 6 2 , 166, 179, 188 188 n. 85 al-Tusi, Muhammad b. Bakr 100 al-Tustari, Sahl b. ‘Abdullah 71, 197 n. 20, 201 n. 25 ‘U lam ä’ 2 5 -2 6 , 110, 140 ‘U nits’ in the K a sh f al-mahjüb 137 Ustuwwa 100, 199 n. 5

Vajda, G. 189 117 Variants and the historical transformation o f segments 32—36, 4 9 -5 6 , 7 5 -7 9 , 85, 109, 137-144, 158 -1 6 2 , 179, 190 n. 123 Verse 2 8 -2 9 Von Schlegell, B. 200 n. 9 Wajd (ecstasy) 81, 85 Walad, Bahä’ al-Din 171 Walad, Sultan 171 Waldman, M .R . 183 n. 3 Wansbrough, J. 183 n. 3, 210 n. 2 Watt, W.M. 195 n. 99 W ildya (sainthood) 139, 207 n. 10 W usül (mystical union) 135

al-Yafi‘T, ‘Abdullah 158-161, 163 -1 6 4 , 209 n. 50 Yafi (‘finding’) 84 al-ZaiiäjI, Abü ‘Amr 14, 2 9 -3 0 , 189 n. 105 al-Zaqqaq al-Kablr, Ahmad 10, 201 n. 2 6 -2 7 al-Zayyät, Abü Ya‘qüb 191 n. 1 3 -1 4 Z u h d (asceticism) & zu h h ä d (ascetics) 26, 82, 105, 110, 113, 119