(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Bukhari.Muslim"

The Canonization of 

Ual-BukharT and L 
^ Muslim 

The Formation and Function of the 
Sunm Hadrth Canon 



^ 



JONATHAN BROWN 



The Canonization of al-Bukhan and Muslim 



Islam History 
and Civilization 

Studies and Texts 



Edited by 

Wadad Kadi 

Rotraud Wielandt 



VOLUME 69 



The Canonization of 
al-Bukharl and Muslim 



The Formation and Function of 
the SunnI Hadlth Canon 



By 

Jonathan Brown 






■/ s 



BRILL 



LEIDEN • BOSTON 

2007 



This book is printed on acid-free paper. 



ISSN 0929-2403 

ISBN 978 90 04 15839 9 

© Copyright 2007 by Koninklijke Brill N^ Leiden, The Netherlands. 
Koninklijke BrUl NV incorporates the imprints Brill, Hotel Publishing, 
IDC Publishers, Martinus Nijhoff Publishers and VSP 

All rights reserved. No part of this publication may be reproduced, translated, stored in 
a retrieval system, or transmitted in any form or by any means, electronic, mechanical, 
photocopying, recording or otherwise, without prior written permission from the publisher. 

Authorization to photocopy items for internal or personal use is granted by Koninklijke BriU NV 
provided that the appropriate fees are paid directly to The Copyright Clearance Center, 
222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. 
Fees are subject to change. 

PRINTED IN THE NETHERLANDS 



And they made their camp near [the tents of] May sun... 
Where the sun forever rose first over the dry land... 

-Jundub b. Su'ud al-Asmari 

To Maisoon, who inspired me to seek knowledge 



CONTENTS 



List of Chart xv 

Acknowledgments xvii 

Dates and Abbreviations xix 

Preface xxi 



PART ONE 

I. Introduction 3 

1.1. Introduction 3 

1.2. Thesis 5 

1.3. Scholarship on the Sahihayn and the Hadlth 

Canon 8 

1.4. Addressing the Sahihayn as a Canon 15 

1.5. Note on the Sources and Approaches of this 

Study 15 

1.6. Problems in Approaches 18 

II. The Study of Canons and Canonization 20 

ILL Introduction 20 

11. 2. Canons in Context and the Emergence of Canon 

Studies 21 

11. 3. Canon Studies and the Islamic Tradition 31 

11. 4. Theoretical Tools and Common Historical 
Processes: Canon Studies and the Hadlth 

Canon 38 

II. 4. a. Canons and Community 39 

II. 4. b. Kandn and the Measure of Revealed 

Truth 41 

II. 4. c. The Principle of Charity and 

Canonical Culture 42 

11. 5. Conclusion 46 



Vlll CONTENTS 

III. The Genesis of al-Bukhari and Muslim 47 

111.1. Introduction 47 

111. 2. The Development of Hadith Literature 47 

III. 3. The Sahih Movement and the Bifurcation of the 

Hadith Tradition 54 

111. 4. The Continuity of the Living Isndd 60 

111. 5. Reality: The Life and Works of al-Bukhari and 
Muslim 64 

111. 6. Reality: Al-Bukhari, Sahib al-Sahih 65 

III. 6. a. The Sahih 69 

III.6.b. Legal Identity and Method 71 

III.6.C. Al-BukharT and the Controversy over the 

Created Wording of the Qur'an 74 

111. 7. Reality: Muslim, the Junior Partner 81 

III. 7. a. Muslim's Methodology in his Sahih 82 

111. 8. Perception: Al-Bukharl, Muslim and the 

Greatest Generation 86 

111. 9. Reception: The Immediate Response to 

al-Bukhari's and Muslim's Works 90 

111.10. Conclusion 98 

IV A "Period of Intense Canonical Process": Imagination and 
the Study of the Scihihciyn in the Long Fourth/Tenth 

Century 99 

IVl. Introduction 99 

IV2. i:\itMustakhrajGcvLre 104 

IV 3. Mustakhraj: The Sahihciyn as Formative Texts 106 

IV 3. a. Al-Isma'ill: Rationalist Muhaddith 109 

IV3.b. Abu Nu'aym al-lsbahani and 

Shiite-Sunni Polemic Ill 

IV. 3. c. Abu 'Awana and an Independent Legal 

Path 113 

IV4. 'Hal and Ilzdmdt: Interaction with the Standards 

of al-Bukharl and Muslim 115 

IV5. Required Study: Clarifying an Unclear Subject .... 120 
IV6. Regional and Temporal Distribution of the 

Sahihciyn Network 124 

IV6.a. Naysabur: The Hometown Cult of 

Muslim 124 



CONTENTS IX 

IV6.b. Jurjan: A Cult of al-Bukhari Among 

Friends 128 

IV6.C. Baghdad: Inheriting the Study of the 

Sahihayn Among the Baghdad Knot 131 

IV.G.d. Other: Isfahan and Central Asia 134 

IVe.e. An End to Regional Cults After 370AH 135 

IV7. The Sahihayn Network: A Shafi'i Enterprise 135 

IV8. Intense Canonical Process: Imagining a New 

Epistemological Status for Hadith Books 144 

IV9. Whythe&Majn? 149 

IVIO. Conclusion: The Eve of Canonization 151 

V Canon and Community: al-Hakim al-Naysaburi and the 

Canonization of the SaMhiyn 154 

VI. Introduction 154 

V2. The Life and Works of al-Hakim al-Naysaburi 155 

V3. Al-Bukhari and Muslim in al-Hakim's Vision of 

Hadith 160 

V4. The Shurut According to al-Hakim: The Requirements 

of al-Bukharl and Muslim 162 

V4.a. Two i?«zfzs and the Elimination of J«M/« 163 

V4.b. Doubling Transmission: 1 ^- 2 ^^ 4 166 

V4.C. A Standard for Authenticity and a 

Standard for the Sahihayn 168 

V5. Admitted Exceptions: al-Mustadrak and the Standards 

of the Shaykhayn as Ideal Rather Than Reality 170 

V6. Al-Hakim's Politics: The Expansion of the 

Authentic Umbrella 172 

V7. Al-Hakim's AfM^te(iz(Z and the Ten Thousand 175 

V8. Al-Hakim's Target Audience: The Mu'tazUites and 

their Criteria for Authentic Hadiths 178 

V9. The Mustadrak as a Common Measure of 

Authenticity 181 

VIO. The Discourse of Legal Theory: The Consensus 

of the Umma on Hadith 183 

VlO.a. The Hanafis 184 

VlO.b. The Later Mutazilites 187 

VlO.c. The Shafi'l/Ash'ari Orthodoxy 188 



CONTENTS 

VlO.d. The Hanbali Orthodoxy: Abu Ya'la 

Ibn al-Farra' 191 

VlO.e. The Malikis 193 

VlO.f. Al-Hakim and the Consensus of the 

Umma 193 

VILA New Common Ground between the Hanbali/ 

Uber-Sunnis and the Shafi'l/Ash'arl Schools 194 

V12. An Articulate Uber-Sunni: Abu Nasr al-Wa'ill 196 

VI 3. Imam al-Haramayn al-Juwaynl: A Consummate 

Shali'i and Ash'ari 200 

V14. The Sahihayn Canon: The Authority of 

Convention and Common Ground 201 

V15. Conclusion: Why the Sahihayn Now? 205 



PART TWO 

VI. The Canon and the Needs of the Community: The 

Sahihayn as Measure of Authenticity, Authoritative Reference 

and Exemplum 209 

VI.l. Introduction 209 

VL2. 1. The Need for a Common Measure of 

Authenticity: The SihWcyn in Scholarly Debate 210 

VL2. Takhrij. Applying the Measure of Authenticity 211 

VL3. The Origins of Takhrij Among the Students of 

al-Hakim al-Naysaburl 217 

VL4. The Historical Application of Takhrij 222 

VI.4.a. Polemics and Debate 222 

VI. 4. b. Bolstering Formative Texts 229 

VI. 5. 1. Misuse of the Sahihayn Canon 239 

VI. 5. 2. The Need for an Authoritative Reference: 

The Sahihayn and Non-Hadith Specialists 240 

VI. 6. 3. The Need for an Exemplum: Aristotle's Poetics 

and the Canon that Sets the Rule 247 

VI. 7. The Limits of the Canon's Authority: The Dialogic 

Power of the SaMhcyn 251 

VL8. Conclusion 260 



CONTENTS XI 

VII. The Principle of Charity and the Creation of Canonical 

Culture 262 

VII.l. Introduction 262 

VII. 2. The Beginnings of Canonical Culture: Between 

390-460/1000-1070 264 

VII. 3. The Character of the Canonical Culture: 
Al-Khatib al-Baghdadi and Defining the 

Personas of al-Bukhari and Muslim 267 

VII. 4. Charity and the Maintenance of Canonical 

Culture 275 

VII. 4. a. Reinventing the Etiology: Charity and 

Legitimizing al-Bukhari's Sahih 276 

VII.4.b. Charity and Maintaining the 
Superiority of al-Bukharl over 

Muslim 278 

VII.4.C. Charity and Muslim's Meeting with 

Abu Zur'a al-Razi 280 

VII. 5. Reconciling the Canon with Convention: the 

Sahihayn and the Rules of Hadith 282 

Vll.S.a. Charity and Tadlis 283 

Vll.S.b. Charity and Transmitters 286 

VII. 6. Rebutting Earlier Criticisms 291 

VII.7. Conclusion 298 

VIII. The Canon and Criticism: Iconoclasm and Rejection of 
Canonical Culture from Ibn al-Salah to the Modern 

Salaft Movement 300 

VIII.l. Introduction 300 

VIII. 2. Rejection of the Canonical Culture: Criticism 

after Ibn al-Salah 301 

VIII. 3. Iconoclasm and Institutional Security in 

Islamic Civilization: The Salafi Tradition 304 

VIII. 3. a. Revival and Reform in the Early 

Modern and Modern Periods 305 

VIII. 3. b. Traditionalist Salafis in the 

Middle East 309 

VIII. 4. Muhammad b. Isma'll al-San'ani: A Yemeni 

Salaft 314 

VIII. 5. Shah Wall Allah and the First Condemnation 

of Criticizing the Canon 318 



XU CONTENTS 

VIIL6. Muhammad Nasir al-Din al-AlbanI: Iconoclast 

Extraordinaire 321 

VIIL7. Against the Canon: Al-Albani's Criticism of 

the Sahihayn and His Detractors 325 

VIII. 8. Conclusion: Al-Albani's Reply and the 

Continuity of Iconoclastic Hadlth 

Criticism 331 

IX. Canon and Synecdoche: The Sahihayn in Narrative and 

Ritual 335 

IX. 1. Introduction 335 

IX. 2. Delimiting the Infinite: Managing the Sunna 

through the Hadlth Canon 336 

IX. 3. Synecdoche in Ritual: Usage of the Sahihayn 

Canon in Ritual Contexts 338 

IX. 3. a. Supplicatory and Medicinal Rituals 340 

IX.3.b. Calendrical Rituals 342 

IX.3.C. Political Rituals 344 

IX. 4. The Ritual Power of the Sahihayn: The 

Muhammadan Blessing 346 

IX.5. The Canon and Synecdoche in Narrative: 

A Salvational Trope in a Narrative of Decline 

and Salvation 349 

IX.5. a. Kh"aje 'Abdallah al-Ansarl and the 

Beginning of Synecdoche in Narrative 352 

IX.5.b. Al-Ghazali's Return to the Straight 

Path: The Sahihayn as Synecdoche 354 

IX.5.C. Al-Dhahabi's Narrative of Islamic 

History: The Sahihayn as Synecdoche 356 

IX. 6. Conclusion 358 

X. Conclusion 360 

X.l. Why the SaMhayn and Not Other Books? 360 

X.2. What Forces Led to the Canonization of the 

Sahihayn? 362 

X.3. Why Did the Canon Form at the Beginning of 

the Fifth/Eleventh Century? 367 

X.4. Did the Canon Emerge from Ferment and 

Strife? 371 



CONTENTS Xlll 

X.5. Was the Canon a Response to Shiism or the 

Product of the Seljuq State? 372 

X.6. Was the Sahihayn Canon the Product of or 

Limited to a Specific Region? 374 

X.7. Conclusion 378 

Appendices 

Appendix I: References for the Sahihayn Network Chart 379 

Appendix II: The Question of the Attribution of the 

SahThoyn 384 

Select Bibliography 387 

Index 411 



LIST OF CHART 



Chart 1.1 Sahthayn Network Chart 103 



ACKNOWLEDGMENTS 



Acknowledgements for this book must begin with Dr. Wadad Kadi of 
the University of Chicago, who served as an excellent teacher, editor and 
role model throughout my graduate career there. Drs. Fred Donner 
and Tahera Qutbuddin also served generously as wonderful professors 
and helpful advisors on this book. I must also thank John VoU, John 
Esposito, John Woods, Heshmat Moayyad, Donald Whitcomb, Cornell 
Fleischer, Gene Gragg, Holly Shissler, Maysam al-Faruqi and Haifaa 
Khalafallah for assisting me in developing this project. Dr Menachem 
Brinker in particular played an enormous role in helping me construct 
the book's theoretical framework, and I am indebted to him for his 
support. My friend Dr Scott Lucas also provided invaluable assistance 
with his rigorous and positive criticism. I must also thank profusely my 
family, in particular my mother, Dr Ellen Brown, for showing me the 
joys of learning and unhesitatingly supporting my interests throughout 
life. My friends in Hyde Park and Washington DC also deserve my 
sincere thanks. 

I am hugely indebted to the financial generosity of the Mellon 
Foundation, the Council for American Overseas Research Centers, 
the American Institute for Iranian Studies, and the Center for Arabic 
Study Abroad. 

I must also acknowledge the indispensable assistance granted by the 
Library of Congress Middle East and North Africa Reading Room; the 
American Research Institute in Turkey for its hospitality; the Khizana al- 
'Amma in Rabat; the Maktabat al-Asad in Damascus; the Siileymaniye 
Library, the Topkapi Sarayi Library and the Istanbul University Rare 
Books Library for allowing me continuous access to their unparalleled 
manuscript collections; and Drs. Gozashte and Pakechi at the Greater 
Islamic Encyclopedia [Da'erat al-ma'aref-e bozorg-e eslami) in Tehran, for 
their valuable assistance. Of course, this book would not exist if not 
for the University of Chicago, its singular Department of Near East- 
ern Languages and Civilizations, and the great Regenstein Library. 
I must also thank Shaykh Osama al-Syed Mahmoud al-Azhari and 
'Imad al-Din 'Abbas Sa'ld in Cairo and Muhammad Mujir al-Khatib 
in Damascus for their patient assistance. 



XVlll ACKNOWLEDGMENTS 

Finally, I must acknowledge the honor of working in the shadow of 
two great minds, Muhammad b. Isma'il al-Bukharl and Muslim b. al- 
Hajjaj, as well as the inimitable generations of scholars who preceded 
and followed them in elaborating the Islamic scholarly tradition. As the 
Seljuq vizier Nizam al-Mulk said: "Indeed I know that I am not worthy 
of this, but I wish to tie myself to the train of those who transmit the 
hadiths of God's Messenger, may the peace and blessings of God be 
upon him." 



DATES AND ABBREVIATIONS 



Dates in this book will follow the Hijii/ Common Era format for all 
dates through the eleventh/ seventeenth century. After that, Hyn dates, 
are of little use, and only CE dates will be provided. 

The phrase "may the peace and blessings of God be upon him (salla 
Allah 'alqyhi wa sallam)," which usually follows the Prophet Muhammad's 
name in Muslim sources wUl be abbreviated as (s). The phrase "may 
God be pleased with him/her/ them [radiya Allah 'an.. .)," which usually 
follows the names of Companions, will be represented with (r). 



PREFACE 



In the most immediate sense, this book consists of a revised version of 
a dissertation submitted to the Department of Near Eastern Languages 
and Civilizations of the University of Chicago under the supervision 
of Dr Wadad Kadi. As a project, however, it represents an attempt to 
answer a question that perplexed me for many years before I ever sat 
down to begin dissertation research: in the history of Sunni Islam, why 
are the Sahihayn of al-Bukhari and Muslim so special, what is their true 
station, and how did they achieve this status? To rephrase this question 
more broadly, what are the origins, nature and applications of authority 
in the Sunni hadith tradition? 

In the West, the study of the Sunni hadith tradition has focused 
mainly on the Authenticity Question' — to what extent does the hadith 
corpus provide a historically reliable documentation of early Islamic 
political, doctrinal and legal history. In its scope (but not in its sources), 
the investigation of the Authenticity Question stops in the early and 
mid third/ninth century with the appearance of extant documentary 
evidence in the form of historical and legal works like the Muwatta ' of 
Malik and the Sahihayn. 

This book is not about the Authenticity Question. It is about the 
Sunni hadith tradition and its role in Islamic civilization after the 
Authenticity Question fades from view. Whether or not the Sahihayn 
or any collection of hadith truly communicate the original teachings 
of Islam across the gulf of time separating us from Muhammad is 
ultimately beyond the ken of historians. It will remain a question 
hobbled as much by the exigencies of faith as a paucity of sources. 
How the hadith tradition reflects, facilitates and informs the choices that 
the Sunni community has made in the thousand some years since its 
emergence lies more squarely within the historian's purview: the study 
of continuity and change in a human tradition. It is my hope that this 
book will assist any reader interested in engaging this topic. 

Tackling the origins, development and function of the Scihihayn 
canon — the two most famous books in Sunni Islam after the Qur'an — 
required casting a very wide net across the diverse and preposterously 
rich historical landscape of Islamic civilization. In order to produce a 
study of any thematic consistency and manageable size, I have almost 



XXll PREFACE 

certainly done great injustice to many genres of Islamicate intellectual, 
literary or religious history I can only hope that this study is worthy 
of correction. 

Finally, this is book is not a criticism of al-Bukhari and Muslim or 
their collections. The genius, rigor and dedication of those two scholars 
stand beyond my reach and abilities. To fully appreciate the Sahihayn 
within the context of the collection and criticism of hadlths is to move 
beyond a common first impression of the hadith tradition — that of an 
erratic and ultimately contrived game of religious telephone — to grasp 
the simple logic and eerie internal consistency of a widely scattered 
but uniformly dedicated community of scholars who, over the past 
1 ,400 years, have repeatedly demonstrated that what we historians have 
deemed the limits of the possible for human memory and attention to 
detail simply need to be rethought. 



PART ONE 



CHAPTER ONE 
INTRODUCTION 

Introduction 

In 465/1072-3, the grand vizier of the Seljuq empire, a statesman so 
spectacularly powerful that he was hailed as Nizam al-Mulk (The Order 
of the Realm), heard of a scholar who possessed a particularly authorita- 
tive copy of the most famous collection of traditions (haditK) related from 
the Prophet Muhammad: the Sahih of al-Bukharl (d. 256/870). Nizam 
al-Mulk ordered this scholar brought to his newly founded college in 
the Iranian city of Naysabur, where the vizier gathered the children 
of the city's judges, scholars and other notables to hear a reading of 
al-Bukhari's Sahih. ' Why did Nizam al-Mulk order such a promulgation 
of the Sahih, and why did he convene the next generation of the Sunni 
Muslim elite in attendance? 

Nizam al-Mulk stood at the intersection of the great forces of 
Islamic religious history at a time when Sunni Islam was coalescing in 
its institutional form. While serving the Seljuq sultans, who were gener- 
ously endowing educational institutions for the Hanafi school of law, 
he established his Nizamiyya college network in the principal cities of 
the empire for the use of the rival Shafi'l school. Yet Nizam al-Mulk 
also held hadith study circles that glorified the 'partisans of hadith 
{ashdb al-hadith)' closely associated with the contending Hanbali school.^ 



' Abu al-Hasan 'Abd al-Ghafir al-FarisI (d. 529/1134-5), selections made by Abu 
Ishaq Ibrahim al-Sarlflnl (d. 64 1 / 1 243-4), Tdrikh Naysabur al-Muntakhab min al-Siydq, ed. 
Mohammad Kazem al-Hamudl (Qpm: Jama'at al-Modarresm, 1403/1983), 65. 

^ Ibn al-jawzl evidendy had seen the founding charter of the Baghdad Nizamiyya; 
Abu al-Faraj 'Abd al-Rahman Ibn al-JawzI (d. 597/1200), al-Muntaiam fi tdnkh al- 
umam wa al-muluk, ed. Muhammad 'Abd al-Qadir 'Ata and Mustafa 'Abd al-Qadir 
'Ata, 19 vols. (Beirut: Dar al-Kutub al-'Ilmiyya, 1412/1992); 16:190-1, 304; 17:32; 
see also 'Abd al-Hadi Rida, "Amali Nizam al-Mulk al-wazir al-salJQql ft al-hadlth," 
Majallat Ma'had al-Makhutdt al-'Arabijya 5, no. 2 (1959): 355. From the material of his 
transmission sessions, it is clear that Nizam al-Mulk made a special effort to hear 
hadlths that were shibboleths of Sunnism as opposed to Mu'tazilism, such as reports 
affirming that the believers will see God on the Day of Judgment; Rida, "Amall;" 
356, 366. See also Richard W. BuUiet, "The Political-Religious History of Nishapur 



4- CHAPTER ONE 

These policies unfolded in the threatening shadow of the Sunni Seljuqs' 
principal rival, the Isma'ili Shiites, whose assassins would eventually 
bring Nizam al-Mulk's career to an end. 

In this divided milieu, Nizam al-Mulk sought to foster a common 
ground of Sunni Islam. In 469/1076-77, when the leading Shafi'i 
scholar of Baghdad tried to win Nizam al-Mulk's support in a bitter 
debate with Hanbali rivals, the vizier sent him a missive refusing to 
intervene on his behalf "We believe in bolstering the Sunni ways [al- 
sunan), not buUding up communal strife (al-fitan)," he explained. "We 
undertook the building of this [Nizamiyya] college in order to support 
and protect the people of knowledge and the welfare of the community, 
not to create divisions amongst Muslims {tqfnq al-kalima)."^ 

By gathering the children of the empire's scholarly and administra- 
tive elite around a reading of al-Bukhari's Sahih, Nizam al-Mulk was 
reinforcing a sense of Sunni communalism. As we shall see, by the 
vizier's time scholars from most of the disputing legal and theological 
schools that would comprise the Sunni fold had together deemed the 
Sahihayn, the two 'Authentic' hadlth collections of al-Bukhari and his 
student Muslim b. al-Hajjaj (d. 261/875), authoritative representations 
of the Prophet's legacy. By convening this reading, Nizam al-Mulk was 
inculcating al-Bukharl's book as a touchstone of Sunni identity in the 
impressionable young minds of the next generation. 

The canonization of al-Bukharl and Muslim thus forms part of 
the greater drama of the formation of Sunni Islam. Nizam al-Mulk's 
fifth/eleventh-century world brought together all the leading characters 
in this saga. Among them were the textualist Hanballs and the more 
rationalist Shafi'is, both heirs to the heritage of 'the partisans of hadlth' 
but divided over the role of speculative theology in Islam. We also find 
the Hanafis, rooted in their own distinct, hadlth-wary hermeneutic 
tradition. These groups composed competing 'orthodoxies,' each inde- 
pendent and self-righteously justified. The canonization of al-Bukhari 
and Muslim is the story of how these and other disjointed segments 
of what became the Sunni community forged a common language for 
addressing the shared heritage of the Prophet's legacy (sunna). 



in the Eleventh Century," in Islamic Civilization 950-1150, ed. D.S. Richards (Oxford: 
Cassirer, 1973), 85 ff. 

' Ibn al-jawzl, al-Muntaiam, 16:190-1. 



INTRODUCTION 5 

This drama began in the classical period, but it has continued into 
modern times. Indeed, the questions that arise in a study of the forma- 
tion, function and status of the Sahihayn canon reflect tensions between 
the competing schools of thought within today's Sunni community Why 
does a modern Hanafi scholar from India seeking to defend his school 
against Salafi critics prominently cite a hadlth from SaMi al-Bukhdn on 
the cover of his book?* Why does a Salafi scholar insist on his right 
to criticize al-Bukhari's and Muslim's collections, while his opponents 
vociferously condemn him for "violating the integrity of these mother- 
books"?'^ These questions, which fuel fierce debates in Muslim discourse 
today, descend from the centuries of historical development that forged 
and maintained the canon of al-Bukhari and Muslim. 

After the Qur'an, the Sahihayn are the two most venerated books in 
Sunni Islam. Yet until now no one has explained this undeniable real- 
ity. This study examines the canonization of al-Bukhari and Muslim 
in order to discover how, when and why the two Sahih^ attained their 
authoritative station. It explores the nature of this authority, the ten- 
sions surrounding it, and the roles that the Sahihayn canon has played 
in Islamic civilization. 



Thesis 

Canons form at the nexus of text, authority and communal identifica- 
tion. Their formation, however, is neither a random nor an inevitable 
process. Canonization involves a community's act of authorizing specific 
books in order to meet certain needs. It entails the transformation of 
texts, through use, study, and appreciation, from nondescript tomes into 
powerful symbols of divine, legal or artistic authority for a particular 
audience. In their own time, al-Bukharl and Muslim were accomplished 
representatives of the transmission-based tradition of Islamic law. Like 
their teacher, Ahmad Ibn Hanbal (d. 241/855), they saw collecting 
and acting on the reports of the early Muslim community as the only 
legitimate means by which believers could ascertain God's will and live 
according to it. Yet they were only two of many such scholars, with 



^ Abdur-Rahman Ibn Yusuf, Fiqh al~Imam: Key Proofs in Hanafi Fiq/i, 2nd ed. (Santa 
Barbara: White Thread Press, 2003), cover 

^ See www.sunnah.org/history/Innovators/al_albani.htm, last accessed 5/31/04. 



b CHAPTER ONE 

al-Bukhari's career in particular marred by scandal. For over two cen- 
turies after al-Bukhari's and Muslim's deaths, the study and collection 
of hadiths continued unabated. Al-Bukhari's and Muslim's remarkable 
contribution came with their decision to compile books devoted only to 
hadiths they considered authentic (sahili). This act broke stridently with 
the practices of the transmission-based school and thus met with sig- 
nificant disapproval in the immediate wake of the authors' careers. 

In the fourth/tenth century, however, the initial controversy surround- 
ing the Sahihayn and their authors dissipated as a relatively small and 
focused network of scholars from the moderate Shafi'i tradition began 
appreciating the books' utility. These scholars found the Sahihayn ideal 
vehicles for articulating their relationship to the Prophet's normative 
legacy as well as standards against which to measure the strength of 
their own hadlth collections. Employing the Sahihayn for these purposes 
required intimate familiarity with the two books and thus spurred an 
intensive study of the works and their authors' methodologies. Simul- 
taneously, between the end of the third/ninth and the middle of the 
fifth /eleventh century, the broader Muslim community began imagining 
a new level of authority for Prophetic traditions. Scholars representing 
a wide range of opinion started to conceive of certain hadiths and 
hadlth collections as providing loci of consensus amid the burgeoning 
diversity of Islamic thought. 

One scholar in particular inherited the body of scholarship on the 
Sahihayn and harnessed the two works as a new measure of authenticity 
for evaluating reports attributed to the Prophet. Al-Hakim al-Naysaburi 
(d. 405/1014) recognized that the Scikihcyn possessed tremendous polemical 
value as common measures of hadlth authenticity that met the require- 
ments of both the transmission-based scholars whom he championed 
and the Mu'tazilites whom he bitterly opposed. He thus conceived of 
the criteria that al-Bukhari and Muslim had used in compiling their 
works as a standard he claimed authorized a vast new body of hadiths 
binding on both parties. A cadre of his students, hailing from the rival 
Hanball and Shafi'l strains of the transmission-based school, agreed 
on the Sahihayn as a commonly accepted tract of the Prophetic past. 
Drawing on developments in legal theory shared by all the major 
non-Shiite schools of the fifth /eleventh century, they declared that 
the community's alleged consensus on the reliability of the Sahihayn 
guaranteed the absolute certainty of their contents. 

This ability of al-Bukharl's and Muslim's collections to serve as an 
acknowledged convention for discussing the Prophet's authenticated 



INTRODUCTION / 

legacy would serve three important needs in the Sunni scholarly culture 
of the fifth /eleventh century As the division between different schools 
of theology and law became more defined, scholars from the compet- 
ing Shafi'i, Hanbali and Maliki schools quickly began employing the 
Sahihayn as a measure of authenticity in debates and polemics. By the 
early eighth /fourteenth century, even the hadlth-wary Hanafi school 
could not avoid adopting this convention. With the increased division 
of labor between jurists and hadith scholars in the mid-fifth/ eleventh 
century, the Sahihayn also became an indispensable authoritative refer- 
ence for jurists who lacked expertise in hadith evaluation. Finally, al- 
Bukharl's and Muslim's works served as standards of excellence that 
shaped the science of hadith criticism as scholars from the fifth/ eleventh 
to the seventh/thirteenth century sought to systematize the study of 
the Prophet's word. 

The authority of the canon as a measure of authenticity, however, 
was an illusion conjured up in the dialogic space of debate and exposi- 
tion. It vanished outside such interactive arenas. Scholars directed the 
compelling authority of the Sahihayn only against others, and within the 
closed doors of one school of law or theology, they had no compunction 
about ignoring or criticizing reports from either collection. 

Although occasional criticism of the So^h^hoyn continued even after 
their canonization at the dawn of the fifth /eleventh century, advocates 
of institutional Sunnism found it essential to protect the two works 
and the important roles they played. Beginning at the turn of the 
fourth/tenth century and climaxing in the mid-seventh/thirteenth, a 
set of predominately Shafi'l scholars created a canonical culture around 
the Scihihcyn that recast the two books' pre-canonical pasts as well as 
those of their authors according to the exigent contours of the canon. 
The canonical culture of the Sahihayn also had to reconcile instances 
in which al-Bukharl's and Muslim's methods had fallen short of what 
had emerged as the common requirements of Sunni hadith criticism 
in the centuries after their deaths. 

While most influential participants in the Sunni tradition accepted 
the canonical culture of the Sahihayn, some hadith scholars refused to 
safeguard the canon at the expense of the critical standards of hadith 
study. The tension between the majority's commitment to the institu- 
tional security of the Sahihayn and this iconoclastic strain came to a 
head with the emergence of the modern hadlth-based Salafi movement 
in the eighteenth century. In a conflict that reflects the anxieties of 
redefining Islam in the modern world, the impermissibility of criticizing 



8 CHAPTER ONE 

the Sahihayn has become a rallying cry for those devoted to defending 
the classical institutions of Islamic civilization against the iconoclastic 
Salafi call to revive the primordial greatness of Islam through the 
hadith tradition. 

Beyond the SahihayrC% roles as a measure of authenticity, an authori- 
tative reference and exemplum among Sunni scholars, the canon has 
played an important role in a variety of ritual domains and broader 
historical narratives about Islamic civilization. Here the Sahihayn have 
become a synecdochic representation of the Prophet himself, essential- 
izing his role as a liminal figure and medium of blessing. The two works 
have also come to serve as a literary trope, symbolizing the Prophet's 
unadulterated teachings in the Sunni tradition's self-perception. 



Scholarship on the Sahihayn and the Hadith Canon 

Western scholars have regularly spoken of 'canonical' hadith collections 
in Islamic civilization.'' This recognition follows the Muslim sources 
themselves, which refer to this canon in a myriad of ways, such as 
'the relied-upon books [al-kutub al-mu'tamad 'alayhd),' 'the Four Books,' 
'the Five Books,' 'the Six Books,' and finally 'the Authentic Collections 



'' For examples, see G.E. von Grunebaum, Classical Islam: A History 600—1258 (London: 
George Allen & Unwin, 1970), 95; Marshall Hodgson, The Venture of Islam (Chicago: 
University of Chicago Press, 1974), 1:332; Norman Calder, Studies in Early Muslim 
Jurisprudence (Oxford: Clarendon Press, 1993), 189; Richard W. BuUiet, Islam: The View 
from the Edge (New York: Columbia University Press, 1994), 19; Uri Rubin, The Eye of the 
Beholder: The Life of Muhammad as Viewed by the Early Muslims (Princeton: Darwin Press, 
1995), 224; Josef van Ess, TheologLC und Gesellschaft im 2. und 3. Jahrhundert Hidschra, 6 
vols. (New York: Walter de Gruyter, 1997), 1:62; Christopher S. Taylor, In the Vicinity of 
the Righteous: ^iyara and the Veneration of Muslim Saints in Late Medieval Egypt (Leiden: Brill, 
1 999), 191; Daphna Ephrat, A Learned Society in a Period of Transition: The Sunni 'Ulama ' of 
Eleventh-Century Baghdad (Albany: State University of New York Press, 2000), 40; Shahab 
Ahmad, "Mapping the World of a Scholar in Sixth/twelfth Century Bukhara: Regional 
Tradition in Medieval Islamic Scholarship as Reflected in a Bibliography," JoKraa/ of the 
American Oriental Society, 120, no. 1 (2000): 25; G.H.A. JuynboU, "Salilh" Encyclopaedia of 
Islam CD-ROM Edition v. 1 .0, henceforth £/''; Jonathan Berkey, The Eormation of Islam: 
Religion and Society in the Near East 600-1800 (New York: Cambridge University Press, 
2003), 116; Sabine Schmidtke, "The ijdza from 'Abd Allah b. Salih al-Samahljl to Nasir 
al-jarodi al-Qatlfl: A Source for the Twelver Shi'i Scholarly Tradition of Bahrayn," 
in Culture and Memory in Medieval Islam: Essays in Honour of Wilferd Madelung, ed. Farhad 
Daftary and Josef W Meri (London: LB. Tauris, 2003), 73; Natana J. DeLong Bas, 
Wahhabi Islam (Oxford: Oxford University Press, 2004), 46; Harald Motzki, "Dating 
Muslim Traditions: a Survey," Arabica 52, no. 2 (2005): 206. 



INTRODUCTION 9 

).' We can discern three strata of the Sunni hadith canon. The 
perennial core has been the Sahihayn. Beyond these two foundational 
classics, some fourth/tenth-century scholars refer to a four-book selec- 
tion that adds the two Sunans of Abu Dawud (d. 275/889) and al-Nasa'l 
(d. 303/915). The Five Book canon, which is first noted in the sixth/ 
twelfth century, incorporates the Jami' of al-Tirmidhi (d. 279/892). 
Finally the Six Book canon, which hails from the same period, adds 
either the Sunan of Ibn Majah (d. 273/887), the Sunan of al-Daraqutni 
(d. 385/995) or the Muwatta' of Malik b. Anas (d. 179/796). Later 
hadith compendia often included other collections as well.' None 
of these books, however, has enjoyed the esteem of al-Bukhari's and 
Muslim's works. 

A study tackling the entirety of the Sunni hadith canon would 
require many more volumes than the present project allows. Because 



' Sa'ld b. al-Sakan of Egypt (d, 353/964) and Ibn Manda of Isfahan (d. 395/1004-5) 
mention the four foundational books of al-Bukharl, Muslim, Abu Dawtld and al-Nasa'l 
(see Chapter 4 ns. 175 and 176). Although he did not denote them as a unit, the 
fifth /eleventh-century Shafi'l scholar Abti Bakr al-BayhaqI (d. 458/1066) stated that 
the six collections of al-Bukharl, Muslim, Abu Dawtld, al-Nasa'l, al-Tirmidhi and Ibn 
Khuzayma (d. 311 /923) had identified a substantial amount of the authentic hadlths 
in circulation. Abu al-Fadl Muhammad b. Tahir al-MaqdisI (d. 507/1 1 13), who spent 
most of his life in Iran and greater Syria, described the Six Books as the Sahihayn, the 
Jdmi' of al-Tirmidhi, and the Sunans of al-Nasa'l, AbQ Dawud and Ibn Majah. 'Abd 
al-KarIm b. Muhammad al-Rafi'l of Qazvin (d. 623/ 1 226) also enumerates this six-book 
series, as does the Indian Hanafl al-SaghanI (d. 650/1252), who also adds the Sunan 
of al-Daraqutnl. The Andalusian Maliki hadith scholar, al-SaraqustI (d. 524/1129), 
on the other hand, counts the Six Books as those of al-Bukharl, Muslim, al-Tirmidhi, 
Abu Dawud, al-Nasa'l and Malik. Al-Rafi'l's father, Muhammad b. 'Abd al-Karlm al- 
Rafi'l (d. 580/ 1 1 84), wrote a book called Hdwi al-usul min akhbdr al-rasul, which included 
all the hadlths from the collections of al-Bukharl, Muslim, al-Tirmidhi, Abu Dawud, 
al-Nasa"!, and Ibn Majah, as well as the Musnad of al-Shafi'l. Al-Silafl of Alexandria 
(d. 576/1 180), Abu Bakr al-HazimI (d. 584/1 188-9) and al-NawawI of Damascus (d. 
676/1277) mention only Five Books: the works of al-Bukharl, Muslim, al-Tirmidhi, 
Abu Dawtld and al-Nasa'l (although al-Silafl notes that these are the works Mushms 
have agreed on after the Muwatta'). See Abu Bakr Ahmad al-BayhaqI, Ma'rifat al- 
sunan wa al-dthdr, ed. Sayyid Kusrawl Hasan, 7 vols. (Beirut: Dar al-Kutub al-'Ilmiyya, 
1412/1991), 1:106; Muhammad b. 'Abd al-KarIm al-Rafi'l, al-Tadwin Ji akhbdr Qazwin, 
ed. 'Aziz Allah al-'Utaridi (Beirut: Dar al-Kutub al-'Ilmiyya, 1408/1987), 1:377; 2:49; 
al-Hasan b. Muhammad al-SaghanI, al-Durr al-multaqat Ji tabyin al-ghalat wayalihi Kitdb 
al-mawdu'dt, ed. 'Abdallah al-Qadi (Beirut: Dar al-Kutub al-'Ilmiyya, 1405/1985), 20; 
Abu Tahir Ahmad b. Muhammad al-Silaft, "Muqaddimat al-hdfiz al-kabir Abi Tdhir al- 
Silafi" in Hamd b. Muhammad al-Khattabi, Ma'dlim al-sunan, 3rd ed., 4 vols. (Beirut: 
al-Maktaba al-'Ilmiyya, 1401/1981), 4:357-8; Muliyl al-Din Abu Zakariyya Yahya b. 
Sharaf al-NawawI, al-Taqnb li'l-}fawawi (Cairo: Maktabat Muhammad 'All Subayh, 
1388/1968), 4; Abu al-Fadl Muhammad al-MaqdisI and Abu Bakr Muhammad al- 
Haziml, Shurut al-aimma al-sitta wa shurut al-a'imma al-khamsa, ed. Muhammad Zahid 
al-Kawtharl (Cairo: Maktabat al-Quds, 1387/ [1967]). 



10 CHAPTER ONE 

the Sahihayn form the unchanging core of the canon, and because 
the roles that the two books have played and the station they have 
achieved differ qualitatively from the other components of the canon, 
this study addresses only the canonization of al-Bukhari and Muslim. 
A comprehensive study of the Sunni hadlth canon as a whole must 
wait until another day. 

Oddly, although the broader hadith canon and the Sahihayn are fre- 
quently mentioned in Western scholarship, neither topic has received 
significant attention. Despite its having been published over a century 
ago, the work of the prescient Orientalist Ignaz Goldziher (d. 1921 CE) 
remains the most profound and detailed study of the hadlth canon. 
His interest in the entire span of the hadlth tradition and his special 
attention to the question of the hadlth canon have made his study 
the most useful to date. Even Muslim authors who regularly criticize 
Goldziher and other elder statesmen of Orientalism quote him in 
order to explain when certain hadlth collections entered the canon." 
Following the predominant Sunni division of the hadith canon into the 
Sahihayn and the four Sunans of al-Tirmidhi, Abu Dawud, al-Nasa'l and 
Ibn Majah, Goldziher devotes separate sections to each of these two 
groups. He fixes approximately where and by what time the four Sunans 
had gained canonical status and the Six Book canon had formed. He 
asserts that this authoritative selection coalesced gradually and was in 
place by the seventh/thirteenth century, perceptively adding that the 
Maghrib and the Islamic heartlands had varying opinions on which 
books constituted the canon." 

Aside from Goldziher's appreciable contributions to our understand- 
ing of the hadlth canon's emergence, his most astute observation was 
that formidable questions about the canon await answers. He evinces 
a particular pessimism about dating the canonization of the Sahihayn: 
"[W]e cannot establish with chronological accuracy the date which 
brought the consensus publicus for the two Sahihs to maturity. . . .'"" Gold- 
ziher also notes the extreme difficulty of determining why the hadith 



" See, for example, Muhammad Zubayr Siddiqi, Hadith Literature: Its Origin, Development 
& Special Features, ed. Abdal Hakim Murad (Cambridge: Islamic Texts Society, 1993), 
73-4. 

" Ignaz Goldziher, Muslim Studies II, trans, and ed. S.M. Stern and G.R. Barber 
(Chicago: Aldine Atherton, 1971), 242, 244. Goldziher's German original, Moham- 
medanische Studien, was published in 1889-90. 

'» Goldziher, 240, 



INTRODUCTION 1 1 

canon was closed and why it excluded certain collections, such as the 
Sahih of Ibn Khuzayma (d. 31 1/923), written in the same period as the 
Sahihayn. ' ' The present study will offer answers to both these questions. 

Goldziher also made a rare foray into the function of the hadith 
canon and the nature of the veneration for al-Bukharl's and Muslim's 
works. He submits that the hadith canon as a whole served as a legal 
"reference in order to find out the traditional teachings about a given 
question."'^ He touches on other functions of al-Bukhari's work in 
particular, alluding to the ritual dimension of the canon and its role 
in defining communal identity. He notes how oaths were sworn on al- 
Bukharl's Sahih, an honor otherwise reserved for the Qur'an.'^ Most 
importantly, Goldziher hints that the canonization of al-Bukhari's and 
Muslim's works was a dynamic process of interaction between the texts 
and the needs of the Muslim scholarly community'* In our discussion 
of the multivalent functions of the Sahihayn canon in Chapters Six and 
Nine, both the insight and limitations of Goldziher's comments will 
become evident. 

Goldziher also makes a unique effort to explain how the Sahihayn were 
both venerated and open to criticism. The heart of the canonical status 
of the books, he explains, was not a claim of infallibility, but rather the 
community's demand that these two works be recognized as legally com- 
pelling indicators of "religious praxis" on the basis of the community's 
consensus on their authenticity. He says: " [v] eneration was directed at 
this canonical work [i.e., al-Bukharl's collection] as a whole but not to 
its individual lines and paragraphs.'"' Goldziher concludes that "the 
veneration [of the SaMh^ of al-Bukhari and Muslim] never went so far 
as to cause free criticism of the sayings and remarks incorporated in 
these collections to be considered impermissible or unseemly...""' As 
we shall see in Chapter Eight, Goldziher's assessment proves correct 
until the early modern period, when criticism of the Sahihayn became 
anathema to many scholars. 

Since Goldziher, scholars investigating Islamic intellectual history 
or evaluating the sources for the formative first three centuries of the 



Goldziher, 239. 
Goldziher, 240. 
Goldziher, 234. 
Goldziher, 222. 
Goldziher, 247. 
Goldziher, 236-7. 



12 CHAPTER ONE 

Muslim community have found acknowledging the existence of the 
hadlth canon inevitable. Few discussions of Islamic thought or society 
fail to mention the canon and the unique status of the Sahihayn. Most 
scholars, however, have been content to either reproduce Goldziher's 
conclusions or devote only cursory remarks to the issue." The superficial 
character of these observations stems from the frequency with which 
they treat the hadlth canon as ancillary to some larger topic, such as 
early Islamic historiography or a survey of the sources of Islamic law. 
Such studies have followed Goldziher by dating the emergence of the 
canon from anywhere between the third/ninth century and the sev- 
enth/thirteenth century, devoting little thought to the actual nature or 
function of the canon. In his unparalleled study of Islamic civilization, 
for example, Marshall Hodgson only notes the existence of "canonical 
collections" of hadlth, adding that al-Bukharl's and Muslim's Sahths, 
"came to be revered as especially holy"'" In his otherwise comprehen- 
sive study of the formation of Islamic dogma and society in the second 
and third centuries AH, Josef van Ess acknowledges the existence of 
the hadlth canon but does not devote further attention to it.'^ Like- 
wise, other excellent studies of Muslim scholarly culture in the classical 
period cast only cursory glances at the hadlth canon, interpreting it as 
a natural product of the salient role that Prophetic traditions played 
in Islamic thought. In A Learned Society in a Period of Transition, Daphna 
Ephrat thus states that "by the third Muslim century, hadith had also 
achieved a central place in Muslim religious life, and the basic canons 
of the prophetic Sunna had been codified."^" 

Scholars have generally perceived the canonical hadith collections as 
representative of the Sunni worldview, and as such they have discussed 
them as a final chapter in the development of Islamic orthodoxy in 
the third/ninth century. Henri Lammens attributed the success of the 
Six Books to "the fact that they came at the right time, at the moment 

when Qpranic religion was about to take definitive shape "^' In the 

conclusion to The Eye of the Beholder, a study on how the Sunni com- 



" For a deferral to Goldziher by one of the leading Western scholars on hadith, see 
Eerik Dickinson, "Ibn al-Salah al-Shahrazurl and the Isnad," Journal of the American 
Oriental Society 122, no. 3 (2002): 488. 

'" Hodgson, The Venture of Islam, 1:332. 

'^ Van Ess, Theologie und Gesellschaft, 1:62. 

^" Ephrat, A Learned Society in a Period of Transition, 40. 

^' H. Lammens, Islam: Beliefs and Institutions, trans., Sir E. Denison Ross (New York: 
E.P. Dutton and Co., [1926]), 79. 



INTRODUCTION 1 3 

munity articulated an image of the Prophet as an act of self- definition, 
Uri Rubin refers to the large collections that appeared in this century 
as "canonical hadlth compilations" that defined orthodox Muslim 
stances. They "served as the venue for the authoritative formulation 
of an Islamic sense of spiritual and legal identity in Umayyad and 

early Abbasid times "^^ Rubin recognizes the intimate connection 

between these canonical works and the question of communal iden- 
tity, but his focus on Islamic origins prevents him from pursuing this 
discussion further 

Other scholars concerned with Islamic historiography and the devel- 
opment of the hadlth tradition have stressed that the Sahihayn and their 
authors represent the culmination of hadlth study In his Arabic Historical 
Thought in the Classical Age, Tarif Khalidi states that in Muslim's time 
"Hadith had reached its quantitative limits and spelled out its method."^'' 
"Bukhari and Muslim," he adds, "gave definitive shape to Hadith."^* 
Both Rubin and Khalidi focus on the writing of the SaMhoyn as one of 
the seals of orthodoxy, paying little attention to their role as a medium 
through which an ongoing process of institutional authorization and 
communal identification would take place. 

Scholarship on the continuing development of hadlth literature after 
the appearance of al-Bukhari's and Muslim's collections has granted 
more space to discussions of the canon. It has not, however, followed 
the promising lead of Goldziher's work. In his Islam: The View from the 
Edge, Richard BuUiet refers to the canonical hadith collections as a 
watershed event in the Muslim community's transition from the oral 
transmission of the Prophet's sunna to limiting it to specific texts. He 
prefers to identify the formation of the canon with this transition rather 
than with the genesis of the Sahihoyn themselves. Following Goldziher, he 
says that the "evolution of hadith culminated in the general acceptance, 
by the thirteenth century, of six books of sound traditions as canonical, 
as least for the Sunni majority of the population."^"' In his valuable 
discussion of the development of hadith literature in the The Cambridge 
History of Arabic Literature, Muhammad Abd al-Rauf straddles the two 
opinions: that the special recognition of the Sahihayn followed on the 



22 Rubin, The Eye of the Beholder, 224. 

2' Tarif Khalidi, Arabic Historical Thought in the Classical Period (Cambridge: Cambridge 
University Press, 1994), 43. 

2* Khalidi, Arabic Historical Thought, 59. 
2'' BuUiet, Islam: The Viewjrom the Edge, 19. 



14 CHAPTER ONE 

heels of their compilation, and that their final canonization took place 
in the seventh/thirteenth century. Thus Abd al-Rauf describes how al- 
Bukharl's book in particular was "almost immediately and universally 
acknowledged as the most authentic work in view of the author's strin- 
gent authentication requirements."^'' But after the famous systematizer 
of the hadlth sciences, Ibn al-Salah (d. 643/1245), announced that the 
Muslim community (umma) had decisively acknowledged the Sahthayn's 
unquestioned authenticity, "no more criticism [of the two books] could 

be tolerated "^' 

Modern Muslim scholarship on this question resembles its Western 
counterpart in its failure to answer questions about the canon's emer- 
gence and functions. This is largely due to the polemic motivation of 
Muslim authors addressing this subject. Khalil MuUa Khatir's Makdnat 
al-Sahthayn (The Place of the Sahlhayn) (1994)^" proceeds from an ortho- 
dox Sunni standpoint and seeks to defend al-Bukhari's and Muslim's 
work from opponents who criticize them. The Ibadi Sa'id b. Mabruk 
al-Q_anubi's ingenious al-Sayf al-hdddfi al-radd 'aid man akhadha bi-hadith 
al-dhdd fi masd'il al-i'tiqdd (The Incisive Sword: A Refutation of Those 
Who Use Ahad Hadlths in Questions of Dogma)-^ (1997-8) and the 
Twelver Shiite Mohammad Sadeq Najml's Sayri dar Sahihayn: sayr va 
barrasi dar do ketdb-e mohemm va madrak-e ahl-e sonnat (A Voyage through 
the Sahlhayn: An Exploration and Examination of two Important 
Books and Sources of the Sunnis) (2001)^" approach the issue of the 
Sahlhayn from non-Sunni stances, seeking to expose what they consider 
undue Sunni reverence for the two works. Although they offer few ana- 
lytical insights into the function or formation of the canon, the invalu- 
able citations found in these three books guide the reader to pertinent 
primary sources. These Arabic- and Persian-language secondary sources 
are thus indispensable aids in studying the Sahlhayn. Without them. 



^'' Muhammad Abd al-Rauf, ^^Hadith Literature — I: The Development of the 
Science of hadith" in The Cambridge History of Arabic Literature: Arabic Literature until the 
End of the Umayyad Period, eds. A.F.L. Beeston et al. (London: Cambridge University 
Press, 1983), 275. 

" Abd al-Rauf, ''Hadith Literature," 285. 

^'' Khalil MuUa Khatir, Makdnat al-Sahihayn (Jeddah: Dar al-Qibla li'1-Thaqafa al- 
Islamiyya, 1415/1994). 

^' Sa'ld b. Mabruk al-Qanubl, al-Sayf al-hdddfi al-radd 'aid man akhadha bi-hadith al- 
dhdd fi masd'il al-i'tiqdd, 3rd ed. (Oman: n.p., 1418/[1997-8]). 

'" Mohammad Sadeq Najml, Sayri dar Sdhihayn: sayr va barrasi dar do ketdb-e mohemm 
va madrak-e ahl-sonnat ([Tehran]: Daftar-e Entesharat-e EslamI, 1379/ [2001]). 



INTRODUCTION 1 5 

navigating the vast expanses of the Islamic intellectual heritage would 
be nearly impossible. 



Addressing the Sahihayn as a Canon 

Scholars of Islamic history have been unsuccessful in addressing ques- 
tions concerning the hadith canon in great part because they have not 
sufficiently articulated what precisely canons are, why they form and 
how they function. As Goldziher sensed, canons are not agents that 
simply leap onto the stage of history. They are created by communities 
in acts of authorization and self-definition because they meet certain 
pressing needs for their audiences. Studies on canons have proven that 
they are complicated creatures, whose emergence and functions must 
be examined as a network of interactions between a community's 
needs, its conceptions of authority, and the nature and uses of specific 
texts. Goldziher realized that to understand the canonical place of the 
Sahihayn, one must appreciate their functions. In the absence of clear 
expectations about what these could be, however, Goldziher's efforts to 
explore the canon could not move beyond a few initial observations. 
A more comprehensive discussion of the emergence and function of 
the Sahihayn canon requires a sensitivity to issues of communal identity, 
institutional authority and the way in which texts can serve as mediums 
for their expression. 

Conversely, some scholars have cultivated an acute sensitivity to 
employing the term 'canon' when treating the Sahihayn and the other 
authoritative hadith collections. The term 'canon' is so culturally loaded 
and so inevitably evokes the Biblical tradition that a commendable com- 
mitment to distinguishing the Islamic tradition from the Occidental has 
led some to deny that any hadith canon existed. Our ability to discuss 
the history of the S^hih^iyn in the language of canons and canonicity 
therefore requires an investigation of these fecund terms and their 
historical application. 



Note on the Sources and Approaches of this Study 

The study of canonization is more a study of historical perceptions 
than of historical reality. Although al-Bukharl, Muslim and their SaMh'^ 
are the centerpieces of this story, they are not its primary actors. It 



16 CHAPTER ONE 

is the community that received, used and responded to their legacies 
that forged the Sahihayn canon. Establishing the background, context 
and historical realities of al-Bukharl's and Muslim's careers is certainly 
essential for appreciating the genesis of the canon. This study, how- 
ever, is not about the Sahihayn as much as it is about the drama that 
unfolded around them. This interest in reception and perception spares 
us a prolonged focus on the questions of textual authenticity that so 
concern scholars of early Islamic history. As we will see in Chapter 
Three, surviving textual sources from the late third/ninth and early 
fourth/tenth centuries provide multi-dimensional and generally reli- 
able biographies of al-Bukhari and Muslim. Sources from this period 
also leave little doubt that the texts of the Sahihayn reached complete, 
although perhaps not polished, forms during their respective authors' 
lives.'' For us, however, the true significance of the details of al-Bukhari's 
and Muslim's lives lies in their roles as stimuli for later Muslims looking 
back at these two personages. 

Of course, our interest in reception and perception does not in any 
way relieve us of our duty to assume a historical critical approach to 
our source material. Because the Sahihayn canon is one of the most 
salient features of Sunni orthodoxy, it has attracted a tremendous 
amount of sacralizing attention from the Sunni tradition. According to 
the historical critical method, we will exert all efforts to rely on multiple 
sources of close temporal proximity to the subjects they address, rely- 
ing on isolated or later works only if the probability of their accuracy 
outweighs that of contrivance. If a source does not meet the require- 
ments of the Principle of Contextual Credibility, which dictates that 
a source must conform to the known features of its historical context, 
and the Principle of Dissimilarity, which states that a non- 'orthodox' 
account probably precedes an 'orthodox' one, then we must treat it as 
suspect from a historical critical standpoint.'^ Such material, however, 
remains tremendously valuable in charting the development of histori- 
cal perceptions about al-Bukhari and Muslim. 

The Sahihayn are arguably the most famous and prominent books in 
the Sunni tradition after the Qjur'an, and al-Bukharl and Muslim are 
titanic figures in Islamic civilization. We must thus cast a very wide 



" See Appendix II. 

'^ For a valuable and very concise discussion of these important principles of the 
historical critical method, see Bart D. Ehrman, The Mew Testament: A Historical Introduction 
to the Early Christian Writings, 2nd ed. (Oxford: Oxford University Press, 2000), 202-7. 



INTRODUCTION 1 7 

net in the sources we examine for tracing the historical development 
of the canon. Narrative sources such as biographical dictionaries and 
local histories provide invaluable source material. The Tdnkh Baghdad 
of al-Khatib al-Baghdadi (d. 463/1071), the Muntazamfi tdnkh al-umam 
wa al-muluk of Ibn al-Jawzi (d. 597/1200), the Siyar a'ldm al-nubald' and 
Tadhkirat al-huffaz of Shams al-Din al-Dhahabi (d. 748/1348), and the 
Daw' al-ldmi'li-ahl al-qarn al-tdsi' of al-Sakhawi (d. 902/1497) exemplify 
these two genres. In addition to providing essential biographical data, 
these works also record the manner in which al-Bukhari, Muslim and 
their books were perceived in different periods and localities. 

Normative sources from the various genres of hadlth literature 
provide another major source for the history of the canon. Hadlth col- 
lections that postdate the Sahihayn, such as al-Baghawl's (d. 516/1122) 
Masdbih al-sunna; works on the technical science of hadlth collection and 
criticism, such as al-Hakim al-Naysaburi's Ma 'rifat 'ulum al-hadith and Ibn 
Hajar al-'Asqalani's (d. 852/1449) al-Nukat 'aldkitdb Ibn al-Saldh; diction- 
aries of hadlth transmitters such as al-Khalill's (d. 446/1054) al-lrshddji 
ma'rifat 'ulamd' al-hadith, and commentaries on the Sahihayn such as Ibn 
Hajar's Fath al-bdn provide the bulk of data on the manner in which 
the Sahihayn were studied and used by the Sunni community. We must 
also draw from a wider range of normative sources. Works on juris- 
prudence, such as the Kitdb al-mabsut of al-Sarakhsi (d. ca. 490/1096); 
legal theory, such as the Kitdb al-burhdn of al-Juwayni (d. 478/1085); 
mysticism, like the 'Awdrif al-ma'drif of 'Umar al-Suhrawardi (d. 
632/1234), and sectarian literature, such as 'Abd al-Jalil Abu al-Husayn 
Qazvlnl's (fl. 560/1162) Ketdb-e naqd, allow crucial glimpses into the 
various usages of the Sahihcyn beyond the limited realm of hadlth study. 

As our investigation reaches the modern period, even the most recent 
Muslim scholarship can serve as a source for grasping the nature and 
function of the Sahihayn canon. Furthermore, the modern period fur- 
nishes oral sources such as lectures from scholarly centers like Cairo's 
al-Azhar University, or the recorded lectures of Salafi shaykhs like 
Muhammad Nasir al-Din al-Albani (d. 1999 CE). 

Historians can work only with what history has preserved for them. 
Like all other historical data, the sources on the origins, development 
and function of the Sahihcyn canon have been subject to the vicissitudes 
of time and fortune. The manner in which we collect and interpret 
such data is similarly prisoner to our own interpretive choices and 
biases. Yet we must have answers, whatever they may be, and for the 
period since the two books emerged as a canon their very prominence 



18 CHAPTER ONE 

in Islamic civilization has preserved a plethora of textual sources in 
manuscript or published form. For the occasionally disreputable period 
of al-Bukharl's and Muslim's pre-canonical gestation, we have only what 
Muslim scholars dutifully preserved for us. That we can even attempt 
a history of this early period is a testament to the integrity of those 
tireless 'seekers of knowledge {talabat al-'ilm)' who for centuries led pack 
animals weighed down with notebooks from teacher to teacher along 
the dusty road between Baghdad and Khurasan. 



Problems in Approaches 

In the coming chapters, our discussion of the Sahihayn canon will 
hinge on themes such as 'standards' and 'convention' and will ulti- 
mately involve the routinization of the Prophet's charismatic authority. 
Although not consciously driven by his theory, this study is perhaps 
irretrievably Weberian. Readers will also note that it is imbued with the 
corporeal language and organic idiom intimated by British scholars like 
E.B. Tylor (d. 1917) andJ.G. Frazer (d. 1941), who described the global 
phenomenon of religion as a stage in the maturation of human con- 
sciousness. In our very biological history of the Sahihayn canon, 'needs' 
will be 'felt' and 'met.' Sunnism will 'mature,' and 'strains' within it 
will 'develop.' The canon 'emerges' and fulfils certain 'functions.' Using 
such phrasal representations to move from one thought to another or 
from particulars to the general betrays certain assumptions about the 
nature of the hadith canon and Islamic civilization. Are we justified in 
treating a human society or a faith tradition as organisms that are born 
and mature until they attain some state of advancement? 

I believe this approach serves us faithfully in a study of Islamic intel- 
lectual history. Inquiring into the history of the Sahihayn is a natural 
reaction to their conspicuous prominence in Sunni Islam today. Yet 
the fact is that Islam existed as a religion and faith tradition before al- 
Bukharl and Muslim and flourished for some time after them without 
paying any remarkable attention to the two books or their authors. 
We are thus inevitably faced with a question of change, of growth or 
emergence. Like the compound of Sunni orthodoxy itself, the canon was 
not then and is now. Faced with such a stark instance of transformation 
or change, examining the canonization of al-Bukhari and Muslim as a 
linear process of maturation and subsequent tensions seems reasonable 
or even inevitable. 



INTRODUCTION 1 9 

Perhaps the most dangerous pitfall of employing a biological metaphor 
for the movement of history is the ambiguous status granted to human 
agency by such an approach. One could describe a 'canon emerging' 
without identifying the specific individuals or class who promulgated 
it. One could mention a community 'feeling needs' without stipulating 
exactly how those needs were expressed. We will try to prevent these 
problems by adhering closely to textual sources and emphasizing the 
role of individuals in the development of the canon. We will rely on 
historical actors to explain their own actions either directly through 
their own words or indirectly by reading their works critically against 
an established context. We will avoid attributing individuals' actions 
to broader political, cultural or economic forces unless there is explicit 
evidence for such a link. Certainly, we may speculate about the man- 
ner in which political context or the allocation of resources affected 
the canon, but we cannot definitively explain the canon as the direct 
result of these factors without some discernable evidence. In this way, 
we hope to avoid what Peter Brown describes as "drawing the net of 
explanation too tightly" around participants in the Islamic scholarly 
tradition.-'^ 



'' Peter Brown, The Making of Late Antiquity (Cambridge: Harvard University Press, 
1987), 4, 



CHAPTER TWO 
THE STUDY OF CANONS AND CANONIZATION 

Introduction 

What happens when a book begins to be read as a classic or part of 
a selection of classics? A sentence or turn of phrase, previously bereft 
of significance beyond its literal import, is suddenly pregnant with 
meaning and worthy of exegesis. What happens if a collection of texts 
is deemed an authentic conduit to God's will or legal right? Its very 
ontological status is raised, and minute inconsistencies within the texts 
themselves or challenges from outside sources can undermine the very 
definition of truth to which a community adheres. In neither of these 
cases are the texts themselves agents. Rather it is their body of readers 
who, out of a need for exemplary literature or select writings through 
which to approach the divine, make the books more than the sum of 
their pages, endowing them with a new authority and significance. This 
elevation binds these texts, their writers and audiences together in a 
new authoritative relationship. It creates a new universe of possible 
meanings and functions for these valorized works. This reverence or 
appreciation for the texts draws lines around the audience, including, 
excluding and defining the community. At this nexus of text, authority 
and communal identity, a canon has been formed. 

Regardless of their specific qualities, canons can be studied as a 
unified phenomenon that appears when communities authorize cer- 
tain texts, radically changing the ways they are interpreted and used. 
The Greek work kanon originally meant 'measuring stick' or a tool 
used to guarantee straightness, thus connoting the notion of a stan- 
dard. Aristotle employed the term in the context of the virtuous man, 
whom he considered to be 'the standard of good measure' in ethics.' 



' Jan Gorak, The Making of the Modern Canon: Genesis and Crisis of a Literary Idea 
(London: Athlone, 1991); 10, 17. For a brief history of the word 'canon,' see Bruce M. 
Metzger, The Canon of the New Testament (Oxford: Clarendon, 1997), 289-93. For a more 
engaged discussion of this historical definition, see Gerald T Sheppard's "Canon," The 
Encyclopedia of Religion, ed. Mircea Eliade (New York: MacMillan, 1987), 3:62-9. 



THE STUDY OF CANONS AND CANONIZATION 2 1 

Epicurus would consider logic to be the 'kanon' of true knowledge.^ 

In the early Christian tradition Paul used the word to refer to the 
'straight path' of correct belief, and 'canon' soon acquired the mean- 
ing of the 'list' of sacred writings that guided the believer. Over the 
centuries the term 'canon' has thus come to indicate a set of authorita- 
tive or exemplary texts within a specific community of readers. Fierce 
debates have raged of late and much ink has been spilled in efforts to 
provide more exact definitions for the word.' Its true and global import, 
however, is best grasped not through restricting it to an exhaustive defi- 
nition, but rather through viewing its reflections in the myriad studies 
on canons and canonicity produced by scholars from different fields. 
By examining the variety of canons, their commonalities, and efforts 
to distill the essence of canonicity, we can identify common historical 
processes and acquire conceptual tools useful for understanding the 
emergence and function of the hadith canon in Islam. 



Canons in Context and the Emergence of Canon Studies 

Canons have emerged in scriptural, literary or legal contexts, and it 
was in these fields that the study of canons and canonization began.* 
In the 1970s, however, the various strands of critical theory and post- 
modernism penetrated these arenas and presented a common challenge 
to the master narrative of canons and objective criteria. Although there 
remains scholarship devoted to religious, literary and legal canons, these 
fields have increasingly adopted the common language of hermeneutic 
studies in a joint investigation of the "politics of interpretation." Lead- 
ing experts such as Frank Kermode and Stanley Fish have exemplified 
this development, as they straddle Biblical studies and literature, and 



^ Harry Gamble, The New Testament Canon (Philadelphia: Fortress Press, 1985), 15. 

' In his study of the canon as a tool of social control, M.B. Ter Borg, for example, 
tries to distill the "primordial definition" for the concept of canon, concluding that 
its essence is that of an "objectified standard rule"; see M.B. Ter Borg, "Canon and 
Social Control," in Canonization and Deeanonization, ed. A. van der Kooij and K. van der 
Toorn (Leiden: Brill, 1998), 41 1-2; see also Jonathan Z. Smith's "Canons, Catalogues 
and Classics" in the same volume, pgs. 299-303. 

* Scholars such asjonathan Z. Smith, H J. Adriaanse and Jan Assmann have sought 
to remind audiences that it is the theological usage of canon that lies at the root of 
all modern discussion of the issues; see Jonathan Z. Smith, "Canons, Catalogues and 
Classics," and H.J. Adriaanse's "Canonicity and the Problem of the Golden Mean" 
in Canonization and Deeanonization; 295, 316. 



22 CHAPTER TWO 

literature and law, respectively. This unified field of canon studies has 
matured sufficiently to produce a series of reflections on debates over 
the notion and value of canons, and works such as Jan Gorak's The 
Making of the Modern Canon (1991) have traced the Western concept of 
'the canon' from its origins in classical Greece until modern times. 

An early attempt to study canonization as a phenomenon in religious 
traditions was AUan Menzies's prescient 1897 article "The Natural His- 
tory of Sacred Books: Some Suggestions for a Preface to the History 
of the Canon of Scripture." Menzies ultimately aims at applauding 
the Christian Biblical canon for its unique excellence and assumes an 
evolution of religion from primitive to advanced, but his work nonethe- 
less possesses remarkable foresight. Indeed, Menzies's description of the 
raw emotive forces that build canons beautifully encapsulates the place 
of hadith in the Muslim worldview. These forces are: 

books which place the believer where the first disciples stood, which 
enable him to listen to the Master's words, and overhear perhaps even 
his secret thoughts and prayers, so that he feels for himself what that 
spirit was which reached the Master from the upper region and passed 
forth from him to other men. . . . ' 

According to Menzies, the two essential conditions for the formation of 
any scriptural canon are, first, "the existence of books which the nation 
is prepared to recognize as the norm of its religion," and, second, "the 
existence of a religious authority of sufficient power to prescribe to the 
nation what books it shall receive as that norm.'"" 

Menzies's approach to canons and canonization touches on themes 
central to later examinations of the issue. Even at this early stage of 
theorizing the canon, we see the importance of communal identity 
(Menzies's "nation"), authority and a standard, or norm, for truth and 
authenticity in a religious community. His stipulation of an extant and 
sufficiently powerful "religious authority" to declare and enforce the 
canon is compelling, raising questions about the potential forms such 
authorities could assume across various communities. 

Further study of scriptural canons owes a great deal to the investiga- 
tion of the formation of the Old and New Testament canons, which 



^ Allan Menzies, "The Natural History of Sacred Books: Some Suggestions for 
a Preface to the History of the Canon of Scripture," American Journal of Theolog)) 1 
(1897): 83. 

'' Allan Menzies, "The Natural History of Sacred Books," 90. 



THE STUDY OF CANONS AND CANONIZATION 23 

began in earnest in Germany during the late eighteenth and early 
nineteenth centuries. The rival works of Theodor Zahn (1888-92) and 
Adolf Harnak (1889) were formative in this field. In the twentieth cen- 
tury, Hans von Campenhausen's Die Entstehung der chnstlichen Bibel (1969) 
is arguably the most frequently cited, although it has been surpassed 
by Bruce Metzger's definitive The Canon of the New Testament (1987). In 
1977 a series of studies on the Old Testament, most notably Joseph 
Blenkinsopp's Prophecy and Canon: A Contribution to the Study of Jewish Origins, 
focused on the canonization of the Hebrew Bible but bound it to the 
universal issues of communal conflict and identity, thus providing an apt 
point of transition into the study of the canon as a phenomenon. 

The approach to canon qua canon owes much to the field of literary 
criticism. Classical Greek literary and aesthetic criticism originated in 
the book Kanon of the mimetic artist Polycletus (fl. 450 BCE). Although 
merely a manual on how to most perfectly mimic the human form in 
sculpture, Polycletus's work was appreciated by later classical figures in 
ways the author never intended, with Pliny the Elder (d. 79 CE) stating 
that Polycletus's exemplary statues were the "canon" or standard for 
artistic expression.' Although he never uses the Greek term kanon in his 
Poetics, Aristotle presents aesthetic criteria for the literary genres of epic 
and tragedy" Each genre culminates in an unsurpassable masterpiece, 
such as the Homeric epics or Sophocles's tragedy Oedipus Rex, which 
embody the standards of excellence for their respective genres. Implied 
is the notion that there exists a set of these exemplary works, a collection 
that one might term a canon. Indeed, later Hellenistic scholars applied 
the term to a group of books whose high level of language made them 
worthy of imitation.'' In the classical Greek and Hellenistic worlds, the 
term canon thus communicated the notion of 'model' or 'exemplum,' 
what Gorak calls "a set of unsurpassable masterpieces to be studied 
and copied by all later practitioners in the field.'"" 

Since the advent of the novel and the bourgeois tragedy in the 
eighteenth century, the fixed canon of classical literature has dissolved 
amid debate over which works of literature merit the title of master- 



' Jan Gorak, The Making of the Modern Canon, 1 1 . 

" Aristotie uses the term in his Mcomachean Ethics in the context of the good person 
as " 'a canon and measure' of the truth." See Metzger, The Canon of the Mew Testament, 
289. 

' Metzger, 289. 

'" Gorak, The Making of the Modern Canon, 1 1 . 



24 CHAPTER TWO 

piece and who possesses the authority to pronounce them canonical. 
Following the post-modernist assault on the cultural systems and nor- 
mative assumptions that framed both scriptural and literary canons, 
the study of canons and canonization as phenomena has progressed 
continuously during the last quarter century. Much of this discussion 
has centered on the proper place of a literary or cultural canon within 
a modern pluralistic society, an issue that Jan Gorak has termed "the 
canon debate." 

The masterful literary and hermeneutic scholarship of Frank Ker- 
mode, exemplified in his book The Classic (1975), made the daring and 
lasting association between the notion of the literary classic, a shared 
historical vision, and empire." For Kermode the exemplification of 
pre-modern canonical literature was Virgil's Aeneid, which embodied 
both the Catholic Church's and European rulers' dream of a Holy 
Roman Empire.'^ Not only was a canon an expression of a shared 
worldview, it could entail the imperial extension and maintenance of 
that vision. In 1979, Kermode adopted a unified approach to literary 
and scriptural canon with his hermeneutic study The Genesis of Secrecy: 
On the Interpretation of Narrative and his article "Institutional Control of 
Interpretation."'^ These studies linked the canon more closely to notions 
of hermeneutic authority, control and the institutional constraint of a 
scholarly or priestly class. 

The 1970s and 1980s saw the publication of a wave of comprehensive 
studies on the formation of the Biblical canon, with a renewed emphasis 
on the role of the canon in forging identity. In his numerous books 
and articles, James Sanders has exerted a strong influence on canon 
studies, adopting the term 'canonical criticism' for the study of the 

"function of authoritative traditions in the believing communities "'* 

Principally aimed at undoing the historical-critical obsession with find- 
ing the original sitz im leben of Biblical texts, his interests lie in the way 
that the needs of a community shape and define a canonical corpus 
over time. Sanders focuses on the "period of intense canonical process" 
between the crafting of a text by its author and the stabilization of a 



" See Frank Kermode, The Classic (New York: Viking Press, 1975), 23 and 28. 

'^ Jan Gorak, Critic of Crisis: A Study of Frank Kermode (Columbia, Missouri: University 
of Missouri Press, 1987), 62. 

'^ See Kermode, "Institutional Control of Interpretation," Salmagundi 43 (1979): 
72-87. 

" James A. Sanders, Canon and Communitj (Philadelphia: Fortress Press, 1984), 24. 



THE STUDY OF CANONS AND CANONIZATION 25 

discrete canon. "It was in such periods that the faithful of believing 
communities . . . shaped what they received in ways that rendered it most 
meaningful and valuable for them."'"' Due to very real and pressing 
needs that appear in this period, a society's conception of the author- 
ity a text could acquire leaps forward. For Sanders, it is not merely 
the canonization of a text that changes its ontological status; rather, 
the pressing needs and dynamics of a faith community lead to a leap 
in that society's conception of what authority a text can attain."' Can- 
onization is therefore not simply a ritual of raising a text's ontological 
status that a community can perform at any time. Rather, communities 
undergo certain processes in which they acquire the imaginative ability 
to canonize. These ideas were further developed in Kermode's article 
"The Canon" (1987) in The Literary Guide to the Bible}' 

Canon studies has also generated a number of studies in comparative 
religion. Miriam Levering's Rethinking Scripture: Essays from a Comparative 
Perspective (1989) tackled issues of canonization and authority in a wide 
range of scriptural traditions. This collection contains a chapter by 
Kendall W. Folkert entitled "The 'Canons' of 'Scripture' " in which 
the author presents a novel distinction between the scriptural power 
of a canonical text and its actual physical presence in ritual. Gerald T. 
Sheppard's influential entry on "Canon" in the Encyclopedia of Religion 
distributes this loaded term out along a continuum between two poles 
that he terms Canon 1 and Canon 2.'" The former represents the 
notion of canon as a criterion between truth and falsehood, inspired 
and uninspired. Canon 2 manifests itself as a list, catalog or "fixed 
collection, and/ or standardized text.'"^ Sheppard proposes these two 
denotations of canon as "an illuminating heuristic device" for examin- 
ing the textual traditions of different faiths.^" 

One of religious studies' most influential contributions came in 
1977 when Jonathan Z. Smith presented a definition of the canon as 



'^ Sanders, 30. 

"^ Sanders, 32-33. 

" See Kermode, "The Canon," in The Literary Guide to the Bible, eds. Robert Alter 
and Frank Kermode (Cambridge: Belknap Press, 1987). 

'" Folkert uses the same distinction with no reference to Sheppard in his "The 
'Canons' of 'Scripture,'" published in 1989; see "The 'Canons' of 'Scripture'," in 
Rethinking Scripture, ed. Miriam Levering (Albany: State University of New York Press, 
1989), 173. 

" Sheppard, "Canon," 66. 

2" Sheppard, 64. 



26 CHAPTER TWO 

a religious phenomenon partially based on several sub-Saharan African 
religious traditions. Smith claims that canonization is "one form of a 
basic cultural process of limitation and of overcoming that limitation 
through ingenuity."^' That ingenuity, he proposes, is the hermeneutic 
process by which a religious community applies the tradition delineated 
by the canon to new problems. "A canon," Smith states, "cannot exist 
without a tradition and an interpreter"^^ Through canonizing a set of 
texts, a tradition can deposit religious authority in a manageable and 
durable form. Later interpreters of that tradition can then bring the 
authority embodied in this canon to bear on new issues. 

A landmark issue of Critical Inquiry in the early 1980s, developed 
into a book in 1984, brought canon studies fully under the rubric of 
critical theory and the postmodernist focus on the politics of expres- 
sion. This volume pursued the structural study of the canon and its 
relationship to power and communal identity by bringing together 
articles on literature, scripture, music and theory. Its editor, Robert von 
Hallberg, built on the recognition that canons had become commonly 
understood as expressions of social and political power Referring to 
questions of aesthetics, he states that "the question is not whether or 
not canons serve political functions, but rather how fully their potential 
functions account for their origins and limit their utility"^^ The most 
striking chapter in this collection is Gerald Burns's "Canon and Power 
in the Hebrew Scriptures," in which Burns addresses the distinction 
between scripture and canon. Against the previous supposition that 
scripture is authoritative and open to additional texts whereas a canon 
is authoritative but closed. Burns asserts that the defining characteristic 
of canons is their power Canons are not simply inspired or authentic 
collections of texts, they are "binding on a group of people."^* Burns 
goes on to link this powerful notion of the canon as binding to the 
act of a public reading of the text. He recalls the story of the dis- 
covery of Deuteronomy in 2 Kings. Circa 621 BCE, a Jewish priest finds 
this bound revelation from God in the Temple and brings it to King 
Josiah, who, after rending his clothes in awe, orders the new text read 



^' Jonathan Z. Smith, "Sacred Persistence: Toward a Redescription of Canon," in 
Imagining Religion (Chicago: University of Chicago Press, 1982), 52. 

^^ Smith, "Sacred Persistence," 49. 

^' Robert von Hallberg, "Introduction," in Canons, ed. Robert von Hallberg (Chicago: 
University of Chicago Press, 1984), 2-3. 

^* Gerald L. Burns, "Canon and Power in the Hebrew Scriptures," in Canons, 67. 



THE STUDY OF CANONS AND CANONIZATION 27 

to the people. ^'^ Burns adds that Ezra was also commanded to read 
the Torah to his people in public places as part of his reconstruction 
of the Jewish community in Palestine.^'' For Burns, the Biblical canon 
is primarily textual power, and the binding act of canonization takes 
place through an authoritative public reading of the text in front of a 
populace it compels to heed and obey 

The 1980s and 1990s witnessed the publication of a series of books 
and articles that turned these new theoretical models back on scriptural 
and literary traditions. Edward Said's The World, the Text and the Critic 
(1983) and Lilian S. Robinson's essay "Treason our Text: Feminist Chal- 
lenges to the Literary Canon"^' (1985), represent attacks on the concept 
of a literary canon from the two dominant trends of feminist and post- 
colonial studies. A conference held at the Leiden Institute for the Study 
of Religion in 1997 produced a massive volume entitled Canonization 
and Decanonization, which includes essays addressing the phenomenon of 
scriptural canonization and also examining the canonical traditions of 
every major religion. In another collection, Guy Stroumsa's fascinating 
essay "The Body of Truth and its Measures: New Testament Canon- 
ization in Context" emphasizes that "[cjanonization processes should 
be understood as part and parcel of religious and social processes of 
identification."^'' This article seconds Metzger's emphasis on the role of 
the Gnostic and Montanist movements in the articulation of the New 
Testament but also points out the effect that Christian -Jewish polemics 
had on the formation of these two communities. Christians and Jews 
each claimed to possess the correct interpretation of the Hebrew Bible, 
the former with the oral teachings of Christ and the latter through the 
hermeneutic tradition descending from the Oral Torah revealed to 
Moses at Sinai. That the New Testament's codification of Christ's words 
and the Mishna's setting down the interpretive methods of the Rabbis 
found written expression in the late second or early third centuries CE 
suggests that both communities were canonizing "secondary" holy texts. 
These were competing keys to understanding and unlocking a shared 



25 Burns, 69-70. 

^^ Burns, 87. 

2' See Lilian S. Robinson, "Treason our Text; Feminist Challenges to the Literary 
Canon," in The Mew Feminist Critieism: Essays on Women, Literature, and Theory, ed. Elaine 
Showalter (New York: Pantheon, 1985). 

2° See Guy G. Stroumsa, "The Body of Truth and its Measures: New Testament 
Canonization in Context," in Gnosisforschung und Religionsgeschichte, eds. Holger Preissler 
and Hubert Seiweret (Marburg: Diagonal- Verlag, 1994), 314. 



28 CHAPTER TWO 

legacy. ^^ In this strongly polemical context, Stroumsa's discussion of 
the Greek expression "kandn Us aletheias," the 'rule of revealed truth,' 
as used by Irenaeus in his writings against what he considered heretical 
Christian sects, illustrates a powerfully normative function of "canon" 
as the criterion distinguishing truth from heresy'" 

Stroumsa also highlights the distinction between cultural and reli- 
gious canons. The cementing of the New Testament as a religious 
canon in the late second century proved a very separate event from its 
emergence in the fourth century as a cultural canon, or selection of clas- 
sics to be studied as part of the curriculum of an educated man in the 
Roman world." The notion of the scriptures functioning as a cultural 
as well as a religious canon highlights the importance of Kermode's 
discussion of "the classic" and its power to extend a communal vision 
through the imperial gravity that 'proper taste' and 'proper edification' 
exert in a society. 

The study of canons in law has proven much more insular than its 
literary or scriptural counterparts. Recently, however, scholars such as 
Stanley Fish have brought legal canons under the aegis of canon studies. 
Lenora Ledwon's collection Law and Literature: Text and Theory (1996) is 
one of the most comprehensive efforts to join these two fields. More 
recently, J.M. Balkin and Sanford Levinson produced a collection of 
essays addressing specific questions of canonicity and law. Although 
these essays deal with topics of an explicitly legal nature, the editors' 
introduction articulates a visionary and overarching aim for canon stud- 
ies: "[t]he study of canons and canonicity is the very key to the secrets 
of a culture and its characteristic modes of thought."-'^ They echo tru- 
isms of canon studies such as the important influence of ferment and 
change on the visibility of a canon, but also explore topics unplumbed 
by other scholars. Balkin and Levinson introduce the idea of "deep 
canonicity," or those canonical modes of thinking, master narratives 
and canonical examples that form the background for a culture's pro- 
cess of expression and argument.'^ Most importantly, however, Balkin 
and Levinson are perhaps the first scholars since Sanders stressed the 



^' Stroumsa, 315—16; see also Sanders, 14. 

'" Stroumsa, 314. See also Elaine Pagels, Beyond Belief: The Secret Gospel of Thomas 
(New York: Vintage Books, 2003), 114-141. 

'■ Stroumsa, 308. 

'^ J.M. Balkin and Sanford Levinson, eds.. Legal Canons (New York: New York 
University Press, 2000), 4. 

'' Balkin and Levinson, 15-18. 



THE STUDY OF CANONS AND CANONIZATION 29 

"multivalency" of canonical texts to explain how canons can function 
differently depending on the audience that they are supposed to guide 
or bind together^* 

The study of legal canons has produced some of the most articu- 
late and incisive observations about the phenomenon of the canon in 
general. Stanley Fish's 1993 article "Not for an Age but for All Time: 
Canons and Postmodernism" identifies the intersection of legal and 
literary canons in the realm of high culture, where both fields stress 
the "valorization of the life of the mind."^"' Fish stresses the probative 
force possessed by canonical works. Addressing a case in which a judge 
rejected a proposed law banning all forms of racist expression because it 
would prohibit teaching Shakespeare's The Merchant of Venice, Fish notes 
that "if Shakespeare is on your side in an argument, the argument is 
over." Much like Irenaeus's kanon as 'rule of revealed truth,' the function 
of the canon. Fish concludes, is not to encourage thought, but rather 
to stop it. His explanation of Shakespeare's compelling power harks 
back to Aristotle's Poetics, for the bard is "the very canon — role, norm, 
measure, standard — in relation to which canonicity is established." A 
text becomes canonical when a community recognizes that it is the 
thing to which "all workers in the enterprise," or, in Aristotle's case, 
the genre, aspire.-''' 

A new standard in canon studies was set by Moshe Halbertal's 1997 
People of the Book: Canon, Meaning and Authority. In this work, Halbertal 
uses the Judaic tradition as a case study to synthesize applicable theory 
on the canon as it pertains to both the Hebrew Bible and the phenom- 
enological study of canonization. In doing so, Halbertal draws on fields 
ranging from jurisprudence to the philosophy of language. Unlike previ- 
ous scholars, he constructs a revolutionary yet practical framework for 
studying the relationship between canonization, authority and identity in 
what he terms "text centered communities," whose members are bound 
together through a common commitment to canonical texts. Halbertal 
explains that a text centered community exhibits several characteristics. 
First, expertise in the canonical text is a source of authority and prestige 
within the community. Second, study of the canonical text is itself an 
act of devotion urged upon all. Third, the text becomes "a locus of 



" Balkin and Levinson, 8. 

'^ Stanley Fish, "Not for an Age but for All Time: Canons and Postmodernism," 
Journal of Legal Education 43 (1993): 13. 
36 Fish, 12-15. 



30 CHAPTER TWO 

religious experience," with those who pore over or imbibe it engag- 
ing in "a religious drama in and of itself." Finally, the canonical text 
defines the boundaries of the community. It is the only recourse and 
source for the justification of ideas.^' "In a text centered community 
the boundaries of a community are shaped in relation to loyalty to a 
shared canon," asserts Halbertal.^" 

Another important concept explored by Halbertal is the notion of 
the formative text, a type of canonical text that serves as a template 
for the development of expression and interpretation within a com- 
munity. Beyond simply being a classic worthy of study and imitation, 
"[a] formative text is one in which progress in the field [, in this case, 
of understanding revealed law] is made through interpretation of that 
text."^" 

Halbertal also proposes a principle by which the vague and intangible 
notion of canonicity can be gauged. Drawing on literary hermeneutics, 
Halbertal employs the well-traveled Principle of Charity (a concept 
whose development and use will be traced later in this chapter), stipulat- 
ing that the canonicity of a scripture can be measured by the charity 
with which it is read and interpreted. If a community reads a text in 
the best possible light, attempting to minimize internal contradictions 
and reconcile notions of truth established by the text with those evident 
in the outside world, their reading is charitable and the text's canonicity 
secure. Readings that either highlight problems within the text or chal- 
lenge its probity by preferring external truths, such as those provided 
by modern science, pose threats to the canon and indicate a decrease 
in the text's holiness. 

Halbertal's work thus constitutes a new stage of canon studies. His 
promulgation of discrete definitions and conceptual tools for the study 
of canons in text centered communities is a corollary to Menzies' 
prescient if parochial work a century earlier. Both scholars grasp that 
canonization in religious communities is an irrepressible reality and 
that our understanding of canonization is nothing more than a tool for 
understanding "the secrets of a culture and its characteristic modes of 
thought."*" As von Hallberg noted, it has been widely acknowledged 



" Moshe Halbertal, People of the Book (London: Harvard University Press, 1997), 
7-8. 

^« Halbertal, 129. 

'" Halbertal, 94. 

*" Balkin and Levinson, 4. 



THE STUDY OF CANONS AND CANONIZATION 3 1 

that sacred canons are intimately bound to the profanity of self-iden- 
tification and authority Given this reality our ability to increase our 
knowledge of what the great Muslim scholar Abu Hamid al-Ghazall 
(d. 505/ 1111) called "the truth of things [haqd'iq al-umur)" hinges on our 
mastery of a lexicon and conceptual framework capable of advancing 
our understanding of how canons are informed by and govern histori- 
cal processes. 



Canon Studies and the Islamic Tradition 

The study of canons emerged in the West. With the exception of 
more global efforts such as those of Kendall Folkert and Jonathan Z. 
Smith, inquiries into canons and canonization have often been directly 
tied to the religious or literary aspects of Christianity or Judaism. To 
what extent can the history of certain authoritative hadlth collections 
in Islamic civilization be read in this light? Scholars of Islam, Islamic 
civilization and its varied genres of literary and religious expression 
have been cautious in applying approaches developed in the Occidental 
tradition to their corresponding fields in Islamic studies. One might 
argue that scholars of other civilizations should not blunder into see- 
ing canons where none exist or assume that they function in the same 
manner as those in the West. As Folkert pointed out. Western scholars 
of South Asian scriptural traditions had been misrepresenting the nature 
and contents of the Jain canon since 1882. Not only had generations 
of scholars based their understanding of the Jain canon on only one 
primary source, their conceptualization of a canon as a discrete and 
complete list of texts distracted them from the fact that "it is not specific 
texts or scriptures" but a specific "class of knowledge" that the Jain 
community considers authoritative.*' 

Tackling the mighty task of summing up the "Muslim Canon" from 
late Antiquity to the modern era, Aziz al-Azmeh is thus duly cautious. 
Al-Azmeh confines himself to a broad discussion of how the Islamic 
scriptural tradition of the Qur'an and the hadith took shape over cen- 
turies as part of a process of communal identification. He admits that 



" John E. Cort, "Svetambar Murtipujakjain Scripture," in Texts in Context: Traditional 
Hermeneutics in South Asia, ed. Jeffrey R. Timm (Albany: State University of New York 
Press, 1992), 171-2. 



32 CHAPTER TWO 

his efforts are hobbled by the primitive state of Islamic studies, which 
leads him to identify more questions than he answers. He concludes 
that the process of canonization in the Muslim tradition is "historically 
obscure except in some of its details."*^ 

Two more directed forays into the study of the canon in the Islamic 
legal and literary worlds have been William Hanaway's article "Is there 
a Canon of Persian Poetry?" (1993) and Brannon Wheeler's Applying the 
Canon in Islam: The Authorization and Maintenance of Interpretive Reasoning in 
Hanafi Scholarship (1996). Hanaway believes that one of a canon's primary 
functions is that of a "heavy weapon to fire at the enemy as well as a 
means of defining the collective self "*^ He thus cites the homogeneity 
of the courtly audience to which classical Persian poetry was addressed, 
and the lack of any "significant other" or "counter canon" contesting it, 
as evidence against the existence of a poetic canon in medieval Persia.** 
Here he echoes the argument of scholars such as Kermode, Blenkinsopp 
and Metzger that it was communal tension and competing identities 
that defined the canons of the Hebrew Bible and the New Testament.*' 
Jonathan Z. Smith's inclusive definition of a religious canon proved 
more easily applicable to the Islamic tradition, and Brannon Wheeler 
employed it to understand how the Hanafi school of legal scholarship 
preserved the authority of the Qiar'anic revelation and the Prophet's 
precedent through its chain of authorized legal interpreters.*'' 

Although extremely valuable, Hanaway's and Wheeler's studies 
nonetheless demonstrate the Scylla and Charybdis of forcing a con- 
ceptual framework onto the complex terrain of textual history. This 
framework may distract a scholar from crucial areas that might other- 
wise be explored, while accommodating the idiosyncrasies of the local 
tradition in question might neutralize a theory's efficacy. Hanaway's 
focus on a very narrow definition of a canon, for example, limits his 
inquiry to determining whether one existed or not. But canon studies 



*'^ Aziz al-Azmeh, "The Muslim Canon from Late Antiquity to the Era of 
Modernism," in Canonization and Decanonization, 197 and 203. 

" William L. Hanaway Jr., "Is there a Canon of Persian Poetry?" Edebiydt 4, no. 
1 (1993): 3. 

'* Hanaway, 3; for a reply, see Julia Rubanovich, "Literary Canon and Patterns 
of Evaluation in Persian Prose on the Eve of the Mongol Invasion," Studia Iranica 32 
(2003): 47-76, esp. 48. 

*'" See Metzger, 90-104, 

* See Brannon M, Wheeler, Applying the Canon in Islam: The Authorization and Maintenance 
of Interpretive Reasoning in Hanafi Scholarship (Albany: State University of New York Press, 
1996). 



THE STUDY OF CANONS AND CANONIZATION 33 

has generated a diversity of approaches to the issue of canonicity 
and identified the manifold functions canons can serve. If, as Moshe 
Halbertal contends, "canon and heresy are twins,"*' must we seek the 
emergence of religious canons only in times of ideological combat or 
sectarian strife? Is its role as a weapon in conflict an essential function 
of a canon? Or, as Menzies alone has argued, is the formulation of a 
religious canon the result of consolidation in the wake of tumult?*" 

Conversely, the definition of canon that Wheeler borrows from 
Smith proves too broad and insubstantial when he tackles the topic of 
the hadith canon. Wheeler's Applying the Canon in Islam is a fascinating 
study of the Islamic legal tradition, affirming von Hallberg's stance by 
concluding that the notion of canon in the Hanafi case "is best under- 
stood as a device to promote the pedagogical agenda of those who use 
certain texts to represent the authority of the past."*^ Wheeler's applied 
definition of canon, however, is so distanced from the physicality of a 
text that in his study the distinction between 'canonicity' and 'authority' 
sometimes collapses. In terms of Sheppard's and Folkert's distinction 
between Canon 1 — the criterion of truth in interpretation — and Canon 
2 — a set of representative texts — Wheeler emphasizes the former to 
the latter's exclusion. 

Wheeler explains that "[t]he Six Books are different attempts to 
delineate in 'written' form what was, at that time, considered to be 
the 'text' of the Sunnah." For Wheeler, however, these attempts do 
not merit mention as a canon. The author follows Schacht and others 
in emphasizing the transition exemplified by al-Shafi'l (d. 204/819— 
20) from local schools of customary law to an exclusive reliance on 
Muhammad's precedent as a source of law. He thus states that it was 
the entirety of the Prophet's sunna that was canonized as opposed to 
certain collections of his hadith. Wheeler warns that "the canonical 
text of the Sunnah ... is not to be equated with a particular book or a 
group of books, nor even necessarily with a written text.""'" This dis- 
tinction between the incalculably vast and amorphous corpus of the 
Prophet's legacy and distinct collections of hadith is valuable. What 



■" Halbertal, 5. 

-'" Menzies, 91. 

« Wheeler, 2. See also page 238. 

™ Wheeler, 59. Here Wheeler repeats the same oversight committed by Sheppard, 
whose very brief discussion of hadith describes the Sunna, as manifested in hadith, 
as providing a "normative and, therefore, 'canonical' (Canon 1) guide to Muslim 
exegesis." See Sheppard, 67. 



34 CHAPTER TWO 

lies unrealized in Wheeler's dismissal of physical tomes, however, is 
that those books that the community recognized as successful efforts 
to "delineate . . . the 'text' of the Sunnah" themselves became a canon 
(Canon 2). As we shall see in Chapter Nine, it was precisely the abil- 
ity of these books to function as physical, manageable symbols of the 
Prophet's sunna that met a need in the Muslim community and created 
one of the canonical dimensions of the Sahihayn. By choosing a defini- 
tion of 'canon' easily divorced from actual physical texts and treating 
'canon' on the ethereal plane of religious authority, Wheeler misses a 
canonical function of the Six Books. 

A skeptic might argue that any Western definition of canon will 
adulterate our perceptions of other traditions. Should we even employ 
the term 'canon' in our reading of hadlth literature and its functions, 
or are we naive in suggesting that they could fit into our compartments 
of canon and canonicity? 

A more germane question might be whether popular senses of 
scriptural canon in the West really acknowledge the potential subtle- 
ties and varied stages of a canon's development. Bernard Weiss, for 
example, dismisses the existence of a hadith canon in Sunni Islam by 
stating that in Islamic civilization "[God] guides no council of elders 

or divines in the formation of a sacred canon ""'' Indeed, at first 

glance the acephalous, consensus-based religious leadership in classical 
Islam might seem completely incomparable to the Pauline authority 
or council-driven first few centuries of Christian history that gave us 
the Biblical canon. As our view shifts, however, these images dissolve 
into one another. It seems evident that neither the Christian nor the 
Jewish scriptural canons were the products of councils or the decrees 
they issued. Rather, they emerged gradually through consensus, external 
pressures and liturgical use within these two believing communities."'^ 
Indeed, the final exercise of papal power that yielded the present canon 
of the Catholic Bible, declaring its text infallible and making any rejec- 
tion of its content anathema, did not occur until as late as the Council 
of Trent in 1546.^' The Biblical canon had thus existed for well over a 



■^' Bernard G. Weiss, The Search for God's Law: Islamic Jurisprudence in the Writings of 
Sayf al-Din al-Amidi (Salt Lake City: University of Utah Press, 1992), 266. 

^^ There is startling agreement on this point. See Metzger, 7; Kermode, "The 
Canon," 601; Stroumsa, 314. 

"■^ Metzger, 246. For more on the various sessions of the Council of Trent and its 
decrees, see Eugene F. Rice Jr. and Anthony Grafton, The Foundations of Early Modern 



THE STUDY OF CANONS AND CANONIZATION 35 

millennium before it reached tlie stringency imposed on the Qjur'anic 
text by the caliph 'Uthman (d. 35/655) roughly two decades after the 
death of the Prophet. 

Even when the long centuries of consensus on the Tanakh were 
sealed with a final debate over the Song of Songs and the Esther scroll, 
it was the tremendous scholarly reputation of Rabbi Akiva and not 
the edict of the Sanhedrin that gained these two books admittance 
into the canon. Biblical scholars like Guy Stroumsa and Blenkinsopp 
even reject the notion that it was the Council of Jamnia ca. 90 CE 
that resulted in the final closure of the Hebrew Bible.'* Indeed, the 
state-sponsored promulgation of the Qur'anic text by 'Uthman, or 
state attempts (even if unsuccessful) to produce official compilations 
of fiscal hadiths or the Prophet's biography under the caliphs 'Umar 
b. 'Abd al-'Aziz (d. 101/720) and al-Mansur (d. 158/775), seem much 
more suited to prevalent Western ideas of a decreed canon than the 
truly gradual maturation of the Biblical canon."'"' Why, then, must we 
tie canonization so firmly to councils? 

Weiss's understanding of canon formation, drawn no doubt from 
a belief that New Testament writings were produced and received as 
canonical texts ab initio, further limits his ability to conceive of a hadlth 
canon. He states that while the Qjur'anic text "may be regarded as a 
canon of sorts, the great compilations of Sunnaic hadith material are 
definitely not canons." Rather, he continues, "they represent a purely 
individual attempt on the part of the renowned compilers to gather 
together what was in their judgment the most reliable of the Sunnaic 
material known to them.""'*" One might ask if the authors of the synop- 
tic gospels were striving to do anything more than set down on paper 



Europe 1460-1559 (New York: W.N. Norton and Company, 1994), 174-5; and Joseph 
G. Prior, The Historical Critical Method in Catholic Exegesis (Rome: Gregorian University 
Press, 1999), 11. 

^' Stroumsa, 308; Joseph Blenkinsopp, Prophecy and Canon (Notre Dame: University 
of Notre Dame Press, 1977), 3; Sanders, 10-11. 

^^ Citing a report about this order that appears in Muhammad b. al-Hasan al- 
Shaybanl's (d. 189/805) recension of the Muwatta', Nabia Abbott states that 'Umar b. 
'Abd al-'AzIz did not order the recording of the whole sunna, but only aspects relating to 
administrative concerns. There are numerous reports that the Abbasid caliphs al-Mansur, 
al-Mahdl and Harun al-Rashid tried to make Malik b. Anas's Muwatta ' the source of 
imperial law; see Nabia Abbott, Studies in Arabic Literary Papyri II: Qur'dnic Commentary 
and Tradition (Chicago: University of Chicago Press, 1967), 2:26; and Muhammad Abu 
Zahra, Malik (Cairo: Dar al-Fikr al-'Arabi, 2002), 184-6. 

^'' Weiss, The Search for God's Law; 260, cf. 266. 



36 CHAPTER TWO 

"what was in their judgment" the most appropriate understanding of 
Christ's life. Ultimately, canon studies has demonstrated unequivocally 
that canonization is not the product of an author's intention, but rather 
of a community's reception of texts. 

Like Wheeler, Weiss concludes that "while the Qur'an was a fairly 
discrete entity with discernible boundaries, the body of hadith narratives 
constituted an amorphous mass whose boundaries no one could hope 
to catch sight of, at least with any degree of clarity." Yet on the same 
page he acknowledges the crucial role of the canonical hadith collec- 
tions. The concept of the Prophet's 'sunna,' he states, "conjures up the 
great compilations of hadith material such as those of al-Bukharl and 
Muslim."" Should we not, then, consider the possibility that the collec- 
tions of al-Bukhari and Muslim played precisely the role of synecdochic 
symbols for the Prophet's sunna in a community that understood the 
need to delimit an otherwise amorphous entity? 

Although canon studies may be a product of the Western intellectual 
tradition, it has been demonstrated that even within one civilization the 
term 'canon' is multivalent. Within this diversity, however, canon stud- 
ies has recognized that when communities authorize texts this involves 
common historical processes that change the way these texts function 
and are used. Addressing concerns about whether or not one can truly 
term the Bible a 'canon,' Kermode states that "works transmitted inside 

a canon are understood differently from those without ""'" It is thus 

ultimately the manner in which the Muslim community has treated the 
Sahihayn and the functions that they have served, not any external and 
sometimes rigid definitions of canon, that determine the two works' 
canonicity More importantly, we turn our backs on any canonical sta- 
tus that these texts may possess — and the discourse of canonization of 
which they may be a part — to the detriment of our own understanding 
of Islamic civilization. 

The existence of a set of authoritative hadith collections is certainly 
not the construct of an outside mind. Its reality as an indigenous product 
of Muslims' understanding of their own scriptural tradition is exempli- 
fied by Rashid al-Din (d. 718/1318), the famous minister and court 
historian of the Ilkhan Mongol sultan Ghazan Khan (d. 703/1304). 
Directing the writing of one of humanity's first world histories in the 



Weiss, The Search for God's Law, 260. 
Kermode, "The Canon," 609. 



THE STUDY OF CANONS AND CANONIZATION 3 7 

wake of Ghazan's conversion to Islam, this Persian scholar, physician 
and historian devotes a section of its introduction to an epistemology 
of historical knowledge. The reports from the past on which historians 
rely, he explains, fall into two categories. The first are so well known 
{tavdtor) that they convey epistemological certainty. The vast majority of 
information, however, falls into the second category of less well-attested 
narrations [ahad), which are subject to uncertainty and distortion. Even 
reports culled from eyewitnesses can transform and eventually become 
cause for disagreement as they pass from person to person. This reality, 
he states, has even affected the Prophet's legacy. "The foremost imams," 
however, "conducted thorough research and made certain selections, 
and they called them the Authentic [Collections] {Sihdh)." "All else," he 
adds, "remains within the sphere of doubt and hesitation. "-^^ 

Rashid al-Din was not writing a religious history. The overpowering 
charisma of the "Golden Family" of Genghis Khan and the dictates 
of classical Persian political theory occupied him far more than the 
distinctly theological or sectarian concerns of the first centuries of 
Islam. The Islam to which the Mongol rulers of Iran and Rashid al- 
Dln himself had converted was a fully mature civilization that initiated 
its citizens into a cosmopolitan worldview and shared vision of history. 
Rashid al-Dln's historical epistemology is itself a product of Hellenistic 
Near Eastern discussions of mediate and immediate (apodictic) knowl- 
edge. Yet even in this context, the Six "authentic" hadlth collections 
represent religious and social order amid the polyglot historical roots of 
Islamic civilization. The Sihdh canonized a tract of the past, securing 
the Prophetic authority so central to Islamic communal identification 
in the medium of specific texts. 

The unique status of al-Bukharl's and Muslim's Sahths similarly con- 
stitutes an undeniable historical reality. From his seat in Delhi, capital 
of the Moghul Empire in the 1700s, Shah Wall Allah (d. 1762) sum- 
marized the legal and doctrinal controversies that had unfolded over 
more than a millennium of Islamic history in his masterpiece, Hujjat 
Alldh al-bdligha (God's Conclusive Argument). In his chapter on hadlth, 
he concludes that "as for the two Sahth^ [of al-Bukharl and Muslim], 
the scholars of hadlth have agreed that everything in them attributed to 



^ Rashid al-Diii Fadlallah, Jame'-e tavankh, ed. Mohammad Rushan and Mostafa 
Musavl (Tehran: Nashr-e Elborz, 1373/[1994]), 1: 9-10. 



38 CHAPTER TWO 

the Prophet is absolutely authentic," adding that "anyone who belittles 
their stature is guilty of corruptive innovation [mubtadi') and not fol- 
lowing the path of the believers."'''' 

The existence of a set of authoritative hadlth collections in general, 
and the exceptional status of al-Bukhari's and Muslim's books in par- 
ticular, are thus historical realities that we ignore at our own peril. As 
this study will demonstrate, in both their capacity as a standard of truth 
(Canon 1) and a set of delimited and representative texts (Canon 2), 
the Sahihayn are in the fullest sense of the word 'canonical.' Not only 
may we dispense with the quotation marks that often so cautiously 
adorn the "canons" of Islam, we can use tools developed in canon 
studies to better understand and articulate the form and functions of 
the hadlth canon. Doing so is nothing more than responding to voices 
from within the Islamic tradition that call us to view the Sahihayn as 
part of the broader phenomenon of canonicity. 



Theoretical Tools and Common Historical Processes: Canon Studies and the 

Hadlth Canon 

The present study is neither theory-driven nor comparative. To the 
extent possible, the story of the hadlth canon must be read on its own. 
This study does, however, contend that any canon represents the inter- 
action of text, authority and communal identification. The foregoing 
discussion of different canons and the phenomenon of canonicity has 
highlighted this common historical process and provided a conceptual 
lexicon that is useful for addressing the hadlth canon. Investigating this 
issue in light of the way other literary and scriptural communities have 
conceived of canonization can bring elements otherwise unperceived 
into relief. In tackling a subject that lies at the nexus of text, community 
and authority, we must expect to address the same themes as studies 
of other canons. It is the extent to which the Muslim community's 
perception and use of al-Bukhari's and Muslim's &y^z"fe meets these 
expectations that justifies this approach. Ultimately, it is the promi- 
nence of questions of self-definition, the institutionalization of religious 



''" Shah Wah Allah al-Dihlawi, Hujjat Allah al-baligha, 2 vols, in 1 (Cairo: Dar al- 
Turath, [1978]), 1:134. 



THE STUDY OF CANONS AND CANONIZATION 39 

authority and a qualitative change in the way the community viewed 
these two works that qualifies them as canonical. 

Having reviewed the development of canon studies, let us now 
elaborate more fully some of the central themes and constructs that 
will be employed in the study of the Sahihayn canon. 

a. Canons and Community 

Texts may become authoritative, but they are not binding on all man- 
kind. Canons are necessarily the creations of specific communities or 
audiences. Because the act of authorizing certain books draws lines 
excluding other works, canons have been understood as tools of inclu- 
sion and exclusion within a broader community. As Gerald Burns and 
Joseph Blenkinsopp have observed in the case of the Hebrew scriptures, 
"what we call 'canon' is intelligible only in the context of conflicting 
claims to control the redemptive media and, in particular, to mediate 
and interpret authoritatively the common tradition.'""' Scriptural canons 
thus form when certain sections of a community attempt to monopolize 
the true interpretation of a religious message shared by all its members, 
excluding those audiences that identify with the non-canonical. 

In the case of the formation of the New Testament canon, one 
of the first to advance a set of authoritative media for understand- 
ing Christ's legacy was the second-century Gnostic Marcion.''^ His 
list of works, one of the first 'canons,' excluded the Hebrew Bible as 
the corrupt revelation of the Old Testament God who had plunged 
the world into darkness. The true salvational teachings of Christ that 
could reunite man's soul with the Divine, Marcion contended, were 
contained solely in a purified version of Luke's gospel and a selection 
of Paul's letters.''-' Championing what would become orthodox Chris- 
tianity, Irenaeus, the second-century bishop of Lyons and inveterate 
enemy of the Gnostics, responded by affirming the unity of the Old 
and New Testaments. More importantly, he proclaimed a closed canon 
consisting of the "four-formed gospel" of Matthew, Mark, Luke and 
John. These books alone, not the myriad of other gospels circulating 
among Christians at the time, captured Jesus's life and teachings; like 



''' Burns, 81; Blenkinsopp, Prophecy and Canon, 96. 
''^ Gerald Sheppard, "Canon," 3:63. 

''' Kermode, "Institutional Control," 77. For an excellent treatment of Marcion's 
beliefs and sources, see Metzger, 90-94. 



40 CHAPTER TWO 

the four directions of the compass, there could be no more and no 
less.''* As Metzger and Elaine Pagels have shown, the formation of the 
New Testament canon cannot be grasped without acknowledging the 
catalyst of Marcion's heretical counter-canon. By declaring that only 
certain books were authentic and binding for Christians, Irenaeus had 
dubbed not only the Gnostics but also the audiences of other innocu- 
ous gospels heretics. Halbertal's stipulation that "canon and heresy are 
twins" succinctly represents this vein of scholarship.'''' 

This conception of canonicity as tied to competing claims to the 
control of a common tradition has so dominated canon studies that 
Hanaway concluded that the absence of such a "significant other" as 
an opponent in medieval Persian literature precluded the existence of 
a canon of Persian poetry. This trend's commanding role in canon 
studies is not difficult to understand. Canons are necessarily vehicles 
for identification. Just as 'non-canonical' works are a byproduct of their 
formation, so canons must delineate a new community of believers 
from the old, wider audience. 

Such assumptions, however, leave unexplored another function of 
canons in community. Canons can also emphasize inclusion and agree- 
ment more than exclusivity. They can function as a tool of reconcili- 
ation, a medium for communication or for creating common ground 
between adversaries. Although one sect might advance a canon as a 
polemical tool in a time of strife, this canon need not serve to exclude 
other forms of redemptive media. Rather, its compelling power could 
dwell in its broad appeal. As Hanaway contends, canons may serve 
chiefly as a "heavy weapon to fire at the enemy'"''' but only evidence 
also accepted by that enemy will prove compelling in debate. Even 
in polemic, a canon's power must spring from its status as part of a 
shared language. Considering the powerful role of the consensus {ijma') 
of the Muslim community in Islamic epistemology we must take care 
to consider the emergence of the Sahihayn canon as an inclusive effort 
to force various sects to recognize a common medium for discussing 
the Prophet's legacy. 



Pagels, 81-5; Metzger, 153-7. 
Halbertal, 5. 
Hanaway, 3. 



THE STUDY OF CANONS AND CANONIZATION 4 1 

b. Kanon and the Measure of Revealed Truth 

Despite its overwhelming denotation of "authoritative list" in modern 
and many pre-modern minds, the kanon that meant "measure" to 
Aristotle and lent itself so readily to the "rule of revealed truth" in 
early Christian polemic has survived as one of the most useful tools for 
conceptualizing canonicity. Canon studies has emphasized canonization 
as an impetus for interpretative activity, with Kermode underscoring 
that authorizing books transforms them into potentially inexhaustible 
mines of interpretation. " 'Licensed for exegesis,' " he concludes, "such 
is the seal we place upon our canonical works. "'*' This focus has some- 
what overshadowed the role of the canon as a categorical measure of 
truth, a tool that Fish notes is designed to end discussion rather than 
encourage it. Here the kanon as measure is "an authority that can be 
invoked in the face of almost any counterevidence because it is its own 
evidence and stronger in its force than any other "''*' 

Indeed, the original purpose of the kanon tes aletheias, or 'measure of 
revealed truth,' advanced by Irenaeus was to limit interpretation of the 
gospels. Just as the early church father had proclaimed an authorized 
collection of four gospels, so too he propounded a hermeneutic lens 
to ensure an orthodox reading of his canon. When reading rich and 
pregnant texts like the Gospel of John, so favored by many Gnostics, 
one must apply "the measure of revealed truth" that interprets them in 
as literal a manner as possible and in the light of Jesus's 'true' teachings. 
To open the doors of esoteric interpretation of the canonical gospels 
would mimic the methods of pagan philosophers such as the Stoics, 
who interpreted Homer's epics allegorically''^ Irenaeus sought to end 
the subversive preaching of the Montanist movement of Asia Minor, 
whose wandering prophets claimed to be seized by the Holy Ghost 
and proclaimed the continuing revelation of Christ in the community. 
The message and authority of Christ thus had to be contained in the 
canon and interpreted properly. As rabbis debating questions of holy 
law had declared when some scholars claimed that God had validated 
their position in a dream, "we do not listen to voices from heaven."'" 



'"' Kermode, "Institutional Control," 83. 
<''" Fish, 12. 
"'' Pagels, 117. 

'" The contemporary Shafi'l scholar Sa'ld 'Abd al-Latif Fuda concurs, stating that 
"inspiration [ilhdm) is not a conduit for revealed knowledge {'ilm) among the people of 



42 CHAPTER TWO 

For Irenaeus, the canon as text and kandn as measure were guarantors 
of an orthodox monopoly on interpretation. In J.Z. Smith's definition 
of the canon as a tool in which the authority of a tradition is depos- 
ited in order to extend its implementation into future circumstances, 
Irenaeus's "measure of truth" would be a trump card in determining 
the authentic vision of Christianity Indeed, the authority of his canon, 
Irenaeus claimed, stemmed from its authenticity. He had chosen his 
"four-formed gospel" because they were the only books supposedly 
written by eyewitnesses of the events they described." 

Like Irenaeus, Muslim scholars of hadith have been preoccupied with 
questions of authenticity. The traditions of the Prophet were certainly 
subject to interpretation as scholars applied them to questions of law, 
morality and doctrine, but in debates over their meaning it was the 
question of authenticity that was paramount in their collection and 
criticism. The more authentic the Prophetic report, the more authorita- 
tive it was. In the elaboration of the faith, and certainly in inter-school 
polemics, "interpretation is a function of authentication {al-ta'wTl far' 
'aid al-ithbat)." While Irenaeus's canon required a canonical lens for 
proper viewing, for hadith collections the kandn of truth was the canoni- 
cal books themselves. A collection deemed an authentic repository for 
the Prophet's hermeneutic authority was the tool through which that 
authority could be employed decisively in the further elaboration of 
Islam. For Kermode the canon is licensed for exegesis; for Muslims a 
canonical hadith collection was licensed for common use. 

c. The Principle of Charity and Canonical Culture 

One of the most useful conceptual tools for studying the emergence 
and development of the hadith canon is the Principle of Charity, a 
notion only recently applied to canonicity In its most general sense, the 
Principle of Charity assumes that people interpret signs in the best pos- 
sible light. It was first developed as a tool of analytical philosophy, and 
later explored by N.L. Wilson in a 1 959 issue of Review of Metaphysics. 
Wilson proposes that, presented with a field of data or propositions, 
humans will choose the designation that makes the maximum number 



truth"; see http://www.al-razi.net/website/pages/warakat.htm, part 10 (last accessed 
9/14/2005). 
" Pagels, 111. 



THE STUDY OF CANONS AND CANONIZATION 43 

of Statements true.'^ Here an individual forced to come to terms with 
a set of propositions treats reality with charity, reading its 'text' in the 
best possible light. He charitably assumes a system must exist, so he 
reasons that one should select the data that best support some notion 
of order 

The Principle of Charity has also found significant use in the study 
of language. Members of a speech community all subscribe to rules 
that govern the common activities of construction and interpretation, 
so every sentence and expression is a new proposition that must fit 
into this shared system. If one's interlocutor says, "I ran the light at 
the introspection," one would automatically assume that he or she 
had meant to say 'intersection.' At a certain point in conversation, it 
becomes more likely that a speaker has simply erred than that he or she 
is trying to subvert grammar or convention." It is not simply due to a 
reliance on the stability of convention that one treats the interlocutor's 
remarks with charity; we automatically view them in the best possible 
light in order to uphold the very conventions of language that allow 
us to understand one another. As Donald Davidson explains, "We do 
this sort of off the cuff interpretation all the time, deciding in favour 
of reinterpretation of words in order to preserve a reasonable theory 
of belief"'* As a result, context can overwhelm isolated or fleeting 
divergences in an otherwise consistent system. 

The Principle of Charity has been similarly applied to the com- 
munication between author and reader through the medium of text. 
In textual interpretation, the Principle involves approaching a work 
with the assumption that its author is rational and that its elements 
of plot, theme and character conform to some sense of order Here 
grammar and semantic convention morph into notions of intra-textual 
uniformity and interpretive harmony. The Principle of Charity mani- 
fests the reader's need for what Kermode calls "that concordance of 
beginning, middle and end which is the essence of our explanatory 
fictions "" 



'^ N.L. Wilson, "Substance without Substrata," Review of Metaphysics 12, no. 4 
(1959): 532. 

" See Willard Quine, Word & Object (Cambridge, Massachusetts: MIT Press, 1960), 
59. 

'* Donald Davidson, Inquines into Truth and Interpretation, 2nd ed. (Oxford: Clarendon 
Press, 2001), 196. 

'"' Kermode, The Sense of an Ending, 3rd ed. (Oxford: Oxford University Press, 2000), 
35-36. 



44 CHAPTER TWO 

Drawing on Ronald Dworkin's Law's Empire, Halbertal extends the 
Principle of Charity to the domain of canonicity. "' Given several pos- 
sible interpretations of a canonical passage, the 'correct' one will be 
the one that supports the text's internal consistency and compatibility 
with accepted notions of truth or propriety Canonizing a legal or 
scriptural text thus "not only endows it with authority but also requires 
a commitment to make the best of it.'"' The Principle of Charity 
recognizes that in the case of a scriptural or legal canon, "there is an 
a priori interpretive commitment to show the text in the best possible 
light. Conversely, the loss of this sense of obligation to the text is an 
undeniable sign that it is no longer perceived as holy." Halbertal thus 
stipulates the principle that "the degree of canonicity of a text corre- 
sponds to the amount of charity it receives in its interpretation.'"" 

The assumed existence of an ordered reality in Wilson's study, and the 
manifest authority of linguistic context and convention in a speech com- 
munity, here become the worldview that a community has constructed 
around a canonized text. One might refer to this surrounding system 
as the text's canonical culture. It is the system that trains readers 
or listeners to interpret a canonical text in a reverential manner and 
with suitable awe. In short, canonical culture obliges readers to treat 
the canon with charity. Unlike grammar or linguistic convention in a 
speech community, however, a canonical culture cannot be taken for 
granted or unconsciously defended. It must be consciously created and 
nurtured through careful control of the manner in which the canon is 
read and discussed. Upholders of this canonical culture must themselves 
actively propagate it and condemn its breaches. A canonical culture 
would demand that interpreters of the canon observe certain respectful 
formalities, accord the text and its authors the proper accolades and 
gloss over possible flaws. Like a language, however, one can identify 
the rules of canonical culture and recognize certain violations of its 
grammar By measuring the charity extended, one can observe the con- 
struction of a canonical culture as it seeks to cast a text, and perhaps 
even its author, in the best possible light. Once one gains a familiarity 
with this canonical culture, one can detect lapses and even perceive its 
participants interacting with its boundaries and demands. 



"" For an analysis and commentary on Dworkin's work, see Andei Marmor, 
Interpretation and Legal Theory (Oxford: Clarendon Press, 1992), 57-60. 
" Halbertal, 28. 
'« Halbertal, 29. 



THE STUDY OF CANONS AND CANONIZATION 45 

The Principle of Charity is ideally suited for studying the canoniza- 
tion of the Sahihayn because the canonical culture surrounding them 
has depended entirely on the compatibility of the two texts and their 
authors with prevailing notions of truth and authenticity" From the 
early second/ eighth century many pious Muslims who collected the 
sayings of their Prophet recognized that an exacting criticism of both 
those who reported these traditions and the traditions themselves was 
necessary to identify forged material. Their opponents from among the 
Muslim rationalists and the more analogy-based legal schools of Iraq, 
however, were very skeptical of their claims to be able to collect and 
authenticate statements transmitted orally. The image that the hadith 
scholars therefore cultivated in the Muslim community highlighted 
their caution, lack of tolerance for lapses in memory or inconsistencies 
in transmission, and an almost pathological devotion to amassing and 
sifting through the Prophet's legacy. The idealized muhaddith (hadith 
scholar) was singularly devoted to mastering the Prophet's word, dismiss- 
ing as corruptive innovation anything that did not extend back to him. 
For them the hadlth's chain of transmission {isndd), the only lifeline to 
the Prophet's teachings and an Islam unpolluted by the cosmopolitan 
religious atmosphere of the Near East, became the center of a cult of 
authenticity. "The isndd for us is religion; were it not for the isndd" they 
claimed, "whoever wanted could say whatever they wanted."™ It was 
the very authenticity of these isndds,, however, that the hadith scholars' 
opponents doubted. To canonize the Sahihayn, the hadith scholars' cult 
of authenticity had to become both more intensified and accepted in 
the wider Sunni community. It was argued, as we shall see, that these 
two demanding books met the whole community's requirements for 
hadith authenticity. The canon thus rested on a claim that required 
the approval of segments of the community that had been perennially 
mistrustful of the hadith scholars' methodology and the ever-critical 
hadith scholars themselves. As we shall see in Chapter Seven, a perpetual 
reinforcing of this cult of authenticity would prove the salient feature 
of the canonical culture surrounding the two works. The two books 



™ For a very brief but parallel discussion of the "critical gentleness" with which 
Muslim scholars treated their canonical texts, see Aziz al-Azmeh, "The Muslim 
Canon," 212. 

°" "Al-isndd 'indand din, law Id al-isndd la-qdla man shd'a md shd'a, wa Idkin idhd qila lahu 
man haddathaka baqiya;" see al-Khatib al-Baghdadi, Tdrikh Baghdad, ed. Mustafa 'Abd 
al-Qadir Ata, 14 vols. (Beirut: Dar al-Kutub al-'Ilmiyya, 1417/1997), 6:164. 



46 CHAPTER TWO 

and their authors had to be lifted above their peers and any possibility 
of error The extent to which different segments of the Sunni commu- 
nity gradually extended the charity of this unblemished authenticity to 
al-Bukhari and Muslim and their works charts the emergence of this 
canonical culture. 



Conclusion 

Whether scriptural, legal or literary, canons lie at the intersection of 
text, authority and communal identification. They are no more unique 
to the Occidental tradition than are these three seminal notions. Indeed, 
canons are undeniable historical realities that change the manner in 
which the books function and are treated by their audiences. Where 
exactly the canon of the Sahih collections of al-Bukhari and Muslim 
fits in this nexus is a question only a study devoted to their unique 
history can answer. The remarkable efforts of scholars such as J.Z. 
Smith, Halbertal and Kermode to understand canons in their various 
contexts, however, must serve as guides in alerting us to the possibili- 
ties and perhaps even the inevitabilities facing the study of a canon's 
emergence and functions. Canon studies has drawn our attention to 
the role of the canon as a possible tool for inclusion in a community. It 
has provided the Principle of Charity as a device to measure canonic- 
ity and chart the development of a canonical culture. Finally, we can 
conceive of the canon as a common measure of truth in which the 
authority of tradition is deposited for later application. As Menzies, 
the earliest student of canonization as phenomenon, so ably pointed 
out, a canon must begin with books.*" What, then, was the genesis of 
those two books that allowed Muslims to stand "where the first disciples 
stood..., to listen to the Master's words, and overhear perhaps even 
his secret thoughts and prayers," feeling "what that spirit was which 
reached the Master from the upper region and passed forth from him 
to other men . . .?"^^ 



Menzies, 90. 
Menzies, 83. 



CHAPTER THREE 
THE GENESIS OF AL-BUKHARI AND MUSLIM 

Introduction 

Leafing through the pages of al-Bukhari's Sahih today, the book seems 
to be the natural culmination of the Muslim study of the Prophetic 
legacy: Muhammad's authenticated words and actions, enclosed in a few 
volumes. For the hadith scholars and pious Muslims of the third/ninth 
century, however, hadlths were not bound tomes taken off the shelf and 
read. They were living links to the Prophet and the manifestation of his 
charismatic authority in everyday life. Although Muslim scholars of the 
first three centuries of Islam strove to prevent forged hadiths from being 
attributed to the Prophet, even in the case of dubious transmissions 
the powerful formula "the Messenger of God said..." made reports 
from Muhammad prima facie compelling to many jurists. Al-Bukhari's 
and Muslim's compilation of works limited to authenticated reports 
was thus a revolutionary act. The two Sahths were destined for eventual 
canonization, but in the decades after their authors' deaths important 
segments of the scholarly community saw them as an insolent departure 
from tradition. The Sahihayn possessed an elitism and finality that clashed 
with the manner in which hadith-based jurists employed the Prophetic 
legacy. Al-Bukharl's and Muslim's work thus constituted a split in the 
hadith tradition; although the Sahihayn would become an authoritative 
institution, they would exist side by side with the continued amassing 
of Prophetic traditions through the living isndd. 



The Development of Hadith Literature 

When he was sixteen years old, Muhammad b. Isma'il al-Bukhari left 
his hometown of Bukhara in Transoxiana with his mother and brother 
Ahmad on a pilgrimage to Mecca. The small party would probably have 
attached themselves to one of the merchant caravans carrying luxury 
goods west along the Silk Road. Traversing the desert, they would have 
passed through the bustling garrison-city of Merv before climbing the 



48 CHAPTER THREE 

mountains to Sarakhs and then descending into the rolling green and 
golden valleys of Khurasan.' They would have made a stop in the 
city of Naysabur, its northernmost orchards lying against the foothills 
of the mountains. As they continued west along the northern edge of 
the Iranian desert, they would have passed through Bayhaq, the great 
commercial and scholarly center of Rayy before voyaging across the 
Zagros Mountains and descending onto the flood plain of Iraq. They 
may have stopped in Baghdad, the "navel of the world" and a throb- 
bing center of trade, scholarship and political intrigue. They would 
have continued along the caravan trail, now crowded with pilgrims, 
across the north Arabian deserts to the rugged mountains of the Hijaz. 
Skirting jagged ridges interspaced by yellow tracts of sand, they would 
have ended their journey where Islam began over two centuries earlier, 
in the dry and rocky valley of Mecca. 

Al-Bukhari, like generations of dedicated and pious Muslims before 
him, devoted his life to answering the question that lies at the heart of 
the Islamic religious tradition: how does one live according to God's 
wUl as revealed in the Qur'an and taught by His Prophet? Almost two 
centuries before al-Bukharl set off on his pilgrimage, the same road 
had carried the Muslim armies into Eastern Iran and Transoxiana 
as they triumphantly spread their new religion outwards in time and 
space from its epicenter in the Hijaz. His voyage back to Mecca, the 
Prophet's home and location of the Ka'ba, fulfilled the duty ordained 
upon all Muslims to return to the place where God had revealed their 
religion and where the Prophet had served as its first authoritative 
interpreter 

In the two hundred years since the beginning of the Islamic tradition, 
Muslims such as al-Bukhari had turned back again and again to the 
authoritative legacy of the Prophet's teachings as it radiated outwards 
through the transmission and interpretation of pious members of the 
community. In Medina, al-Qasim b. Muhammad b. Abi Bakr (d. 108/ 



' ^Khurasan' as a topographical and administrative term has had a wide range of 
meanings. In the early Islamic period the name was often used to denote the region 
extending from Western Iran to Transoxiana. Today it is a relatively contained province 
in Eastern Iran with its capital at Mashhad. We will use the name as the geographer 
al-MuqaddisI (d. after 380/990) did: to describe the area in Eastern Iran centered on 
the four major cities of Naysabur, Merv, Herat and Balkh. We will distinguish this 
region from Transoxiana, with its Zarafshan River cities of Bukhara and Samarqand; 
Paul Wheately, The Places Where Men Pray Together (Chicago: University of Chicago 
Press, 2001), 172-90. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 49 

726-7), the grandson of the first caliph, and Sa'ld b. al-Musayyab (d. 94/ 
713), the son-in-law of the most prolific student of the Prophet's 
legacy, AbQ Hurayra, became two of the leading interpreters of the 
new faith after the death of the formative first generation of Muslims. 
Their interpretations of the Qjur'an and the Prophet's legacy, as well 
as those of founding fathers such as 'Umar b. al-Khattab, were col- 
lected and synthesized by the seminal Medinan jurist Malik b. Anas 
(d. 179/796). In Kufa, the Prophet's friend and pillar of the early 
Muslim community, 'Abdallah b. Mas'ud (d. 32/652-3), instructed his 
newly established community on the tenets and practice of Islam as it 
adapted to the surroundings of Christian, Jewish and Zoroastrian Iraq. 
His disciple 'Alqama b. Qays (d. 62/681) transmitted these teachings 
to a promising junior, Ibrahim al-Nakha'T (d. 95/714), who in turn 
passed his approaches and methods of legal reasoning to Hammad 
b. Abi Sulayman (d. 120/738). His student of eighteen years, Abu 
Hanifa (d. 150/767), would become a cornerstone of legal interpre- 
tive effort in Iraq and the eponym of the Hanafi school of law. Unlike 
Medina, the Prophet's adopted home where his legacy thrived as living 
communal practice, the polyglot environment of Kufa teemed with 
ancient doctrines and practices foreign to the early Muslim community. 
Many such ideas found legitimation in spurious reports attributed to 
the Prophet, and Abu Hanifa thus preferred a cautious reliance on 
the Qjur'an and his own reasoning rather than to risk acting on these 
fraudulent hadiths. 

By the mid-second century, two general trends in interpreting and 
applying Islam had emerged in its newly conquered lands. For both these 
trends, the Qur'an and the Prophet's implementation of that message 
were the only constitutive sources of authority for Muslims. The practice 
and rulings of the early community, who participated in establishing the 
faith and inherited the Prophet's hermeneutic authority, were the lenses 
through which scholars like Abu Hanifa and Malik understood these 
two sources. 'Abd al-Rahman al-Awza'i of Beirut (d. 157/773-4) thus 
stated that "religious knowledge ('ilm) is what has come to us from the 
Companions of the Prophet; what has not is not knowledge."" When 
presented with a situation for which the Qiar'an and the well known 



^ Abu 'Umar Yusuf Ibn 'Abd al-Barr al-Qurtubi, j'ami' bayan al-'ilm wafadlihi, ed. 
'Abd al-Rahman Muhammad 'Uthman, 2 vols. (Medina: al-Maktaba al-Salahyya, 
[1968]), 2:36. 



50 CHAPTER THREE 

teachings of the Prophet and his Companions provided no clear an- 
swer, scholars like AbQ Hanifa relied on their own interpretations of 
these sources to respond. Early Muslim intellectuals like Ibn Qutayba 
(d. 276/889) referred to such scholars as 'ahl al-ra'y' or practitioners 
of individual legal reasoning.' Other pious members of the community 
preferred to limit themselves to the opinions of the earliest generations 
and more dubious reports from the Prophet rather than to opine in 
a realm they felt was the exclusive purview of God and His Prophet. 
The great Baghdad scholar Ahmad b. Hanbal (d. 241/855) epitomized 
this transmission-based approach to understanding law and faith in his 
famous statement: "You hardly see anyone applying reason {ra)}) [to 
some issue of religion or law] except that there lies, in his heart, some 
deep-seated resentment {ghill). A weak narration [from the Prophet] is 
thus dearer to me than the use of reason."* Such transmission-based 
scholars, referred to as 'the partisans of hadith [ahl al-hadith)' preferred 
the interpretations of members of the early Islamic community to their 
own. For them the Muslim confrontation with the cosmopolitan atmo- 
sphere of the Near East threatened the unadulterated purity of Islam. A 
narcissistic indulgence of human reason would encourage the agendas 
of heresy and the temptation to stray from God's revealed path. Only 
by clinging stubbornly to the ways of the Prophet and his righteous 
successors could they preserve the authenticity of their religion. 

It was in this milieu that the tradition of hadith literature emerged. 
Although Muslims had been memorizing or writing down the words 
of the Prophet and his followers from an early period,^ the first major 
hadith collections, called musannafi, were essentially transcripts of the 



^ For more on this subject, see Christopher Melchert, "Traditionist-Jurisprudents 
and the Framing of Islamic Law," Islamic Law and Society 8, no. 3 (2001): 383-406, 
esp. 385. 

' Muhammad Abu Zahra, Ibn Hanbal (Cairo: Dar al-Fikr al-'Arabi, [1965]), 239. 

^ An example of an early collection of hadith is the sahifa of Hammam b. Munabbih 
(d. circa 130/747), a disciple of Abu Hurayra, which includes 138 hadlths; for more 
information on the unsystematic collection of written hadith in the first two centuries 
of Islam, see Abd al-Rauf, ^'Hadith Literature," 272. For more on the emergence of 
historical writings, see Nabia Abbott, Studies in Arabic Literary Papyri II: Qur'dnic Commentary 
and Tradition; Fuat Sezgin, Geschichte des arabischen Schrifttums, 12 vols. (Leiden: E.J. Brill, 
1967), 1:53-84; Fred M. Donner, Narratives of Islamic Origins: The Beginnings of Islamic 
Historical Writing (Princeton: Darwin Press, 1998), 279; Muhammad al-A'zamI, Studies 
in Early Hadith Literature (Kuala Lumpur: Islamic Book Trust, 2000); Harald Motzki, 
The Origins of Islamic Jurisprudence: Meccan Fiqh before the Classical Schools, trans. Marion 
H. Katz (Leiden: Brill, 2002), 158. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 5 1 

legal discourse that had developed during the first two centuries of 
Islam. Arranged into chapters dealing with different legal or ritual ques- 
tions, they were topical records of pious Muslims' efforts to respond 
to questions about faith and practice. Malik b. Anas's Muwatta' is thus 
a mixture of Prophetic hadlths, the rulings of his Companions, the 
practice of the scholars of Medina and the opinions of Malik himself^ 
Likewise, the musannaf of Ibn Jurayj (d. 150/767) is a collection of 
reports from the Prophet, Companions and Successors such as 'Ata' 
b. AblRabah(d. 114/732).' 

During the late second and early third centuries, however, the 
prevalence of specious hadlths attributed to the Prophet led to the 
emergence of a shared three-tiered process of authentication among 
transmission-based scholars in Medina, Basra, Baghdad and Naysabur. 
In the first tier, scholars such as Abu Dawud al-TayalisI (d. 204/818) 
and Ibn Hanbal strove to anchor core doctrine and practice in the 
teachings of the Prophet. They thus compiled collections limited to 
reports possessing explicit chains of transmission (isnad) going back to 
Muhammad. These musnad collections would have proven a very effec- 
tive first line of defense against material entering the Islamic tradition 
from outside sources; Ibn Hanbal and other early transmission-based 
scholars paid no heed to material lacking an isndd/' 

These isndds, however, could be forged or inauthentic material simply 
equipped with one and then circulated. In what constituted the sec- 
ond tier of hadith criticism, Iraqi scholars like Ibn Hanbal, Ibn Sa'd 
(d. 230/845) and 'All b. al-Madinl (d. 234/849) evaluated the quality of 
these isndds by collecting opinions about the transmitters who comprised 
them. As Scott C. Lucas has established in his study of Ibn Sa'd and 
Ibn Hanbal's work, they drew on two previous generations of hadith- 
transmission critics: that of Malik and his contemporaries like Shu'ba 



'' Yahya b. Yahya al-Laythl's recension of the Muwatta', which was transmitted to 
the West into Andalusia, contains 1,720 narrations, of which 613 are statements of 
the Companions, 285 of the Successors and 61 with no isndd at all; Abd al-Rauf, 
"Hadith Literature," 273. 

' For more on Ibn Jurayj, see Harald Motzki, "The Musannaf oi Abd al-Razzaq 
al-San'anl as a Source of Authentic Ahddith of the First Century AM." Journal of Near 
Eastern Studies 50 (1991): 1-21. 

' Al-Hakim al-Naysaburl quotes the famous early muhaddith Shu'ba b. al-Hajjaj 
(d. 160/776) as saying, "all religious knowledge {'ilm) which does not feature 'he nar- 
rated to me' or 'he reported to me' is vinegar and sprouts [khall wa baqlf; al-Hakim 
al-Naysaburl, Kitab al-madkhal ild ma'rifat Utah al-iklil, ed. Ahmad b. Faris al-Sulum 
(Beirut: Dar Ibn Hazm, 1423/2003), 58, 



52 CHAPTER THREE 

b. al-Hajjaj (d. 160/776), and that of the next generation of the great 
Basran critics 'Abd al-Rahman b. Mahdl (d. 198/814) and Yahya b. 
Sa'id al-Qattan (d. 198/813).^ Ibn Sa'd amassed a huge dictionary 
of hadith transmitters, his Tabaqat, which included statements from 
respected hadith authorities rating transmitters for honesty, piety and 
their command of the material they conveyed. In addition, works like 
the Tabaqat and 'All b. al-Madlni's 'Hal also tried to ascertain the per- 
sonal links between different narrators in order to assure the continuity 
of transmission and establish the most secure links to the Prophet. A 
liar, a forgetful person or a break in the isndd could thus weaken the 
reliability of a hadith. 

Finally, the third tier consisted of demanding corroboration for 
hadlths being circulated among the network of hadith transmitters that 
spread from Yemen to Transoxiana. Even though a hadith narration 
might possess a sound isndd, it was considered unreliable if only one 
out of several students of a famous transmitter reported it from him. 
Reports that either conflicted with similar reports or lacked corrobora- 
tion were deemed likely errors. A genre of books identifying these 'Hal 
(flaws) thus arose with the work of 'All b. al-Madlnl and Ibn Hanbal. 

A hadith that passed this three-tiered test was considered sahih, or 
an authentic saying of Muhammad.'" Although scholars applied these 
three tiers of criticism to their corpora of hadlths, they did not dispense 
with weaker material or require a report to be sahih in order to function 
in deriving laws. Ibn Hanbal's massive Musnad of approximately thirty 



' See Scott C. Lucas, Constructive Critics: Hadith Literature and the Articulation of Sunni 
Islam (Leiden: Brill, 2004). 

'" There has been some contention over the proper translation of the term sahih. 
G.H.A. JuynboU renders it 'sound' or 'healthy.' Although this translation adheres most 
closely to the etymological root of the word, in her dissertation on the term sahih, Asma 
Helah Miiller argues that 'authentic {authentigue}' more accurately represents the term's 
significance in the hadith worldview. The tradition of hadith criticism, she explains, was 
ultimately wholly preoccupied with the ideal of 'authenticity {sihhdf and establishing it. 
Declaring a report sahih was thus to announce that its authenticity had been determined, 
and hadith critics like al-Bukharl and Muslim envisioned this as the highest rating for 
a report. Although Muslim juxtaposes the 'sound [sahiKf hadith with the 'ailing (saqirrif 
one on a semantic level, on the conceptual level this discussion occurred between the 
poles of 'authentic' and 'unauthenticated.' He thus contrasts "well-known and sahih 
reports with weak hadlths and unacceptable narrations (al-akhbdr al-sahiha al-mashhura 
with al-ahddith al-da'ifa wa al-riwaydt al-munkara)." A report rated saMh thus constituted 
the authenticated words of the Prophet; G.H.A. JuynboU, "Sahih," EP; Asma Helali 
Miiller, "Etude sur la tradition prophetique: La question de I'authenticite du I/VIP™ 
au VI/XII*""^ siecle," (Doctoral diss., I'Ecole Pratique des Hautes Etudes, 2004), 19; 
Sahih Muslim: muqaddima, introduction. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 53 

thousand hadiths represented a lifetime of collection and review, with 
the compiler adding or removing reports as he became aware of their 
strengths and weaknesses. Ibn Hanbal himself, however, admitted that 
his collection contained lackluster hadiths, which he readily employed 
in situations where no stronger reports could be found." Of course, 
that a scholar like Ibn Hanbal could suffice with a report lacking a 
strong isnad in no way entails that he was comfortable with forged 
hadiths. Ibn Hanbal himself cited a report in which the Prophet said, 
"Whoever narrates a hadlth from me thinking that it is a lie, then he 
is among the liars.'"" But any hadlth with a passable isndd back to the 
Prophet had a chance of truly being his words and thus outweighed 
any individual's feeble reasoning. Transmission-based scholars like Ibn 
Hanbal thus looked to the report attributed to 'Abdallah b. 'Abbas: 
"Indeed this knowledge is [our] religion, so incline towards hadiths as 
long as they have isndd% to your Prophet.'"' 

Here, one must not fall into the trap of conflating the epistemo- 
logical worldview of transmission-based scholars in this period with 
that of later Sunni legal theorists. As we will discuss in Chapter Five, 
later legal theorists considered the most reliable form of reports to be 
those that were mutawdtir, or reports so massively transmitted that they 
could not possibly have been forged and thus conveyed epistemological 
certainty. Ahdd hadiths, those reports that were transmitted from the 
Prophet by a less impressive number of isndds and constituted the bulk 



' ' Ibn Hanbal is reported as saying that none of the twenty-eight narrations of the 
famous hadlth in which the Prophet tells 'Ammar b. Yasir that he will be killed by the 
rebellious party {al-Ji'a al-bdghiya, i.e., Mu'awiya), several of which he includes in his 
Musnad, are correct; see Muwaifaq al-Din Ibn Qudama (d. 620/1223), al-Muntakhab 
min al-'ilal li'l-Khalldl, ed. Abu Mu'adh Tariq b. Awad Allah (Riyadh: Dar al-Raya, 
1419/1997), 222. For a famous Hanball's rebuttal of this attribution to Ibn Hanbal, see 
Ibn Rajab, Path al-bdn, ed. Mahmud Sha'ban Abd al-Maqsud et al. (Medina: Maktabat 
al-Gharaba al-Athariyya, 1417/1996), 3:310. For a more general statement on this 
from a later hadlth scholar, see Ibn al-Salah al-Shahrazurl, Muqaddimat Ihn al-Saldh wa 
Mahdsin al-istildh, ed. 'A'isha Abd al-Rahman (Cairo: Dar al-Ma'arif, 141 1/1990), 286. 
Ibn Hanbal is quoted by later scholars as saying that "if we are narrating [hadiths] 
about prohibition or permissibility {al-halal wa al-hardm) we are strict, but if we are nar- 
rating them in matters of the virtues [of the early community] and similar matters, we 
are lax"; Ibn Hajar al-'AsqalanI, al-Qawl al-musaddad fi al-dhabb 'an al-Musnad li'l-imdm 
Ahmad (Hyderabad: Da'irat al-Ma'arif al-'Uthmaniyya, 1386/1967), 12; cf al-Khatib 
al-Baghdadi, al-KifayaJi ma'rifat 'ilm usul al-riwdya, ed. Abu Ishaq Ibrahim al-DimyatI, 
2 vols. (Cairo: Dar al-Huda, 1423/2003), 1:399. 

'2 Musnad Ibn Hanbal: 5:14. 

" Ibn Adi al-jurjani, al-Kdmilji du'ajd' al-rijdl, 1 vols. (Beirut: Dar al-Fikr, 1405/1985), 
1:156. 



54 CHAPTER THREE 

of the hadith corpus dealt with by Ibn Hanbal and his cohort, only 
yielded 'legally compelling probability (zann)' in the eyes of later legal 
theorists. The epistemological category of mutawatir, however, did not 
exist in the discourse of hadith critics in the third/ninth century. The 
word Hawatara' and synonyms like Hazahara' simply meant that a hadith 
'appeared widely' and even the early Sunni legal theorist al-Shafi'l 
(d. 204/819—20) knew no technical definition for the word.'* Morever, 
the term zdnn, a positive concept for later legal theorists, was associated 
with 'other than truth' and tantamount to 'falsehood {batil)' among 
transmission-based scholars of the third/ninth century''^ For transmis- 
sion-based jurists of Ibn Hanbal's time, the ultimate epistemological 
rating a hadith could achieve was that it was 'authentically from the 
Prophet (sahha 'an al-nabi)' or 'well-known (mashhur).' 



The Sahih Movement and the Bifurcation of the Hadith Tradition 

Two of Ibn Hanbal's students found his latitude in the use of weak 
hadlths unnecessary. Muhammad b. Isma'll al-Bukharl (d. 256/870) 
and Muslim b. al-Hajjaj (d. 261/875) were the first to produce musannaf 
collections devoted only to hadlths they felt met the requirements of 
authenticity (sihhd). Their books were the first wave of what Muham- 
mad Abd al-Rauf terms "the saMh movement.""' Unlike Ibn Hanbal, 
Muslim felt that there were enough sahih hadlths in circulation that 
transmission-based scholars could dispense with less worthy narrations 
in elaborating Islamic law and doctrine.'' Such thinking represented a 



" Muslim b. al-Hajjaj al-Naysaburi, Kitab al-tamyiz, ed. Muhammad Mustafa al- 
A'zaml (Riyadh: Matba'at Jami'at Riyad, [1395/1975]), 134, 136; Muhammad b. Idrls 
al-Shafi'l, al-Umm (Cairo: Dar al-Sha'b, 1968-), 7:258—9. Al-Shafi'l does note that there 
are two kinds of reports transmitted from the Prophet, those narrated by "masses from 
masses {'dmtna 'an 'dmma)" and those narrated by individuals ikhdssa). For him, however, 
these two species of reports deal with the importance of the material they convey, not 
their epistemological strength. 'Mass' reports transmit the practices and beliefs that all 
Muslims must know, while khdss reports pertain to the more obscure questions that should 
only occupy scholars; ibid., 7:255; idem, Mawsu'at al-imdm al-Shdfi'i, ed. 'All Muhammad, 
'Adil Ahmad et als,, vol. 10 (Beirut: Dar Ihya' al-Turath al-'Arabi, 1422/2001), 154 
(this cited from al-Shafi'l's Ikhtildf al-hadith, which is included in this volume). 

'' Abuja'far al-TahawI (d. 321/933), Shark mushkit al-dthdr, ed. Shu'ayb Arna'ut, 16 
vols. (Beirut: Mu'assasat al-Risala, 1415/1994), 1:375. 

"^ Muhammad Abd al-Rauf, "Hadith Literature," 274. 

" Muslim b. al-Hajjaj, Sahih Muslim (Cairo: Maktabat wa Matba'at Muhammad 
'All Subayh, [1963]), 1:22. Al-Bukharl is also quoted as rejecting the use of non-sahih 



THE GENESIS OF AL-BUKHARI AND MUSLIM 55 

new stage in the critical study of hadlth but continued the transmission- 
based legal strain in Islamic scholarly culture. Al-Bukharl and Muslim 
made the authenticity always prized by hadlth scholars paramount in 
their books, but the works themselves were still musannafi designed for 
use as comprehensive legal and doctrinal references. 

This notion of legal and ritual utility strongly influenced other 
scholars who soon followed in al-Bukharl's and Muslim's footsteps. 
Their students and colleagues Abu Dawud al-SijistanI (d. 275/889), 
Muhammad b. Isa al-Tirmidhi (d. 279/892) and Ahmad b. Shu'ayb 
al-Nasa'l (d. 303/915),'" as well as Muhammad b. Yazid Ibn Majah 
(d. 273/887), aimed at providing collections of hadlths that combined 
this utility with high standards of authenticity. These collections none- 
theless did feature some reports that their authors acknowledged as 
weak but included either because they were widely used among jurists 
or because they, like Ibn Hanbal, could find no sahih hadlth addressing 
that particular topic. '^ Sa'id b. 'Uthman Ibn al-Sakan (d. 353/964), who 
lived mostly in Egypt, also collected a small sahih book consisting of 
hadlths necessary for legal rulings and whose authenticity he claimed 
was agreed on by all.^" 

Other contemporaries of al-Bukhari and Muslim adhered more to the 
requirement of authenticity than to legal utility. Muhammad b. Ishaq 



hadlths in issues of prohibition [tahlil wa tahnrri); Muhammad b. Ibrahim Ibn al-Wazir, 
Tanqih al-anzdrfi ma'rifat 'ulum al-dthdr, ed. Muhammad Subhl b. Hasan Hallaq (Beirut: 
Dar Ibn Hazm, 1420/1999), 72. 

'" There is some doubt as to whether al-Nasa'l studied with al-Bukharl: al-NawawI 
affirms this while al-Dhahabi says that al-Nasa'i never transmitted from al-Bukharl; 
see al-Dhahabi, Tdrikh al-isldm wa wafaydt al-mashdhir wa al-a'idm, ed. Bashshar 'Awwad 
Ma'ruf, Shu'ayb al-Arna'ut and Salih Mahdl 'Abbas (Beirut: Mu'assasat al-Risala, 
1988-present), 19:241. 

" See AbQ Dawud al-Sijistanl's letter to the scholars of Mecca, where he states that 
he alerts the reader to any hadlth with a "serious weakness {wahn shadid)"; "Risalat 
al-imam Abl Dawud al-SijistanI ila ahl Makka ft wasf Sunanihi," Thaldth msd'ilji 'ilm 
mustalah al-hadith, ed 'Abd al-Fattah Abu Ghudda (Aleppo: Maktab al-Matbu'at al- 
Islamiyya, 1417/1997), 37; Ibn Manda (d. 395/1004-5) also states that Abu Dawud 
included weak hadlths if he could find no reliable reports on a certain subject; see 
Muhammad b. Ishaq Ibn Manda, Shurut al-a'imma/RisdlaJi bajdnfadl al-akhbdr wa shark 
madhdhib ahl al-dthdr wa haqiqat al-sunan wa tashih al-riwdydt, ed. 'Abd al-Rahman b. 'Abd 
al-Jabbar al-Farlwa'l (Riyadh: Dar al-Muslim, 1416/1995), 73. 

^^ This book was called al-Muntaqd and was highly esteemed by Ibn Hazm. See 
Muhammad b. Ja'far al-KattanI, al-Risala al-mustatrafa fi baydn mashhur kutub al-sunna 
al-musharrafa, 2nd ed. (Beirut: Dar al-Kutub al-'Ilmiyya, 1400/[1980]), 20; al-Dhahabi, 
Tadhkirat al-hujfdz, ed. Zakariyya 'Umayrat, 4 vols, in 2 (Beirut: Dar al-Kutub al-'Ilmiyya, 
1419/1998), 3:231 (biography of Ibn Hazm). 



56 CHAPTER THREE 

Ibn Khuzayma (d. 311/923), an early pivot of the Shaii'i school who 
both studied with and transmitted hadlths to al-Bukhari and Muslim, 
compiled a sahih work he entitled Mukhtasar al-mukhtasar min al-musnad 
al-sahih 'an al-nabi (The Abridged Abridgement of the Sahih Musnad 
from the Prophet).^' Abu Hafs 'Umar b. Muhammad al-Bujayri of 
Samarqand (d. 31 1/924) produced a collection called al-J ami' al- sahih P 
Even the famous historian and exegete Muhammad b. Jarlr al-Tabari 
(d. 310/923) attempted a gigantic saMh musnad called Kitdb tahdhib al-dthdr 
but died before he finished it.^-' Ibn Hibban al-Bustl's (d. 354/965) mas- 
sive Sahih has been highly esteemed by Muslim scholars and is usually 
considered the last installment in the sahih movement (though three sahih 
works were evidently produced in the fifth/ eleventh century).^* 

Although in retrospect the sahih movement may appear to be a 
natural progression of the collection and criticism of Prophetic hadlths, 
it possessed an inherent elitism and a definitiveness that clashed with 
underlying characteristics of hadith transmission. Since the early days 



^' This work would later become known as Sahih Ibn Khuzayma. Al-Khallll (d. 446/ 
1054) calls this book Mukhtasar al-mukhtasar because Ibn Khuzayma had made it out of 
a bigger collection; al-Khalll b. 'Abdallah al-Khallll, al-lrshddji ma'rifat 'ulamd' al-hadith, 
ed. 'Amir Ahmad Haydar (Mecca: Dar al-Fikr, 1414/1993), 313. In his very brief 
introduction to his Sahih, Ibn Khuzayma says that this book contains material "that an 
upright {'adl) transmitter narrates from another upstanding transmitter continuously to 
[the Prophet] (s) without any break in the isndd nor any impugning [jarh) of the reports' 
transmitters"; see Abu Bakr Muhammad b. Ishaq b. Khuzayma, Sahih Ibn Khuzayma, 
ed. Muhammad Mustafa al-A'zamI, 5 vols. (Beirut: al-Maktab al-Islaml, [1970?]), 1:3. 
Al-Khapb al-Baghdadl felt that Ibn Khuzayma's collection should be ranked closely 
after al-Bukharl's and Muslim's Sah^hi because the author also demanded authenticity 
{sihhdj; al-Khatib al-Baghdadi, al-Jdmi' li-ikhtitdf al-rdwi wa dddb al-sdmi', ed. Mahmud 
Tahhan (Riyadh: Maktabat al-Ma'arif, 1403/1983), 2:185. 

^^ 'Umar b. Muhammad al-Nasaft (d. 537/ 1 142-3), al-Qandfi dhikr 'ulamd' Samarqand, 
ed. Yusuf al-Hadi (Tehran: Ayene-ye Mirath, 1420/1999), 472; al-Khallll, al-Irshdd, 
387. 

^' The full work would have included legal, linguistic and other kinds of commentary; 
see al-Dhahabi, Tadhkirat al-hujjaz, 2:202. The surviving work has been published as 
Tahdhib al-dthdr wa tafsil al-thdbit 'an Rasul Allah min al-akhbdr, ed. Mahmud Muhammad 
Shakir, 5 vols. (Cairo: Matba'at al-MadanI, 1982), idem, Tahdhib al-dthdr: al-juz' al-mafqud, 
ed. 'All Rida b. 'Abdallah (Beirut: Maktabat al-Ma'mun li'1-Turath, 1995). 

^* It is difficult to determine whether or not these works were actually collections 
devoted to authentic hadlths or just utilized the word sakih in the tide. Abu al-Qasim 
'All b. al-Muhassin al-Tanukhl (d. 407/1016), a Shiite hadith scholar, evidentiy had a 
Sahih- Ibn Hazm had a book called al-Jdmi' Ji sahih al-hadith bi-ikhtisdr al-asdnid, and Abtl 
Muhammad al-Hasan b. Ahmad al-KukhmaythanI (?) (d. 491/1098) wrote book of 
SOOjM^'s called Bah al-asdnid Ji sahih al-masdnid ihM was never studied; see al-Dhahabl, 
Siyar a'ldm al-nubald', ed. Shu'ayb al-Arna'ut (Beirut: Mu'assasat al-Risala, 1982), 17:650; 
idem, Tadhkirat al-hujfdz, 3:230 and 4:21. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 57 

of Islam, the transmission of hadiths was a means for everyday Mus- 
lims to bind themselves to the inspirational authority of the Prophet 
and incorporate his charisma into their lives."' Like all early Muslim 
scholarship, the collection and study of hadiths was not the product of 
institutions of learning; it was undertaken by devout individuals whose 
eventual knowledge and pious allure earned them positions of respect 
and authority in their communities.^^ In the late Umayyad and early 
Abbasid periods, however, a new perspective emerged in Muslim society. 
A self-aware scholarly and educated class [al-khassd) appeared which 
began distinguishing itself from the masses [al-'dmma)}' The great legal 
theorist Muhammad b. Idris al-Shafi'l (d. 204/819-20) thus divided 
knowledge of Islamic law and ritual into that which is demanded of 
the masses {'dmm) and that which is the purview of the scholars (khass). 
This bifurcation between laymen and specialists also appears in the 
introduction to Muslim's SaAih collection. Just as al-Shafi'l articulates the 
domain and duties of a scholarly elite, so too Muslim urges a special- 
ized corps of hadlth scholars to study the sunna and guide the regular 
folk, who should not concern themselves with amassing hadiths beyond 
a few authentic reports. Abu Dawud al-Sijistani evinces the same legal 
paternalism in a letter to the scholars of Mecca explaining the content 
and structure of his Sunan. He may not, he warns, alert the reader to 
all the weaknesses of a hadith because "it would be harmful to the 
masses {al-'dmma)" to reveal minor flaws that might undermine their 
faith in the report's legal applicability^'' 

Furthermore, for Muslim and Abti Dawud, their authentic collec- 
tions provided all the legal and ritual knowledge an ordinary Muslim 
required. Abu Dawtid states confidently that he knows of "nothing 
after the Qur'an more essential for people to learn than this book [his 
Sunan] , and a person would suffer no loss if he did not take in any more 
knowledge after this book."^" If the masses of Muslims should leave 



^'' For the function of Prophetic hadltii as a rehc of the Prophet, see Eerik Dickenson, 
"Ibn al-Salah al-Shahrazurl and the Isnad," 481-505. 

^^ This did not mean that one could not earn money studying hadlth. Some scholars 
asked fees for narrating hadiths, but this was the subject of much controversy in the 
scholarly community. 

^' For more on this development, see Jonathan A.C. Brown, "The Last Days of 
al-Ghazzall and the 'dmm, khdss and khdss al~khawdss of the Sufi World," Muslim World 
96, no. 1 (2006): 97 ff. 

2« Abu Dawud, "Risala," 50. 

2' Abtl Dawud, "Risala," 46. 



58 CHAPTER THREE 

the collection and criticism of hadlths to a class of specialists, and this 
elite had now provided them with definitive references, what use were 
the activities of other hadith scholars? 

This elitism and definitiveness was directed not only at the Muslim 
masses, but also at more serious hadith collectors, whose laxity in 
criticism and irresponsible leadership had motivated Muslim to write 
his Sahih in the first place. He believed that those scholars who strove 
to collect as many hadiths as possible regardless of their quality were 
doing so only to win the acclaim of the masses, who would express 
in awe, "How numerous are the hadiths so and so has collected!"''" In 
the introduction to his Sahih, Muslim expresses serious concern over 
would-be hadith scholars who transmitted material of dubious nature 
to the exclusion of well-known and well-authenticated hadiths. They 
provide this material to the common people and thus mislead them in 
their faith. It is this fact, he says, that has made him feel comfortable 
about producing a work restricted to only authentic material. ■'' It is in 
fact the duty of those who understand the science of hadith to leave the 
common folk with trustworthy reports only. To do otherwise would be a 
sin [athinf"), for the masses would believe and act on these hadiths.'^ 

The sahih movement therefore marked a departure from the main- 
stream transmission-based scholars and from the masses whose amateur 
hadith collection was a means of tying themselves to their Prophet. In 
fact, there were some who opposed the very notion of criticizing isnad's, 
and the narrators in them. Muslim addresses his Kitdb al-tamyiz (Book 
of Distinguishing) to someone who had been censured for distinguish- 
ing between sahih and incorrect hadiths, or asserting that "so and so 
has erred in his narration of a hadith." Muslim explains that these 
skeptics accuse those who attempt to distinguish between correct and 
incorrect narrations of "slandering the righteous forefathers [al-sdlihin 
min al-salaf al-mddin)" and "raising accusations [mutakharris) in things 
of which they have no knowledge, making claims to knowledge of the 
unknown ighayh) which they cannot attain. "'' 



™ Muslim, Sahih, 1:22. 

" Muslim, Sahih, 1:6. 

'^ Muslim, SaKh, 1:22. 

'' Muslim, Kitdb al-tamyiz, 123. Muslim's younger contemporary al-Tirmidhi also 
notes objections to critically evaluating narrators; Ibn Rajab, Sharh 'Ilal al-Tirmidhi, ed. 
Nur al-Din 'Itr ([n.p.]: [n.p.], 1398/1978), 1:43. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 59 

Although such an outright rejection of the ethos of the sahih move- 
ment is extreme, it differs only in degree from the practice of tradi- 
tionists like Ibn Hanbal. Reports traced back to the Prophet, bearing 
his name and conveying his authority, were, prima facie compelling.^* 
Not even a problematic isnad necessarily undermined the authority the 
Prophet commanded. Even in legal issues, scholars like Ibn Hanbal 
and Abu Dawud depended on weak or mediocre hadlths, and such 
hadlths were indispensable in fields like the history of the Prophet's 
campaigns, contextualizing Qur'anic verses or recounting the virtues 
of the Prophet's Companions.^"' 

From a modern perspective it seems difficult to understand why the 
study or legal use of hadlths did not culminate naturally in the sahih 
movement. Why would scholars elaborating law and doctrine, both 
ostensibly rooted in revelation, rely on questionable reports when they 
had authentic collections at their disposal? Answering this question 
a century after the sahih movement, the seminal systematizer of the 
hadlth tradition, al-Hakim al-Naysaburl (d. 405/1014), explained that 
using hadlths with problematic isndd'i to interpret law was an established 
practice going back as far as Abu Hanlfa. If the early Muslims had 
acted on a report from the Prophet, for example, then the fact that 
later hadith critics could not find a strong isnad for the report should 
not affect its legal reliability — practice had already proven its authen- 
ticity. Furthermore, different hadlth critics employed different criteria 
for authenticity; just because one strict scholar considered a narration 
weak does not mean that a less demanding legal scholar might not 
find it acceptable.^'' 



'■* Abu Zahra, Ibn Hanbal, 243. 

'^ Ibn Hanbal, for example, is reported not to have demanded full isndd% for hadlths 
relating to Qiar'anic exegesis, the campaigns of the Prophet (maghdzi) and apocalyptic 
prophesies (maldhim); see Ibn Taymiyya, Majmu fatdwd shaykh al-isldm Ibn Taymiyja, ed. 
'Abd al- Rahman b. Muhammad b. Qasim al-'Asimi, vol. 13 (Riyadh: Matabi' al-Riyad, 
1382/1963), 346; Ibn Rajab, Shark Hal al-Tirmidhi, 1:74. Other early scholars like 
Sufyan al-Thawrl (d. 161/778) and Sufyan b. 'Uyayna (d. 196/811) also allowed the 
use of lackluster hadlths in issues not related to obligation and prohibition {al-haldl wa 
al-hardrri); al-Khatib al-Baghdadi, al-Kifdya fi ma'rifat usul 'ilm al-riwdya, 1:398. 

"" It is important to note that such weak hadlths were problematic from the stand- 
point of hadlth scholars, not for Abu Hanlfa; al-Hakim, al-Madkhal ild ma'rifat kitdb 
al-iklil, 66-8. 



60 CHAPTER THREE 

The Continuity of the Living Isnad 

The sahih movement thus marks a bifurcation in hadith literature. In 
the wake of the sahih collections, particularly the works of al-Bukhafi 
and Muslim, the study of hadith diverged into two parallel streams 
that would clash and interact as the centuries progressed. Their rela- 
tionship with one another would remain one of tension, sometimes 
complementary and sometimes destructive, between the transmission 
of individual hadiths through living isnad'i back to the Prophet and 
the definitive and institutional power acquired by authentic hadith 
collections. The canonical destiny of the Sahihayn, the two works that 
inaugurated and epitomized the sahih movement, will be discussed in 
the following chapters. Here at the genesis of the Sahihayn, however, 
we must not allow the canonical status these works would acquire to 
distract us from their powerful alter-ego in the hadith tradition: the 
continuity of hadith transmission through the living isnad. 

The hadith tradition from which the SaMhoyn emerged remained pre- 
occupied with the continued transmission of hadiths through personal 
study long after al-Bukhari and Muslim. The strong legal and pietistic 
attachment to the living isnad of transmitters back to the Prophet con- 
tinued to drive the hadith tradition, and both the oral transmission 
of hadiths and the compilation of major non-saMh works continued 
unabated. Scholars with strong affiliation to legal schools, such as the 
Shafi'T Abu Bakr al-Bayhaqi (d. 458/1066), compiled hadith collections 
supporting their madhhab'?, positions. His massive al-Sunan al-kubrd is a 
landmark in the Shafi'i legal school, supporting its detailed case law 
with a myriad of reports from the Prophet and his Companions. Dur- 
ing the fourth/tenth century several Hanafi scholars produced musnad 
collections of the hadiths used by Abu Hanifa and his students. Even 
non-Hanafis like Abu Nu'aym al-Isbahani (d. 430/1038) participated 
in efforts to find chains going back to the Prophet for Abu Hanlfa's 
reports." The Maliki scholar Ibn al-Jabbab (d. 322/934) even created 
a musnad of Malik's hadiths.^'' 

The personal collection of hadiths expanded after and even despite 
the sahih movement, with hadith collectors amassing titanic works in the 
fourth/tenth century. Abu al-Qa.sim Sulayman al-Tabarani (d. 360/971) 



See Fuat Sezgin, Geschichte des arabischen Schnfttums, 1:414-6. 
AI-Dliahabl, Tadhkirat al-hujjai, 3:25. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 6 1 

of Isfahan compiled a huge collection, his Mu'jam al-kabir, that amounted 
to two hundred fascicules (juz').''^ He took pride in gathering rare hadlths 
found nowhere else as well as their relatively short isndds. Authenticity 
was not one of his concerns.*" 'All b. Hamshadh of Naysabur (d. 338/ 
950) produced a personal musnad twice as large as al-Tabarani's, and 
al-Hasan b. Muhammad al-MasarjisI of Naysabur (d. 365/976) com- 
piled a musnad of an astounding 1,300 fascicules.*' 

Even as late as the sixth/twelfth century, for some it was the con- 
tinued transmission of hadiths through living isndds, not the study of 
existing hadith collections, that defined the muhaddith. In his history of 
his native Bayhaq and its prominent citizens, for example, Ibn Funduq 
'All Abu al-Hasan al-BayhaqI (d. 565/1169-70) states that "a hadith 
from the Prophet (s) will be given for each of the scholars and imams, 
of hadith."*^ Even in very brief entries, Ibn Funduq usually provide a 
narration of a hadith that goes directly back to the Prophet. His focus 
on living isndds for individual hadiths dominates his Tdnkh-e Bayhaq; in 
a history largely devoted to hadith scholars, only once does he men- 
tion an actual hadith collection: the Sunan al-kubrd of the city's tower- 
ing native doyen, Abu Bakr al-Bayhaql.*^ We know that many of the 
scholars featured in Tdnkh-e Bayhaq, including Abu Bakr al-Bayhaqi, 
heard and mastered major hadith collections such as the Sahihayn. Yet 
so dominant is the role of personal transmission from the Prophet in 
the worldview of Ibn Funduq that the study or communication of such 
hadith books goes undocumented. Soon after Ibn Funduq, however, in 
the early seventh/thirteenth century, producing compilations consist- 
ing of hadiths whose isndds extended back to the Prophet generally 
ceased and scholarly energy was totally devoted to studying existing 
collections. 

These living isndds had flourished for so long, however, because they 
carried significant pietistic weight due to both their Prophetic origin 



'' AjM^' seems to have been a fascicule of about 20 folios. To contextualize what this 
meant in terms of size, Jamal al-Din al-Mizzl's (d. 741/1341) well-known biographical 
dictionary of hadith transmitters Tahdhib al-kamdl, whose present-day published form 
consists of thirty-five volumes and occupies two library shelves, was 2507W^'s; see al- 
Dhahabl, Tadhkirat al-hujfai, 4:194. 

'» Al-Dhahabi, Tadhkirat al-hujfdz, 3:85-7. 

*' Al-Dhahabi, Tadhkirat al-huffdz, 3:50, 111. 

''^ Ibn Funduq al-BayhaqI, Tdrikh-e Bayhaq (Tehran: Chapkhane-ye Kanun, 1317/ 
[1938]), 137. 

" Ibn Funduq, Tdnkh-e Bayhaq, 183. 



62 CHAPTER THREE 

and their ability to trace Muhammad's authority outward through the 
venerated heirs to his legacy The staunchly orthodox seventh/thir- 
teenth-century Sufi 'Umar al-Suhrawardi (d. 632/1234) began most of 
the chapters of his popular manual on Sufism, 'Awarif al-ma'arif, with 
hadlths whose isnads extend from him to the Prophet. Many of these 
chains reach the Prophet through major figures in the Sufi tradition, 
such as Abu al-Qasim al-Qushayri (d. 465/1072) and Abu Nu'aym 
al-Isbahanl.** 

This is not to suggest that books played no role in the continuation of 
living isnads. A hadith scholar's book could simply serve as a vehicle for 
passing on his transmitted material. Hadith collections like al-Bukharl's 
Sahth or Malik's Muwatta ' were transmitted from teacher to student in 
the same manner as individual hadiths. For hadith scholars, any refer- 
ral to such books was contingent upon hearing them from a chain of 
transmitters back to the author A book could not simply be taken 
off the shelf and used. Like a single report, only a student copying a 
text in the presence of his teacher could protect against the vagaries 
and errors of transmission.*'^ Furthermore, for hadith scholars this act 
of becoming part of the text's isndd to the author is what rendered 



" Abu Hafs 'Umar b. Muhammad al-Suhrawardi, 'Awarif al-ma'arif, ed. Adib 
al-Kamdanl and Muhammad Mahmud al-Mustafa, 2 vols. (Mecca: al-Maktaba al- 
Makkiyya, 1422/2001), 1:49, 60. 

'' Abu Bakr Ahmad b. Mahk al-Qatt'l (d. 368/979), who was die principal transmitter 
of Ibn Hanbal's Musnad from his son 'Abdallah, was severely criticized for transmit- 
ting one of Ibn Hanbal's books from a copy which he had not heard directly from his 
teacher. Although al-Qatl'l had in fact heard this book from his teacher previously, the 
copy he had used was destroyed in a flood, leaving him with only the other copy. This 
case demonstrates the sensitivity of hadith scholars to the question of aural transmission 
[samd'); even a respected scholar who had actually heard a book from his teacher could 
be criticized for relying on another copy of that same book if he had not received 
samd' for that copy; al-Khatib, Tdrikh Baghdad, 4:293-4. The scholar who transmitted 
the Musnad from al-Qatl'l, al-Hasan b. 'All Ibn al-Mudhhib (d. 444/1052-3), was also 
accused of lax transmission practices. Specifically, he did not have samd' for certain 
sections of the Musnad. Al-Dhahabi (d. 748/1348) thus explains that, because of this, 
"material with unrehable texts [matn) and isndds entered into the Musnad"; al-Dhahabi, 
Mizdn al-i'tidalfi naqd al-rijdl, ed. 'All Muhammad al-BajawI, 4 vols. ([Beirut]: Dar Ihya' 
al-Kutub al-'Arabiyya, n.d. Reprint of the Cairo edition published by 'Isa al-Babi al- 
Halabl, 1963-4), 1:51 1-12. Another fourth/tenth-century scholar, the Hanballlbn Batta 
(d. 387/997), was also criticized for poor 5am«' practices. A scholar who had received 
Abu al-Qasim al-Baghawl's (d. 3 1 7/929-30) Mu'jam al-sahdba through Ibn Batta refused 
to grant any hadlths he found in that book a sahih rating because Ibn Batta's isndd to 
the book's author was broken. This demonstrates the continuity between the isndds, in 
a book and the isndds, to a book in this period — as al-Dhahabi points out, a problem 
in the manner in which a book was transmitted affected the reliability of the material 
in the book; al-Khatib, Tdrikh Baghddd, 10:373. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 63 

the book legally compelling. Speaking from this transmission-based 
perspective, AbQ Bakr Muhammad b. Khayr al-IshbilT (d. 575/1179) 
said that no one could introduce a statement with the formula "the 
Prophet said..." without possessing some personal chain of transmis- 
sion back to the Prophet for that report.**" Scholars like al-Q_ushayri 
and al-Isbahanl, through whom al-Suhrawardi linked himself by isnad 
back to the Prophet, had recorded their hadlths in book-form. The 
religious capital gained by providing living ismdi for hadiths transmit- 
ted through them, however, proved more compelling to al-Suhrawardi 
than simply citing their books. 

The tension between this centrality of living transmission for hadlth 
books and the emerging independent authority of the sahih collections 
had important implications for the development of legal institutions 
in the fifth /eleventh century. In this period (and later on), both jurists 
and hadlth scholars found it necessary to respond to the question, "If 
you find a well-authenticated copy of a sahih collection, can you act 
on or transmit its contents?" Summarizing the majority opinion of the 
transmission-based scholars, Majd al-Din Ibn al-Athir (d. 606/1210) 
states that in the absence of a formal transmission of the text (samd'), 
one should neither narrate any of the book's contents to others nor 
feel obligated to act on its legal implications.*' Without transmission, 
the text simply had no power 

Scholars articulating legal theory (usul al-fiqh) and the majority of 
Sunni jurists disagreed totally with this transmission-based stance. 
Acknowledging the prohibition of the muhaddithun, the great Shafi'i 
jurist and theologian Abu Hamid al-Ghazali (d. 505/ 1111) asserts that 
one can utilize a hadlth collection even without hearing it through an 
isndd.*^ Here he follows his teacher Imam al-Haramayn 'Abd al-Malik 
al-Juwaynl (d. 478/1085), who states that if a hadlth appears in Sahih 



'"' Muhammad b. Khayr al-Ishbfll, Fahrasat ma rawdhu 'an shuyukhihi min al-dawdwin 
al-musannaja fi dumb al-'ilm wa anwd' al-ma'drif {BeimV. al-Maktab al-Tijarl, 1963), 17. 
On the issue of the orality of knowledge in Islamic civilization and its tension with 
the written book, see BuUiet, Islam: The View from the Edge, 13-22; Paul L. Heck, "The 
Epistemological Problem of Writing in Islamic Civilization: al-Khatib al-Bagdadl's 
(d. 463/1071) Taqyid al-'ilm,'' Studia hlamica 94 (2002): 85-1 14, esp. 96. 

" Majd al-Din al-Mubarak b. Muhammad Ibn sd-Aihir, Jdmi' al-umlji ahadith al-rasul, 
ed. 'Abd al-Qadir al-Arnaut, 15 vols. ([Beirut]: Dar al-Mallah 1389/1969), 1:88. 

'" Al-Ghazall qualifies this by demanding that the copy be well-authenticated; Abu 
Hamid Muhammad al-Ghazall, al-Mankhul min ta'liqdt al-uml, ed. Muhammad Hasan 
Hltu ([Damascus]: n.p., [1970]), 269. 



64 CHAPTER THREE 

al-Bukhari one can transmit it, act on it and ask others to do so as 
well.*^ This opinion concurs with the Maliki jurist Abu al-Walld al-Baji 
(d. 474/1081) and the vast majority of jurists and legal theorists.'''' The 
legal utility of the Sahihayn as institutions distinct from the continued 
tradition of hadlth transmission will resurface later in discussions of 
the two works' canonization. 



Reality: The Life and Works of al-Bukhan and Muslim 

Although this study focuses on the perception of al-Bukharl and Muslim 
as icons, it is important to understand the historical reality from which 
the Sahihayn romance developed. Because al-Bukharl and Muslim were 
eventually canonized, any accurate portrait of them in their own context 
must depend on the earliest possible sources and on the evidence they 
themselves left behind. As we will see later in Chapter Seven, it was 
not untU the beginning of the fifth /eleventh century that a canonical 
culture formed around al-Bukharl and Muslim. By referring to their 
own works and consulting early biographies that preceded this shift 
towards hagiography we can broadly outline al-Bukhari's and Muslim's 
careers as well as the immediate reactions to their work. 

Very brief biographies or references to al-Bukhari and Muslim ap- 
pear in fourth/tenth century works such as Ibn Abi Hatim al-Razl's 



'^ Imam al-Haramayn 'Abd al-Malik al-Juwaym, Kitab al-burhan fi usul al-fiqh, ed. 
'Abd al-'AzIm al-Dlb, 2 vols. (Cairo: Dar al-Ansar, 1400/[1980]), 1:647^ 

''" Abu al-Walld Sulayman b. Khalaf al-Bajl al-Qurtubl, al-hhdra Ji usul al-fiqh, ed. 
'Adil Ahmad 'Abd al-Mawjud and 'All Muhammad 'Awad (Riyadh: Maktabat Nizar 
Mustafa al-Baz, 1418/1997), 162-3; Speaking on behalf of all jurists (fuqaha), Ahmad 
b. 'All Ibn Barhan al-Shafi'l (d. 518/1 124) repeats al-Ghazall's above quote. Al-SuyutI 
(d. 9 1 1 / 1 505) states that the earlier Shafi'l/ Ash'arl legal theorist Abu Ishaq al-Isfaraylnl 
(d. 418/1027) claimed a consensus on this stance. There is also a report from al-Shafi'i 
himself allowing this; Muhammad b. 'Abd al-Rahman al-SakhawI, Fath al-mughith, ed. 
'All Husayn 'All, 5 vols. (Cairo: Maktabat al-Sunna, 1424/2003), 1:83; Jalal al-Din 
al-SuyutI, Tadrib al-rdwi fi shark Taqrib al-Nawdwi, ed. 'Abd al-Wahhab 'Abd al-Latif, 
3rd ed. (Cairo: Maktabat Dar al-Turath, 1426/2005), 119. Ibn al-Salah, however, 
reports that some Maliki scholars reject narrating from a hadlth book for which one 
lacks sama; Ibn al-Salah, Muqaddimat Ibn al-Saldh, 360; see also Ibn al-WazIr, Tanqih 
al-anidr, 241-2. Al-Khatib al-Baghdadi, in an apparent attempt to bridge the gap 
between hadlth scholars and jurists, provides no definitive stance in his al-KifayaJi 'ilm 
al-riwdja. He includes many citations from early masters like Waki' b. Jarrali and Ibn 
Sirln condemning even reading a book without having heard it from a trustworthy 
transmitter, but notes that many have allowed this; al-Khatib al-Baghdadi, al-Iafdya Ji 
ma'rifat 'ilm usul al-riwdya, 2:358-6. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 65 

(d. 327/938) al-Jarh wa al-ta'dil, Ibn Hibban's (d. 354/965) Kitdb al- 
majruhin, and Ibn al-Nadlm's (d. after 385-8/995-8) al-Fihrist. More 
detailed early information for al-Bukhari's life and career occurs in 
sources like Ibn 'Adi al-Jurjanl's (d. 365/975-6) two books: al-KdmilJi 
du'ajd' al-rijdl and As ami man rawd 'anhum Muhammad b. Ismd'U al-Bukhdn 
min mashdyikhihi alladhina dhakarahum Jt Jdmi'ihi al-sahih. For both al- 
Bukharl and Muslim, the Tdnkh Naysdbur of al-Hakim al-Naysaburi 
(d. 405/1014) provides our earliest comprehensive source. Although 
now lost, this work is quoted at length by al-Khatib al-Baghdadi 
(d. 463/1071) in his Tdnkh Baghddd and by Shams al-Din al-Dhahabi 
(d. 748/1348) in his Tdnkh al-isldm. Fragments of Tdnkh Naysdbur sur- 
vive in an eighth/fourteenth-century abridgement by Muhammad b. 
al-Husayn Khalifa (fl. 720/1320).''' But since al-Hakim was one of the 
central figures in the canonization of the Shaykhayn (the 'two shaykhs,,' 
an honorific for al-Bukharl and Muslim), we must be very wary of 
relying on his work for reconstructing pre-canonical perceptions of 
the Sahihayn. Unfortunately, with regard to Muslim, he represents the 
only real source for early information. Both Muslim and al-Hakim 
were citizens of Naysabur, however, and al-Hakim's father met the 
great traditionist. We may thus feel comfortable relying on al-Hakim 
in outlining Muslim's life and work in their native city. 



Reality: al-Bukhan, Sahib al-Sahih 

Abu 'Abdallah Muhammad b. Isma'll b. Ibrahim b. al-Mughira b. Bar- 
dizbeh al-JuTi al-Bukhari was born in Bukhara in 194/810. His fam- 
ily were wealthy landowners [dehqdn), and his great-grandfather had 
converted to Islam from Zoroastrianism at the hands of Yaman al-JuTi, 
the Arab governor of the city" Al-Bukhari himself lived off proper- 
ties he rented out on a monthly or yearly basis." He started studying 



^' Al-Hakim al-Naysabun, recension and translation by Mohammad b. Hosayn 
Kliallfe-ye Nishaburl, Tdnkh Kshdbur, ed. Mohammad Reda Shaft'l KadkanI (Tehran: 
Agah, 1375/[1996]). 

^^ Abu Ahmad 'AbdaUah Ibn 'Adi al-jurjani, Asdmi man rawd 'anhum Muhammad b. 
Ismd'il al-Bukhdri min mashdyikhihi alladhina dhakarahum Jt Jdmi'ihi al-sahih, ed. Badr b. 
Muhammad al-'Ammash (Medina: Dar al-Bukharl, 1415/[1994-5]), 59. 

^' Al-Dhahabi cites Muhammad b. Abl Hatim al-Warraq, al-Bukharl's secretary, 
as saying that al-Bukharl had a piece of land that he would rent every year for 700 
dirhami. He quotes al-Bukharl as saying: "I used to acquire (astaghillu) every month 



66 CHAPTER THREE 

hadith at a young age, learning from local Bukharan experts, and in 
his late teens he began writing books on the sayings of the Compan- 
ions and the Successors. His pilgrimage to Mecca at age sixteen was 
the beginning of a long career of traveling that connected him to the 
most vaunted hadith scholars of his day In Khurasan he visited Balkh, 
Merv and Naysabur, where he studied with Ishaq b. Rahawayh (d. 238/ 
853). In western Iran he stayed in Rayy and made numerous trips to 
Baghdad, where he studied with Ibn Hanbal and Yahya b. Ma'in. In 
Basra he heard from 'All b. al-Madini, who would become one of his 
main teachers, and Abu 'Asim Dahhak al-Nabll (d. 212/827). He also 
studied in Wasit, Kufa and Medina. In Mecca he heard from 'Abdallah 
b. al-Zubayr al-Humaydi (d. 219/834), and also went to Egypt and cit- 
ies like 'Asqalan and Hims in greater Syria. There is some debate on 
whether he visited the cities of upper Mesopotamia (al-Jazira),''* and 
it is unclear whether he reached Damascus.'"' 

In his Tdrikh Naysabur, al-Hakim al-Naysaburl reports that al-Bukhari 
arrived in Naysabur for the last time in 250/864—5. Later Muslim 
sources convey the impression that he quickly gained the enmity of 
Naysabur's senior hadith scholar, Muhammad b. Yahya al-Dhuhli 
(d. 258/873), who had him expelled from the city due to his statement 
that the physical recitation [lafz) of the Qur'an was created. Indeed, we 
do know from Ibn Abi Hatim al-Razi's (d. 327/938) al-Jarh wa al-ta'dil, 
our earliest source on al-Bukharl, that al-Dhuhli publicly condemned 
al-Bukhari for his beliefs about the lafz of the Qur'an.''^ Furthermore, 
our sources are also unanimous that al-DhuhlT used this as a pretext 
to demand al-Bukhari's expulsion from Naysabur 

Early information from al-Hakim and Ibn 'Adi, however, suggests 
that the tension between al-Bukhari and al-Dhuhli was multifaceted 



500 dirhams, and I spent it all in the quest for knowledge"; al-Dhahabi, Tankh al-islam, 
19:263-4; Ibn Hajar, Hady al-sdn, ed. Muhammad Fu'ad 'Abd al-BaqI and 'Abdallah 
b. 'Ubaydallah b. Baz (Beirut: Dar al-Kutub al-'Ilmiyya, 1418/1997), 664. 

''* Al-Subkl cites his teacher al-Mizzl's rejection of al-Hakim's claim that al-Bukharl 
had entered the Jazira and heard from people like Isma'll b. 'Abdallah b. Zurara al- 
Raqql; Taj al-Din 'Abd al-Wahhab b. 'All al-Subkl, Tabaqdt al-shdfi'ijya al-kubrd, ed. 
Mahmud Muhammad al-Tanahl and 'Abd al-Fattah Muhammad al-Halw, 10 vols. 
([Cairo]: 'Isa al-Babi al-Halabl, 1383-96/1964-76), 2:214.' 

'"' Ibn 'Asakir lists al-Bukhari in his history of Damascus. For more on al-Bukhari's 
teachers, see Fuat Sezgin, Buhdri'nin Kaynaklan (Istanbul: Ibrahim Horoz Basimevi, 1956); 
A.J. Arberry, "The Teachers of Al-Bukhari," Islamic Quarterly 11 (1967): 34—49. 

^^ 'Abd al-Rahman Ibn Abl Hatim al-RazI, al-Jarh wa al-tadil, 6 vols. (Hyderabad: 
Da'irat al-Ma'arif al-'Uthmaniyya, 1959), 4:1:182-3. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 67 

and grew over some time. The earliest detailed report mentioning the 
lafi scandal, given by Ibn 'Adi, includes no mention of al-Dhuhli or 
of al-Bukharl's expulsion. It certainly portrays al-Bukhari falling into 
disfavor with hadith scholars due to his views on the Qur'an, but con- 
cludes with him retiring to his residence in Naysabur, not leaving the 
city. This is not surprising, as al-Hakim states that al-Bukhari's last stay 
in Naysabur was lengthy, lasting five years.^' 

Ibn 'Adi furnishes another reason for al-Dhuhll's animosity towards 
al-Bukharl. He reports third-hand from al-Dhuhll's son, Haykan b. 
Muhammad al-Dhuhlf'*' (d. 267/881), that he asked his father: 

What is with you and this man — meaning Muhammad b. Isma'il — when 
you are not one of those from whom he transmits iwa lasta min rijdlihi fi 
al-'ilm)? He said, "I saw him in Mecca and he was following Shamkhada" 
(Ibn 'Adi: Shamkhada is a Kufan Qadarite). When I reached [al-Bukhan], 
he said, "I entered Mecca and I didn't know anyone from among the 
hadith scholars, while Shamkhada knew them, so I would follow him so 
that he would acquaint me with them; so what is the shame in that?"^" 

Interestingly, with the exception of the encyclopedic Ibn 'Asakir (d. 571/ 
1 176), Ibn 'Adi's report appears in none of the later sources, and there 
is no evidence that Ibn 'Adi's younger contemporary al-Hakim took 
it into consideration in his discussion of al-Bukhari's relationship with 
al-Dhuhli.'''' Since later apologists for al-Bukhari never acknowledged 
the report, and it was the laf^ scandal and not this accusation which 
attracted detractors, we have no reason to doubt the provenance and 
veracity of Haykan's story. It thus seems likely that the Iqfz incident 
was not the immediate cause of al-Dhuhli's dislike for al-Bukhari or 
of the latter's expulsion. It was merely a pretext, the last episode in an 
aversion that al-Dhuhli had developed for al-Bukhari earlier during his 
lengthy tenure in Naysabur 

After his consequent expulsion from Naysabur, al-BukharT returned 
to his native Bukhara in what would prove to be the last year of his life. 
He was soon driven from there as well. The Tahirid amtr of Bukhara, 
Khalid b. Ahmad (coincidentally also surnamed al-Dhuhll), entertained 
many hadith scholars, such as Muhammad b. Nasr al-MarwazI (d. 294/ 



'■' As cited by al-Dhahabi; al-Dhahabi, Tankh al-islam, 19:250. 
'" Khatib, Tdrikh Baghdad, 14:220. 
™ Ibn 'Adi, Asdmi, 66-7. 

''" Ibn Asakir, Tdnkh madinat Dimashq, ed. Muhibb al-Din Abu Sa'ld 'Umar al- 
Amrawl, 80 vols, (Beirut: Dar al-Fikr, 1418/1997), 52:95. 



68 CHAPTER THREE 

906), as guests at his court/'' He even ordered the hadith scholar Nasr 
b. Ahmad al-Kindl 'Nasrak' (d. 293/905-6) to come to his court and 
make him a musnad.'^'^ When the amir asked al-Bukhari to provide his 
children with a private reading of the Sahih and the Tdrikh al-kabir, the 
scholar refused to extend him preferential treatment. Using al-Bukharl's 
controversial stance on the Qur'an, the amir ordered his expulsion from 
Bukhara. Tired and intimidated, al-BukharT passed through the city 
of Nasaf before dying in the village of Khartank a few miles from 
Samarqand.*"^ 

Al-Bukhari's early works consisted of musings on the sayings of the 
Companions and the Successors. These writings later matured into 
a much more ambitious project. He began his al-Tdrikh al-kabir (The 
Great History) while a young man in Medina. The extant work is a 
massive biographical dictionary of over 12,300 entries.''* He is reported 
to have revised it at least three times over the course of his life, as 
Christopher Melchert corroborates in his analysis of the Tdrikh.^''' Al- 
Bukharl consistently provides neither full names nor evaluations of the 
persons in question, focusing instead on locating each subject within 
the vast network of hadith transmission. The Tdrikh seems to have 
no connection to the Sahih.'-''' Al-Bukhari produced another smaller 
dictionary of hadith transmitters, one large book of weak transmitters 
[Kitdb al-du'afd' al-kabir, now lost) as well as a smaller book on weak 
narrators.'" In addition, he wrote several smaller topical works, such 
as his Khalq afdl al-'ibdd (On the Createdness of Men's Actions) and 
Kitdb raf' al-yadaynfi al-saldt (Book on Raising One's Hands in Prayer). 



" Ibn al-Jawzi, al-Munta^am, 12:225-6. 

''^ Al-Khatib, Tdrikh Baghdad, 8:310-1 1 (biography of Khalid b. Yahya); Ibn al-JawzI, 
al-Muntazam, 13:48. 

I*' J. Robsoii, "al-Bukharl," EP. 

" Melchert, "Bukharl and Early Hadith Criticism," Journal of the American Onental 
Soeiety, 121, no. 1 (2001): 8. Oddly, extant copies of al-Tdrikh al-kabir feniure no female 
transmitters. Al-Hakim, however, quotes Abu 'All al-Husayn al-MasarjisI as saying that 
the book contains approximately forty thousand (sic!) "men and women." It thus seems 
likely that at some crucial point in the transmission of our extant manuscript tradition, 
a last volume containing women was lost. See al-Hakim, al-Madkhal ild al-sahih, ed. Rabi' 
b. Hadi 'Umayr al-Madkhall (Beirut: Mu'assasat al-Risala, 1404/1984), 111. 

''•'' See Melchert, "Bukharl and Early Hadith Criticism," 9; al-Khatib, Tdrikh Baghddd, 
2:7. 

^'^ Melchert, "Bukharl and Early Hadith Criticism," 12. 

'*' Al-Bukharl's Tdrikh al-awsat and his Kitab al-du'afd' al-saghir have both been published 
in several editions. Al-Dhahabi notes his Kitdb al-du'afd' al-kabir, now lost; al-Dhahabi, 
Mizdn al-i'tiddl, 2:570, 598; 3:31 1. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 69 

Some reports indicate that al-Bukharl produced an 'Hal book as well 
as a large musnad, both now lost/'" 

a. The Sahlh 

Al-Bukharl's Sahth, actually titled al-Jdmi' al-musnad al-sahih al-mukhtasar 
min umur Rasul Allah wa sunanihi wa ayydmihi (The Abridged Authentic 
Compilation of the Affairs of the Messenger of God, his Sunna and 
Campaigns),''^ was a mammoth expression of his personal method of 
hadlth criticism and legal vision. It covers the full range of legal and 
ritual topics, but also includes treatments of many other issues such 
as the implication of technical terms in hadith transmission and the 
authority of dhdd hadlths (reports transmitted by only a few chains of 
transmission) in law.™ The Sahih consists of ninety-seven chapters (kutub, 
sing, kitdb), each divided into subchapters (abwdb, sing. bdb). The sub- 
chapter titles indicate the legal implication or ruling the reader should 
derive from the subsequent hadiths, and often include a short comment 
from the author'' Such short legal discussions often feature hadlths not 
naming al-Bukhari's immediate source (termed ta 'liq or hadith mu 'allaq) 
or a report from a Companion for elucidation. Al-Bukharl often repeats 
a Prophetic tradition, but through different narrations and in separate 
chapters. Opinions have varied about the exact number of 'hadlths' in 
the Sahth, depending on how one defines 'hadith': e.g. as a 'tradition' (a 
saying attributed to the Prophet) or a 'narration' (one version of that 
saying narrated by a specific isndd). Generally, experts have placed the 
number of fuW-isndd narrations at 7,397, with Ibn Hajar (d. 852/1449) 
counting a total of 9,082 including all the incomplete isndds. Of these 
around 4,000 are repetitions, placing the number of Prophetic traditions 



''" Ibn Hajar, Hadj al-san, 679. 

''' Abu Nasr Ahmad al-Kalabadhi, Rijdl Sahih al-Bukhdn, ed. 'Abdallah al-Laythi, 
2 vols. (Beirut: Dar al-Ma'rifa, 1407/1987), 1:23. For a discussion of the tide of the 
Sahih, see 'Abd al-Fattah Abu Ghudda, Tahqiq ismay al-Sahihayn wa ism Jdmi' al~Tirmidhi 
(Aieppo: Maktab al-Matbu'at al-Islamiyya, 1414/1993), 9-12. 

'" Ibn Hajar al-'AsqalanI, Fath al-bdri sharh S^hih al-Bukhdri, ed. 'Abd al-'AzIz b. 
'Abdallah b. Baz and Muhammad Fu'ad 'Abd al-BaqI, 15 vols. (Beirut: Dar al-Kutub 
al-'Ilmiyya, 1418/1997), 1:191-2; Sahih al-Bukhdri: kitdb al-'ilm, bdb 4; and Fath al-bdri, 
13:302, #7267; Sahih al-Bukhdri: kitdb akhbdr al-dhdd, bdb 6. 

" The best discussion to date of the nature of al-Bukharl's legal commentary 
is Mohammad Fadel's "Ibn Hajar's Hady al-Sdri: A Medieval Interpretation of the 
Structure of al-Bukharl's al-Jdmi' al-SMh: Introduction and Translation," Journal of 
Near Eastern Studies 54 (1995): 161-197. 



70 CHAPTER THREE 

between 2,602 (Ibn Hajar's lowest count) and the more widely accepted 
figure of 3,397-4,000.'2 

Unlike Muslim, al-Bukharl provides no methodological introduc- 
tion to his Sahih. As we shall see in Chapter Five, later scholars spilled 
a great deal of ink attempting to reconstruct his requirements [rasm 
or shurut) for authenticity [sihhd) from his Sahih and al-Tdnkh al-kabir. 
Except for some statements gleaned from his extant works, however, 
our understanding of al-Bukharl's methods depends totally on either 
these later analyses or on statements attributed to al-Bukhari in later 
sources.'^ It is generally believed that in his Sahih al-Bukhari followed 
his teacher 'All b. al-Madini in requiring some proof that at each link 
in the isnad the two transmitters had to have narrated hadlths to one 
another in person at least once. Later scholars like al-Qadi 'lyad b. 
Musa (d. 544/1149) verified this by locating an occurrence of "he 
narrated to us [haddathanaj" between every two transmitters at each 
link in al-Bukharl's isndds,.^* This is crucial for isndds in which transmis- 



'2 Abd al-Rauf, ''Hadith Literature," 274-5; Ibn Kathir Isma'll b. Abl Hafs (d. 774/ 
1373), al-Baith al-hathith shark Ikhtisdr ulum al-hadith, ed. Ahmad Muhammad Shakir 
(Cairo: Dar al-Turath, 1423/2003), 22. Ibn al-Salah states that al-Bukharl's book con- 
tains 4,000 Prophetic traditions (usul); Ibn al-Salah, Sijdnat Sahih Muslim min al-ikhlal wa 
al-ghalat, ed. Muwafiaq b. 'Abdallah b. 'Abd al-Qadir (Beirut: Dar al-Gharb al-lslami, 
1408/1987), 101-2; Ibn Hajar, Hady al-sdri, 648-53; MuUa Khatir, 41. 

" An example of al-Bukharl revealing his methods would be his statement in Kitdb 
raf al-jadayn that one narration adding a phrase in the matn of a hadlth (literal matn 
addition) is allowed if the narration is authentic [idhd thabatd); al-Bukharl, Mtdb raf'al- 
jadaynji al-saldt, ed. Badi' al-Din al-Rashidl (Beirut: Dar Ibn Hazm, 1416/1996), 131—3. 

'* The most exhaustive work on this issue from a medieval Muslim scholar is 
Muhammad b. 'Umar Ibn Rushayd's (d. 721/1321) al-Sarmn al-abjan wa'l-mawrid at-am'an 
fi al-muhdkama bayn al-imdmaynji al-sanad al-mu'an'an, ed. Muhammad Habib b. Khawja 
(Tunis: Matba'at al-Dar al-Tunisiyya, 1397/1977), esp. 22-32. The first scholar known 
to have attributed this stance to al-Bukharl and 'All b. al-Madlnl was al-Qadi 'lyad.; 
al-Qadi 'lyad- b. Musa, Ikmdl al-mu'lim bi-fawd'id Muslim, ed. Yahya Isma'll, 9 vols. 
(Mansura, Egypt: Dar al-Wafa', 1419/1998), 1:164. See also, Abu al-Husayn 'All b. 
Muhammad Ibn al-Qattan al-FasI (d. 628/1231), Baydn al-wahm wa al-ihdm al-wdqi'ayn 
Ji kitdb al-Ahkdm, ed. al-Husayn Ayat Sa'ld, 5 vols. (Riyadh: Dar al-Tayba, 1418/1997), 
2:576-7. Several modern Muslim scholars have devoted extensive studies to the question 
of al-Bukharl's and Muslim's stances on hadlths transmitted by 'an. In his innovative 
work Ijmd' al-muhaddithin 'aid 'adam ishtirdt al-'ilm bi'l-samd' Ji al-hadith al-mu'an'an bayn 
al-muta'dsinn, al-Sharif Hatim al-'Awnl argues that al-Bukharl never actually required 
proof of personal contact, but that this had been incorrectly inferred by al-Qadi lyad, 
whose conclusion about al-Bukharl's methods were parroted uncritically by virtually all 
later scholars of hadlth; al-Sharif Hatim al-'Awnl, Ijmd' al-muhaddithin 'aid 'adam ishtirdt 
al-'ilm bi'l-samd' fi al-hadith al-mu'an'an bayn al-muta'dsirin (Beirut: Dar 'Alam al-Fawa'id, 
142 1 /2001). See also, Khalid Mansur 'Abdallah al-Durays, Mawqif al-imdmayn al-Bukhdn 
wa Muslim min ishtirdt al-luqyd wa al-samd 'fi al-sanad al-mu 'an 'an bayn al-muta 'dsirin (Riyadh: 
Maktabat al-Rushd and Sharikat al-Riyad, 1417/1997). 



THE GENESIS OF AL-BUKHARI AND MUSLIM 7 1 

sion is recorded by the vague phrase "from/on the authority of {'an)." 
Unlike the transmission terms "he narrated to us" or "he reported to 
us (akhbarana)," "from/on the authority of" could be used by someone 
who never met the transmitter of the hadith in question. This means 
that in al-Bukharl's Sahth any isnad with "from/on the authority of {'an) 
so and so" in the isnad is theoretically equivalent to "so and so narrated 
to us directly." 

b. Legal Identity and Method 

Al-Bukhari never explicitly adhered to any of the nascent schools of 
law, though he was eventually claimed by all four madhhabs. He studied 
with several scholars closely associated with al-Shafi'i, like al-Husayn 
al-Karabisi (d. 245/859) and Abu Thawr (d. 240/854). Although al- 
Bukhari never narrates hadiths through al-Shafi'i, the Shafi'i biographers 
Abu 'Asim Muhammad al-'Abbadi (d. 458/1066) and Taj al-Din al- 
Subkl (d. 771/1370) use these scholarly links to tie al-Bukhari to the 
school's founder'^ Ibn Abi Ya'la al-Hanball (d. 526/1 131-2) claims al- 
Bukharl was a Hanball because he transmitted hadiths and legal rulings 
from Ibn Hanbal, and some Malikis have considered him one of their 
own because he transmitted the Muwatta'. Even later Hanafis claim 
al-Bukhari, since they argue that one of his teachers, Ibn Rahawayh, 
was Hanafi.'*' 

An examination of al-Bukharl's Sahih reveals that he was an inde- 
pendent scholar unconstrained by any particular school." In contrast 
to all four Sunni schools of law, he allows those who have had sexual 
intercourse {junub) during the Ramadan fast to expiate their sin by 
performing charity but does not require them to repeat the day of 
fasting. In another break with the schools, he allows someone who has 
had intercourse and not performed ablutions to read the Qur'an.™ He 



'^ Abu 'Asim Muhammad b. Ahmad al-'Abbadi, Kitab Tabaqat al-Fuqaha' as-Safi'iyya, 
ed. Gosta Vitestam (Leiden: Brill, 1964), 53-4; al-Subkl, Tabaqat al-shdfi'iyya al-kubrd, 
2:214. 

"" Abu al-Husayn Muhammad Ibn Abl Ya'la, Tabaqat al-handbila, ed. Abu Hazim 
Usama b. Hasan and Abu al-Zahra' Hazim 'All Bahjat, 2 vols. (Beirut: Dar al-Kutub 
al-'Ilmiyya, 1997), 1:254-9; al-Husaynl 'Abd al-Majld Hashim, al-Imdm al-Bukhari 
muhaddith" wafaqih"" (Cairo: Misr al-'Arabiyya, n.d.), 167. 

" J. Robson agrees in his entry on al-Bukharl; see J. Robson, "al-Bukharl, Muham- 
mad b. Isma'll," EP. 

"^ Hashim, al-Imdm al-Bukhdri muhaddith"" wafaqih"", 190-1. 



72 CHAPTER THREE 

also permits reading the Qiar'an in the bathroom, declares 'umra to be 
mandatory just like hajj, and allows women not to veil themselves [ihtijab) 
in the company of slaves.'^ 

Al-Bukhari obliquely sets forth his legal methodology in the penul- 
timate chapter of the Sahih, the Kitdb al-i'tisam bi'l-kitab wa al-sunna (the 
Book of Clinging to [God's] Book and the Sunna).^" From the author's 
often detailed subchapter headings and the Prophetic and Companion 
traditions that he includes, the reader gleans a minimalist approach to 
law closely tied to the revealed sources. The Prophet was sent with the 
totality of guidance to mankind, and adhering to his message is the key 
to salvation. The precedent in the community, from the time of the 
first caliph Abu Bakr, is not to deviate from the Prophet's sunna. The 
next subchapter, however, is entitled "Concerning what is hated about 
asking too many questions," including a hadlth in which the Prophet 
states that the believer's greatest crime is to inquire about something 
previously unmentioned and thus cause its prohibition for the whole 
community"' Al-Bukhari's opposition to the use of excessive legal 
reasoning and speculation manifests itself in his subchapters on "the 
condemnation of ra'y and excessive qiyas {takalluf al-qiyds)'" and how the 
Prophet himself would not answer a question until God had revealed 
the answer to him."^ Al-Bukhari does, however, allow limited analogi- 
cal reasoning based on the Prophet's answer to a man who refused to 
acknowledge a black child to whom his wife had recently given birth. 
The Prophet enlightens the man by asking him rhetorically if his camels 
are always the same color as their parents."^ 

In the conflict between the ahl al-hadith and the ahl al-ra'y, al-Bukhari 
clearly identified himself with the transmission-based jurists. In the 
Sahih, he uses his chapter headings and brief comments to differ on 



" 'Abd al-Khaliq 'Abd al-Ghani, al-Imam al-Bukhan wa Sahihuhu (Jedda: Dar al- 
Manara, 1405/1985), 146. 

"" For an in-depth discussion of this chapter, see Scott C. Lucas, "The Legal Principles 
of Muhammad b. Isma'll al-Bukharl and their Relationship to Classical Salafi Islam," 
Islamic Law and Society 13, no. 3 (2006): 291 If. 

" Ibn Hajar, Fath al-bdn, 13:328; Sahih al-Bukhari: kitdb al~i'tisdm hi'l-kitdh wa al-sunna, 
bdb 3 / #7289. 

'^ Ibn Hajar, Fath al-bdri, 13:349-359; S<ihih al-Bukhdri: kitdb al-i'tisdm bi'l-kitdb wa 
al-sunna, bdb 7-8. 

°' Ibn Hajar, Fath al-bdri, 13:366-7, #7314. This section is entided bdb man shabbaha 
asl" ma'lum" bi-asl" mubin wa qad bayyana al-Nabi (s) hukmahumd li-yafhama al-sd'il (He 
who compares a known basis {asl) to another clear basis {asl mubin), and the Prophet 
(s) has clarified their ruling so that one can understand). 



THE GENESIS OF AL-BUKHARI AND MUSLIM 73 

twenty-seven occasions with "a certain person [ba'd al-nas)." Fourteen 
of these instances occur in a chapter devoted solely to rebutting the 
use of legal devices [hiyal) employed predominantly by Hanafis to 
circumvent the literal requirements of their school's law."* Al-Bukhari 
condemns hiyal using the famous hadith that all deeds are judged by 
their intention.""' In this al-Bukharl was following the precedent of 
transmission-based jurists such as Ibn Hanbal and Ibn al-Mubarak 
(d. 181/797), who vehemently rejected the use of hiyal}^' Since the 
positions he rejects are associated with the Hanafi school, it seems 
almost certain that al-Bukhari was referring to Abu Hanifa. Al-Bukhari, 
for example, disagrees with the well-known Hanafi laxity on defining 
intoxicants: he considers tild' (reduced grape juice) to be a type of wine 
(nabtdh), while Hanafis do not."^ 

Outside his Sahih, however, al-Bukhari's disagreement with Abu 
Hanifa and the ahl al-ra'y manifests itself in virulent contempt. He 
introduces his Kitdb raf al-yadaynfi al-saldt as "a rebuttal of he [man) who 
rejected raising the hands to the head before bowing" in prayer and 
"misleads the non-Arabs on this issue [abhama 'aid al-djam fi dhdlikd) . . . 
turning his back on the sunna of the Prophet and those who have 



"' 'Abd al-Ghanl al-Ghunayml al-Maydanl al-Dima.shqi (d. 1298/1880-1), Kashf 
al-iltibds 'ammd awrada al-imdm al-Bukhdri 'aid bad al-nds, ed. 'Abd al-Fattah Abu Ghudda 
(Aleppo: Maktab al-Matbu'at al-Islamiyya, 1414/1993), 19; see Ibn Hajar, Path al-bdn, 
12:404-425. 

'^'' Ibn Hajar, Path al-bdri, 12:405; Sahih al-Bukhdri: kitdb al-hiyal, hdb 1. For a recent 
discussion of hiyal in the Hanafi school and Islamic legal thought in general, see Satoe 
Horii, "Reconsideration of Legal Devices (hiyal) in Islamic Jurisprudence: The Hanafis 
and their 'Exits' {rmLkhs.nj}," Journal of Islamic Law and Society, 9, no. 3 (2002): 312-357. 
The author describes how the Hanaft tradition used hiyal to provide people means by 
which to escape the more difficult sanctions of law in everyday life. It is also probable, 
in my opinion, that the emphasis that the early Hanafis placed on the formal structure 
of qiyds, where the ruling must inhere whenever its immediate cause {'ilia) appears, 
made hiyal attractive. They allowed scholars to preserve the logical continuity of the 
qiyds system while avoiding some of its admittedly unjust or unfairly difficult results; a 
scholar could maintain the system of qiyds by acknowledging that the ruling inhered 
in the case, but then use a hila to deal more justly with it. The two manners in which 
hiyal were misunderstood by their opponents, that they were a means to cheat God's 
law or that they represented inappropriate rational gymnastics, would both have 
offended al-Bukharl. 

"^ Al-Khatib, Tdrikh Baghddd, 13:404 (biography of Abu Hanifa), where Ibn al- 
Mubarak is quoted as saying, "Whoever looks into the Book of hiyal of Abu Hanifa 
has made permissible the impermissible and forbidden what is allowed." See also 
Christopher Melchert, "The Adversaries of Ahmad ibn Hanbal," Arabica 44 (1997): 
236. 

*' Ibn Hajar, Path al-bdri, 1 1:696, #6685; Sahih al-Bukhdri: kitdb al-aymdn wa al-nudhdr, 
bdb 21. 



74 CHAPTER THREE 

followed him " He did this "out of the constrictive rancor [haraja] 

of his heart, breaking with the practice [sunan) of the Messenger of 
God (s), disparaging what he transmitted out of arrogance and enmity 
for the people of the sunan; for heretical innovation in religion (bid'a) 
had tarnished his flesh, bones and mind and made him revel in the 
non-Arabs' deluded celebration of him.""'' The object of this derision 
becomes clear later in the text, when al-Bukharl includes a report of 
Ibn al-Mubarak praying with Abu Hanifa. When Ibn al-Mubarak raises 
his hands a second time before bowing, Abu Hanifa asks sarcastically, 
"Aren't you afraid you'll fly away? {ma khashita an tatiraF)," to which 
Ibn al-Mubarak replies, "I didn't fly away the first time so I won't the 
second.""^ 

c. Al-Bukhdn and the Controversy over the Created Wording of the Qur'dn 

In light of al-Bukharl's strong identification with the ahl al-hadith, it 
seems difficult to believe that radical members of that camp ostracized 
him for his stance on the Qur'an. The issue of the createdness of the 
Qur'an had begun in the early Abbasid period, when a group of Muslim 
rationalists referred to by transmission-based scholars and later Sunni 
orthodoxy as the Jahmiyya began asserting that God did not speak in 
the anthropomorphic sense of the word, for this would necessitate His 
having organs of speech. Since this would belittle a power beyond the 
scope of human comparison, the Jahmiyya said that the Qur'an and 
other instances of God's speech (such as His speaking to Moses) were 
sounds that He created in order to convey His will to His domain.^" 
These rationalists similarly opposed other manifestations of anthropo- 
morphism, such as the notion that God will be seen by the believers 
on the Day of Judgment or that He sits on a throne or descends to the 



"" Al-Bukharl, Kitdb raf' al-yadaynji al-saldt, 20. This virulence is absent in Bukhari's 
chapters on this issue in his Sahih; see Ibn Hajar, Path at-bdn, 2:277-84. Note that the 
above-mentioned edition of this text contains an error on this page; the editor read 
as ^^mustahiqq"" what can only be ^'mustakhijf" ." 

"' Al-Bukharl, Kitdb raf al-yadayn, 107. 

'" Wilferd Madelung, "The Origins of the Controversy Concerning the Creation of 
the Koran," Orientalia Hispanica Volumen 1, ed. J.M. Barral (Leiden: Brill, 1974), 506. For 
interesting discussions of the debate over the nature of the Qur'an and its Iqf^ from 
within the Muslim tradition, see al-Subkl, Tabaqdt al-shdfi'iyya, 2:117-20 (biography of 
al-Husayn b. 'All al-Karablsl); Ibn al-Qayyim al-Jawziyya, Mukhtasar al-sawd'iq al-mursah, 
2 vols, in 1 (Cairo: Matba'at al-Madanl, [n.d.]), 2:304—17; al-Dhahabi, Tadhkirat al-hujfdi, 
2:223; 'Abd al-Khaliq 'Abd al-GhanI, Al-Imdm al-Bukhdn wa Sahihuhu, 156-67. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 75 

lowest heavens at night. ^' They also rejected ideas equally incompatible 
with a rationalist demeanor, like the punishment of the grave ['adhdb 
al-qabr).^"^ However, Muslims who maintained that the community 
should rely on the literal revelation received from the Prophet and his 
interpretation of the Qur'an as preserved in the sunna of the early 
Muslim community saw this rationalist movement as an attack on the 
textual authenticity of Islam. These traditionalists, who believed that 
one should not discuss these issues speculatively, opposed all instances 
of what they saw as the rationalist denial of God's attributes (ta'tU). 
Relying on the text of the Qur'an, hadiths and the stances of prominent 
members of the early community, books such as Ahmad b. Hanbal's 
al-Radd 'aid al-zanddiqa wa al-jahmiyya (Refutation of the Heretics and 
the Jahmiyya) asserted that God does in fact speak, that the Qur'an is 
one of His uncreated attributes, that He does mount His throne and 
that the believers will receive the beatific vision. 

The traditionalists' objections were not simply academic; they equated 
the assertion that the Qiar'an was created with calling God Himself 
created. Yahya b. Sa'id al-Qattan asked rhetorically of those who said 
the Qur'an is created, "How do you create (tasna'un) [the Qur'anic verse] 
'say He is the One God (qui huwa Alldh ahad; Qur'an 112:1),' how do 
you create [the verse] 'indeed I am Allah, there is no deity besides Me 
{innam and Alldh, Idildh Hid and; Qur'an 14:20).' "'' Moreover, the Qur'an 
had become a bulwark of social capital in the emerging civilization of 
Islam. When a famous Hanafi judge, 'Isa b. Aban (d. 221/836), who 
upheld the createdness of the Qur'an, was presiding over a dispute 
between a Muslim and a Jew, he asked the Muslim to swear "By God 
besides whom there is no other deity [wa'lldh alladhi Id ildha Hid huwa).'" 
His opponent objected, demanding that the judge make him swear by 
the real Creator, since these words were in the Qur'an, which Muslims 
claimed was created.^* The circulation of this story among traditionalists 



" There is some indication that the third caliph to preside over the mihna, al-Wathiq, 
added a denial of the beatific vision to the agenda of the inquisition; Abu Zahra, Ibn 
Hanbal, 143. 

'^ Madelung, "The Origins of the Controversy Concerning the Creation of the 
Koran," 510. See also Martin Hinds, "Mihna," EP. 

'' Al-Bukharl, Khalq af'dl al-'ibad, ed. 'Abd al-Rahman 'Umayra (Riyadh: Dar al- 
Ma'arif al-Suudiyya, 1398/1978), 33; cf Josef van Ess, "Ibn KuUab et la Mihna" 
Arabica "il (1990): 198. 

'* Al-Khatib, Tdrikh Baghdad, 11:160 (biography of Isa b. Aban). For another ref- 
erence to the controversy over this type of verse, see al-Dhahabi, Tadhkirat al-hujfdi, 
2:195 (biography of al-Nasa*!). 



76 CHAPTER THREE 

indicates that they felt that a belief in the createdness of the Qjur'an 
threatened its paramount role in society. 

In the early third/ninth century, however, the Abbasid caliph al- 
Ma'mun (d. 218/833) instituted a purge of these traditionalist beliefs 
from the empire's corps of judges. His Inquisition [mihna) was directed 
at those people who claimed to be the upholders of the Prophet's sunna 
and defenders of the community's unified identity, but, he claimed, 
were in reality demeaning God's greatness by putting the Qur'an on 
par with His essence. The rationalists behind this movement, including 
many of the Hanafi judges of Baghdad and Samarra, rejected the idea 
upheld by the traditionalists that the Qur'an was co-eternal with God, 
for that would mean that God is not the only eternal being.'^ Many 
of these rationalists were primarily concerned with polemics against 
Christian scholars who attempted to corner Muslims into accepting the 
divine nature of Christ by comparing him with the Qjur'an. If God 
states in the Qur'an that Jesus is the Word of God, just like the holy 
book itself, and that book is uncreated and co-eternal with God, then 
is Jesus not also co-eternal with God?^'' Is it so absurd, then, to believe 
that in the beginning he was the Word, and that the Word was with 
God? In addition to rejecting the anthropomorphic claim that God 
spoke in the literal sense, these rationalists thus also insisted that the 
Qur'an was created [muhdatli] as opposed to being an eternal attribute 
{qadirri) of God. 

The grueling torture, imprisonment or humiliation of prominent 
and widely respected hadith scholars such as Ahmad b. Hanbal, Yahya 
b. Ma'ln and 'All b. al-Madini in the Baghdad Mihna left an endur- 
ing and bitter impression on the hadith scholar community. Although 
the inquisition conducted by al-Ma'mun and his two successors did 



'"' Madelung, "The Origins of the Controversy Concerning the Creation of the 
Koran," 516; Hinds, "Mihna"; Melchert, "The Adversaries of Ahmad Ibn Hanbal," 
238-9. For a critique of current scholarship on the mihna, see Lucas, Constructive Critics, 
192-202. 

"' Abu Zahra, Ibn Hanbal, 64; Madelung, "The Origins of the Controversy 
Concerning the Creation of the Koran," 517. Madelung believes that the Muslim 
rationalist argument that the traditionalists were unintentionally abetting their Christian 
adversaries was more of an excuse for their attacks on the ahl al-hadith. Muhammad 
Abu Zahra, however, holds that the Mu'tazila and al-Ma'mun were in fact sincerely 
concerned with defending Islamic doctrine from Christian and other rationalist oppo- 
nents. There is also an interesting story about the distinction between muhdath (created) 
and qadim (eternal) being integral to an interfaith discussion between Harun al-Rashid 
and the sovereign of India; see Ibn Hajar, Fath al-bdri, 13:340. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 7 7 

not have as powerful a presence in Khurasan and Transoxiana, it 
did increase the enmity between the ahl al-hadith scholars and the 
Jahml/Mu'tazilite/Hanafi rationalists who had prosecuted it. During 
the lifetime of al-Bukhari and Muslim and in the decades after their 
deaths, the question of the nature of the Qiar'an in particular remained 
a touchstone for the resentment between these groups. In Iraq, Ibn 
Qutayba (d. 276/889) wrote al-Ikhiildf fi al-lafz wa al-radd 'aid al-Jahmiyya 
wa al-mushabbiha (Disagreement over the Lafz and the Rebuttal of the 
Jahmiyya and the Anthropomorphists),^' and Ibn Abl Hatim also wrote 
a book refuting the Jahmiyya.^" Even as late a scholar as al-Tabarani 
(d. 360/971) wrote a book condemning those espousing a belief in the 
created Qur'an.^^ In Naysabur, when someone who upheld the created- 
ness of the Qjur'an arrived in town, the hadlth scholar Abu al- 'Abbas 
al-Sarraj (d. 313/925) ordered the people in the market to curse him, 
and they complied.'"" 

The tremendous tension surrounding this issue led the most con- 
servative section of the traditionalists to declare anathema anyone 
who asserted that the wording of the Qur'an {lafz), the physical sound 
of the book being recited or its written form on a page, was created. 
This most intolerant end of the traditionalist spectrum, what George 
Makdisi called "ultra-conservatives,"'"' included the standard por- 
trayal of Ahmad b. Hanbal, Abu JaTar Muhammad Ibn al-Akhram 
(d. 301/913-4), Muhammad b. Yahya al-Dhuhli of Naysabur and others. 
These iiber-Sunnis repudiated any traditionists who did not declare 
that the Qur'an was God's eternal speech and utterly increate. Those 
who simply proclaimed that the Qur'an was God's speech and then 
were silent, even those like 'All b. al-Madlnl who collapsed under the 
weight of the Inquisition, were dubbed "Those who stopped short 
{wdqifyyd)" and often equated with Jahmls.'"^ As Christopher Melchert 



^' Al-Bukhan is not mentioned in this book, although Ibn Hanbal is; see Ibn 
Qutayba, al-Ikhtildf fi al-laji wa al-radd 'aid al-jahmiyya wa al-mushabbiha, ed. Muhammad 
Zahid al-Kawtharl (Cairo: Maktabat al-Sa'ada, 1349/ [1930]). 

™ Al-Dhahabi, Tadhkirat al-huffas, 3:34. 

'^ Abu Zakariyya Yahya Ibn Manda, "Manaqib al-Shaykh Abl al-Qasim al- 
Tabaranl," MS Esad Efendi 2431, Siileymaniye Library, Istanbul: 14b. 

'»" Al-Dhahabi, Tadhkirat al-hujfdz, 2:215. 

"" George Makdisi, "Ash'arl and the Ash'arites in Islamic Religious History," Studia 
Idamica 17 (1962): 39. 

'"2 Wilferd Madelung, "The Origins of the Controversy Concerning the Creation of 
the Koran," 521. Although Ibn Hanbal narrates some hadlths from 'All b. al-Madlnl 
in his Musnad, one of his son's students, al-'Uqayll, said that when he studied Ibn 



78 CHAPTER THREE 

observes, the iiber-Sunnis saw them as doubly dangerous because they 
were "self-proclaimed traditionalists" who identified themselves with the 
ahl al-hadith/ ahl al-sunna camp. The iiber-Sunnis thus reserved some of 
their fiercest invective for these folk.'"^ Melchert has astutely identified 
this group between the iiber-Sunnis and their rationalist adversaries, 
dubbing them "the semi-rationalists." He includes a diverse selection 
of scholarly figures, from al-Shafi'T's most famous disciple, al-Muzani, 
to the great historian and exegete al-Tabarl. '"* The identifying charac- 
teristic of what Melchert admits is a loosely-knit group is their belief 
that the lafz of the Qur'an is created. He includes al-Bukhari in this 
number because he upheld this stance. 

Yet it is not very accurate to employ the term "rationalist" in any 
sense when describing al-Bukharl, who was a diehard traditionalist. 
Rather, we should view him as a representative of Ibn Hanbal's original 
traditionalist school who fell victim to its most radical wing. Indeed, al- 
Bukhari's Khalq af 'al al-'ibad contains the earliest representation of the 
position taken by Ibn Hanbal, a figure often co-opted by later groups 
to legitimize their stances.'"'^ Al-Bukharl wrote this work within fifteen 



Hanbal's Kitdb al-'ilal with Ibn Hanbal's son 'Abdallali he saw that Ibn Hanbal had 
crossed out 'All's name in many isndda and replaced it with "a man." Nonetheless, 
al-'Uqayll affirms that 'All's hadlths are reliable; Muhammad b.'Amr al-'Uqayll, Kitdh 
al-du'aja' al-kabir, ed. 'Abd al-Mu'tl Amin Qal'ajl, 4 vols. (Beirut: Dar al-Kutub al- 
'Ilmiyya, 1404/1984), 3:239. 

"" Melchert, "The Adversaries of Ahmad Ibn Hanbal," 252. 

'"* Melchert's evidence for al-Tabarl's stance on this issue (see Ibn Hajar, Lisdn al- 
mizdn {Hyderabad: Da'irat al-Ma'arif al-'Uthmaniyya, 1330/[1912]}), 3:295 [biography 
of Ibn Abl Dawud al-Sijistanl] is meager (as Melchert admits elsewhere, the charge 
"looks anachronistic"). In his al-Tabsir Ji ma'dlim al-din, al-Tabarl cleverly avoids discuss- 
ing the issue of the lafi of the Qur'an. He explicitiy states that the Qiar'an is neither 
created nor a creator — the ahl al-hadith position — supporting his stance with a long 
logical argument. On the issue of the lafi of the Qur'an, however, al-Tabarl refers 
the reader to his discussion of the acts of humans {afdl al-'ibdd). In this discussion, he 
rejects the Qadarl and Jahml position (the latter that men have no control over their 
acts) and embraces the third position, that of ihejamhur ahl al-ithbdt (the majority of 
those who affirm God's power over destiny), namely that God guides those destined 
for faith to faith and vice versa. He does not clearly state, however, whether or not 
men's acts are created. His exact position on the Iqfi issue thus remains unclear See 
al-Tabarl, al-Tabsu fi ma'dlim al-din, ed. 'All b. Abd al-'AzIz al-Shibl (Riyadh: Dar al- 
•Asima, 1416/1996); 167-76, 200-5; cf Melchert, "The Adversaries of Ahmad Ibn 
Hanbal," 245—7; idem. The Formation of the Sunni Schools of Law, 9th and 10th Centuries 
C.E. (Leiden: Brill, 1997), 195. 

'"' Ibn Hanbal's role as a figure on which different schools of thought have pro- 
jected their particular stances is well known. Ibn Hanbal is most famous for stating 
that "he who says my wording of the Qur'an is created is Jahmi, and he who says it is 
not created is guilty of bid'a." Another, less likely, report through Ibn Hanbal's student 



THE GENESIS OF AL-BUKHARI AND MUSLIM 79 

years of Ibn Hanbal's death in 241/855, and he incisively identified 
the polemical circus that had already grown up around Ibn Hanbal's 
persona: 

And as for the two sects [of the rationalists and hadith scholars] that claim 
proof for themselves from Ahmad, many of their reports [from him] are 
not reliable. Perhaps they have not understood the precise subdety of his 
stance {diqqat madhhabihi). It is known that Ahmad and all the people of 
knowledge hold that God's speech is uncreated and that all other speech 
is created. Indeed they hated discussing and investigating obscure issues, 
and they avoided the people of dialectical theology ikaldm), speculation 
{al-khawd) and disputation (tandzu') except on issues in which they had 
[textual] knowledge. ""^ 

Al-Bukhari's allegiance to the ahl al-hadith camp and to Ibn Hanbal 
himself is thus obvious. Indeed, he quotes Ibn Hanbal as evidence for 
his position on the /a^."" 

Melchert concedes that the semi-rationalists were a diverse group, 
but it seems more accurate to group al-Bukharl with the traditionalist 
camp of Ibn Hanbal than with al-Tabari, whose lengthy explanation 
of why the Qur'an is uncreated consists of a formalized logical discus- 
sion of accidents and whether or not speech can inhere in the essence 
{dhdi) of a thing. Also, Melchert's description of the semi-rationalists 
as "insinuating the tools of the rationalists into traditionalist practice" 
would hardly place al-BukharT in the environs of the rationalist camp. 
None of al-Bukharl's extant works employs Islamicate logic or the 
philosophical jargon found in al-Tabarl's discussion. ""* 



Ibrahim al-Harbi tells of someone asking Ibn Hanbal about a group of people who 
say that "our wording of the Qiar'an is created." He replied, "The slave approaches 
God through the Qur'an by five means, in which [the Qiar'an] is not created: memo- 
rizing in the heart, reading by the tongue, hearing by the ear, seeing with the eye, and 
writing by the hand. The heart is created and what it memorizes is not; the reading 
(tildwd) is created but what is read is not; hearing is created but what is heard is not; 
sight is created but what is seen is not; and writing is created but what is written is 
not"; Ibn al-Qayyim, Mukhtasar al-sawd'iq al-mursala, 2:313-4. For another example of 
attributions to Ibn Hanbal, see Zayn al-Din al-'iraqi, al-Taqyid wa al-iddh li-md utliqa 
wa ughliqa min Muqaddimat Ibn al-Saldh, ed. Muhammad 'AbdaUah Shahln (Beirut: Dar 
al-Kutub al-'Ilmiyya, 1420/1999), 205. 

""5 Al-Bukharl, Khalq af'dl al-'ihdd, 62. 

'»' M-^ukhm, Khalq af'dl al-'ibdd, 108. 

""^ Al-Bukharl's Khalq af'dl al-'ibdd is little more than a collection of proof texts fi"om 
Prophetic hadlths and earlier Muslim authorities, including Ibn Hanbal himself Only 
at the very end of his book does al-Bukharl resort to what could be termed dialectics, 
such as the use of constructions like "if someone says ... let it be said to him" or terms 
like bajdn. Often when this work does resort to dialectical arguments, they center on 



80 CHAPTER THREE 

It is more accurate to describe al-BukharT as a conservative tradi- 
tionalist trying to navigate the contradictions inherent in the blunt ahl 
al-sunna creed touted by the iiber-Sunnis like al-Dhuhli. Al-Bukhari 
knew that the Qur'an was God's uncreated speech, but he also knew 
that God creates human actions, as the ahl al-sunna had insisted in their 
attacks on the free-will position of their Qadarite opponents. What, 
then, does one say of the Qur'an when it becomes manifested in a 
human act such as recitation or writing? 

The earliest sources on al-Bukharl's life suggest that he was very 
reluctant to discuss this issue at all. He would understandably have 
viewed it as speculation (khawd) and thus tried to avoid it. Our earliest 
substantial source on al-Bukhari, Ibn 'Adi, includes a story he heard 
from a group of his teachers that tells of al-Bukharl refusing to answer 
questions about the nature of the Qur'an's wording until absolutely 
pressed, saying, "The Qur'an is God's speech, uncreated, and the acts 
of men are created, and inquisition (imtihan) is heresy (bid'a)."^"^ 

Al-Bukhari's defense against the accusations of the iiber-Sunnis, his 
Khalq af'dl al-'ibdd, displays this same caution. The first section of the 
book is devoted solely to narrations from earlier pious authorities such 
as Sufyan al-Thawri that affirm the increate nature of the Qjur'an 
and condemn anyone who holds the contrary position as a Jahmi or 
unbeliever. The second section argues that the acts of men are created, 
relying on Qur'anic verses and reports from such vaunted traditional- 
ists as Yahya b. Sa'ld al-Qattan. Al-Bukharl himself rarely comments, 
but does assert that men's actions, voices and writing are created. He 
then begins introducing narrations from the Prophet that suggest that 
it is permissible to sell and buy written copies of the Qur'an."" Finally, 
he provides a hadlth of the Prophet enjoining Muslims to "beautify 
the Qjur'an with your voices" and a report from 'All b. Abl Talib that 
there will come a time when nothing remains of the Qur'an except its 
written form. ' ' ' These reports insinuate that physical manifestations of 
the Qjur'an do indeed belong to the material world. The author then 



combating his opponents' use of hadlths. See al-Bukhari, Klialq af'dl al-'ibdd, 105-6; 
al-Subkl, Tabaqdt al-shdfi'iyya, 2:229. 

'"" Ibn 'Adi, Asdmi, 64—5. This story also appeared in al-Hakim's Tdrikh Maysdbur, 
narrated from Ibn 'Adi. See al-Dhahabi Tdrikh al-isldm, 19:266. 

"" Al-Bukharl, Ualg af'dl al-'ibdd, 59-60. 

" ' "Ya'ti 'aid al-nds zamdn Idyabqd min al-isldm Hid ismuhu wa Id min al-qur'dn ilia rasmuhu"; 
al-Bukharl, EJialq af'dl al-'ibdd, 66-7. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 8 1 

returns to refuting the rationalists, emphasizing that the belief that 
human acts are created is not heresy [bid'd)}^'^ Only at this point does 
al-Bukharl begin actively arguing that the sound of the Qur'an being 
recited is created. 



Reality: Muslim, the Junior Partner 

Abu al-Husayn Muslim b. al-Hajjaj al-Qushayri was born in 206/821 
in Naysabur. He first learned hadlth from Ishaq b. Rahawayh and 
Yahya b. Yahya al-Tamlml (d. 224—6/839-41) in his hometown before 
leaving for a pilgrimage to Mecca in 220/835. In the Hijaz he heard 
from 'Abdallah b. Maslama al-Qa'nabi (d. 220-1/835-6), a favorite 
transmitter of Malik's Muwatta', and others. He later visited Baghdad 
to hear from Ibn Hanbal and also went to Basra. He went to greater 
Syria, Egypt and Rayy where he met several times with Abu Zur'a al- 
Razi (d. 264/878) and Abu Hatim al-RazI (d. 277/890). A few years 
before his death he settled in Naysabur, where he became one of the 
senior hadlth scholars in the city and a central figure for study. "^ It was 
in Naysabur that he studied and became acquainted with al-Bukhari. 
Al-Hakim al-Naysaburl, whose father met Muslim, recalls that Muslim's 
"place of business (matjar) was Khan Mahmash," where his father saw 
him narrating hadlths. Muslim's livelihood also came from his proper- 
ties at Ustu which came from "the progeny (a'qab) of the females of 
his family""* He died in 261/875 at the age of fifty-five. 

Muslim left many more works than his elder contemporary. His most 
famous, of course, was his Sahih, originally titled al-Musnad al-sahih}^^ 
Muslim also produced two larger collections, a musannaf and a musnad, 
representing the sum total of the hadlth corpus from which he selected 
his Sahth. Ibn al-Jawzi does not believe that anyone ever transmitted 



"2 M-Bukhan, Malq af at al-ibdd, 102-4. 

"' In his biography of Abu 'All al-Husayn al-Qabbam (d. 289/901-2), al-Dhahabi 
notes Abu 'Abdallah b. al-Akhram (d. 344/955) saying, "The people of hadlth used to 
gather around him {'indahu) after Muslim"; al-Dhahabi, Tadhkirat al-hujfdi, 2:183. 

"* Cited in al-Dhahabi, Tdnkh al-isldm, 20:187. 

"^ This is somewhat misleading, since Muslim's work is topically organized, not a 
musnad. Ibn Khayr al-Ishblll recorded the full title as al-Musnad al-sahih al-mukhtasar min 
al-sunan bi-naql al-'adl 'an al-'adl 'an rasul Allah s; Abu Ghudda, Tahqiq ismay al-Sahihayn, 
33-4. 



82 CHAPTER THREE 

this large musnad from Muslim.'"' He also produced several biographical 
dictionaries. The largest one, his Tabaqat, simply provides the names 
of the hadlth transmitters in the generations after the Prophet. Other 
smaller works, such as the Munfariddt, the Wihddn and the Dhikr man laysa 
lahu ilia raw'" wdhid min ruwdt al-hadith, detail people who lack more than 
one transmitter from them.'" Like al-Bukhari and many other hadlth 
masters of his age, Muslim produced a book of criticized narrations 
{Kitdb al- 'Hal) and a work of the same ilk but designed for a more gen- 
eral audience, the Kitdb al-tamyiz. This latter work has survived in part, 
and along with Muslim's involved introduction to his Sahih, provides 
invaluable information about its author and his leanings. 

a. Muslim's Methodology in his Sahlh 

One of the most prominent statements Muslim makes about his 
methodology is his comparatively lax requirement for ascertaining 
whether a link in an isndd marked by "from/on the authority of ( (zn)" 
actually represents personal contact. When " 'an'' is used, Muslim does 
not require affirmative proof that the two transmitters actually met. 
Instead he requires only that they were contemporaries with no "clear 
indication (daldla bayjina)" that they did not meet. Here Muslim invokes 
the example of Malik, Shu'ba, Yahya b. Sa'id al-Qattan and 'Abd al- 
Rahman b. Mahdi, who "only felt compelled to find a guarantee of 
direct transmission {samd') if the narrator was known to conceal his 
immediate source (mudallis)."^'^' In this, Muslim openly breaks with the 
position attributed by scholars to al-Bukharl and his teacher 'All b. al- 
Madini. Muslim acknowledges that there are those who uphold that 
position, but he angrily asserts that they lack precedent from earlier 
hadlth masters."" The notion that affirming one meeting between two 



'"' Ibn al-jawzl, al-Munta^am, 12:171. 

'" One such work has been published under the title al-Munfariddt wa al-wahddn, ed. 
'Abd al-GhalFar Sulayman al-Bandari (Beirut: Dar al-Kutub al-'Ilmiyya, 1408/1988). 

"« Muslim, Sahih, 1:26. 

"" Muslim, Sahih, 1:23, 28. The majority of later commentators assumed that 
Muslim meant al-Bukharl, but Ibn Kathir believes he intended 'All b. al-Madlnl. Several 
modern Muslim scholars have also dealt with this question. In his comprehensive treat- 
ment of this question in the third appendix to his edition of al-Dhahabl's al-Muqiza, 
Abd al-Fattah Abtl Ghudda states that the person in question cannot be al-Bukharl. 
Assuming Muslim wrote his introduction before he completed the book, he would not 
even have met al-Bukharl at the time; he only met his teacher in 250-1 AH when al- 
Bukhari came to Naysabur; Ibn Kathir, al-Bd'ith al-hathith, 45; al-Dhahabi, al-Muqi^aJi 



THE GENESIS OF AL-BUKHARI AND MUSLIM 83 

transmitters somehow assures direct transmission for all their hadiths, he 
states, is absurd. He provides examples of isndd% in which two narrators 
who had met nonetheless occasionally transmitted via an intermedi- 
ary concealed by a '"an" link in the isndd.^^^ Moreover, the adherents 
of this position unnecessarily dismiss many authentic hadiths. "If we 
were to count the authentic reports {al-akhbdr al-sihdij) . . .," he says, "that 
would be maligned by the claim of this claimant, the number would 
be inestimable."'^' 

In his introduction, Muslim divides hadiths and their concomitant 
transmitters into three groups, stating that he will rely on two of them 
in his Sahih. The first consists of the well-established hadiths whose 
transmitters do not lapse into the "excessive confusion" [takhlTt fihish) 
into which many muhaddith's, stumble. Having exhausted this group, he 
will proceed to the reports of transmitters who are not as masterful as 
the first group but nonetheless "are characterized by pious behavior 
{sat), honesty and the pursuit of knowledge." He will not take reports 
from the third group, which consists of those who either forge hadiths 
or whose material differs beyond reconciliation with that of superior 
scholars. '^^ 

Muslim's Sahih contains far fewer chapters (only 54) than al-Bukharl's 
and lacks al-Bukharl's legal commentary. It has many more narrations, 
numbering about 12,000, with 4,000 repetitions. According to Muslim's 
companion Ahmad b. Salama al-Bazzar (d. 286/899), who was with 
Muslim for fifteen years while he wrote the Sahih, this number is based 
on Muslim's very unarf-based definition of a hadlth. If he heard the 
same tradition from two shaykhs, he considered it to be two hadiths. '^^ 
Ibn al-Salah (d. 643/1245) places the number of Prophetic traditions 
in the Sahih at around 4,000.'^* Unlike al-Bukharl, Muslim keeps all the 
narrations of a certain hadlth in the same section. Muslim also diverges 
significantly from al-Bukhari in his exclusion of Companion hadiths 
and narrations without full isndds {ta'liqdt) as commentary. '^'^ 



'utum mustalah al-hadith, ed. 'Abd al-Fattah Abu Ghudda (Aleppo: Maktab al-Matbu'at 
al-Islamiyya, 1405/1084), 122-140. 

'™ Muslim, Sahih, 1:24-5. 

'2' Muslim, Safe-A, 1:26. 

'22 Muslim, Sahih, 1:4-5. 

'2' Al-Dhahabi, Tdnkh al-isldm, 20:186; Abd al-Rauf, "Hadlth Literature," 275. 

'2' Ibn al-Salah, Siydnat Sahih Muslim, 101-2. 

'2^ Scholars have generally counted only 12-14 instances of incomplete isndds, (ta'liq) 
used for commentary in Muslim's book; cf Ibn al-Salah, Siydnat Sahih Muslim, 77. 



84 CHAPTER THREE 

There is considerable overlap between Muslim's Sahih and that of his 
teacher al-Bukharl; according to Abu Bakr Muhammad b. 'Abdallah 
al-Jawzaqi (d. 388/998), whose book al-Muttafaq combined the two 
books, there are 2,326 common traditions.'^'' The two scholars drew 
on essentially the same pool of transmitters, sharing approximately 
2,400 narrators.'^' Al-Bukhari narrated from only about 430 that 
Muslim did not, while Muslim used about 620 transmitters al-Bukhari 
excluded.'^" 

Scholars have generally devoted much less attention to Muslim's legal 
positions, perhaps because his Sahih is more simply a hadith book than 
al-Bukharl's legally charged work. Not only does Muslim's book cover 
many fewer legal topics than his teacher's, his chapters often provide 
support for both sides of a particular issue. Indeed, he seems to have 
left his subchapters without titles, and he never raged as angrily as al- 
Bukhari in any of his extant works. '^^ Muslim thus does not appear in 
al-'Abbadi's or al-Subki's roster of the Shafi'i school. Ibn Abi Ya'la, on 
the other hand, does include him in the Tabaqat al-hanabila, emphasiz- 
ing his narrations from Ibn Hanbal and his discussing hadith narrators 
with him.''" 

These sources leave little doubt concerning Muslim's identification 
with the transmission-based school. Like most of the ahl al-hadith, 
Muslim reportedly criticized Abu Hanifa and the ahl al-ra'y, but his 
comments certainly lack al-Bukhari's ferocity. Al-Jawzaqi quotes him 
as saying that Abu Hanifa was "a practitioner of independent legal 
reasoning whose hadiths are problematic [sahib ra'y, mudtarib al-hadith).'"^'^^ 
In the introduction to his Sahih, Muslim also gives a report condemning 
answering questions for which one has no textual recourse {'ilm) or nar- 



'^*' Ibn Hajar, al-Mukat 'ala kitab Ibn al-Salah, ed. Mas'ud 'Abd al-Hamid al-Sa'dafi 
(Beirut: Dar al-Kutub al-'Ilmiyya, 1414/1994), 69-70. Ibn Hajar states that al-Jawzaqi 
considers the same tradition from two different Companions to be one hadith. This 
would mean that his account of the number of hadiths common to both the Sahihi, is 
probably much lower than other Muslim scholars might consider. 

'" This number was arrived at by Abu al-Fadl Muhammad b. Tahir al-MaqdisI b. 
al-QaysaranI (d. 507/1113); MuUa Khatir, Makdnat al-Sahihayn, 182. 

128 'Yhh number was arrived at by al-Hakim al-Naysaburl and quoted by Ibn al- 
Salah; Ibn al-Salah, Siydnat Sahih Muslim, 84. 

'2^' Al-Nawawl, 5AaM Sahih Muslim, 15 vols. (Beirut: Dar al-Qalam, 1407/1987), 
1:129. 

"" Ibn Abl Ya'la, Tabaqat al-handbila, 1:311-2. 

'" Ibn al-Najjar, Kitdb al-radd 'aid Abi Bakr al-Khatib al-Baghdddi, ed. Mustafa 'Abd 
al-Qadir 'Ata (Beirut: Dar al-Kutub al-'Ilmiyya, 1417/1997), 101. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 85 

rating from untrustworthy people. '^^ Like al-Bukhari, Ibn Hanbal and 
other ahl al-hadith, this position represents the rejection of speculation 
{khawd) on issues of dogma. 

Unlike al-Bukhari, Muslim managed to avoid the controversy that 
plagued the latter part of his senior's career. Although later sources 
report that Muslim explicitly shared al-Bukharl's stance on the created 
lafz of the Qur'an, there is no early evidence for this. Ibn AbT Hatim 
al-RazI, who notes al-Bukharl's lafi scandal, mentions nothing of the 
sort in his entry on Muslim. When al-Hassan b. Muhammad al-QazwInI 
(d. 344/955) of Naysabur asked his father whose book he should imi- 
tate, al-Bukharl's or Muslim's, his father directed him towards Muslim's 
Sahih because he was not tainted by the lafz issue. '^^ 

Nonetheless, Muslim also fell out with al-Dhuhll, who seems to have 
been unable to bear serious competition in Naysabur As in al-Bukharl's 
case, al-Dhuhli's animosity towards Muslim was not sudden. Al-Hakim 
reports from Tahir b. Ahmad, who heard Muslim's student Makki b. 
'Abdan say that when Dawud b. 'All al-Zahiri (d. 270/884) came to 
Naysabur to study with Ishaq b. Rahawayh they held a discussion 
(al-na^ar) session for him. Al-Dhuhll's son Haykan (d. 267/881) and 
Muslim, at that time no older than thirty-two, attended. Haykan gave 
his opinion on an issue, and Dawud scolded him (zabarahu), saying, "Be 
silent, youth!" Muslim did not rally to his side. Haykan then went back 
to his father and complained about Dawud. Al-Dhuhli asked who was 
with him in the debate, and Haykan replied, "Muslim, and he did not 
support me." Al-Dhuhli bellowed, "I take back all that I transmitted 
to him [raja'tu 'an hull md haddathtuhu bihi)." When Muslim heard this 
he "collected all that he had written from him in a basket and sent it 
to him, saying, 'I will never narrate from you,' " then left to study with 
'Abd b. Humayd (d. 249/863).'^* According to al-Hakim, the last part 
of this story is inaccurate. He states that Muslim continued to associate 
and study with al-Dhuhli until al-Bukharl's lafz scandal some twenty 
years later When al-Dhuhli prohibited his students from attending al- 
Bukhari's lessons, Muslim stood up and left al-Dhuhli's circle, sending 



"2 Muslim, Sahih, 1:13. 

'^^ Al-Dlialiabl, Tadhkirat al-hujfa^, 3:75; idem, Tdrikh al-isldm, 25:417-8. 
'^' Cited from al-Hakim 's Tdnkh Naysdbur, al-Dhahabi, Tdrikh al-isldm, 20:187; Ibn 
'Asakir, Tdrikh madinat Dimashq, 58:93. 



86 CHAPTER THREE 

a porter to him with all the material he had received from him.'^' That 
the tension between Muslim and al-Dhuhli was longstanding dovetails 
with an otherwise bizarre quote from Abu Zur'a al-RazI, who criticized 
Muslim as unreasonable, saying, "If he had tended properly to [dara) 
Muhammad b. Yahya [al-Dhuhli] he would have become a man!"'^'' 



Perception: al-Bukhan, Muslim and the Greatest Generation 

To the ahl al-hadith, in the decades after their deaths al-Bukhari and 
Muslim were simply two accomplished scholars among many. They 
studied at the feet of titans and were survived by cohorts who often 
outshone them in the eyes of fourth/tenth-century hadlth authori- 
ties. To best understand their place in this context, we shall compare 
perceptions of al-Bukhari and Muslim with those of their teachers, 
such as 'All b. al-Madlni, Ishaq b. Rahawayh and Ibn Hanbal; and of 
their peers, like al-Dhuhll, Abu Zur'a al-Razi and his colleague Abu 
Hatim al-RazI. 

Our earliest sources leave no doubt that al-Bukhari and Muslim 
were certainly respected authorities whose talents were widely recog- 
nized. Al-Hakim narrates from Muhammad b. Ahmad al-Mudhakkir 
that Ibn Khuzayma (d. 311/923) said, "I have not seen beneath the 
heavens one more knowledgeable in hadlth than Muhammad b. Isma'il 
al-Bukhari.'"^' Ibn 'Adi heard al-Bukhari's student Muhammad b. 
Yusuf al-Firabri (d. 320/932) say that al-Najm b. al-Fadl had seen the 
Prophet in a dream, with al-Bukharl walking behind him exactly in 
his footsteps.'-'" Oddly, there is little explicit praise for Muslim in the 
early sources. In a rare Persian-language quote, al-Hakim cites Ishaq 
b. Rahawayh saying, "What a man [Muslim] is!'"'^ 

Later sources, of course, overflow with reports about both men's 
abilities, phrased in the hyperbolic style so common to Muslim schol- 
arly expression. Al-Khatib quotes Ibn Hanbal's saying that the mastery 
of hadlth [hifz) ends with four people from Khurasan: Abu Zur'a, al- 



Al-Hakim as quoted in al-Dhahabi, Tankh al~islam, 20:188, cf. al-Khatib, Tankh 
13:103 for the same narration with the same isndd through al-Hakim. 
'* Cf. al-Dhahabi, Tankh al-isldm, 12:187; 19:341. 

'" Al-Hakim al-Naysabun, Ma'rifat 'ulum al-hadith, ed. Mu'azzam Husayn (Hyderabad: 
Da'irat al-Ma'arif al-'Uthmaniyya, 1385/1966), 93. 
"^ Ibn 'Adi, al-KdmilJi du'aja' al-rijdl, 1:140. 
139 '%ardi keh in bud"; al-Hakim, Ma'rifat 'ulum al-hadith, 98. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 87 

Bukhari, 'Abdallah b. 'Abd al-Rahman al-Darimi (d. 255/869) and al- 
Hasan b. Shuja' al-Balkhi (d. 266/880).'*" In the Tarikh Baghdad we also 
find a quote from al-Bukhari's Basran teacher Muhammad b. Bashshar 
Bundar (d. 252/866) saying that "the hadith masters [huffaz) of the 
world are four...:" Abu Zur'a al-Razi in Rayy, Muslim in Naysabur, 
al-Darimi in Samarqand and al-Bukharl in Bukhara.'*' 

Yet in our earliest sources, instances of such hyperbolic praise often 
ignore al-Bukharl and Muslim. Even Muslim's colleague Ahmad b. 
Salama (d. 286/899) is reported to have said, "I have not seen after 
Ishaq [b. Rahawayh] and Muhammad b. Yahya [al-Dhuhll] someone 
with a greater command of hadith [ahfaz li'l-hadith), nor more knowl- 
edgeable as to their meanings, than Abu Hatim Muhammad b. Idris 
[al-Razi].'"*^ In his book on al-Bukharl's teachers, Ibn 'Adi records a 
statement from another of their contemporaries, 'Uthman b. 'Abdallah 
b. Khurrzadh (d. 281-4/894-8). He says that "the most prodigious in 
memory {ahfaz) I have seen are four: Muhammad b. Minhal al-Darir, 
Ibrahim b. Muhammad, b. 'Ar'ara, Abu Zur'a and Abu Hatim [al- 
Razi].'"'" Even reports found only in later sources often neglect the 
two scholars. In al-Dhahabl's Tadhkirat al-huffaz, Abu Ishaq Ibrahim Ibn 
Urama of Isfahan (d. 266/880) is quoted as saying during al-Bukharl's 
and Muslim's lifetimes that "now there remain only three in the world: 
al-Dhuhll in Khurasan, Ibn al-Furat in Isfahan, and [al-Hasan b. 'All] 
al-Hulwani (d. 243/857-8) in Mecca.'"** 

But how did hadith scholars in the century after al-Bukharl and 
Muslim view these two in holistic surveys of the hadith tradition? The 
earliest impression we have comes from Abu Hatim's son, Ibn Abl 
Hatim (d. 327/938), who wrote a monumental treatise on the disci- 
pline of hadith criticism, al-Jarh wa al-ta'dil (Criticism and Approval). 
At the beginning of the work, the author provides lengthy and lauda- 
tory chapters devoted to pillars of the hadith tradition such as Sufyan 
al-Thawri and Wakf b. al-Jarrah. This section ends with the great 
scholars Ibn Hanbal, Yahya b. Ma'ln, and 'All b. al-Madlnl, but also 



'"' Al-Khatib, Tarikh Baghdad, 2:21, 10:326 (biography of Abu Zur'a al-Razi); Yaqut 
b. 'Abdallah al-HamawI (d. 626/1229), Mu'jam al-bulddn, 6 vols. (Tehran: Maktabat 
al-Asadi, 1965), 1:714. 

'" Al-Khatib, Tarikh Baghdad, 2:16; Ibn 'Asakir, Tarikh madinat Dimashq, 58:89. 

'" Al-Hakim, Ma'rifat 'ulUm al-hadith, 95-96; al-Khatib, Tarikh Baghdad, 2:73. 

'" Ibn 'Adi,Asdmi, 138; idem, al-Kdmil, 1:143. 

'" Al-Dhahabl, Tadhkirat al-huffds, 2:80. 



88 CHAPTER THREE 

includes Abu Zur'a al-Razi and the author's father. Although al-Bukhari 
and Muslim both died before the two Razis, Ibn Abi Hatim devotes 
only short and unremarkable entries to them in the main biographi- 
cal body of his dictionary For al-Bukhari he states that his father and 
Abu Zur'a rejected his hadlths after al-Dhuhli wrote informing them 
of his view on the Qur'an.'*'' Muslim too receives a perfunctory entry 
with the compliment "trustworthy, one of the hadlth masters [hujfaz] 
with knowledge of hadith."'*'' Neither al-Bukhari nor Muslim merited 
a place in the last great generation of their teachers. 

Of course, Ibn Abi Hatim's view is very biased — his inclusion of 
his father and his close associate Abu Zur'a in the pantheon of great 
hadlth scholars was no doubt an act of discretion. In examining the 
initial reception of al-Bukharl's and Muslim's works, however, it is 
precisely such biased perception that interests us. For Ibn Abi Hatim, 
one of the most influential figures in the development of hadlth criti- 
cism, Muslim is negligible and al-Bukharl anathema. As we shall see, 
the cadre of RazI hadlth scholars based in Rayy provided the earliest 
and most vocal reaction to al-Bukharl's and Muslim's careers. 

In his Kitdb al-majruhin (Book of Criticized Narrators), Ibn Hibban 
al-BustI (d. 354/965) includes a review of the various generations of 
hadlth scholars who had toiled to preserve the legacy of the Prophet. 
The generation that inherited this trade and learned from masters like 
Malik b. Anas and Shu'ba b. al-Hajjaj consists of Ibn Hanbal, Yahya 
b. Ma'ln, 'All b. al-Madini (the three biggest), Ishaq b. Rahawayh, 
'Ubaydallah al-Qawariri (d. 235/850) and Abti Khaythama Zuhayr b. 
Harb (d. 234/848). The next generation, which "took from them this 
path of criticism," he lists as al-Dhuhll, al-Darimi, Abu Zur'a al-Razi, 
al-Bukhari, Muslim and Abu Dawud al-Sijistani.'*' Here we clearly 
see a division between al-Bukharl's and Muslim's generation and that 
of the teachers from whom they derived their skills. The two scholars, 
however, receive no special attention. 

In his early work on the discipline of hadlth transmission, al-Muhaddith 
al-Jadil (The Virtuous Hadlth Scholar), al-Hasan b. 'Abd al-Rahman 
al-Ramahurmuzi (d. 360/970—1) lists five generations of great hadlth 



"' Ibn Abi Hatim, al-Jarh wa al-ta'dil, 2:3:191. 

'*« Ibn Abi Hatim, al-Jarh wa al-ta'dil, 4:1:182-3. 

'*' Abu Hatim Muhammad Ibn Hibban al-BustI, Kitdb al-majruhin min al-muhaddithin 
al-du'afd' wa al-matmkin, ed. Mahmud Ibrahim Zayid (Aleppo: Dar al-Wa'y, 1396/1976), 
1:54-7. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 89 

collectors who brought together the transmitted materials of various 
regions. His third generation includes men like Ibn Hanbal and Ishaq b. 
Rahawayh, his fourth the likes of al-Dhuhli, Abu Zur'a and Abu Hatim 
al-Razi, and Abu Dawud. The fifth and final generation includes Ibn 
Abi Hatim, al-Nasa'i, al-Tabari and others.'*" Al-Bukharl and Muslim 
appear nowhere. 

In his al-KdmilJi du'afa' al-rijdl (The Complete Book on Weak Trans- 
mitters), Ibn 'Adi (d. 365/975-6) places al-Bukhari at the beginning of 
the final generation {tabaqa) of hadith scholars. Although this genera- 
tion includes Abu Hatim and Abu Zur'a al-RazI as well as al-Nasa'l, 
Muslim never appears. These scholars follow the era of men like Ibn 
Hanbal, Ishaq b. Rahawayh and 'All b. al-Madlnl. Ibn 'Adi quotes 
the litterateur cum hadith scholar Abu 'Ubayd al-Q_asim b. Sallam 
(d. 224/839) of Naysabur on the definitive place of this greatest gen- 
eration: "[Mastery of] hadith stopped at four people: Abu Bakr b. Abi 
Shayba (d. 235/849), Ahmad b. Hanbal, Yahya b. Ma'ln, and 'All b. 
al-MadinI."'« 

Muslim scholars outside the Sunni traditionalist fold also grasped the 
prominence of the greatest generation of Ibn Hanbal and his contem- 
poraries. The Mu'tazilite Abu Qa.sim al-Balkhi (known as al-Ka'bi, d. 
319/931) wrote his Qubul al-akhbdr (The Acceptance of Reports) as a 
weapon against the ahl al-hadith. In it he gathered damning judgments 
on respected Sunni hadith transmitters from prominent members of 
the ahl al-hadith themselves. Yet al-Balkhi never refers to Muslim and 
does not mention al-Bukharl in the chapter citing evaluations of Sunni 
transmitters.''^" Instead, he relies principally on Ibn Hanbal, 'All b. al- 
Madlnl, Abu Khaythama, al-Shafi'l, Malik, and Yahya b. Ma'in. 

In his Fihrist, written in 377/987-8, Ibn al-Nadim (d. after 385- 
8/995-8) lists al-Bukharl and Muslim as two of sixty-three transmission- 
based jurists in Islamic history. Along with others like Sufyan al-Thawrl, 
'All b. al-Madlnl and al-Tirmidhi, he describes them simply as ex- 
perts and trustworthy narrators {thiqa)}''^ Neither of their biographies. 



'*" Al-Hasan b. 'Abd al-Rahman al-Ramahurmuzi, al-Muhaddith al-fddil hayn al- 
rdwi wa'l-wdy, ed. Muhammad 'Ajjaj al-Khatib {[Beirut]: Dar al-Fikr, 1391/1971), 
229-31. 

'« Ibn 'Adi, al-Kdmil, 1:129, 

''"" Abu al-Qasim Abdallah al-Ka'bl al-Balkhl, Qubul al-akhbdr wa ma'rifat al-rijdl, ed. 
Abu Amr al-Husaynl b. Muhammad b. Abd al-Rahman, 2 vols. (Beirut: Dar al-Kutub 
al-'Ilmiyya, 1421/2000), 2:149. 

'^' Abu al-Faraj Muhammad b. Ishaq Ibn al-Nadim, The Fihrist, ed. and trans. Bayard 



90 CHAPTER THREE 

however, matches that of the later Kufan chief judge and hadith 
scholar Abu 'Abdallah al-Husayn b. Isma'il al-Mahamill (d. 330/942); 
Ibn al-Nadlm states that no one was more knowledgeable than him 
in hadith. '^2 



Reception: the Immediate Response to al-Bukhan's and Muslim's Works 

Al-Bukhari and Muslim functioned as magnets for hadith transmission 
during their lives, selecting choice narrations for the Sahihs that formed 
their lasting legacy. But strikingly enough, they themselves proved 
insignificant in the continuing transmission of hadith through living 
isndds. In his annals listing the significant hadith scholars who died in 
the second half of the third/ninth century and the first few decades of 
the fourth/tenth, Ibn al-jawzl (d. 597/1200) lists seventeen who studied 
with Ishaq b. Rahawayh, twenty-two with 'All b. al-Madlnl, but only one 
with al-Bukharl or Muslim. Indeed, other contemporaries of al-Bukharl 
and Muslim completely obviated their role in the transmission of 
hadlths. Abu al-Qasim 'AbdaUah b. Muhammad al-BaghawI of Baghdad 
heard from what al-Khatib al-Baghdadi terms "uncountable masses" 
of hadith transmitters, including Ibn Hanbal, 'All b. al-Madlnl and 
Yahya b. Ma'in. He died at the age of 104 or 110 in 317/929-30 
and was thus much sought after for his elevated isndd to that greatest 
generation. The major scholars who heard from al-BaghawI directly, 
such as al-Daraqutnl (d. 385/995), or through his isndd, like al-Khatib 
al-Baghdadi, had no need to refer to transmitters like al-Bukharl or 
Muslim for living transmission.'"'^ Even in the case of hadlths that 
appeared in Muslim's Sahih, for example, later hadith scholars like al- 
Dhahabl preferred to narrate them through al-BaghawI in their own 
hadith collections.'''* 



Dodge (New York: Columbia University Press, 1970; Chicago: Kazi Publications, 1998), 
555-6. Citations are to the Kazi edition. 

'^2 Ibn al-Nadim, The Fihrist, 560; cf. al-Khallll, al-Irshdd, 193; al-Khatib, Tdrikh 
Baghdad, 8:19-22. 

''' Al-BaghawI is often referred to as Ibn Manl' or even Ibn Bint al-Mani'. Some 
were skeptical of al-Baghawl's narration from Yahya b. Ma'in. Al-Khallll says that 
he could narrate from one hundred shaykhs that no one else in his time had met; 
al-Khallll, al-Irshdd, 192. The last of al-Baghawl's students, Muhammad b. Ahmad b. 
•All al-Baghdadi, died in 399/1008-9. 

''' Al-Dhahabi, Tadhkirat al-hujfaz, 4:159. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 9 1 

This focus on the living isndd and the veneration paid to previous 
generations of hadlth scholars also dominates the immediate reception 
of al-Bukhari's and Muslim's works in the hadith community. The 
hadlth scholars' conception of their own tradition, as shown in the 
early and mid-fourth/tenth-century works of Ibn Abi Hatim al-Razi, 
Ibn Hibban and Ibn 'Adi, distinguishes between the colossal genera- 
tion of Ibn Hanbal and 'All b. al-Madlni and that of their students 
al-Bukhari and Muslim. Many in the hadith community, such as the 
influential bloc of Razi scholars in Rayy immediately balked at what 
they perceived as the elitism and finality of the two works, accusing 
al-Bukharl and Muslim of insolence. 

The reaction of the Rayy scholars to Muslim's Sahih during his own 
lifetime portrays his work as an act of egoism that could undermine 
the legal methodology of the transmission-based scholars. The chief 
critics of Muslim's Sahih were Abu Zur'a al-Razi and his colleague 
Muhammad b. Muslim Ibn Wara al-Razi (d. 270/884). Along with 
Abu Hatim, Abu Zur'a was an institution of hadith study in Rayy. 
Even at middle age he had earned the respect of prominent scholars 
such as Ishaq b. Rahawayh, who said that "any hadlth that Abu Zur'a 
al-Razi does not know has no basis." ''^^ Muslim met several times with 
the two Razis and their colleague Ibn Wara in Rayy. Their reaction 
to his Sahih clearly communicates the initial shock that the notion of a 
book of purely authentic hadlths had on some scholars in the hadith 
community. It has been preserved in Abu Zur'a's Kitdb al-du'afa' wa 
ajwibatuhu 'aid as 'Hat al-Bardha'i, a compilation of both Abu Zur'a's and 
Abu Hatim's opinions on transmitters as transcribed by their student 
Abu 'Uthman Sa'id b. 'Amr al-Bardha'i (d. 292/905), who also studied 
with Muslim: 

I saw Abu Zur'a mention the Sahih book written by Muslim b. al-Hajjaj, 
then [that of] al-Fadl al-Sa'igh'"''' based upon it {'aid mithdlihi). Abu Zur'a 
said to me, "These are people who wanted prominence {taqaddum) before 
their time, so they did something for which they show off {yatashawwajun 
bihi); they wrote books the likes of which none had written before to gain 
for themselves precedence [riydsa] before their time." One day, when I 
was present, a man came to [Abtl Zur'a] with the Sahih transmitted from 



"' Ibn 'Adi, al-Kdmil, 1:141. 

'^1^ This is Abu Bakr al-Fadl b. al- Abbas al-Sa'igh al-RazI (d. 270/883). I have found 
no other mention of this book. See al-Khatib, Tdnkh Baghdad, 12:363; al-Dhahabi, 
Tadhkirat al'hujfdz, 2:132^3; idem, Tdnkh al-isldm, 20:149-50. 



92 CHAPTER THREE 

Muslim, and Abu Zur'a started to look through it. When he came across 
hadlths from Asbat b. Nasr he said to me, "How far this is from sahihl 
He includes Asbat b. Nasr in his book!" Then he saw in the book Qatan 
b. Nusayr, so he said to me, "This is even more overwhelming than the 
first one! Qatan b. Nusayr [incorrecdy] attributed hadlths from Thabit 
[al-Bunani] to Anas [b. Malik]." Then he looked and said, "[Muslim] 
narrates from Ahmad b. 'Isa al-Misri in his Sahih book: did you not see 
the people of Egypt complaining that Ahmad b. 'Isa," and he pointed 
to his tongue as if to say, 'lies,' then said to me, "[Muslim] narrates from 
the likes of them and leaves out [hadlths] from Muhammad b. 'Ajlan 
and those like him. He is making a path for the people of heresy {bida') 
against us, for they see that they can respond to a hadlth that we use as 
proof against them by saying 'That is not in the SahihV" 

I saw him denigrating the book and censuring it, so when I returned 
to Naysabur on the second occasion I mentioned to Muslim b. al-Hajjaj 
AbQ Zur'a's rejection of his narrations in the book from Asbat b. Nasr, 
Q_atan b. Nusayr and Ahmad b. 'Isa. Muslim said to me, "Indeed I 
did deem [the book] 'Sahih,^ and what hadlths I included from Asbat, 
Qatan and Ahmad have been narrated by [other] trustworthy narrators 
(thiqdt) from their [Asbat, Qatan and Ahmad's] shaykhs, except that these 
[that I included] came from [Asbat and them] through shorter isndds 
ibi'l-irtifd'). But I also have these [hadlths] from those who are more 
reliable than them [Asbat et al.] via longer isndd'i [bi-nuzul] . . . and the 
core report of the hadlth is well known through the transmission of 
trustworthy transmitters." 

Muslim came to Rayy and it reached me that he went out to Abil 
'Abdallah Muhammad b. Muslim b. Wara, and he received him coldly 
{fa-jajdhu) and chastised him for the book, saying essentially what AbU 
Zur'a said: this opens us up to the people of bida '. So Muslim apologized 
to him and said, "Indeed I produced this book and declared it authentic 
{sihdh), but I did not say that that hadlths I did not include in this book 
are weak. Rather, I produced this from sahth hadlths to be a collection 
for me and those who transmit from me without its authenticity being 
doubted. I did not say that everything else is weak..." and Ibn Wara 
accepted Muslim's apology and transmitted [the book].'"' 



'^' This quote is found in its entirety in Abu Zur'a 'Ubaydallah b. 'Abd al-Karlm 
al-RazI, Abil ^ur'a al-Rdzi wajuhuduhuji al-sunna al-nahawiyja ma'a tahqiq kitdbihi al-Du'aJd' 
wa ajwibatihi 'aid as'ilat al-Bardha'i, ed. Sa'dl al-Hashimi, 3 vols. (Medina, Cairo: Dar 
al-Wafa' and Maktabat Ibn al-Qayyim, 1409/1989), 2:674-6; al-Khatib al-Baghdadl, 
Tdnkh Baghdad, 5:28-30 (biography of Ahmad b. 'Isa al-Tustarl al-Misri); al-MaqdisI 
and al-HazimI, Shurut al-a'imma al-sitta wa shurut al-a'imma al-khamsa, 60-3; al-NawawI, 
Shark Sahih Muslim, 1:135-6; cf. for partial quotes, Ibn al-Salah Siydnat Sahih Muslim, 
99-iob; cf. Abu Muhammad Muhyl al-Din 'Abd al-Qadir Ibn Abl al-Wafa', al-Jawdhir 
al-mudiyyaji tabaqdt al-hanafiyya, ed. 'Abd al-Fattah Muhammad al-Halw, 5 vols. (Giza: 
Mu'assasat al-Risala, 1398-1408/1978-1988), 4:569. 



THE GENESIS OF AL-BUKHARI AND MUSLIM 93 

So charged is al-Bardha'i's report that it seems miraculous that we have 
received it from a provenciated source.'"'" Indeed, Abu Zur'a's and Ibn 
Wara's reaction to the Sahih as well as Muslim's concessions highlight 
issues that would later prove some of the most hotly debated questions 
in the hadlth tradition. The Rayy scholars raise three objections to 
Muslim's Sahih. First, they decry it as impertinent glory-seeking. Second, 
they disagree with Muslim's judgment concerning the reliability of 
some transmitters, arguing that his criteria are flawed and subjective. ''^^ 
Finally, they worry that producing a sahih compilation could hinder the 
use of other hadiths that would be considered lackluster in comparison. 
Absolute authenticity had never been the determining factor in the use 
of hadiths in either elaborating law or polemics with the ahl al-hadith'i, 
rationalist foes. We thus detect the immediate and palpable fear that a 
definitive sahih book would be used to exclude all other materials. 

The concerns of the Razis seem to have been pervasive, with al- 
Bukhari also attracting criticism from younger experts like al-Nasa'i for 
the seemingly arbitrary omission of hadiths from respected transmit- 
ters like Suhayl b. Abl Salih."'" Both al-Bukharl and Muslim were thus 
forced on more than one occasion to deny that their works encom- 
passed all authentic hadiths. Muslim did so in the body of his Sahih 
in a rare response to a question, saying that his book only contains 
those authentic hadiths that "were agreed upon (ajma'u 'alayhaj" and 
excludes other nonetheless worthy ones.'^' Ibn 'Adi provides an early 



'"''' Sa'dl al-Hashiml's edition of al-Bardha'i's text is based on a manuscript from the 
Kopriilii Library in Istanbul (#3/40 in a 2 7M<:' notebook). This report appears in the 
above sources but it is always narrated through the same initial imdd from al-Bardha'l. 
Al-Khatib al-Baghdadi and al-HazimI have isnddi to Abu Bakr Ahmad b. Muhammad 
al-BarqanI <- Abu al-Husayn Ya'qub b. Musa al-Ardablll <- Ahmad b. Tahir b. al- 
Najm al-Mayyaniji <- Sa'ld b. Amr al-Bardha'l. Al-KhaMi (d. 446/1054), who does 
not mention this story, tells us that al-Bardha'l studied with Abtl Zur'a al-RazI. The 
isnad of AbU Zur'a — > al-Bardha'l — > Ahmad b. Tahir b. al-Najm al-Mayyaniji is also 
established elsewhere separately by al-Khallll; cf. al-Khallll, al-Irshdd;l09 , 129, 286. 

'^^ Interestingly, Muslim is quoted by his student Makkl b. 'Abdan as supposedly say- 
ing, "I showed my book to AbQ Zur'a al-RazI and everything that he indicated as having 
a flaw {'ilia) I left out. And what he said, 'This is sahih with no 'ilia,' I included." The 
earliest appearance of this quote I have found is in the work of Abu 'All al-GhassanI 
al-Jayyam of Andalusia (d. 498/1 105); al-Tanbih 'aid al-awhdm al-wdqi'aji Sahih al-imdm 
Muslim, ed. Muhammad Abu al-Fadl (Rabat: Wizarat al-Awqaf wa al-Shu'un al- 
Islamiyya, 1421/2000), 39; al-Qadi lyad, Ikmdl al-mu'lim bi-fawd'id Muslim, 1:82; Ibn 
al-Salah, Sijdnat Sahih Muslim, 68; al-NawawI, Sharh Sahih Muslim, 1:121. 

'60 Muhammad b. al-Husayn al-SulamI, "Su'alat Abl 'Abd al-Rahman al-SulamI 
li'1-Daraqutni," MS Ahmet III 624, Topkapi Sarayi, Istanbul: 162a. 

'*' Sahih Muslim: kitdb al-saldt, bdb al-tashahhud. Later analysts believed that the group 



94 CHAPTER THREE 

quote from al-Bukhari that he had left many sahih reports out of his 
collection, which he entitled an "abridged {mukktasar)" compilation, in 
order to keep its size manageable.'®^ We shall see in Chapter Five how 
prescient the Razis' concerns were. 

Muslim's response to Ibn Wara provides a fascinating glimpse into 
the pre-canonical life of his Sahth. If a canon is a text endowed with 
authority and made binding on a community, its converse is a powerless 
text that reaches no farther than its author Yet this is precisely how 
Muslim is forced to describe his Sahih in order to placate Ibn Wara. 
He is forced to reduce his book to a private "collection for me and 
those who transmit from me." In the face of resistance, we thus see 
that Muslim was obliged to deny his work the features that would one 
day accord it canonical status. ""' 

One of the earliest recorded reactions to al-Bukhari's Sahih seconds 
the accusation of impudence leveled at Muslim by Abu Zur'a. Maslama 
b. Qasim al-Qurtubi (d. 353/964)"'* recorded a story about al-Bukhari 



that Muslim was referring to as "having agreed upon" these hadiths consisted of Ibn 
Hanbal, Yahya b. Ma'ln, 'Uthman b. Abl Shayba and Sa'ld b. Mansur al-KhurasanI; 
Abu Hafs 'Umar b. Raslan al-BulqInl (d. 805/1402-3), Mahdsin al-istildh, in Muqaddimat 
Ibn al-Saldh wa Mahdsin al-istildh, 162. 

"'^ Ibn 'Adi, Asdmi, 68. 

"'' Al-Bukharl is also reported to have shown his Sahth to senior scholars such as 'All 
b. al-Madlnl and Ibn Hanbal. This report only appears in a very late source, however: 
Ibn Hajar's (d. 852/1449) Hady al-sdn. He quotes Abu Ja'far Muhammad b. 'Amr al- 
'Uqayll's (d. 323/934) statement that these scholars acknowledged the authenticity of 
the Sahih with the exception of four hadiths. This information does not appear in the 
one work that has survived from al-'Uqayll, his Kitdb al-du'ajd' al-kabir. Ibn Hajar had 
access to at least one other work by al-'Uqayll, his Kitdb al-sahdba, so he might have 
had a source for this quote. Al-'Uqayll was very familiar with al-Bukharl's al-Tdrikh 
al-kabir (one of his principal sources in his Kitdb al-du'ajd') and his So-hih, and he had 
studied with Ibn Hanbal's son 'Abdallah. It is thus not improbable that he could have 
transmitted this information about the evaluation of the Sahih- But since 'All b. al- 
Madlnl died in 234/849, whatever al-Bukharl might have showed him was probably 
only a very early draft of the work. See Ibn Hajar, Hady al-sdri; 7, 676; al-'Uqayll, Kitdb 
al-du'ajd' al-kabir, 1:48—9 (editor's introduction). 

"'* In his Tahdhib al-tahdhib, the only place I have found this story, Ibn Hajar cites the 
source only as "Maslama." We know that this is Maslama b. Qasim, however, because 
in his al-Mu 'lim bi-shuydkh al-Bukhdri wa Muslim, Abu Bakr Muhammad b. Isma'll Ibn 
Khalfun (d. 636/1238-9) duplicates the first line of the story [allqfa 'Ali b. al-Madini 
Kitdb al-'ilal wa kdna danirf" bihi...) exactly in a quote from Maslama b. Qasim. Ibn 
Hajar's version then continues with the insulting story above, while in Ibn Khalfun's 
version Maslama goes on to tell how 'All did not lend his book to anyone or narrate 
it because of its valuable content, then states "and he [Maslama] mentioned the story 
{wa dhakara al-qissd)." See Abu Bakr Muhammad b. Isma'll Ibn Khalfun, al-Mu'lim 
bi-shuyUkh al-Bukhdri wa Muslim, ed. Abu 'Abd al-Rahman 'Adil b. Sa'd (Beirut: Dar 
al-Kutub al-'Ilmiyya, 1421/2000), 464. For an article discussing Maslama b. Qasim 's 



THE GENESIS OF AL-BUKHARI AND MUSLIM 95 

that paints him as a plagiarist whose brilliant Sahih was truly the work 
of his famous teacher 'All b. al-Madlni. Maslama reports that 'All had 
a book detailing the flaws in various hadlth narrations [Kitab al-'ilaiy^'^ 
that represented his mastery of hadith criticism. One day when 'All 
had gone to view some of his properties, al-Bukharl came to one of his 
sons and bribed him to lend him the book, which al-Bukhari promptly 
had duplicated by a copyist. When 'All returned and held a session for 
hadlth study, al-Bukhari's knowledge rivaled his teacher's. 'All grasped 
what had occurred from his student's exact imitation of his own work 
and was so saddened that he eventually died of grief Having no further 
need of his teacher, al-Bukhari returned to Khurasan and compiled 
his Sahih, gaining fame and followers."*^ 

Maslama b. Qasim was from Cordova, but sometime before 320/932 
he traveled east to Egypt, greater Syria, Mecca, Wasit, Basra, Baghdad 
and Yemen before returning to Spain after losing his vision."'' He cer- 
tainly had a copy of al-Bukhari's al-Tdnkh al-kabir, since Ibn Hajar states 
that Maslama compiled a one-volume book on hadith transmitters {tdnkh 
fial-rijal) intended to cover those not mentioned in al-Bukharl's diction- 
ary (including some of Maslama's own contemporaries). "''' Maslama 



heresiographical contributions, see Maribel Fierro, "Batinism in al-Andalus. Maslama 
b. Qasim al-Qurtubl (d. 353/964), author of the Rutbat al-Hakim and the Ghdyat al- 
Hakim," Studia hlamica 84,2 (1996): 87-112. 

""^ This book could not possibly be 'All's Kitdb al-'ilal that has come down to us 
today. While the book Maslama describes contains what seems to be the sum total 
of 'All's corpus of hadlth criticism, his extant work is very small and only deals with 
several dozen narrations. It is possible that the book mentioned here is a work of 'All's 
that Ibn al-Nadim describes as a musnad accompanied by 'Hal commentary; see Ibn 
al-Nadim, The Fihrist, 556. 

'^^ Ibn Hajar, Tahdhib al-tahdhib, ed. Mustafa 'Abd al-Qadir 'Ata (Beirut: Dar al-Kutub 
al-'Ilmiyya, 1415/1994), 9:44; Najml, Sayn dar Sahihayn, 72. 

'^' Maslama was criticized as a weak transmitter, but was defended by others 
who said that he simply was not very intelligent (da'if al-'aqt). He was also accused 
of anthropomorphism, but, in light of the controversial material he recorded about 
al-Bukharl, these are probably reactionary ad hominem attacks by later commentators; 
see Muhammad b. al-Futuh al-Humaydl, Jadhwat al-muqtabis Ji dhikr wuldt al-Andalus wa 
asmd' Tuwdt al-hadith wa ahl al-fiqh wa al-adab, ed. Muhammad b. Tawit al-Tanji (Cairo: 
Maktabat al-Nashr al-Thaqafi al-Islami, 1371 /[1 952]), 324; al-Dhahabi, Tdnkh al-isldm, 
26:98; idem, Sijar a'tdm al-nubatd', 16:1 10; idem, Mizdn al-i'fiddlji naqd al-rijdl, 4:1 12; cf 
Ibn Hajar, Lisdn, 6:35-6; cf Tahir al-Jaza'irl al-Dimashql (d. 1338/1919-20), Tawjih 
al-na^ar ild usul al-athar, ed. 'Abd al-Fattah Abu al-Ghudda, 2 vols. (Aleppo: Maktab 
al-Matbu'at al-Islamiyya, 1416/1995), 1:302. Although he visited Baghdad, al-Khatlb 
does not mention him in his history. 

""" Ibn Hajar, Lisdn, 6:35. Here Ibn Hajar quotes Abu Ja'far al-Maliqi's Tdnkh. We 
know that Maslama's Tdnkh included such contemporaries as Abu Ja'far al-'Uqayll 



96 CHAPTER THREE 

probably heard the story about al-Bukhari stealing his teacher's work 
after his arrival in the Islamic heartlands (i.e. after 320/932) but before 
his death in 353/964. We can thus assume that it was in circulation by 
the early 300s/900s, if not before. 

The story of 'All b. al-Madini and the SaMi is almost certainly untrue, 
since refusing to transmit one's work to students would be extremely 
unusual among scholars of hadith. Maslama's own preoccupation 
with al-Bukharl's Tdrikh and the fact that the story recognizes that the 
Sahih was a major accomplishment points to a more subtle motivation. 
Regardless of the high quality of his Sahih, al-Bukhari's work clashed 
with the atavistic traditionalism endemic among the ahl al-hadith. For 
them the community was always in decline as it grew more distant 
from the Prophet, and students could do no more than try to preserve 
their masters' knowledge. The creator of Maslama's story could only 
interpret al-Bukhari's unprecedented contribution as an act of insub- 
ordination. 

Maslama's Tdnkh, however, illustrates another important aspect of 
the community's reception of al-Bukharl's works: for decades after his 
death, al-BukharT was much better known for his Tdrikh than for his 
Sahih. In his Muntaiam, Ibn al-Jawzi mentions someone narrating al- 
Bukharl's Tdrikh fully a century before the first person is mentioned as 
narrating his Sahih-^^'' Also, almost seventy years before the first scholar 
compiled a hadith collection using the Sakih as a template, al-Husayn b. 
Idrls al-Ansarl (d. 301/913-4) used the Tdrikh as a format for his own 
biographical dictionary'™ When al-Bukhari's student and a compiler of 
a famous hadith collection himself, Abu 'Isa al-Tirmidhi, said that he 
had never seen anyone with al-Bukhari's command of the narrations 
of hadith and the lives of their transmitters, he was referring explic- 
itly to the scholar's Tdrikh al-kabir}'^ Muhammad b. 'Abd al-Rahman 
al-Daghull (d. 325/936-7) of Sarakhs, who had studied hadith with 
al-Bukhari's rival al-Dhuhli, nonetheless said that al-Bukharl's Tdrikh 
was one of the four books with which he never parted.''^ Abu JaTar 



(d. 323/934), since this is one of the sources al-Dhahabi relies on for his biography 
of al-'Uqayll in Tadhkirat al-huffa^. 

''•"' Ibn al-Jawzl, al-Muntazam, 13:362 and 15:270. 

'™ Al-Dhahabi, Tadhkirat al-huffai, 2:192. 

'" Ibn Rajab, Shark 'Hal al-Tirmidhi, 1:32. Al-Tirmidhl's hadith collection also 
includes, however, the earliest actual mention of al-Bukharl's "Jdmi" (i.e., his Sahih); 
Jdmi' al-Tirmidhi: kitdb al-tahdra, bah mdjd'ajial-istinjd'bi'l-hajarajn. 

"^ The others were al-Muzanl's Mukhtasar, Khalll b. Ahmad's dictionary Mtdb al- 



THE GENESIS OF AL-BUKHARI AND MUSLIM 97 

al-'Uqayli's (d. 323/934) Kitab al-du'afd' al-kabir (Great Book of Weak 
Transmitters) relies on al-Bukhari as the single largest source of 
evaluations for transmitters. Al-'Uqayli frequently refers to al-Bukhari's 
al-Tdnkh al-kabir, which he calls the scholar's "great book [al-kitdb al- 
kabtr)," but never mentions the Sahih}''' The only occasion on which 
al-Ramahurmuzi mentions al-Bukhari in his al-Muhaddith al-Jadil is in 
relation to his Tdnkh."* 

While it was Muslim's Sahih that attracted the critical ire of the 
hadith scholars in Rayy al-Bukharl's Tdnkh became the locus of drama 
and debate for the Razis. In the first written response to any aspect 
of al-Bukharl's oeuvre, Ibn Abi Hatim penned a short book correcting 
errors he detected in the Tdnkh al-kabir. The involvement of Ibn Abi 
Hatim, his father and Abu Zur'a with the Tdnkh became even more 
problematic when a prominent muhaddith of Naysabur, Abu Ahmad 
Muhammad b. Muhammad al-Hakim (d. 378/988), accused them of 
plagiarizing al-Bukharl's work. Al-Hakim al-Naysaburi, Abu Ahmad's 
friend and student, reports from him that when he was in Rayy once 
he saw Ibn Abi Hatim reading his al-Jarh wa al-ta'dil to students. He 
recognized its contents as that of al-Bukharl's Tdnkh and inquired as 
to why Ibn Abi Hatim had attributed this work to his father and Abu 
Zur'a. A student replied that al-Bukharl's Tdnkh had so impressed Abu 
Hatim and Abu Zur'a that they had taken it as the basis of their work, 
sitting with Ibn Abi Hatim so that he could record some modifications 
to the work and then ascribe it to them."'^ 



'ayn, and the cultured political treatise Kabla wa dimna; al-Dhahabi, Tadhkirat al-huffdz, 
3:30. 

'" Al-'Uqayll, Kitdb aMuafi' al-kabir, 1:285, 3:345, 4:292. 

'" Al-Ramahurmuzl, al-Muhaddith al-Jadil, 310. 

"' Al-Khapb, MUdih awhdm al-jam'wa al-tafifq, 2 vols (Hyderabad: Da'irat al-Ma'arif al- 
'Uthmaniyya, 1378/1959), 1:8-9; Yaqutal-Hamawi,Af«}'am«/-A«i/a/z, 2:799; cf. al-Dliahabl, 
Tadhkirat al-huffa^, 3:124. Yaqut and al-Dhahabl's reports are taken from al-Hakim al- 
NaysabQrl, but al-Dhahabl's lacks the last concluding statement that Ibn Abi Hatim 
attributed the book to his father and Abtl Zur'a al-RazI. Abu Ahmad al-Hakim also 
voices his accusations in his own Kitdb al-kund, which al-Dhahabl quotes in his biogra- 
phy of al-Bukharl and which is also partially and lazily quoted in al-Khallll's al-Irshdd; 
see al-Khallli, al-Irshdd, 380; cf. al-Dhahabi, Tdrikh al-isldm, 19:259; Ibn Hajar, Hady 
al-sdri, 11-12. 



98 CHAPTER THREE 

Conclusion 

As the next chapter will demonstrate, the Sahihayn, and Muslim's 
Sahih in particular, quickly became objects of study and imitation in 
Khurasan, Eastern Iran and, eventually, Baghdad. We have seen, how- 
ever, that during their lives and in the immediate wake of their deaths 
al-Bukharfs and Muslim's Sahths met with rejection and scorn among 
important elements of the hadlth scholar community. The tradition 
of hadlth collection and study rested on a veneration for the past as 
the repository of the Prophet's sunna and the only authentic source 
for interpreting Islam. Although they had developed a methodology 
for distinguishing between authentic and forged hadiths, for transmis- 
sion-based scholars the Prophet's charismatic authority rendered even 
weaker hadiths legitimate tools for understanding the faith. For scholars 
like Abu Zur'a al-Razi, a collection limited to purely authentic hadiths 
unnecessarily delimited the potential application of the Prophet's sunna 
in Muslim life and debate. Furthermore, hadlth scholars cultivated a 
worldview in which later generations could at best struggle to preserve 
their predecessors' transmission of the normative past. During al- 
Bukharl's and Muslim's lives and the century after their deaths, hadlth 
scholars' native perception of their tradition viewed them as merely 
two experts among many, placing them in positions junior to their 
teachers. Al-Bukhari in particular was also tainted with scandal and 
accusations of heresy. For Abu Zur'a, for his colleagues in Rayy and 
for whomever first circulated accusations of al-Bukharl's plagiarism, the 
Sahihayn were acts of insubordination by students seeking to supplant 
their teachers and defy tradition. For common Muslims and scholars 
alike the collection and transmission of hadiths through living isndds 
back to the Prophet remained a dominant pious and legally significant 
activity for centuries after the sahth movement. Al-Bukhari and Muslim 
would prove insignificant in the continued transmission of hadiths, but 
their Sahihs, became institutions that soon rivaled it. 



CHAPTER FOUR 

A 'PERIOD OF INTENSE CANONICAL PROCESS': 

IMAGINATION AND THE STUDY OF THE SAHlHATN 

IN THE LONG FOURTH/TENTH CENTURY 



Introduction 

With the exception of Deuteronomy's revelation to the court of King 
Josiah in II Kings, canonical texts do not fall intact from the heavens. 
Whether scriptural or literary, they pass through phases of use and study 
within a community before their canonization. Scripture must earn the 
devotion of a congregation before priests can declare it authoritative, 
and a body of critics must first study and explore literary works before 
dubbing them classics. Books are thus not written as canons. This sta- 
tus is bestowed upon them by a community engaged in a process of 
self-identification or authorizing institutions. The books of the New 
Testament were not all written as scripture, a role already played in 
early Christian communities by the Greek edition of the Hebrew Bible. 
What became the canonized New Testament was a diverse selection 
of writings used in services that eventually became widely recognized 
guides to Christian devotion. The usage of the word canon as 'list' 
in the first centuries CE originated in this roster of familiar books.' 
The books of the New Testament canon had therefore already proven 
effective at conveying a particular understanding of Christ's mission to 
a certain audience. 

This process of use and familiarization was not limited to passive 
reception. Paul's canonical epistle to the Corinthian congregation (2 Corin- 
thians) probably originally consisted of at least two separate letters writ- 
ten at different times and later pasted together for circulation amongst 
Paul's churches.^ Such editorial activity highlights the role of clerics 
or scholars in molding proto-canonical texts after they have left the 
hands of their authors. In the words of James Sanders, this "period 



Gamble, The New Testament Canon, 17-lf 
Ehrman, The New Testament, 299. 



100 CHAPTER FOUR 

of intense canonical process" between the crafting of a text and the 
stabilization of a discrete canon represents a crucial interaction between 
text and audience. It is in these periods that audiences "shaped what 
they received in ways that rendered [the texts] most meaningful and 
valuable for them."^ 

Periods of intense canonical process are thus periods of intensive 
study. Before the emergence of a canon, texts must receive critical 
attention from scholars who catalog their contents, detail their merits 
and build around them that edifice of oral or written scholarship that 
distinguishes the familiar and valuated from the banal or unknown. 
Beyond the valorization that a scholarly class bestows on written works, 
in pre-modern times intense study was required merely to produce 
a coherent text. The folkloric tradition of the Trojan War thrilled 
multitudes of small Greek audiences for most of the first millennium 
BCE. Yet as a scattered and diverse body of oral epic the Iliad and 
Odyssey could never have become classics of Hellenistic literature or 
cornerstones of the Western literary canon. The first 'edition' of the 
Homeric epics was produced by Antimachus of Colophon (fl. 410 
BCE) after centuries of fermenting as an oral-formulaic tradition. In 
the great Hellenistic Library of Alexandria, scholars like Zenodotus of 
Ephesus (fl. 270 BCE) initiated the first studies of the Homeric epics, 
editing and collecting manuscripts, creating lexicons and producing a 
standardized vulgate tradition. Alexandrian scholarship on Homeric 
works continued unabated in the following decades, with great writers 
and critics such as ApoUonius of Rhodes and Rhianus of Crete debat- 
ing and producing critical editions.* It was these relatively standardized 
texts that Hellenistic scholars declared the 'canons' of Greek language 
worthy of imitation. 

Certain Muslim scholars recognized that an intensive familiarization 
with a text was a prerequisite for its canonization. Shah Wall Allah of 
Delhi (d. 1762) felt that the treatment a book received after its composi- 
tion was a crucial characteristic of a mainstay authentic hadlth collec- 
tion. In addition to its author purposing a work of authentic hadlths 
and succeeding in that task, such a book must be studied, its rare or 
difficult ighanb) words explained and its legal implications derived. It 



' Sanders, 30. 

' Rudolph PfeiiFer, History of Classical Scholarship: From the Beginnings to the End of the 
Hellenistic Age (Oxford: Oxford University Press, 1968), 148-9. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 101 

must be edited and refined [tahdhib), and historians must identify all 
its transmitters as well as their death dates.' Thus in the century after 
al-Bukhari's death, scholars strove painstakingly to understand his 
methodology, identify his obscure transmitters (sometimes only referred 
to by their first names) and locate all the narrations of one Prophetic 
tradition scattered throughout his work. 

Yet periods of intense canonical process do not only involve this req- 
uisite study and familiarization with a text. Separately, they involve the 
community developing the conceptual ability to endow texts with some 
binding authority. For a canon to form, a community must imagine texts 
that have transcended the normal status of books as objects of study or 
usage and are able play some loftier role. Periods of intense canonical 
process are times in which communities' conception of the authority a 
text can acquire leaps forward due to real and pressing needs.'' 

Although the Sahihayn met with resistance during the lives of their 
authors and in the wake of their deaths, these two works quickly 
emerged as formative texts in certain areas of the Nile-Oxus region. 
Beginning in Muslim's home city of Naysabur and later in Jurjan and 
Baghdad, scholars began viewing the Sakths not as threats to the living 
transmission of the Prophet's sunna but rather as vehicles for express- 
ing their personal link to his authority and interpreting his teachings 
according to their own local agendas. Hadith scholars began using the 
Sahihayn and the methods of their authors as templates for their own 
hadith collections. These mustakhraj books, however, required a detailed 
mastery of al-Bukhari's and Muslim's transmitters, the permutations of 
the hadiths they included as well as their requirements for authenticity. 
The mustakhraj cults that formed in Naysabur around Muslim's Sahih, in 
Jurjan around al-Bukhari's, and finally in Baghdad around the conjoined 
Sahihayn thus sparked a flurry of studies on the two books and their con- 
stitutive elements. Scholars not only detailed al-Bukharl's and Muslim's 
works, they also interacted with their methodologies. Just as Abu Zur'a 
al-Razi had questioned Muslim's right to delimit authentic traditions, 
so did later scholars apply their own requirements for authenticity to 
the Scihihayn, identifying what they considered errors and questioning 
why other hadiths had not merited a place in the collections. 



Shah Wall Allah, Hujjat Allah al-baligha, 1:133. 
Sanders, 32-33. 



102 G HAPTER FOUR 

As we shall see, the network of scholars who devoted themselves 
to employing and studying al-Bukhari's and Muslim's Sahihs between 
the last quarter of the third/ninth century and the first half of the 
fifth/ eleventh was distributed with remarkable geographic and chrono- 
logical consistency. Equally important, however, was their ideological 
makeup. The study of the Sahihayn fell to neither the iiber-Sunnis who 
had ostracized al-Bukhari nor the historically hadlth-wary HanafTs. It 
was a more moderate group of transmission-based scholars belonging 
to the nascent Shafi'i school that forged the proto-canon. 

In this chapter we will examine this network of scholars and their 
accomplishments during what one might term the long fourth century, 
that period between the deaths of the Shaykhayn and the widespread 
acknowledgment of the canon in the mid-fifth/ eleventh century. This 
periodization is not merely heuristic. As we shall see, it reflects the 
uniqueness of a time characterized by fleeting genres and an often 
frustrating liminality in Islamic intellectual culture. 

The long fourth century also proved a period in which important 
elements of the broader Muslim community began articulating the 
notion of a hadith collection acting as a locus of communal consensus. 
Whether as common ground between different schools of thought or 
simply common references in an increasingly diverse hadith tradition, 
this period of intense canonical process left the Muslim community 
with the imaginative capability of endowing hadith works with a new 
epistemological status. 



Sahihayn Network Chart: 
Study and Usage in the Long Fourth Century 
Key: 

• ► : Personal study relationship/teacher-student relationship 

• ► : Transmission of a scholar's books to another scholar 

^ y ; Transmission or transmitter of al-Bukharl's Sahih 

• ► : Transmission or transmitter of Muslim's Sahih 

The following chart describes the location, dates, written works and scholarly 
relationships of the network of scholars who studied and employed the Sahihayn 
between 270 and 450AH. When required, some later figures are included with 
their death dates noted. For references, see Appendix I. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 



103 













104 CHAPTER FOUR 

The Mustakhraj Genre 

The phenomenon of the mustakhraj forms a bizarrely short and circum- 
scribed chapter in the history of Islamic religious thought. These works 
were produced from about 270/880 to 480/1085 in the Nile-Oxus 
region and then exited the stage of cultural expression.' They mark 
a transitional period between the time when one could realistically 
cultivate one's own isnadi, to the Prophet and the time when books of 
hadith replaced this direct connection. A scholar produced a mustakhraj 
by compiling a book of hadlths based on an existing collection that 
he used as a template. For each of the hadiths in the template book 
the author would use his own narration of the hadith, with the isnad 
extending from him back to the Prophet. The very term mustakhraj con- 
notes 'seeking to include' certain narrations from the Prophet. Isndds, 
in these mustakhrajs would generally join with the isndds of the template 
collection at the teacher of the original collector, following the same 
isndd from that point to the Prophet." 

Mustakhrajs could vary in the degree to which they adhered to the for- 
mat and contents of the template collection. Abu Nu'aym al-Isbahani's 
(d. 430/1038) mustakhraj of Muslim's Sahih is remarkably faithful to 
the contents of the original, generally replicating them down to the 
details of each narration. Abu Bakr Ahmad b. Ibrahim al-Isma'lll's 
(d. 371/981-2) mustakhraj of Sahih al-Bukhdri, now lost, appears to have 
been so faithful that if he could find no other transmission of a hadith 
he would narrate it through al-Bukharl and his student al-Firabri, the 
transmitter from whom al-Isma'ili received the SaMh-^ Abu Ja Tar Ahmad 
b. Hamdan al-Hiri of Naysabur (d. 311/923-4) spent years working 
on a mustakhraj meeting Muslim's requirements for authenticity to the 



' There may be one exception to this. Al-Dhahabi say.s that Abd al-Ghani b. 
'Abd al-Wahid al-MaqdisI (d. 600/1203) wrote a 48>^'bool5: entitled al-Misbdh fi 
'ujun ahadith al-sihdh in which he reproduced the hadlths of the Sahihajn with his own 
isndds. This is the only mention of this book, however; al-Dhahabi, Sjyar, 21:446-7. 
Al-Rafi'l also notes that in the sixth/twelfth century one AbQ al-Mansur Nasr b. 'Abd 
al-Jabbar made a mustakhraj of al-Khallll's Fadd'il Qazwin from his own hadlths; al-Rafi'l, 
al-Tadwin, 3:449. 

° For useful discussion of the mustakhraj genre and related topics, see MuUa Khatir, 
Makdnat al-Sahihayn, 167; Ibn al-Wazir, Tanqih al-anidr ji ma'rifat 'ulum al-dthdr, 40-2; 
Muhammad b. 'All Ibn Daqlq al-'Id, al-Iqtirdh ji baydn al-istildh, ed. Qahtan 'Abd al- 
Rahman al-Durl ([Baghdad]: Wizarat al-Awqaf wa al-Shuun al-Dlniyya, 1982), 317; 
Ibn Hajar, al-Mukat 'aid kitdb Ibn al~Saldh, 86-7; al-SakhawI, Path al-mughith, 1:57. 

' Ibn Hajar, Path al-bdri, 13:319. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 105 

extent that he voyaged to Iraq and the Hijaz for a few hadiths needed 
to complete it.'" Other mustakhrajs were far more lenient. Ya'qQb b. 
Ishaq Abu 'Awana al-Isfaraylnl's (d. 312/924—5) work departs from 
Muslim's Sahih on many occasions in both content and structure." 
Although the great Moroccan hadlth scholar of the early twentieth 
century, Muhammad b. JaTar al-Kattani (d. 1927), asserts that Ibn al- 
Jarud al-Naysaburi's (d. 307/919-20) al-Muntaqa is a mustakhraj of Ibn 
Khuzayma's Sahih, it is less than a fifth of the Sahth's, size and bears 
only the most superficial structural similarities.'^ Joint mustakhrap of the 
Sahihayn were also more lax in following the format of the template 
collections, generally just listing hadiths found in the works and noting 
how al-Bukhari or Muslim included them. 

A genre of hadlth literature similar to the mustakhraj is that of atraf, 
or an index of hadiths by the key components of their matns. A book of 
the atraf of the Sahihayn would list all their hadiths by the beginning of 
the matn or its key component, and then provide all the transmissions 
of that tradition found in the two works. '^ Unlike mustakhrajs, which 
are organized along the chapter structure of the template book, atraf 
books usually present the hadiths according to the Companion at the 
beginning of the isndd. 

From a modern standpoint it seems difficult to discern the purpose 
or utility of producing a mustakhraj. Why reproduce a copy of an exist- 
ing hadlth collection? Why not boast one's own corpus of hadiths or 
express one's own legal or doctrinal vision? Mustakhrcjs, certainly did 
not replace original hadlth collections. Many hadlth scholars from the 
long fourth century, such al-MasarjisI, produced gargantuan personal 
musnads alongside mustakhrajs of the Sahihayn. 



'» Al-Khatib, Tdrikh Baghdad, 4:337-8; cf. al-Dhahabi, Tdrikh al-isldm, 23:402-3. 

" It is interesting to note that the great Mushm analyst of the hadlth tradition, Ibn 
Hajar al-'AsqalanI (d. 852/1449) notes that although Abu 'Awana's book has been 
dubbed a mustakhraj of Sahih Muslim, it deviates from it a great deal, and that even the 
author notes that on some occasions; Ibn Hajar, al-Nukat 'aid kitdb Ibn al-Saldh, 67. 

'^ Al-Kattanl, al-Risdla al-mustatrafa, 20. Ibn al-Jarud's text contains no introduction 
explaining the nature of his work. See Abtl Muhammad 'Abdallah b. 'All Ibn al-Jarud 
al-NaysabOrl, Kitdb al-muntaqd min al-sunan al-musnada 'an Rasul Allah (s), ed. 'Abdallah 
Hashim al-Yamani al-Madani (Cairo: Matba'at al-Fajjala al-Jadlda, 1382/1963). 

" Al-Kattanl, al-Risdla al-mustatrqfa, 125; Abu Mas'ud Ibrahim al-Dimashql, "Atraf 
al-Bukharl wa Muslim," MS 1 164, Maktabat al-Asad, Damascus; Khalaf b. Muhammad 
al-WasitI, "Atraf Sahih al-Bukhari wa Muslim," MS 1162, Maktabat al-Asad, 
Damascus. 



106 CHAPTER FOUR 

The motivation for producing a mustakhraj lies on two levels. First, we 
must remember that for transmission-based scholars a hadith collection 
could not simply be opened up and cited; one needed to have heard it 
from an authorized chain of transmitters who in turn had heard it from 
its author Abu Muhammad Qasim b. Asbagh al-MalikT of Cordova 
(d. 340/951) traveled east in 274/887-8 to study in Iraq and access the 
wealth of transmitted material in the heartlands of Islam. When he 
discovered that he had "missed" his chance to hear the Sunan of Abu 
Dawtid from its author, he produced a mustakhraj of the work.'* Abu 
Nu'aym al-Isbahani states that he composed his mustakhraj of Muslim for 
the benefit of those who had "missed" hearing that book.'"' When Q_asim 
b. Asbagh realized he had missed his opportunity to be incorporated 
into the chain of transmitters of Abu Dawud's book, he reconstructed 
his own version of his Sunan. Abu Nu'aym, who died about 170 years 
after Muslim, similarly offered his own version of Sahih Muslim to his 
contemporaries with his own intact link to the Prophet. Yet how could 
a scholar "miss" his chance to hear a book when all he had to do was 
find an authorized transmitter of the work?"" As we shall see, this would 
entail relying on an unappealingly long chain of transmission back to 
the Prophet, an act that a hadith scholar was loathe to do. 



Mustakhraj: The Sahihayn as Formative Texts 

The second level on which the mustakhraj attracted hadith scholars of 
the long fourth century was the manner in which the template coUec- 



" Al-Dhahabi, Tadhkirat al-hujfai, 3:49; idem, Tdrikh al~isldm, 25:192-3. He also 
produced a short collection called al-Muntaqd, which al-Dhahabi says is tiie equal of 
Muslim's Sahih in authenticity and is based on the chapter structure of Ibn al-Jartld's 
al-Muntaqd. See al-KattanI, al-Risdla al-mustatrafa, 20. 

'^ Abu Nu'aym al-lsbahani, 3.1-Musnad al-mustakhraj 'aid Sahih al-imdm Muslim, ed. 
Muhammad Hasan Isma'll al-Shafi'l, 4 vols. (Beirut: Dar al-Kutub al-'Ilmiyya, 1417/ 
1996), 1:89-90. 

"^ We must certainly acknowledge the possibility that a scholar in the fourth/tenth or 
fifth/eleventh century may not have been able to find an authorized transmitter for a 
work, especially a more obscure hadith collection. While in Baghdad in 478/1085, for 
example, Abu Bakr Muhammad b. al-Walld al-TurtQshl (d. 520/1 126) could not find a 
transmitter from whom to hear al-Khattabl's commentary on the Sunan of Abu Dawud, 
Ma'dlim al-sunan. In the case of Qasim b. Asbagh, who associated with Abu Dawud's 
students, and works as widely studied as the Sahihayn at the time of AbQ Nu'aym, this 
seems unlikely. See al-Silafi, "Muqaddimat al-hdfiz al-kabir Abi Tdhir al-SilaJi" in Hamd 
b. Muhammad al-Khattabi, Ma'dlim al-sunan, 4:358—9. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 107 

tion served as a formative text through which scholars could engage 
the Prophet's authoritative legacy. Formative texts are those works that 
serve as textual fora for members of a community to express their own 
relationship with the source of authority in their tradition. In Judaic 
law, the elaboration of ritual law or its adaptation to new challenges 
takes place through the rabbi's interpretive interaction with the Torah, 
Mishna and Talmud. They provide the formative texts through which 
he establishes a relationship between the Lawmaker and the needs of 
his community. Formative texts not only embody the authority of the 
Lawmaker, but also serve as a vehicle for the believer to extend that 
authority into his own context. 

The potential for a hadlth collection to function as a formative text 
stems from the essential magnetism that the hadith medium exerted on 
Muslims. A direct transmission from Muhammad, the living isndd to 
his legacy, tied Muslims to the Prophetic charisma. The isndd incorpo- 
rated the transmitter into the hermeneutic chain of interpreters. The 
transmitter could then draw on the Prophet's normative precedent and 
manifest it in daily life, where his exemplum dominated the arenas of 
law and social mores. The Prophet's message had moved out from 
Islam's epicenter in space and time through generations of interpreters 
who had inherited and transformed his teachings, and the isndd was 
the tie that bound the scholar to that one true source of authority. 
In essence, the mustakhraj was a collection of these transmissions, a 
vehicle for expressing and establishing one's relationship to the source 
of hermeneutic authority. 

Scholars of the Islamic tradition thus placed great value on proximity 
to the Prophetic legacy. In the face of Abu Zur'a's barbed critiques, 
Muslim defended his use of flawed narrations in his Sahih by asserting 
that their short isndd'i made them attractive options (in addition, he 
argued, he had more reliable versions of the same Prophetic traditions 
with longer isndds). Muslim's aspiration for elevated isndds echoed his 
senior contemporary Abti Bakr b. Abi Shayba's (d. 235/849) exhortation 
that "seeking elevated isndds, is part of religion (talab al-isndd al- 'dli min 
al-din)."^' Mustakhrg's, represented a forum in which hadlth scholars could 
display the elevation or quality of their personal narrations from the 
Prophet. Abu Nu'aym 'Abdallah al-Haddad (d. 517/1123) of Isfahan 
once faced criticism from an opponent who faulted him for not having 



" Al-Khalill, al-lrshad, 6. 



108 CHAPTER FOUR 

an elevated isndd to Muslim's Sahih. Al-Haddad replied that while he did 
not have an elevated isndd for the book itself, he had heard Abu Nu'aym 
al-Isbahani's Mustakhraj of the Sahih from his father He boasted: 

If you heard [the Mustakhraj^ from my father it would be as if you had 
heard [Muslim's hadiths] from 'Abd al-Ghafir al-FarisI (a famous transmit- 
ter of Muslim's Sahih); and if I wanted I would say: as if you had heard 
them from al-Juludi (an earlier transmitter of Muslim's Sahih); and if I 
wanted to say: it would be as if you had heard them from Ibn Sufyan 
(who transmitted the Sahih from Muslim) — I would not be lying. And 
if I wanted I would say: it was as if you had heard them from Muslim 
himself [The Mustakhraj^ has some even more elevated hadiths, so that 
if you heard them from my father it would be as if you, al-Bukhari and 
Muslim had all heard them from the same teacher'" 

Here al-Haddad used Abii Nu'aym al-Isbahani's Mustakhraj of Muslim's 
collection to assert his own proximity to the Prophet. This conversation 
occurred in the sixth/twelfth century, long after the canonization of al- 
Bukhari and Muslim, and al-Haddad uses the two icons as benchmarks 
for rating his own link to the Prophet. Abii Nu'aym's AfwiteMra; features 
such elevated isndds, al-Haddad implies, that by reading it even in his 
own time one could become al-Bukhari's or Muslim's equal. When 
Qasim b. Asbagh "missed" his opportunity to hear Abii Dawiid's Sunan 
from its author, what he had missed was the chance to transmit the 
work with a respectably short isndd to the Prophet. When faced with 
hearing the work from one of Abii Dawiid's students, and thus adding 
another transmitter between himself and the Prophet, he felt it was 
more appealing to reconstitute the work with his own, shorter isndds. 

In addition to affording the opportunity to prove the elevation of 
isndds, mustakhraj?, also provided a stage for demonstrating their authen- 
ticity. For twelve out of the thirty-six known mustakhraj?, of the Sahihciyn 
we have explicit evidence that the authors attempted to meet certain 
requirements for authenticity [sihho), often imitating those of al-Bukhari 
or Muslim. This sometimes became a cause of much concern and ten- 
sion for scholars. Abii Bakr Ahmad b. Muhammad al-Barqani (d. 425/ 
1033-4), a premier student of the Sahihayn, admitted with regret to 
having used one person in his mustakhraj who was not up to the stan- 
dards of al-Bukhari and Muslim.'^ Abii al-'Abbas Muhammad b. Ishaq 



AI-Dhahabi, Tadhkirat al-hujfai, 4:43. 
Ibn al-jawzl, al-Muntazam, 14:333. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 109 

al-Sarraj (d. 313/925) generally tried to stand by Muslim's standards 
but relaxed them order to get more hadiths from 'All b. Abl Talib.^° 

Yet the mustakhraj was not simply a vehicle for demonstrating the 
quality of one's link to the Prophet. It served as a stage for interpreta- 
tion according to the specific needs and leanings of the scholar who 
produced it. The narrations that scholars chose as counterparts to al- 
Bukharl's and Muslim's hadiths often differed in significant ways from 
those of the Sahihayn, expressing the authors' own stances on the topic. 
The compilers of these mustakhrajs, could also alter the organization or 
chapter titles of their works in addition to adding their own commen- 
tary. The following examples demonstrate how the Sahihayn served as 
formative texts that enabled later scholars to interpret and apply the 
Prophetic legacy according to their own specific needs. 

a. Al-Isma'ili: Rationalist Muhaddith 

Abu Bakr al-Isma'lll (d. 371/981-2) built up his corpus of hadiths in 
Baghdad, Rayy and Khurasan before returning to his native Jurjan, 
where he became a local institution of hadlth study^' Along with a 
vast musnad, he displayed his legal acumen by composing the Tahdhib 
al-na^ar, a work on Shafi'l legal theory (usul), and writing a rebuttal 
of the Hanafi legal theorist al-Jassas (d. 370/982). Al-Isma'lll seems 
to have shared a great deal in common with what would emerge as 
Ash'arl doctrine in the decades after his death. The Mu'tazUite Buyid 
vizier al-Sahib b. 'Abbad (d. 385/995) sent him a very complimentary 
letter, an honor usually reserved for those scholars the vizier considered 
acceptably rationalist.-^ It is thus not surprising that al-Isma'lll, like Abu 
al-Hasan al-Ash'arl himself and later Ash'arites, found it necessary to 
publicly affirm his identification with the ahl al-sunna. Al-Dhahabi pro- 
vides a transmission in which al-Isma'lll upholds what he calls the ahl 
al-hadith creed, including the duty "to accept without deviation what 
God spoke in His book and what has been transmitted authentically 
{sahhat bihi al-riwdyd) from His Messenger (s)." In line with the standard 



2" Al-Dhahabl, Tadhkirat al-huffBz, 2:215. 

^' Al-Khallll, al-Irshdd, 291. Al-Khallll says al-Isma'lll wrote books on al-Bukharl 
and Muslim, 

^^ Ibn al-Salah, Tabaqdt al~fuqaha al-shdfi'iyya, ed. Yahya al-Zayn 'All Najib, 2 vols. 
(Beirut: Dar al-Basha'ir al-Islamiyya, 1413/1992), 1:417-418. For more about al-Isma'lll 
and his family, see BuUiet, Islam: The View from the Edge, 107 ff. 



110 CHAPTER FOUR 

Sunni creed, he also describes God "by those attributes by which He has 
described Himself and His Prophet described Him . . . with no question 
as to how [bi-la kqyfa)."'^-' 

Al-Isma'lll's insistence on such matters belies an aversion to anthro- 
pomorphism consistent with the more rationalist traces we have of his 
personal leanings. His mustakhraj of al-Bukhari's Sahih reveals how he 
used the work as a forum to argue his own stances on hadlths dealing 
with subjects traditionally problematic for Muslim rationalists. In a 
hadlth describing the Day of Judgment, al-Bukharl narrates from Abu 
Sa'ld al-Khudrl: 

I heard the Prophet say: Our Lord [will] reveal His shin ['an sdqihi) and 
every believing man and woman will prostrate to Him. But he who 
prostrated in the worldly life for the sake of reputation, he will go to 
prostrate, but his back will merely straighten again.'* 

Al-Isma'ili notes that in the Qjur'anic verse to which this hadith alludes, 
"[God] will reveal a shin, and they will be called to prostrate but will 
not be able to (Qur'an 68:42)," features the indefinite, "a shin {'an sag)" 
rather than the narration's definite "His shin {'an sdqihi)." Al-Isma'lll 
then provides another narration with the original Qjur'anic wording 
"jiukshafu 'an sag," which he favors because of "its agreement with the 
wording of the Qjur'an in that sentence." Ibn Hajar, one of our best 
sources for al-Isma'lll's work, explains the scholar's stance, "He does 
not think that God is possessed of members and limbs due to what 
that entails of resemblance to created beings {mushdbahat al-makhluqin)." 
Al-Isma'lll was not the only scholar of his time to feel discomfort with 
al-Bukharl's narration. In his commentary on al-Bukharl's work, Abu 
Sulayman Hamd al-Khattabi (d. 388/998) wrote that this hadith refers 
metaphorically to God revealing His power {qudra)P 

Al-Isma'ili's rationalist streak reveals itself elsewhere in his Mus- 
takhraj to the extent that he even questions the authenticity of one of 
al-Bukhari's hadlths. Describing how Abraham will meet his polytheist 
father on the Day of Judgment, the Prophet says, "Abraham [will] meet 
his father and say, 'O Lord, indeed you promised not to humiliate me 
{tukhzini) on the day they are all resurrected.' God [will] reply, 'Indeed 



Al-Dhahabi, Tadhkirat al-hujfaz, 3:106-7. 

Ibn Hajar, Fath al-bdn, #4919; Sahih al-Bukhdri: kitdb al-tafsir, sura 68, bah 2. 

Ibn Hajar, Fath al-bdn, 8:857-8; cf. al-Qanubl, al-Sayf al-hddd, 146. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 111 

I have prohibited Heaven to the disbelievers {al-kdfinn).'"'^^' Ibn Hajar 
notes that al-Isma'ili found the very basis of this hadlth problematic 
{istashkala . . . hadha al-hadith min aslihi) and criticized its authenticity (sihhd) 
after he included it in his Mustakhraj. Al-Isma'ili notes: 

This hadlth contradicts the evident meaning (^dhir) of God's words that 
"Abraham's praying for his father's forgiveness was but the fulfillment 
of a promise [Abraham] had made to him, and when it became clear 
to him that [his father] was an enemy of God he disassociated himself 
from him . . . (Qur'an 9: 1 14)."^' 

Al-Isma'ill thus concludes: 

There is some question as to the authenticity of this report from the 
standpoint that Abraham knew that God does not renege on His prom- 
ises [Idyukhlifii al-mi'ad), so how could he consider what happened to his 
father humiliation when he knew that [God would punish him on the 
Day of Judgment for his disbelief]?^" 



b. Abu JVu 'qym al-Isbahani and Shiite-Sunni Polemic 

Muslim's Sahih includes a subchapter that has generally been titled 
"Proof that loving the Ansar and 'All (r) is a part and indication of 
faith and that hating them is a sign of hypocrisy {al-dalTl 'aid anna hubb 
al-ansdr wa 'Alt (r) min al-tmdn wa 'aldmdtihi wa bughdahum min 'aldmdt al- 
nifdq)." This title reflects the subchapter's contents: five narrations about 
the importance of loving the Ansar (the Muslims native to Medina as 
opposed to those who immigrated from Mecca), four of them using the 
love ->■ believer vs. hatred -> hypocrite distinction. The subchapter ends 
with one narration in which the Prophet uses exactly the same construc- 
tion to assert the importance of loving an early Muslim who was not 
one of the Ansar, his cousin 'All b. Abi Talib. In his Mustakhraj, Abii 
Nu'aym al-Isbahani (d. 430/1038) provides hadiths that perfectly mir- 
ror the layout and content of Muslim's chapter, with five for the Ansar 
and one for 'All. The significant difference appears in the subchapter 



^^ Ibn Hajar, Fath al-bdri, #4768-9; Sahih al-Bukhdri: kitdb al-tafsir, sura 26, bdb 2. This 
hadlth is a narration of another hadlth found in Fath #3350; Sahih al-Bukhdri: kitdb ahddith 
al-anbiyd] bdb 8, which discusses the story in more detail. See also Qur'an, 26:87. 

^' " Wa md kdna istighfar Ibrahim li-abihi ilia 'an maw 'ida wa 'adahd iyydhufa-lammd tabayyana 
lahu annahu 'aduww"' li-Alldh tabarra'a minhu..." 

^^ Ibn Hajar, Fath al-bdri, 8:641—2; see also al-Jaza'irl, Tawjih al-na^ar ild usdl al-athar, 
1:332. 



112 CHAPTER FOUR 

title: "On Love for the Ansar as a Sign of Faith (ayat al-iman)." There 
is no mention of 'Ali.^^ 

This small difference might seem unimportant until one views it in the 
context of Abu Nu'aym's other writings. Most importantly, he cultivated 
an ongoing interest in debating the Imami Shiites using hadlths. Abu 
Nu'aym's Kitab al-imama wa al-radd 'aid al-rdfida (Book of the Imamate 
and a Rebuttal of those who Reject the Caliphates of Abu Bakr and 
'Umar) provides a manual for debating the Shiite claim that 'All should 
have been the first caliph. The book is organized along dialectic lines, 
with the structure "if your opponent says . . . then you say." Many of the 
debates in the work revolve around the tensions between the different 
hadlths used as proof texts by Shiites and Sunnis. Abu Nu'aym tells 
his opponent that "if you use reports [akhbdr) as proof then it follows 
that you must accept them from your opponents . . .; reports (akhbdr) are 
thus for you and against you."^" One of the main proof texts employed 
by Shiites was Muslim's above-mentioned hadlth about the believers' 
duty to love 'All and the hypocrites' disregard for him.^' Abu Nu'aym 
rebuts this proof text by alerting his opponent to the other reports in 
which the Prophet says the same thing about the Ansar ^^ The pro-'All 
hadlth thus has no probative force in issues of succession, for "if [the 
opponent] says, 'That has been narrated from so-and-so and so-and-so,' 
let it be said to him, '[Material] opposing that has [also] been related. 
So if you use reports (akhbdr) as proof, since [all] the reports contest one 
another, [the reports] fail (saqatat).'"-'^ The subtle polemic embodied in 
Abu Nu'aym's subchapter title in his Mustakhraj now becomes evident, 
since it buries the pro-'Ali hadlth in the folds of a chapter he defines 
as strictly addressing love of the Ansar For Abu Nu'aym al-lsbahani, 
minimizing the importance and visibility of this hadlth and highlight- 
ing the similar compliments paid the Ansar is a critical part of his 
anti-Shiite polemic. 



^^ Abu Nu'aym al-Isbahani, al-Musnad al-mustakhraj , 1:156—157. 

'" Abu Nu'aym al-Isbahanl, Kitdb al-imdma wa al-radd 'aid al-rdfida, ed. 'All b. Muham- 
mad al-Faqlhl (Medina: Maktabat al-'Ulum wa al-Hikam, 1415/1994), 217. 

" For a modern example of the polemical use of this hadlth, see Mohammad Sadeq 
Najml, Sayri dar Sahihayn, 77. 

'^ Abu Nu'aym al-lsbahani, Kitdb al-imdma, 244. 

'' Abu Nu'aym, Kitdb al-imdma, 230. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 113 

c. Abu 'Awana and an Independent Legal Path 

Abu 'Awana Ya'qub b. Ishaq al-Isfarayinl (d. 312/924-5) studied the 
legal scholarship of al-Shafi'l at the hands of the latter 's two most 
renowned Egyptian students, Rabf b. Sulayman al-Muradi (d. 270/883) 
and Abu Ibrahim Isma'il al-Muzani (d. 264/878). Al-Dhahabi describes 
Abu 'Awana as the first to introduce that school to the famous Khurasani 
city of Isfaraym, later home to generations of great Shafi'i scholars.^* 
Abu 'Awana's al-Sahih al-musnad al-mukharraj 'aid Sahih Muslim (The 
Authentic Musnad Collection Based on Sahih Muslim), however, reveals 
an independent legal mind unconstrained by rigid loyalty to Muslim's 
book or al-ShafiYs opinions. On the famous issue of what invalidates 
prayer if it passes in front of one, al-Shafi'T had rejected a Prophetic 
hadith stating that a black dog, a woman or a donkey invalidates prayer 
We know from a source that predates Abu 'Awana, Muhammad b. Nasr 
al-Marwazl's (d. 294/906) Ikhtildf al-fuqahd' (The Differing Opinions 
of Jurists), that al-Shafi'i based his opinion on a report from 'A'isha in 
which she objects to this notion, angrily telling the Companion who 
narrated the hadith that "you've compared us to dogs!"''' Although 
three narrations of 'A'isha's objection appear in Muslim's SahihJ'^' he 
also includes a lengthy section of hadlths that support the idea that 
these three things do indeed invalidate prayer In Muslim's work these 
conflicting reports are buried among a range of other topics, such 
as hadlths enjoining physically obstructing people who refuse to stop 
passing in front of someone engaged in prayer Other hadlths in this 
subchapter state that one can protect oneself by constructing a small 
mound or placing something the size of the back of a saddle in front 
of oneself while praying. '' The material that Muslim puts forth thus 
offers the reader no concrete conclusion, while al-Shafi'i acts definitively 
on 'A'isha's report. 

In Abu 'Awana's Mustakhraj, this issue is greatly simplified. The 
author, who disagrees with al-Shafi'i, includes a chapter called "The 
Size of the Barrier [by which] Nothing that Passes in Front of Someone 
Praying Can Harm Him [miqddr al-sutra allati la yudirru al-musalli man 



" Al-Dhahabi, Tadhkirat al-httffaz, 3:3. 

'^ Muhammad b. Nasr al-MarwazI, Ikhtildf al-fuqahd ', ed. Muhammad Tahir Hakim 
(Riyadh: Adwa' al-Salaf, 1420/2000), 161; cf. Sahih al-Bukhdn: kitab al-saldt, bdb man 
qdla Idjaqta 'u al-saldt shay '. 

'^ Sahih Muslim: kitdb al-saldt, al-i'tirdd baynyaday al-musalli. 

'' Sahih Muslim: kitdb al-saldt, qadr mdyustaru al-musalli. 



114 CHAPTER FOUR 

yamurru baynyadayhi)." He states immediately after the chapter heading 
that if one does not have this barrier, then a black dog, a woman or 
a donkey does indeed violate prayers if it passes in front of one, and 
that a line drawn in the dirt is not sufficient protection (as Ahmad b. 
Hanbal claimed).^" He then provides seven narrations backing up his 
point, most of which also appear in Muslim's Sahih. They instruct the 
reader to build these saddle-back-sized barriers in front of himself to 
prevent his prayer from being invalidated.^^ 

Here we see that Abu 'Awana has taken a large, assorted and ulti- 
mately legally inconclusive chapter of Muslim's Sahth and compressed 
it into a treatment of one problem: women, black dogs and donkeys 
invalidate prayer To this he supplies an immediate solution: placing 
something in front of you while you pray As we have mentioned earlier, 
it was the often inconclusive character of Muslim's Sahih that diverted 
legal attention from the work. Abu 'Awana's mustakhraj not only greatly 
simplifies this topic, but also transforms it into a legal text expressing 
the author's independent thought. Despite his ties to al-Shafi'i, Abu 
'Awana breaks with him on other salient issues as well, such as al-Shafi'T's 
insistence on saying "In the name of God, the most Merciful, the most 
Compassionate {bismilldh al-Rahmdn al-Rahimj" aloud in certain prayers.*" 
As Wael Hallaq has demonstrated, in this period madhhabs were not yet 
rigid sets of legal stances. They were common hermeneutic traditions 
still in the process of being elaborated by the scholars who followed 
them. Al-Shafi'l himself was thus only primus inter pares among the jurists 
who followed his tradition.*' Abu 'Awana's work demonstrates how a 
mustakhraj could function as an independent hermeneutic expression of 
the Prophet's legal authority within the nascent Shafi'l school. 



'" Abu ' Awana Ya'qub b. Ishaq al-Isfaraylnl, Musnad Abi 'Awdna Ya 'qub b. Ishdq al- 
Isfardyim, 4 vols. [vol. 3 missing] (Hyderabad: Matba'at Jam'iyyat Da'irat al-Ma'arif 
al-'Utlimaniyya, 1362-85/1942-63), 2:49. The missing sections of the Mitsnad have now 
been published as al-Qism al-mafqud min Musnad Abi Awdna, ed. Ayman Arif al-Dimashql 
(Cairo: Maktabat al-Sunna, 1995); cf. Mansur b. Yunus al-BuhutI, al-Rawd al-murbi', ed. 
Bashir Muhammad Awn (Damascus: Maktabat Dar al-Bayan, 1420/1999), 79. 

'^ Abu Awana, Musnad, 2:30-1. 

" Abu Awana, Musnad, 2:133-5. 

*' Wael Hallaq, "From Geographical to Personal Schools?: A Reevaluation," Islamic 
Law and Society 8, no. 1 (2001): 24-5. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 115 

Hal and Ilzamat: Interaction with the Standards of al-Bukhdri and Muslim 

When Abu Zur'a al-Razi read through Muslim's Sahth, he criticized the 
lines its author had drawn in compiling his collection. He found flaws 
in some of the narrations Muslim had declared authentic and criticized 
his failure to include other worthy material. Abu Zur'a's reaction to 
the Sahih foreshadowed the emergence of two closely related genres of 
hadlth literature addressing the SaMhayn during the long fourth century: 
books of 'Hal (flaws) and ilzamat (recommended additions). 

Books detailing Hal, or the obscure flaws of transmission, represented 
the third tier of hadlth criticism discussed in the previous chapter. Such 
books existed since at least the early third/ninth century. The long 
fourth century, however, saw the appearance of 'Hal works devoted 
specifically to weeding out such flaws from the Sahihayn. These works 
illustrate the multiplicity of approaches existing in the hadlth-critic 
community; a scholar critiquing the Sahihayn was effectively juxtapos- 
ing his methods and standards of hadlth criticism with those used by 
al-Bukharl and Muslim, critically applying his definition of 'authentic' 
to their works. We have two surviving criticisms of the Sahihayn from 
this period. The earliest is Muhammad b. Ahmad Ibn 'Ammar al- 
Shahid's (d. 317/929-30) 'Hal of Muslim's Sahih- The most famous and 
comprehensive work, however, is the Kitdh al-tatabbu ' of the dominant 
Baghdad hadlth scholar 'All b. 'Umar al-Daraqutni (d. 385/995). 

As the third tier of hadlth criticism, the study of 'Hal had always 
targeted two categories of flaws: independent and comparative. Critics 
first focused on flaws that independently undermined the strength of 
an isndd. A sahih hadlth should possess an uninterrupted chain of trust- 
worthy and competent transmitters that reached back to the Prophet.*^ 
Hadlth critics thus searched for weak or error-prone transmitters as 
well as breaks between links in the isndd ijnqitd'). Broken transmissions 
included reports that someone who had never met the Prophet attributed 
directly to him (termed mursal) or that were actually the statements of 
the Prophet's Companions (termed mawquf)*"^ This stage of criticism 



" For appropriate expressions of this definition, see Muslim, Sahih, 1:23; Ibn Khu- 
zayma, Sahih Ibn Khuzayma, 1:3; Muhammad Ibn Hibban al-BustI, Sahih Ibn Hibbdn, ed. 
Ahmad Muhammad Shakir (Cairo: Dar al-Ma'arif, [1952]), 1:112. 

*' For examples of these flaws in our earliest extant 'Hal work, see All b. al-Madml, 
al-Ilal, ed. Muhammad Mustafa A'zamI ([n.p.]: al-Maktab al-lslami, 1392/1972); 81, 
104, 110. 



116 CHAPTER FOUR 

was subjective, as different critics applied different standards to their 
material. Muslim's decision to accept the narration of two transmitters 
joined by the vague phrase "from/on the authority of {'an)," provided 
they were contemporaries, proved controversial for later scholars who 
upheld more rigid standards for transmission. Al-Bukhari's inclusion 
of a hadlth narrated by the extremist Kharijite 'Imran b. Hittan, who 
praised the caliph 'All's murderer in poetry, would prove similarly 
problematic for critics less forgiving of such 'heresies.' 

The second breed of flaws on which 'ilal criticism focused was com- 
parative. Scholars acknowledged two comparative signs of unreliable 
narrations: disagreement [khilaf) and a lack of corroboration (tqfarrud). 
These two concepts existed in relative space, for both rested on the 
critic gathering all the available narrations of a hadith and examining 
which were the most well-established. If a specific narration differed 
with the bulk of other transmissions or with that of a master hadith 
scholar, it was generally deemed weak. If one student transmitted a 
narration of a hadith without the corroboration of his colleagues, it 
too was declared unreliable. 

A central theme in this comparison of isnads was the layered notion 
of Addition' (ziyddd), a concept that Muslim scholars of this period 
commonly considered unified but which actually subsumed three very 
different phenomena. The first can be termed Isndd Addition, which 
occurred when one narration of a hadith added a transmitter not found 
in the other isndd?,. The second, termed Literal Main Addition, 
involved one narration of a hadith adding material to the text of the 
report. Finally, Normative Main Addition occurred when one nar- 
ration of a report that was generally considered to be the statement of 
a Companion [mawquf) was elevated and attributed to the Prophet.** 

This comparison of narrations was also a subjective process. If, out 
of a selection of ten narrations of a tradition from reliable transmit- 
ters, only one was attributed to the Prophet while the others were 
the words of a Companion, most hadith critics would consider the 
exception defective. This tradition would thus not be sahih, since it had 
been established as not extending back to the Prophet. Another critic, 
however, might trust the lone transmitter and choose his as the correct 



" For a detailed discussion of the phenomenon of Addition [ziydda], see Jonathan 
A.C. Brown, "Criticism of the Proto-Hadith Canon: al-Daraqutnl's Adjustment of the 
Sahihayn" Journal of Islamic Studies 15, no. 1 (2004): 8-11. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 117 

narration of the hadith, declaring it an authentic Prophetic statement. 
Muslim often seems to have been more lax on such matters than his 
fourth/tenth-century critics. In the introduction to his Sahih he states 
that he accepts a transmitter's uncorroborated material provided he not 
deviate blatantly from his cohorts.*"' As the works of Ibn 'Ammar and 
al-Daraqutnl demonstrate, on a number of occasions Muslim's desire to 
locate a reliable, uninterrupted narration of a hadith from the Prophet 
led him to ignore the best established versions of the report, which often 
showed that the hadith was actually of limited reliability. 

Many of the flaws that Ibn 'Ammar identifies in Muslim's Sahih thus 
revolve around demonstrating how the best established version of one 
of Muslim's hadiths is actually a broken or weak transmission. Out of 
a total of thirty-six criticized narrations from the Sahih, Ibn 'Ammar 
locates thirteen instances of inappropriate Addition (4 Isndd Addition, 
4 Literal Matn Addition, 5 Normative Matn Addition), and nine instances 
of a break in the isndd {inqitd'). Ibn 'Ammar differs with Muslim's meth- 
odology in other areas as well. He finds fault with one narration because 
an earlier hadith scholar could find no trace of it in the transmitter's 
personal notebooks.*® For another narration Ibn 'Ammar explains that 
an error occurred because the transmitter had buried his books and 
begun narrating from memory. Here we see that Ibn 'Ammar adhered 
more to al-Bukhari's school of thought, which appreciated written 
sources as an invaluable bulwark against error despite the emphasis of 
the hadlth-scholar community on oral transmission.*' 

While Ibn 'Ammar's relatively early 'Hal work tackled only Muslim's 
Sahih, fifty years later al-Daraqutni critiqued both the Scih^hciyn. His Kitdb 
al-tatabbu ' criticizes two hundred and seventeen narrations, one hundred 
from Muslim's Sahih, seventy-eight from al-Bukhari's and thirty-two 
shared by both collections.*^ Like Ibn 'Ammar, al-Daraqutnl's com- 
ments frequently involve instances of inappropriate Addition, especially 
in Muslim's work. Unlike Muslim, he only accepted Addition, either 
Isndd or Matn, when it enjoyed the support of a preponderance of 



« Muslim, Sahih, 1:6. 

*'' Ibn 'Ammar Abu al-Fadl al-Shahid, 'Hal al-ahddith Ji kitdb al~sahih li-Muslim h. al- 
Hajjaj, ed. 'All b. Hasan al-Halabi (Riyadh: Dar al-Hijra, 1412/1991), 109. 

" Al-Bukhan states that "books are more accurate {akfai) for the people of knowledge 
(ahl al- 'Urn), since a person could transmit something and then return to a book and [it 
turns out] that it is as in the book"; see his Kitdb raf al-yadaynji al-saldt, 82. 

"^ For a more exact breakdown of these narrations, see Brown, "Criticism of the 
Proto-Hadith Canon," 11. 



118 CHAPTER FOUR 

experts.*^ Al-Daraqutni also reveals a stringency absent in al-Bukhari's 
method. The Baghdad scholar chastises al-Bukhari for narrating a 
hadlth from the arch-Kharijite 'Imran b. Hittan, citing his deviant beliefs 
{sW i'tiqadihi)/'^ 

Unlike Hamd al-Khattabi and later hadith critics such as Ibn Qayyim 
al-Jawziyya (d. 751/1350) and MuUa 'All Qarl (d. 1014/1606), neither 
Ibn 'Ammar nor al-Daraqutni criticized any hadlth found in the Sahihayn 
for ideological or polemical reasons.'^' In only one instance does either 
scholar even directly address the legal implications of any hadlth. Ibn 
'Ammar rejects a narration from Muslim's Sahih stating that the Prophet 
did not perform 'umra after the battle of Hunayn because it contradicted 
another authentic hadlth asserting that he did." In fact, al-Daraqutni 
demonstrates astonishing objectivity in his critique: although he had 
compiled an entire book of hadiths devoted to affirming that God 
would grant the believers a vision of Himself on the Day of Judgment, 
al-Daraqutni explicitly rejects a unique narration in Sahth Muslim sup- 
porting exactly that belief"'' 

The second genre of hadlth literature closely related to 'Hal was 
that of ilzamdt. These works listed hadiths that the authors believed 
al-Bukharl and Muslim should have included in their two collections. 
Only four ilzamdt works, also known as mustadrah, were produced, all 
of them based on both al-Bukharl's and Muslim's Sahihs, in tandem. 
The remarkable Mustadrak of al-Hakim al-Naysaburl will receive suf- 
ficient attention in the next chapter 'Abdallah b. Ahmad Abu Dharr al- 
Harawl's (d. 430/1038) one-volume mustadrak of the Sahihayn appears not 
to have survived.^^* Ahmad b. 'All al-'Awall of Naysabur (fl. 420/1030?) 



'^ For more on al-Daraqutni's stance on AAi^ition/ ziyada, see Brown, "Criticism of 
the Proto-Hadith Canon," 31-4. 

^" 'All b. 'Umar al-Daraqutnl, Kitdb al-ilzdmdt wa al-tatabbu', ed. Muqbil b. Hadi b. 
Muqbil (Medina: al-Maktaba al-Salafiyya, [1978]), 333. 

^' See, for examples, Ibn Hajar, Fath al-bdn, 13:591; Shams al-Din Muhammad Ibn 
Qayyim al-Jawziyya, al-Mandr al-muntf ji al-sahih wa al-da 'if, ed. ' Abd al-Fattah Abu 
Ghudda (Aleppo: Maktab al-Matbu'at al-Islamiyya, 1970), 78; Nur al-Din MuUa 'All 
b. Sultan Qarl, al-Asrdr al-marju'a fi al-akhbdr al-mawdu'a, ed. Abu Hajir Muhammad 
al-Sa'ld Zaghlul (Beirut: Dar al-Kutub al-'Ilmiyya, 1405/1985), 319. 

'' Ibn 'Ammar, 93. 

" See Brown, "Criticism of the Proto-Hadith Canon," 21. 

^* Al-FarisI, Tdrikh Naysdbur al-muntakhab min al-Siydq, 607. Here the author states that 
Abu Dharr produced a mustakhraj of both Sahihi. Al-Harawl's mustakhraj of Muslim was 
criticized for narrating from transmitters unworthy of Muslim's standards; al-Dhahabi, 
Tadhkirat al-hujfai; 3:201-3, 244. The large hadith collection, amounting to thirteen 
printed volumes, of the later scholar Diya' al-Din al-MaqdisI (d. 643/1245) also con- 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 119 

made a sahih selection of hadiths from his teacher AbQ Muhammad 
'Abd al-Rahman b. Muhammad al-Balawi (d. 410/1019) that met the 
requirements of al-Bukhari and Muslim {'ala shart al-shaykhayn).-" The 
only other extant work from this genre comes from al-HarawT's teacher, 
al-Daraqutni. Scholars have closely identified his Kitab al-ilzamat with 
his above-mentioned Kitdb al-tatabbu', and they have often been trans- 
mitted as one unit. 

Ilzdmdt works applied al-Bukhari's and Muslim's own standards to 
hadiths left out of their works. Unlike 'Hal works, this entailed a further 
application of the Shaykhayn's methods and not a juxtaposition with the 
methods of later critics. As with his critique of the Sahihayn, al-Daraqutnl 
did not use his ilzdmdt as a means for advancing his own legal or doc- 
trinal positions. There is an almost total separation between the hadiths 
that al-Daraqutnl addended to the Sahihayn and those that he selected 
for his own legal reference, his Sunan. At no point, for example, does 
he claim that one of the narrations included in his Sunan should have 
been featured in the Sahih^.'''^' 

What remains unclear is how these scholars understood and articulated 
al-Bukhari's and Muslim's requirements for authenticity. Al-Daraqutnl's 
Kitdb al-ilzdmdt implies he considered himself well acquainted with the 
two scholars' methodologies, and his student Abu Mas'ud al-Dimashqi 
(d. 401/1010-1 1) confidently refers to Muslim's "usual methods (roym)."" 
The only explicit studies devoted to this subject, however, seem to be 
al-Hakim al-Naysaburl's separate monographs on al-Bukhari's and 
Muslim's requirements."'" Neither work, however, is extant. 

Both ilzdmdt and 'Hal activities seem to have been fairly informal 
among scholars of the long fourth century. Al-Khatib al-Baghdadl's 
teacher HibataUah b. al-Hasan al-Lalaka'l (d. 418/1027—8), for example, 
noted incidentally in his Shark usul i'tiqdd ahl al-sunna wa'l-jamd'a (Exposi- 
tion of the Principles of the Ahl al-Sunna wa al-Jamd'a Creed) that a cer- 
tain hadlth met Muslim's requirements and should have been included 



sisted of reports the author states "are not found in the Sahihayn," but the author makes 
no claim that they meet al-Bukharl's and Mushm's standards of authenticity; Diya' 
al-Din Muhammad b. 'Abd al- Wahid al-MaqdisI, al-Ahddith al-mukhtdra, ed. 'Abdallah 
b. 'Abdallah Duhaysh, 13 vols. (Beirut: Dar Khidr, 1421/2001), 1:69-70. 

'"'' 'Abd al-Ghafir al-FarisI, Tdrikh Naysdbur al-muntakhab min al-Siydq, 472. 

^1^ Brown, "Criticism of the Proto-Hadith Canon," 20-2 1 . 

'■' Abu Mas'ud al-Dimashqi, Kitdb al-ajwiba, ed. Ibrahim b. All Kulayb (Riyadh: 
Dar al-Warraq, 1419/1998), 298, 

■^" Al-Hakim al-Naysabtlrl, al-Madkhal ild man/at kitdb al-Iklil, 72. 



120 CHAPTER FOUR 



in his Sahih [yalzamuhu ikhrdjuhu)."^ In addition to his Kitab al-tatabbu', 
al-Daraqutni criticized at least thirteen other narrations from Muslim's 
Sahth. These were not set down in any extant books, but have survived 
in a rebuttal by al-Daraqutni's student Abu Mas'ud al-Dimashql/'" 



Study: Clmijying an Unclear Subject 

As templates for mustakhrap, al-Bukhari's and Muslim's collections served 
as formative texts for scholars to interpret and implement the Prophet's 
normative legacy in new times. Through 'Hal and ilzdmdt works, hadith 
scholars of the long fourth century critically engaged the standards of 
authenticity established by the Shaykhayn. Both the mustakhraj and the 'ilaU 
ilzdmdt genres required an exhaustive knowledge of al-Bukhari's and 
Muslim's collections. Scholars seeking to partially reproduce their isnddi, 
or understand their requirements for authenticity needed to identify 
all of al-Bukhari's and Muslim's chains of transmission. These genres 
of scholarly activity thus spurred a myriad of subsidiary studies on the 
Sahihayn. Mustakhrcy's themselves often included elucidations of obscure 
transmitters. Al-Isma'lll's work, for example, identifies a narrator in one 
isndd whom al-Bukharl refers to simply as 'al-Maqburl' as the famous 
Successor Sa'id al-Maqburi.''' 

Those who transmitted al-Bukhari's and Muslim's Scihih's, also con- 
tributed to clarifying some of the collections' indistinct features and 
deciphering textual vagaries. Ibn al-Sakan (d. 353/964) of Baghdad 
settled in Egypt after years of travel and became an important trans- 
mitter of al-Bukharl's S(Mi-'^ He received his text of the Sohih directly 
from al-Bukharl's student al-FirabrI (d. 320/932) and attempted to 
clarify, through his own research, as many of the ambiguous transmit- 
ters as possible. As a result, his recension of the Sahih became one of 



■'■' Abu al-Qasim Hibatallah b. al-Hasan al-Lalaka'i, Shark usul i'tiqad ahl al-sunna 
wa al~jamd'a, ed. Ahmad b. Sa'd b. Hamdan al-Ghamidi, 4 vols. (Riyadh: Dar Tayba, 
1415/1994), 4:878. 

™ See Abu Masud al-Dima.shqI, Kitdb al-ajwiba; 187, 195, 198, and 203, for 
examples. 

" Ibn Hajar, Path al-bdn, 13:371. 

^'^ Al-Dhahabi, Tadhkirat al-hujfd^, 3: 100; idem, Tdnkh al-isldm, 26:88-9. He transmitted 
Sahih al~Bukhdri to Ibn Asad al-juhani, Muhammad b. Ahmad b. Yahya b. Mufarrah 
and Abu Ja'far b. 'Awn. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 121 

the most definitive studies of al-Bukhari's transmitters/'' Abu Dharr al- 
Harawi was a Maliki who settled among the Bedouin near Mecca and 
visited the city every year for pilgrimage as well as to narrate hadiths. 
He brought together the three disparate transmissions of al-Bukhari's 
Sahih from Abu Ishaq al-Mustamli of Balkh, al-KushmihanI of Merv 
and Abu Muhammad al-Hamawayh of Sarakhs. These were the three 
most prominent students of al-Firabrl, the primary transmitter of the 
Sahth from its author''* More importantly, al-Harawi noted the variations 
among the three transmissions and attempted to accurately reconstitute 
the original text.'''^ 

Differences between various narrations of al-Bukhari's Sahih occasion- 
ally proved noticeable. Besides the transmission of al-Firabri, those of 
Ibrahim b. Ma'qil al-Nasafi (d. 295/907-8) and Hammad b. Shakir 
(d. 290/902-3) also survived for several centuries. Hammad b. Shakir's 
recension, however, contained two hundred fewer narrations than that 
of al-Firabri, while Ibrahim's was three hundred fewer''*' 

Transmitters could also play more substantial editorial roles. Abu 
al-Walld al-Bajl reports that when Abu Ishaq al-Mustamll examined 
al-Firabrl's copy of the Sahih he noticed that some sections were still 
in draft form, with a number of chapter headings lacking hadiths, or 
hadiths with no chapter headings. Al-Mustamll states that he and his 
fellow students attempted to arrange unsorted material in its proper 
place {fa-adajhd ba'd dhdlik ild ha'd)^'^ 

Most importantly, the long fourth century saw the emergence of 
studies specifically devoted to identifying and describing al-Bukharl's 



''•'• Later scholars testify to the importance of Ibn al-Sakan's work; see Abu 'All al- 
Husayn al-jayyani al-Ghassani, al-Ta'rif bi-shuyukh haddatha 'anhum Muhammad b. Ismd'il 
al-Bukhdnfi kitdbihi iva ahmala ansdbahu wa dhikr mdju'rafun bihi min qabd'ilihim iva buUdnihim, 
ed. Muhammad al-Sa'ld Zaghlul (Beirut: Dar al-Kutub al-'Ilmiyya, 1418/1998), 11. 

''' Al-Dhahabi, Tadhkirat al-hujjdi, 3:201; cf. Ibn al-jawzl, al-Muntaiam, 15:287. 

''^ Ibn Daqlq al-'Id, al-Iqiirdhji baydn al-istildh, 299. 

r>i> Al-'IraqI, al-Taqyidwa al-iddh, 26—7. Ibn Hajar explains that Ibrahim and Hammad 
heard incomplete versions of the Sahih from al-Bukharl and that al-Firabrl's recension 
represents the final product (ad al-tamif); Ibn Hajar, al-Mukat 'aid kitdb Ibn al-Saldh, 69. 
For more information on the details of the transmission of al-Bukharl's and Muslim's 
Sahihi, see Chapter 7 n. 100. For a discussion of the attribution and textual authenticity 
of the two works, see Appendix II. 

'*' Abu al-Walld Sulayman b. Khalaf al-Bajl, Abu al-Walid Sulaymdn b. Klialaf al-Bdji 
wa kitdbuhu al-Ta'dil wa al-tajnh ti-man kharraja lahu al-BukhdriJi al-Jdmi' al-sahih, ed. Abu 
Lubaba Husayn, 3 vols. (Riyadh: Dar al-Liwa', 1406/1986), 1:310-1; Muhammad b. 
Yusuf al-KirmanI (d. 786/1384), al-Kawdkib al-dardri fi sharh Sahih al-Bukhdn, 25 vols. 
(Cairo: al-Matba'a al-Bahiyya al-Misriyya, 1358/1939), 1:5. 



122 G HAPTER FOUR 

and Muslim's transmitters. The earliest examples of this genre are 
limited to identifying al-Bukharl's immediate sources. Ibn 'Adi's Asdmi 
man rawd 'anhum Muhammad b. Ismd'il al-Bukhdn and Muhammad b. Ishaq 
Ibn Manda (d. 395/1004—5) of Isfahan's Asdmi mashdyikh al-imdm al- 
Bukhdn represent the first two generations of these transmitter studies. 
Abu Nasr Ahmad al-Kalabadhi (d. 398/1008) of Bukhara produced 
the most comprehensive listing of all al-Bukharl's transmitters."" Yet it 
was not until the early fifth/eleventh century that a work was compiled 
on the transmitters used in Muslim's Sahih'- this was the book of Abu 
Bakr Ahmad b. 'All Ibn Manjawayh of Naysabur (d. 428/1036-7). 
Al-Daraqutnl was the first to write a biographical dictionary covering 
both the Sahihayn. His student al-Hakim al-Naysaburi and the Baghdad 
scholar al-Lalaka'i each repeated this task several years later*'" 

Al-Daraqutni's oeuvre constituted the first and most impressive holistic 
study of the Sahihoyn as two complementary texts. He authored no less 
than eleven books detailing various aspects of al-Bukharl's and Muslim's 
work. In addition to his biographical dictionary of their transmitters, 
he compiled separate lists of the transmitters after the generation of 
the Companions who comprised al-Bukharl's and Muslim's isnddi.J'^ 



''" Although originally titled al-Hiddya wa al-irshdd fi ma'rifat ahl al-thiqa wa al-saddd 
alladhina akhraja lahum al-Bukhdri ft Sahihihi, this work is often referred to as Rvjdl Sahih 
al-Bukhdn. 

^^ Al-Hakim 's small work is entitled Tasmiyat man akhrajahum al-Bukhdn wa Muslim 
wa md infarada bihi kull minhumd, ed. Kamal Ytlsuf al-Htlt (Beirut: Mu'assasat al-Kutub 
al-Thaqafiyya and Dar al-Jinan, 1407/1987). This genre continued beyond the scope 
of our long fourth century. AbQ 'All al-Jayyanl al-Ghassanl (d. 498/1 105) made elForts 
to complete the task of identifying al-Bukharl's obscure transmitters (see above note 
63). The Maliki jurist Abu al-Walld Sulayman b. Khalaf al-Bajl wrote a book collect- 
ing critical opinions on al-Bukharl's men entitied Kitdb al-ta'dil wa al-tajrih li-man rawd 
'anhu al-Bukhdn Ji al-Sahth (see al-KattanI, al-Risdla al-mustatrafa, 154; n. 67 above). Abtl 
al-Fadl Muhammad b. Tahir al-MaqdisI (d. 507/1113) combined Ibn Manjawayh 
and al-Kalabadhl's two works in Kitdb al-jam' bayn kitdbay Abi Nasr al-Kaldbddhi wa Ahi 
Bakr al-Isbahdni, 2 vols. (Hyderabad: Matba'at Majlis Da'irat al-Ma'arif al-Nizamiyya, 
1323/[1905]). 'Abdallah b. Ahmad al-Shantarmi of Cordova (d. 522/1128) wrote 
a book correcting some of al-Kalabadhl's oversights called Mtdb baydn 'ammdji kitdb 
Abi Nasr al-Kaldbddhi min al-nuqsdn as well as a work on Muslim's men entitled Kitdb 
al-minhdj. Ahmad b. Ahmad al-Hakkarl (d. 763/1362) also wrote a book on the men 
of al-Bukharl and Muslim. Finally, one of the most useful studies on this topic is Abu 
Bakr Muhammad b. Isma'll Ibn Khalfun's (d. 636/1238-9) work on al-Bukharl's and 
Muslim's teachers, al-Mu'lim bi-shuyukh al-Bukhdri wa Muslim, ed. Abu 'Abd al-Rahman 
'Adil b. Sa'd (Beirut: Dar al-Kutub al-'Ilmiyya, 1421/2000); al-Dhahabi, Tadhkirat al- 
hujjdi, '\:'A1; Sezgin, Geschichte des arabischen Schnfttums, 1:131. 

'" These two works, Dhikr asmd' al-tdbi'in wa man ba'dahum mimman sahhat riwdyatuhu 
min al-thiqdt 'ind Muhammad b. Ismd'il al-Bukhdri xaA Dhikr asmd' al-tdbi'in wa man ba'dahum 
mimman sahhat riwdydtuhu 'ind Muslim, have been published together as Dhikr asmd' al- 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 123 

He emphasized the complementary relationship of the two works in 
his listing of the Companions featured in both Sahihs as well as those 
that each book used exclusively. He also made a study of the differ- 
ent transmissions of the Sahihayn after their authors' deaths." The 
functional nature of these studies reveals itself in the book that al- 
Daraqutni tailored to his interest in expanding the number of verified 
authentic hadiths through ilzamdt work. He composed a book solely on 
the Companions through whom reliable hadiths were transmitted but 
who were not included in the Sahihayn [Dhikr al-sahaba alladhina sahhcit 
al-riwdya 'anhum wa laysufi al-SahThayn)J'^ 

An examination of the studies devoted to al-Bukharl's and Muslim's 
transmitters reveals a gradually increasing mastery of the two Sahth^ as 
the long fourth century progressed. Moreover, we are alerted to another 
central feature of the network of Sahihayn scholars in this period: the 
serious regional boundaries that still constricted the movement of texts 
and information. In Jurjan, Ibn 'Adi was unable to identify one of al- 
Bukhari's teachers mentioned in the S^hih, Sa'ld b. Marwan, listing him 
as unknown [Idyu'raf).^'' Even Ibn Manda, who died some thirty years 
after Ibn 'Adi, fails to mention this Sa'ld b. Marwan in his book on 
al-Bukharl's sources. It is not until Abu Nasr al-Kalabadhi, who died 
a mere three years after Ibn Manda but lived mainly in Bukhara, that 
we find a listing for Sa'id b. Marwan b. 'All Abu 'Uthman al-Baghdadi 
(d. 252/866), who lived and died in Naysabur.'* 

Why was neither Ibn 'Adi nor Ibn Manda able to identify this trans- 
mitter? Sa'ld b. Marwan had narrated hadiths to two major scholars 
in his adopted home city of Naysabur, Ibn Khuzayma and his dis- 
ciple Ibn al-Jarud. Ibn 'Adi, however, never traveled to the Khurasan 
region, and neither he nor his close friend al-Isma'lll had any contact 
with Ibn Khuzayma or his student. It is therefore not surprising that 



tdbi'in, ed. Burhan al-DanawI and Kamal Yusuf al-Hut, 2 vols. (Beirut: Mu'assasat 
al-Kutub al-Thaqafiyya, 1985). 

" For the unpublished works, Asmd' al-sahdba allati ittajaqafihd al-Bukhdn wa Muslim 
wa md infarada bihi hull minhumd, Kitdb Ji dhikr riwdydt al-Sahihayn and al-Daraqutnl's 
dictionary of al-Bukharl's and Muslim's transmitters, see Fuat Sezgin, Geschichte des 
arabischen Schrifitums, 1:207-9. 

'^ This work remains unpublished, al-Daraqutnl, "Dhikr asma' al-sahaba alladhina 
sahhat al-riwaya 'anhum wa laysu ft al-Sahihajn" MS 7 1 59, Maktabat al-Asad, Damas- 
cus: fols. 197b-198a. 

" Ibn 'Adi, Asdmi, 1 10. 

" Al-Kalabadhi, Rijdl Sahih al-Bukhdn, 2:872. Al-Hakim benefited from al-Kalabadhi; 
see his Tasmijat man akhrajahum al-Bukhdn wa Muslim, 123. 



124 CHAPTER FOUR 

Ibn 'Adi ignores Ibn Khuzayma completely in the list of great hadith 
scholars in his al-Kamil.'^ Conversely, Ibn Manda visited both Bukhara 
and Naysabur But we know from al-Hakim, however, that he had 
completed his book on al-Bukhari's teachers before staying in Naysabur 
and possibly before arriving in Bukhara.'^ It seems that, like Ibn 'Adi, 
Ibn Manda never had access to information about Sa'id b. Marwan 
of Naysabur. 



Regional and Temporal Distribution of the Sahihayn Network 

Ibn 'Adi's and Ibn Manda's failure to identify Sa'id b. Marwan illus- 
trates one of the salient characteristics of the study of the Sahihayn 
in the long fourth century Although hadith scholars traversed the 
Islamic world from Andalusia to Central Asia, resilient regional cults 
still developed according to material constraints like the availability of 
certain texts as well as the functionalist and ideological preferences of 
local scholarly communities. The Sahihayn Network of the long fourth 
century revolved around three of these regional schools: Naysabur, 
Jurjan and Baghdad. 

a. Naysabur and the Hometown Cult of Muslim 

Naysabur was the birthplace of the mustakhraj phenomenon, and it was 
in this city and its environs that the genre flourished most intensively. 
From the time of Muslim's death until the close of the long fourth 
century, scholars devoted mustakhrajs to the Sc^h^h of the city's native 
son. In addition, Naysabur scholars crafted mustakhrajs of Abu Dawud's 
Sunan, al-Tirmidhi's j^amr and Ibn Khuzayma's Sahih. It was only in the 
mid 300s/900s, however, that the city's scholars developed an interest 
in al-Bukharl's collection. 



" For a biography of Sa'id b. Marwan al-Baghdadl, see Ibn Khalfun, al-Mu'lim 
bi-shuyukh al-Bukhdn wa Muslim, 514-5. Ibn Khalfun lists another Sa'ld b. Marwan as 
well, namely Sa'ld b. Marwan b. Sa'ld Abu 'Uthman al-Azdl from the Jazlra. Ibn Wara 
and Abo Hatim al-RazI narrated from him, and al-Bukharl notes him in his Tdrikh 
al-kabir. It is very unlikely that this was the Sa'ld b. Marwan to which Ibn 'Adi was 
referring, since he was very familiar with Ibn Wara and Abfl Hatim, both of whom 
appear in his al-Kdmil. 

™ Al-Dhahabi, Tdrikh al-isldm, 27:320-4. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 125 

Naysabur was the linchpin of the eastern Islamic lands during the 
classical period. Astride the road that ran from Baghdad to Central Asia 
and beyond, it was an inevitable commercial way station and a bustling 
center of scholarly activity The city's intellectual landscape was sharply 
divided between the Hanafi school, with its strong ties to Mu'tazilite 
doctrine, and the transmission-based ahl al-sunna, who generally identi- 
fied with the teachings of al-Shafi'i." In the decades after the city laid 
Muslim to rest at the head of one of its major squares, Naysabur's 
transmission-based legal culture was dominated by Muhammad b. 
Ishaq Ibn Khuzayma. Declared "mam of the imams," Ibn Khuzayma 
was described by al-Hakim al-NaysaburT as "the foremost [scholar] by 
agreement of all of his age," an authority on the teachings of al-Shafi'i 
and a source of religious rulings [fatwas).'"^ He studied with al-ShafiYs 
most illustrious students, al-Rabf and al-Muzani, and was relied upon 
greatly by Ibn Surayj (d. 305/917-18), the Baghdad scholar around 
whom the Shafi'i legal school coalesced.'^ Ibn Khuzayma rigidly upheld 
the iiber-Sunni stance on the nature of the Qur'an, stating that anyone 
who believed it to be created was an unbeliever or heretic."" A poem 
by Muhammad b. Ibrahim b. Yahya of Naysabur testifies to the high 
positions of Muslim and Ibn Khuzayma in the city's pantheon of 
scholars: 

So set aside ail thought of Jurjan, for indeed our scholars 

In the land of Naysabur are more illustrious by far; so why the sadness? 

No one can be compared to Yahya b. Yahya."' 

If tested his glory would suffice you. 

And his student Ishaq [b. Rahawayh], how great he is (li-lldh darruhu)\ 

Indeed, along with al-RibatI, their virtue is not hidden. 

Abu al-Azhar al-Mifdal, then Ibn Hashim, 

And Muslim, they are the lords of hadith so do not deny it. 

And who is their equal in prodigious memory and station? 

And from us, too, Ibn Ishaq the Khuzaymi, our shaykh, 
Our source of pride, shaykh of all shaykh?, in his time. 



" See Richard Bulliet, The Patricians of Mskapur (Cambridge: Harvard University 
Press, 1972), 36-40. 

'" Al-Hakim al-Naysaburl, Tdrikh Nishdbur, 120; Bulliet, Patricians, 62. 

'' Al-Hakim al-Naysaburl, Ma'rifat 'ulum al-hadith, 104; al-Khallli, al-Irshdd, 312-3; 
Ibn al-Jawzl, al-Muntazam, 12:233-6. 

»» Al-Dhahabi, Tadhkirat al-huffd^, 2:211 ff. 

"' Abu Zakariyya Yahya b. Yahya al-Tamlml al-NaysabOrl (d. ca. 220/835); see Ibn 
Hajar, Tahdhib al-tahdhib, 11:259. 



126 CHAPTER FOUR 

Indeed he was for Islam a pillar and pivot. 

May God water well a grave with such a shaykh buried within. 



82 



One of Ibn Khuzayma's colleagues also exercised a tremendous amount 
of influence in Naysabur Abu al- 'Abbas Muhammad b. Ishaq b. Ibrahim 
al-Sarraj (d. 313/925) was one of the city's leading scholars. A student 
of Ishaq b. Rahawayh and a teacher of Ibn Khuzayma, both al-Bukhari 
and Muslim studied hadith with al-Sarraj. He was an inveterate critic 
of the Hanafi school and active prosecutor of those who upheld the 
created wording of the Qur' an.^^ Al-Sarraj also produced one of the 
earliest mustakhrqjs of Muslim's Sahih. 

Scholars in Naysabur began using Muslim's collection as a template 
for mustakhrajs almost immediately after his death. Abu Zur'a al-RazI 
mentioned that Abu Bakr al-Fadl b. al-'Abbas al-Sa'igh of Rayy (d. 270/ 
883) had done so during Muslim's lifetime.''* Abu Bakr Muhammad Ibn 
Raja' (d. 286/899) studied with many of the same teachers as Muslim 
did. He nonetheless used his coevals's remarkable collection as the 
basis for a mustakhraj called al-Sahih al-mukharraj 'aid kitdb MuslimP Abu 
al-Fadl Ahmad b. Salama al-Bazzar (d. 286/899), Muslim's companion 
to whom the Sahih is dedicated, also wrote a mustakhraj.^'' As the Sahihayn 
Network Chart demonstrates, scholars studying or living in Naysabur 
and its immediate environs continued to produce waves of mustakhraj?, 
on Muslim's collection. No less than ten had been compiled before Abu 
'Abdallah Muhammad b. Ya'qub Ibn al-Akhram (d. 344/955) finally 
produced one of the Sahihayn together"' Almost two decades later al- 
MasarjisI (d. 365/976) devoted another mustakhraj to the Sahihayn.^^ Yet 
in the century after Ibn al-Akhram 's death, Naysabur produced eight 
more mustakhrcjs of Muslim and four of the combined Sahihayn, but 
only one devoted solely to al-Bukharl's S^hih- 



"^ Al-Hakim al-Naysabun, Tankh Mshabur, 177-8. 

«' Al-khallll, al-Irshdd, 310-11; al-Khatib, Tdnkh Baghdad, 1:264-7; al-Dhahabi, 
Tadhkirat al-hujfdi, 2:215; idem, Tdnkh al-isldm, 23:462-4. 

'* Abu Zur'a al-RazI, Kitdb al-du'a/d' wa ajwibatuhu 'aid as'ilat al-Bardha'i, 2:674. 

'^ Ibn al-Salah, Siydnat Sahih Muslim, 89; al-Dhahabl, Tadhkirat al-huffa^, 2:186; idem, 
Tdnkh al-isldm, 2 1 :288. 

"'' Al-Dhahabi states that people like Abu Nu'aym al-Isbahanl also called the work 
Sahih Ahmad b. Salama; al-Khatib, Tdnkh Baghddd, 4:408; cf. al-Dhahabi, Tdrikh al-isldm, 
21:59-60; idem, Tadhkirat al-hujjdi, 2:156. 

"' Al-Dhahabi, Tadhkirat al-hujfd^, 3:55; idem, Tdrikh al-isldm, 25:312—3; cf. al-Khallll, 
al-Irshdd, 315. 

'" Al-Dhahabi, Tadhkirat al-hujfaz, 3:110-11; idem, Tdrikh al-isldm, 26:337-8. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 127 

Although al-Bukharl was not a native of Naysabur like Muslim, he 
resided in the city for approximately five years, during which time he 
narrated his Sahih to circles of hadith students."^ Why then did scholarly 
activity in the city seem so oblivious of al-Bukharl's work until Ibn al- 
Akhram's and al-Masarjisl's writings? The answer lies in the qualitative 
preference Muslim enjoyed in his hometown as well in the accusations 
of heresy that had tainted al-Bukharl's name. When Abu al-'Abbas 
b. Sa'ld Ibn 'Uqda (d. 332/944), who taught many Naysaburls, was 
asked who was more knowledgeable, al-Bukharl or Muslim, he even- 
tually replied that al-Bukhari occasionally made mistakes with reports 
transmitted from Syrians because he had only received these in writ- 
ten form. He thus sometimes thought that a person mentioned once 
by his name and once by his patronymic was two people. Conversely, 
Ibn 'Uqda notes, Muslim rarely made errors concerning transmission 
{'Hal) because he avoided al-Bukhari's practice of including additional 
hadiths with incomplete isnads^^ Abu 'All al-Naysaburl (d. 349/960), 
who had traveled widely in Egypt, Jurjan and Merv, concluded that 
"there is not beneath the heavens {taht adim al-samd') [a book] more 
authentic than the book of Muslim.""' Ibrahim b. Muhammad Abu 
Ishaq al-Muzakki (d. 362/973), a student of Ibn Khuzayma and Ibn 
Abi Hatim al-Razi, proved to be a major link between Naysabur and 
scholarly circles in Baghdad and Isfahan. He instructed al-Daraqutni, 
al-Barqanl, al-Hakim al-Naysaburl as well as Abu Nu'aym al-Isbahani. 
Although al-Muzakkl transmitted a number of Muslim's works (pre- 
sumably his Sahih was among them) on his many visits to Baghdad, of 
al-Bukharl's works he transmitted only the Tdnkh al-kabir.^"^ 

This delayed attention to al-Bukhari's Sahth also stemmed from the 
scandal of the lafz of the Qur'an. Two of the most influential trans- 
mission-based scholars in the city, Ibn Khuzayma and al-Sarraj, both 



'" We know from al-Kalabadhi that al-Bukharl had been narrating his work to stu- 
dents since at least 248 AH. He arrived in NaysabUr in about 250 AH; al-Kalabadhi, 
Rijal Sahih al-Bukhdri, 1:24. 

™ Al-Hakim NaysabUrl, Tdnkh Nishdhur, 101; Ibn Asakir, Tdrikh madinat Dimashq, 
58:90. 

" Ibn Manda heard this direcdy from Abu 'All; see Ibn Manda, Shurut, 7 1 ; al-Khatib, 
Tdnkh Baghddd, 8:70-2; cf al-Dhahabi, Tadhkirat al-hujfdz, 3:80. 

'^ Al-Muzakkl must have visited Baghdad more than once, since at the time of his 
recorded visit in 316/928-9 both al-Daraqutnl and al-BarqanI would have been too 
young to have heard from him; al-Daraqutnl never voyaged east from Iraq. See al- 
Khatlb, Tdnkh Baghddd, 6:165-7; al-Dhahabi, Tdnkh al-isldm, 26:289-90. 



128 CHAPTER FOUR 

aggressively attacked anyone who upheld a belief in the created wording 
of the holy book. Even Ibn al-Akhram, who composed the first joint 
al-Bukhari/ Muslim mustakhraj, did so only after responding to al-Sarraj's 
request to complete one based solely on Muslim's SahihP Abu al-Walld 
Hassan b. Muhammad al-UmawI (d. 344/955) expressed a desire to 
craft a mustakhraj of al-Bukhari's work, but his father instructed him to 
follow Muslim due to al-Bukharl's lafz scandal.''* It is thus no surprise 
that, with the exception of Ibn al-Akhram and al-Masarjisi, all the 
conjoined Sahihayn mustakhrajs in Naysabur and the only one devoted 
solely to al-Bukhari appeared only after the generation of scholars who 
had studied with Ibn Khuzayma and al-Sarraj had died (see Sahihayn 
Network Chart). Only at that point could scholars like Abu Ahmad 
al-Hakim (d. 378/988), a judge who worked in Naysabur's environs 
and whom al-Hakim al-Naysaburl calls one of most knowledgeable 
concerning the requirements of authenticity {shurut al-sahih), state, "May 
God bless imam Muhammad b. Isma'il [al-Bukhari] , for it was he who 
set forth the foundations {al-usul) [of hadlth] and elucidated them to the 
people. All those who have come after him, like Muslim b. al-Hajjaj, 
have taken from his book (the Sahth)."^'' 

h. Jurjan: A Cult of al-Bukhan Among Friends 

On a map, the small province of Jurjan on the southeast coast of the 
Caspian Sea does not seem far from Naysabur and its satellite cities of 
Tus, Juvayn and Isfaraym. The intimidating Elborz Mountains, however, 
separate Jurjan's littoral marshes and thickly forested mountainsides 
from these KhurasanI centers as well as from the great city of Rayy. Yet 
during the mid-fourth/tenth century, Jurjan constituted an important 
center of hadlth study in its own right. More specifically, it was home to 
three friends who formed a bastion of scholarly interest in al-Bukharl's 
Sahth- The region produced no mustakhrap of any other hadlth work. 
Two of these scholars emerged as extremely influential figures in the 
historical development of hadlth literature. We have already relied on 
'Abdallah Abu Ahmad Ibn 'Adi (d. 365/975-6) as the earliest significant 



'' Al-Dhahabl, Tadhkirat al-hujfaz, 3:55; idem, Tdrikh al-isldm, 25:312-3. 

'* Ibn al-Salah, Siydnat Sahih Muslim, 90; al-Dhahabl, Tadhkirat al-huffai, 3:75; idem 
Tdrikh al-isldm, 25:417-8. 

'"* Al-Hakimal-Naysabun, Tdrikh Nishdbur, 187; al-Dhahabl, Tadhkirat al-hujfdz, 3:123—4. 
For Abu Ahmad's quote see al-Khallll, al-Irshdd, 380. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 129 

source on al-Bukhari's life and work. He gained renown, however, for 
his voluminous dictionary of problematic hadlth transmitters, al-Kamil 
fi du'afi' al-rijal, that became the foundation for many later works in 
that genre. The Kdmil enjoyed immediate popularity and quickly spread 
among scholarly circles in major cities like Baghdad. Ibn 'Adi's younger 
contemporary in Baghdad, al-Daraqutni, said that the work sufficed for 
all needs in that genre.^'' Ibn 'Adi traveled widely in Iraq, Syria, the 
Hijaz and Egypt and was deeply versed in the school of al-Shafi'i. He 
wrote a juridical manual called al-Intisdr based on the chapter structure 
of al-Muzani's Mukhtasar, the most famous abridgment of the Shafi'i 
tradition's formative text, al-Shafi'l's Umm (The Motherbook).^' Ibn 
'Adi not only served as an important transmitter of al-Bukharl's Sahih 
from al-FirabrI in Jurjan,''" he also wrote the aforementioned first work 
on al-Bukhari's sources. 

When Ibn 'Adi died, his close friend and colleague al-Isma'ili (d. 371/ 
981-2) led his funeral prayer ^^ As we have noted in the preceding dis- 
cussion of al-Isma'ili's Mustakhraj, this scholar adhered to al-Shafi'i's 
transmission-based legal tradition and also exhibited marked rationalist 
tendencies. Al-Isma'ill was so well-respected that several hadith schol- 
ars, including al-Daraqutni, felt that he should have compiled his own 
sahih instead of following in al-Bukhari's footsteps. It was reported that 
when news of his death reached Baghdad, over three hundred hadith 
scholars, merchants and jurists from both the Shafi'i and Hanbali 
schools gathered in the main mosque to mourn him for several days.'"" 
Although al-Isma'llT produced no independent study of al-Bukhari's 
work, his Mustakhraj remained an indispensable reference for students 
and scholars of the Sahih, even late ones such as Ibn Hajar 

Abu Ahmad Muhammad b. Ahmad al-Ghitrifi (d. 377/987-8) 
was the least accomplished of the Jurjan scholars. He was a very close 



"* Ibn al-Jawzi, al-Munta^am, 14:245. 

"" Al-Khallll, al-Irshdd, 291-2; cf. al-Dhahabi, Tadhkirat al-huffdi, 3:102-3. The 
various recensions of the Umm are most likely collections of all the works narrated by 
Rabf b. Sulayman from al-Shafi'l; Abu Zahra, al-Shdfi'i (Cairo: Dar al-Fikr al-'Arabi, 
1416/1996), 148-50. 

'" Ibn 'Adi transmitted Sahih al-Bukhan to people like ' Amr b. Ahmad b. Muhammad. 
al-Astarabadhi; Abu al-Qasim Hamza b. Yusuf al-Sahml (d. 427/1035-6), Tdnkh 
Jurjan, ed. Muhammad 'Abd al-Mu'ld Khan et al. (Hyderabad: Da'irat al-Ma'arif 
al-'Uthmaniyya, 1387/1967), 106. 

'' Al-Dhahabi, Tdnkh at-isldm, 26:241. 

""" Al-Sahml, Tdnkh Jurjdn, 87; cf. al-Subkl, Tabaqdt al-shdfi'iyya, 3:8; Ibn al-jawzl, 
al-Munta^am, 14:281-2. 



130 CHAPTER FOUR 

associate of al-Isma'ill as well as his son's tutor."" Like his friend, al- 
Ghitrlfi composed a mustakhraj of al-Bukharl's Sahih. Although his father 
was from Naysabur, he lived almost his entire life in Jurjan. He visited 
Rayy and Baghdad, and was the only Jurjan scholar to have heard 
from Ibn Khuzayma in Naysabur.'"^ 

Why did this cluster of Jurjan scholars prove such redoubt partisans 
of al-Bukharl's Sahih to the exclusion of Muslim's and the other major 
fruits of the sahih movement? This phenomenon may have resulted in 
part from a limited exposure to Muslim's work. As the Sahihayn Network 
Chart demonstrates, there were almost no personal links between Jurjan 
and Naysabur, where the cult of Muslim's Sahih matured. Ibn 'Adi thus 
excludes both Muslim and Ibn Khuzayma from his list of noteworthy 
hadith scholars and does not seem to have had access to valuable 
information about al-Bukharl's Naysabur sources. As with Muslim's col- 
lection in Naysabur, however, the Jurjan scholars also considered SaMh 
al-Bukhdn to be a more accurate representation of the Prophet's legacy. 
Al-Isma'lll argues in the introduction to his Mustakhraj (his Madkhal) that 
al-Bukhari's book is superior to Muslim's because the latter "set out to 
do what [al-Bukhari] sought to do, and took from him or from his books, 
except that he did not restrict himself [in what he included] as much 
Abu 'Abdallah [al-Bukharl] did, and he narrated from a large number 
from whom Abu 'Abdallah would not deign to narrate {lamyata'arrad. . . 
li'l-riwaya 'anhum)." He adds that al-Bukharl's Sahih also bested Abu 
Dawtid's Sunan because the former had higher standards for selecting 
hadiths as well as better explanations of their legal implications.'"^ Abu 
al-Q_asim Hamza b. Yusuf al-Sahmi (d. 427/1035-6), author of the 
local history of Jurjan {Tdnkh Jurjan), relies on al-Bukhari ten times in 
his history for information about hadith transmitters.'"* Although al- 
Sahml interacted with several scholars who cultivated equal interests in 
al-Bukharl and Muslim, including al-Daraqutnl, Abu Bakr al-BayhaqI 
and al-Khatib al-Baghdadi, he never mentions Muslim in his work. He 
does, however, note two people as hearing Scihih al-Bukhari. 



'"' Al-Dhahabi, Tadhkirat al~hujjmz, 3:120. 

'"^ Al-Dhahabi, Tdnkh al-isldm, 26:614-5. 

'"' Ibn Hajar, Hady al-sdn, 11; al-Jaza'irl, Tawjih al-naiar ild usiil al-athar, 1:305. For 
a short summary of this, see Muhyl al-Din al-Nawawi, Tahdhib al-asmd' wa al-lughdt, 
3 vols. (Beirut: Dar al-Kutub al-'Ilmiyya, [1977]), 1:74. 

'"' Al-Sahml, Tdnkh Jurjdn, 488. Al-Sahml is connected to al-Bukharl by the isndd 
of Abu Bakr Ahmad b. 'Abdan <- Muhammad b. Sahl <- al-Bukharl. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 131 

c. Baghdad: Inheriting the Study of the Sahihayn Among the Baghdad Knot 

Baghdad inherited the study of al-Bukharl's and Muslim's collections 
from both Jurjan and Naysabur. From the mid-fourth/tenth century to 
the mid-fifth/eleventh, the capital of the Abbasid caliphate hosted a 
knot of scholars who pioneered the study of the two works as comple- 
mentary units. The genesis of this close association of experts lay in 
the seminal work of 'All b. 'Umar al-Daraqutnl, whose eleven treatises 
on the Sahihayn have proven some of the most influential books on the 
subject. In particular, his joint critical study, Kitdb al-ilzdmdt wa al-tatahhu\ 
has attracted scholarly attention up to the present day. Al-Daraqutni 
brought these two previous centers of study together through his per- 
sonal scholarly relationships with Abu Sa'id al-Hirl, Ibrahim al-Muzakki, 
al-Hakim al-Naysaburl and Ibn Dhuhl of Naysabur, and Ibn 'Adi of 
Jurjan. He also interacted with scholars from farther afield in Central 
Asia, such as al-Kalabadhi. He received at least two transmissions of 
Muslim's SoMi, one from Ibn Mahan in Egypt and one from Ibrahim 
al-Muzakki. He heard S^hih al-Bukhdn from Abu Sa'Td Ahmad Ibn 
Rumayh (d. 357/967-8) and most probably from others as well.'""' 

Al-DaraqutnT mentored another of the most influential scholars on 
the Sahihoyn in the long fourth century. Originally from Khwarazm in 
Transoxiana, Abu Bakr Ahmad b. Muhammad al-Barqani, (d. 425/ 
1033-4) traveled extensively throughout Khurasan before settling in 
Baghdad, accompanied by a massive personal library. It was al-BarqanI 
who set down and assembled one of al-Daraqutni's most famous and 
voluminous works, his prodigious Kitdb al-'ilal.^"'^' Unlike his teacher, 
however, al-Barqani managed to study extensively with al-Isma'ili and 
became the most important transmitter of his, Mustakhraj.'^' Al-Barqanl's 
interest in the Sahihayn led him to compile a musnad version of the two 
works as well as a joint mustakhraj.^^^' Al-BarqanI fell into the gray area 
of the transmission-based tradition that was gradually separating into 



"" Al-Dhahabi, Tadhkirat al-huffaz, 3:96; cf. al-Khatib, Tdnkh Baghdad, 5:210-1. 

'"i^ Ibn al-jawzl, al-Muntaiam, 14:379. 

"" For al-Barqanl's transmission of the Mustakhraj, see Ibn al-jawzl, al-Munta^am, 
14:281—2; for al-Barqanl's role in transmitting al-Isma'ill's teachings, see al-'Iraqi, al- 
Taqytd wa al-iddh, 187. 

'™ The first part of this mustakhraj has been published as al-Juz ' al-awwal min al-takhnj 
li-sahih al-hadith 'an al-shujukh al-thiqdt 'aid shart kitdb Muhammad b. Ismd'il al-Bukhdri wa 
kitdb Muslim b. al-Hajjdj al-Qushayri aw ahadihimd, ed. Abu 'Abd al-Barl Rida Bushshama 
al-Jaza'irl (Riyadh: Dar Ibn Hazm, 1420/1999). 



132 G HAPTER FOUR 

the iiber-Sunni Hanbali school and the more moderate Shaii'i strain. 
He was later identified as a Shafi'i, no doubt due to his apprenticeship 
with al-Daraqutni but more probably because of his role as a teacher 
to three of the most prominent Shafi'i scholars of the fifth/ eleventh 
century: Abu Ishaq al-Shirazi (d. 476/1083), Abu Bakr al-Bayhaqi 
and al-Khatib al-Baghdadi (who relies heavUy upon him as a source 
for his history of Baghdad). Yet al-Barqani also had strong ties to the 
tradition evolving around Ibn Hanbal: he studied with Abu Bakr b. 
Malik al-Q_atfi (d. 368/978-9), the main transmitter of Ibn Hanbal's 
Musnad from his son 'Abdallah.'"^ 

Another important member of the knot of Baghdad hadlth schol- 
ars studying the two Sahihs was al-Daraqutni's student Abu Mas'ud 
Ibrahim al-Dimashqi (d. 401/1010-11). Al-Khatib describes him as 
having a "strong interest in the Sahthayn," which he expressed in his 
famous Atraf of the two works."" Although this book exists today in 
only partial and unpublished form, hadith scholars as far-flung as Abu 
'All al-jayyani al-GhassanI (d. 498/1105), who never left Andalusia, 
and the ninth/fifteenth century Cairene Ibn Hajar regularly drew on 
it.'" In addition to the Atraf, the only book of Abu Masud to have 
reached us alludes to an interesting tension between the author and 
his teacher, al-Daraqutni. Abu Mas'ud's Kitab al-ajwiba 'amma ashkala 
al-shaykh al-Ddraqutni 'aid Sahih Muslim b. al-Hajjdj (Book of Responses 
to what al-Daraqutni Criticized from the Sahih of Muslim b. al-Hajjaj) 
contains rebuttals to twenty-five narrations that al-Daraqutni points 
out as problematic as well as to several ilzdmdt the latter suggested."^ 
In addition, Abu Mas'ud rejects al-Daraqutnl's referral to Abu Zur'a's 
criticism of four of Muslim's narrators."' Although we know little 
about his legal stances, Abu Mas'ud clearly cultivated a close personal 



'™ Al-Khatib, Tdrikh Baghdad, 5:137-40; Ibn al-jawzl, al-Muntazam, 14:333; Ibn al- 
Salah, Tabaqdt al-Jiiqahd' al-shdfi'iyya, 1:363-5; 15:242; cf. al-Dhahabi, Siyar, 17:464-8; 
idem, Tadhkirat al-huffaz, 3:183. 

"» Al-Khatib, Tdrikh Baghdad, 6:170-1. 

'" Jamal al-Din al-MizzI (d. 742/1341) states that he relied on al-Dimashqi's and 
al-Wasitl's Atrdfoi the Sahihajn in his index of the Six Books; al-MizzI, Tukfat al-ashrdf 
ft ma'rifat al-atrdf, ed. Bashshar 'Awwad Ma'ruf (Beirut: Dar al-Gharb al-lslami, 1999), 
1:102. 

"^ These ilzdmdt do not appear in al-Daraqutnl's Kitdb al-ilzdmdt wa aHatabbu'; see 
Abu Mas'ud al-Dimashql, Kitdb al-ajwiba, 287-303. 

'" See Abu Mas'ud al-Dimashql, Kitdb al-ajwiba, 331. These criticized narrators are 
Asbat b. Nasr, Qatan, Ahmad b. 'Isa al-MisrI, and Ja'far b. Sulayman, three of whom 
Abn Zur'a mentioned in his criticism of Muslim's Sahih. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 133 

relationship with the scholar later considered the third reviver of the 
Shafi'l school, Abu Hamid al-Isfaraylni (d. 406/1016)."* When Abu 
Mas'ud died, Abu Hamid led his funeral prayer and was the executor 
of his will (as his wasty).^^'' 

One of Abu Mas'ud's colleagues, Khalaf b. Muhammad al-Wasiti 
(d. ca. 400/1010) also produced a three- or four- volume atraf of the 
Sahihayn (one volume, seven j'm^'s, of which has survived in manuscript 
form)."® He studied with al-Isma'ili as well as many scholars in Baghdad 
but eventually abandoned scholarship and devoted himself to business. 
Nonetheless, prominent experts such as al-Hakim al-Naysaburi and 
Abu Nu'aym al-Isbahani studied at Khalaf 's hands."' 

The last noteworthy scholar of the Baghdad knot was Hibatallah b. 
al-Hasan al-Lalaka'l (d. 418/1027-8). Born in Rayy he studied hadith 
there before moving to Baghdad, where he studied with the city's pillar 
of the Shafi'i tradition, Abu Hamid al-Isfaraylnl. Al-Lalaka'l compiled 
a biographical dictionary of the Sahihayn, which has since been lost, 
but his most famous work was his Kitdb al-sunna}^'^ 

Along with Abu Muhammad al-Hasan b. Muhammad al-Khallal, 
(d. 439/1047), who wrote a mustakhraj of the Sahihayn,^ ^'^ these scholars 
constituted a relatively close-knit society characterized by an adher- 
ence to the Shafi'l tradition and a shared interest in al-Bukharl's and 
Muslim's works. Three of the five studied directly with al-Daraqutni, 
the progenitor of an approach to the Sahihayn as complementary texts. 
Al-BarqanI describes the close scholarly association among this cluster 
as follows: One day al-Lalaka'T approached him because he had heard 
Abu Mas'ud al-Dimashql mention that Muslim had included a certain 



"' Mahdi Salmasi, "Abu Hamid al-Isfarayini," Da'erat al-ma'aref-e bozorg-e eslatrn, 
ed. Kazem Bojnurdi (Tehran: Merkez-e Da'erat al-Ma'aref-e Bozorg-e EslamI, 1368/ 
[1989])', 5:318; al-Khatib, Tdnkh Baghdad, 5:132-4. 

"' Al-Khapb, Tdnkh Baghdad, 6:170-1; al-Dhahabi, Tadhkirat al-huffaz, 3:180. Reports 
that Abo Mas'nd studied with Ibn Khuzayma seem difficult to believe, since the latter 
died in 31 1/923. 

'"' Al-KattanI, al-Risdla al-mustatrafa, 125. 

'" Al-Khatib, Tdnkh Baghdad, 8:329-30; al-Dhahabi, Tadhkirat al-huffdi, 3:179-80. 

'"^ This has been published as Sharh usul i'tiqdd ahl al-sunna wa al-jamd'a, ed. Ahmad 
b. Sa'd b. Hamdan al-Ghamidi, 5 vols. (Riyadh: Dar Tayba, 1415/1994); al-Dhahabi, 
Tdnkh al-isldm, 28:456-7; idem, Tadhkirat al-huffaz, 3:189. Al-Lalaka'i's book on the men 
of the Sahihayn is referred to as a book of Muslim's transmitters by Ibn Abl al-Wafa' 
(d. 775/1374); Ibn Abl al-Wafa', al-Hdwi ji baydn dthdr al-Tahdwi, ed. Ytlsuf Ahmad, 
3 vols. (Beirut: Dar al-Kutub al-'Ilmiyya, 1419/1999), 1:60. 

'" Al-Khatib, Tdnkh Baghdad, 7:437-8; al-Dhahabi, Tadhkirat aHujfdz, 3:205; idem, 
Tdnkh al-isldm, 29:471-2. 



134 CHAPTER FOUR 

narration of the hadith "the signs of a hypocrite are three. . .," and he 
wanted al-Barqani to find it for him in the Sahih. Al-BarqanI looked 
through his combined musnad of the Sahihayn and discovered that the 
narration did not exist. This vindicated al-Lalaka'i suspicion that Abu 
Mas'ud had mixed up one of the names in the isndd. Al-Barqani recalls 
how Khalaf al-Wasiti was also mistaken about this narration.'^" 



d. Other: Isfahan and Central Asia 

Not all studies of the Sahihayn during the long fourth century emerged 
from Naysabur, Jurjan or Baghdad. Several important scholars worked 
independently of these regional camps. Al-Kalabadhi (d. 398/1008) 
traveled to Khurasan and Iraq, but he spent most of his life in Tran- 
soxiana.'^' The first scholar to produce a commentary on one of the 
SaMhoyn, that of al-Bukhari, was Abu Sulayman Hamd b. Muhammad 
al-Khattabi of Bust (d. 388/998). Although he studied in Baghdad and 
narrated hadlths to Abu Hamid al-Isfaraylni, Abu Dharr al-Harawi and 
al-Hakim al-Naysaburi, he remained a relative outsider in the main 
regional centers of study. He spent most of his time in Bust, in the far 
east of Khurasan. Even there his pietistic inclinations kept him far from 
public life. In one poem he wrote, "Indeed I am a stranger among Bust 
and her people . . . though my family and kin are there." '^^ Al-Khattabl's 
primary hadith interest lay in the Sunan of Abu Dawud, on which he 
wrote a famous commentary. It was only after some of his students in 
Balkh pressured him to write a commentary on al-Bukharfs work that 
he composed his A 'lam al-hadith fi sharh SaMh al-Bukhdri. Al-Khattabi 
also wrote a work on the vocabulary of al-Muzani's Mukhtasar, and his 
opinions on legal theory became a source for later Shah 'I scholars. '^^ 
Several important scholars from the Sahihayn Network also hailed 
from Isfahan. In addition to his being one of the most influential hadith 
scholars of his time, we have already noted Ibn Manda's contribution 
to the study of al-Bukharl's sources. Before him Abu Bakr Ahmad 
b. 'Abdan al-ShlrazI (d. 388/998) moved between Khurasan and the 



™ Al-Khatib, Tarikh Baghdad, 14:71-2. 

'^' Al-Khapb, Tdrikh Baghdad, 5:201; al-Dliahabi, Tadhkirat al-hu]fdi, 3:154-5; idem, 
Tarikh al-isldm, 27:355. 

'^^ Al-Subkl, Tabaqdt al-shdfi'iyya, 3:284; cf. al-Dhaliabi, Tdnkh al-isldm, 27:166-7; 
idem, Tadhkirat al-huffaz, 3:149-150; Ibn al-jawzl, al-Munla^am, 14:129. Ibn al-jawzl errs 
in al-Khattabl's death date; he includes him among those who died in 349 AH. 

''^ Al-Subkl, Tabagdt, 3:289-90. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 135 

western Iranian cities of Ahwaz and Isfahan. He produced a joint 
mustakhraj and also narrated al-Bukhari's al-Tankh al-kabir}'^* Abu Bakr 
Ahmad b. Musa Ibn Mardawayh (d. 416/1025-6) wrote a mustakhraj 
of al-Bukharl,'^"' and Abu Nu'aym al-Isbahani's separate mustakhrajs of 
al-Bukharl and Muslim have already been discussed. As the Sahihayn 
Network Chart demonstrates, however, Isfahan never became a united 
camp or developed a local tradition of studying al-BukharT or Muslim. 
Its scholars lived at different times and were more connected with the 
centers of Naysabur and Baghdad than with each other. 

e. An End to Regional Cults After 370 AH 

The study of the Sahihayn in the long fourth century thus breaks down 
along clear chronological and geographical lines. The initial popularity 
that Muslim's work enjoyed as a template for mustakhrap in his home 
city of Naysabur later developed into a more diverse interest that 
subsumed al-Bukhari's collection as well as other products of the saMh 
movement. The cluster of colleagues in Jurjan remained relatively 
isolated from Khurasan and thus cultivated an exclusive interest in al- 
Bukhari. Beginning with al-Daraqutnl, the network of Baghdad scholars 
inherited the legacies of both regions and pioneered the study of the 
two works as a pair 

By the 3 70s/ 980s, however, the regional cults of al-Bukhari and 
Muslim had disappeared. After the death of al-Ghitrifi, Jurjan faded 
into geographical and historical obscurity. The Baghdad knot was built 
on the study of the two works together, and even in Muslim's native 
Naysabur by 370 AH a study of the conjoined Sahihayn as well as 
other major products of the sahih movement eclipsed the strict focus 
on his Sahih. 



The Sahihayn Network: A Shafi'i Enterprise 

The S^ihihayn Network of the long fourth century exhibits another 
striking characteristic: study of the two works seems to have been an 
exclusively Shafi'i endeavor Although the profound work of George 



Al-Dhahabi, Tankh al-islam, 27:161; cf. al-Khalili, al-Irshad, 335. 
Al-Dlialiabl, Tadhkirat al-hujfaz, 3:169. 



136 CHAPTER FOUR 

Makdisi, Wael Hallaq, Nurit Tsafrir and Christopher Melchert has 
shed light on the formation of the Sunni madhhab\ discussing trends 
in legal and ritual identification still proves very difficult in the third/ 
ninth and fourth/tenth centuries. The inchoate intellectual landscape 
of this period resists attempts to apply the construct of the clearly 
defined Sunni madhhabs, in part because it preceded institutions like 
the madrasa that would later play important roles in their expression. 
Hallaq therefore describes this period as one of "indistinguishable 
plurality"'^'' This period retains the startling diversity of early Islam, 
as schools of law usually dismissed as phenomena of the second and 
third centuries survived. It was only in 347/958-9, for example, that 
the last mufti of the Awza'i school died in Damascus.'^' One of the 
most important transmitters of Muslim's Sahih, al-Juludi (d. 368/979), 
followed the moribund madhhab of Sufyan al-Thawri.'-" 

Indeed, the undeniable presence of the regularized four Sunni schools 
marks the end of the long fourth century. With a cadre of scholars such 
as Abu Bakr al-Bayhaqi, al-Khatib al-Baghdadi, Imam al-Haramayn 
al-Juwaynl (d. 478/1085) and Abu Ishaq al-ShirazI (d. 476/1083), for 
example, we can for the first time feel totally at ease discussing a broad 
and unshakable guild-like loyalty to a Shafi'i school. Only in the ample 
wake of the long fourth century can we rely on the well-worn stereo- 
types invoked by al-Hasan b. Abi Bakr al-Naysaburl in 536/ 1 142 when 
he told a congregation, "Be Shafi'i but not Ash'arl, be Hanafi but not 
Mu'tazilT, be Hanball but not anthropomorphist.'"^^ 

In the long fourth century the arena for the study of the Sahihayn 
extended from Transoxiana to the Hijaz. There the enduring distinc- 
tion between "the two sects [al-fanqdrij" of the transmission-based and 
reason-based scholars still ruled. The Hanafis/«^/ al-ra'y were develop- 
ing a keener interest in hadlth, but the school retained its link with the 
Mu'tazilite doctrine so anathema to the ahl al-hadith. The doyen of the 
Hanafi hadlth tradition, Abu Ja Tar al-Tahawi of Egypt (d. 321/933), 
seems to have been in a minority with his distance from Mu'tazilism. 
Abu al-Hasan 'Ubaydallah b. al-Husayn al-Karkhi (d. 340/952), the 



'^^ Hallaq, Authority, Continuity and Change in Istamic Law (Cambridge: Cambridge 
University Press, 2001), 61. 

'" Abu Zahra, at-Shdfi'i, 339, 

128 xhis according to al-Hakim al-Naysabflrl. See, Ibn al-Salah, Siydnat Sahih Muslim, 
107; al-Dhahabi, Sijar, 16:302. 

™ Ibn al-jawzl, at-Munta^am, 18:31. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 137 

most prominent Iraqi Hanafi of his time, is described as a leading 
Mu'tazilite [kdna ra's^'Jial-i'tizal).'-'" Ahmad b. Yusuf al-Tanukhl, who 
learned^^A from al-Karkhl, was from a "house of hadith" but was none- 
theless Mu'tazilite.'^' 'All b. Muhammad al-Tanukhl (d. 342/953) was 
also a Hanafi hadith scholar knowledgeable in Mu'tazilite kaldm.'^^ 

It was the monolithic construct of the ahl al-hadith that was becoming 
increasingly insufficient for describing the divisions among transmis- 
sion-based scholars. Two distinct strains were emerging. Al-Bukharl's 
persecution at the hands of fellow hadith scholars illustrated a break 
between the conservative iiber-Sunni interpretation of Ibn Hanbal's 
legacy and a more moderate transmission-based approach, which 
Melchert has dubbed "semi-rationalist." These two strains would later 
emerge as two competing parties in the Sunni Islamic heartlands, 
the Hanbali/iiber-Sunni school and its rival Shafi'i/Ash'ari camp. In 
the long fourth century, however, these two budding schools shared a 
common heritage. Abu Zur'a al-Razi heard the entirety of al-Shafi'l's 
oeuvre from Rabf, yet he is claimed as a Hanbali.'^'' Ibn Abl Hatim 
devoted a work to the virtues of al-Shafi'i but is similarly claimed by 
Hanballs."* This ambiguity was deeply rooted in the career of Ibn 
Hanbal himself, for it is reported that he considered al-Shafi'l to be 
his century's reviver of the faith. '^'' The Maliki school, based in Egypt 
and the lands of the Maghrib, proves tangential to the Sahihayn Net- 
work. Only Qasim b. Asbagh of Cordova and Abu Dharr al-Harawi 
belonged to the Maliki school. 

Identifying the porous boundaries between the emerging Hanbali 
and Shafi'i strains is challenging in the long fourth century. In the early 
stages one cannot yet consistently identify legal schools through telltale 
shibboleths like the Shafi'l insistence on the voiced basmala (saying 
'bismilldh al-Rahmdn al-Rahim' out loud in prayer). An early scholar like 
Abu 'Awana is considered the person who brought the Shafi'i school 
to Isfarayin, but he broke with what became important madhhab stances 
such as the basmala and the issue of what invalidates prayer. 



i^" Ibn al-jawzl, al-Muntazam, 14:85; cf. Ahmad b. Yahya Ibn al-Murtada (d. 839/ 
1437), Tabaqdt al-mu'tazila, ed. Suzanna Diwald-Wilzer (Beirut: Dar Maktabat al-Hayat, 
[198-]), 130. 

''' Ibn al-Murtada, Tabaqdt al-mu'tazila, 108. 

"^ Ibn al-jawzl, al-Muntazam, 14:90. 

"' See Abu Zalira, al-Shdji'i, 148; Henri Laoust, "Hanabila," EP. 

'" Ibn Abl Ya'la, Tabaqdt al-handbila, 2:47-8. 

"^ Abu Zahra, Ibn Hanbal, 29. 



138 CHAPTER FOUR 

The distinction between the two transmission-based strains becomes 
more evident in their attitudes towards rationalism in perennial con- 
troversies such as the lafz of the Qur'an and the use of speculative 
theology [kalam). Melchert describes how by the early fourth/tenth 
century a "vague Shafi'l school" had emerged that "comprised both 
a particular system of jurisprudence and a particular theological ten- 
dency." "It was a compromise," he states, espousing traditionalist tenets 
but very often defending them rationally'^'' In the early 300s/900s this 
distinction is problematic, since an incontrovertibly Shafi'i scholar like 
Ibn Khuzayma proved one of the most ruthless critics of those who 
upheld the created wording of the Qur'an. Yet by the time of al-Khatlb 
al-Baghdadi in the mid-400s/ 1000s, this intransigence on questions 
of rationalism had become a hallmark of the Hanball school, not the 
Shafi'l. Al-Khatlb began his scholarly career as a Hanball, but moved 
to the Shafi'i camp after his Hanball cohorts relentlessly criticized his 
indulgence in Ash'ari rationalist discourse. Ibn al-Jawzi, a later Hanball 
openly offended by al-Khatlb's defection, notes how the newly christened 
Shafi'i began mocking Ibn Hanbal's legendary intransigence on the 
issue of the created Qjur'an.'^' An incontestable Shafi'l, al-Daraqutni 
distrusted a reliance on reason and rejected famous hadlths praising it. 
Yet he also evinced an appreciation for the use of kalam. He reportedly 
told Abu Dharr al-HarawI that one of the founding members of the 
Ash'arT school, Abu Bakr al-BaqillanI (d. 403/1013), was "the imam of 
Muslims and the defender of the religion [al-dhdbb 'an al-diri).'"^-'^ Despite 
his personal aversion to speculation, al-Daraqutnl had himself written 
a refutation of the Mu'tazila and probably understood its utility in 
defending against rationalist opponents. 

Perhaps the most effective way to identify the two strands, however, 
is through personal relationships and textual transmission. Daphna 
Ephrat asserts that even after the dawn of the madrasa and the distinct 
Sunni madhhabs, in the late fifth /eleventh century, it was the bonds of 
personal loyalty between teachers and their students that proved the 
most cohesive.''^ In the long fourth century both the emerging Shafi'i 



'"■ Melchert, The Formation of the Sunni Schools of Law, 70. 

"' Ibn al-jawzl, al-Munta^am, 16:132. 

"» Al-Dhahabi, Tadhkirat al-hufBz, 3:202. 

"" Ephrat, A Learned Society in a Period of Transition, 88. For a fascinating study on the 
tight links between the development of Sufism in Khurasan and the Shafi'l tradition, 
see Margaret Malamud, "Sufi Organizations and Structures of Authority in Medieval 
Nishapur," International Journal of Middle East Studies 26, no. 3 (1994): 427-442, esp. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 139 

and Hanball camps expressed themselves most clearly through the 
teachings of specific individuals with strong attachments to the legacies 
of the two eponymous founders. The nascent schools extended out 
from these individuals, whom Mel chert refers to as "local chiefs,'"*" 
through teacher/ student relationships and through the study of for- 
mative texts. 

At the epicenter of the Shafi'i pedagogical and textual tradition 
were his most prominent students, Rabf and al-Muzani. Their student 
Ibn Khuzayma became a bastion of the Shafi'i tradition in his native 
Naysabur Another student of Rabf, Muhammad b. Nasr al-Marwazi 
(d. 294/906) of Samarqand, became one of the first scholars to dis- 
cuss the "madhhab" of al-Shafi'l and elaborate his stances on legal 
theory'*' Later Baghdad scholars such as Ibn Surayj and Abu Hamid 
al-Isfarayini also served as pivots for the Shafi'T tradition during the 
long fourth century. In addition to scholarly relationships with these 
pillars, the Shafi'l tradition propagated itself through the transmission 
of its formative text, al-Muzani's Mukhtasar of al-Shafi'l's Umm. While 
the Shafi'l scholar al-Isma'lll produced an independent treatise on legal 
theory, many of the nascent school's adherents preferred to write com- 
mentaries or studies on the Mukhtasar. 

The tradition of Ibn Hanbal likewise propagated itself through a 
network of scholars tied closely to the school's two formative texts, Ibn 
Hanbal's Musnad and what developed as the definitive collection of his 
legal opinions. Ibn Hanbal's son 'Abdallah served as the most commit- 
ted transmitter of his teachings, crafting a finished draft of his father's 
Musnad. Abu Bakr al-Qati'i transmitted the Musnad from Ibn Hanbal's 
son and became a central figure in disseminating his teachings. The 
earliest extant collection of Ibn Hanbal's legal and doctrinal responsa, 
the Kitdb al-masd'il, was the work of Abu Dawud al-Sijistani.'*^ Abu 
Hatim al-RazI also collected a selection of Ibn Hanbal's responsa, and 
later the school claimed his son Ibn Abl Hatim as a member Abu Bakr 



430. For a discussion of the formative period of the Hanafi school, see Nurit Tsafrir, 
The History of an Islamic School of Lam: The Early Spread of Hanafism (Cambridge, Mass.: 
Harvard University Press, 2004). 

"" Melchert, The Formation of the Sunni Schools of Law, 87. 

'" Muhammad b. Nasr al-MarwazI, al-Sunna, ed. 'Abdallah b. Muhammad al-BasIrl 
(Riyadh: Dar al-'Asima, 1422/2001), 231. The entire second half of this work consists 
of a discussion of al-Shafi'l's school of thought on the issue of abrogation inaskh). 

112 -pj^jj work has been published as AbU Dawud al-SijistanI, Kitdh masd'il al-imdm 
Ahmad, 16 vols. (Beirut: Muhammad Amin Damaj, [197—]). 



140 CHAPTER FOUR 

al-Khallal (d. 311/923-4) traveled extensively in a quest to unite Ibn 
Hanbal's legal legacy and compiled a massive collection of his opinions 
as well as other works, such Ibn Hanbal's 'Hal. He also compiled the 
first roster of Hanbalis. Al-Khallal's student Abu al-Qa.sim al-Khiraql 
(d. 334/945-6) edited his master's work and produced the school's 
formative legal text, the Mukhtasar}*'^' 

The intellectual landscape of Iraq and Iran in the long fourth cen- 
tury thus consisted of three dominant schools: the Hanafi ahl al-ra'j, 
the Hanbali/iiber-Sunnis and the nascent Shafi'l tradition. In order to 
place the network of Sahihayn scholars in this milieu, we can identify 
Shafi'ls as exhibiting three major characteristics. Firstly, they are not 
Hanafi. Secondly, they tend to be more moderate than their liber-Sunni 
counterparts. Finally, they exist within a network of personal and textual 
relationships with bastions of the school such as Ibn Khuzayma and 
al-Muzani's Mukhtasar. 

Oddly, not a single scholar from the Sahihayn Network is claimed as 
Hanafi in the definitive rosters of the school.'** While Hanafi scholars 
did not participate in the study of al-Bukharl's and Muslim's works, 
they did play noted roles in the transmission of the two texts. '*'^ 
According to Ibn al-Salah, the critical transmitter of Muslim's SaMh, 
Ibn Sufyan, was probably Hanafi.'*'' Abu al-Khayr Muhammad b. 
Musa al-Saffar (d. 471/1078-9), one of the most prolific transmitters 



"' For more information, see Laoust, "Hanabila," EI'; Abu Zahra, Ibn Hanbal, 
179— 188; Melchert, The Formabon of the Sunni Schools of Law, 144— 6; Nimrod Hurvitz, Tlie 
Formation of Hanbalism: From Piety to Power (London: Routledge-Gurzon, 2002), 78-90. 

'** The most comprehensive is the Jawdhir al-mudtyya of Ibn Abl al-Wafa' (d. 775/ 
1374). For an earlier hst, al-'Abbadl's Tabaqdt alfuqahd' al-shdfi'tyya includes a lengthy 
list of scholars whom this fifth/eleventh-century scholar considered Hanafi; al-'Abbadi, 
Tabaqdt alfuqahd', 2 if 

'^' Here we must note the work of Abn al-Layth al-Nasr b. Muhammad al- 
Samarqandl (d. 373/983-4 or 393/1002-3), a Hanafi jurist and exegete of Transoxiana. 
One of his lesser known works, al-Latd'if al-mustakhraja min Sahih al-Bukhdri (Useful 
Niceties Derived from Sahih al-Bukhdri), would seem to have been small collection of 
the author's musing on elements from the Sahih but could not have qualified as either 
a commentary on the work or a study of its hadlth science dimensions. The unique 
manuscript of the Latd'fwas in the rare books library at Istanbul University, and was 
"lost" after the terrible 1999 earthquake. Some Turkish scholars debate whether the 
work ever existed. 

'**' Ibn al-Salah, Siydnat Sahih Muslim, 107; cf Ibn al-jawzl, al-Muntaiam, 14:267. Ibn 
Sufyan is not, however, included in Ibn Abi al-Wafa"s al-Jawdhir al-mudiyyaf tabaqdt 
al-hanafiyya. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 141 

of Sahih al-Bukhdff from al-Kushmihani, was Hanafi.'*' Abu Talib al- 
Husayn b. Muhammad al-Hashimi (d. 512/1118-1119), one of the 
main transmitters of the Sahih from the famous Meccan female student 
of al-Kushmlhani, Karlma al-Marwaziyya, was also Hanafi.'*" 

Why would Hanafis actively and enthusiastically transmit al-Bukhari's 
and Muslim's Sahihs but not study the works? One possible explanation 
lies in the function of the mustakhrcy's that sparked the flurry of interest 
in the Sahihayn. Mustakhrap were interpretations of formative texts that 
allowed transmission-based scholars to express and elaborate their rela- 
tionship with the source of hermeneutic authority in Islam. For Hanafis 
this role was already played by the school's formative legal texts. For 
them the chain of legal scholars emanating from Abu Hanlfa and his 
students provided that link to the Prophet's message. 

Neither did the network of Sahihayn scholars identify with the 
Hanbali/iiber-Sunni tradition. Only one member of this group, Ibn 
Manda, is listed as Hanbali in Ibn Abi Ya'la's Tabaqat al-hanabila.'*'^ 
The Hanbali school seemed to prefer critics of al-Bukharl or Muslim 
such as Abu Hatim al-Razi and his son Ibn Abi Hatim.'^" None of 
the well-known Hanbalis of the period, such as Abu Bakr al-Najjad 
(d. 348/959-60) of Baghdad, Abu Bakr al-AjurrI (d. 360/971) and al- 
Hasan b. Hamid al-Warraq (d. 403/1012-13), appears in the Sahihayn 
Network. Given al-Bukhari's pariah status among iiber-Sunnis, it is not 
difficult to understand why they did not participate in the study and 
transmission of his Sahih- We have already discussed how the dominant 
scholarly presence in Naysabur of the iiber-Sunnis Ibn Khuzayma and 
al-Sarraj played a central part in preventing the study of al-Bukhari's 
collection in that city. The attitude of iiber-Sunni members of the Bagh- 
dad scholarly community did not differ Al-Hasan b. 'All al-Barbahari 
(d. 329/940-1) was one of the Hanbali tradition's most outspoken 
advocates in Baghdad. He never mentions al-Bukhari in his manifesto 
of the ahl al-hadith creed, the Sharh al-sunna (Explanation of the Sunna), 
but he does assert that anyone who says that the lafz of the Qur'an is 



'" Ibn Abi al-Wafa', al-Jaivahir al-mudijja, 3:215; al-Dhahabi, Tadhkirat ahhuffaz, 
3:245. 

"« Al-Dhahabl, Tadhkirat al-hujfdz, 4:32. 

'" Ibn AblYa'la, Tahaqdt al-handbila, 2:142-3. 

''" Laoust, "Hanabila," EP . 



142 CHAPTER FOUR 

created is a heretic (mubtadi').'-'^ Although he did not officially belong to 
the Hanbali madhhab, AbQ Hafs 'Umar b. Ahmad Ibn Shahin (d. 385/ 
996) provides another interesting example of this scholarly strain in 
the Abbasid capital. Ibn Shahin heard from many of the same teach- 
ers as his contemporary al-Daraqutni, whom he enlisted at least once 
to review his hadlth corpus.'"'^ Yet Ibn Shahin is completely absent in 
the network of Sahihayn scholars. In his Shark madhdhib ahl al-sunna wa 
ma'rifat shardV al-din wa al-tamassuk bi'l-sunan (Explanation of the Ways 
of the Ahl al-Sunna, Knowledge of Religious Law and Clinging to the 
Sunna), he echoes al-Barbahari by narrating that anyone who says that 
the lafz of the Qur'an is created is Jahml, or worse. '"'^ 

Still, how do we explain the absence of iiber-Sunni interest in 
Muslim's Sahth? Unlike al-Bukharl, he was not tainted by the lafz 
scandal. It seems most likely that in the first half of the fourth/tenth 
century Muslim's collection was simply not well-circulated in the 
Hanball/iiber-Sunni bastion of Baghdad. We do know that the work 
had limited circulation in places like Jurjan and seems to have been 
relatively unknown in the Hijaz through the first half of the fourth/tenth 
century. Al-'Uqayli (d. 323/934) of Mecca knew al-Bukharl's al-Tdnkh 
al-kabir intimately but never refers to Muslim in any form in his Kitdb 
al-du'ajd'. That al-'Uqayli totally rejects a hadlth found in Muslim's 
SaMh without mentioning the work reinforces the notion that he was 
ignorant of it.'"'* Another notable non-Khurasani hadlth scholar of the 
mid 300s/900s, al-Hasan b. 'Abd al-Rahman al-RamahurmuzI, likewise 
makes no mention of Muslim. 

Unlike the Hanbali/iiber-Sunnis, members of the Shafi'i tradition 
actively accommodated al-Bukharl and Muslim. In their treatises on 
the Sunni creed and proper ahl al-sunna stances, both al-Barbahari and 
Ibn Shahin had implicitly condemned al-Bukhari for his stance on the 



'"'' Abu Muhammad al-Hasan al-Barbahan, Shark al-sunna, ed. Khalid b. Qa.sim 
al-Raddadi (Beirut: Dar al-Sumay'l; Riyadh: Dar al-Salaf, 1421/2000), 92. 

''^ Al-Khatib, Tdrikh Baghdad, 11:264-7; al-Dhahabi, TdrM al-isldm, 27:107. 

''' Abu Hafs 'Umar b. Ahmad Ibn Shahin, Sharh madhdhib ahl al-sunna wa ma'rifat 
shard'i' al-din wa al-tamassuk bi'l-sunan, ed. 'Adil b. Muhammad (Cairo: Mu'assasat 
Qurtuba, 1415/1995), 32. 

'^* This hadlth is, "If two caliphs receive allegiance kill the second of them . . . {idhd 
bUyi'a li-khalifatayn fa-'qtuld al-dkhir minhumd...), and al-'Uqayll criticizes it in his biog- 
raphy of Fadala b. Dinar, saying: "Narration on this topic is not sound [wa al-riwdya 
fihddhd al-bdb ghayr thdbit}." We know this represents a blanket dismissal of the hadlth 
because when al-'Uqayll merely criticizes narrations he uses the term 'wajh'; al-'Uqayll, 
Kitdb al-du'ajd', 3:1144. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 143 

lafz issue. The Shafi'i al-Lalaka'l, however, affirms both al-Bukharl's and 
Muslim's worthiness as commendable Sunnis. His Kitab al-sunna focuses 
overwhelmingly on the controversial sectarian issues of the nature of the 
Qjur'an and the definition of faith {imdn). Yet he cites al-Bukharl as one 
of a small set of exemplary figures who upheld the Sunni definition of 
faith as including both a profession of belief and proper practice [qawl 
wa 'amal). Al-Lalaka'i lists al-Bukharl in the company of al-Awza'l, Ibn 
Hanbal, al-Shafi'i and al-Muzani, even including two quotations from 
him.'-'"' He also lists both al-Bukharl and Muslim as two of the scholars 
who upheld the uncreated nature of the Qur'an, along with Abu Zur'a, 
Abu Hatim al-Razi and Abu Dawud.''^'' Al-Lalaka'i's book, in fact, 
represents the first work in the Sunni creed genre to accept al-Bukhari. 
The Sahihayn Network proved fairly accommodating to rationalists as 
well. Both Abu Nu'aym al-lsbahani and Abu Dharr al-HarawT were 
Ash'aris, and al-Isma'ili had marked rationalist tendencies. 

Of the forty-four scholars in the network who composed works on 
the Sahihayn, fuUy fourteen (32%) directly studied with or instructed Abu 
Hamid al-Isfarayinl, Ibn Khuzayma, Ibn Surayj, Rabi' al-Muradi or 
al-Muzani. Six (14%) of them either wrote books based on al-Muzanl's 
Mukhtasar or composed their own works on al-Shafi'l's legal method. Ten 
(23%) are later explicitly referred to as Shafi'ls by al-Dhahabl. He calls 
Abu al-Nadr Muhammad b. Muhammad al-Tusi (d. 344/955) ^^shaykh 
al-shafi'iyya," which should not surprise us since he studied extensively 
with Muhammad b. Nasr al-MarwazI in Samarqand.'" Abu al-Walid 
Hassan b. Muhammad al-UmawI of Naysabur (d. 344/955) studied^^^ 
in Baghdad with Abu al-'Abbas Ibn Surayj and composed legal rulings 
{ahkarn) for the madhhab. He even had a ring patterned after those worn 
by Rabf b. Sulayman and al-Shafi'i.'"* 



157 



Al-Lalaka'l, Shark usul i'tigad ahl al-sunna wa al-jamd'a, 5:959. 

Al-Lalaka'l, Shark usul i'tigad ahl al-sunna wa al-jamd'a, 1:302. 

yiu&S.Y^a.tn:, Makdnat al-Sahihayn, 176; al-Dhahabi, Tadhkirat al-hujfai, 3:73; idem, 
Tdrikh al-isldm, 25:311—12; cf. al-'Abbadi, Kitdb tabaqdt al-fuqahd', 11. 

'^' Ibn al-Salah, Siydnat Sahih Muslim, 90; al-Dhahabi, Tadhkirat al-hujjdi, 3:75; idem, 
Tdrikh al-isldm, 25:417-8; cf. al-'Abbadi, Kitdb tabaqdt al-fuqakd', 74. 



144 CHAPTER FOUR 

Intense Canonical Process: 
Imagining a New Epistemological Status for Hadith Books 

The long fourth century had not simply seen a profound interest in the 
Sahihayn among a relatively limited network of scholars. In this period 
before the canonization of the two works, we also see the appearance 
of what Frank Kermode called a "canonical habit of mind" in the 
Muslim community in general.'"'^ For the first time Muslim scholars 
began discussing the hadith tradition in terms that endowed certain 
books with a sense of communal and epistemological preeminence. 
Among hadith scholars this derived from personal convictions about the 
broad acceptance and overwhelming utility of certain books. For legal 
theorists this resulted from an increased application of the notion of the 
community's authoritative consensus, ijma\ to the hadith corpus. What 
lay behind both these perceptions, however, was a new conception of 
what kind of authority certain hadlths and specific hadith collections 
could exercise. It was in this period that the Sunni community imagined 
a new epistemological status for hadith works. 

The notion of authoritative consensus {ijmd') has ancient origins in 
Islam. In addition to functioning as one of the primary means of justi- 
fying decisions during the time of the Companions and their followers, 
it arose quickly as a tool in debates between the early schools of law 
in cities like Kufa.'™ By the time of the eponymous founders of the 
four madhhabs, hadlths were circulating that established the consensus 
of the community as a source of legal and doctrinal authority. One 
of the most famous was the tradition in which the Prophet says, "My 
community will not agree on error (Id tajtami'u ummati 'aid al-daldla)."""'^ 
In correspondences between al-Awza'l and Abu Hanlfa's chief disciple 
Abu Yusuf (d. 182/798), each contested the other's claim that his 
stances enjoyed the consensus of the Muslim community"'^ Later, al- 



'™ Kermode, "The Canon," 601. 

"*" Hallaq, A History of Islamic Legal Theories (Cambridge: Cambridge University 
Press, 1997), 20. For more discussion on the development of ijmd', see idem, "On the 
Authoritativeness of Sunni Consensus," International Journal of Middle East Studies 18 
(1986): 427-54. 

'*'' Wahba al-Zuhayll, UsUl al-fiqh al-isldmi, 2 vols. (Beirut: Dar al-Fikr al-Mu'asir, 
1406/1986), 1:488. See also, Abu Bakr Muhammad b. Ahmad al-SarakhsI, Usui al- 
Sarakhsi, ed. AbQ al-Wafa' al- Afghani, 2 vols. (Beirut: Dar al-Kutub al-'Ilmiyya, 1414/ 
1993, reprint of the Lajnat Ihya' al-Ma'arif al-Nu'maniyya edition from Hyderabad, 
citations are to Beirut edition), 1:299. 

'^^ Abu Zahra, Ibn Hanbal, 260—1; Zafar Ishaq Ansari, "Islamic Juristic Terminology 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 145 

Shafi'i and Ibn Hanbal grew very skeptical of such claims about ijma'. 
Although they acknowledged that ijma' existed as a source of author- 
ity among Muslims, they limited it to fundamental issues, such as the 
ordination of the five daily prayers, that truly enjoyed total communal 
consensus. Their skepticism was well-founded, as the later Shafi'i jurist 
Abu Ishaq al-Isfarayini (d. 418/1027) estimated that "the questions 
on which ijma' has been invoked {masd'il al-ijma') number more than 
twenty thousand.""'-' 

By the time of al-Shafi'i in the early third/ninth century the notion 
of universally agreed-upon precedent from the Prophet was manifest- 
ing itself in scholarly discourse. Al-Shafi'i placed "sunna on which 
consensus has been achieved" on the same level of legal compulsion 
as the Qur'an. As opposed to hadith with limited attestation (khass), 
those who knowingly rejected such reports must repent immediately"** 
Even later in the thought of the Ibn Surayj, however, this articulation 
remained primitive."''^ 

Al-Tabari discussed these most authoritative instances of the Prophet's 
sunna in the more technical terms of hadith study. These were reports 
so widely transmitted [mustajid qati'"") that they are epistemologically 
certain. Indeed, rejecting them places one outside the pale of Islam. 
These include reports such as the hadith ordering stoning as a pun- 
ishment for adultery'* More importantly, however, on two occasions 
al-Tabari refers to certain reports that are not massively transmitted 
but nonetheless convey a great deal of certainty. Al-Tabarl describes 
a hadith in which God states that He will remove certain people from 
Hellfire after they have been appropriately punished for their sins as 
coming from "someone whose transmission prohibits error, oversight or 
lying and yields certainty {'ilm) ""*' We thus see nascent in al-Tabarl's 



before Safi'i: A Semantic Analysis with Special Reference to Kufa," Arabica 19 (1972): 
282-7. 

"^' Al-Zuhayli, Usui al-fiqh, 1:489. 

'^^ See Norman Calder, '^ Ikhtilaf xad Ijmd' in al-Shafi'i's Risala," Studia Islamica 58 
(1983): 60, 74-8; al-Shafi'l, al-Umm, 7:255. 

"•^ Ibn Surayj, "al-Wada'i' li-mansus al-shara'i'," ed. Salih al-Duwaysli (unpublished 
manuscript), 2:672-3. Here Ibn Surayj states that the consensus of the umma on a 
report is merely one way in which a hadith is established as legally compelling. I am 
totally indebted to my friend and colleague Ahmed El Shamsy of Harvard University 
for this citation and for providing me with the text itself. 

'«« Al-Tabarl, at-Tabm, 161. 

"'' Al-Tabarl, al-Tabsir, 185. For the other instance, see 212. Although he does not 
cite it from any sources, this hadith appears in the Sahihayn. See Sahih al~Bukhdn: kitdb 



146 CHAPTER FOUR 

thought the idea that certain transmitters or collectors could themselves 
guarantee the authenticity and epistemological yield of non-massively 
transmitted [ahad) hadiths. 

The concept of universally agreed-upon hadiths extended beyond 
Sunni circles. The Mu'tazilite Abu al-Qasim al-Balkhi writes in his 
Qubul al-akhbdr that the ultimate test for determining a good narrator 
or report is its accordance with the Qur'an, the sunna "agreed upon 
by consensus imujma"alayhi)," the ijmd' of the umma, the ways of the 
early community and the Mu'tazilite slogans of justice ('adl) and God's 
unicity [tawhid)}'''^ 

Although Abu Nu'aym al-Isbahani lived a century later than these 
scholars, his work nonetheless affords an interesting glimpse into the 
place of hadlth consensus in sectarian debates. One of the chief impedi- 
ments he faced in his dialectical handbook for debating Imami Shiites 
was the different repertoires of hadiths from which the two sides drew 
proof texts. As a solution to this lack of common ground, al-Isbahani 
proposed that "the recourse at that point is to what the umma has agreed 
on after the Prophet (s), and those authentic [sahih) reports [akhbdr) from 
him that the scholars have transmitted and are uncontested [Id ddfi' 
lahd)."^^''' Abu Nu'aym is not admitting any parity between Sunni and 
Shiite hadiths; quite the opposite, he maintains that Sunnis actually 
uphold standards for using hadiths as proof texts, while Shiites use 
forged reports.'™ But here we see the notion of shared and commonly 
accepted material that neither camp can contest. 

The epistemological status of these universally accepted reports and 
their role in deriving law also began receiving more attention in the 
long fourth century. Unlike al-Shafi'l and Ibn Hanbal, who believed 
dhdd traditions of the Prophet could be used to determine issues of 
dogma and abrogate Qur'anic verses, the Hanafi tradition remained 



al-riqdq, bdb sijat al-janna wa al-ndr; Sahih Muslim: kitdb al-imdn, hdb ithhdt al-shajd'a wa 
ikhrdj al-muwahhidin min al-ndr. Another hadlth he cites in this context appears in the 
collections of Ibn Hibban and Ibn Khuzayma. 

'«" Al-Balkhl, Qubul al-akhbdr, 1:17, Even earlier, al-Jahiz (d. 255/868-9) had men- 
tioned a report accepted by consensus {khabar mujtama"alayhi) as one of the four sources 
of knowledge, citing the founder of the Mu'tazilite school, Wasil b. 'Ata' (d. 131/750), 
as the originator of this idea; Marie Bernand, "la Notion de 'Ilm chez les premiers 
Mu'tazilites," Studia Islamka 36 (1972): 26. 

"'" Al-IsbahanI, Kitdb al-imdma, 244. Although he does not cite any collections, the 
hadiths he then presents are all found in either al-Bukhan or Muslim, with one in 
al-Tirmidhl's collection. 

'™ Al-Isbahanl, Kitdb al-imdma, 241. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 147 

very wary of endowing these relatively uncorroborated reports with 
such authority The concept of universally accepted hadlths, however, 
emerged as a common ground acceptable to Hanafis. Like al-Tabari, 
the early Hanafi legal theorist Abu Bakr Ahmad al-Jassas of Rayy 
(d. 370/982) acknowledged that there exists a category of reports that 
lack massive transmission {tawdtur, istifada) but nonetheless convey epis- 
temological certainty'^' For these ahad hadlths to yield such knowledge 
and function in abrogating Qiar'anic verses, for example, certain indica- 
tions (daldld) must accompany them assuring their authenticity These 
include reports that enjoy the consensus (ijmd') of the umma's scholars, 
such as the report denying members of a family guaranteed a portion 
of the deceased's estate from receiving additional inheritance (Id wasiyya 
li-wdrith)}''^ Following the earlier Hanafi scholar 'Isa b. Aban, al-Jassas 
states that dhdd reports that are used in important issues of dogma and 
ritual [umur al-diydndt) must be widespread [shd'i'a musta/tdd) in the umma, 
which accepts [talaqqatha] and acts on them."^ 

Among hadith scholars, this new epistemological status attainable 
by hadlths is evident in a revised historical conception of the hadith 
tradition. This new vision viewed the sahih movement in general and 
certain collections in particular as loci of scholarly consensus. While 
previously we have seen that scholars such as Ibn Abi Hatim identi- 
fied the pinnacle of the hadith tradition with the greatest generation 
of Ibn Hanbal and ignored the existence of the sahih movement, Ibn 
Manda's perspective is very different. Like Ibn Abi Hatim, Ibn 'Adi 
and Ibn Hibban, he lists the generations [tabaqdi) of hadith scholars 
up to the generation of Ibn Hanbal, 'All b. al-Madinl and Ibn Ma'in. 
In a novel step, however, he then mentions the "four imams," who 



'" For a discussion of al-Jassas's legal theory, see Marie Bernand, "Hanafi Usui 
al-Fiqh through a Manuscript of al-Gassas, " J'oMraa/ of the American Oriental Society 105, 
no. 4(1985): 623-35. 

'" Abu Bakr Ahmad al-Jassas, Usui al-Jassas, al-musammd al-Fusul fi al-usul, ed. 
Muhammad Muhammad Tamir, 2 vols. (Beirut: Dar al-Kutub al-'Ilmiyya, 1420/2000), 
1:532—5. The numerous narrations of this hadith have been individually criticized, but 
scholars have generally agreed that the text of the hadith is too widely attested and has 
been accepted too widely to be false. Al-Shafi'l even described it as eifectively mutawdtir; 
Ibn Hajar, Fath, 5:467-9; cf Abu Ibrahim Muhammad b. Isma'll al-AmIr al-San'anI, 
Tawdih al-ajkdr li-ma ani Tanqih al-an^dr, ed. Abu Abd al-Rahman Ibn 'Uwayda, 2 vols. 
(Beirut: Dar al-Kutub al-'Ilmiyya, 1417/1997), 1:229. See also, David Powers, Studies 
in Quran and Hadith: The Formation of the Islamic Law of Inheritance (Berkeley: University 
of California Press, 1986), 159-64. 

"' Al-Jassas, Usui, 1:548. Such reports include the hadith of the Prophet accepting 
the word of one Bedouin that the new moon of Ramadan was visible. 



148 CHAPTER FOUR 

produced the sahih books: al-Bukhari, Muslim, Abu Dawud and 
al-Nasa'i. He notes other, less impressive installments of the sahih 
movement as well, such as the works of al-Darimi, al-Tirmidhi, Ibn 
Khuzayma and Ahmad b. Abl 'Asim al-Nabil. Although they followed 
in the footsteps of the four imams, "they were less skilled.'"'* This 
generation that Ibn Manda describes as studying at the hands of Ibn 
Hanbal and his cohort, however, has achieved an unprecedented station. 
"Al-Bukhari, al-Hasan b. 'All al-HulwanI, al-Dhuhll, Abu Zur'a, Abu 
Hatim, Muslim, Abu Dawud, and al-Nasa'i . . . make up the generation 
{tabaqd) accepted [by all] by consensus, and their knowledge trumps 
all others [wa bi-'ilmihim yuhtajju 'aid sd'ir al-nds)."^''' Ibn Manda thus 
articulates the notion that the generation of al-Bukhari and Muslim 
represents a compelling concentration of knowledge agreed upon by 
all. More importantly, this mastery is articulated in the sahih collections 
of four scholars who embody the authority of their age. 

Implicit in Ibn Manda's genealogy of the hadith tradition is the 
same problem that Abu Nu'aym faced in his polemic: the vast cor- 
pus of hadiths had become too broad and diverse to be succinctly 
studied and employed. Specific outstanding collections that embody 
the utility of the hadith tradition should thus be viewed as common 
references. Ibn Manda echoes a statement attributed to the Egyptian 
hadith scholar and transmitter of al-Bukhari's Sahih, Ibn al-Sakan 
(d. 353/964). Disturbed by the great number of hadith collections 
flooding the book markets, a group of hadith scholars gathered at Ibn 
al-Sakan's house asking him to direct them to what books they should 
study at the expense of others. Ibn al-Sakan entered his house and 
reemerged with four books, saying "these are the foundations (qawd'id) of 
Islam: the books of Muslim, al-Bukhari, Abu Dawud and al-Nasa'i.'"''' 
These four collections are thus not only the most important for students 
of hadith, they also provide the common references to be shared by 
all. Ibn al-Sakan's own sahih work, in fact, may have been little more 
than a digest of these four books.''' 



"* Ibn Manda, Shurut al-aimma, 42^43; cf. al-Dhahabi, Siyar, 14:135 (biography of 
al-Nasa"!). 

"■' Ibn Manda, Shurut al-a'imma, 67-8. 

'"* Ibn Hazm 'All b. Ahmad, "[Two Hadiths from the Sahihajn — One from al- 
Bukharl and One from Muslim — that Ibn Hazm Considers Forgeries]," MS Ahmet III 
624, Topkapi Sarayi, 28b; al-MaqdisI, Shurut, 16; Ibn 'Asakir, Tdrikh madinat Dimashq, 
58:93. 

'" Shams al-Din al-SakhawI, Bughyat al-rdghib al-mutamanni fi khatm al-Nasd'i, ed. Abu 
al-Fadl Ibrahim b. Zakariyya (Cairo: Dar al-Kitab al-MisrI, 1991), 38. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 149 

The notion that a hadith collection can serve as the locus for con- 
sensus and as legal and doctrinal common ground appears even more 
clearly in the work of Ibn Manda's contemporary, al-Khattabi (d. 388/ 
998). In the introduction to his commentary on Abu Dawud's Sunan, 
he states that the collection is: 

a noble book unique in the science of religion . . . approved by all people. 
It has become the ultimate recourse for differences of opinion amongst 
the various sects of the learned and the generations of scholars . . . the 
people of Iraq, Egypt, the lands of the West, and still more from among 
the cities and regions of the Earth, rely upon it.'™ 

Acknowledging the Khurasani cradle of the Sahihayn Network, he notes 
that the scholars of that region preferred those two works and books 
based on their requirements, although he personally considers Abu 
Dawud's Sunan more legally useful. ' '^ Al-Khattabi describes al-Bukharl's 
Sahih in language similar to but less grandiose than his accolades of the 
Sunan, with an emphasis on authenticity as opposed to legal utility: 

It has become a treasure for [our] religion, a mine for [its] sciences. It 
has become, due to the quality of its criticism inaqdihi) and the severity 
of its articulation [sabk), a judge {hakam) in the umma in what is sought 
out from among hadiths as authentic or weak.'"" 

Ibn Manda, Ibn al-Sakan and al-Khattabi provide no extensive or 
concrete explanations for their evaluations of these works as loci of 
consensus in law and hadith. Neither do they articulate their specific 
authority or epistemological yield. What is nonetheless clear, however, is 
that the community of transmission-based legal scholars was beginning 
to see a proto-canon of hadith collections as extant and necessary. 



Why the Sahihayn? 

When examining the mustakhraj and 'ilal/ilzdmdt phenomena, one can- 
not help but ask why these fleeting genres focused so predominantly on 
the Sahihayn. The resilient regional barriers of the first half of the long 
fourth century cannot provide a full explanation for the nature of the 



'"^ Al-Khattabi, Afo a/im a/-iMna«, 1:6. 
'" M-KhsOSbi, Ma'dlim al-sunan, 1:6. 

""' Al-Khattabi, A'ldm al-hadithji shark Sahih al-Bukhdn, ed. Muhammad b. Sa'd Al- 
Su'udi, 4 vols. (Mecca: Mu'assasat Makka li-al-Tiba'a wa al-Flam, [n.d.]), 1:102. 



150 G HAPTER FOUR 

mustakhraj genre, since the Sahihayn were not the only collections used 
as templates even within one region. Muslim's Sahih enjoyed favored 
status in his home city of Naysabur, but the city and its environs also 
saw the production of three mustakhrap based on Abu Dawud's Sunan, 
two on al-Tirmidhl's Jdmi', and one on Ibn Khuzayma's Sahih (with 
Ibn al-Jarud's Muntaqd a possible second). Scholars in Naysabur thus 
could and did see other collections as attractive and available forma- 
tive texts. 

Having exhausted the path of material constraint, we must ulti- 
mately turn to matters of functionalism and scholarly preference. As 
al-Isma'lll's, Ibn 'Uqda's and Abu 'All al-Naysaburi's testimonies prove, 
many scholars of the Sahihayn Network simply felt that a specific work 
was the most accurate and useful presentation of the Prophet's legacy. 
Al-Isma'lli favored al-Bukharl's collection over Muslim's Sahih, Abu 
Dawud's Sunan and the Sunan of al-Hulwani (d. 243/857-8) because 
in his eyes it provided a more authentic selection of hadiths and a bet- 
ter analysis of their legal content. Conversely, Ibn 'Uqda felt Muslim's 
work outshone al-Bukharl's because it was more purely a collection 
of hadiths without the incomplete narrations and commentary added 
for legal elucidation. Al-Isma'ili and Ibn 'Uqda were attracted to the 
differing functional methodologies of al-Bukhari and Muslim, but why 
did Abu 'All al-Naysaburi favor Muslim's work above all others? Such 
matters of scholarly preference lie beyond our ken. 

Certainly, if hadlth scholars of the long fourth century hoped to 
prove the quality of their isnddi, by composing mustakhrajs, it seems logi- 
cal to choose the most rigorous collections as templates. This explains 
why all the template collections were products of the sahih movement 
and not earlier works like Malik's Muwatta'. In fact, the only work one 
might call a mustakhraj of the Muwatta', the Kitdb al-tamhid of Ibn 'Abd 
al-Barr (d. 463/1070), was effectively an attempt to place Malik's work 
on equal footing with other sahih books. Because the Muwatta ' is replete 
with hadiths lacking complete isndds, Ibn 'Abd al-Barr set out to collect 
complete narrations. As Ibn 'Abd al-Barr makes clear in his introduction, 
one of his goals in the Tamhid is to establish Malik's book according to 
the language and requirements of the sahih movement."" 



"" Ibn 'Abd al-Barr, al-Tamhid li-md Ji al-Muwatta' min al-ma'dni wa al-asdnid, ed. 
Mustafa Ahmad al-'AlawI and Muhammad 'Abd al-Kablr al-BakrI, 2nd ed., 26 vols. 
(Ralaat: Wizarat 'Umum al-Awqaf wa al-Shuun al-Islamiyya, 1402/1982), 1:7. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 151 

The nature of the Sahihayn also partly explains why they were the 
only works to prompt 'Hal or ilzamat studies in this period. Al-Bukharl 
and Muslim were two of the only scholars to purpose works devoted 
solely to sahih hadlths. Others such as Abu Dawud and al-Tirmidhi 
acknowledged that they relied on weak or lackluster narrations when 
necessary. Consequently, as al-Khattabi noted, the Sahihayn and the 
notion of their authors' "conditions [shart, rasmj" proved attractive tar- 
gets for study. Only with works that set uniform standards could one 
apply these standards elsewhere. Only with authors who claimed to 
include only authentic material could one object that certain hadlths 
fell short of this measure. 

Even in this matter, however, we cannot escape the aesthetics of 
critical preference. Ibn Khuzayma also sets up a clear requirement 
for authenticity [sihhd) on the first page of his Sahih- But despite the 
arguably unparalleled accolades al-Hakim grants him, al-Hakim found 
Ibn Khuzayma an unsatisfactory judge of authentic reports (j-iMa).'"^ 
Although some scholars like al-Khatib said that Ibn Khuzayma's work 
deserved mention alongside the Sahihayn, his collection never accumu- 
lated critical studies.'"' 



Conclusion: The Eve of Canonization 

Our analysis of the Sahihayn Network of the long fourth century brings 
us to the eve of the canonization of these two texts. Among Mu'tazilites, 
hadith-minded Sunnis like al-Tabari, the hadlth-wary Hanafi theorist 
al-Jassas and even in the realm of Sunni-Shiite polemic, there had 
arisen the idea that hadlths could enjoy the consensus of the umma 
and thus wield tremendous epistemological authority. Among transmis- 
sion-based scholars this concept expressed itself in a proto-canon of 
hadith collections that certain scholars felt provided loci of legal and 
narrative consensus. 



'«2 Al-Khallll, al-Irshdd, 313. 

'"' Al-Khatib al-Baghdadi, al-Jdmi' li-ikhtilaf al~rdwi wa dddb al-sdmi', 2:185. It was 
not until the eighth/fourteenth century that 'Umar b. 'All Ibn Mulaqqin (d. 804/ 1 40 1 ) 
added the men of Ibn Khuzayma to al-Mizzl's ever-expanding biographical dictionary 
of hadith transmitters; TaqI al-Din Muhammad Ibn Fahd al-Makkr, Lah^ al~lihdi, ed. 
Zakariyya 'Umayrat (Beirut: Dar al-Kutub al-'Ilmiyya, 1419/1998), 130. 



152 G HAPTER FOUR 

But how did this period of intense study affect al-Bukhari's and 
Muslim's works? One can best answer this question by referring to 
sahih hadith collections that never attained canonical status. In his 
brief explanation of why Sahih Ibn Hibbdn did not become one of the 
famous Six Books, the Azhar scholar Muhammad al-Qj'i states curtly 
that Ibn Hibban (d. 354/965) narrated from unknown transmitters 
(jnajdhil)}^* This negative evaluation of Ibn Hibban's work originated 
as early as the writings of his own student, al-Hakim al-Naysaburi. "*' 
Yet as our review of transmitter studies has shown, the earliest work on 
al-Bukhari's teachers freely admits that at least one of his sources in the 
Sahih was also unknown. It was only after another two generations of 
study that al-Kalabadhi discovered the identity of this transmitter. Ibn 
Hibban died almost a century after al-Bukhari and lived in an era that 
he himself bemoaned as a sad time, when people no longer wrote sahih 
books.""" Had his Sahih received the generations of scholarly attention 
devoted to the Sahihayn during the long fourth century, it too might 
have been purged of unknown transmitters, in which case al-Hakim 
would have read it with glowing approval. Indeed, later scholars such 
as Ibn Taymiyya (d. 728/1328), Ibn Kathir (d. 774/1373) and Zayn 
al-Din al-'Iraqi (d. 806/1404) did champion Ibn Hibban's work as an 
exceptional source for authentic hadith."" As we will see in the next 
chapter, they were simply too late. 

Conversely, the extraordinary efforts of the Sahihayn Network schol- 
ars to produce definitive texts of al-Bukhari's collection and identify 
his methods and transmitters made the work an ideal candidate for 
canonization. As we shall see in the next chapter, it was claims about 
al-Bukhari's and Muslim's methods and transmitters that lay at the 
center of the case for their authority. 

We must now also ask: How did this "period of intense canonical 
process" involve the community shaping and appreciating these texts in 
ways that made them "most meaningful and valuable?"""* A number of 
scholars in the long fourth century immediately seized on the SaMhoyn as 



"" Muhammad al-Qi'i, Qanun al-Jikr al-islami (Cairo: Dar al-Basa'ir, 1424/2004), 
145. 

'"■' See al-San'am, Tawdih al-afkar, 1:66; cf. al-Sakhawi, Fath al-mughith, 1:56. 

'"« Ibn Hibban, Sahih Ibn Hibbdn, 1:58. 

"" Ibn Taymiyya, Majmu fatdwd, 1:256; Ibn Kathir, al-Bd'ith al-hathith, 23; al-'iraqi, 
al-Taqyid wa al-iddh, 30; Ibn al-Salah, Muqaddima, 164-5. 

""' Sanders, 30. 



A 'PERIOD OF INTENSE CANONICAL PROCESS' 153 

formative texts for engaging the Prophetic legacy and expressing their 
relationship with it. Their interest spawned the period's concentrated 
studies of the two works. It was not, however, the need that drove the 
mustakhraj genre that would result in the canonization of al-Bukhari 
and Muslim. Expressing one's relationship to the Prophet's legacy and 
interpreting his teachings through living isnad's, remained the unique 
obsession of hadlth scholars. The canonization of the Sahihayn would 
have to involve a broader Muslim community. 

It would be the ilzdmdt genre, which extended al-Bukhari's and 
Muslim's standards for authenticity to new hadiths, that proved crucial. 
It was the standards of the two scholars that served as the measure of 
truth in which the authority of the lawmaker could be deposited and 
then extended into new territory. It is no surprise that the one scholar 
of the long fourth century to have dealt exclusively with the standards 
of the Shaykhayn is the one scholar we have conspicuously avoided until 
now. He is the focal point of the Sahihayn Network to whom all roads 
lead. Prior to al-Hakim al-Naysaburl's seminal career, the nexus of 
canonicity — that of text, authority and communal identification — had 
not yet coalesced. Transmitters like Ibn al-Sakan, Abu Dharr al-Harawi 
and the various scholars who produced studies of the Sahihcyn in effect 
succeeded in producing definitive, accessible texts of the two works. But 
the Sahihayn were not authoritative even for their local mustakhraj cults. 
Unlike many post-canonization critics, al-Isma'ili, Ibn 'Ammar and 
al-Daraqutni included no word of apology or explanation for criticiz- 
ing the two works. Before al-Hakim the Sahihayn were simply tools and 
objects of interest for local communities of transmission-based scholars. 
After him, the canon had formed. 



CHAPTER FIVE 

CANON AND COMMUNITY: 

AL-HAKIM AL-NAYSABtJRl AND THE CANONIZATION 

OF THE SAHIHATN 



Introduction 

Around the end of the fourth/tenth century, the Sahih collections of 
al-Bukharl and Muslim first emerged as kandns of authenticity. Rep- 
resentatives from the two divergent strains of the transmission-based 
school, the Hanbali/iiber-Sunnis and the nascent Shafi'i/Ash'ari camp, 
agreed on the Sahihayn as common references for the Prophet's authen- 
tic legacy The study and exploration of the Sahihayn took place at the 
hands of a network of devoted hadlth scholars, but the canonization 
of the two works would result from the activities of a different cadre. 
Al-Hakim al-Naysaburl was the common link. He both inherited and 
participated in the study of al-Bukharfs and Muslim's collections, yet he 
employed the ilzdmdt genre for a new ideological purpose. Al-Hakim's 
vision of the critical standards that the two scholars had followed in 
compiling their works was designed to meet the demands of both 
Sunni hadith scholars and the hadith-wary Mu'tazilites who rivaled 
them. Al-Hakim used the "standards of al-Bukharl and Muslim" as a 
measure of authenticity to extend this common requirement to a vast 
new body of hadlths. 

In the long fourth century, the broader Muslim community developed 
a new vision of the authority that Prophetic hadlths could attain when 
validated by communal consensus. By the mid-fifth/ eleventh century, 
this leap had led legal theorists from the Hanafi, Maliki, Mu'tazilite, 
Hanbali and Shafi'i/Ash'ari schools to a common belief that hadlths 
accepted by the umma yielded epistemological certainty. It was this 
principle that two of al-Hakim's close associates, one from the budding 
Shafi'i/Ash'ari tradition and the other from the Hanbali/iiber-Sunni 
school, would use to declare the Sahihayn a common body of authentic 
hadlths agreed on by these two vying groups. 



CANON AND COMMUNITY 155 

The Life and Works of al-Hdkim al-Naysdbun 

Abu 'Abdallah Muhammad b. 'Abdallah al-Hakim al-Naysaburl was 
born in 321/933 in Naysabur and began studying hadlth at the age 
of nine. He studied extensively with over two thousand teachers in 
Kufa, Rayy Baghdad, Abadan, Hamadhan, Merv, Transoxiana, and 
his native Naysabur' His primary mentors in the sciences of hadlth 
collection and criticism were three major members of the Sahihayn Net- 
work: Abu 'All al-Naysaburl, Abu Ahmad al-Hakim and al-Daraqutni, 
as well as Muhammad b. 'Umar Ibn al-Ji'abi (d. 355/966).^ Al-Hakim 
traveled twice to Baghdad for his studies, once as a youth and again 
in 368/978-9.' Throughout his career he and his Baghdad teacher al- 
Daraqutnl had an uneasy and tense relationship. Al-Hakim's student 
al-Khalill mentions that his teacher sat and discussed {ndzara) hadlth 
with al-Daraqutni and that the latter was pleased with the student from 
Naysabur* In another report, however, it is said that when al-Hakim 
arrived in Baghdad he asked to see al-Daraqutni's collection of hadlths 
from a certain shaykh. When the young scholar looked at the first hadlth 
and saw it was from a transmitter whom he considered weak, he threw 
down the papers and never looked at them again. ^ As we shall see, al- 
Hakim and al-Daraqutni would remain in an ongoing correspondence 
characterized by serious disagreements over the nature of al-Bukhari's 
and Muslim's methods. 

In Naysabur's rigid division between the Hanafi school and the 
transmission-based scholars, al-Hakim adhered firmly to the latter's 
moderate Shafi'i strain. He studied the Shafi'i tradition with Abu Sahl 
al-Su luki (d. 369/980) as well as others and even composed a book on 
the virtues of the school's eponymous founder [Fadd'il al-Shdfi'i).'' He 
complained about the way in which the Hanafi Muhammad b. Sa'ld 
al-Bawraqi used to forge hadlths for that school, such as the following 
report attributed to the Prophet: "There will be in my umma a man 
named Abu Hanifa, and he will be its lamp . . . and there will be in my 



' Al-Dhahabi, Siyar, 17:163. 
2 Al-Dhahabi, Svyar, 17:165. 
' Al-Khallll, al-Irshdd, 324. 

' Al-Khallll, al-Irshdd, 324, Al-Subkl frankly admits that al-Hakim and al-Daraqutnl 
were often at odds; al-Subkl, Tabaqdt, 4:164. 

^ Al-Khatlb, Tdrikh Baghddd, 3:94. Al-Khatib adds, "Or so he said [aw kamd qdt)." 
'' Cf al-Dhahabi, Tadhkirat al-huffd^, 3:164; al-Subkl, Tabaqdt, 4:156. 



156 CHAPTER FIVE 

umma a man named Muhammad b. Idris [al-Shafi'i] whose strife {fitna) 
is more harmful than that of Satan (Iblis)."'' 

Like many participants in the early Shafi'i tradition, al-Hakim cul- 
tivated relationships with practitioners of dialectical theology In fact, 
he studied extensively with two of the architects of the Ash'ari school. 
He attended the lessons of Ibn Furak (d. 406/1015), who held him in 
high regard, and also produced a sizable selection [intakhaba 'alayhi) of 
hadiths from the famous Shafi'l jurist, legal theoretician and theologian 
Abu Ishaq al-Isfarayini (d. 418/1027).'' 

Al-Hakim eventually became a leading member of the hadith scholar 
community in Naysabur Not only was he sought out for opinions on 
the authenticity of hadiths and the reliability of narrators, he also exer- 
cised a great deal of authority in the community. One of al-Hakim's 
main teachers assigned him as the supervisor for his pious endowment 
[waqf) and charged him with running a small hadith school called 
Dar al-Sunna.^ Al-Hakim towered over the multitudes of students 
who flocked to the city to study the Prophet's legacy. The famous Sufi 
exegete, Muhammad b. al-Husayn al-SulamI (d. 412/1021), who was 
accused of forging hadiths for the Sufi cause, had heard a small number 
of hadiths from the great Naysabur muhaddith Abu al-'Abbas al-Asamm 
(d. 346/957). After al-Hakim's firm oversight had ended with his death 
in 405/1014 at the age of eighty-four, however, al-Sulami felt free to 
exaggerate dramatically to students the amount of material he had 
heard from al-Asamm.'" 

Al-Hakim's interest in hadith dominated his oeuvre. Aside from his 
book on al-Shafi'i, a contribution to the Proofs of Prophecy [Dald'U al- 
nubuwwa) genre, and his landmark biographical dictionary of Naysabur, 
al-Hakim's works revolved around the science of hadith criticism. Well 
before he reached the age of seventy he had written a selection of one 
hadith from each of his teachers [mu'jam al-shuyukK), a book of 'Hal, and 
a hadith work called Kitdb al-iklTl about the Prophet's campaigns for the 
local military governor [Sahib al-jaysh)}^ Much more important, however, 
was the introduction to that work, which served to familiarize the lay 



' Al-Khatib, Tdnkh Baghdad, 2:379. 

* Cf. al-Subkl, Tabaqdt, 4:162; al-Dhahabi, Tadhkirat al-huffd^, 3:164; idem, Tdrikh 
al-isldm, 28:438. 

' 'Abd al-Ghafir al-FarisI, Tdnkh Naysabur al-muntakhab min al-Siydq, 6. 
'° Al-Khatib, Tdrikh Baghddd, 2:245. 
" Al-Khailll, al-Irshdd, 325. 



CANON AND COMMUNITY 157 

reader with the types of authentic and defective (saqim) reports as well 
as the levels of narrator criticism.'- He also wrote an introduction to 
his treatments of al-Bukharl's and Muslim's works, called al-Madkhal 
ild al-Sahih (or al-Sahihayn), in which the author gives a tantalizing 
indication of his vision of the Shaykhayn's, criteria and their range of 
acceptable narrators. In addition, he states that he wrote one book on 
each of al-Bukhari's and Muslim's criteria for authenticity as well as 
a work on those reports that one of the two scholars had included to 
the exclusion of the other. '^ 

Probably around the age of sixty-five, al-Hakim penned his famous 
and comprehensive treatise on the sciences of hadlth, the Ma'rifat 'ulum 
al-hadith (Knowledge of the Sciences of Hadlth). Divided into fifty-two 
chapters, this book discusses the technical terms used in hadlth criticism 
and transmission, lists the different generations of transmitters, gives 
brief biographies of major hadlth scholars and outlines material essen- 
tial for a hadlth student. Al-Hakim's opinions and the chapter structure 
of his Ma'rifa would exercise tremendous influence on the genre of 
hadlth's technical discipline [mustalah al-hadith) for centuries.'* 

The work with which we are most concerned in this chapter was 
evidently one of the last al-Hakim composed: a voluminous ilzdmdt of 
the Sahihayn entitled al-Mustadrak. This work differed both qualitatively 
and quantitatively from the ilzdmdt works of al-Hakim's teacher al- 
Daraqutnl and his student Abu Dharr al-Harawi. Unlike al-Daraqutnl's 
diminutive Mtdb al-ilzdmdt, which consists of only one hundred and 
nine hadiths, and Abu Dharr al-HarawT's lost Mustadrak, which was 
only one volume, al-Hakim's Mustadrak is a multivolume work. Unlike 



' ^ Al-Hakim, al-Madkhal ild ma 'rifat kitdb al-Iklil, 5 1 . We know al-Hakim had composed 
the Mil, its introduction, his Madkhal ild al-Sahih and his Muzakki al-akhbdr well before 
389 AH, because we know his Ma'rifat 'ulUm al-hadith was being transmitted widely as 
early as that date, and in that work the author refers the reader to the aforementioned 
books; al-Subkl, Tabaqdt, 4:157; al-Dhahabi, Tadhkirat al-hujfdz, 3:162. 

" This last work was titled Ma infarada kull wdhid min al-imdmayn bi-ikhrdjihi. For lists 
of al-Hakim's oeuvre, see Ibn al-Salah, Tabaqdt, 1:199-200; al-Dhahabi, Siyar, 17:170; 
al-Hakim, Tdrikh Mshdbur, 38-42 (editor's introduction); al-Subkl, Tabaqdt, 4:156. Al- 
Hakim had other small books on legal matters, such as a work called Kayfiyyat saldt 
al-duhd (How to Pray the Late Morning Prayer), a work called Fard'id al-fawdid and 
a forty hadlth collection, which was widely studied in Qazvin; al-Rafi'l, al-Tadwin ji 
akhbdr Qazwin; 1:337, 341, 346; 2:45, 58. 

" Ibn al-Salah's famous Muqaddima, for example, is influenced by the chapter 
structure of the Ma'rifa, to the extent that Ibn al-Salah included a certain chapter (on 
afrdd) which he felt was covered elsewhere simply because al-Hakim had a chapter on 
it; al-'iraqi, al-Taqyid wa al-iddh, 95. 



158 CHAPTER FIVE 

al-Daraqutni's random and incidental collection of hadiths, the Mus- 
tadrak is organized topically in musannaf iovva}" 

Al-Hakim's works on the technical discipline of hadlth study were 
widely read even during his own lifetime, and several scholars responded 
to his work. His student al-KhaMi notes that al-Hakim was sometimes 
not sufficiently discriminating or clear in his writings. The criticisms 
of his colleagues thus led him to review and clarify his work."' 'Abd 
al-Ghani b. Sa'id of Egypt (d. 409/1019), for example, wrote to al- 
Hakim with some criticisms of his al-Madkhal ild al-Sahih, for which 
al-Hakim thanked him. ' ' Farther west than Egypt, we know that even 
within the author's lifetime (by 389/998-9) some hadith scholars in 
Andalusia possessed copies of his Ma'rifa}^ Al-Hakim was well-known 
enough in the region within several decades of his death for Ibn Hazm 
(d. 456/1064), who never left Andalusia, to prominently note his opin- 
ion in the debate over who was the most virtuous of the Prophet's 
Companions.'^ In the Islamic heartlands of Iraq and Iran, al-Hakim's 
student Abu Nu'aym al-Isbahani had a copy of his Tdnkh Naysdbur, his 
Madkhal ild al-Sahih and probably many of his other books. ^^ Although 
al-Khatib al-Baghdadi never met al-Hakim, he relies on information 
and reports from him extensively through a myriad of intermediaries 
in his Tdnkh Baghddd}^ 

Yet al-Hakim's adherence to the moderate Shafi'i tradition and some 
of his interpretive choices in his Mustadrak precipitated a clash with 



'' The Cairo edition of the Mustadrak fills five volumes; al-Hakim, al-Mustadrak 'aid 
al-Sahihayn, ed. Muqbil b. Hadl al-Wadi'l, 5 vols. (Cairo: Dar al-Haramayn, 141 7/ 1997). 
See also Brown, "Criticism of the Proto-Hadith Canon," 11. The Mustadrak has fewer 
chapters (47) than al-Bukharl's or Muslim's Sahihi, but seems to be inspired by both 
works' ordering. Only 3 chapters appear in the Mustadrak that do not appear in either 
of the Sahihayn {kitdb al-hijra, kitdb qism al-fay ' and kitdb tawdrikh al-mutaqaddimin min al- 
anbiyd'). 

"• Al-Khallll, al-Irshdd, 324. 

" Al-Dhahabi, Tadhkirat al-huffai, 3: 168; this work has survived in manuscript form, 
entitled "Bayan awham al-Hakim ft al-Madkhal," MS Ahmet III 624, Topkapi Sarayi, 
Istanbul: fols. 200a-206a. 

'" Al-Dhahabi, Siyar, 17:165-6. 

'^ Abo Muhammad 'All Ibn Hazm al-Zahirl, Kitdb al-fisalfi al-milal wa al-ahwd' wa 
al-nihal, 5 vols, in 2 (Baghdad: Maktabat al-Muthanna, [1964]), 4:1 1 1. Ibn Hazm notes 
that al-Hakim upheld the unusual position that 'Umar b. al-Khattab was the foremost 
Companion of the Prophet. Considering the controversy over al-Hakim's supposedly 
Shiite views (see below), however, such a report was most likely a product of polemics 
surrounding his position. 

^" See, for example, al-Khatib, Tdnkh Baghddd, 2:73. See also n. 98 below. 

2' See, for example, al-Khatib, Tdnkh Baghddd, 2:438; 10:147; 11:385. 



CANON AND COMMUNITY 159 

more conservative members of the transmission-based community. 
Specifically, al-Hakim's statement that two pro-Alid hadiths known as 
the hadlth al-Tayr"^- and the hadlth of Ghadir Uiumrn^^ met the require- 
ments of al-Bukharl and Muslim led certain hadlth scholars to accuse 
him of Shiism. These accusations are well documented; writing not 
long after al-Hakim's death, al-Khatib al-Baghdadi notes several reports 
about the hadith-al-Tayr incident and al-Hakim leaning towards Shi- 
ism.^* Al-Hakim's student al-Khalili alludes to the accusations leveled 
against his teacher when he writes, "For me he was an ocean, and all 
that was hurled at him could not detract from that [ra'aytuhu fi hull ma 
ulqiya 'alayhi bahr"" layu'jizuhu 'anhu).""^-' More extreme reports have also 
survived, such as stories that hadlth scholars blockaded al-Hakim in 
his house and that he disliked Mu'awiya so much that he could not 
bring himself to narrate a hadlth praising him in order to placate his 
opponents. Such reports, however, appear only in later sources compUed 
by al-Hakim's critics, such as Ibn al-Jawzi's Muntazam}'^ 

This accusation of Shiism was probably baseless, like the scandal that 
had earlier tarnished al-Bukharl's reputation. Both he and al-Hakim 
were attacked by extreme members of the transmission-based school 
for their more moderate stances. Al-Hakim's most vocal critics were 
all prominent iiber-Sunnis: the Hanball Kh"aje 'Abdallah al-Ansarl 
(d. 481/1089), Muhammad b. Tahir al-Maqdisi (d. 507/1 1 13) and Ibn 
al-jawzl.^' Much like al-Shafi'i himself, al-Hakim's Shafi'i identity led to 



^^ In this hadlth the Prophet is eating a fowl and calls on God to "bring me the 
most beloved of your creation, {kuntu akhdamu Rasul Allah [s] fa-quddima li-Rasul Allah 
{s} farakh mashwi..)" at which point 'All enters and eats with the Prophet. See Jdmi' 
al-Tirmidhi: kitdb al-mandqib, bdb mandqib Alt. 

^' In this hadlth the Prophet says, "Whoever's master I am, 'All is his master {man kuntu 
mawldhu fa-All mawldhu)r See AbQ 'Abdallah Muhammad Ibn al-Najjar (d. 643/1246), 
al-Radd 'aid Abi Bakr al-Khatib al-Baghdddi, 129; al-Dhahabi, Siyar, 17:168. For these 
hadiths, see al-Hakim, al-Mustadrak: kitdb ma'rifat al-sahdba, bdb ba'dfadd'il All. 

^' Al-Khatlb, Tdnkh Baghddd, 3:94; cf Ibn al-Jawzi, al-Muntazam, 15:109; Abu Tahir 
Ahmad b. Muhammad al-Silafi (d. 576/1180), Mu'jam al-safar, ed. 'Abdallah 'Umar 
al-Barudi (Beirut: Dar al-Fikr, 1414/1993), 99. 

^^ Al-Khallll, al-Irshdd, 325. The editor of this text vowels the word 'yu'jizhu,^ which 
I think is incorrect. 

^'' Ibn al-jawzl, al-Munta^am, 15:109—10. In addition, there is a record of al-Hakim 
narrating a pro-Umayyad report in which Mu'awiya performs the pilgrimage and urges 
people to give their oath of loyalty to Yazld; al-Husayn b. Ibrahim al-jawzaqani (d. 
543/ 1 148-9), al-Abdtil wa al-mandkir iva al-sihdh iva al-mashdhir, ed. Muhammad Hasan 
Muhammad (Beirut: Dar al-Kutub al-'Ilmiyya, 1422/2001), 142. 

^' See al-Dhahabi, Siyar, 17:174—5; Ibn Hajar, Lisdn al-mizdn, 5:233; Ibn al-jawzl, 
al-Munta^am, 15:110. 



1 60 CHAPTER FIVE 

accusations of Shiism. Al-Shafi'l had based his legislation on issues of 
rebellion (al-bughdt) on the premise that 'All had dealt righteously and 
appropriately with Mu'awiya's uprising against the caliphate. Combined 
with his affection for the family of the Prophet, such thinking led to a 
trial before the Abbasid caliph in which al-Shafi'l had to defend himself 
against accusations of Shiism.^" Al-Hakim upheld this Shafi'l position, 
quoting the great Shafi'l Ibn Khuzayma as saying that anyone who 
fought 'All on the issue of the caliphate was a rebel [bagk")}'^ 

The furor caused by al-Hakim's approval of the two pro-Alid hadiths 
also seems to have been accidental. The hadiths themselves had been 
verified by earlier Sunni scholars such as al-Nasa'l and al-Tirmidhi. In 
al-Hakim's time, however, the reports had become anathema to certain 
elements of the hadith community. Whereas al-Nasa'l was only vaguely 
criticized for not praising Mu'awiya sufficiently, when a contemporary of 
al-Hakim, Ibn al-Saqqa' (d. 371/981—2), narrated the hadith al-Tayr'va 
a mosque he was expelled, confined to his house, and the place where 
he sat in the mosque washed clean.'" It thus seems probable that the 
accusations of Shiism resulted from al-Hakim's Shafi'i approval of 
'All's position against Mu'awiya and his authentication of two hadiths 
that had become touchstones for anti-Shiite sentiment among the ahl 
al-hadith. 



Al-Bukhan and Muslim in al-Hakim's Vision of Hadith 

As the Sahihayn Network Chart in the previous chapter demonstrates, 
al-Hakim acted as a magnet for studies of al-Bukharl's and Muslim's 



^" AI-Dhahabi, Ma'rifat al-ruwat al-mutakallam fihim bima la yujibu al-radd, ed. Abu 
'AbdaUali Ibrahim Sa'iday (Beirut: Dar al-Ma'rifa, 1406/1986), 49-50; cf. Abu Zahra, 
al-ShafiX 22-3. 

^^ This is based on the famous hadith in which the Prophet tells 'Ammar b. Yasir 
that he will be killed by the rebellious party (i.e., Mu'awiya); al-Hakim, Ma'rifat 'utum 
al-hadith, 105. 

'" Al-Dhahabi, Tadhk,irat al-liuffa^, 3:1 17. For the accusations of al-Nasa'l, see ibid., 
2:194-5; al-San'anl, Tawdih al-qflidr, 1:199. That these two pro-Alid hadiths were par- 
ticularly controversial in al-Hakim's time is also evidedent from the fact that scholars 
of this period devoted specific treatises to these reports. Abu al-'Abbas Ibn 'Uqda (d. 
332/944) wrote a work on the hadith of Ghadir Khumm, and al-Tabarl (d. 310/923) 
and al-Hakim's student Ahmad b. Hamdan (d. ca. 440/1048-9) wrote works on the 
hadith of al-Tayr; al-Dhahabi, Tadtikirat al-liuffd^, 3:206; cf. Ahmad al-Ghumarl, Fath 
al-matilc al-'ati bi-sihhat hadith bdb madinat al-'ilm 'Alt, ed. 'Imad Surur ([n.p.]: [n.p.], 
1426/2005), 11-12. 



CANON AND COMMUNITY 161 

work. Like his teacher, al-Daraqutni, al-Hakim's scholarly activities 
revolved around the Sahihayn and the methods of their authors. Unlike 
earlier scholars such as al-Isma'lll, however, al-Hakim's appreciation for 
the Sahihayn did not involve their legal merits. For al-Hakim, al-Bukhari 
and Muslim represented the pinnacle of skill and achievement in the 
realm of hadlth criticism in particular. He writes in his al-Madkhal ila 
al-Iklil, "All regions testify to the superiority of Khurasan in the knowl- 
edge of authentic hadiths . . . due to the precedence of the two imams, 
Abu 'Abdallah al-Bukharl and Abu al-Husayn [Muslim] al-Naysaburi, 
and their lone mastery itafarrudihima) of that science."^' Unlike the 
other members of the Sahthqyn Network who viewed the works only 
as formative texts or objects of study, al-Hakim endowed them with a 
loftier station. Al-Bukharl's and Muslim's books embodied the highest 
level of critical stringency, and for him they were key pillars of the 
science of hadith criticism itself In the Ma'rifa's, chapter on authentic 
hadiths, al-Hakim begins with a description of reports that seem to 
have authentic isndds but in fact possess fatal weaknesses perceptible 
only to master critics. He concludes that if a hadlth does not have 
an isndd found in one of the Sahihayn, one must subject it to thorough 
examination for such hidden flaws {'ilia pi. 'Hal):'- Inclusion in one or 
both of the Scihihayn thus tremendously bolsters the credibility of a 
narrator or his reports. In al-Hakim's chapter on how hadith scholars 
have treated narrators with non-Sunni beliefs, he uses the Sahihayn to 
demonstrate that mild heretics are acceptable sources. Aban b. Taghlib 
(d. 140—1/757—9), for example, was a known Shiite who once narrated 
a hadlth attacking the caliph 'Uthman. But al-Hakim states that he is 
nonetheless "trustworthy, with his hadiths included in the Sdhihoyn." 
Despite Malik's rejection of Ibrahim b. Tahman (d. 168/784) for being 
a Murji'ite, al-Hakim defends him in the same manner. ^^ 

Al-Hakim did not, however, consider al-Bukhari's and Muslim's col- 
lections infallible. He himself criticizes some of Muslim's selections. He 
mentions a narration of the famous hadlth in which the Prophet states 
that the best generations are the first three generations of Muslims, 



" Al-Hakim, al-Madkhal ila ma'rifat kitdb al-Iklil, 72. 

'^ Al-Hakim, Ma'rifat 'ulUm al-hadith, lb. 

'' Al-Hakim, Ma'rifat 'ulUm al-hadith, 168-9. Al-Hakim lists Ibrahim as a one of the 
famous trustworthy imams of his generation; ibid., 308. Al-Hakim himself states that 
one has to be a proselytizer of heresy to be placed outside the pale of 'addla; al-Hakim, 
Ma'rifat 'uldm al-hadith, 67. 



162 CHAPTER FIVE 

adding, "That hadith is included in the Sahih of Muslim b. al-Hajjaj, 
but it has a remarkable flaw {'ilia 'ajiba)."-'* Such critiques come as no 
surprise, since al-Hakim did not feel that al-Bukhari and Muslim had 
designed their works to be totally free of error In the introduction to 
his Mustadrak, he states that his work will consist of hadlths meeting al- 
Bukharl's and Muslim's standards but that "it is not possible to include 
[only] what has no flaws, for indeed they [al-Bukharl and Muslim] 

did not even claim this for themselves "■'■' Here we see the first of 

several inconsistencies in al-Hakim's methodology If the Sahihayn are 
secure sources whose isndd's, require little critical attention, how can he 
so readily admit that they contain flawed reports? We will be better 
able to solve this riddle once we have addressed al-Hakim's purpose in 
employing the standards of al-Bukhari and Muslim. 



The Shurut According to al-Hakim: The Requirements of al-Bukhan 

and Muslim 

Although scholars such as Abu Mas'ud al-Dimashqi and al-Daraqutni 
regularly refer to the standards (shart/shurut/rasm) of al-Bukhari or 
Muslim in their extant works, al-Hakim seems to be the only scholar 
of the long fourth century to have devoted specific treatises to this 
subject. These works have unfortunately been lost, but it appears that 
they did not succeed in clearly explaining al-Hakim's school of thought 
on the topic. The scholar's ambiguous and inconsistent writings on 
the requirements for sahih hadiths in general and al-Bukhari's and 
Muslim's methodologies in particular have confounded hadith experts 
from al-Hakim's time to the present day^'' It is therefore necessary to 
establish the most accurate understanding of al-Hakim's stance, which 
has generally been interpreted in one of three ways. First, al-Hakim's 
writings have led many scholars to believe that he considered the 
elimination of unknown transmitters from the isndd of a hadith to be 
essential for its inclusion in both the general category of sahih and in 



'* Al-Hakim, Ma'rifat 'ulum al-hadith, 52; cf. al-Daraqutm, Kitab al-ilzamat iva al- 
tatabbu', 501-2. 

'' Al-Hakim, al-Mustadrak, 1:39. 

''^ One of the more recent attempts to grasp al-Hakim's definition of the shurut comes 
fi-om Muhammad Abd al-Hayy al-LaknawI. See his Z'^far al-amdni, ed. TaqI al-Din 
al-Nadawi (United Arab Emirates: Dar al-Qalam, 1415/1995), 69-71. 



CANON AND COMMUNITY 163 

the Sahihayn. Other scholars have interpreted al-Hakim's vision of al- 
Bukhari's and Muslim's standards as requiring what we will define as 
'doubling transmission.' Finally, the third and most accurate camp has 
understood that al-Hakim intended both of the above meanings in his 
definition of the Shaykhayn's conditions. 

a. Two Rawls and the Elimination o/" Jahala 

The first interpretation of al-Hakim's writings on the requirements of 
al-Bukhari and Muslim centers on the qualities of the transmitters they 
employed. The notion that a narrator needed to be well-established as 
a transmitter in order to form part of a sahih isndd exerted a tremen- 
dous influence among hadlth scholars. The presence of an unknown 
transmitter in a report's isndd was one of the foremost obstacles to 
its achieving a sahih rating.^' By the time of al-Khatib al-Baghdadi 
(d. 463/1071), Sunni scholars had agreed almost unanimously that 
a person needed at least two established narrators [rdwi) transmitting 
from him in order to avoid being condemned as "unknown (majhtil)."^^' 
The first explicit formulation of this principle is usually attributed to 
al-Bukharl's great adversary al-Dhuhll.^^ This concept, however, was 
clearly already applied in practice during al-Dhuhll's time. Muslim 
had dedicated an entire work to listing transmitters who only had one 
transmitter {rdwi) from them, thus falling short of the requirements 
necessary for a sahih isndd. Al-Nasa'l (d. 303/915) also composed a short 
work on this subject, and al-Hakim himself devoted a chapter to it in 
his Ma 'rifat 'ulum al-hadith. The opposite of unknown transmitters were 
"well-known (mashhdr)" ones whose testimony and transmission could 
validate those of others.*" 



" For a discussion of this, see Ibn al-WazIr, Tanqih al-an^dr, 102. 

'" Al-Khatib, al-Kifaya, 1:290. Later scholars such as Ibn 'Abd al-Barr and AbQ al- 
Hasan b. al-Qattan al-FasI (d. 628/1230-1) attempted to qualify this generally consistent 
rule. For a discussion of such attempts, see Ibn al-WazIr, Tanqih al-an^dr, 192-198; Ibn 
al-Salah, Muqaddima, 296; al-'iraqi, al-Taqyid wa al-iddh, 117-8; al-LaknawI, al-Raf iva 
al-takrml fi al-jarh wa al-ta'dil, ed. 'Abd al-Fattah Abu Ghudda, 8th ed. (Beirut: Dar 
al-Basha'ir al-Islamiyya, 1425/2004), 256-60. Al-Shafi'l (d. 204/819-20) himself is 
attributed with the quote that one cannot accept the narration of an unknown; al- 
Bayhaql, Ma'nfat al-sunan wa al-dthdr, 1:75, 81. 

'" See al-Khapb, al-Kifdya, 1:290; Ibn Rajab, Shark 'Hal al-Tirmidhi, 1:82. 

'" See Ahmad b. Shu'ayb al-Nasa'l, Thaldth rasd'il hadithiyya, ed. Mashhur Hasan 
Malimud Salman and 'Abd al-KarIm Ahmad al-Warlkat (al-Zarqa', Jordan: Maktabat 
al-Manar, 1408/1987), 27-50; al-Hakim, Afa 'n/af 'ulum al-hadith, 195-200. The technical 



1 64 CHAPTER FIVE 

Al-Hakim's work leaves little doubt that he intended the elimination 
of anonymity to be an essential feature of a sahih hadith as well as a 
requirement of al-Bukhari and Muslim. In the Madkhal ild al-Iklil, al- 
Hakim describes ten levels of sahih hadiths. He notes how the first five 
levels are agreed on by all and are found in the collections of estab- 
lished experts used as proof texts {kutub al-a'imma al-muhtajj biha).*^ The 
bottom five levels, on the other hand, faU to meet the requirements 
for authenticity of certain schools of thought. The highest level of 
sahih, he explains, consists of reports narrated by a Companion whose 
identity and reputation as a narrator of hadiths has been established. 
This occurs, al-Hakim elaborates, when one proves that two known 
Successors have narrated hadiths from that Companion, thus freeing 
him of "anonymity {jahdld)." This report is then narrated from that 
Companion by a Successor who is equally well established as a transmit- 
ter The same follows for the ensuing generations until al-Bukhari's and 
Muslim's teachers. As this last clause suggests, al-Hakim concludes by 
stating that this is the level of hadiths found in the Sahihayn, and that 
their number does not exceed ten thousand.*^ Al-Hakim then proceeds 
to define the other levels of authentic hadiths, which do not include 
those featured in the Sahihayn}"^ 

In the Ma 'nfat 'ulum al-hadith, written long after the Madkhal ild al-IklU, 
al-Hakim provides only one definition for sahth hadiths. Abandoning 
the multiple levels of authentic narrations, he restates his definition 
of the highest level: a sahih hadith is narrated from the Prophet by a 
Companion freed of anonymity by having two upright Successors (tdbi' 
ddil) who generally transmit from him. The hadith is then accepted 
and transmitted widely among [yataddwaluhu . . . bi'l-qubul) scholars from 
that point on. He likens this mass transmission to continuous levels of 
testimony by witnesses in court [shahddd).** Invoking this analogy between 



term mashhur was already in use during the first half of the third/ninth century and 
appears in Muslim's writings; Muslim, al-Munfariddt wa al-wahddn, 88. 

*' Al-Hakim, al-Madkhal ild ma'rifat kitdb al-Iklil, 107. 

'^ Al-Hakim, al-Madkhal ild ma'rifat kitdb al-Iklil; Ti, 78. Scholars like al-LaknawI have 
admitted that this passage and the following description oi sihha from the Manfa could 
support the notion of doubling transmission. See al-LaknawI, ^a/ar al-amdm, 69-7 1 . 

" Again falling into inconsistency, al-Hakim notes that al-Bukharl and Muslim 
include one narration each that belongs in the fourth level of universally accepted 
hadiths; see James Robson, trans., An Introduction to the Science of Tradition (London: 
Luzac and Co., 1953), 19. 

** Al-Hakim, Ma'rifat 'ulum al-hadith, 77. 



CANON AND COMMUNITY 165 

bearing witness and transmitting hadlths on the topic of eliminating 
anonymity was odd for a Sunni muhaddith, although it was especially 
common among Mu'tazilites.*' The reason for this bizarre comment 
will became clear when we discuss al-Hakim's target audience. 

Support for this interpretation of al-Hakim's vision of the SahThaynh 
criteria comes from one of his senior students, Abu Bakr al-Bayhaql. 
He held that al-Bukhari and Muslim demanded that each narrator in 
the isndd have the two transmitters required to eliminate anonymity 
Although one would expect that this close student of al-Hakim would 
have provided more productive insights into his school of thought, al- 
Bayhaql's comments are frustratingly brief In his al-Sunan al-kubrd he 
states definitely that al-Bukhari and Muslim did not narrate reports from 
a Companion or Successor with only one transmitter from him. For 
this reason, they did not include hadiths from one Mu'awiya b. Hida 
because only one person ever narrated material from him.*'' Another 
scholar very familiar with both al-Hakim's works and the Sahihayn, 
Abu 'All al-Jayyani al-Ghassani of Andalusia (d. 498/1105), states 
that Hakim's definition of sahih aimed at the elimination of majhuh. 
He therefore required each Companion and Successor to have two 
narrators establishing him as a viable transmitter.*' 

This definition of al-Bukhari's and Muslim's criteria and the require- 
ments for authentic hadiths in general, however, was very controversial. 
Even during his lifetime, al-Hakim's colleagues attempted to correct his 
understanding. In fact, in his Mustadrak, al-Hakim quotes the text of 
a letter al-Daraqutni sent him debating his claim that al-Bukharl and 
Muslim included hadiths only from narrators with two transmitters 
from them. Al-Daraqutni objects, "Indeed al-Bukhari, God bless him, 
included a hadith from . . . Qays b. Abu Hazim from Mirdas al-Aslami 



*^ The invocation of the notion of witnessing (shahdda) was more common in the 
context of establishing the upstanding character {'addla) of a transmitter; see Mushm, 
Sahih, 1:7 and al-Khatib, al-Kifiya, 1:285. For an excellent discussion of rejecting the 
analogy with regards to the number of transmitters needed to eliminate j'aAa/a, with 
references to all the Ash'arl theorists who rejected this analogy as the basis for requir- 
ing two transmitters, see al-'iraqi, al-Taqyid wa al-iddh, 117-8. For a Hanaft rejection, 
see al-Jassas, Usui, 1:567-8. 

*^ Abu Bakr al-BayhaqI, al-Sunan al-kubrd, ed. Muhammad 'Abd al-Qadir 'Ata 
(Beirut: Dar al-Kutub al-'Ilmiyya, 1420/1999), 4:176. See also see Ibn Hajar, Tahdhih 
al-tahdhib, 10:187. It is interesting to note that this Mu'awiya is not included in Muslim's 
Munfariddt. 

"" Al-Qadi 'lyad, Ikmdl al-Mu'lim bi-fawd 'id Muslim, 1:83; al-Dhahabi, Tdnkh al-isldm, 
20:189. 



1 66 CHAPTER FIVE 

(r) from the Messenger of God..., and Mirdas has no transmitter 
other than Qays." Al-Daraqutni provides three more cases in which 
al-Hakim's rule fails to apply, but the scholar gives no response.*" 

b. Doubling Transmission: 1 -^2 -^ 4 

A second interpretation of al-Hakim's writings on the requirements of 
the Sahihayn revolved around the transmission of the actual report and 
not the status of its transmitters. This school of thought interpreted 
the same passages mentioned above as requiring what we can term 
'doubling transmission,' namely a report whose narrators doubled at 
each stage of transmission: one Companion narrated to two Succes- 
sors, who together narrated to four from the next generation, and so 
on. Al-Hakim's colleague and student Ibn Manda upheld this criterion, 
calling for two to three narrators at the level of Successor He added 
that al-BukharT and Muslim based their books on this requirement, 
falling short on only a few occasions {ilia ahruf"). Abu al-Fadl b. Tahir 
al-Maqdisi, who wrote the first comprehensive book on the requirements 
of the Six Books, believed that this was the proper interpretation of 
al-Hakim's description of the ultimate level of sahih hadiths and those 
found in the Sahihayn.*'^ The great Andalusian scholar and traveler Abu 
Bakr b. al-'Arabi (d. 543/1 145) also explicitly states in the introduction 
to his commentary on Bukhari's Sahih that the author required doubling 
transmission for each hadith."'" Abu Bakr al-HazimI (d. 584/1188—9) 
similarly interprets al-Hakim's definition in the Madkhal ila al-IklilJ'^ 
Majd al-Din Ibn al-Athir explains sahth narrations by replicating al- 
Hakim's list of the five universally accepted levels, echoing him further 
by adding that fewer than ten thousand reports meet the highest level. 



'" Al-Hakim, al-Mustadrak, 4:558—9. Generations of scholars such as Abu Bakr 
Muhammad b. Musa al-HazimI (d. 584/1188-9), Ibn al-Salah, al-NawawI, al-'iraqi 
and Ibn Hajar have echoed al-Daraqutnl's disapproval of al-Hakim's claim about al- 
Bukhan's and Muslim's standards. See Abu Bakr Muhammad al-Hazimi, Shurut al-a'imma 
al-khamsa, 35-36; Ibn al-Salah, Muqaddima, 554—6; al-Nawawl, Shark Sahih Muslim, 1 : 140; 
al-'IraqI, al-Taqyid wa al-idah, 122; Ibn Hajar, al-Nukat 'aid kitdb Ibn al~Saldh, 110. 

" Al-MaqdisI, Shurut al-a'imma al-sitta, 15. 

''" Although it seems that Abu Bakr b. al-'Arabl's commentary is lost, his statement 
was repeated by Ibn Rushayd in his rebuttal of this opinion based on the example of 
the hadlth, "Actions are by intentions [innamd al-a'mdl bi'l-niyjdtf'; Ibn Hajar, JVuzhat 
al-nag,arji tawdih nukhbat al-fikar Ji mustalah ahl al-athar, ed. 'Abd al-Samf al-AnIs and 
'Isam Faris al-Harstam (Amman: Dar 'Isam, 1419/1999), 23-24. 

'■' Al-HazimI, Shurut al-a'imma al-khamsa, 24. 



CANON AND COMMUNITY 167 

He considers the possibility that al-Hakim meant the requirement of 
eliminating unknowns, but ultimately deems the doubling transmis- 
sion interpretation more likely. Many scholars, Ibn al-Athlr explains, 
did indeed require this for authenticity [sihhd). He adds that this is the 
highest standard of authenticity, "so who is more deserving of it [ajdar) 
than al-Bukharl and Muslim?""'^ 

We can appreciate these scholars' interpretation of al-Hakim's defini- 
tion of the Sahihayn'% requirements by examining an underappreciated 
source for al-Hakim's thought: a question and answer session recorded 
by his student Masud b. 'All al-SijzI of Naysabur (d. 438-9/1046-8). 
It goes as follows. When al-Hakim is asked why al-Bukharl and Mus- 
lim narrated from Humayd al-Tawll •<— Anas and not from YazTd [b. 
Tahman] al-Raqashi •<— Anas, he replied that other men corroborated 
Humayd's narrations from Anas while Yazld was on his own." In this 
work al-Hakim is also mentioned as saying that, for al-Bukhari, "hadiths 
do not become well-known except by being narrated by two trustworthy 
transmitters who agree on the narration (al-hadith Idyashtahiru 'indahu Hid 
bi-thiqatayn yattajiqdn 'aid riwdyatihi."^'^ Finally, al-Hakim's description of 
a sahth hadlth as being transmitted like a series of testimonies (shahddd) 
leaves little doubt that he intended doubling transmission as a criterion. 
Islamic law required the testimony of two upstanding males in most 
legal matters. It thus seems clear that al-Hakim felt that al-Bukhari 
and Muslim required hadiths to be transmitted by the same number 
at every stage of transmission. 

With the exception of Ibn Manda, Abu Bakr b. al-'Arabi, Ibn al- 
Athir and, oddly, the later Yemeni hadith scholar 'Abd al-Khaliq al- 
Mizjajl (d. 1786—7), commentators who followed this interpretation of 
al-Hakim's work vehemently rejected it as an inaccurate expression of 
the Sahihayn's criteria." Al-MaqdisI exclaims that doubling transmission 
was an admirable ideal, but one that totally faUs to describe the reality 



^^ Ibn al-Aihir, Jami' al-usul, 1:161—3. Ibn al-Athir adds that this requirement would 
be impossible to meet in his own time, since hadlth transmissions had become far too 
diffuse. Here he echoes al-Ghazall a century earlier; Ibn al-Aihir, Jdmi' al-usul, 1:70; 
al-Ghazall, al-Mankhul, 255. 

^' Al-Hakim, Su'dldt Mas'ud b. 'All al-Sijzi ma'a asilat al-baghdddiyyin 'an ahwdl al- 
ruwdt, ed. Muwaffaq b. 'Abdallah b. 'Abd al-Qadir (Beirut: Dar al-Gharb al-Islami, 
1408/1988), 223-4. 

^* Al-Hakim, Su'dldt Mas'ud b. 'All al-Sijzi, 209. 

^^ 'Abd al-Khaliq al-Mizjajl, Muzhat riydd al-ijdza al-mustatdha, ed. Mustafa 'Abd al- 
Karlm Khatib (Beirut: Dar al-Fikr, 1418/1997), 43. 



1 68 CHAPTER FIVE 

of al-Bukhari's and Muslim's books. Al-HazimI says that he has been 
shocked how this palpably false notion had become so widespread, 
demolishing al-Hakim's claim with a long list of examples."'*' These 
scholars note that the very first hadlth in al-Bukharl's Sahih has only 
one transmitter for the first three levels of the isnad):" Ibn Hajar roundly 
rejects all scholars who interpret al-Hakim's explanations as meaning 
doubling transmission. ■^'' He believes that al-Hakim's Madkhal ild al-Iklil, 
where he identifies the top level of sahih with al-Bukharl and Muslim, 
and his Ma 'lifa, which universalizes this definition, both clearly intend 
the elimination of anonymity. Like earlier scholars, he rejects both 
these standards as patently inaccurate representations of al-Bukhari's 
and Muslim's criteria. '^^ 

Ibn Hajar 's teacher, Zayn al-Din al-'Iraqi, invokes the authoritative 
testimony of al-Hakim's senior disciple al-Bayhaql to disprove the notion 
of doubling transmission. He quotes a letter in which al-Bayhaqi skep- 
tically mentions that one Abu Muhammad al-Juwayni (d. 438/1047) 
had cited a hadlth scholar who had required doubling transmission 
for authenticity. No scholars of the ahl al-hadith, al-'iraqi asserts, ever 
upheld that opinion.''" 

c. A Standard for Authenticity and a Standard for the Sahihayn 

In my opinion, the most accurate interpretation of al-Hakim's definition 
of the Sahihayn criteria comes first from a scholar that many later com- 
mentators underestimated. The North African 'Umar b. 'Abd al-Majid 
al-Mayyanishi (d. 583/1187) recognized that al-Hakim distinguished 
between the requirements for authentic reports in general and the 



* Al-Hazimi, Shumf, 15, 24. 

^' Ibn a[-Al\ii'[,Jdmi' al-usul, 1:161—3. Ibn al-Athir acknowledges these criticisms, but 
retorts that al-Hakim knew what he was doing and must have come to this conclusion 
after intensive study. Turning to principles of Islamicate logic, he argues that whoever 
objects to al-Hakim's position could certainly have delved no deeper than he did. A 
critic is thus merely negating al-Hakim's statement. Invoking the principle that the 
alRrmative supersedes the negative (al-muthbit muqaddam 'aid al-nafi), he concludes that 
al-Hakim's position prevails. In any case, it may be that al-Hakim had more informa- 
tion at his disposal, so later scholars should assume the best of him. 

^^ Ibn Hajar, al-Nukat 'aid kitdb Ibn al-Saldh, 110. 

^^ Ibn Hajar, al-Nukat 'aid kitdb Ibn al-Saldh, 41-42. 

^ Al-'IraqI, al-Taqyid wa al-iddh, 21. No mention of doubling transmission appears 
in the text of a letter preserved from al-BayhaqI to al-Juwaynl in al-Subki's Tabaqdt 
al-shdfi'ijya; al-Subkl, "Tabaqdt al-shdfi'iyya, 5:77-90. 



CANON AND COMMUNITY 169 

Standards employed by al-Bukhari and Muslim in particular. Al- 
Mayyanishi's definition for a sahih hadith quotes al-Hakim's Ma'rifa 
verbatim, even citing him clearly as the source. As for the criteria of 
al-Bukharl and Muslim, al-Mayyanishi states (obviously) that they lim- 
ited their works to authentic hadlths, namely reports narrated from the 
Prophet by two Companions, four Successors etc.^' Here the scholar 
provides an unmistakable description of doubling transmission. 

Al-Mayyanishl's younger contemporary, Ibn al-Jawzi, also understood 
that al-Hakim had intended two separate definitions. First, he required 
the elimination of majhul narrators for sahih hadiths in general. Second, 
he defined the Sahihaynh criteria as doubling transmission, with the 
hadith being relayed by "two upstanding narrators from two upstand- 
ing narrators {'adlayn 'an 'adlayn)." Like al-MaqdisI, al-Hazimi and Ibn 
Hajar, however, Ibn al-jawzl deems both these standards reprehensible 
{qabiK) assessments of al-Bukharfs and Muslim's standards. Instead, Ibn 
al-Jawzi says that al-Bukhari and Muslim required simply "a reliable 
transmitter and a well-known report [al-thiqa wa al-ishtihar)^^'^ 

At first glance, the writings of al-Hakim's most well-known student, 
al-BayhaqI, present the one opposing piece of evidence to the argument 
that al-Hakim intended two separate definitions. In his al-Sunan al-kubrd 
al-BayhaqI clearly states that the Sahihayn excluded narrators with only 
one transmitter. This does not necessitate, however, that al-Hakim 
believed that al-Bukhari and Muslim added no other requirements, 
such as doubling transmission. Since al-Bayhaqi never provides any 
systematic discussion of al-Hakim's school of thought or the standards 
of the Shaykhayn, we cannot dismiss anything due to absence of evidence. 
Al-'Iraqi's reading of al-Bayhaqi's letter to Abu Muhammad al-Juwayni 
suggests that al-Bayhaqi questioned whether doubling transmission 
was an existing requirement for authenticity among hadith scholars. 



''' 'Umar al-Mayyanishi, "Ma tajasa'u al-muhaddithjahlahu" in Khams rasa'ilji 'ulum al- 
hadith, ed. 'Abd al-Fattah Abu Ghudda (Beimt: Dar al-Basha'ir al-Islamiyya, 1423/2002), 
266. The text of al-Mayyanishl's work seems to have been corrupted slightly at some 
crucial point in the transmission process, since it reads "and four Successors from each 
one of the Companions {wa ma naqalahu 'an hull wdhid min al-sahdba arba'a min al-tdbi'm)." 
Doubhng transmission would entail four Successors from every two Companions. All 
later scholars reacting to this passage gloss it as meaning l->- 2, not l->- 4. It thus 
seems possible that some copyist mistakenly added "from each one" to the text; cf al- 
Mayyanishl, Md Idjasa'u al-muhaddithjahlahu, ed. Subhl al-Samarra'l (Baghdad: Sharikat 
al-Tab' wa al-Nashr, 1387/1967), 9. 

''^ Ibn al-jawzl, Kitdb al-mawdu'dt, ed. Abd al-Rahman Muhammad 'Uthman, 3 vols. 
(Medina: al-Maktaba al-Salafiyya, 1386-88/1966-68), 1:33-34. 



170 CHAPTER FIVE 

Yet al-'iraqi admits that his explanation interpolates a great deal. He 
cautiously states that "it is as if al-BayhaqI saw [this requirement] in 
Abu Muhammad al-Juwayni's words and was alerting him that it is not 
known among transmission-based scholars.'"'^ 

Al-Mayyanishl's and Ibn al-Jawzi's interpretation of al-Hakim's work 
seems to be the most convincing. Considering the well-established 
principle of rejecting reports through majhul narrators, it is very rea- 
sonable to conclude that al-Hakim considered their elimination to be 
an essential feature of an authentic chain of transmission. In light of 
al-Hakim's statements to al-Sijzi and the legion of hadith scholars who 
upheld the interpretation of doubling transmission, it seems equally 
certain that al-Hakim also considered this to be part of al-Bukharl's 
and Muslim's requirements. 



Admitted Exceptions: al-Mustadrak and the Standards of the Shaykhayn 
as Ideal Rather than Reality 

Al-Hakim's writings leave no doubt that he was aware that many hadlths 
from the Sahihayn did not live up to his definition of their authors' cri- 
teria. Indeed, as al-Daraqutni's letter proves, al-Hakim faced criticisms 
of his definition of their criteria during his own lifetime. He nonetheless 
retained total faith in his "requirements of al-Bukharl and Muslim." 
What is evident is that al-Hakim understood these "requirements" as 
an ideal that the two masters strove to achieve in their work rather 
than a consistent reality. In the Mustadrak al-Hakim thus concedes that 
al-Bukhari and Muslim did not always meet their own requirements 
for eliminating majhukJ'* In his responses to Mas'ud al-Sijzi's questions, 
al-Hakim admits that one of Muslim's transmitters, Fudayl b. Marzuq, 
did not meet Muslim's own standards for authenticity and that he 
should not have narrated from him in his Sahih [fa- 'iba 'aid Muslim bi- 
ikhrdjihifi al-sahTK)P 



'''' Al-'Iraqi, al~Taqyid wa al-idah, 21. 

" Al-Hakim, al-Mustadrak, 1:47. 

"^ Al-Hakim, Su'dlut Mas'ud b. 'Ali al-Sijzi, 109. Scholars like al-NawawI, Abu Hafs 
'Umar al-Bulqlnl and al-SakhawI felt that al-Hakim exempted the Companions from 
the Shaykhajris, requirement for two rdwiy, see al-NawawI, Shark Sdhih Muslim, 1:327; 
'Umar al-Bulqml, Mahdsin al-istildh, in Muqaddimat Ibn al-Saldh wa mahdsin al-istildh, 
296—7; al-SakhawI, Fath al-mughith, 1:68. 



CANON AND COMMUNITY 171 

How could al-Hakim compile an entire hadith collection replicat- 
ing al-Bukhari's and Muslim's methodologies when he acknowledged 
that even these two giants could not always meet their own standards? 
Although al-Hakim envisioned the Sahihayn's, requirements as very 
restrictive and claimed that the contents of his Mustadrak fulfilled them, 
his actual application of them proved latitudinarian. As he notes in the 
introduction to his Mustadrak, he simply compiled the work from hadlths 
narrated by transmitters that appeared in one or both of the SaMiayn, or 
those "like" them. He adds haphazardly that Addition by a trustworthy 
transmitter [ziyddat al-thiqd) does not constitute a flaw {'ilia) in hadith/''' 
As we discussed in Chapter Three, however, selecting reliable isndd% 
represented only half of the critical methodology of hadith scholars; 
even reports narrated via such transmitters had to be examined for 
corroboration or irregularities such as inappropriate Addition. 

Al-Hakim 's vague and lax methods led many later scholars to severely 
criticize the authenticity of material found in the Mustadrak. The 
consummate Hanafi hadith scholar Jamal al-Din 'Abdallah b. Yusuf 
al-Zayla'l (d. 762/1361) struck at the heart of al-Hakim's strategy: he 
had relied on the same transmitters as al-Bukharl and Muslim, but he 
did not thoroughly examine his material to sift weak narrations from 
those enjoying corroboration. "Simply because a transmitter is used in 
[one of] the Sahihs," al-Zayla'i explains, "does not entail that if he is 
found in another hadith, that hadith meets al-Bukharl's or Muslim's 
standards.'"'' Al-Dhahabi thus concluded that the Mustadrak was seriously 
flawed and detracted from al-Hakim's reputation.'"" According to him, 
only one-fourth of the work's contents actually meet the standards of 
the Sahihayn, with another quarter of its hadlths being authentic but 
not meeting their requirements. The remaining half, he states, is of 
dubious reliability*'^ Along the same lines, Ibn Hajar admits that he 
cannot comprehend how al-Hakim could have included certain material 



'* Al-Hakim, al-Mustadrak, 1:39-40. 

''' Jamal al-Dlii 'Abdallah b. Yusuf al-Zayla'l, Masb al-rdya li-ahddith al-Hiddya, ed. 
Muhammad 'Awama, 5 vols. ( Jeddah and Beirut: Mu'assasat al-Rayyan and Dar al- 
Qibia al-Thaqafiyya al-Islamiyya, 1418/1997), 1:342. 

'^" Al-Dhahabi, Tadhkirat al-hujfaz, 3:166. Al-Dliahabl states, "Would that he had not 
composed the Mustadrak, for his poor comportment in it detracted from his virtues {iva 
laytahu lamyusannif al-Mustadrak, fa-innahu ghadda min fadd'ilihi bi-su' tasarrufihi." 

"' Ibn al-Wazir, Tanqih al-an^dr, 38. Al-Bulqlnl states that approximately one hundred 
hadlths in the Mustadrak are forgeries {mawdu); al-BulqInl, Mahdsin al-istildh, 164. 



172 CHAPTER FIVE 

in his Mustadrak. He notes how al-Hakim even used transmitters he 
himself considered weak and had thus consigned to his Kitdb al-du'afd' 
(Book of Weak Narrators). Ibn Hajar believes that al-Hakim was too 
skilled a scholar to make such simple mistakes, but if he knew that 
some material was unreliable and yet included it anyway, then "this is 
a tremendous betrayal {khiydna d^tmd)." Ibn Hajar tried to excuse the 
great scholar on the grounds that he wrote the Mustadrak near the end 
of his life when senility had taken its toll.'" 



Al-Hakim's Politics: the Expansion of the Authentic Umbrella 

The motivation behind al-Hakim's controversial definition of the 
requirements of the Sahihciyn as well as the cause of his inconsistency 
in applying them become clear, however, when one appreciates the true 
purpose of the Mustadrak. He did not compose this work as a legal refer- 
ence, as Abu Dawud had, or as an expression of the body of hadlths 
he had personally collected in his career, as al-Tabarani had. Rather, 
al-Hakim's objective was polemical. 

The unbroken thread running throughout al-Hakim's career was his 
concerted drive to increase the number of hadiths considered authentic 
in the wider Muslim community. Yet this was a matter of great con- 
troversy even among Sunni hadith scholars. In the generation after 
al-Hakim, his own student al-Bayhaqi would make an unprecedented 
declaration that all the reliable hadiths of the Prophet had been docu- 
mented, and thus any previously unrecorded attributions to Muhammad 
should be considered de facto forgeries." Already in al-Hakim's time, 
prominent scholars maintained that the umma had grown too distant 
from the Prophet to identify authentic hadiths. Al-Hakim's colleague Ibn 
Manda, for example, thus stated that "anyone who produces [yukharriju) 
sahih hadiths today is either relying on too lengthy an isndd [yanzilu) 
or is lying."" On the other hand, many shared al-Hakim's vision of 
expanding the number of reports considered authentic. Ibn al-Akhram 
once admitted that he had wasted his life working on his mustakhraj of 
Muslim and regretted having written a joint mustakhraj of the Sahihcyn 



Ibn Hajar, Lisdn al-mizdn, 5:233. 
Ibn al-Salah, Muqaddima, 307. 
AI-Dhahabi, Tadhkirat al-huffaz, 3:158. 



CANON AND COMMUNITY 173 

[Mukhtasar al-sahih al-muttafaq 'alayhi) because "it is our obligation [min 
haqqina) to strive in increasing the sahih hadiths.'"^ 

Al-Hakim's opponents among the hadlth scholars, however, were not 
his principal concern. Relatively early in his career, he had asked how it 
was possible that some groups believed that the hadlths of the Prophet 
amounted to no more than ten thousand reports. The Companions, 
he exclaimed, numbered at least four thousand and spent over twenty 
years in the company of the Prophet! One hadith scholar alone had 
memorized over five hundred thousand hadlths. '* Such ludicrous claims 
limiting the number of reliable hadlths disconcerted al-Hakim terribly, 
and he thus urged hadlth scholars to avoid circumscribing the body of 
authentic reports. He objected, for example, to his teacher al-Masarjisl's 
research on the total number of transmitters in the Sahihayn. A group of 
"heretics and deniers (mubtadi'a wa mulhida),'"'' he explained, were using 
these statements made by transmission-based scholars against them to 
defame [yashtumuna) the use of hadlths.'*' Much later in his career, in 
his very succinct introduction to the Mustadrak, al-Hakim reiterated the 
same complaint. "There has emerged in our time a group from among 
the heretics [mubtadi'a) who defame the narrators of traditions, [saying]: 
the totality of your hadlths that are authentic (yasikhu) does not reach 



" Al-Dhahabi, Tadhkirat al-hujfag., 3:55. 

" Al-Hakim, al-Madkhal ild ma'rifat kitdb al-Iklil, 81-3. 

'^ The term mulhida here should probably neither be understood in its true techni- 
cal sense of "atheists" or "religious skeptics," nor in the later denotation of Isma'ills. 
As Madelung has discussed, al-Ash'arl described mulhid as a term encompassing those 
who deny God's attributes {mu'attil), crypto-Zoroastrians (zanddiqd) as well as other 
bizarre heresies. In the sixth/twelfth century in Iran the term had come to denote 
Isma'ills. The Maturldl theologian Abu al-Mu'ln al-Nasafi (d. 508/1114) thus wrote 
a refutation of the sect entitled Kitdb al-ifsdd li-khudd' ahl al-ilhdd. Al-ShahrastanI (d. 
548/1153) concurs that in this time in Khurasan Isma'ills were also called mulhida. 
Although even in the early fourth/tenth century there was Isma'lll missionary activity 
in Naysabur, we should not assume that al-Hakim intended this group with his refer- 
ence. He was neither a theologian nor a heresiographer, so his addition of the label 
mulhida to mubtadi'a probably just represents another denigration of his opponents. 
Considering that transmission-based scholars of Rayy felt that the Mu'tazilites of the 
city had joined forces with Isma'lll rebels in an uprising in the city in 420/1029, a 
hadlth scholar of al-Hakim's time may not have even distinguished between Mu'tazilites 
and Isma'ills. See S.M. Stern, "The early Isma'lll missionaries in North-West Persia 
and in Khurasan and Transoxania," Bulletin of the School of Oriental and African Studies 
23 (1960): 56-90, esp. 76; W. Madelung, "Mulhid," EF; Ibn al-jawzl, al-Munta^am, 
15:196; see also n. 84 below. 

"" Ai-ankim, al~Madkhal ild al'Sahth, 112. 



174 CHAPTER FIVE 

ten thousand, and all these [other] isnad's, amount to only about one 
thousand fascicules (Juz'), all of them weak, not authentic."'' 

Although al-Hakim reverently describes the Sahihayn as two works 
"whose mention has spread far and wide [intashara dhikruhumdji al-aqtar)," 
he based his mission to expand the umbrella of authentic hadiths on 
the premise that al-Bukhari and Muslim had neither intended to nor 
succeeded in including all of the authentic reports in their works.'" 
Thus, someone's exclusion from the Sahihayn must not be interpreted 
as a criticism of his reliability'^ A wide body of hadiths and hadith 
transmitters still existed outside the two books that met the standards of 
the Shaykhayn, and al-Hakim proved this through an innovative reading 
of Muslim's introduction to his Sahih- He concluded that of the two 
levels of narrators upon which Muslim said he would draw in compil- 
ing his collection, the author had only exhausted the first and had died 
before he could include hadiths from the second level.'"' 

Al-Hakim's interpretation of al-Bukharl's work is even more creative. 
Because that scholar had provided no introduction to his Sahih, al- 
Hakim treated al-Bukhari's cumulative oeuvre as the key to understand- 
ing his requirements. He viewed al-Bukharl's biographical dictionary 
al-Tdnkh al-kabir as the total body of transmitters who comprised the 
scholar's hadith worldview Based on the research conducted earlier by 
al-MasarjisI, he set the number of transmitters in the Tdrikh at about 
forty thousand. But all the reliable transmitters who narrated authentic 
material and appear in the Scihihciyn amount to only about two thou- 
sand. Al-Hakim then turned to al-Bukharl's list of weak transmitters 
(his Kitdb al-du'afd'), which included about seven hundred names, as a 
list of those whom al-Bukharl considered unacceptable. After subtract- 
ing the narrators al-Bukharl used in the Sahth and those he considered 
weak from the forty thousand transmitters included in the Tdrikh al-kabir, 
al-Hakim concluded that more than thirty thousand acceptable trans- 
mitters "remain between the house and the gate." By drawing on this 
untapped body of reliable transmitters and also targeting subjects that 



" Al-Hakim, al-Mustadmk, 1:39. 
™ Al-Hakim, al-Mustadmk, 1:39. 
'^ Al-Hakim, al-Madkhal ild al-Sahih, 114. 

'^'' Al-Hakim, al-Madkhal ild ma 'rifat kitdb al-Iklil, 78; idem, al-Madkhal ild al-Sahih, 112; 
Ibn al-Salah, Siydnat Sahih Muslim, 9 1 . 



CANON AND COMMUNITY 1 75 

al-Bukharl had omitted in his Sahih, one could thus add to the number 
of traditions meeting al-Bukhari's standards.*" 



Al-Hakim's Mubtadi'a and the Ten Thousand 

Who were these "heretics (mubtadi'a)" whose claim that there existed 
only ten thousand authentic hadlths so plagued al-Hakim throughout 
his career? Unfortunately, the scholar provides little description of them 
beyond the brief complaints found in his works. But he does offer two 
important clues as to their identity. First, he quotes al-Bukharfs teacher 
Ahmad b. Sinan al-Qattan (d. 259/872-3) using the term mubtadi' to 
indicate those who oppose hadlth and transmission-based scholars. ''^ 
We could infer from this that during al-Hakim's time mubtadi'a served 
as a transmission-based nomenclature for the reason-based Hanafis or 
Mu'tazilites who constantly criticized the ahl al-hadith's heavy reliance 
on dhdd reports. 

Other evidence for usage of the term suggests it denoted the 
Mu'tazilites more specifically. According to Ibn al-JawzT (d. 597/1200), 
in 408/1017-18 the Abbasid caliph al-Q,adir (d. 422/1031) pub- 
licly demanded, in the famous Qadiri creed, the repentance of the 
"mubtadi'a." Ibn al-Jawzi elaborates that the caliph was requiring "the 
Mu'tazilite-Hanafi jurists [Juqahd') to repent" and disassociate themselves 
from Mu'tazilism ial-i'tizdl), which, like Shiism [al-rafd), the caliph called 
"counter to Islam. ""^ In a letter written to the caliph in 420/1029-30, 
the Buyid amir Yamin al-Dawla mentions the twin perils of "the sin- 
ful Batinis [al-bdtiniyya al-fajard)" and "Mu'tazilite heretics [mu'tazila 
mubtadi'a).'"'^* Mubtadi'a thus appears to have indicated Mu'tazilites and 
not Shiites in these contexts. Ibn al-Jawzi writes that in 460/1067-8 
the jurists and hadith scholars [al-Juqahd' wa ahl al-hadith) of Baghdad 
congregated and demanded that the Qadiri doctrine be publicly pro- 
mulgated once again, because the Mu'tazilite teacher Abu al-Walld 



"' M-iiskim, al-Madkhal ila al-Sahth, 112. 

^'^ "There is not a mubtadi' in the world who doe.s not hate the ahl al-hadith, and when 
a man becomes a mubtadi' the sweetness of hadlth is torn from his heart"; al-Hakim, 
Ma'rifat 'ulUm al-hadith, 5. 

"' "al-mukhdlifa li'l-isldm..."; Ibn al-jawzl, al-Muntazam, 15:125; cf. al-Khatib, Tdrikh 
Baghdad, 4:257-58. Al-Khatib, who saw the caliph many times, explains that the ruler 
wrote treatises declaring the Mu'tazila infidels (ikfar). 

"' Ibn al-jawzl, al-Muntaiam, 15:195. 



176 CHAPTER FIVE 

was insisting on teaching his school's doctrine. One scholar stood up 
in the gathering and cursed the Shiites (Rafida), then another rose to 
separately curse the "mubtadi'a."^' 

Ibn al-jawzl was writing almost a century and a half after these 
events, but his Muntazam often relies on earlier histories such as Tdnkh 
Baghdad. The promulgation of the Qadiri creed in 408/1017-8 was a 
well-known event, and Ibn al-jawzl had documentary evidence for its 
wording."'' Moreover, he was a member of the ahl al-hadith extraordi- 
naire and was even more vehemently opposed to the ahl al-ra'y than 
al-Hakim had been. We may safely assume that he understood the 
term in approximately the same manner as al-Hakim. From this evi- 
dence, we can thus deduce that the term mubtadi'a frequently denoted 
the Mu'tazilites. 

The second clue that al-Hakim provides for identifying these mubtadi'a 
is their claim that there are only ten thousand sahth hadiths. The most 
obvious candidate for such a group would be the Mu'tazilites, who cul- 
tivated a continuous skepticism about the flood of dhdd hadiths adduced 
by transmission-based scholars. The Fadl al-i'tizdl (Virtue of Mu'tazilism) 
of the Shafi'i Mu'tazilite al-Qadi 'Abd al-Jabbar of Rayy (d. 415/1025) 
supports this conclusion. He states that he and his Mu'tazilite colleagues 
are very critical of those who employ significant numbers of hadiths 
in scholarly discourse."' Although he uses such dhdd hadiths in debates 
with his transmission-based opponents, he does so only so they would 
not doubt his affection for the Prophet's sunna. In their own theology, 
however, Mu'tazilites limit themselves to epistemologically certain evi- 
dence {adilla qat'iyja) such as the Qur'an."" Al-Q_adi 'Abd al-Jabbar refers 
to the Mu'tazilites' discriminating standards in his rebuttal of a serious 
transmission-based accusation: that Mu'tazilites use too few hadiths. The 
only reason, he states, that the Mu'tazilites limit their use of hadiths is 
that dhdd reports have too high a probability of being false. "^ 



"' Ibn al-Jawzl, al-Munta^am, 16:106. 

'"' Ibn al-jawzl, al-Muntaiam, 15:279-82. The actual wording of the creed as provided 
by Ibn al-jawzl, however, does not include the term mubtadi'a. 

'^ Al-Qadi 'Abd al-Jabbar b. Ahmad, AbQ al-Qasim al-Balkhl and al-Hakim 
al-jushami, Fadl al-i'tizdl wa Tabaqdt al-mu'tazila, ed. Fu'ad Sayyid (Tunis: al-Dar al- 
Tunisiyya, 1393/1974), 193. 

'" Al-Qadi Abd al-Jabbar, Fadl al-i'tizdl, 156. 

»5 Al-Qadl Abd al-Jabbar, Fadl al-i'tizdl, 195. 



CANON AND COMMUNITY 177 

Ibn al-Jawzl's Muntazam provides similar evidence for this outstanding 
ahl al-hadith grievance with the Mu'tazilites.^" In 456/1064 partisans of 
the transmission-based school physically attacked the Mu'tazilite Abu 
'All Muhammad b. Ahmad al-Mu'tazili (d. 478/1085-6), whom Ibn 
al-Jawzi mocks as having narrated only one hadlth.^' Ibn al-jawzl hurls 
the same accusation at the famous Shafi'i Mu'tazilite Abu al-Husayn 
al-Basri (d. 436/1044).82 

But why did the Mu'tazilites to whom al-Hakim refers set the number 
of authentic hadlths at ten thousand and not some other number? This 
is so because it was the number of hadiths considered to be contained 
in the Sahihayn. Al-Hakim's mubtadi'a opponents told him that this was 
the number of sahth hadiths "in your school i'indakum)," namely the ahl 
al-hadith. Al-Hakim himself stated that the top level of authentic hadiths 
identified with the Sahihayn did not exceed ten thousand. ^^ Al-Hazimi 
concluded from this that the Mu'tazilites' number was based on estima- 
tions of how many hadiths the Sahihayn contained.^* This number must 
indicate the number of Prophetic traditions , since Ahmad b. Salama 
had counted twelve thousand narrations in Muslim's Sahih alone, and 
al-Hakim's teacher al-jawzaqi had placed the total number of narra- 
tions ituruq) in the Sahihayn at 25,480.^'^ Ibn al-Salah placed the number 
of traditions [usut) in each of the SdhWcyn at four thousand, amounting 
to a total of eight thousand.*' Considering that scholars generally put 
the number of Prophetic traditions in al-Bukharl's book at 3,397—4,000 
and in Muslim's at between 4,000 and 8,000, the average number for 
the Sodfihoyn combined would be approximately 9,700.^' 

Abu Nu'aym al-lsbahani provides further evidence that the Sahihayn 
were an important tool in the Mu'tazilites' polemics against the trans- 
mission-based school. He reports that someone who "belittles the 



"' Conflict between the transmission-based school and their opponents on this matter 
seems to have extended back to the time of al-Bukharl and Muslim themselves. Ibn 
al-Salah quotes someone telling Abu Zur'a al-RazI, "Is it not said that the hadiths of 
the Prophet are only four thousand?" He rephes, "Whoever says that, may God jar his 
teeth, this is the claim of the heretic crypto-Zoroastrians {zanddiqdj, for who can account 
[all] the hadiths of the Messenger of God (s) . . .?"; Ibn al-Salah, Muqaddima, 494. 

" Ibn al-jawzl, al-Munta^am, 16:247. 

'2 Ibn al-jawzl, al-Muntaiam, 15:300. 

'' See n. 42 above. 

'* Al-HazimI, Shurut al-a'imma al-khamsa, 32. 

'"' Ibn Hajar, al-Nukat 'ala kitdb Ibn al-Saldh, 70; al-SakhawI, Fath al-mughith, 1:50. 

"" Ibn al-Salah, Sijdnat Sahih Muslim, 101-2. 

'' For the wide range of opinions on this, see Chapter 3, nn. 72, 124, 125. 



178 CHAPTER FIVE 

acceptance of reports" said that al-Bukharl's Sahih only uses some two 
thousand transmitters; all the others are thus clearly unreliable for 
hadlth scholars. Abu Nu'aym responds with a lengthy quotation from 
al-Hakim's Madkhal ild al-Sahih, reiterating al-Hakim's argument that 
al-Bukhari's al-Tdnkh al-kabir contains over thirty thousand acceptable 
but untapped transmitters. ^'^ 

This Mu'tazilite attack was a recurring theme in al-Hakim's career 
and almost certainly served as his primary motivation in composing the 
Mustadrak. Just as Abti Zur'a al-Razi had feared over a century earlier, 
the opponents of the Sunnis had made use of the esteemed standards 
set by al-Bukharl and Muslim in order to object to reports lying outside 
the Sahihayn. Indeed, al-Hakim's Mu'tazilite interlocutors condemned 
the thousands of hadiths not included in the two works as defective 
{saqimd). In order to understand how the Mustadrak embodied al-Hakim's 
response to this attack, we must trace the history of the Mu'tazilite 
treatment of Prophetic traditions until al-Hakim's time. 



Al-Hakim's Target Audience: The Mu'tazilites and their Criteria 
for Authentic Hadiths 

As Josef van Ess has demonstrated, Mu'tazilites found themselves forced 
to adjust the place of Prophetic traditions in their legal and doctrinal 
epistemologies following the Sunni victory in the Baghdad Inquisition 
{Mihna). When Dirar b. 'Amr (fl. 195/810) established Mu'tazilism as 
a cosmological system, hadlth played no major role. He rejected the 
dhdd reports adduced as evidence by his transmission-based opponents 
in favor of the Qur'an and reason, and this position was taken up by 
Abu Bakr al-Asamm (d. 201/816) of the Basran Mu'tazilite school. Van 
Ess postulates that in the wake of al-ShafiYs championing the use of 
dhdd hadiths in law as well as the compilation of major hadlth collec- 
tions in the late second/eighth century, Mu'tazilites found themselves 
forced to meet the challenges posed by the transmission-based school. 
Another early member of the Basran school, Abu Hudhayl (d. 200/815), 
thus tackled the epistemological problem of hadlth with numerical 
requirements. With him we see Mu'tazilites beginning to limit the use 
of hadiths to those they considered massively transmitted beyond the 



Abu Nu'aym al-Isbahani, Mustakhraj, 1:52. 



CANON AND COMMUNITY 1 79 

scope of error (mutawatir). For a hadith to be accepted in discussions of 
dogma, Abu Hudhayl required twenty separate transmitters to meet the 
conditions of tawatur. For legal matters, he demanded only four.^^ The 
Basran Mu'tazilite and polymath al-Jahiz (d. 255/869) also required 
four narrations for a report to qualify as authentic.'"" 

With the end of the Inquisition [Mihna) in 234/848, the Mu'tazilite 
position against the transmission-based scholars was further weakened.'"' 
Ironically, it was during the classical period of Mu'tazilism from the late 
third/ninth century to the early fifth /eleventh that the school had to 
increasingly compromise with its opponents. In this period Mu'tazilites 
began serious studies of hadith comparable to those of their transmis- 
sion-based adversaries. Although Muhammad b. 'Imran al-MarzubanI 
of Baghdad (d. 384/994) was Mu'tazilite, hadith scholars considered 
him reliable as a transmitter, and he composed a book on the hadith 
of the Mu'tazila.'"^ Abu Sa'ld Isma'll b. 'All al-Samman of Rayy (d. 
434 or 445/1042-3 or 1053-4) was one of al-Khatib al-Baghdadl's 
hadith teachers but was a Hanafi imam of the Mu'tazilites.'"' Al-Qadi 
'Abd al-Jabbar was an active student and transmitter of hadith who, 
in a series of dictation sessions (amalt) in Rayy and Qazvin, transmitted 
twenty fascicules (Juz') of hadlths with his own isnads.^^* 

In matters of law, both the Baghdad and Basran schools of 
Mu'tazilism dropped their requirements for authenticating legal hadlths 
to two narrators at each link in the isndd — the same doubling transmis- 
sion required by al-Hakim. The doyen of the Basran school, Abu 'All 
al-Jubba'l (d. 303/915-6) explicitly demanded doubling transmission for 
dhdd hadlths to be admitted in "legal matters (al-shar'iyjdt)."^"-' Abu al- 



^'' Josef van Ess, "L'Autorite de la tradition prophetique dans la theologie mu'tazilite," 
in La Motion d'autoriti au Moyen Age: Islam, Byzance, Occident, ed. George Makdisi et al. 
(Paris: Presses Universitaires de France, c. 1982), 216-7. For a short but comprehen- 
sive discussion of the different rationalist requirements for accepting hadlths, see Ibn 
Qiatayba, Ta'wil mukhtalif al~hadith, ed. Muhammad Zuhrl al-Najjar (Beirut: Dar al-Jll, 
1393/1973), 65-66. 

"*' Ibn Hajar, al-Mukat 'aid kitdb Ibn al-Saldh, 43. 

"" Van Ess, "L'Autorite de la tradition," 220. 

'»2 Al-Khatib, Tdrikh Baghddd, 3:353. 

"" Al-Dhahabi, Tadhkirat al-hujfdz, 3:213. 

"" Al-Rafi'l, al-TadwinJi akhbdr Qazwin, 3:40. 

"''' Abu al-Husayn Muhammad b. 'Ah al-BasrI, (d. 436/1044), Kitdb al-mu'tamad 
ji usul al-jiqh, ed. Muhamed HamiduUah et al., 2 vols. (Damascus: Institut Frangais 
de Damas, 1964), 2:623; al-Juwaynl, Kitdb al-burhdn, 1:607; Abu Ya'la Ibn al-Farra' 
Muhammad b. al-Husayn al-Hanball (d. 458/1066), al-'Udda fi usul al-jiqh, ed. Ahmad 
b. 'All Sir al-Mubarak, 3 vols. (Beirut: Muassasat al-Risala, 1400/1980), 3:861; Abu 



1 80 CHAPTER FIVE 

Q_asim al-Balkhi (d. 319/931), who lived mosdy in Naysabur and whose 
works gained a wide readership in the region, compromised similarly.""' 
In his Qubul al-akhbar, he still demanded massively transmitted hadlths 
(mutawatir) for theological doctrine (usul al-kaldm) and "general legal indi- 
cations [al-amr al-'amm)." For deriving laws {Juru'), however, he believed 
that one need only provide a report transmitted by two or three people 
to two or three upstanding {'adl) people at each level of the isnad. He 
equates this with the requirements for testimony in court.'*" 

The Mu'tazilites' final compromise to the transmission-based Sunnis 
occurred during al-Hakim's lifetime. This brings us to the career of 
al-Q_adi 'Abd al-Jabbar of Rayy which represented a major shift in the 
Mu'tazilite school. Whereas Mu'tazilites had previously associated with 
the hadith-wary Hanafi madhhab, al-Q_adi 'Abd al-Jabbar retained his 
loyalty to the Shafi'T school after embracing Mu'tazilite doctrine.'"" As 
a Shafi'i, he was obliged to accept rulings from dhdd hadiths in matters 
of law even if they lacked the multiple narrations required by earlier 
Mu'tazilites such as al-Balkhi and al-Jubba'l. In his al-Usul al-khamsa, 
al-Q_adi 'Abd al-Jabbar thus states that, while discussing issues of dogma 
and theology [diydnd) requires massively transmitted reports [mutawatir), 
deriving law [Juru' al-fiqh) demands only one or two narrations.'"^ 

By the time al-Hakim was writing in the second half of the fourth/ 
tenth century, the Mu'tazilites' standard for authentic hadith admissible 
in discussions of law thus generally demanded doubling transmission. 
Previously, al-Hakim's teacher and author of a famous sahih work, 
Ibn Hibban, had railed against this stance."" Responding to those 
who rejected dhdd hadiths lacking doubling transmission, Ibn Hibban 
exclaims, "There exists no report from the Prophet (s) narrated by two 



Ishaq al-ShirazI, al-Tabsira fi usul al-fiqh, ed. Muhammad Hasan Hitu (Damascus: Dar 
al-Fikr, 1400/1980), 312; al-Ghazall, al-Mankhul, 255; Ibn Hajar, ahYukat 'aid kitdb Ibn 
al-Saldh, 43; idem, Muzhat al-na^ar, 23. 

""" Cf. Ibn al-Nadlm, The Fihnst, 425-30; al-Khatib, Tdrikh Baghddd, 9:392; Ibn 
al-Murtada, Tabaqdt al-mu'tazila, 88-9. 

'"' Al-Balkhl, Qubul al-akhbdr, 1:17-18. For a short discussion of al-amr al-'dmm, see 
Aron Zysow, "Mu'tazilism and Maturldism in Hanaft Legal Theory," in Studies in Islamic 
Legal Theory, ed. Bernard Weiss (Leiden: Brill, '2002), 252 ff. 

'"" Richard C. Martin, Mark R. Woodward and Dwi S. Atmaja, Defenders of Reason in 
Islam: Mu'tazilism from Medieval School to Modern Symbol (Oxford: Oneworld, 1997) 43; cf. 
Ibn al-Murtada, Tabaqdt al-mu 'tazila, 112-113. Al-Qadi 'Abd al-Jabbar was not the first 
Shafi'l Mu'tazilite, Abu al-Hasan 'All b. Sa'ld al-Istakhrl (d. 404/1014) preceded him, 
but he was certainly the first influential one; al-Khatib, Tdrikh Baghddd, 1 1 :429-30. 

'"" Martin, Defenders of Reason in Islam, 108. 

"" For al-Hakim's link to Ibn Hibban, see al-Subki, Tabaqdt, 4:156. 



CANON AND COMMUNITY 181 

upstanding transmitters {'adlayri), each one of them from two upstanding 
transmitters until it ends at the Prophet (s)!" Those who uphold such 
stringent requirements, he adds, "have intended to abandon all of the 
sunna (j-wnan).'"" Al-HazimI says that the Mu'tazila were in fact the 
only group to require a certain number of transmitters for the accep- 
tance of ahad hadlths. As al-Balkhi had stated, they based this on the 
requirements for court testimony"^ 

Al-Hakim was no doubt extremely familiar with the Mu'tazilite 
demands for authentic hadlths as expressed by both al-Balkhi and 
al-Q_adi 'Abd al-Jabbar. Not only did al-Balkhl reside in Naysabur 
for many years just before al-Hakim's birth, his writings also enjoyed 
popularity in the city. Al-Qadi 'Abd al-Jabbar lived in Khurasan at 
the same time as al-Hakim, and several of his students also lived in 
Naysabur"^ We cannot know exactly where al-Hakim encountered the 
Mu'tazilites whose criticism he noted in his al-Madkhal ild al-Iklil, his 
al-Madkhal ild al-Sahih and finally his Mustadrak, but he would have had 
ample opportunity to do so in his native Naysabur. 



The Mustadrak as Common Measure of Authenticity 

The polemical aim of al-Hakim's Mustadrak and the underlying reason 
for his inclusion of doubling transmission in al-Bukhari's and Muslim's 
criteria now becomes clear. Al-Hakim devoted his career to increasing 
the number of authentic Prophetic traditions in circulation. For him 
the work of al-Bukharl and Muslim provided the highest standards of 
critical rigor, but their two collections had by no means exhausted the 
pool of sahih hadlths. The threat that worried, and motivated, al-Hakim 
throughout his career was the Mu'tazilite claim that only the Sahihayn 
were admissible as authentic. For al-Hakim, the response to this criti- 
cism lay in the standards of al-Bukhari and Muslim. By defining their 
criteria as requiring reports free of transmitters deemed unknown by 
Sunni hadith scholars and possessing the doubling transmission that 
Mu'tazilites required, al-Bukhari's and Muslim's standards became a 
measure of authenticity accepted by all. The Mustadrak constituted the 



Ibn Hibban, Sahih Ibn Hibbdn, 1:118. 
Al-HazimI, Shurut al-a'imma al-khamsa, 47. 
Ibn al-Murtada, Tabaqat al-mu 'tazila, 116-7. 



182 CHAPTER FIVE 

fruit of al-Hakim's efforts; it applied standards he believed compelled 
the acceptance of Sunnis and Mu'tazilites alike to a massive new corpus 
of Prophetic traditions. 

In this new light, al-Hakim's non-sequitur remark that authentic 
hadlths must circulate among scholars like "testimony upon testimony" 
now also becomes clear. Since the Mu'tazila were a key target audi- 
ence of his expansion of authentic hadiths, his definition of sahih had 
to satisfy their requirements. Ibn Hajar alludes to this matter while 
discussing the doubling transmission requirement of the Mu'tazilite al- 
Jubba'l. He says, "This is what al-Hakim was getting at [wa ilayhiyumi'u 
kalam al-Hakim)."^'* Ibn Hajar was certainly justified in concluding that 
al-Hakim's standards somehow involved the Mu'tazila. As Ibn Hibban 
had angrily explained, the notion of requiring doubling narration was 
totally alien to Sunni transmission-based scholars. 

We can now better understand why al-Hakim conceived of the 
standards of al-Bukharl and Muslim more as an ideal than a reality, 
and why he adhered so fiercely to his definition of their requirements 
in the face of tremendous opposing evidence. For him, the two schol- 
ars' requirements embodied a kandn of authenticity accepted by the 
broader community of Sunnis and Mu'tazilites. Unlike earlier hadith 
collections, the purpose of the Mustadrak was not simply to record al- 
Hakim's personal corpus of hadiths or to compile a legal reference for 
transmission-based scholars. Al-Hakim's effort was political. It aimed 
at demonstrating that both the Sahihayn and material that measured 
up to al-Bukharl's and Muslim's standards met the requirements of 
two opposing scholarly camps. This notion of the Sahihayn as common 
ground was to prove central in the two works' canonization. 

Yet how could al-Hakim have expected his audience to grasp the 
requirements of al-Bukhari and Muslim as he defined them if they 
caused later scholars so much difficulty? Al-Hakim's extant works suggest 
that the answer lies in the immediacy of his intended audience. Both al- 
Hakim's responses to Mas'ud al-SijzI and his elliptical analogy between 
transmission and court testimony illustrate that the scholar relied more 
on his personal interaction with others and their familiarity with context 
than on detailed expositions of his theories. The introduction to the 
Mustadrak is thus no manifesto; in fact, it consists of slightly more than 
a single page of disorganized text. Only in another text does al-Hakim 



Ibn Hajar, Muzhat al-na^ar, 23. 



CANON AND COMMUNITY 183 

make his sole reference to his two treatises on the methodologies of 
al-Bukharl and Muslim."' But these also appear to have been ephem- 
eral, and not a single later scholar mentions them. This explains why 
the Mustadrak was never treated as a polemic by later analysts. Only 
by reconstructing the context of al-Hakim's works and reading them 
against the grain could a later scholar understand his motivations and 
target audience. Just as he felt comfortable providing only the most 
tantalizing references to the dreaded "mubtadi'a" and his "standards of 
al-Bukhari and Muslim," so must he have assumed that the bustling 
scholarly circles of Naysabur would have grasped his intent. 



The Discourse of Legal Theory: The Consensus of the Umma on Hadith 

Al-Hakim pioneered the notion of the Sahihoyn as a commonly accepted 
measure of authenticity and a tool for extending this authority to hadiths 
outside the works of al-Bukharl and Muslim. The wider acceptance 
of the Sahihayn in this role, however, depended on the status that the 
various Muslim schools of thought were willing to grant dhdd hadiths. 
By the late fourth/tenth and early fifth/ eleventh centuries, the broader 
Muslim community, including transmission-based scholars, Hanafis, 
Mu'tazilites and even mainstream Shiites had accepted the notion that 
certain Prophetic traditions had received uniform approval and were 
above doubt. Shortly thereafter, by the mid-fifth/eleventh century, 
the major legal schools in Iraq and Iran had acknowledged this class 
of reports and incorporated it into their epistemological systems."^ 



"^ See Chapter 4 n. 58. 

'"' The issue of the epistemological yield of dhdd hadiths and their potential uses in 
deriving law and dogma is long and complicated. The oldest aspect of the debate centers 
on whether or not dhdd hadiths are admissible in deriving laws and are legally compel- 
ling. This debate raged between Mu'tazihtes like Ibrahim Ibn 'Ulayya (d. 218/833) 
and transmission-based scholars like al-Shafi'l. Even among those who accepted that 
dhdd hadiths were legally compelling, however, there was debate over whether or not 
they yield religious knowledge strong enough to elaborate dogma (i'tiqdd) and/or gov- 
ern worship (ta'abbud). Hanafis, Malikis and transmission-based Shafi'l and Hanball 
scholars further disagreed over what kind of dhdd hadiths could delineate or specify 
Qiar'anic rulings such as cutting off the hand of a thief In addition, scholars debat- 
ing the subject did not adhere to a rigid set of terminology. In other debates, scholars 
used the terms 'ilm al-yaqin and 'Urn al-iann to indicate certain knowledge and probable 
knowledge respectively. In the debate over the yield of dhdd hadiths and the effect of 
the community's consensus, however, the term 'ilm denoted certain knowledge (i.e., 
equivalent to the epistemological strength of the Qur'an in deriving law and dogma) 



1 84 CHAPTER FIVE 

A shared conceptual and even linguistic notion of the umma's "accep- 
tance {al-talaqqi bi'l-gubul)" appeared among later Mu'tazilites, Hanafis, 
Malikis, Hanballs/iiber-Sunnis and Shafi'ls/Ash'aris. These agreed-upon 
reports formed a new middle tier: one that yielded an epistemological 
certainty below the almost unattainable confidence conveyed by unim- 
peachable mass-transmission (tawatur) but above the mere probability 
izann) yielded by ahad hadiths. The ahdd hadlths that had received the 
consensus of the community produced a level of certainty sufficient for 
such lofty and restricted tasks as abrogating the Qiar'an and elaborating 
dogma. ' ' ' This widely accepted notion of the epistemological transfor- 
mation that dkdd hadlths could undergo when agreed upon by all would 
prove an essential element in the canonization of the Sahihayn. 

a. The Hanafis 

Systematic discussions of the role of hadlth in the Hanafi epistemo- 
logical system seem to have originated with the writings of the early 
Hanafi judge 'Isa b. Aban (d. 221/836). Later Hanafi legal theorists 
such as al-Jassas regularly quoted his works at length. Our earliest extant 
works of Hanafi legal theory trace their discussions of hadith back 
to Ibn Aban, who originated the tripartite distinction of reports into 
those massively transmitted (mutawdtir), well-known (mashhur) and dhdd. 
Unfortunately, we must depend on later scholars such as al-Jassas and 
Muhammad b. Ahmad al-Sarakhsi of Khurasan (d. ca. 490/1096) for 
explanations of Ibn Aban's thought. Since these two scholars gener- 
ally adhered to Ibn Aban's theories, we can treat their expositions as 
illustrations of Hanafi legal theory in Rayy and Khurasan during the 
fourth/tenth and fifth/ eleventh centuries. 

Al-Sarakhsl states that Ibn Aban believed that mutawdtir hadiths 
yielded epistemologically certain apodictic knowledge {'ilm daruri); 



and i^ann meant probable knowledge (i.e., sufficient only for deriving substantive law). 
For a discussion of the epistemological yield of mutawdtir, mashhur and dhdd hadlths as 
well as the general historical development of these concepts, see Wael Hallaq, "On 
Inductive Corroboration, Probability and Certainty in SunnI Legal Thought," in Islamic 
Law and Jurisprudence, ed. Nicholas Heer (Seattie: University of Washington Press, 1990), 
3-31; idem, "The Authenticity of Prophetic Hadith: a Pseudo-problem," Studia Islamica 
89(1999): 75-90, esp. 80-1. 

'" Ibn Taymiyya was the first to collect a list of scholars from various schools who 
upheld this stance; Ibn Taymiyya, Majmu fatdwd, 13:351-2; Ibn Kathir, al-Bd'ith al-hathith, 
31; al-BulqInl, Mahdsin al-istildh, 172; Ibn Hajar, al-Mukat 'aid kitdb Ibn al-Saldh, 113. 



CANON AND COMMUNITY 185 

anyone who heard the report was immediately certain, without any 
consideration, that its contents were authentic. Mashhur hadlths yielded 
epistemologically certain acquired knowledge {'ilm muktasab); only those 
able to properly contemplate the report's transmission would grasp 
its total authenticity"" Ahad hadlths provided mere probability {^ann), 
which was suitable only for elaborating law in certain circumstances. 
Al-Sarakhsi, who also upholds this opinion, states that mashhur reports 
begin as ahad hadlths but then spread out like mutawdtir hadlths. Their 
epistemological strength stems from the fact that the umma has accepted 
them [qubut). Such hadlths include the famous Prophetic tradition 
allowing believers to wipe water on their socks during ablution instead 
of having to remove them to wipe their feet {al-mash 'aid al-khuffayri). 
Because mashhur reports yield certain knowledge, Hanafis allow their 
use to abrogate, modify or supplement Qjur'anic rulings. Although al- 
Sarakhsi concedes that mashhur reports cannot produce the same level 
of certainty associated with mutawdtir reports, scholarly consensus on 
their reliability [talaqqat bi'l-qubul) endows mashhur reports with "assuring 
knowledge [Him al-tuma'nTniyyd)}^'^ 

Although few of his works have survived, we know from later sources 
that the great Mu'tazilite Hanaft master of the first half of the fourth/ 
tenth century, Abu al-Hasan 'Ubaydallah al-Karkhl (d. 340/952), also 
elevated dhdd hadlths agreed upon by the scholars to a higher level of 
proof Unlike others, however, he believed that the consensus {ijmd') of 
the umma, in and of itself, caused no epistemological change in the 
hadlth. It simply indicated the existence of some compelling proof 
{htgja) for the authenticity of the report, since consensus would not have 
occurred in the first place without such evidence.'-" 

Another Hanafi legal theorist of the fourth/tenth century follows 
Ibn Aban in his tripartite distinction. In his brief treatise on Hanafi 
legal theory, Abu 'All Ahmad b. Ishaq al-Shashl (d. 344/955-6) defines 
mashhur as a report that begins as dhdd and becomes widespread in 
the second and third generations {'asr) until, finally, the umma accepts 
it by consensus [talaqqathu bi'l-qubul). Mashhur reports yield "assured 
knowledge {'ilm al-tuma'mniyya)," and those who reject them are heretics 



"" Al-Sarakhsl, Usui al-Sarakhsi, 1:292. 

'" Al-Sarakhsl, Usui al-Sarakhsi, 1:292-3; cf. al-Jassas, Usui, 1:548. 

'^" Abn al-Husayn al-Basn, Kitdb al-mu'tamad, 2:556. This information does not appear 
in al-Karkhl's short extant m5m/ work. See Abu al-Hasan 'Ubaydallah al-Karkhl, al-Usul 
allati 'alajhd maddrfuru al-hanafiyja (Cairo: al-Matba'a al-Adabiyya, [n.d.]). 



1 86 CHAPTER FIVE 

{mubtadi'). Unlike dhdd hadiths, al-Shashi states, scholars do not differ 
over whether or not such reports are legally compelling. As examples, 
he provides the hadith of wiping over the socks as well as the hadith 
enjoining stoning as a punishment for adulterers.'^' 

We have already discussed al-Jassas's opinions on dhdd hadiths enjoy- 
ing the consensus of the umma and on which scholars have acted in 
law; he admits them as compelling evidence in issues of law and dogma 
{umur al-diydndi)}'^'^ Al-Jassas describes such reports as "widespread 
(mustafidd)."^'^'' His discussion of reports, in fact, devotes significant space 
to defending the use of dhdd hadiths from groups such as the Mu'tazila 
who attack them.'^* 

A significant development seems to have occurred in the Hanafi 
use of the term mashhur between the times that al-Jassas was writing 
in the mid-fourth/tenth century and al-SarakhsI in the second half 
of the fifth/ eleventh. While al-Sarakhsi felt that mashhur reports could 
abrogate or adjust Qur'anic rulings, al-Jassas limited that power to 
mutawdtir hadiths.'^"' Abu al-Hasan al-Karkhi also maintained that only 
mutawdtir hadiths could abrogate the holy book. Yet it appears that this 



'^' Abu 'All Ahmad b. Muhammad Nizam al-Din al-Shashi, Usui al-Shashi, ed. 
Muhammad Fayd al-Hasan al-Kankuhl (Beirut: Dar al-Kitab al-'Arabi, 1402/1982), 
269-72. For his biography, see Ibn Abl al-Wafa', al-Jawdhir al-mudiyya, 1:262. There is 
significant debate over the identity of the author of this text as well as when he lived. 
Three editions of the work have been published, each attributed to a different Shashl. 
In addition to the above-mentioned work, one is attributed to Ishaq b. Ibrahim Abu 
Ya'qub al-Shashi al-Khurasani (d. 325/937), who lived mostly in Egypt (see Ibn Abl 
al-Wafa', al-Jawdhir al-mudiyya, 1:364) and has been published as Usui al-Shdshi (YiAhv. 
Kotob-khane-ye Rashldeyye, [1963]). Finally, the most recent edition attributes the 
work to another Nizam al-Din al-Shashi (fl. 700s/ 1300s) and is published as Usiil 
al-Shdshi: mukhtasar Ji usiil al-fiqh al-isldmi, ed. Muhammad Akram Nadwl and Yusuf 
al-QaradawI (Beirut: Dar al-Gharb al-lslami, 2000). Murteza Bedir has argued that 
the UsUl al-Shdshi cannot have predated the work of the Hanafi legal theorist Abu al- 
Hasan 'All b. Muhammad al-BazdawI of Samarqand (d. 482/1089). The edition used 
here contains some references to figures (al-Dabusi {d. 430/1038}, for example) who 
died after the fourth/tenth century, so at the very least we can be sure that additions 
were made to the text. The bulk of the work, however, seems to be representative of 
other Hanafi usiil treatises from the late fourth/tenth to mid-fifth/ eleventh centuries, 
so there is littie reason to assume the whole work dates from a later time. Suggestions 
that Usui al-Shdshi is a work of Shafi'l usUl are untenable given the distinctly Hanafi 
contents and format of the book. See Murteza Bedir, "The Problem of Usiil al-Shdshi" 
Islamic Studies 42, no. 3 (2003): 415-36. 

"' See Chapter 4, nn. 171 and 173. 

"' Al-Jassas, Usui, 1:548. 

"* See al-Jassas, Usdl, 1:560 and 1:568-73. 

""■ Al-Jassas, ifsul, 1:449. 



CANON AND COMMUNITY 187 

change involved a semantic shift in the usage of the term mashhur rather 
than any revolution in Hanafi epistemology. All these scholars believed 
that the hadith of wiping one's socks was sufficiently well-attested to 
abrogate the Qiar'an. But while Abu al-Hasan al-Karkhi and al-Jassas 
had considered it mutawatir,'"^^' al-Shashi and al-SarakhsI considered it 
mashhur. 

h. The Later Mu 'tazilites 

Abti al-Husayn al-BasrI (d. 436/1044) was a product of late Mu'tazilism. 
Like his teacher, al-Qadi 'Abd al-Jabbar, he espoused Mu'tazilite the- 
ology while belonging to the Shafi'i school of law. His work on legal 
theory, the Kitab al-mu 'tamad, would become one of the most influential 
works in that genre and provide a framework for many later Shafi'l usul 
books. '^' Abu al-Husayn's stance on the epistemological yield of dhdd 
hadiths reflected the Shafi'i/Ash'ari position embraced as orthodox 
among almost all Sunnis: such hadiths yield only probable knowledge 
{^ann), but are nonetheless legally compelling {mujib al-'amal)™ The 
consensus of the umma, however, alters this completely. He explains, 
"As for the wdhid [i.e., dhdd hadith], when the umma has come to con- 
sensus as to what it entails [muqtaddhu) and deemed it authentic, then 
its authenticity is epistemologically certain (juqta'u 'aid sMcitihi).'"^'^'' 

There does not appear to be any evidence that the later Mu'tazilites 
endowed the term mashhur with any technical meaning. In his Fadl al- 
i'tizdl, however, al-Qadi 'Abd al-Jabbar does use the term to describe 
a "well-known" hadith that he employs as a proof text.'^" 



'^" Al-Jassas, Usui, 1:467, 518, 

'^' This is the opinion of the later Mu'tazilite Abu Sa'id al-Muhassin b. Muhammad 
al-Hakim; Ibn al-Murtada, Tabaqdt al-mu'tazila, 119. 

'^" Abu al-Husayn al-Basri, Kitdb al-mu'tamad, 2:570. For what became the stance 
of the Ash'arl orthodoxy, see al-Khatib al-Baghdadi, al-KiJdya, 2:557; idem, Kitdb al- 
faqih iva al-mutafaqqih, ed. 'Adil b. Ytlsuf al-'AzzazI, 2 vols. (Riyadh: Dar Ibn al-jawzl, 
1417/1996), 1:278; al-Juwaynl, Shark al-Waraqdt fi ilm usul al-fiqh (Cairo: Maktabat 
Muhammad 'All Subayh, [1965]), 12; al-ShlrazI, al-Tabsira, 315; al-Ghazall, al-Mankhul, 
252. For a similar Maliki opinion, see Abu al-Walld al-Bajl, al-Ishdra Ji usiil al-fiqh, 
207-8, and Ibn 'Abd al-Barr, al-Tamhid, 1:2, 8. For a Hanball discussion of the school's 
stance and an explanation of the conflicting quotes of Ibn Hanbal on this matter, see 
Abu Ya'la Ibn al-Farra', al-'Udda, 3:861, 900. For the Hanafi position, see Ahmad b. 
Muhammad al-GhaznawI, Usui fiqh al-Ghaznawi, ed. Muhammad Tu'mat al-Qudat 
(Amman: n.p., 1421/2001), 31. 

'^' Abu al-Husayn al-BasrI, Kitab al-mu'tamad, 2:555. 

'™ Al-Qadi 'Abd al-Jabbar, Fadl al-i'tizdl, 195. 



i 00 CHAPTER FIVE 

c. The Shqfi'i/Ash'an Orthodoxy 

Although Abu al-Hasan al-Ash'ari served as the eponym and inspiration 
of the Ash'ari school of speculative theology, its tenets and doctrines 
took shape mainly through the work of three scholars who lived in the 
late fourth/tenth and early fifth /eleventh century: the Baghdad MalikT 
Abu Bakr Muhammad al-Baqillam (d. 403/1013), Abu Ishaq Ibrahim 
b. Muhammad al-Isfarayinl (d. 418/1027) and Abu Bakr Muhammad 
Ibn Furak (d. 406/1015). The influential Buyid vizier and intellectual 
al-Sahib Ibn 'Abbad described these three figures colorfully thus, "Al- 
Baqillani is an engulfing sea, Ibn Furak a silent serpent {sail mutriq) 
and al-Isfaraylni a burning fire.'"^' Here we will focus only on Ibn 
Furak and al-Isfarayini, the two scholars who played salient roles in 
the articulation of the Shafi'i/Ash'ari orthodoxy that would compete 
with the Hanball/iiber-Sunni orthodoxy for ascendancy in fifth/elev- 
enth-century Baghdad. 

Abu Ishaq al-Isfaraylni was born in 337/949 in the city of Isfaraym, 
a town nestled in the gateway to the northern mountains of Khurasan 
and separated from the main road running from Bayhaq to Naysabur 
by a grassy valley and a chain of hills. He studied hadlth intensively 
with scholars such as al-Isma'ili and also attended the lessons of his 
older contemporary Ibn Furak. He was sought out as a hadith expert, 
and among the students to whom he transmitted hadith were al-Hakim 
al-Naysaburi, Abu Bakr al-Bayhaqi and the great Shafi'l of Baghdad 
Abu al-Tayyib al-Tabari (d. 450/1058). Al-Hakim and al-Bayhaqi 
in particular studied Abu Ishaq's works in depth. Among the other 
noteworthy figures who studied law, legal theory, hadith and theology 
at Abu Ishaq's hands were the other great Shafi'ls of the age: Abu 
Ishaq al-Shirazi, 'Abd al-Qahir al-Baghdadi (d. 429/1037) as well as 
the famous Sufi systematizer Abu al-Qasim 'Abd al-Karim al-QushayrI 
(d. 465/1072).'^^ 

Abu Ishaq spent many years studying in Baghdad, but retired to 
his native Isfaraym to teach. He also undertook a visit to the court of 
Mahmud al-Ghaznavi in Ghazna in order to debate the Karramiyya. 



'" "al-Baqillam bahr mughriq wa Ibn Furak sail mutriq wa al-Isfaraymi nar muhriq"; 'Abd 
al-Ghafir al-FarisI, Tdrikh Naysabur al-muntakhab min al-Siydq, 152; al-Dhahabi, Tdrikh 
al-isldm, 28:438; al-Subkl, Tahaqdt, 4:257. 

"^ Al-Dhahabi, Siyar, 17:353-5; cf. Mohammad Javad HqjjetlKermanl, "Abu Ishaq 
Isfaraylnl," Ddr'erat al-ma'dref-e bozorg-e esldrm, 5:158-9; 'Abd al-Ghafir al-FarisI, Tdrikh 
Naysdbur al-muntakhab min al-Siydq, 151-2; al-Subkl, al-Tabaqdt, 4:259. 



CANON AND COMMUNITY 189 

Upon the request of the scholars of Naysabur, he traveled to that city 
and taught at a school built there for his use. When he died, his body 
was carried back to Isfaraym for burial. '^^ 

In his addendum to al-Hakim's Tdrikh Naysabur, 'Abd al-Ghafir al- 
Farisi (d. 529/1134-5) says that Abu Ishaq's works "will last until the 
Day of Judgment, God willing.'"^* God's will was not forthcoming, 
however, and almost nothing of Abu Ishaq's writings has survived. Al- 
Nawawl (d. 676/1277) said that his books were too vast to be contained 
in tomes; '^'^ he wrote treatises on legal theory. Shah 'I substantive law 
and the art of dialectic, but it seems that he devoted a great deal of 
attention to attacking the Mu'tazila. He penned one work entitled al- 
Mukhtasar fi al-radd 'aid ahl al-i'tizdl wa al-qadar (Abbreviated Refutation 
of the Mu'tazila and those Believers in Free Will) and another named 
al-Jdmi' al-haly fi usul al-din wa al-radd 'aid al-mulhidtn (The Ornamented 
Concordance of the Principles of Dogma and a Refutation of the 
Nonbelievers). In addition, Abu Ishaq engaged in several debates with 
the Mu'tazilite al-Qadi 'Abd al-Jabbar'^'^ 

Despite the fact that none of these works have survived, Abu 
Ishaq's scholarly opinions appear frequently in later Shafi'l works on 
legal theory, and figures like al-ShlrazI and Ibn al-Salah recognized 
the importance of Abu Ishaq's role in formulating the Shafi'T/Ash'ari 
stances on issues like abrogation and consensus.'^' Later Shafi'i legal 
theorists have thus preserved Abu Ishaq's stance on the issues of the 
epistemological yield of hadlths and the effect of consensus. From the 
works of Imam al-Haramayn al-Juwaynl and al-GhazalT, we know that 
Abu Ishaq matched the Hanafi tripartite division of reports, identi- 
fying hadlths as mutawdtir, dhdd and a middle tier called mustafid (remi- 
niscent of al-Jassas's terminology). While mutawdtir reports yielded 
certain apodictic knowledge {'ilm daruri) and dhdd hadlths mere prob- 
ability izann), these mustafid reports conveyed "epistemologically certain 
discursive knowledge {'ilm nazari)." Like the 'ilm muktasab that Hanafis 



133 p^ermani, "Abu Ishaq al-Isfaraylnl," Ddr'erat al-ma'dref-e bozorg-e esldmi, 5:158—9. 

'" 'Abd al-Ghafir al-FarisI, Tdrikh Naysabur al-muntakhab min al-Siydq, 151-2. 

'35 Al-Nawawl, Tahdhib al-asmd', 1:170. 

'"^ KermanI, "Abu Ishaq Isfaraylnl," 5:158-9; al-'Abbadi, Mtdb Tabaqdt al-Jiiqahd', 
104, Partial transcripts or quotations from some of these debates seem to have sur- 
vived. See al-Subkl, Tabaqdt, 4:261; MuUa 'All Qarl, Shark al-Fiqh al-akbar, ed. Marwan 
Muhammad al-Sha"ar (Beirut: Dar al-Nafa'is, 1417/1997), 123. 

"' See, for example, Abu Ishaq al-ShlrazI, Sharh al-luma', ed. 'Abd al-Majld Turki 
(Beirut: Dar al-Gharb al-lslami, 1988), 1:573; al-NawawI, Tahdhib al-asmd', 1:170. 



1 90 CHAPTER FIVE 

attributed to mashhur reports, this discursive knowledge resulted from 
a consideration of the report's transmission. Abu Ishaq defined this 
middle tier as those reports on which the imam?, of hadith [a'immat al- 
hadith) had reached consensus.'^" 

In many respects, Abu Ishaq al-Isfarayini's career mirrors that of his 
senior colleague Abu Bakr Muhammad Ibn Furak, who also belonged 
to the Shafi'i school. Ibn Furak studied in Baghdad, spent a period in 
the Buyid capital of Rayy and then moved to Naysabur to teach at a 
madrasa built specifically for him. There he remained until the last years 
of his life, when he accompanied Abu Ishaq to the Ghaznavid court 
to debate the Karramiyya sect.'''^ Unlike Abu Ishaq's books, several 
of Ibn Furak's writings have survived. Like him, though, the main 
opponents that he addresses are the Mu'tazila. The most noteworthy 
is his exposition of Abu al-Hasan al-Ash'arl's school of speculative 
theology, entitled Mujarrad maqdldt al-Ash'an (The Essential Positions 
of al-Ash'arl). In addition, he authored a condensed work on usul 
entitled Mtdb al-hududji al-usul (Definitions in Legal Theory). Finally, he 
devoted a book to interpreting problematic hadlths in a manner that 
trod a middle path between Mu'tazilite rationalism and iiber-Sunni 
anthropomorphism. '*" 

In his Mujarrad maqdldt al-Ash'ari, Ibn Furak employs Prophetic 
traditions very carefully. He admits authentic hadiths as evidence in 
describing God's attributes if they can convey the requisite epistemologi- 
cal certainty, denying that He is Hanndn because "there has not been 
established any authentic report [khabar sahilj) that could dependably 
predicate that attribute to Him.'"*' Ibn Furak concedes the ambiguity 
in the Ash'ari stance on the ability of hadiths to abrogate the Qur'an. 
He states that al-Ash'arl required that a report be mutawdtir or have the 



"" Cf. al-Juwaynl, al-Burhdn, 1:584; al-Ghazall, al~Mankhul, 244. Both al-Ghazall and 
al-Juwaynl disagree with AbQ Isliaq on this matter; cf. al-Juwaynl, al-KdfiyaJi al-jadal, 
ed. Fawqiyya Husayn Mahniud (Cairo: Matba'at 'Isa al-Babi al-Halabi, 1399/1979), 
55-6. 

™ W. Montgomery Watt, "Ibn Furak," EF; M.A.S. Abdel Haleem, "Early Islamic 
Theological and Juristic Terminology: Kitdb al-Hududfi 'l-usul, by Ibn Furak," Bulletin 
of the School of Oriental and African Studies 54, no. 1 (1991): 5-41. 

'■"' These works have been published as: Abu Bakr Muhammad Ibn Furak, Kitdb 
al-hududfi al-usul, ed. Mohamed al-Sulaymani (Beirut: Dar al-Gharb al-lslami, 1999); 
idem, Mugarrad maqdldt al-As'ari: expose de la doctrine d'al-As'ari, ed. Daniel Gimaret (Beirut: 
Dar al-Machreq, 1987); idem, Baydn muskil al-ahddit des Ibn Furak, ed. Raimund Kobert 
(Rome: Pontificium Institutum Biblicum, 1941). Cf. Watt, "Ibn Furak," EF'. 

'*' Ibn Furak, Mugarrad maqdldt al-As'ari, 57. 



CANON AND COMMUNITY 191 

ruling of tawdtur in order to abrogate the holy book, although he admits 
that in its capacity as a restriction or specification {takhsis) of Qur'anic 
rulings, abrogation can in effect occur with dhdd hadlths as well.'*^ In 
his Kitdb al-hudud fi al-usul, Ibn Furak bisects reports into mutawdtir and 
dhdd; the first conveys epistemologically certain apodictic knowledge {Him 
damn), while dhdd hadlths are all those that do not meet the require- 
ments of tawdtur and thus do not yield certain knowledge.'**' 

Later sources, however, provide an impression of a more nuanced 
understanding of reports that allows for the tripartite division pres- 
ent in Abu Ishaq's thought. Al-Juwayni states that Ibn Furak believed 
that reports that scholars had accepted by consensus were "of assured 
authenticity {mahkum bi-sidqihi)," even if these scholars did not act on 
their legal implications.'** Ibn Hajar states that Ibn Furak believed that 
if an dhdd hadlth became ^^mashhur" with well-established transmission, 
it could yield certain discursive knowledge {'ilm nazari)}*-' 

d. The Hanbali Orthodoxy: Abu Ta'ld Ibn al-Farrd' 

During the late fourth/tenth and the fifth/ eleventh centuries in major 
cities tension between the two increasingly divergent strains of the 
transmission-based school became more intense. In Baghdad, partisans 
of the conservative Hanbalis/iiber-Sunnis and those of the Shafi'i/ 
Ash'ari camp competed with one another for intellectual ascendancy 
and state patronage. Both were and remain competing orthodoxies in 
Sunni Islam. 

Abu Ya'la Ibn al-Farra' al-Hanbali (d. 458/1066) of Baghdad served 
as the pivot for the Hanbali school in the fifth/ eleventh century and 
was the single most influential formulator of its legal theory. He wrote a 
commentary on the Hanbali formative text, the Mukhtasar of al-Khiraqi, 
and authored the school's first significant usul text, al-'Udda.^*^ Through 
his writings on issues such as God's attributes and the fundamentals 
of doctrine (usul al-din), he proved himself an inveterate opponent of 



'■" Ihn Furak, Mugarrad maqdldt al~As'an, 199. 

'■" Ibn Furak, Kitdb al-hudud fi al-usul, 150. 

'" Al-Juwaynl, al-Burhdn, 1:585. 

"^ Ibn Hajar, Muzhat al-nazar, 29-30. 

'"" Ibn al-Farra' himself notes that an ealier Hanbali, al-Hasan b. Hamid al-Warraq 
(d. 403/1012-13), wrote a work on usul al-fiqh, which seems not to have survived; al- 
Khatlb, Tdnkk Baghddd, 7:213 (biography of al-Hasan). 



192 CHAPTER FIVE 

the Mu'tazila and the burgeoning Shafi'i/Ash'ari orthodoxy. Among 
his many works we thus find a rebuttal of Ash'arism [al-Radd 'aid al- 
Ash'ariyyd)}*'' This Hanball-Ash'ari disagreement centered on the proper 
interpretation of Qur'anic verses and hadiths dealing with God's attri- 
butes and movement. Ibn al-Farra' believed that true proponents of 
the Prophet's legacy accept the meaning of such reports at face value, 
while Ash'aris deigned to interpret them figuratively'*^ Ironically, this 
enmity masked a growing rapprochement between the Ash'aris and 
leading elements of the Hanball school. Ibn al-Farra', for example, 
found himself forced to admit that the wording of the Qur'an was 
indeed created, and by penning a work of usul structured like those of 
his opponents he was in effect agreeing to join in the discourse estab- 
lished by the Hanafis, Mu'tazilites and Shafi'ls/Ash'arls.'*^ 

In his work on Hanbali legal theory, al- 'Udda fi usul al-jiqh, Ibn al- 
Farra' explains that while dhdd hadiths convey only probability (zann), 
when the umma reaches consensus {ijmd') on some piece of evidence 
such as a hadlth [anyatalaqqdhu bi'l-qubut), the report then yields certain 
knowledge {'ilm). According to the general rules of reality {'add), no 
hadlth enjoying this level of credibility could be incorrect.'^" In another 
work attempting to reconcile Ibn Hanbal's contrasting statements on 
issues of dogma, Ibn al-Farra' reveals that he shares the other schools' 
view on the special capacity of these consensus-approved dhdd hadiths. 
For an dhdd hadlth to be considered as proof on an issue such as see- 
ing God on the Day of Judgment, he explains, the umma must have 
accepted it with consensus [talaqqathu bi'l-qubul).'-'^ 

Ibn al-Farra' does not acknowledge a middle tier of reports, men- 
tioning only dhdd and mutawdtir. Interestingly, however, he does refer 
to the term mashhur in his effort to translate the jargon used by earlier 
hadlth scholars such as Ibn Hanbal into terms comprehensible in the 
arena of legal theory. He explains that hadlth scholars employed mashhur 
for "a report whose transmissions have become massively widespread 
(tawdtara)."'^'^ 



'" For a list of Ibn al-Farra"s works, see Ibn Abl Ya'la, Tabaqdt al-handbila, 2:175. 

''"' Ibn Abl Ya'la, Tabaqdt al-handbila, 2:179. 

'*" Ibn al-Farra', al-Masd'il al-'aqdiyya min Kitdb al-riwdyatayn wa al-wajhayn, ed. Su'fld 

'Abd al-'Aziz al-Klialaf (Riyadh: Adwa' al-Salaf, 1419/1999), 77 ff. 

'™ Abu Ya'la Ibn al-Farra', al-'UddaJi usul al-jiqh, 3:900-1. 

''' Ibn al-Farra", al-Masd'il al-'aqdiyya, 70. 

''^ Ibn al-Farra', al-'Udda fi usul al-jiqh, 3:930. 



CANON AND COMMUNITY 193 



e. The Malikis 



Although Abu Bakr al-BaqillanI and later Ash'arls such as Abu Dharr 
al-HarawI belonged to the Maliki school of law, Malikis were not as 
prominent as the Shafi'is in their contributions to the discourse on 
epistemology or legal theory. Al-Baqillani seems to be the exception in 
not mentioning any special status for ahad hadlths on which the com- 
munity had agreed. Nonetheless, Ibn Hajar mentions that al-Q_adi 'Abd 
al-Wahhab al-Malikl of Baghdad (d. 422/1031-2) insisted in his Kitab 
al-Mulakhkhas (which has probably not survived) that the authenticity 
of reports that the umma accepted with consensus was absolute. '"'■' For 
him tawatur and the consensus of the umma were the only means by 
which transmitted material could yield epistemological certainty'-^* Abu 
al-Walid al-Baji, another prominent Maliki of the fifth /eleventh cen- 
tury, also stated that there are six circumstances in which dhdd hadiths 
can yield 'Urn, one of which is when the umma has accepted the dhdd 
hadith with consensus [talaqqathu bi'l-qubul)}'''^ 

f Al-Hdkim and the Consensus of the Umma 

Although al-Hakim attended the lessons of Ibn Furak, studied closely 
with Abu Ishaq al-Isfarayini and transmitted hadlths from him, his 
work bears little trace of this ubiquitous agreement on the effect of 
consensus on the epistemological yield of hadlths. Furthermore, he does 
not employ the widespread terms mashhur or mustafid in the technical 
sense explored above. Perhaps the closest he comes to acknowledging 
the role of ijmd' or utilizing its associated jargon is his statement that 
authentic reports must be "circulated with acceptance (bi'l-qubul)" 
among hadith scholars.''"'' Such feeble evidence, however, does not 
establish any link between al-Hakim's methodology and that of the 
legal theorists of his time. Although al-Hakim associated with giants 



'^' Ibn Hajar, al-Nukat 'ala kitab Ibn al-Salah, 113. 

'^' Abu Muhammad 'Abd al-Wahhab b. 'All al-Maliki, al-Ishrdf 'aid nukat masd'il at- 
khildf, ed. al-Habib b. Tahir, 2 vols. (Beirut: Dar Ibn Hazm, 1420/1999), 1:233. 

'^^ Abu al-Walld Sulayman al-Bajl, Ihkdm al-fusulfi ahkdm al-usul, ed. Abdel-Magid 
Turki (Beirut: Dar al-Gharb al-Islami, 1407/1986), 330. 

'^^ Al-Hakim, Ma'nfat 'ulum al-luuBth, 77. Al-Hakim did sometimes employ the concept 
of the umma coming to consensus on issues of hadith, such as the unreliability of a 
certain narrator, in other works. Al-Dhahabi, who had access to al-Hakim's lost Tdnkh 
Naysdbur, reports that he wrote, "the umma has come to consensus that ['Abdallah b. 
Muslim] al-Qutabi is a liar"; al-Dhahabi, Mizdn al-i'tiddl, 2:503. 



1 94 CHAPTER FIVE 

in the fields of law, legal theory and theology, he was ultimately only 
a hadlth scholar. He offered the standards of al-Bukhari and Muslim 
as a kandn of authenticity binding for hadlth scholars and Mu'tazilites 
alike, but it was his students and colleagues from among the ranks of 
the legal theorists who truly declared the two works common ground. 
For them the widely accepted notion that ahad hadiths that had earned 
the acceptance of the umma could be declared epistemologically certain 
would provide the key to canonizing the Sahihayn. 



A New Common Ground between the Hanbali/tJber-Sunni and the Shafi'i/ 

Ash 'an Schools 

The role of the SaMhoyn as an authoritative common ground between 
two of the major scholarly camps of the early fifth/ eleventh century 
expressed itself in the careers of two of al-Hakim's close associates: his 
teacher and colleague Abu Ishaq al-Isfaraylnl (d. 418/1027) and his 
student Abu Nasr 'Ubaydallah b. Sa'id al-Wa'ill al-SijzI (d. 444/1052). 
A slightly later figure. Imam al-Haramayn 'Abd al-Malik al-Juwaynl 
(d. 478/1085), soon reiterated this new standing for the two books. 
Beyond their belief in the Qjur'anic revelation and a general Sunni 
loyalty, a common reverence for al-Bukhari or the Sahihayn constituted 
the only firm common ground between figures whose relationships with 
one another were otherwise characterized by bitter enmity. 

A discussion of the role of the Sahihayn as a common denominator 
in the scholarly community must begin with three landmark quota- 
tions from Abu Ishaq, Abu Nasr al-Wa'ili and al-Juwayni.'" Al-Subkl 



''" Although we have no extant proof of these quotes from the books of these 
three scholars themselves, this should not lead us to reject their provenance. Only one 
of al-Wa'ill's works has survived; none of Abu Ishaq al-Isfaraylnl's books is extant. 
Furthermore, both al-Wa'ill's and al-Juwaynl's quotes are of a decidedly oral nature, 
and we should not be surprised not to find the quote in the many works of al-Juwaynl 
that have survived. Ibn al-Salah provides an ismd back to al-Juwaynl for his quote, which 
suggests at least some documentation. Al-Juwaynl's contemporary, Abu al-Muzaffar 
Mansur al-Sam'anl of Naysabur (d. 489/1096), describes Sahih al-Bukhdn wiih the 
statement, "It has been said that the authenticity from the Prophet of what is in it is 
absolutely certain." This proves that this claim was known during al-Juwaynl's lifetime, 
providing a firm terminus ante quern that is relatively close chronologically to the earliest 
quote, namely that of al-Isfaraylnl. In light of these circumstances, we should not equate 
an absence of documentary evidence for these quotes with evidence of absence. One 
claim does exist for a declaration about al-Bukharl's and Muslim's works before that 
of al-Isfaraylnl, but this lacks credibility: Ibn Hajar states elliptically that al-jawzaqi 



CANON AND COMMUNITY 195 

(d. 771/1370) cites the following statement from Abu Ishaq's lost Kitdb 
fi usul al-jiqh: 

The authenticity of the reports in the Sahihayn is epistemologically certain 
in terms of their texts iusulihd wa mutmiha), and no disagreement can occur 
concerning them. If disagreement does occur, it is over the transmissions 
and narrators [turuq wa ruwatiha). Anyone whose ruling disagrees with a 
report and does not provide some acceptable interpretation (ta'wil sd'igh) 
for the report, we negate his ruling, for the umma has accepted these 
reports with consensus.'"'" 

Abu Nasr al-Wa'ill is attributed with the following statement: 

Scholars {ahl al-'ilm), the jurists among them and others, have reached 
consensus {ajma'dj that, if a man swears that if anything in al-Bukharl's 
collection that has been reported from the Prophet (s) is not authentic 
and that the Prophet (s) indeed did not say it, he will divorce his wife, 
he would not be breaking his word and the wife would stay as she was 
in his custody {hibdlatihi)P'^ 



(d. 388/998) also declared the material in the Sahihayn to be absolutely authentic due 
to the consensus of the umma, but we have no other mention or evidence of this. The 
quote does not appear in al-Jawzaqi's al-Muttafaq. Furthermore, why would al-Jawzaql's 
student al-Hakim never mention his teacher's statement among his accolades of the 
Sahihayn? Another figure who supposedly made this claim somewhat later was Abu Nasr 
Abd al-Rahlm b. Abd al-Khaliq al-Yusufi (d. 574/1178-9) of Mecca, about whom 
we know very littie. See Abu al-Muzaffar Mansflr b. Muhammad al-Sam'anI, Qawdti' 
al-adilla fi usul al-fiqh, ed. Abdallah b. Hafiz al-HakamI, 5 vols. (Riyadh: Maktabat 
al-Tawba, 1418/1998), 2:500; Ibn Hajar, al-Nukat 'aid kitdb Ibn al-SaM, 116; Abd 
al-Hayy b. Ahmad Ibn al-'Imad, Shadhardt al-dhahab, 8 vols, in 4 (Beirut: al-Maktab 
al-tijarl, [I960]), 4:248. 

''" Al-Subkl, Tabaqdt, 4:261. 

'^' Ibn al-Salah, Muqaddima, 168. Abu Nasr's statement was echoed later by some- 
one whom Ibn al-'Imad identifies only as Ibn al-Ahdal; see Ibn al-'Imad, Shadhardt 
al-dhahab, 2:135 (biography of al-Bukharl). I have found only one instance of the 
divorce oath trope being used to testify to the authenticity of a hadlth collection 
other than the S^ihihayn, namely the Muwatta ' of Malik. In his Tartib al-maddrik, al-Qadl 
'lyad quotes Abil Zur'a al-RazI as saying, "If a man swore by divorce that Malik's 
hadlths that are in the Muwatta' are all authentic (sihdh), he would not be violating his 
oath. If he swore by the hadlths of another he would be." Although this source is 
late, it is entirely possible that this attribution is correct. As we shall see in the next 
chapter, such statements gave voice to the Maliki desire to put the Muwatta' on par 
with or above the Sahihayn; al-Qadi lyad, Tartib al-maddrik fi taqrib al-masdlik li-ma'rifat 
a'ldm madhhab Mdlik, ed. Ahmad Bakir Mahmud, 5 vols, in 3 (Beirut: Dar Maktabat 
al-Hayat, 1387/1967), 1:196. Yossef Rapoport notes how, in the Islamic culture of 
the Middle Period, the divorce oath was "the most solemn form of oath" and was 
frequently invoked by participants in scholary and political culture when they wanted 
to underscore their certainty or commitment on an issue; Yossef Rapoport, Marriage, 
Money and Divorce in Medieval Islamic Society (Cambridge: Cambridge University Press, 
2005), 90 ff. For examples of scholars using divorce oaths in debates, see al-Khatlb, 



1 96 CHAPTER FIVE 

Finally, al-Juwayni is quoted as saying: 

If a man swore that he would divorce his wife if something in the books 
of al-Bukharl and Muslim that they had declared authentic were not 
[really] from the words of the Prophet (s), I would not oblige him to 
divorce her and he would not be violating his oath due to the consensus 
of the Muslim umma on the authenticity of the two books."'" 



An Articulate Uber-Sunni: Abu JVasr al-Wa'ili 

We are already familiar with the life and career of the great Shafi'i 
theorist, hadith scholar and Ash'ari theologian Abu Ishaq al-Isfarayini, 
for the Shafi'i tradition has sufficiently recorded and honored his legacy. 
Conversely, the Hanball/iiber-Sunni Abu Nasr al-Wa'ili has never 
received his due from the school to which he belonged and for which 
he battled so fiercely. Ibn Abi Ya'la devotes no entry to him in the 
Tabaqdt al-hanabila, although he does respectfully mention a letter Abu 
Nasr wrote to Ibn al-Farra' from Mecca praising one of the latter's 
books.'*"' Abu Nasr's sole surviving work, however, leaves no doubt as 
to his allegiances. He was an iiber-Sunni who viewed Ibn Hanbal as 
the culmination of the Islamic religious tradition. After al-Shafi'i's con- 
voluted attempts at theorizing Islamic law had left Muslims confused, 
Ibn Hanbal took what he could from al-ShafiYs work as well as that of 
Malik and Abu Hanifa, and restored the pure tradition of complying 
with the Prophet's sunna."'^ 

Abu Nasr extends the budding Ash'ari school no mercy. He condemns 
al-BaqillanI, Abu Ishaq al-Isfarayinl and Ibn Furak as the "imams of 
misguidance {a'immat al-dalal)" of his time. For, although they reject 
some opinions of the Mu'tazila, they reject more from the partisans of 
hadith (ahl al-athai)}^''' Abu Nasr is unconvinced by the Ash'ari use of 
speculative reasoning to trump the Mu'tazila, whom he is convinced 



Tankh Baghdad, 7:306; 10:333; cf. Ibn 'Adi, al-Kamil, 1:141; al-Nawawi, Fatawa al-imam 
al-Nawawi C&evml: Dar al-Kutub al-'Ilmiyya, 1402/1982), 140; Abu Zahra, Ibn Taymijya 
(Cairo: Dar al-Fikr al-'Arabi, [1964]), 428-430. 

'*'" Ibn al-Salali, Siydnat Sahih Muslim, 86. 

'*'' Ibn Abl Ya'la, Tabaqdt al-handbila, 2:173. 

"■^ Abu Nasr 'Ubaydallah b. Sa'ld al-Wa'ill al-Sijzl, Risdlat al-Sijzi ild ahl Zobidfi al- 
radd 'aid man ankara al-harf wa al-sawt, ed. Muhammad b. Karim b. 'Abdallah (Riyadh: 
Daral-Raya, 1414/1994), 215. 

'«' Al-Wa'ill, al-Radd, 223. 



CANON AND COMMUNITY 197 

are a spent force. He explains that while Ash'aris purport to debate 
the Mu'tazila, they are in fact with them. Indeed, "they are viler than 
them [akhass hdt")r^'^'^ 

Abo Nasr al-Wa'ili was born in the Iranian province of Sijistan to 
a family that followed the Hanafi madhhaby^'' He soon split from his 
father's school, however, and traveled to Khurasan and Ghazna. In 
404/1014 he undertook the pilgrimage to Mecca, then visited Bagh- 
dad, Egypt and Basra before returning to Mecca, where he remained 
until his death."'*' 

Abu Nasr studied hadith with al-Hakim al-Naysaburi, probably in 
Naysabur, and clearly respected him a great deal. He seems to have 
viewed him as an exemplary hadith scholar. Abu Nasr would tell a story 
about his teacher's encounter with the famous litterateur Badl' al-Zaman 
al-Hamadhanl (d. 398/1008) upon his arrival in Naysabur to a crowd 
of admirers. When al-Hamadhani awed onlookers by memorizing a 
hundred lines of poetry after one hearing and then belittled the memo- 
rization of hadlths, al-Hakim decided the time had come to put this 
bonvivant litterateur in his place. He approached him and asked him 
to memorize dijuz' of hadiths. When he returned a week later to test 
al-HamadhanI, he could not remember the specifics of the isnads. Al- 
Hakim scolded him for mocking something more difficult to memorize 
than poetry and told him, "Know your place (i'raf nafsak)."^'^''' 

Abu Nasr seems to have produced very few works, only one of which 
has survived. His al-Radd 'aid man ankara al-harf wa al-sawt (Rebuttal of 
Those who Deny [that God's Speech Consists of] Words and Sounds), 
written as a letter to the people of Zabid in Yemen, is probably a sum- 
mary of his magnum opus, the Kitdb al-ibdna al-kubrd. Al-Dhahabi praises 
both this work and its author, whom he lauds with the unique accolade 
"the imam of the knowledge of the sunna [imam 'ilm al-sunna)."^''^' He 



'" Al-VVa'ill, al-Radd; 81, 222. He considers the last generation of Mu'tazilites to 
be 'Abd al-Jabbar and al-Sahib Ibn 'Abbad. 

'^^ This is the cause of Abu Nasr al-Wa'ill's outrageous inclusion in Hanafi bio- 
graphical dictionaries, see below n. 166. 

"''' Cf. Ibn al-Athir, al-Lubdb Ji tahdhib al-ansdb, 3:351-2; Ibn al-jawzl, al-Munta^am, 
16:187 (Ibn al-Jawzl errs in his death date, which he has as 469 AH); al-Dhahabi, 
Siyar,n :654-6; idem, Tdrikh al-isldm, 30:95-97; al-Safadi, al-WdJi bi-al-wajdydt, vol. 
19, ed. Ridwan al-Sayyid (Beirut: Steiner Verlag, 1413/1993), 19:372-3, "Abu Nasr 
SiizT," Dd'erat al-madref-e bozorg-e esldmi, 6:318-9; Ibn Abl al-Wafa', al-Jawdhir al-mudiyja, 
2:495. 

'" Al-Dhahabi, Siyar, 17:173. 

'f'" Al-Dhahabi, Tadhkirat al-hujfdz, 3:211. 



1 98 CHAPTER FIVE 

explains that the work dealt incisively with questions of the Qjur'an's 
nature and God's attributes."''' The Rebuttal itself addresses numerous 
topics, such as the nature of the Qur'an, God's speech, His sitting on 
the throne, the beatific vision, and His descending to the lowest heav- 
ens at night. The Ibana was read during its author's lifetime, for Ibn 
Taymiyya tells us that when Abu Nasr and the Ash'arl Abu Dharr al- 
HarawT were both in Mecca they fell into a serious argument over the 
nature of the Qjur'an and the Ibana} "^ In addition, later scholars such 
as Ibn al-Salah cite Abu Nasr's hadith work on the narration of sons 
from their fathers as the definitive book in that genre.'" 

The Radd indicates that Abu Nasr possessed a deep understanding 
of both Ash'arT and Mu'tazilite thought as well as the Ash'ari mis- 
sion of defending Sunnism using the Mu'tazilites' rational tools. The 
Mu'tazila claimed that speech consists of words and sounds, which are 
created. Since Sunnis believed that the Qur'an was God's speech, it 
must also be created. The Ash'arls circumvented this trap by denying 
that God spoke in sounds; rather. His speech was figurative. His words 
were "meaning inhering in the essence of the Speaker [ma'nd qd'im bi- 
dhdt al-mutakallim)." Abu Nasr rejects the Ash'ari position, stating that 
it was well-understood amongst Arabs that the term "speech (kaldm)" 
denoted actual words. "^ The Ash'arls claimed that God "spoke" only 
in the figurative (majdzi) sense because, if He were actually to articulate 
words, this would be anthropomorphism itajsTm, tashbth)."^ 

Against this, Abu Nasr defends the iiber-Sunnis' literalist interpreta- 
tion of God speaking or moving in space. He states that his party is the 
true ahl al-sunna "who stand fast on what the early generations [salaf) 
had transmitted to them from the Messenger of God (s)" and rely on 
the traditions of the Companions where God and His Prophet are 
silent.'^* Reports about God speaking, ascending His throne or descend- 
ing to the lowest heavens have been bequeathed to the Muslims of the 
present day by upstanding and trustworthy imams like Malik through 
many corroborating reports (turuq mutasdwiyd)}'^ 



"''> Al-Dhahabi, Siyar, 17:654. 

"" "Abu Nasr Sijzl," Dd'erat al-ma'dref-e bozorg-e esldmi, 6:318. 

'" Al-'IraqI, al-Taqyid wa al-iddh, 273; Zakariyya al-Ansarl, Fath al-bdqi bi- shark alfijjat 
al-'Irdqi, ed. Thana'allah al-Zahidi (Beirut: Dar Ibn Hazm, 1420/1999), 562. 
'" Al-Wa'ill, a/-&(/rf, 81-2. 
'" Al-Wa'ill, al-Radd, 82. 
'" Al-Wa'ill, al-Radd, 99. 
"' Al-Wa'ill, al-Radd, 186. 



CANON AND COMMUNITY 199 

Abu Nasr's position on the epistemological yield of dhad hadlths 
reveals an acute and cunning approach to dialectic. He acknowledges 
that most scholars believe that dhdd hadlths are only compelling in 
law {'amal). Unlike massively transmitted reports (mutawdtir), they do 
not yield certainty {'ilm). He replies using the Ash'aris' own position 
that tawdtur is not defined by a fixed number of reports, but rather by 
circumstances that lead to the total alleviation of doubt concerning 
the authenticity of the message. This could occur with one hundred 
narrations, four or even less depending on circumstances. Most hadlths 
dealing with God's attributes, he continues, have been transmitted in 
sufhcient number to alleviate doubt and make the heart feel at ease."^ 
He mocks the Ash'aris' attempts to parry the Mu'tazila using rational 
argumentation without recourse to hadlths that are "dhdd and do not 
yield 'ilm." How can they say that a sahih dhdd hadith does not yield 
'ilm but their reason does!?''' 

Although Abu Nasr never provides a systematic discussion of the 
different levels of hadlths and their epistemological yields, he employs 
the notions of consensus and other terminology of the legal theorists 
of his day. This should not surprise us, for we know that he read Ibn 
al-Farra"s works.''" He describes one hadith as ''sahih mashhdr" and as 
having been "accepted by the umma [talaqqathu al-umma bi'l-qubul)}'"^ 
In fact, in a brief listing of the different kinds of Prophetic traditions, 
he lists reports that enjoy the consensus of the umma as the opposites 
of those that scholars have abandoned and not acted on.'"" 

As Abu Nasr's quotation about the umma's consensus on al-Bukhari's 
Sahih indicates, he respected the work highly. On the controversial 
issue of God speaking audibly, he cites al-Bukhari for his inclusion of 
a hadith in which God calls to the believers on the Day of Judgment 
with a voice."" On another occasion he describes a hadith as "occur 



"f' Al-Wa'ill, al'Radd, 187. 

'" Al-Wa'ili, «/-&</(/; 81, 101. 

"" See n. 161 above. 

' " Al-Wa'ill, al-Radd, 151. This hadith, "/nna Allah tajdwaza li-ummati md haddathat bihi 
anjusuhd md lam tatakallam aw ta'mal bihi" appears in Muslim's SaMh- See Sahih Muslim: 
kitdb al-imdn, bdb 58. 

""' Al-Wa'ill, al-Radd, 206, 

"" "istashhada bihi al-BukhdriJi kitdbihi al-Sahlh"; al-Wa'ill, al-Radd, 164. Hadith: 

yahshuTu Allah al-nds yawm al-qiydma For a discussion of this Prophetic tradition, see 

Ibn Hajar, Fath al-bdn, 13:555-561; SaMh al-Bukhdri: kitdb al-tawMd, bdb 32. 



200 CHAPTER FIVE 

ring in the Sahih {jd'aji al-Sahih)."^^''^ His work makes no specific men- 
tion of Muslim's Sahih. When urging Muslims to resort to the hadith 
collections of those who have stood out as experts on Islam and the 
Prophet's legacy, he names as examples the Sunam of Abu Dawud, 
Ibn al-Athram, 'Uthman b. Sa'ld al-Darimi (d. 280/894) and Harb b. 
Isma'il al-Sirjani (d. 280/893-4).'"^ Given his esteem for al-Bukharl's 
collection, it seems odd that he does not include his Sahih in this list. 
But Abu Nasr al-Wa'ill was first and foremost a loyal Hanbali, and 
the four collections that he mentions are all the works of Ibn Hanbal's 
close associates. 



Imam al-Haramayn al-Juwaym: A Consummate Shafi'i and Ash' an 

Born in 419/1028 in the constellation of villages called Jovayn astride 
the winding road from Bayhaq to Isfaraym in the hills near Naysabur, 
'Abd al-Malik b. 'Abdallah al-Juwaynl studied Shafi'i law and Ash'ari 
theology in Naysabur until the new Seljuq administrator of the city 
declared that "[Abu al-Hasan] al-Ash'arl is guilty of innovation in 
religion (mubtadi') worse than the Mu'tazilites.""'* Al-Juwaynl thus fled 
to Baghdad and then to the Hijaz in 450/1058. He became one of the 
most sought-after masters of his school, teaching in Mecca and Medina 
and earning the honorary title "imam of the two Sanctuaries [al-hara- 
mayn)." When the great administrator Nizam al-Mulk came to power, 
al-Juwaynl became one of his favorites. The vizier invited the scholar 
to return to Naysabur and teach at his state-sponsored college, the 
Nizamiyya. He remained in the city until his death in 478/1 085. '''■^ 

Al-Juwayni produced extremely important works in the fields of 
legal theory, Shafi'i substantive law and Ash'arl theology. His Waraqdt 



'82 Al-Wa'ill, al-Radd, 174. This hadith, "Tahmilu al-samdwdt 'aid asba' wa al~ardajn 
'aid asba'..." appears in the Sahihayn; Sahth al-Bukhdn: kitdb al-tawhid, bdb qawl Alldh 
limd khalaqtu bi-yadi; Sahih Muslim: kitdb si/at al-mundfiqin, bdb sifat al-qiydma wa al-janna 
wa al-ndr. 

'«' Al-Wa'ill, al-Radd, 223, 

'"* Ibn al-jawzl, al-Muntazam, 15:340; see also, BuUiet, "The Political-Religious History 
of Nishapur in the Eleventh Century," 82 IF. 

'"' 'Abd al-Ghafir al-FarisI, Tdrikh Naysdbdr al-muntakhab min al-Siydq, 508; al-Dhahabi, 
Siyar, 18:468-77; al-Subkl, "Tabaqdt, 5:171-88; al-Safadi, al-Wdfi bi-at-wqfdydt, 19:171-5; 
C. Brockelmann and L. Gardet, "al-Djuwaynl," EP; Hallaq, "Caliphs, Jurists and the 
Saljuqs in the Pohtical Thought of Juwayni," Muslim World 74, no 1 (1984): 27-8. 



CANON AND COMMUNITY 201 

(Pages) and his Kitdb al-burhdn (Book of Demonstration) have remained 
two standard texts for teaching the principles of jurisprudence in the 
Shafi'T school. In addition, his massive twenty- volume ^^A work entitled 
Mhdyat al-matlabfi dirdyat al-madhhab (The End of the Question for Know- 
ing the Path) served as the formative text around which all later legal ref- 
erences in the Shafi'i school would revolve.""' Al-Juwaynl also composed 
a seminal work on Ash'ari theology entitled al-Shdmil (The Compre- 
hensive Book) as well as another book rebutting the Mu'tazilite school. 
The study of hadith was certainly al-Juwayni's weakest field. He did 
receive an ijdza from Abu Nu'aym al-Isbahani (although as a child) and 
was very familiar with the Sunan of al-Daraqutnl, which he employed 
as a source of legal hadiths and narrator criticism (jarh wa ta'dil)}^^ We 
also know that he received a copy of Muslim's Sahih from Abu 'Abdallah 
al-Husayn b. 'All al-Tabari (d. 499/1105-6).""' Al-Dhahabi, however, 
questioned his mastery of the sahih collections. He points out that in 
the Kitdb al-burhdn al-Juwayni describes the hadith in which the Prophet 
approves of Mu'adh b. Jabal's decision to use his own reasoning in the 
absence of any Qur'anic or Prophetic injunctions as "recorded in the 
sahih%, with its authenticity agreed upon [mudawwan fi al-sihdh muttafaq 
'aid sihhatihi)." Al-Bukhari and al-Tirmidhi, however, expressly reject 
this hadith as unreliable. '^^ 



The Sahihayn Canon: The Authority of Convention and Common Ground 

The above three quotations of al-Isfarayinl, Abu Nasr al-Wa'ili 
and al-Juwayni provide the first historical evidence that the Sahihayn 



'"*' Al-Safadi, al-Wafi l)i-al-wafayat, 19:173; 'Alijum'a, al-Imam al-Shcifi'i wa madrasatuhu 
al-fiqhiyja (Cairo: Dar al-Risala, 1425/2004), 80-82. 

"" Al-Subkl, Tabaqdt, 5:171, 182. 

'"" 'Abd al-Ghafir al-FarisI, Tdrikh Majsdbur al-muntakhab min al-Siydq, 305. 

'«" Al-Dhahabl, Siyar, 18:471-2; al-Safadi, al-Wdfi bi-al-wafiydt, 19:173; al-Juwaynl, 
Burhdn, 2:882. Al-Subkl contests his teacher al-Dhahabl's condemnation of Juwaynl's 
hadith skills, saying that the hadith, in which Mu'adh b. Jabal tells the Prophet what 
steps he would take in deciding the correct course of action while traveling to Yemen 
(i.e., consulting the Qur'an, the Prophet's precedent, then his own reason), is in al- 
Tirmidhl's collection; al-Subkl, Tabaqdt, 5:187-8. This is immaterial, however, since 
al-Juwaynl had claimed that the authenticity of the hadith was agreed upon by all — a 
statement that al-Bukharl's dismissal undermines. Al-BukhaiT considered the hadith to be 
weak because one of the narrators, al-Harith b. 'Amr al-Thaqaft, was majhul; Ibn Hajar, 
Tahdhib al-tahdhib, 2:139-40. In addition, al-Tirmidhi criticizes the report for lacking a 
continuous isndd; Jdmi' al-Tirmidhi: kitdb al-ahkdm, bdb md jd'a ft al-qddi kayfa yaqdi. 



202 CHAPTER FIVE 

functioned as texts authorized by a certain community. In these three 
cases, representatives from the two opposing strains of the transmis- 
sion-based school affirm a common source for discussing the authentic 
legacy of the Prophet. For one Hanbali/iiber-Sunni and two Shafi'i/ 
Ash'arls, the works of al-Bukhari and Muslim had authenticated a 
common tract of the Prophetic past. This agreement authorized the 
Sahihayn by demonstrating that the three scholars all acknowledged a 
common body of proof texts that had been guaranteed by a mutually 
recognized scholarly consensus. 

We must note that the quotations of al-Wa'ill and al-Juwayni do 
not directly identify the authority of the Sahihayn as that of legal com- 
pulsion. Rather, they focus on the two works' total authenticity and 
the authority that this created for the books as a convention within a 
community of discourse. These two statements took place in a context 
that was uniquely interactive. The formula of swearing to divorce one's 
wife in order to prove the truth of a statement was a trope among 
scholars in the classical Islamic world. '^^ It was a rhetorical statement 
made in a dialectical context. Al-Juwaynl's and Abu Nasr's statements 
were responses to stimuli designed to test the conventions to which 
they subscribed. They made these statements because some questioner 
or adversary had elicited them. Perhaps someone had probed the two 
scholars for their opinion on the Sahihoyn or questioned the authenticity 
of al-Bukharl's or Muslim's collections. Their responses showed that 
the scholars acknowledged a common convention to which both were 
accountable. They recognized a new canon regarding sources for the 
Prophet's sunna. 

This role of drawing inclusive lines for a community that certainly 
encompassed the Hanbali/iiber-Sunnis and the Shafi'i/Ash'arls but 
also may have included other groups such as the declining Mu'tazilites 
was unique to the Sahihayn. Al-Isfaraylni, who penned polemical works 
against the Mu'tazilites, felt he could claim the Scih^hcyn as an authorita- 
tive common ground in his work on legal theory. Abu Nasr al-Wa'ili, 
who denigrated Abu Ishaq al-Isfarayini as one of the most destructive 
religious forces of his time, nonetheless seconds his evaluation of &M 
al-Bukhan's reliability. Years later, al-Juwayni echoed Abu Nasr al-Wa'ill's 
evaluation, including Muslim's Sahih as well. What is truly shocking is 
that al-Juwaynl detested Abu Nasr both personally and ideologically. 



Rapoport, Marriage, Money and Divorce in Medieval Islamic Society, 90 fF. 



CANON AND COMMUNITY 203 

Once while strolling through the book market in Mecca, he found al- 
Wa'ill's book Mukhtasar al-baydn (probably an abbreviation of his Ibdna). 
In a lost refutation entitled Naqd kitdb al-Sijzi (Refutation of al-Sijzi's 
Book), he describes the work as dealing with the nature of the Qur'an 
and "saying that Ash'arls are unbelievers (kuffdr)." Al-Juwaynl states, "I 
have never seen an ignoramus [jdhil"") more daring in calling people 

unbelievers and hastier in judging the imams "'^' Considering that 

Abu Nasr and al-Juwaynl viewed each others' positions as anathema on 
issues ranging from ritual law to the nature of the Qjur'an and God's 
attributes, the Sahihayn (or, for Abu Nasr, Sahih al-Bukhdn) were one of 
the few articles on which they actually agreed. 

Bridging the chasm between these two strains of transmission-based 
scholars was not merely a personal matter In the fifth/ eleventh century, 
Baghdad was plagued by internecine violence between the Hanball/ 
iiber-Sunnis and the Shafi'i/Ash'aris. Throughout 469/1076-7 and 
470/1077—8, for example, debates between Abu Ishaq al-Shlrazi and 
his Hanball opponents spilled into the streets, where mobs supporting 
the two groups ruthlessly hurled bricks at one another"'^ Only state 
intervention could end the quarrel. On the level of doctrine and public 
religious symbolism, the Sahihayn could thus serve as one of the few 
threads joining these two parties, the canon that bound both together 
as one community. 

The notion of consensus [ijmd' or talaqqi al-umma bi'l-qubut) provided 
the key to authorizing these two works within the expanded boundar- 
ies of a widened Sunni Islam. As we have seen, the augmenting effect 
of communal consensus on dhdd hadlths proved a common discourse 
among the Hanafi, Maliki, Mu'tazilite, Shafi'i/Ash'ari and Hanball 
schools in the first half of the fifth/ eleventh century. It was to this 
epistemological authority that Abu Ishaq, Abu Nasr and al-Juwaynl 
turned in order to empower the new hadith canon. 

Clearly, however, the entire Muslim world did not consider the two 
works totally authentic. Imami Shiites, for example, would never have 
subscribed to this opinion. How, then, should we understand these claims 
of consensus? Ijmd' is, fundamentally self-centered, invoked and defined 
by scholars attempting to make their beliefs normative by ascribing 



"' Taqi al-Dm 'All b. 'Abd al-Kafi al-Subki (d. 756/1356), al-Sayf al-saqilfi al-radd 
'aid ibn al-Zafil, ed. Muhammad Zahid al-Kawtharl and 'Abd al-Haflz Sa'd 'Atiyya 
([Cairo]: Matba'at al-Sa'ada, 1356/1937), 19-20. 

"^ Ibn al-jawzl, al-Munta^am, 16:171-2. 



204 CHAPTER FIVE 

them to a wider community. This 'community' rarely actually applies 
to the entire Muslim world. Rather, it encompasses those Muslims who 
uphold correct belief or practice as imagined by the scholar invoking 
ipna' in that moment. As al-Juwaynl states, ijma' does not include those 
Muslim heretics {mubtadi'd) whom "we have declared unbelievers.'"^' A 
claim of ijmd' is thus always 'accurate' from the point of view of the 
scholar invoking it, since anyone who disagrees with it is, according to 
the claimant, not truly part of the Muslim community at that moment. 
Claims of ijmd' are thus inherently subjective, and their efficacy in a 
debate thus depends entirely on the opponents' willingness to consider 
themselves beholden to the same "we," the same community, and the 
same terms invoked by the claimant. 

In essence, then, ijmd' is prescriptive and not a description of real- 
ity.'^* Someone who invokes the authority of consensus is attempting 
to force another to heed evidence he considers universally compel- 
ling. In this sense, the actual boundaries of the umma mentioned 
by Abu Ishaq al-Isfarayini, Abu Nasr al-Wa'ill and al-Juwaynl prove 
immaterial. In reality, asserting the authenticity of the hadiths in the 
Sahihayn could convince only those willing to accept the premises of 
mainstream Sunni hadith criticism as it existed in the fifth /eleventh 
century. This claim of consensus would not even have convinced a 
great Sunni muhaddith like al-Daraqutni, whose standards for Addition 
had proven more stringent than al-Bukhari's or Muslim's.'^' On the 
rhetorical plane, however, invoking the authority of consensus on the 
Sahihayn could prove compelling provided one's opponent also upheld 
the status of the two books. Claims made about ijmd' on the Sahihayn 
thus depended on an opponent's commitment to imagining the same 
authoritative station for the two books and acknowledging the same 
conventions of argument. 



"" Al-Juwaynl's requirements for inclusion in ijma are vague and highly subjective, 
generally restricting it to qualified jurists and legal theorists (usuli). He states that the 
opinions of vaguely named "heretics (mubtadi'a)" may be considered depending on the 
circumstances; al-Juwaynl, al-Burhdn, 2:684-5, 689. 

194 -pj^jj follows Snouck Hurgronje, Goldziher and Makdisi. See Makdisi, "Hanbalite 
Islam," in Studies on Islam, ed. and trans. Merlin L. Swartz (Oxford: Oxford University 
Press, 1981), 253. 

'^^ Brown, "Criticism of the Proto-Hadith Canon," 31-34. 



CANON AND COMMUNITY 205 

Conclusion: Why the Sahlhayn Now? 

As the long fourth century came to a close around 450/1058, a cadre 
of hadith scholars and legal theorists from the transmission-based 
schools had put forth al-Bukhari's and Muslim's collections as texts 
wielding the authority of a common convention. Yet the Sahlhayn were 
not necessarily the most widely used hadith collections. Malikis could 
rely on the Muwatta ', Hanballs on the Musnad. Even Abu Nasr al-Wa'ill 
clearly favored Abu Dawud's collection; al-Juwayni relied more on al- 
Daraqutni's Sunan in his everyday work. Moreover, when Abu Ishaq 
al-Isfaraylnl made his proclamation about the Sahihcyn many decades 
had passed since hadith scholars such as Ibn al-Sakan and jurists like 
al-Khattabi had articulated the possibility and need for hadith works 
that could act as loci of consensus. Why canonize the Sahihcyn, and 
why now? 

It was al-Hakim al-Naysaburl who provided the necessary catalyst for 
the transformation of al-Bukharl and Muslim into kandns, of authentic- 
ity. He served as a magnet for studies of the Sahihcyn, inheriting two 
works the contents of which had been thoroughly studied and whose 
transmitters had been painstakingly identified. No other hadith collec- 
tions had received the ceaseless attention devoted to the Sahihcyn and 
their authors' methods, and no other works had consistently earned the 
admiration of the community of hadith scholars. Most importantly, no 
other collections could conceivably bear the claims that al-Hakim made 
about their authors' methods and the status of their transmitters. 

The genre of ilzdmdt had been established by al-Daraqutnl, but 
al-Hakim transformed it from an obscure and personal activity into 
a polemical tool. The mission of expanding the number of authentic 
hadlths in circulation motivated al-Hakim throughout his career, and 
the concept of the "requirements of al-Bukharl and Muslim" furnished 
the vehicle for doing so. He identified the methodologies that the two 
scholars employed in compiling their works with the highest level of 
critical stringency. Apparently conscious that he was acting more on 
ideals than reality, al-Hakim defined their standards in a manner that 
met the requirements of both Sunni hadith scholars and the Mu'tazilites 
whose attacks on the transmission-based school had irked him through- 
out his career In his Mustadrak, al-Hakim presented the standards of 
al-Bukharl and Muslim as a kanon of authenticity that could endow a 
vast new body of hadlths with the reliability of the ScihWcyn. Al-Hakim's 



206 CHAPTER FIVE 

work became very influential very quickly, attracting commentary and 
spreading as far as Andalusia during the author's lifetime. 

Al-Hakim, like most of the Sahihayn Network, worked within the 
realm of hadith collection and criticism, but his colleague Abu Ishaq 
al-Isfaraylni and his student Abu Nasr al-Wa'ill participated in the wider 
discourse of epistemology, law and legal theory Indeed, the broader 
Muslim community had earlier imagined the authority with which 
ijmd' could endow hadiths, and hadith scholars had begun conceiving 
of the hadith collection as a possible locus of communal consensus. 
It was only during the late fourth/tenth and early fifth/ eleventh cen- 
turies, however, that legal discourse among a wide variety of schools 
had collectively articulated that the ijmd' of the umma could raise dhdd 
hadiths from yielding mere probability to total certainty. Abu Ishaq and 
Abu Nasr al-Wa'ili combined these notions of the hadith collection as 
a common ground and the authority endowed by ijmd' in their proc- 
lamation of the absolute authenticity of al-Bukharl's and/or Muslim's 
Sahths. Al-Juwaynl seconded this declaration, proving that the Sahihayn 
could bridge the serious enmity between the Hanball/iiber-Sunni and 
Shafi'l/Ash'arl camps. 

These developments endowed the Sahihayn with a new potential 
authority within the body of transmission-based scholars. They had 
been acknowledged as a common ground and a convention recognized 
by both the Hanball/iiber-Sunni and the Shafi'l/Ash'arl schools. More- 
over, both al-Hakim and the scholars who declared the community's 
authoritative consensus on the two books envisioned a canon that 
reached beyond the boundaries of the transmission-based schools. With 
the end of the long fourth century we thus find that members of the 
transmission-based schools had authorized two texts that both defined 
an existing convention for discussing the Prophet's legacy and carried 
the potential to extend that convention to a wider community. What 
would come of this potential beyond the three figures of Abu Ishaq al- 
Isfaraylnl, al-Wa'ill and al-Juwaynl? Only by meeting widespread needs 
within the scholarly community could the Sahihayn canon take root. 



PART TWO 



CHAPTER SIX 

THE CANON AND THE NEEDS OF THE COMMUNITY: 
THE SAHIHATMAS MEASURE OF AUTHENTICITY, 
AUTHORITATIVE REFERENCE AND EXEMPLUM 



Introduction 

At some moment around the dawn of the fifth/ eleventh century, the 
Sahihayn emerged as authoritative representations of the Prophet's sunna 
among the transmission-based Shafi'i and Hanball schools. Beyond that 
theoretical singularity when a book becomes more than the sum of its 
pages, however, canonization involves forces greater than the career 
of one remarkable individual, like al-Hakim al-Naysaburi, or the iso- 
lated declarations of a few, like Abu Ishaq al-Isfarayini or Abu Nasr 
al-Wa'UT. It represents the choice of a community to transform texts 
into authoritative institutions, to endow them with authority because 
doing so allows them to meet certain needs or perform certain essential 
functions. 

The authorization of the Sahihayn indeed met three important needs 
in the Sunni scholarly community of the mid-fifth/ eleventh century. 
First, the canon provided a common measure of authenticity for 
scholars from different legal schools engaged in debate, exposition of 
their doctrines or efforts to bolster the hadiths they employed as proof 
texts. Spreading out from al-Hakim's students and prominent mem- 
bers of the Sahihayn Network to leading scholars among the Shafi'i, 
Hanball and Maliki schools in Iraq and Iran, the two works became 
an authoritative convention for evaluating attributions of the Prophet's 
interpretive authority. This canon would become indispensable for 
scholars, for citing a hadith as being included in one or both of the 
Sahihayn endowed it with an authenticity guaranteed by the umma's 
consensus. By the mid-eighth/fourteenth century, even the hadith- 
wary Hanafi school found it essential to acknowledge this convention. 
Second, in a time when jurisprudence was growing increasingly distant 
from the specialization of hadith criticism, the institution of the canon 
also began playing an important role as an authoritative reference for 
jurists who lacked the expertise necessary to independently evaluate 



210 CHAPTER SIX 

hadiths. Finally, the Sahihayn canon was not simply a conventional tool 
for authorizing Prophetic reports. Al-Bukhari and Muslim also became 
the exemplum that could shape the science of hadith collection and 
criticism itself Therefore, as institutions such as the madrasa formed, 
schools of law solidified and the field of legal theory fully matured, 
the Sahihayn emerged as powerful institutions for jurists searching for 
conventions of debate or authoritative references, as well as for hadith 
scholars struggling to systematize the study of the Prophet's word. 

The nature of the authority that the Sahihayn canon wielded, however, 
was far from absolute. The power of the canon was bound intimately 
to the interactive functions it fulfilled. It was an illusion conjured up as 
convention in the dialogic space of debate and exposition. Within the 
closed circles of legal or theological schools, however, scholars had no 
compunction about rejecting al-Bukharl's and Muslim's hadiths. 



1 . The Need for a Common Measure of Authenticity: 
The Sahihayn in Scholarly Debate 

Traditions of the Prophet were prima facie compelling for Muslim schol- 
ars. Certainly among their own colleagues, the jurists of a particular 
legal school felt no pressure to provide rigorous chains of transmission 
for hadiths used in elaborating their common body of law. In such 
circumstances, it was not necessary to go beyond simple attributions 
of Prophetic authority. The issue of a hadlth's authenticity arose only 
when opinions clashed, when competing parties challenged the reli- 
ability of one another's evidence. 

The Baghdad Shafi'l Abu Ishaq al-ShlrazI (d. 476/1083) empha- 
sized this need for a common measure of authenticity in his manual 
on juridical debate, the Kitdb al-ma'una f al-jadal. Engaging his Hanafi 
counterparts proved an alluring interest for al-ShlrazI, and he authored 
two other works on issues of disagreement between the two schools. ' In 
the Kitdb al-ma'una, al-ShlrazI addresses the possibility of a situation in 
which a Shafi'l scholar faces demands to produce an isndd for a hadith 
he has adduced as evidence. If an opponent demands that one provide 



' Abu Ishaq al-ShlrazI, Kitdb al-ma'una fi al-jadal, ed. 'Abd al-Majld Turki (Beirut: 
Dar al-Gharb al-lslami, 1408/1988), 55 (editor's introduction). These two works are 
al-Nukatfi al-masd'il al-mukhtalaf fihd bayn al-imdmajn Abi Hanifa wa al-Shdfi'l and Tadhkirat 
al-mas'ulm Ji al-khildf bayn al-Hanaji wa al-Shdfi'i. 



THE CANON AND THE NEEDS OF THE COMMUNITY 211 

a chain of transmission, one should simply refer him to "a relied-upon 
book [kitdb mu'tamad)." The difficulty in providing or rebutting evidence 
only arises when one's own hadith is not found in "the sunan."- 

It was this need for a common measure of authenticity in the context 
of debate or exposition that the Sahihayn canon so effectively fulfilled. 
Indeed, al-Bukhari's and Muslim's works had acquired a powerful air 
of legal compulsion by al-ShlrazT's time. As Abu Ishaq al-Isfarayinl had 
declared, to rule against a hadith found in the Sahihayn without some 
convincing excuse was to oppose the consensus of the Muslim com- 
munity. Writing some sixty years after al-Isfarayini's death, al-Ghazali 
emphasized how widespread the notion that the contents of the two 
books were legally compelling had become. In his al-Mankhul min ta 'liqat 
al-usul, a work on legal theory directed against Hanafi opponents of 
the Shafi'i/Ash'ari school, al-Ghazali states casually: 

We know that if a mufti, if a question proves too difEcult for him and he 
looks through one of the Sahihayn, comes across a hadith that addresses his 
aim, it is not permitted for him to turn away from it, and he is obligated 
to rely on it [al-ta'wU). He who permits [turning away from the hadith] 
has broken with the consensus [of the umma] {kharaqa al-ijmd')? 

That al-Ghazali does not feel obliged to prove this claim, but rather 
employs it axiomatically to argue a separate point, illustrates how 
compelling an institution the SaMhcyn had become by the late fifth/ 
eleventh century. It was thus in debates or polemical writings that the 
Sahihayn canon functioned most clearly as a vehicle by which a scholar 
could wield the authoritative consensus of the community against his 
opponent. 



Takhrij: Applying the Measure of Authenticity 

The Sahihcyn canon thus found its most salient application in the takhrij 
of hadiths, or citing the various collections in which a report appears. 
In theory, a scholar seeking to provide such validating references for 
his hadiths could cite any hadith collection he wished. The attempt 



^ Abu Ishaq al-ShlrazI, Kitdb al-ma'unaji al-jadal, 160. 

^ Al-Ghazall, al-Mankhul, 269. For the importance of consensus in the formation and 
maintenance of orthodoxy in Islam, and the equation of breaking it with disobeying 
the Prophet, see Devin Stewart, Islamic Legal Orthodoxy (Salt Lake City: University of 
Utah Press, 1998), 48-53. 



212 CHAPTER SIX 

to prove the reliability of a report, however, hinged inevitably on the 
quality of the collections to which he referred. For this reason, takhrij 
generally involved the products of the sahih movement, especially the 
Six Books and later the Sahihs, of Ibn Khuzayma, Ibn Hibban and the 
Mustadrak of al-Hakim. As we shall see, referring to the Sahihayn canon 
differed qualitatively from citing these other respected collections. Not 
only did al-Bukharl's and Muslim's works alone enjoy the claim of 
the community's consensus on the authenticity of their contents, they 
also better accorded with the rules of Sunni hadith criticism as they 
coalesced in the mid-fifth/ eleventh century and beyond. 

Takhrij using al-Bukhari and Muslim, however, did not merely serve 
as a stamp of approval for the relatively limited quantity of material 
featured in their collections. Taking advantage of the differing narra- 
tions or multiform permutations of a single Prophetic tradition, scholars 
like the Shafi'i Abu Bakr al-Bayhaqi (d. 458/1066) were able to extend 
the measure of authenticity to material that differed significantly from 
the actual contents of the Sahihayn. Later scholars such as al-'iraqi, 
Ibn Hajar and al-SakhawT thus took al-Bayhaqi and others to task 
for telling their readers that a hadith appears in the Sahihayn when in 
fact al-Bukhari or Muslim included only the basic isndd [ad al-isndd) or 
general text of the report.* 

More importantly, the critical standards of al-Bukhari and Muslim, 
however a scholar might choose to define them, continued as a stamp 
of legitimacy that could extend the consensus on the SaMhcyn to new 
bodies of hadith. In his treatise on Sufism, entitled Scijwat al-tasawwuf 
(The Essence of Sufism), Muhammad b. Tahir al-Maqdisi (d. 507/1 113) 
proudly states that he will not use any poorly attested ighanb) hadlths 
in arguments against opponents. Rather, he will rely only on those 
found in the Sahihayn, which "the umma of Muslims has accepted with 
consensus, as well as that which meets [al-Bukharl's and Muslim] 's 
requirements (shartihimd) but that they did not include.""' Here the dual 
power of the S^hihcyn canon is clear in the authority of al-Bukhari's 
and Muslim's texts themselves and in their capacity as a kanon by which 
their authority could be extended to outside hadiths. 



* Ibn Hajar, al-Mukat 'ala kitab Ibn al-Salah; 81; al-Sakhawi, Fath al-mughith, 1:60-1. 
'' Al-MaqdisI, Sajwat al-tasawwuf, ed. Ghadah al-Muqaddam 'Adrah (Beirut: Dar 
al-Muntakhab al-'Arabi, 1995), 133. 



THE CANON AND THE NEEDS OF THE COMMUNITY 213 

To the present day, the "requirements of al-Bukharl and Muslim" 
have retained this function as a vehicle in which the authorizing con- 
sensus of the community can be deposited for later application. In the 
perennial debate over seeking the intercession of dead saints (tawassul), 
the modern scholar Yusuf Hashim al-Rifa'i defends this practice against 
detractors by invoking a hadlth in which the caliph 'Uthman tells a 
man seeking aid to call upon the late Prophet for assistance in gain- 
ing God's favor Al-Rifa'i avers that this hadlth meets the criteria of 
al-Bukharl and Muslim, "so there remains nothing one could criticize 
or denounce in the authenticity of the hadlth.'"' 

The array of sources that could be invoked in takhrij led hadlth 
scholars to contemplate a system of ranking the various respected 
hadlth collections. As we have seen above, al-Hakim had pioneered 
this by associating the Sahihayn and their requirements with the highest 
level of authentic hadlths. In his Shurut al-a'imma al-khamsa, al-Hazimi 
(d. 584/1188-9) uses the students of the early hadlth transmitter al- 
Zuhri (d. 124/743) as a template for ranking the critical stringency of 
al-Bukhari, Muslim, Abu Dawud, al-Tirmidhi and al-Nasa'i. Al-Bukhari 
drew only from the top level, which consisted of scholars like Malik, 
while Muslim also relied on the second tier Abu Dawud and al-Nasa'l 
resorted to the third level, while al-Tirmidhi plumbed the depths of 
the fourth.' 

Since debate often pitted al-Bukharl and Muslim or one of these two 
scholars' critical requirements against one another, there gradually devel- 
oped a more detailed ranking strictly for the Sahihayn. Al-Mayyanishi 
(d. 583/1187) concluded that the highest level of reliability belongs 
to hadlths on which both al-Bukhari and Muslim agreed. The second 
level consists of reports that only one of them included. The third level 
features reports that meet their requirements but do not appear in the 
Sahihayn, and the lowest level consists of hadlths that fail to meet those 
conditions but nonetheless possess good isndds.^ Ibn al-jawzl followed 
al-Mayyanishi, adding several lower levels of hadlths such as forged 
reports.^ Ibn al-Salah developed the final form of this ranking system, 
which consisted of hadlths: 



'' Yusuf al-Sayyid Hashim al-Rifa'i, Adillat ahl al-sunna wa al-jama'a (Cairo: Matba'at 
al-Sa'ada, 1405/1985), 96. 

' Al-HazimI, Shurut al-a'imma al-khamsa, 43—4. 

" Al-Mayyanishl, 262-3. 

' Ibn al-jawzl, al-Mawdudt, 1:32-5. 



214 CHAPTER SIX 

1) Agreed on by al-Bukharl and Muslim 

2) Only included in al-Bukhari 

3) Only included in Muslim 

4) Meeting the requirements of al-Bukhari and Muslim 

5) Meeting only the requirements of al-Bukharl 

6) Meeting only the requirements of Muslim 

7) Hadiths that are sahih but do not meet al-Bukharl's or Muslim's 



requirements 



10 



These rankings were not simply exercises in empty contemplation. If we 
understand these evaluations as judgments about the functional value 
of hadith collections, we must appreciate that they arose as responses to 
pressing questions within the scholarly community As Monroe Beardsley 
states in his discussion of instrumentalism in aesthetics, "Statements 
of value are to be regarded as proposed solutions to problems of value, 
that is, situations in which choices have to be made."" Scholars faced 
situations in which they had to choose between competing authentic 
hadiths. As Ibn al-WazIr notes incisively in his comparison between the 
critical methods of Muslim and Abu Dawud, "Know that the purpose 
of this discussion is to demonstrate that the hadiths of Muslim are 
preferable to those of Abu Dawud in the case of competition {ta'drud) 

between them "'^ 

Indeed, these comprehensive rankings emerged in the wake of 
seminal attempts to systematize the Sunni study of hadith. Although 
scholars such as Abu 'All al-Naysaburl (d. 349/960) and al-Isma'lll 
(d. 371/981-2) had been evaluating collections such as the Sahihayn from 
a relatively early date, concerted efforts to rank the various products of 
the sahih movement seem to have started suddenly in the early and mid- 
sixth/twelfth century'^ This followed works like al-Khatib al-Baghdadl's 



'" Ibn al-Salah, Muqaddima, 169. This ranking has been followed by almost all later 
scholars, some of whom have discussed the levels in more detail; see Abu al-Fayd 
Muhammad al-Hanaft al-Faslh al-HarawI (d. 837/1434), Jawdhir al-usul Ji 'Urn hadith 
al-Rasul, ed. Abu al-Ma'all Athar al-Mubarakfurl (Medina: al-Maktaba al-'Ilmiyya, 
[1973?]), 19; Ibn Hajar, al~Nukat did kitdb Ihn al-Saldh, 107; MuUa Khatir, Makdnat 
al-Sahihayn, 98-102. 

" Monroe C. Beardsley, Aesthetics: The Philosopliy of Criticism (New York: Harcourt, 
Brace and World, 1958), 543. 

'^ Ibn al-WazIr, Tanqih al-an^dr, 81. 

" Ibn Hazm (d. 456/1064) seems to have been an exception. Al-Dhahabi reports 
that he ranked the best hadith collections as the Sahihayn, the Muntaqd of Ibn al-Sakan, 
the Muntaqd of Ibn al-Jarud, the Muntaqd of Qasim b. Asbagh, then the Sunans of Abu 
Dawud, al-Nasa'l and then thirty other books; al-Dhahabi, Tad/ikirat al-hujfdz, 3:231. 



THE CANON AND THE NEEDS OF THE COMMUNITY 215 

al-MJayaJi 'ilm al-riwaya (The Sufficient Work on the Science of Trans- 
mission), which were attempts to authoritatively recognize choices that 
Sunni hadith scholars, jurists and legal theorists had made about the 
transmission, evaluation and usage of hadiths. Scholars like al-Hazimi 
found themselves forced to see where the methods of al-Bukharl and 
Muslim fit within the shared rules of hadith study articulated in the 
writings of systematizers like al-Hakim, al-Khatib and Ibn 'Abd al-Barr 
(d. 463/1070). 

Ranking al-Bukharl's critical stringency above that of Muslim, for 
example, acknowledged significant and practical principles that had 
emerged as predominant among Sunni hadith critics. On the issue of 
when one could accept the vague phrase "from/on the authority of {'arij" 
in an isnad as not masking a break in transmission, it was the school of 
thought associated with al-Bukharl and 'All b. al-Madlnl that became 
the mainstream stance. These two masters had required proof that the 
transmitter employing "from/on the authority of" had actually met at 
least once the person from whom he claimed to narrate. Muslim, on 
the other hand, had only required that they be contemporaries with a 
possibility of having met one another'* In his al-KiJaya, al-Khatib al- 
Baghdadl declares that the community of hadith scholars had come 
to consensus that requiring at least one meeting was correct. When 
Ibn 'Abd al-Barr sought to apply the criteria of the sahih movement 
to Malik's Muwatta', he therefore turned to al-Bukharl's requirements 
as the prevailing rule. Most major hadith scholars or critics since then, 
such as Ibn al-Salah (d. 643/1245), have followed Ibn 'Abd al-Barr's 
and al-Khatlb's formulations of the rules governing the use of "from/on 
the authority of ( (zn)."'"' Ranking Muslim slightly below al-Bukharl in 



" See above Chapter 3, section on Muslim's Methodology in his Sahih. 

'^ For the majority (al-Bukharl's stance), see Abu al-Hasan 'All b. Muhammad al- 
QabisI, Muwatta' al-imdm Malik, ed. Muhammad b. 'Alawl b. 'Abbas al-Maliki (Abu 
Dhabi: al-Majma' al-Thaqafi, 1425/2004), 38 (I have interpreted al-Qabisl's phrase 
^idrdk bayjin' as 'proof of direct transmission;' this could also mean 'clear contempora- 
neousness); Ibn 'Abd al-Barr, al-Tamhid, 1:12; al-Khatib, al-Kifaya, 2:229; Abu al-Husayn 
b. al-Qattan (d. 628/1231), al-Iqnd' fi masd'il al-ijmd', ed. Husayn b. Fawzl al-Sa'idl, 2 
vols. (Cairo: al-Faruq al-Hadlthiyya li'1-Tiba'a wa al-Nashr, 1424/2004), 1:66-7; idem, 
Bajdn al-wahm wa al-ihdm, 3:287; Ibn al-Salah, Muqaddima, 220; Ibn Rushayd, al-Sanan 
al-abyan, 32; al-Dhahabl, al-Muqi^a, 45-6; Khalll b. Kaykaldl al-'Ala"! (d. 761/1359), 
Jdmi' al-tahsiljt ahkdm al-mardsil, ed. Hamdl 'Abd al-Majld al-Salafi (Baghdad: al-Dar al- 
'Arabiyya li'1-Tiba'a, 1 398/ 1 978), 1 34 ff.; Ibn Kadilr, al-Bd'ith al-hathith, 44-5; al-Bulqinl, 
Mahdsin al-istildh, 224-5; Ibn Rajab, Shark 'Ital al-Tirmidhi, 1:360-5; al-'iraqi, al-Tabsira 
wa al-tadhkira, ed. Muhammad b. al-Husayn al-'iraqi al-Husaynl (Fez: al-Matba'a al- 
Jadlda, 1353/[1935]), 1:162; al-SakhawI, Path al-mughith, 1:202-13; al-San'anl, Tawdih 
al-ajkdr, 1:299. Al-Nawawl seems to favor Muslim's stance in his Taqnb, but states that 



216 CHAPTER SIX 

critical stringency thus amounted to tailoring the canon to the contours 
of convention among hadlth scholars. 

The superiority of the Sahihayn over other respected hadlth collec- 
tions used for takhnj also had palpable implications in scholarly debate. 
This shines forth clearly in a seventh/thirteenth century debate that 
raged between the towering Shafi'i hadlth scholar Ibn al-Salah and his 
contemporary al-'Izz b. 'Abd al-Salam (d. 660/1261-2)"* over the per- 
missibility of a type of supererogatory prayer known as saldt al-ragha'ib. 
The evidence for this type of prayer hinged on a hadlth adduced by 
al-Ghazali in his Ihya' 'ulum al-din (The Revival of the Religious Sci- 
ences). Although both Ibn al-Salah and Ibn 'Abd al-Salam agreed that 
this report was weak, the former felt that people should stUl be allowed 
to perform the prayer, while Ibn 'Abd al-Salam argued that "paving 
the way for lying about the Messenger of God is not permitted [al- 
tasabbub ila al-kadhib 'aid Rasul Allah lajajuz)."" In the course of letters 
these two scholars wrote to one another publicly debating the issue, Ibn 
al-Salah defended his point of view by arguing that "the hadlth has 
sahth narrations," citing a hadlth from Ibn Majah's Sunan as evidence.'" 
Ibn 'Abd al-Salam, however, refuted him by pointing out that one of 
the transmitters in Ibn Majah's isndd was a known liar (i.e., Ya'qub b. 
al-Walld al-Madlnl)."' 

Although by the time of al-Maqdisi in the early sixth/twelfth century 
many scholars in the Islamic heartlands considered Ibn Majah's Sunan 
to be part of the well-respected "Six Book" hadlth canon, the work 



al-Bukharl's is correct in his Shark of Muslim; al-NawawI, al-Taqnb, 10; idem, Sharh Sahih 
Muslim, 1:145. Ibn Daqlq eiFectively favors Muslim's stance; Ibn Daqlq, al-Iqtirdh, 207. 
Ibn Jama'a favors Muslim's stance; Badr al-Din Muhammad b. Ibrahim Ibn Jama'a, 
Manhal al-rdimji 'ulum al-haMth al-nabawi, ed. Muhammad al-Sayyid Nuh (Mansoura, 
Egypt: Dar al-Wafa', 1402/1981), 175. As does the Hanafi al-Faslh nl-Harawi, Jawdhir 
al-usul, 29. The later Hanaft MuUa 'All Qarl also favors Muslim's school; MuUa 'All 
Qarl, Shark Musnad Abi Hanifa, ed. Khalll Muhyl al-Din Malls (Beirut: Dar al-Kutub 
al-'Ilmiyya, [n.d.]), 10. Al-Hakim does not address the issue of requiring a meeting; 
al-Hakim, Ma'rifat 'ulum al-hadith, 43-4. For more modern analyses of this debate, see 
al-LaknawI, Z'ifo^ al-amdni, 235-40; Khaldun al-Ahdab, Asbdb ikhtildf al-muhaddithin, 
2 vols, geddah: Dar Kunuz al-'Ilm, 1422/2001), 1:179-96; al-Sharif Hatim al-'Awnl, 
Ijmd' al-muhaddithin. 

"^ See al-Safadi, al-Wdfi bi'l-wafaydt, vol. 18, ed. Ayman Fu'ad Sayyid (Wiesbaden 
and Beirut: Steiner Verlag, 1408/1988), 18:520-2. 

" Al-AlbanI and Muhammad Zahir al-ShawIsh, eds., Musdjala 'ilmiyya bayn al-imamayn 
al-jalilayn al-'Izz Ibn 'Abd al-Saldm wa Ibn al-Saldh (Damascus: al-Maktab al-IslamI, 
[I960]), 5. 

'" Al-AlbanI et al, Musdjala 'ilmiyya, 17. 

'^ Al-AlbanI et al, Musdjala 'ilmiyya, 32. 



THE CANON AND THE NEEDS OF THE COMMUNITY 217 

could not deliver the decisive authority of the Sahihayn. A rigorous critic 
like al-Daraqutnl had disapproved of only two hundred and seventeen 
narrations from al-Bukharl's and Muslim's books and only two of their 
narrators. Al-Dhahabi, however, counted no less than one thousand 
weak narrations from the approximately 4,341 hadiths in Ibn Majah's 
Sunan.-^ Ibn 'Abd al-Salam was thus on much steadier ground when 
he cited a hadith from Sahih Muslim to support his position.^' Given 
the possible implications of choosing one collection over another for 
takhnj in a debate, it is not surprising that scholars in Baghdad asked 
al-Maqdisi to write a book explaining the differing criteria of the Six 
Books. 22 



The Origins of Takhnj Among the Students of al-Hakim al-Naysabun 

In light of al-Hakim al-Naysaburi's leading role in the canonization of 
the Sahihayn, it seems natural that we find the first concerted applica- 
tion of this new measure of authenticity in the work of his students. 
The actual earliest known use of al-Bukharl and Muslim for the takhnj 
of hadiths, however, occurs in a small hadith collection compiled by a 
prominent member of the SaMhcyn Network who was both al-Hakim's 
teacher and senior colleague: al-Daraqutnl.^^ Another member of the 



^" Al-Dhahabi, Siyar, 13:279. For another instance in whicli tlie Sliafi'i Taj al-Din al- 
Subla confidently states tiiat a liadlth fi'om Ibn Majah is inauthentic, see his Tabaqdt, 4: 1 3 
(biography of al-Bayhaqi); also, Abu al-Fayd Ahmad al-Ghumarl (d. 1960), al-Mughir 
'aid ahddith al-Jdmi' al-saghir (Beirut: Dar al-Ra'id al-'Arabi, 1402/1982), 89-90. 

2' Al-Albanl et al., Musdjala 'ilmiyya, 8. 

22 Al-Maqdisi, Shurut al-a'imma al-sitta, 10. 

2' Al-Daraqutnl, Kitdbfihi arba'un hadith" min musnad Burayda b. 'Abdalldh b. Abi Burda 
'anjiddihi 'an Abi Musd al-Ashan, ed. Muhammad b. 'Abd al-Karim b. 'Ubayd (Mecca: 
Ma'had al-Buhuth al-'Ilmiyya, 1420/[2000]). I have found one earlier occurrence of 
takhnj, but I believe it to be a later addition to the text. In his work on the differences 
of opinions amongst jurists, Ibn al-Mundhir (d. 318/930-1) cites a hadith and then 
says "akhrajahu al-Bukhdn iva Muslim." This is probably a later addition, since in the 
early fourth/tenth century people did not generally refer to al-Bukharl as such (if they 
referred to him at all), calling him Muhammad b. Isma'll or Abu 'Abdallah. Using 
'al-Bukharl' as shorthand was a result of the mustakhraj period, and no mustakhraji of 
al-Bukharl had been produced during Ibn al-Mundhir's time; Muhammad b. Ibrahim 
Ibn al-Mundhir, al-Ishrdf 'aid madhhab ahl al-'ilm, ed. Muhammad Sa'ld Mubayyad (Idilb, 
Syria and Doha, Qatar: Maktabat al-Ghazall and Maktabat Dar al-Fath, 1415/1994), 
96. The early Hanaft hadith scholar Abu JaTar al-TahawI (d. 321/933) also mentions 
that al-Bukharl narrated a hadith. This hadith, however, does not appear in the Sahih, so 
al-TahawI was probably referring to al-Bukharl's Tdnkh al-kabir, which he cited several 
times in his works; Abu Ja'far Ahmad al-TahawI, Mushkil al-dthdr, 25 vols. (Hyderabad: 



218 CHAPTER SIX 

Baghdad knot followed closely on al-Daraqutni's heels. At several points 
in his Shark usul i'tiqdd ahl al-sunna, HibataUah al-Lalaka'i (d. 418/1027-8) 
adduces hadlths as evidence and then supports them by stating that 
al-Bukhari and/or Muslim included them [akhrajahu) in their Sahih'-,}* 

The takhrij format was a natural outgrowth of the mustakhraj tech- 
niques of al-Daraqutnl's contemporaries and students such as al-Jawzaqi 
(d. 388/998) and al-BarqanI (d. 425/1033-4). Like the mustakhraj, takhrij 
functioned to display the quality of a scholar's hadiths. Instead of fol- 
lowing the format of other mustakhraj authors like Abu 'Awana or Abu 
Nu'aym al-Isbahani, who simply replicated the template collection with 
their own isnads,, al-Jawzaqi's and al-Barqani's joint Mustakhrqjs of the 
Sahihayn list their authors' narration of a hadith and then note that 
al-Bukhari, Muslim or both "included it [akhrajahu).'"'^-' Takhrij simply 
involved using this tactic when composing other books. 

The use of al-Bukhari and Muslim to consistently and confidently 
affirm the authenticity of hadlths or the reliability of transmitters, how- 
ever, can be traced to two of al-Hakim's students: Abu Ya'la Khalil b. 
'Abdallah al-Khalili (d. 446/1054) and Abu Bakr Ahmad b. al-Husayn 
al-BayhaqI (d. 458/1066). Al-Khalili employed the Sahihayn as a tool 
for establishing the reliability of transmitters in his short but valuable 
biographical dictionary of hadith scholars, al-Irshdd ji ma'rifat 'ulamd' 
al-hadith (Guidance for Knowing the Scholars of Hadith). Al-KhalTlT 
hailed from Qazvin, where he worked for a time as a judge, but studied 
extensively with al-Hakim in Naysabur From among the other members 
of the Sahihiyn Network, he only studied with al-Ghitrifi.^*' His link 
to the Jurjan cult of al-Bukhari might explain his favoring al-Bukhari 
over Muslim as a source for citation. His admiration for al-Bukhari is 
clear, for he calls him "the imam agreed on by all without contest. "^^ 



Da'irat al-Ma'arif al-'Uthmaniyya, [1968]), 1:278-9. For this citation, I am indebted to 
the extremely useful study by 'Abd al-Majld Mahmud, Abuja'far al-Tahdwi wa atharuhu 
fial-hadith (Cairo: al-Maktaba al-'Arabiyya, 1395/1975), 119, 228-9. 

^' Al-Lalaka'l, Shark usul i'tiqdd ahl al-sunna, 1:108 (for al-Bukharl), 1:87, 4:876 (for 
al-Bukharl and Muslim), 1 :85 (for Muslim). On one occasion "al-Bukharl included it . . ." 
is added in the margin by a later copyist. That this addition is noticable bolsters the 
reliability of the remaining instances as parts of the author's original work. 

^ ' See al-Barqanl, al-Juz ' al-awwal min al-takhnj li-saMh al-hadith; Abu Bakr Muhammad 
b. ' Abdallah al-jawzaqi, al-Jam ' bayn al-Sahihayn, MS 118 Awqaf, Khizana al-' Amma 
Library, Rabat. 

^^ Al-Rafi'l, al-TadwinJi akhbdr Qazwin, 2:501-4; al-Dhahabi, Tadhkirat al~huffdz, 3:214; 
idem, Tdrikh al-isldm, 30:120-1; idem, Siyar, 17:666-8. 

2' Al-Khallll, al-Irshdd, 'ill. 



THE CANON AND THE NEEDS OF THE COMMUNITY 219 

Al-Khalili introduces at least nineteen men as transmitters al-Bukharl 
included in his Sahih. He cites another eighteen as transmitters from 
both the Sahihayn. He relies on Muslim's Sahih independently only twice, 
however, and mentions no other works as a means of takhrij. 

Using al-Bukhari and Muslim as a measure of authenticity for 
hadiths began in earnest with Abu Bakr al-BayhaqI, who was well- 
known as one of al-Hakim's most senior students. When later scholars 
such as Ibn al-jawzl and Ibn al-Salah cited al-Hakim's opinions or his 
works, it was most frequently through a chain of transmission from al- 
Bayhaql. Al-Hakim provided one of al-Bayhaqi's primary reservoirs of 
hadiths, since, according to al-Dhahabi, he did not have the books of 
al-Tirmidhl, Ibn Majah or al-Nasa'l at his disposal. He did, however, 
possess a camel load of hadlth books from al-Hakim. In addition to 
al-Hakim, he also studied extensively with Abu Ishaq al-Isfarayini, 
al-Barqani and Ibn Furak, who served as another major source of al- 
Bayhaqi's hadiths.^" 

Al-BayhaqI was an amazingly prolific scholar, who, according to al- 
Dhahabl, was capable of founding his own madhhab had he so wished. 
Instead, al-Bayhaqi authored an oeuvre that became such a bastion 
of the Shafi'i school that Imam al-Haramayn al-Juwayni considered 
al-BayhaqI to be the only person to whom al-Shafi'l was indebted. 
Al-BayhaqI organized al-Shafi'i's statements and proof texts in the 
massive Ma'rifat al-sunan wa al-dthdr and then compiled his al-Sunan 
al-kubrd, a huge hadlth collection backing up every detail of Shafi'i 
substantive law with Prophetic traditions as well as opinions from the 
Companions. Al-Bayhaqi was sought out as an expert on ^YiaH'i fiqh 
and al-Muzani's MukhtasarP Both later Shafi'i/Ash'aris and Hanball/ 
iiber-Sunnis respected and relied on his work. The staunch Ash'arl 
Ibn 'Asakir heard his entire oeuvre from his students, and the Hanbali 
Kh"aje 'Abdallah had ya^as from him.^" 

Al-Bayhaqi's output was representative of the new Shafi'l/Ash'ari 
orthodoxy. Works such as his al-Madkhal ild al-Sunan al-kubrd (Introduction 
to the Great Sunan) and the Sunan itself champion the Shafi'l transmis- 
sion-based legal methodology and the school's body of substantive law. 
In works like his Khildjiyydt (The Disagreements), al-Bayhaqi defends the 



2" Al-Dhahabi, Siyar, 18:165. 

2' 'Abd al-Ghafir al-FarisI, Tdnkh Najsdbur al-muntakhab min al-Siydq, 1 2 7-8 

™ Al-Dhahabi, Tdnkh al-iddm, 30:438-41; idem, Siyar, 18:163-70. 



220 CHAPTER SIX 

school's positions against its Hanafi opponents. He affirms the transmis- 
sion-based trust in the revealed text of the sunna for understanding 
dogma, while simultaneously validating Ash'ari efforts to interpret God 
and His attributes rationally. Discussing the hugely divisive controversy 
over the wording {lafz) of the Qur'an, for example, he states simply that 
all transmission-based scholars believe that the Qur'an is the uncreated 
word of God. While some scholars might prefer not to discuss the 
issue, others like al-Bukhari (and al-Bayhaqi himself) have chosen to 
distinguish between the physical manifestation of the Qjur'an and the 
text itself Nonetheless, all belong to the same unified school.^' 

We can clearly appreciate the manner in which al-Bayhaqi employed 
the Sahihayn as a measure of authenticity in a sample of four works 
intended to affirm his Shafi'l/Ash'ari position. Stylistically, his use of 
the phrase "al-Bukharl and/or Muslim included it" after a hadlth 
reflects the works of his teacher al-Barqani and also that of al-Lalaka'i. 
Beginning with the first hadith in his Kitab al-Asma' wa al-sifit, a trea- 
tise on God's names and attributes, and thereafter wherever possible, 
al-BayhaqI uses inclusion in al-Bukharl's and Muslim's collections to 
establish reliability^^ He pursues the same tactic in his KhilajiyyatP In 
a work intended to provide hadlths proving the existence of the bete 
noire of Muslim rationalists, the punishment of the grave {'adhdb al-qabr), 
al-Bayhaqi uses the canonical formula "al-Bukhari and/ or Muslim 
included it (akhrajahu)" for eighty-eight out of the four hundred and 
thirty (20%) narrations in the book. He only twice mentions other 
collections such as Abti Dawud's Sunan and Ibn Hanbal's Musnad."'* 
Al-Bayhaqi's al-Sunan al-kubrd represents the most extensive use of the 
Sahihayn canon for takhrij. In a sample of the 1,472 narrations consti- 
tuting his lengthy chapter on ritual purity (tahdra), al-Bayhaqi refers to 
inclusion by al-Bukhari, Muslim or both 23.5% of the time. The only 
other work he refers to for takhrij, Abu Dawud's Sunan, appears only 
0.6% of the time (9 instances). 



" Abu Bakr al-Bayfiaqi, Kitab al-asma' wa al-sifit, ed. 'Abdallah b. Muhammad al- 
Hashidl, 2 vols. (Jedda: Maktabat al-Sawadi, 1413/1993), 2:17. 

'^ Abu Bakr al-BayhaqI, Kitab al-asma' wa al-sifit, 1:17-18. 

'^ Abu Bakr al-BayhaqI, al-Khilafyydt, ed. Mashhur b. Hasan Al-Salman, 2 vols. 
(Riyadh: Dar al-Sumay'l, 1415/1995), 1:48. 

^' See Abu Bakr al-BayhaqI, Ithbdt 'adhdb al-qabr, ed. Sharaf Mahmud al-Qudat 
(Amman: Dar al-Furqan, 1403/1983). 



THE CANON AND THE NEEDS OF THE COMMUNITY 221 

Another student and follower of al-Hakim's school of thought, Abu 
Nu'aym al-Isbahani, also provides some of the earliest usages of the 
Sahihayn canon as a measure of authenticity. In his biographical dic- 
tionary of Isfahan, Dhikr akhbdr Isbahdn, he uses the phrase "the hadith 
is authentic by agreement [al-hadith sahih muttafaq 'alayhi)" to validate 
his own narration of a Prophetic hadith. ^^ Here he follows an earlier 
member of the Sahihayn Network, Ibn al-Akhram, who had entitled 
his joint mustakhraj of the Sahihoyn "The Sahih by Agreement [al-SaMh 
al-muttafaq 'alayhi)."^''-' In his landmark biographical dictionary of Sufism 
and asceticism, Hilyat al-awliyd', Abu Nu'aym also uses al-Bukhari and 
Muslim as direct stamps of approval for hadiths he includes in the 
work's entries.^' 

We know that employing the canon for takhnj had also begun in 
Baghdad by the mid-fifth/eleventh century. Abu Nu'aym's student 
and a main inheritor of the Sahihayn Network (see Soh^oyn Network 
Chart), al-Khatib al-Baghdadi, used the Sahihayn canon dramatically to 
establish the authenticity of a selection of 173 of his hadiths that he 
narrated in a hadith dictation session. He invokes the inclusion of al- 
Bukharl, Muslim or both for 57% of his reports. He invokes no other 
work for takhnj, and only declares one hadith to be sahih that does not 
appear in one of the Sahihoyn?^ Al-Khatib reiterates the paramountcy 
of the Sahihayn in his vision of the hadith sciences when he instructs 
students that the two works should form the basis of any curriculum 
in hadith study. ^^ 



'"' Abu Nu'aym al-Isbahani, Tankh Ishahan /Dhikr akhbar hbahan, ed. Sayyid Khusrawi 
Hasan, 2 vols. (Beirut: Dar al-Kutub al-'Ilmiyya, 1410/1990), 1:21. We know that Abu 
Nu'aym used the term 'muttafaq 'alayi' to refer to al-Bukharl's and Muslim's agree- 
ment because he uses it in the midst of critiquing several transmitters whom he says 
al-Bukharl and Muslim did not use in their Sahihs; al-Dhahabi, Mizdn al-i'tiddl, 1:166 
(bio of Ahmad b. Yusuf al-Manbiji). 

''' Al-Dhahabi, Tadhkirat al-hujjai, 3:55. 

'' See, for examples, Abtl Nu'aym al-Isbahani, Hilyat al-awliyd' iva tabaqdt al-asfiyd', 
10 vols. (Cairo: Maktabat al-Khanji and Matba'at al-Sa'ada, [1351-1357/1932-1938]), 
3:205 (al-Bukharl), 8:261 (Muslim). 

'° See al-Khapb al-Baghdadi, al-Fawd'id al-muntakhaba al-dhdh iva al-ghard'ib, ed. Khalil 
b. Muhammad al-'Arabi (Giza: Maktabat al-Taw'iyya al-Islamiyya, 1415/1995). See 
p. 206 for the one instance. 

'' Al-Khapb, al-Jdmi' li~ikhtildf al-rdwi iva dddb al-sdmi', 2:185. 



222 CHAPTER SIX 

The Historical Application of Takhrij 

We have located both the epicenter of the Sahihayn canon and its initial 
use as a measure of authenticity in the seminal work of al-Hakim al- 
Naysaburi and his students from the Shafi'l school. We will now examine 
how and when the canon spread to the Hanbali, Maliki, Hanafi and 
ImamT Shiite schools. We will focus on the two most salient means 
in which scholars used the Sahihayn canon as a common measure of 
authenticity: polemics, and employing the canon to fortify a school's 
formative legal or hadlth texts. 

a. Polemics and Debate 

In the mid-fifth/ eleventh century, prominent adherents of the Shafi'l, 
Hanbali and Maliki schools all began employing the SaMhoyn canon as 
a measure of authenticity in polemics and expositions of their schools' 
doctrines. It was not until the eighth/fourteenth century, however, that 
the Hanafis also adopted the canon for this use. 

Al-Bayhaqi's categorical reinforcement of the Shafi'l/Ash'ari catalog 
stands out as both the earliest and most stunning application of the 
canon in his school's history. It seems clear, however, that this intensive 
recourse to the Sahihayn hinged on al-Bayhaqi's proximity to al-Hakim 
and the canonization of the two works. Although other Shafi'i jurists 
of this period did employ the Sahihayn canon, no one matched the 
concentrated use found in al-Bayhaql's or al-KhalTlT's works. Abu al- 
Hasan al-Mawardi (d. 450/1058), for example, was a contemporary 
member of the Shafi'l school in Baghdad who was also engaged in the 
process of explicating and establishing Shafi'i substantive law. However, 
he made very limited use of the Sahihayn canon for takhnj in his legal 
reference, al-Hdwi al-kabir fi jiqh madhhab al-imdm al-Shdji'T (The Great 
Compendium of the Shafi'l School of Law). On only two occasions in 
his voluminous explanation of the school's law does he use inclusion 
in al-Bukharl's or Muslim's collections to support the authenticity of 
hadiths that al-Shafi'i had invoked as proof texts.*" 



'" See Abu al-Hasan 'Ali b. Muhammad al-Mawardi, al-Hawi al-kahr fi fiqh madhhab 
al-imdm al-Shdfi'i, ed. 'All Muhammad Mu'awwad and 'Adil Ahmad 'Abd al-Mawjud 
(Beirut: Dar al-Kutub al-'Ilmiyya, 1414/1994), 1:140; 17:71. 



THE CANON AND THE NEEDS OF THE COMMUNITY 223 

It is not surprising that one of the earliest employers of the Sahihayn 
as a measure of authenticity came from the Hanball camp, which 
cooperated with the Shafi'i/Ash'aris in canonizing the two works. Like 
his correspondent, Abu Nasr al-Wa'ill, the great Hanbali Abu Ya'la Ibn 
al-Farra' (d. 458/1066) was an inveterate opponent of the Ash'arls and 
their figurative interpretation of God's attributes. Like al-Bayhaqi, how- 
ever, he used the canon to bolster the authority of the hadiths he cited 
as proof texts on such controversial issues. In 456/1064, Ibn al-Farra' 
held a session for dictating hadiths to students [majlis imld') and tackled 
the perennially divisive issue of seeing God on the Day of Judgment 
{ru'yat al-Bdri'), rejected by Mu'tazilites and interpreted figuratively 
by Ash'aris. He narrated a hadith in which the Prophet looks at the 
full moon and then tells his followers, "Indeed you will see your Lord 
with your own eyes {'iydn'")." Ibn al-Farra' adds, "This hadith is sahih; 
al-Bukhari included it..., and it is as if I heard it from al-Bukhari."*' 
Here Ibn al-Farra' uses both his own proximity in the isndd to al-Bukhari 
and the latter's inclusion of the hadith in his Sahih as a means for 
augmenting its authority. In his treatise on legal theory, al-'Udda, Ibn 
al-Farra' similarly uses al-Bukharl's Sahth to validate a report proving 
that a five-year-old could effectively hear hadith transmitted.*^ 

Ibn al-Farra' also utilizes the canon in his work on issues of dogma 
(usul al-din), the Kitdb al-mu'tamad. The author devotes his attention in 
this work primarily to his Mu'tazili and Ash'ari opponents, treating 
controversial topics such as God's attributes, the punishment of the 
grave, and the issue of appropriate rule in Islam (imdmd). In his sub- 
chapter on the existence of magic (sihr), he argues against the Mu'tazila, 
saying that both the Qjur'an and the hadith affirm it. He invokes the 
hadith in which 'A'isha recounts how a Jewish sorcerer once cast a 
spell on the Prophet, adding that "this is a well-known [mashhur) hadith 
that al-Bukharl and others from the hadith scholars {muhaddithin) have 
mentioned."*' He also mentions that some hadiths are "included in the 



" Ibn Abl Ya'la, Tabaqdt al-handbila, 2:172; Fath # 7435; Safiih al-Bukhdri: kitdb al- 
tawhld, bdb 24. 

*^ Ibn al-Farra', al-'Udda, 3:950. This is the hadith from the Companion Mahmud 
b. Rabi' saying, "'Aqaltu min al-Nabi (s) majjaf" majjahd fi wajhi wa and ibn khamas sinin"; 
Fath #77; Sahih al-Bukhdri: kitdb al-'ilm, bdb matdyasihhu samd' al-saghir. 

*' Abu Ya'la Ibn al-Farra', Kitdb al-mu 'tamadfi uml al-din, ed. Wadi' Zaydan Haddad 
(Beirut: Dar al-Mashriq, 1974), 168. This specific version of the hadith "sahara al-nabi 
(s) yahudi min al-yahud. . .," appears in Sahih Muslim, see Sahih Muslim: kitdb al-saldm, bdb 



224 CHAPTER SIX 

Sahih,'' a phrase that generally denotes inclusion in one or both of the 
Sahihayn (here it evidently refers to Muslim's work).** Besides al-Bukhari, 
he only once mentions another hadith scholar as narrating a report, 
namely al-Daraqutni; in this case, however, he places no emphasis on 
the source as a guarantor of authenticity. Ibn al-Farra"s son, Ibn Abl 
Ya'la, also occasionally uses al-Bukhari and Muslim as a measure of 
authenticity in his discussion of the differences between Hanballs and 
Ash'aris on issues such as God's attributes.*^ This use of the canon 
continues in later Hanbali works such as Ibn 'Aqll's (d. 513/1119) al- 
Wadihfi usul al-jiqh, until the end of the sixth/twelfth century. **■ 

Among Hanballs, it was the Neo-Hanbalite cadre of Ibn Taymiyya 
(d. 728/1328) and his student Ibn Qayyim al-Jawziyya (d. 751/1350) 
that exhibited the most cunning and aggressive usage of the Sahihayn 
canon. The two works served as powerful weapons in polemics against 
Ash'arls over issues such as God's attributes, the nature of the Qur'an 
and invoking the intercession of dead saints. Asserting the literalist posi- 
tion that one should accept the outward meaning of Qur'anic verses or 
Prophetic hadiths describing God's movements, Ibn al-Q_ayyim calls the 
attention of his Ash'arl opponents to al-Bukharl's narrations of hadiths 
asserting that God is indeed physically above us in the heavens. He 
exploits al-Bukharl's position of extreme respect among both Ash'aris 
and Hanbali/iiber-Sunnis to his advantage, sarcastically implying that 
his opponents would condemn this venerable figure as an anthropo- 
morphist. Ibn al-Qayyim states in a verse of poetry: 

And from among you, al-Bukhari the 'anthropomorphist' has narrated 
it, Nay, an anthropomorphist who attributes to God a [physical] position 
above us {mujassim fawqdni).*^ 

On the issue of visiting the graves of prophets and seeking their 
assistance, Ibn al-Q_ayyim challenges the orthodox tenet that they are 



al-sihr. A slightly dilFerent wording appears in Sahih al-Bukhan, see Sahih al-Bukhan: kitab 
al-tibk hdb 47 / Path # 5763. 

''* Ibn al-Farra', Kitdb al-mu'tamad, 224; Sahih Muslim: kitdb al-imdra, bdb al-istikhldf 
wa tarkihi. This hadith goes as follows: 'Abdallah b. 'Umar <- 'Umar b. al-Khattab: In 
atruku fa-qad taraka khayr minni, rasul Alldh, wa in astakhlifu fa-qad istakhlafa man huiva khayr 
minni, ya'ni Aba Bakr. Ibn al-Farra"s version inverts Muslim's word order 

" Ibn Abl Ya'la, Tabaqdt al-handbila, 2:182. 

^ See, for example, Abu al-VVafa' 'All Ibn Aqll, al-Wddih fi usUl al-fiqh, ed. George 
Makdisi (Wiesbaden and Beirut: Steiner Verlag, 1423/2002), 3:191; 4b:200, 436. 

" TaqI al-Din al-Subkl, al-Sayf al-saqil, 65. 



THE CANON AND THE NEEDS OF THE COMMUNITY 225 

indeed alive in their graves and able to respond to the invocation of 
pilgrims.*" One of the hadiths that scholars had produced as evidence 
for this stance describes Moses praying in his grave. Ibn al-Q_ayyim, 
however, argues that al-Bukharl's decision to exclude the hadith from 
his Sahih demonstrates its weakness, as does al-Daraqutnl's claim that 
it is actually the opinion of a Companion (hence, mawquf).*'^ Not only 
does Ibn al-Qayyim use al-Bukharl as a measure of truth to reinforce 
his position, he also exploits exclusion from the work to undermine his 
opponent's evidence. 

Like others, Malikis employed the Sahihayn canon in debates or 
expositions of their school's positions. It is little surprise that the first 
Maliki to employ the Sahihayn canon as a measure of authenticity had 
studied extensively at the hands of a member of the Sahihiyn Network, 
Abu Dharr al-HarawT. Abu al-Walid al-Baji (d. 474/1081) of Cor- 
dova traveled east in 426/1035 and studied with al-HarawI for three 
years in Mecca before moving to the Abbasid capital to study with 
al-Khatib al-Baghdadi and others.'" With such prolonged exposure to 
one of the most prominent members of the SaMhoyn Network, al-Bajl 
confidently employed the canon in his book defending Maliki usul, the 
Ihkdm al-Jusulfi ahkam al-usul. This work is an aggressive exposition of 
Maliki legal theory, often targeting Hanafi or iiber-Sunni opponents. 
Although al-Bajl makes only a few references to al-Bukhari or Muslim, 
or any other hadith collections for that matter, these references clearly 
illustrate the function of the Sahihayn canon in the author's thought.^' 
One of al-Baji's primary concerns in the Ihkdm is mounting a defense 
of analogical reasoning (qiyds) against those iiber-Sunnis who reject 
any rulings not based directly on revealed text (nass). He lists the 
various Prophetic reports that his opponents cite as evidence against 
the use of reason, but rebuts them by stating that these are defective 
and too unreliable to be compelling. He asks his opponents how they 
could invoke such feeble hadiths in the face of the reports that he had 
advanced as evidence, "most of which the two imams [al-Bukharl and 
Muslim] have agreed on including in the Sahih[a)in] ." "This is what the 



'* For a discussion of Ibn Tayniiyya's and Ibn al-Qayyini's argument against visiting 
graves, and an Ash'arl response, see Taylor, In the Vicinity of the Righteous, 168-94. 

"" TaqI al-Dln al-Subkl, al-Sayf al-saqil, 155. 

^° D.M. Dunlop, "al-Badjl, Abu al-Walld," EF . 

^' For these instances, see Abu al-Walld al-Bajl, Ihkdm al-fusul fi ahkam al-usul; 591, 
744. 



226 CHAPTER SIX 

people have agreed on as authentic," he adds, noting that only one of 
his opponents' hadlths appears in the Sahihayn."'^ 

Abu al-Walid al-BajT's al-Muntaqa, a commentary on the Muwatta', 
shares many of the same concerns as his usul work. Although it pri- 
marily seeks to explain and elaborate on the positive law laid out by 
Malik, the author's perspective is consistently both comparative and 
polemical. He is as eager to prove the correctness of Malik's school as 
to explain it. Al-Baji thus occasionally relies on the Sahihayn to validate 
Malik's legal positions. Defending his stance against Hanafi opponents 
on the necessity of the taslim (turning one's head and saying 'peace 
be upon you' at the end of prayer) for exiting a prayer, al-Bajl states, 
"The proof of the correctness {sihhd) of Malik's position is [a hadlth] 

that al-Bukhari narrated " He also employs the canon conversely to 

cast doubt on the authenticity of opposing hadlths. He rejects reports 
that offer more information on the Prophet's taslim than those found 
in the Muwatta' by stating, "Al-Bukhari did not include any of them, 
and what Muslim included are reports that allow for interpretation 
[yahtamilu al-ta'wil)."^'^ 

The Hanafi school seems to have been much slower to adopt the 
Sahihayn canon as a measure of authenticity. Although, as we discussed 
in Chapter Four, Hanafi scholars played an active role in transmitting 
al-Bukhari's and Muslim's collections during the fifth /eleventh and 
sixth/twelfth centuries, they did not develop the strong interests in 
studying or utilizing the two works demonstrated by the Shafi'i Sahihayn 
Network or later scholars like al-Bayhaql. The earliest Hanafi scholar- 
ship on the works of al-Bukhari and Muslim appears in the seventh/thir- 
teenth century with the pioneer of Indian Islamic scholarship, al-Hasan 
b. Muhammad al-SaghanI (d. 650/1252), who produced a combined 
edition of the Sahihayn, a commentary on al-Bukharl's Sahih and a 
work on his transmitters. The Damascene Hanafi Abu al-Hafs 'Umar 
b. Badr al-Mawsili (d. 622/1225) produced a simplified digest of the 
Sahihayn, and Muhammad b. 'Abbad al-KhilatI (d. 652/1254) devoted 
a book to Muslim's collection.-^* It was not until the eighth/fourteenth 



■^^ Al-Bajl, Ihkdm al-fusulji ahkdm al-usul, 610. 

^' Al-Bajl, al-Muntaqd shark al-Muwattd', 7 vols, in 4 ([Cairo]: Dar al-Fikr al-'Arabi, 
[1 982]), 1 : 1 69. For an extensive discussion of the tastm in early works of law and hadlth, 
see Yasin Button, "An Innovation from the Time of the BanI Hashim: Some Reflections 
on the Taslim at the End of the Frayer," Journal of Islamic Studies 16 (2005): 147-8. 

'■' Ibn Abl al-Wafa', al-Jawdhir al-mudiyya, 3:180. Al-Mawsill's work is published as 



THE CANON AND THE NEEDS OF THE COMMUNITY 227 

century, however, that Hanafis began using the Sahihayn to vaHdate 
hadlths. Writing in the Chagataied and Ilkhanid Mongol realms of 
Iran and Central Asia, 'Ala' al-Din 'Abd al-'AzIz b. Ahmad al-Bukharl 
(d. 730/1329-30)''^ employs them briefly but effectively in his Kashf al- 
asrar (Revealing the Secrets), a commentary on the Hanaft usul treatise 
by Abu al-Hasan 'All b. Muhammad al-Bazdawi of Samarqand (d. 
482/1089). Responding to criticisms that one of the transmitters of a 
hadlth he uses was weak, 'Abd al-'AzIz retorts that al-Bukharl "is a pillar 
to be followed in that science [of hadlth] , the imam of that craft, so his 

including that [hadlth] suffices as proof of its authenticity [sihha) "'*' 

The author thus leaves his readers no doubt about the legitimating 
power of al-Bukharl's Sahih. In general, however, 'Abd al- 'Aziz's Kashf 
al-asrar makes very limited use of the Sahihayn in this manner 

By the time al-Bayhaqi and Ibn al-Farra' were putting the SaMhoyn 
canon to use as a measure of authenticity, Imami Shiism had taken 
crucial steps in articulating its doctrine and outlining its sources. In 
329/940 the twelfth imam's absence was declared permanent, and 
leadership in the community fell into the hands of scholars pending 
the imam's, return. The collections that would become the Imami hadlth 
canon had all been produced: Muhammad b. Ya'qub al-Kulayni's (d. 
329/940) al-Kafi, Ibn Babawayh's (d. 381/991) Man la yahduruhu al- 
faqih and Muhammad b. al-Hasan al-Tusl's (d. 460/1067) two works, 
al-Tahdhib and al-IstibsarJ'^ 

In the same period, tensions between Imam! Shiites and Sunnis 
rose markedly with the rise of Fatimid Isma'ilT power in Egypt and 
Syria, the terror wreaked by the Isma'ill assassins, and the impending 
threat of the sect's missionary activities in the central Islamic lands 
of the Seljuq empire. For the Imami Shiite minorities living in the 
Karkh district of Baghdad or in the great Iranian cities of Rayy and 
Naysabur, identification with the Isma'ili threat presented a constant 
danger Imami scholars like Nasir al-Din Abu al-Rashid b. 'Abd al-Jalll 



al-Jam' bayn al-Sahihayn, ed. Salih Ahmad al-Shami, 2 vols. (Beirut: al-Maktab al-Islami, 
1416/1995). 

^^ For his biography, see Ibn Abi al-Wafa', al-Jawdhir al-mudiyya, 2:428. 

'''' Al-Ansarl, Path al-bdqi, 76. 

'" Stewart, Islamic Legal Orthodoxy, 5. For a discussion of the contents and uses of 
the canonical Shiite hadlth collections, see Robert Gleave, "Between Hadith and Fiqh: 
the 'Canonical' Imami Collections of Akhbdr," Islamic Law and Society 8, no. 3 (2001): 
350-382. 



228 CHAPTER SIX 

Abl al-Husayn Qazvini (d. ca. 560/1165) thus expended great efforts 
in trying to both defend ImamI doctrine in the face of Sunni critiques 
and educate Sunnis on the important differences between their own, 
ImamI school and the Isma'ilis. 

ImamI Shiites like Q_azvlnl did not identify with Sunni hadlth col- 
lections at all, for they considered the Companions on whom collectors 
like al-Bukharl had relied most heavily, such as Abu Hurayra, to be 
brazen liars.'''' Nonetheless, the authority commanded by the Sahihayn 
within the Sunni community provided QazvInl with an important tool 
for defending his school. His Ketab-e naqd (The Refutation) represents a 
comprehensive effort to validate ImamI doctrine and practice in Sunni 
eyes as well as to educate his readers on the trenchant differences 
between ImamI and Isma'lll Shiites. Qazvinl frequently cites famous 
Sunni works such as al-Tabarl's TafsTr as proof texts, obliging Sunnis to 
heed "one of their own imdms."^^ In response to Sunni accusations that 
Shiites rely on weak hadlths and lies, he says that they are narrated via 
reporters who are mostly "Sunnis" and "Hanafis" and are to be found 
in the books of these "two sects {fanqayn)." Qazvlnl adds that the Sunni 
hadlth scholars {ashab al-haditii) accept many of these reports.''" 

Q_azvlnl often refers to the consensus (ijmd') of the umma and of 
the hadlth scholars in his arguments for Shiite stances. ''' Responding to 
Sunni criticisms of Shiite claims that 'All was the first person to ever 
have that name, he invokes as evidence the Sahihayn and other books 
of the ashab al-hadith that "are relied upon [keh mo'tamad-ast)." Q_azvlnl 
tells his opponents to "take up the Sahihayn" and find the hadlth that 
says that 'All's name is written on the leg of God's throne and on the 
doorway to Paradise as the brother of Muhammad. Since both these 
structures existed before the creation of the world, 'All is doubtless the 
first person to have been so named. ''^ 



'''' For a Shiite study of Abu Hurayra, see 'Abd al-Husayn Sharaf al-Din al-Musawi, 
Abu Hurayra (Beirut: Dar al-Zahra', 1397/1977). 

™ Nasir al-Dln 'Abd al-Jalll Abu al-Husayn Qazvinl Razi (fl. 560/ 1 1 62), Kitdb-e naqd-e 
ma 'refat beh ba 'd-e mathdleb al-navdseb fi naqd ba 'dfadd 'eh al-ravdfed, ed. Jalal al-Dln Hosaynl 
Ormavl ([Tehran]: Chap-khane-ye Sepehr, 1331-1371/[1952]), 392. 

^ Nasir al-Din Qazvinl, Ketdb-e naqd, 654-5. 

" For example, see Nasir al-Dln Qazvinl, Ketdb-e naqd, 557. 

''- Nasir al-Din Qazvinl, Ketdb-e naqd, 576-8. Neither of these two hadlths actually 
appears in the Sahihayn or the other Six Books: "I saw on the night I was taken up to 
the heavens, inscribed on the leg of the throne and the doorway of Paradise, 'The 
garden of Eden was planted by the hands of Muhammad, the purest of My creation, 
and I have supported him with 'All' {ra'aytu laylat usriya biild al-samd' muthabbat"'ald sdq 



THE CANON AND THE NEEDS OF THE COMMUNITY 229 

The Sahihayn and other respected Sunni hadith collections also 
provided the later ImamI theologian of Baghdad, Radi al-Din 'All b. 
Musa Ibn TawQs (d. 664/1266), with authoritative proof texts to use 
against Sunnis. In his study of Ibn Tawus's library, Etan Kohlberg 
states that he possessed copies of the Sahihayn "for polemical pro-Alid 

traditions included in them " He also relied on Muhammad b. Futuh 

al-Humaydfs (d. 488/1095) combination of al-Bukharl's and Muslim's 
collections, al-Jam' bayn al-Sahthqjin, as a more convenient source/'^ 

There can be no quantitative comparison between al-Bayhaqi's 
overwhelming employment of the Sahihayn canon to validate his hadlths 
and the more limited use of Ibn al-Farra', al-Mawardi, al-Bajl, 'Abd 
al-'Aziz al-Bukhari or Qazvini. In general, these scholars employed the 
Sahihayn canon only sparingly. Unlike al-BayhaqI and other students 
of al-Hakim, their work does not overflow with authorizing references 
to al-Bukhari and Muslim. As 'Abd al-'Aziz's reverential invocation of 
al-Bukharl's authority and al-Baji's explicit referral to the community's 
consensus on the Sahihayn demonstrate, however, these scholars were 
aware of the SaMhcyn canon's etiology and utility even if they only 
invoked it occasionally. 



b. Bolstering Formative Texts 

Although al-Bayhaqi had used the canon to comprehensively buttress 
Shafi'l substantive law in the mid-fifth/ eleventh century, the remaining 
three Sunni madhhabs followed very different paths in their recourse 
to the Sahihayn to bolster their formative hadith or legal texts. Their 
approaches to the canon for this purpose would depend on either the 
nature of their formative text or their attitude towards the Scihihiyn 
canon itself 



al- 'arsh wa bah al-janna an ghwisat jannat 'Adn hi-yadaj Muhammad safooati min khalq ayyadtuhu 
bi-'Ali)" and "It was written on the doorway to Paradise, 'There is no god but God, 
Muhammad is the Messenger of God, and 'All is the brother of Muhammad,' before 
God created the heavens and the earth by two thousand years (rnaktub 'aid bdb al-janna 
'la ildh Hid Allah Muhammad rasul Alldh 'Alt akhu Muhammad qabla an yakhluqa Alldh al- 
samdwdt wa al-ard bi-alfay'dm)." AI-Dhahabi includes permutations of both these reports 
in his work on criticized transmitters and their hadlths, Mizdn al-i'tiddt; al-Dhahabi, 
Mtzdn, 1:269, 530. 

"' Etan Kohlberg, A Medieval Muslim Scholar at Work (Leiden: Brill, 1992), 324-5. 



230 CHAPTER SIX 

It was only at the beginning of the seventh/thirteenth century that 
Hanball scholars like Ibn Qudama (d. 620/1223) started to seriously 
reinforce the hadiths used in elaborating their school's substantive law 
by takhrij through al-Bukharl, Muslim and other products of the sahih 
movement. In his commentary on the Hanbali formative legal text, 
al-Khiraqi's Mukhtasar, Ibn Qjudama mentions that one of his goals in 
explicating Ibn Hanbal's madhhab is the takhnj of the hadiths al-KhiraqI 
had used as proof texts. He states that he will cite them "from the books 
of the imams, from among the scholars of hadith, so that [these reports] 
might inspire trust in what they indicate, and to distinguish between 
the authentic and flawed [reports], so that what is well-established can 
be relied upon and what is unknown can be abandoned."" 

The task of undertaking takhry on the school's most prominent hadith 
collection, Ibn Hanbal's Musnad, daunted scholars for centuries. The 
sheer inertia of Ibn Hanbal's massive work has thwarted almost every 
scholarly attempt to systematically evaluate the authenticity of its con- 
tents or make the work more accessible. The Musnad, which consists of 
over forty thousand narrations (thirty thousand excluding repetitions), 
clearly contains a great deal of material that does not warrant a saktk 
rating. Discussions over its authenticity have thus generally revolved 
not around the question of whether the Musnad was totally reliable, 
but on whether or not its more lackluster narrations ever reached the 
level of fatal weakness or forgery. Because a systematic analysis would 
be a titanic feat, claims on this matter were often mere guesswork. Al- 
Dhahabl attempted to cast the Musnad in a good light by optimistically 
asserting that there are only a "few (qalil)" hadiths found in the Sahihayn 
that do not appear in the Musnad. He could not conceal the question- 
able status of the rest of the book's contents, however, and added that 
one should not take the Musnad'% contents as proof {^ujjd) because it 
has many reports that are too weak and even forged.*""' Ibn al-jawzl (d. 
597/1200) and Zayn al-Din al-'iraqi (d. 806/1404) also listed numerous 
hadiths from the Musnad that they believed were clearly forgeries. 



''' Ibn Qudama, al-Mughni, ed. 'Abdallah b. 'Abd al-Muhsin al-Turki and 'Abd al- 
Fattah Muhammad al-Halw, 15 vols. (Cairo: Hajr, 1406/1986), 1:5. 

1^' Al-Dhahabi, Siyar, 11:329 (biography of Ibn Hanbal). Al-SuyutI (d. 911/1505) 
asserted that everything in Ibn Hanbal's Musnad was "tnaqbul," or strong enough for 
use at the very least in pious preaching. The collection's weak hadiths, he argues, are 
close to the acceptable hasan grade; al-Suyuti, Jam' al-jawdmi' at-ma'ruf bi'l-Jdmi' al-kahir, 
29 vols. ([Cairo]: Majma' al-Buhuth al-Lslamiyya, 1390/1970), 1:3. 



THE CANON AND THE NEEDS OF THE COMMUNITY 231 

It was not until the career of al-'Iraqi's student Ibn Hajar (a Shaii'l) 
that a scholar succeeded in performing at least a preliminary takhrij 
of the contents of Ibn Hanbal's Musnad. This feat, however, was only 
subsidiary to Ibn Hajar's primary purpose in the work: rendering the 
Musnad more accessible to scholars by compiling a huge index [atrdf) of 
its contents. He did note, however, in which other main hadlth collec- 
tions Ibn Hanbal's material appears, identifying al-Bukharl and Muslim, 
among others, to bolster the authenticity of the Musnadh hadiths/''' Ibn 
Hajar tackled the issue of authenticity in the Musnad more directly by 
writing a rebuttal of al-'Iraqi's list of nine forged hadiths found in the 
work, often referring to al-Bukhari and Muslim to back them up.'" 

In theory, the Sahihayn canon would have proven extremely useful to 
Maliki efforts to bolster their school's formative text, Malik's Muwatta'. 
The feat that al-Bayhaqi performed for hadiths supporting the Shafi'l/ 
Ash'ari school, al-Baji's student Abu 'Umar Yusuf b. 'Abdallah Ibn 
'Abd al-Barr (d. 463/1070) accomplished for the Muwatta'S''' The 
Cordovan scholar's gargantuan Kitdb al-Tamhtd li-md Jt al-Muwatta' 
min al-ma'dnt wa al-masdnid, twenty-four printed volumes, constitutes 
a comprehensive commentary on Malik's magnum opus. In addition to 
discussing the legal, doctrinal and ritual implications of the material 
contained in the Muwatta', Ibn 'Abd al-Barr attempts to establish the text 
in the language of the sahih movement. Because the Muwatta ' predated 
the exclusive focus on Prophetic hadiths and uninterrupted chains of 
transmission emphasized by the sahih?, and sunan books, the work's large 
number of Successor opinions and incomplete isndd'i compromised its 
strength as a hadith reference. Ever a fly in the ointment, the Zahiri 
maverick Ibn Hazm (d. 456/1064) thus attacked the Maliki opinion 
that the Muwatta ' was the best hadith book by listing it as thirty-first in 
his own ranking of thirty-six books. He placed it well below collections 



'''' The wide net Ibn Hajar cast in his attempt at the takhnj of the Musnad?, contents 
includes: the Sahihayn, the Sunam of Abtl Dawtld, al-NasaX al-Tirmidhi, Ibn Majah, 
al-Dariml and al-Daraqutnl, the SaMhi of Ibn Khuzayma, Ibn Hibban and Abu 'Awana, 
and al-Hakim's Mustadrak; Ibn Hajar, Atraf Musnad Ahmad ibn Hanbal, al-musammd Itrdf 
al-musnid al-mu'tali bi-Atraf al-Musnad al-hanbali, ed. Zuhayr b. Nasir al-Nasir, 10 vols. 
(Damascus: Dar Ibn Kathir and Dar al-Kalim al-Tayyib, 1414/1993). 

''' See, for example, Ibn Hajar, al-Qawl al-musaddadji al-dhabb 'an al-Musnad li'l-imdm 
Ahmad, 39. 

''" Al-Bajl himself produced a larger commentary on the Muwatta' from which he 
drew his Muntaqd. This larger text dealt with Malik's isnddi more than the abridgement; 
Abd al-Rauf, '"Hadith Literature," 280. 



232 CHAPTER SIX 

containing only Prophetic reports, amid books that mix "the words of 
the Prophet with those of others."'''' 

Oddly, although Ibn 'Abd al-Barr had the Sahihayn, the Sunans of 
Abu Dawud, al-Nasa'i and other hadith collections at his disposal, he 
made little use of them in bolstering Malik's reports.^" In fact, Ibn 'Abd 
al-Barr rarely resorts to takhrij at all. On only a handful of occasions 
does he refer to major hadith collections." Instead, Ibn 'Abd al-Barr 
relies on his own mastery of the criteria established by "those requir- 
ing authentic [hadlths] in their compilations" to rate and reinforce 
material in the Muwatta'J'^ Each narration discussed in the Tamhid 
begins with a rating such as muttasil musnad (extending to the Prophet 
with an uninterrupted isndd) or musnad sahih (extending to the Prophet, 
authentic). Occasionally Ibn 'Abd al-Barr reiterates the strength of 
Malik's hadlths with statements such as "this hadith is authentic, its 
authenticity agreed upon by all" or ''musnad muttasil according to the 
people of knowledge.'"^ In the case of mursal reports (those in which a 
Successor quotes the Prophet without citing a Companion) and other 
defective chains of transmission, the author musters sound hadith nar- 
rations to support them. 

Ibn 'Abd al-Barr's contribution proved formidable. He found com- 
plete isndd% for all except four of the hadlths in the Muwatta ' that had 
lacked them. It was not until two centuries later that Ibn al-Salah, 
a Shafi'l by allegiance, succeeded in reinforcing the remaining four 
hadlths. In his Risdlajt wasl al-baldghdt al-arba', Ibn al-Salah argues that 
al-Bukhari and Muslim included a hadith conveying the same meaning 
as Malik's report, "Inni la-ansd aw unassd li-asunn (indeed I forget or am 



^'^ Al-Dhahabi, Tadhkimt al-hujfai, 3:231. It is interesting that Abu al-Hasan 'All b. 
Muhammad al-QabisI (d. 403/1012), one of the first scholars to take Sahih al-Bukhan 
to the Maghrib, compiled a collection of the material in the Muwatta' with complete 
isnadi in his Kitah al-mulakhkhas\ it amounted to only 527 hadlths. This work has been 
published as: Abu al-Hasan 'All b. Muhammad al-QabisI, Muwatta' al-imdm Malik, 
ed. Muhammad b. 'Alawl b. 'Abbas al-Maliki (Abu Dhabi: al-Majma' al-Thaqafi, 
1425/2004); cf. al-KattanI, al-Risdla al-mustatrafa, 12. 

'" Ibn 'Abd al-Barr even had a book entitled al-Ajwiba did al-masd'U al-mustaghraba 
min al-Bukhdn (Answers to Peculiar Questions in al-Bukhan); Ahmad b. Muhammad 
b. Abl Bakr al-Qastallani (d. 923/1517), Irshdd al-sdn li-sharh Sahih al-Bukhan, 10 vols. 
(Beirut: Dar Sadir, [1971], reprint of an 1886-8 edition), 1:43. 

" Ibn 'Abd al-Barr occassionaly notes that a hadith was included by al-Nasa'i, Abu 
Dawud, or al-Bukharl. For examples, see Ibn 'Abd al-Barr, al-Tamhid, 3:265; 4:194—5, 
313; 5:227, 253. 

'^ Ibn 'Abd al-Barr, al-Tamhid, 1:12. 

'3 Ibn 'Abd al-Barr, al-Tamhid, 6:17; 8:11. 



THE CANON AND THE NEEDS OF THE COMMUNITY 233 

caused to forget, so that I create sunna)," and finds narrations from the 
SiK Books for the three other hadiths.'* Ibn 'Abd al-Barr's work and 
the final addition of Ibn al-Salah elicited so much confidence among 
Malikis that the famous Egyptian commentator on the Muwatta', Abu 
'Abdallah Muhammad b. 'Abd al-Baqi al-Zurqani (d. 1710) stated 
unequivocally, "The truth is that the Muwatta' is sahih with no excep- 
tions.'"'^ The twentieth-century Mauritanian scholar of the Sahihayn, 
Muhammad Habib Allah al-Shinqiti (d. 1944) exclaimed that there was 
now "no difference between al-Bukhari and the Muwatta'.'"'^' 

Yet why did Ibn 'Abd al-Barr, al-Baji, and other early commenta- 
tors on the Muwatta' such as Abu Bakr b. al-'Arabi (d. 543/1145) not 
employ the Sahihayn canon to systematically validate Malik's reports?" 
Al-Bukharl's Sahih could certainly have proven invaluable for this task, 
for Malik's transmissions in the Muwatta ' furnished perhaps the largest 
single source for al-Bukhari's work. No fewer than six hundred (35.3%) 
of the Muwatta'''?, narrations appear in the SahihJ^ The answer to this 
conundrum may lie in that very fact: Malikis realized that the Sahihayn 
were effectively built upon the Muwatta '. To use the Sahihayn to shore up 
Malik's work would thus be circular, tantamount to referring to a repro- 
duction to prove the worth of an original. Indeed, Malikis frequently 
cited early reports of al-Shafi'i saying, "There is no book after the 
book of God most high that is more useful (anfa ') than the Muwatta ' of 
Malik," or of the great Basran hadith critic 'Abd al-Rahman b. Mahdl 
(d. 198/814) saying, "We know of no book in Islam after the book 
of God most high that is more authentic (asahh) than the Muwatta' of 



'■' Ibn al-Salah, Risdlaji wad al-baldghdt al-arha', ed. 'Abdallah b. al-Siddlq al-Ghumarl 
(Casablanca: Dar al-Tiba'a al-Hadlthiyya, 1400/1979), 15; Sahih al-Bukhdn: kitdb al-saldt, 
bdb 31; Muwatta': kitdb al-sahw. 

'^ Muhammad b. 'Abd al-BaqI al-ZurqanI, Shark Muwatta' al-imdm Mdlik, 5 vols. 
([Cairo]: Matba'at Mustafa al-Babl al-Halabi, 1381/1961), 1:13. We will see below 
that this claim exceeded even those made about the Scihlhayn, where some exceptions 
were made for flawed hadlths. Some earlier figures such as the Hanafl al-Mughultay 
(d. 762/1361) brought the Muwatta' to the same level as al-Bukharl's SaM not by 
praising the former but by denigrating the latter Al-Mughultay states that the ta'liq 
hadlths in al-Bukharl's book are far more compromising than Malik's incomplete 
isndds; ibid., 1:12. 

"" Ibn al-Salah, Risdla, 3-4 (editor's introduction). 

" In his commentary on the Muwatta', Abu Bakr b. al-'Arabl frequentiy uses the 
Sahihayn as well as other famous sunans such that of al-Nasa"! for takhnj of hadlths he 
mentions in his comments, but not to back up the hadlths of Malik himelf; see Abu 
Bakr b. al-'Arabi, Kitdb al-qabas Ji sharh Muwatta' Mdlik b. Anas, ed. Muhammad 'Abdallah 
Walad-Karlm (Beirut: Dar al-Gharb al-lslami, 1992). 

'" Fuad Sezgin, Buhdn'nin Kaynaklan, 305. 



234 CHAPTER SIX 

Malik."" Ibn 'Abd al-Barr sets forth this myriad praise of the Muwatta' 
in the introduction to his Tamhid, adding other reports such as 'Abdallah 
b. Wahb's (d. 197/813) statement that "whoever has copied [katabd) the 
Muwatta ' of Malik need write nothing more on what is permissible and 
forbidden [al-halal wa al-haram)."^'° 

Among Malikis, the Muwatta' was thus the true foundation of the 
sahth movement on which later masterpieces like the Sahihayn were built. 
Abu Bakr b. al-'Arabi states in the introduction of his commentary on 
al-Tirmidhi's Jamr that al-Bukhari's Sahih "is the second basis [ad) in 
the realm [of hadlth], but the Muwatta' h the first basis [al-asl al-awwal), 
and on them have been built all others" such as the collections of 
Muslim and al-Tirmidhi."' Al-Qadi 'lyad thus speaks of the Muwatta' 
and the Sahths of al-Bukhari and Muslim as "the three mother-books 
{al-ummahat al-thalatli)'' "the authentic collections of reports (athar) that 
have been agreed upon as foremost throughout the ages, and that the 
scholars have accepted in all the rest of the regions {sa'ir al-amsar)." 
These works are "the usul of every asL. . and the principles of the sci- 
ences of traditions [mabadi' 'ulum al-dthdr) ""^ 

Like Ibn 'Abd al-Barr, neither al-Baji nor Abu Bakr b. al-'Arabl's 
commentaries on the Muwatta ' make use of the Sahihayn canon to sup- 
port the authenticity of Malik's material. Rather, al-Baji exudes con- 
fidence in the foundational role of the Muwatta ' and the unanimity of 
the community's approval of Malik's hadiths. He admits, for example, 
that Malik's report about 'Abdallah b. 'Umar's never attending Friday 
prayer without perfuming and anointing himself with oils lacks a reli- 
able isndd (i.e., in this case it does not extend back to the Prophet). 
But al-Baji argues that this is unnecessary, since the umma had acted 
on this hadlth and "accepted it with consensus [talaqqathu bi'l-qubul)." 
The report thus enjoyed a guarantee of authenticity far beyond that 
provided by a mere sahth isndd.^^ 



'^ Ibn Hibban, Kitdh al-majruhin, 1:41—2. 

"" Ibn 'Abd al-Barr, Kitdb al-tamhid, 1:78. For the other quotes praising the Muwatta', 
see ibid., 1:76-79; cf. al-Qadi 'lyad, Tartib al-maddrik, 1:191. 

"' Abu Bakr b. al-'Arabi, Sahih al-Tirmidhi bi-sharh al-imdm Ibn al-'Arabi al-Mdliki, 13 
vols, in 5 (Cairo: al-Matba'a al-Misriyya bi'1-Azhar, 1350/1931), 1:5. 

°^ Al-Qadi 'lyad b. Musa, Mashdriq al-anwdr 'aid sihdh al-dthdr, ed. Bafamshl Ahmad 
Yagan, 2 vols. ([Rabat]: Wizarat al-Awqaf wa al-Shu'un al-Islamiyya, 1402/1982), 
1:27. 

»' Al-BajI, al'Muntaqd, 1:203. 



THE CANON AND THE NEEDS OF THE COMMUNITY 235 

As with their late recourse to the Sahihayn canon in debate and exposi- 
tion, it was only in Mamluk Cairo of the eighth/fourteenth century that 
Hanafis turned to al-Bukhari and Muslim to bolster their school's forma- 
tive legal and hadith texts. With the exception of al-Saghani, al-Mawsili 
and al-Khilati in the seventh/thirteenth century, only at this time did 
Hanafi hadith scholars begin systematically studying and employing 
the Sahihayn. 'All b. 'Uthman Ibn al-Turkumani (d. ca. 747/1347), a 
Hanafi judge in Egypt, was a prominent teacher of al-Bukharl's SaMh', 
even Zayn al-Din al-'Iraqi numbered among his students."* Another 
Hanafi teacher of al-'Iraqi's in Cairo, 'Ala' al-Din 'Abdallah b. Qalij 
al-Mughultay (d. 762/1361), wrote a famous commentary on Sahih 
al-Bukhdn.^' It was Ibn al-Turkumanl's students, however, who first 
systematically employed the SaMhoyn canon to legitimize major Hanafi 
hadith collections. 

Muhyl al-Din Abu Muhammad 'Abd al-Qadir Ibn Abi al-Wa£a' 
(d. 775/1374) served as a Hanafi mufti '\n Mamluk Cairo and eventu- 
ally produced the most comprehensive biographical dictionary of the 
Hanafi school."'' In a personal addendum to this dictionary, Ibn Abl 
al-Wafa' explains how he was assigned the task of validating Hanafi 
hadlths using canonical collections. His teacher Ibn al-TurkumanI had 
been approached by a Mamluk amir who, like most of the Turkish mili- 
tary elite, subscribed to the Hanafi madhhabf'^ This amir, who evidently 
enjoyed debating issues of religious law with scholars from an opposing 
school (probably the dominant Shafi'l madhhab), consistently stumbled 
before his adversaries' demands for his hadith sources. The amir would 
reply "We have the book of [AbuJaTar] al-TahawI (d. 321/933)," but 
complained to Ibn al-TurkumanI that "if we mention a hadith from it 
to our opponents they say to us, 'We will not listen to anything except 

what is in al-Bukharl and Muslim ' " Ibn al-TurkumanI replied to the 

amir, "Most of the hadlths in al-TahawI are [also] in al-Bukharl and 
Muslim or the Sunans [of Abu Dawud, al-Tirmidhi, al-Nasa'l and Ibn 

Majah], and other books of the hadith masters [huffaz) " The amir 

thus asked him to find citations for all of al-Tahawl's material based 



"■' Ibn Fahd, Lahi al4ihd^, 91, 93-4. 

"'• Ibn Fahd, Lahi al-lihd^, 87. 

"<'' Ibn Fahd, lafe al-tihdr., 105. 

"' Ulrich Haarmann, "Joseph's law — the careers and activities of Mamluk descen- 
dents before the Ottoman conquest of Egypt," in The Mamluks in Egyptian Politics and 
Society, ed. Thomas Philipp and Ulrich Haarmann (Cambridge: Cambridge University 
Press, 1998), 78. 



236 CHAPTER SIX 

on those books. In typical scholarly manner, the judge replied, "I do 
not have the time for that, but one of my students [ashabi) can do it." 
Ibn al-Turkumani handed the task to his son, Jamal al-Din al-Maridlni, 
who then assigned it to a younger student: Ibn Abi al-Wafa'.^" Provided 
with reference books from the (Zotzt's own library, Ibn Abl al-Wafa' pro- 
ceeded to supplement the contents of al-Tahawl's Shark ma 'dm al-dthdr 
with narrations from "well-known hadlth books [al-kutub al-mashhurd), 
namely the Sahihayn, the Four Sunam as well as other musnads, detailing 
what is authentic, acceptable or weak."**^ 

Although Ibn Abi al-Wafa"s finished work, al-Hdwi fi baydn dthdr 
al-Tahdwi, occasionally refers to other works, such as Ibn Khuzayma's 
Sahih, it is inclusion in the Sahihayn in particular, or meeting al-Bukhari's 
and Muslim's standards, that furnishes the author's principal means 
for validating al-Tahawl's hadiths. Indeed, Ibn Abi al-Wafa' bends the 
Sahihayn canon to maximum use. Even when a hadith appears with 
a chain of transmission not approved by al-Bukhari or Muslim, Ibn 
Abi al-Wafa' asserts that "the basic text {asl) of the hadith is in the 
Sahihayn."^^ Conversely, if the text of one of al-Tahawl's hadiths does 
not appear in the Sahihayn but its isndd does, he states that "its isndd is an 
isndd from the Sahihayn."'^^ Ibn Abi al-Wafa' proves even more flexible in 
employing the legitimizing power of the canon: if one narrator in the 
isndd did not earn a place in al-Bukhari's or Muslim's works, Ibn Abi 
al-Wafa' still insists that "the rest of the isndd is men of the Sahihayn. "^'^ 
He also makes use of al-Hakim's application of "the requirements of 
al-Bukharl and Muslim" in the Mustadrak to authorize reports, sometimes 
declaring in his own opinion that certain hadiths meet the conditions 
of the ShaykhaynP 

The task of reinforcing the hadiths cited in one of the Hanafi school's 
leading legal references, the Hiddya of Abu al-Hasan 'All b. Abu Bakr 
al-MarghlnanI (d. 593/1196-7), fell to another of Ibn al-Turkumanl's 



"" Ibn Abl al-Wafa', al-Jawahir al-mudiyya, 4:571. 

"' Ibn Abl al-Wafa', al-Hdwifi baydn dthdr al-Tahdwi, ed. Ynsuf Ahmad, 3 vols. (Beirut: 
Dar al-Kutub al-'Imiyya, 1419/1999), 1:24. 

* Ibn Abl al-Wafa', al-Hdwi, 1:94. 

" Ibn Abl al-Wafa', al-Hdwi, 1:50, where it occurs twice. 

'^ Ibn Abl al-Wafa', al-Hdwi, 1:61, 142. 

'' Ibn Abl al-Wafa', al-Hdwi, 1:49, 64, 75, 85, 120. He notes, for example, that "al- 
Hakim narrated through him [Fahd b. Sulayman] in his Mustadrak, so he meets the 
requirements of the Shaykhayn.'" 



THE CANON AND THE NEEDS OF THE COMMUNITY 237 

Students: 'Abdallah b. Yusuf al-Zayla'i of Cairo (d. 762/1361)."'^ A 
friend and colleague of the Shafi'i Zayn Dm al-'Iraqi, al-Zayla'i's Nasb 
al-rdya fi takhnj ahddith al-Hiddya stands out as one of the clearest and 
most accessible works of hadlth literature. The great Indian Hanafi 
hadlth scholar of Cairo, Muhammad Murtada al-Zabidi (d. 1 205/ 1 79 1), 
later performed the same service for a selection of hadlths on which 
Hanafls had historically relied for deriving law [ahkdm). In his Kitdb 
'uqud al-jawdhir al-munifa, he states that he will validate these hadiths 
by showing their narrations in the Six Books. ^'^ 

Why did the Hanafls begin employing the canon almost three 
centuries after their Shafi'i counterparts? With al-Hakim's Mustadrak 
and the declarations of his associates from the Shafi'i/Ash'arl and 
Hanball/iiber-Sunni camps, the Sahihayn emerged as authoritative texts 
within the transmission-based community. The Hanafi school, however, 
constituted the bulk of the reason-based school to which the transmis- 
sion-based scholars remained in steadfast opposition. Just as hadlth 
scholars like al-Bukharl and al-Hakim had condemned Hanafls for 
departing from the Prophet's true sunna, so did the Hanafls like Abu 
Mutf Makhul al-Nasafi (d. 318/930) consider the ahl al-hadith brainless 
literalists, capable of merely parroting the Prophet's words but not of 
understanding his message.*' 

This Hanafi contempt for transmission-based scholars tainted the 
school's view of al-Bukhari. This comes as no surprise in light of the 
muhaddith''s, virulent criticism of Abu Hanlfa in his Kitdb raf al-yadayn 
and his general criticism of the reason-based school in his Sahih. In the 
chapter on the issue of milk-relationships {ridd') in his mammoth work 
of Hanafi substantive law, the famous Hanafi jurist and legal theorist 
al-Sarakhsi (d. ca. 490/1096) produces an amazingly insulting story 
about al-Bukharl. He tells how al-Bukhari upheld the opinion that if 
two children drink milk from the same ewe they would become milk- 
siblings, prohibited from ever marrying one another [hurmat al-ridd'). 
When the great muhaddith supposedly visited his native Bukhara and 
began answering the legal questions of its citizens, the leading Hanafi 



"* Ibn Hajar, al-Durar al-kamina fi a'yan al-mi'a al-thamina, ed. 'Abd al-Warith 
Muhammad 'All (Beirut: Dar al-Kutub al-'Ilmiyya, 1418/1997), 2:188-9. 

'^ Muhammad Murtada al-Zabldl, Kitdb 'uqud al-jawdhir al-munija, ed. Wahbi 
Sulayman Ghawjl al-Albani (Beirut: Mu'assasat al-Risala, 1406/1985), 17. 

"* Marie Bernand, "Le Kitab al-radd ala 1-bida' d'Abu Mutf Makhul al-Nasafi," 
Annates Islamologiques 16(1980): 121-2. 



238 CHAPTER SIX 

of the city, Abu Hafs Ahmad b. Hafs (d. 217/832), told him that he 
was unqualified to give expert legal opinions. Al-Bukhari ignored him 
and continued to answer questions. When someone asked about the 
issue of drinking milk from the same ewe, the people found al-Bukharl's 
response so preposterous that they expelled him from the city. 

It goes without saying that al-Bukhari probably did not espouse this 
opinion and that the story is apocryphal; earlier sources make clear 
that al-Bukhari's expulsion from Bukhara came at the amir'i orders at 
the end of his life, and at any rate, Abu Hafs died before al-Bukhari 
reached full maturity'" The story, however, provides a comic foil for 
al-Sarakhsi, who proceeds to explain that if two youths drink the milk 
of the same animal, in no way do they become milk-siblings. The 
milk-sibling relationship is analogous to kinship, and just as humans 
cannot be related to animals, so that relationship cannot be established 
by an animal's milk.^" Over two hundred years later, the Hanafi legal 
theorist Abu Barakat 'Abdallah b. Ahmad al-Nasafi (d. 710/1310) 
reproduced the same insulting story to prove a fundamental principle 
in the Hanafi school: "a hadlth scholar who is not a jurist [al-muhaddith 
ghayr al-faqih) errs often." In other words, only specialized jurists are 
qualified to derive laws from Prophetic traditions.''^ Ibn Abl al-Wafa' 
includes the same story about al-Bukhari in his Hanafi biographical 
dictionary, al-Jawdhir al-mudiyya}^^' 

Hanafis seem to have maintained a skeptical distance from the 
Sahihayn canon into the eighth/fourteenth century. Yet it was an ines- 
capable feature of the scholarly environment with which they had to 
come to terms. As his account of how he came to apply the Sahihayn 
canon to a Hanafi hadith collection suggests, Ibn Abi al-Wafa' was 
responding to outside polemical pressures rather than acting on any 
reverence for al-Bukharl's or Muslim's work. In fact, Ibn Abi al-Wafa' 
reveals a deep cynicism towards the canonical culture surrounding the 



'' Also, al-Bukharl's Sahih lacks a chapter on milk- relationships (al-rida). He cov- 
ers the topic in four subchapters in the book on marriage, but makes no claim about 
animal's milk; Ibn Hajar, Fath al-bdri, 9:174. On al-Bukharl's expulsion from Bukhara, 
see above, Chapter 3, n. 63. 

"" Al-Sarakhsl, Kitdb al-mabsut, 2nd ed., 30 vols. (Beirut: Dar al-Ma'rifa, 197-), 
30:297; Ibn Abi al-Wafa', al-Jawdhir al-mudiyya, 1:166 (biography of Ahmad b. Hafs 
Abu Hafs al-Kablr). 

'^ Jamal al-Din Muhammad al-Qasimi al-Dimashqi, Haydt al-Bukhdri, ed. Mahmtid 
al-Arna'ut (Beirut: Dar' al-Nafa'is, 1412/1992), 48. 

'™ See n. 98 above. 



THE CANON AND THE NEEDS OF THE COMMUNITY 239 

two collections. Discussing how Shafi'ls assert the authenticity of a 
hadith that al-TahawT had declared weak by arguing that it is included 
in Muslim's Sahih, Ibn Abl al-Wafa' states that Shafi'ls "cannot show 
off [the hadith] [yatajawwahund) to us because it comes from Muslim, 
for [many] things appear in Muslim, and showing it off does not bolster 
[their position] in situations of conflicting [narrations] {istiddm)." Ibn 
Abl al-Wafa' then embarks on what may be the lengthiest and most 
comprehensive existing enumeration of the types of flaws appearing 
in the Sahihayn, detailing consistently weak chains of transmission as 
well as the problematic texts of certain hadiths. Referring to Abu Zur'a 
al-Razl's warning to Muslim upon reading his Sahih, Ibn Abi al-Wafa' 
concludes, "God bless Abu Zur'a, for he spoke the truth." In Ibn 
Abl al-Wafa"s opinion, the Sahihayn had indeed "made a path for the 
people of bid'd' and been bent to polemical and partisan purposes."" 
A more playful contempt for the canon appeared in the career of a 
slightly earlier Hanafi hadith scholar who visited Cairo, Shams al-Din 
Mahmud b. Abl Bakr al-Kalabadhi al-Bukhari (d. 700/1300). When 
this scholar would see a handsome youth, he would say, playing on his 
own name (al-Bukhari), "that is sahih according to the requirements 
of al-Bukhari."'"2 



Misuse of the Sahihayn Canon 

The authority that the Sahihayn or the "requirements of al-Bukhari and 
Muslim" carried in debates was very alluring. In the time before stan- 
dardized texts and easily accessible indices, and long before searchable 
databases, knowing the exact contents of capacious hadith collections 
like the Sahihayn proved impossible to all but the most accomplished 
scholars. Both among the less masterful of the scholarly class and less 
literate segments of society, it was difficult to limit the legitimizing 
authority of the Sahihayn to the actual contents of the books. It was 
tempting to claim that a hadith supporting one's position had met al- 
Bukhari's and Muslim's standards. 



"" Ibn Abl al-Wafa', al-Jawdhir al-mudiyya, 4:565-69. 

'"^ Ibn Abl al-Wafa', al-Jawdhir al-mudiyya, 3:455. Invoking religious idiom in homo- 
erotic literature was common; see J.W. Wright Jr, "Masculine AUusion and the Structure 
of Satire in Early Abbasid Poetry," in Homoeroticism in Classical Arabic Literature, ed. J.W. 
Wright Jr. and Everett K. Rowson (New York: Columbia University Press, 1997), 10. 



240 CHAPTER SIX 

Qazvini had made a valiant attempt to defend Imami beliefs by 
claiming that certain pro-'Alid reports were included in the Sahihayn. 
Unfortunately, the hadiths he cites stating that 'All's name is written on 
the leg of God's throne or above the doorway to Paradise are nowhere 
to be found in the two collections, nor do they appear in any of the Six 
Books, as was mentioned above. ""^ This overstepping of the boundaries 
of the canon was not limited to non-Sunnis who may not have been 
well-acquainted with Sunni hadlth collections. The prominent Cairene 
Hanafi Badr al-Din Mahmud b. 'Ubaydallah al-Ardabill (d. 875/1471) 
approached the Shafi'i hadith scholar Abu 'Abdallah Muhammad b. 
'Abd al-Rahman al-Sakhawi (d. 902/1497) with a list of hadiths the 
status and citations of which he was unsure. In the majority of al- 
Sakhawi's responses in his al-Ajwiba al-'aliyya 'an al-as'ila al-Dimydtiyya, 
the scholar replies that the hadiths have been falsely ascribed to some 
hadlth collection or critic. Seven hadiths had been falsely cited from 
Sahth al-Bukhdn, eight from Sahih Muslim and three from al-Tirmidhl's 



2. The Need for an Authoritative Reference: 
The Sahihayn and Non-Hadith Specialists 

The Sahihayn met a second important need exhibited by the Sunni 
community in the mid-fifth/eleventh century: that of a common 
authoritative hadlth reference for non-specialists. This need stemmed 
from an increasing division of labor between jurists and hadlth scholars 
in the mid-fifth/ eleventh century. With the establishment of madrasah 



'»' See n. 62. 

'"* Shams al-Dlii Muhammad b. 'Abd al-Rahman al-SakhawI, al-Ajwiba al-'aliyya 'an 
al-as'ila al-Dimydtiyya, ed. Mish'al b. Barn al-MutayrI (Beirut: Dar Ibii Hazm, 1420/1999), 
al-Bukharl: 81, 87, 101, 149, 112, 131, 145; Muslim: 99, 110, 139, 134, 143, 145, 
151; al-Tirmidhi: 76, 108, 131. The authority of the Sahihayn canon continues to be 
misapplied in the modern period as well. The Moroccan hadlth scholar 'Abdallah b. 
al-Siddlq al-Ghumarl (d. 1993) criticized Ahmad al-Baqurl, who had previously been 
a high-ranking member of the Muslim Brotherhood before being co-opted by the 
Egyptian government, for incorrectly attributing a hadlth to al-Bukharl. Al-Baqurl 
had tried to defend the practice of mixed-gender dinner gatherings by claiming that 
al-Bukharl's SaMh included a hadlth in which the Prophet refused a dinner invitation 
because his wife 'A'isha was not invited. Al-Ghumarl objects that no such hadlth exists 
in the SaMh or any other collection; 'Abdallah b. al-Siddlq al-Ghumarl, al-Khawdtir al- 
diniyya, 2 vols, in 1 (Cairo: Maktabat al-Qahira, 1425/2004), 1:33. 



THE CANON AND THE NEEDS OF THE COMMUNITY 241 

in cities like Baghdad, Naysabur and Merv in this period, a space 
had been created that primarily emphasized the study of law [fiqli) as 
opposed to the pietistic or scholarly transmission of hadiths.'"'^ Unlike 
the transmission-based scholars of al-Bukharl's time, who had compiled 
their musannafi as expressions of their own legal thought, many of the 
mid-fifth/ eleventh century denizens of the madrasas lacked expertise in 
hadlth criticism. Although Abu Ishaq al-Isfaraylnl had been sought out 
as a hadlth scholar, legal theorist and theologian alike, two generations 
later Shafi'i scholars like al-Shirazi and al-Juwayni were focusing more 
narrowly on elaborating substantive law, theology and legal theory. As 
al-BayhaqI (d. 458/1066) noted in his letters, the breed of jurists who 
were also masters of hadlth criticism had all but died out.""" Legal 
scholars needed to turn to established hadlth collections with widely 
respected standards in order to validate their legal stances or hadlths. 

The role of the Sahihayn as an authoritative reference was embryonic 
in al-Hakim al-Naysaburl's work, where he proffered the Sahihayn as 
a protective canopy for authentic Prophetic reports."" In his lengthy 
treatise on usul, the Sharh al-luma', al-ShirazI builds on this theme in 
an attempt to meet the jurists' needs. He explains that Shafi'l jurists 
accept hadlths from "senior hadlth scholars (kibdr ashdb al-haditK)" with- 
out research or question. Just as a judge trusts a witness once he has 
proven his reliability, so too jurists can trust the authenticity of these 
critics' material. Al-ShlrazI mentions al-Bukharl, Muslim, Abu Dawud 
and Yahya b. Ma'ln as examples, as well as major jurists who had also 
mastered hadlth, such as Malik and Ibn Hanbal.'"" 

The articulation of this need for authoritative references and the suit- 
ability of the Sahihayn to meet it appear most clearly in discussions on 
the office of wz^f (jurisconsult, a term often conflated with mujtahid), the 
legal expert from whom the population sought rulings. In his descrip- 
tion of the necessary qualifications for a mufti, al-ShirazI states that he 
must possess a command of the four sources of Islamic jurisprudence: 
the Qjur'an, the Prophet's sunna, consensus and analogical reasoning 
{qiyds). In terms of the sunna, the mujii must know which hadlths to 



'"' George Makdisi, "Muslim Institutions of Learning in Eleventh-Century Baghdad," 
BulkUn of the School of Oriental and African Studies 24:, no. 1 (1961): 10-1 1; idem, "Hanbalite 
Islam," in Studies on Islam, ed. and trans. Merlin L. Swartz (Oxford: Oxford University 
Press, 1981), 230. 

'"" See al-Subkl, Tabaqdt al-shafi'ijya, 5:82. 

'"' Al-Hakim al-Naysaburl, Ma'nfat 'ulUm al-hadith, 75. 

'"" Abfl Ishaq al-ShlrazI, Sharh al-luma', 2:634. 



242 CHAPTER SIX 

accept and which to reject. But al-Shirazi exempts the mufti from the 
requirement of mastering the intricacies of isnad or hadith criticism, 
for "if we made knowing that [hadith] by its isndd obligatory for each 
mujtahid, this would lead to great difficulty, for that requires a lifetime." 
Instead, a muffi should rely on "the imams of the ashdb al-hadith" like 
al-Bukhari, Muslim, Ibn Hanbal, al-Daraqutni and Abu Dawud.'"" A 
contemporary Shafi'i in Naysabur, Abu al-Muzaffar al-Sam'anI, (d. 489/ 
1096), lists "the relied- upon books" for such purposes as the Sahih of 
al-Bukhari first and foremost, then that of Muslim, Abu Dawud, al- 
Tirmidhi, al-Nasa'l, the Mustakhraj of Abu 'Awana and finally the Sahihs 
of Abu 'Abbas al-Daghuli and Ibn Hibban."" 

Al-Ghazall concurs, stating that a mufit or mujtahid must rely on critical 
collections of hadlths that distinguish between authentic and unreli- 
able material.'" When working with hadlths that have been accepted 
as authentic by the umma, one need not scrutinize their chains of 
transmission [Id hdja bihi ild al-nazar fi isnddihi). The mujii should thus 
follow al-Bukharl and Muslim in the evaluation of narrators, since 
these two critics only narrated from those whose uprightness ('addld) 
they had established. Al-Ghazall cautions that if one does not defer 
to these two experts on issues of isndd evaluation, one would have to 
master that science oneself He adds, "This is a tall order (tawtl), and 
is, in our time, with the massive number of intermediaries {wasd'it) [in 
the chains of transmission], very difficult (Wr).""^ 

In his discussion of the requirements for a muffi m the Hanafi school, 
'Abd al-'AzIz al-Bukhari echoes this division of labor and reliance on 
canonical hadith collections. Like al-Shirazi, he requires the mujtahid or 
mufti to have command of the sunna and know the hadlths dealing with 
legal rulings [hadith al-ahkdm). The jurist, however, need not memorize 
this material. Rather, he must have at his disposal a vetted copy [asl 
musahhiK) of one of the ahkdm hadith collections such as al-Bukharl, 
Muslim or Abu Dawud as a reference."' 



Abu Ishaq al-Shlrazl, Shark al-luma', 2:1033-4. 

Al-Sam'anl, Qawdti,' al-adilla, 2:499-500; cf. al-Juwaynl, al-Burhdn, 2:1333. 

Al-Ghazall, al-Mankhul, 459. 

Al-Ghazall, al-MustasJa, ed. Muhammad Yusuf Najm, 2 vols. (Beirut: Dar al- 
Sadir, 1995), 2:200-2. 

' '^ 'Ala' al-Din Abd al-'Aziz b. Ahmad al-Bukharl, Kashf al-asrdr 'an usul Fakhr al-hldm 
al'Bazdawi, 4 vols, in 2 (Beirut: Dar al-Kitab al-Arabi, 1394/1974), 4:15. 



112 



THE CANON AND THE NEEDS OF THE COMMUNITY 243 

Abu al-Walid al-Baji expresses the same opinion for the Maliki 
school. He states that those who have achieved the expertise necessary 
to critically examine hadiths can evaluate reports on their own, just as 
al-Bukharl and Muslim did. "But he who has not achieved that condi- 
tion," he adds, "must follow those two [al-Bukharl and Muslim] for 
hadiths he claims to be authentic, pausing [tawaqquf) at what they did 
not include in their Sahths."^^* 

It is at this point that the split in the hadlth tradition initiated by 
the sahih movement again comes into focus. The canonization of the 
Sahihayn and their use as measures of authenticity transformed them 
into institutions of authority in the Muslim community. This institutional 
role emerged as a counterweight to the focus on the chain of transmis- 
sion as the sole vehicle for tying Muslim scholars to the hermeneutic 
authority of the Prophet's words. The consensus of the umma on the 
Sahihayn and their subsequent use as a reference in implementing the 
Prophet's authority meant that books could replace the authoritative 
source provided by the living isndd. When al-Shirazi explains that jurists 
can replace a direct link to the Prophet and a mastery of evaluating 
its authenticity with reference books vetted and authorized for that 
purpose, he obviates the need for an intensive study of isnads. 

The diverging paths of the jurists and hadith scholars becomes 
evident when we juxtapose al-Shlrazl's discussion of muftis with that 
of two of his Shafi'l contemporaries more rooted in hadlth study than 
legal theory or substantive law. In Abu Bakr al-Bayhaqi's discussion of 
the mufti's, requirements we find no mention of resorting to reference 
works. He merely repeats al-Shafi'l's original requirement that a mufti 
himself master the sources of legislation and know which hadiths to 
accept or reject.""' Al-Khatib al-Baghdadi also repeats these fundamen- 
tal requirements, stating that "a mujii will not be able to [meet these 
requirements] unless he has been excessive {akthara) in writing the reports 
of the early generations and hearing hadiths." The chasm separating 
him from al-ShlrazI widens further when al-Khatib recounts, rhetori- 
cally no doubt, how Ibn Hanbal required someone to know at least five 
hundred thousand hadiths before he could act as a mufti J^''' 



"■' Al-Baji, Abu al-Wabd Sulayman b. Khalaf al-Baji iva kitabuhu al-Ta'dil wa al~tajnh, 
1:310. 

"' Abu Bakr al-Bayhaqi, al-Madkhal ild al-Sunan al-kubrd, ed. Muhammad Diva' 
al-Rahman al-A'zamI, 2nd ed., 2 vols. (Rivadh: Adwa' al-Salaf, 1420/[1999-2o'o6]), 
1:169. 

'"" Al-Khatib, Kitdb al-faqih wa al-mutafaqqih, 2:330, 344-5. 



244 CHAPTER SIX 

The most dramatic step in proposing the Sahihayn as institutions of 
authority to which scholars seeking to evaluate hadiths could turn came 
almost two centuries later, with the work of Ibn al-Salah (d. 643/1245). 
By Ibn al-Salah's time, Muslims no longer compiled massive collections 
of hadith with living isnad'i back to the Prophet, like al-Bayhaqi's Sunan. 
In a time when the critical rigor of giants like al-Bukharl seemed to be 
fading into history, Ibn al-Salah pondered how jurists or even hadith 
scholars should evaluate previously unrated hadiths they came across in 
the course of study or debate. He argued that, "If we find some report 
in a hadith notebook that seems to have a sahih isnad but is neither in the 
Sahihayn nor indicated as sahth in a book of the relied-upon, well-known 
imams, we do not dare insist that it is authentic (la natajdsaru 'aldjazm 
al-hukm bi-sihhatihi)." Ibn al-Salah's call rested on his belief that hadith 
transmission in his time had deteriorated so much from the rigorous 
standards of yesteryear that hadith scholars were no longer able to 
trust their transmissions from earlier sources. Consequently, "knowing 
the hasan and sahth depended on the imams of hadith having specified 
this in their well-known, relied-upon works that. . .have been preserved 
against alteration and scribal error (tahnf)" "Most of what is sought 
out from the isndds circulating [today]," he concludes, "falls outside this 
pale.""' Beginning with his follower al-NawawI, scholars understood 
Ibn al-Salah's position as calling for an end to the evaluation of hadiths 
in favor of a total reliance on sahth collections."" 

This dramatic call to equate all sahth hadiths with the contents of the 
Sahihayn and other sahth books embraced the jurists' need for authorita- 
tive references at the expense of the hadith scholars' methodology."^ 



'" Ibn al-Salah, Muqaddima, 159-60. 

"" Ibn al-Salah probably meant that one could no longer declare hadiths transmit- 
ted by living isndds, and not found in major collections authentic. As for hadiths found 
in earher compilations that included reports of various levels of reliability, such as 
al-Tabaranl's Mu'jam, Ibn al-Salah was probably not arguing against ruling on the 
authenticity of this material. It was in this sense, however, that Ibn al-Salah's comments 
were understood from the time of his follower al-NawawI on. Al-SuyutI (d. 91 1/ 1505) 
devoted a small treatise to this subject, entitled al-Tanqih fi mas'alat al-tashih, in which 
he clarified Ibn al-Salah's statement but then proceeded to himself declare an end to 
the authentication of hadiths due to the inability of later scholars to conduct proper 
'ilal criticism. See the edited text of this treatise in Badi' al-Sayyid al-Lahham, al-Imdm 
al-hdfiz Jaldl al-Din al-Suyuti wajuhuduhufi al-hadith wa 'ulumihi (Damascus: Dar Qutayba, 
1415/1994), 460-3. 

"' Ibn al-Salah states that one could also find sahih hadiths in the books of Abu 
Dawud, al-Tirmidhi, al-Nasa'l and al-Daraqutnl, but that one could not assume that 
all their contents were authentic, since this was not the criterion of their compilers. 



THE CANON AND THE NEEDS OF THE COMMUNITY 245 

The function of the two books as authoritative institutions therefore 
emerged as a source of tension between scholars whose chief affilia- 
tion was to the study of law and others who focused more on hadith. 
Although Ibn al-Salah was first and foremost a Shafi'i hadith scholar, 
as his efforts to eliminate the last vestiges of doubt from the Muwatta' 
suggest, his interests lay in strengthening scholarly institutions. His call 
indeed amounted to declaring the victory of the authoritative institu- 
tion of the sahih book over the living isnad. Reacting with predictable 
tension to Ibn al-Salah's argument, almost all later hadith scholars 
understandably rejected the notion that they were unqualified to inde- 
pendently evaluate hadlths; as Zayn al-Din al-'Iraqi explained, "this 
was the hadith scholars' job." '^^ 

What emerged as a consensus among scholars in the wake of Ibn 
al-Salah's provocative claim was a balance between the jurists' needs 
for authorized institutions housing the Prophet's legacy and the hadith 
scholars' focus on the living isndd as the link to his authority. The Sahihayn 
would serve as the primary reference for non-specialists, while qualified 
hadith scholars could continue evaluating material they came across. 
Ibn Hajar thus instructs jurists who are browsing through a musnad or 
sunan work but are not hadith experts to refer to the Sahihayn to see if 
a report is authentic or not. If al-Bukharl or Muslim did not include 
the report, one should see if some other imam declared it authentic.'^' 
Other hadith scholars, like al-NawawI, al-BulqInl (d. 805/1402-3) and 
Ibn al-WazIr seconded the notion that those who have the expertise 
must independently evaluate isndds, but those who do not must rely on 
the Sahihayn, their mustakhrap and ilzdmdt works. '^- 

The role of the Sahihcyn as a reference for non-specialists evaluat- 
ing the reliability of Prophetic reports had profound implications for 
pietistic literature: if a hadith had earned al-Bukharl's or Muslim's 



Sahih books, however, such as that of Ibn Khuzayma, could provide this security; Ibn 
al-Salali, Muqaddima, 163-4. Al-SuyutI went far beyond Ibn al-Salah, claiming that 
any hadith cited from the Sahihayn, the Sahihs of Ibn Khuzayma, Ibn Hibban, the 
Mustadrak of al-Hakim (with the exception of material criticized by scholars like al- 
Dhahabl), the two Muntaqdi of Ibn al-Sakan and Ibn Jarud, al-Ahddlth al-mukhtdra of 
Diya' al-Din al-MaqdisI, Malik's Muwatta and the mustakhraj works of the So-hihayn is 
authentic; z^-i^uyWJ, Jam' al-jawdmi', 1:2. 

120 Al-'IraqI, al-Taqyid iva al-iddh, 27; idem, al-Tabsira wa al-tadhkira, 1:67; al-NawawI, 
al-Taqrib, 6; Ibnjama'a, 130; al-Bulqlnl, 159; 'A-ii'Arayii, Jawdhir al~usul, 21; al-SakhawI, 
Fath al-mughith, 1:63-4. 

'^' Ibn Hajar, al-Nukat 'aid kitdb Ibn al-Saldh, 149. 

'22 Ibn al-WazIr, Tanqih al-anidr, 40; see n. 120. 



246 CHAPTER SIX 

Stamp of approval, one need not provide an isndd when citing it. The 
Shafi'i hadlth scholar AbQ Muhammad al-Husayn b. Mas'ud al- 
Baghawi (d. 516/1122), dubbed "the Reviver of the Sunna [muhyi al- 
sunna)," demonstrated how the Sahihayn canon could simplify the use 
of hadiths in the religious life of regular Muslims. He explains that 
his most famous work, the pietistic manual Masdbih al-sunna, is culled 
from the books of the great hadlth imams, to help people implement 
the Prophet's sunna in daily life. The work is small and portable, for a 
very simple reason: al-Baghawi omits the contents' isndds. Instead, the 
author divides the hadiths in each chapter into two sections, "authentic 
[sihdlij" and "good (hisdn)." The authentic section consists only of reports 
from al-Bukhari and Muslim, while the less reliable ^^hisdn" hadiths 
come from the collections of al-Tirmidhi, Abu Dawud, al-Nasa'i and 
other respected compilers. The reader thus relies on the source of the 
hadiths to know their reliability. Those coming from the Sahihayn are 
considered automatically authentic, whereas al-Baghawi states that he 
will alert the reader to any weaknesses in the hadiths of the "good" 
section. '^^ 

It is clear that in cities like Damascus in the early seventh/thirteenth 
century, inclusion in the Sahihcyn exercised potent authority among the 
everyday Muslims al-BaghawI was targeting. Even the laity held the 
contents of the two works in unique veneration. A common layman, 
for example, asked Ibn al-Salah for a legal ruling about the hadlth "He 
who repents for a sin is like one without sin [al-td'ib min al-dhanb hu- 
man Id dhanb lahu)," inquiring whether or not it was in the Sahihcyn and 
how it relates to the issue of that person's legal competence.'^* Of the 
twenty-one recorded requests that the Shafi'i prodigy al-Nawawi (who 
began his studies in the wake of Ibn al-Salah's death and remained 
firmly within his orbit in hadlth study), received from everyday citi- 
zens of Damascus asking if a certain hadlth was authentic or not, the 
scholar employs the Sahihcyn in four responses (most are negative). '^'^ 
One questioner even inquires directly if the Sahihcyn or other famous 



'^' Abu Muhammad al-Husayn b. Mas'ud al-Baghawi, Masabih al-sunna, 2 vols in 
1 vol. (Beirut: bar al-Qalam, [197-]), 1:2. 

'^' Ibn al-Salah, Fatdwa Ibn al-Salah (Cairo: Maktabat Ibn Taymiyya, [1980]), 19. 
Ibn al-Salah replies that the hadith was not in al-Bukharl's or Muslim's collections nor 
does it have a firm isndd {isndd thabt). 

'^^ Al-Nawawl, Fatdwd al-imdm al-Nawawi, 177-192. For example, one person asks 
about whether the hadlth "/a saldt li-jdr al-masjid ilia fi al-masjid'' is in the Sahihayn; 
ibid., 191. 



THE CANON AND THE NEEDS OF THE COMMUNITY 247 

collections include any non-authentic hadlths. Al-Nawawi replies that 
all the hadlths of al-Bukhari and Muslim are authentic, while the 
Sunam of Abu Dawud, al-Tirmidhi and al-Nasa'i include varied levels 
of weak and sound hadlths.'"'' 

The referential role of the Sahihayn canon even facilitated the study of 
hadith among aspiring young students. Zayn al-Din al-'iraqi produced 
a manual using the Sahihayn in the same manner as al-BaghawI but 
designed it for students of hadith. In the introduction to this book, his 
Taqnr al-asanidfi tartib al-masanid, al-'Iraqi explains that he has collected 
a selection of hadlths for his son, since a student of hadith needs to 
memorize a number of reports in order to dispense with carrying heavy 
loads of books. Since in his time chains of transmission had grown too 
long to have any significant number of one's own living isndds, to the 
Prophet, al-'iraqi states that he has collected hadlths from the books 
of early scholars [al-mutaqaddimun) instead. If the hadith appears in 
the Sahihayn, he states, he provides no isndd, because its authenticity is 
"agreed on [muttafaq 'alcyhi)." If the report is not found in al-Bukharl's 
or Muslim's works, he provides isndds from other major collections.'^' 



3. The Need for an Exemplum: 
Aristotle's Poetics and the Canon that Sets the Rule 

Al-Bukharl and Muslim were not just used to prove the authenticity of 
Prophetic reports, but also to authoritatively shape the study of hadith. 
Just as the SdhWcyn canon served as a trump card in debates over individ- 
ual hadlths, so did scholars like al-Khatib al-Baghdadl and Ibn al-Salah 
employ it to elaborate the tenets of hadith transmission, criticism and 
its applications in deriving law. As Stanley Fish notes in his discussion 
of the durability of literary canons, "If Shakespeare is on your side in 
an argument, the argument is over'"-'' In this sense both Shakespeare's 
works and the Scihihoyn are canonical in that they are standards that 
can be employed to set the rules of a genre. They are the kanon to be 
imitated, the exemplum in whose ingenious pages lie the methods of 
mastering a science. Aristotle thus employs Homer's Iliad and Odyssey 



"'• M-n&wawi, Fatdwd, 177. 

'^' Al-'IraqI, Taqrib al-asdnidji tartih al-masdmd, ed. 'Abd al-Mun'im Ibrahim (Riyadh: 
Maktabat Nizar Mustafa al-Baz, 1419/1998), 14. 
'28 Fish, 12-15. 



248 CHAPTER SIX 

in his exposition of the proper components and characteristics of epic 
poetry. Amid his discussion of how well Homer embodied excellence 
in this genre, he states, "Homer deserves acclaim for many things, but 
especially because he alone among [epic] poets is well aware of what 
he himself should do.'"-^ For Aristotle, Homer's conscious mastery of 
his art provides the ultimate example for appreciating and writing epic. 
Homer's unparalleled methods themselves act as Aristotle's proof texts. 
As Fish realizes, a text thus becomes canonical when a community 
recognizes that it is the thing to which "all workers in the enterprise," 
or, in Aristotle's case, the genre, "aspire."'^" 

Just as Aristotle invoked Homer, prominent architects of the hadlth 
tradition declared al-Bukharl and Muslim the exemplum that sets the 
rule. Ibn Hajar states that "there is no doubt about the preeminence 
of al-Bukharl and Muslim over both the people of their own time and 
those who came after them from among the imams of that science in 

terms of knowledge of authentic and flawed hadlths " If someone 

opposes their work or their judgment on authenticity, "there is no doubt 
that [al-Bukharl and Muslim] supersede all others in this." "Objec- 
tion," he adds, "is thus fended off from them globally. . . ."'^' Al-Hazimi 
describes al-Bukharl as the best of his time in hadith collection and 
criticism, "and in light of the certainty of his station in these matters 
there is no way to object to him in that subject.'"^^ Ibn Taymiyya (d. 
728/1328) states that not even Ibn Khuzayma or Ibn Hibban approach 
al-Bukhari's mastery. As the result of his consummate skill, in the vast 
majority [jumhur) of instances in which someone criticized material 
that al-Bukhari approved, "his [al-Bukhari's] opinion is more favored 
than those of his detractors." '^^ Al-MaqdisI stated that the Sahihayn had 
become "proofs for the people of Islam [hujia li-ahl al-islam)." He claims 
that hadlth scholars since their time have thus focused on commenting 
on and studying the two books, since it is not possible to add anything 
more to that science [san'd)}''* 



'^' Aristotle, "Poetics," in The Norton Antholog)) of Theory and Criticism, ed. Vincent B. 
Leitch (New York: WW. Norton Co., 2001), 1 12. 

"» Fish, 12-15. 

'" Ibn Hajar, Hadj al-sdri, 502. 

"^ Al-HazimI, ShurUt al-a'imma al-khamsa, 59. 

"' Ibn Ttiymiyy 3., Majmu' al-fatawd, 1:256. 

"* Muhammad b. Tahir al-MaqdisI, Kitdb al-jam' bayn kitabay Abi Jiasr al-Kaldbddhi 
wa Abi Bakr al-Isbahdni, 2. 



THE CANON AND THE NEEDS OF THE COMMUNITY 249 

One of the most obvious areas in which al-BukharT and Muslim 
impacted the rules of hadith criticism was the definition of 'authentic' 
reports. Al-BaghawT testified to this when he equated the Sahihayn with 
authentic hadlths in general. One of the flaws that could undermine 
the authenticity of a hadith was "irregularity (shudhUdh)." The definition 
of 'irregular [shddhdli)' hadiths, according to the consensus of Sunni 
hadith scholars by the eighth/fourteenth century, was a report that 
contradicted a more reliable source, such as a better-attested hadith or 
a verse of the Qjur'an.''-^ Some earlier scholars like al-Khallll, however, 
had defined shddhdh much more broadly, and thus more dangerously, as 
a report whose only flaw is that it is narrated through only one chain 
of transmission. Here al-Khallll had followed his teacher al-Hakim 
al-Naysaburl, who wrote that shddhdh hadlths are those narrated by a 
trustworthy [thiqd) transmitter but whose text is not corroborated [ad 
mutdba') from his source.'^'' Later scholars such as Ibn al-Salah and 
Ibn Hajar fiercely rejected al-Khallll's definition because it would 
compromise prevailing understandings of the definition for authentic 
hadlths. Ibn al-Salah uses two hadlths "included in the Sahihayn" that 
would fall under al-Khallll's definition to prove that his definition was 
flawed. Ibn Hajar underscores this objection, arguing that not even al- 
Bukharl's and Muslim's methodologies could live up to what al-Khallll 
had proposed.'''' Ibn Hajar offers his final definition for sahih hadlths 
thus: "a report whose isndd connects to the Prophet via the narration 
of totally upstanding transmitters in command of what they transmit 
or, if not totally, supported by others like them, and is not shddhdh or 
afflicted with a flaw (mu'all)." Significantly, he immediately adds that he 
has tailored this definition specifically to al-Bukharl and Muslim. He 
explains: "I say this because I have considered many of the hadlths of 
the Sahihayn and have found that the ruling of saMh cannot be conferred 
upon them without this [definition].'"^" 



"'' See al-NawawI, al-Taqrib, 12; al-Dhahabi, al-Muqi^a, 42; Ibn Kathir, al-Bd'ith al~ 
hathith, 48-50; al-'iraqi, al-Taqyid wa al-iddh, 88; Ibn al-WazIr, Tanqih al-an^dr, 150-4; 
al-Sakhawl, Fath al-mughith, 1:244-8. 

'''' Al-Khallll, al-Irshdd, 13. Here al-Khallll states that, contrary to al-Shafi'l's opin- 
ion (and that of later orthodoxy), a shddhdh hadith is not one that disagrees with a 
more reliable source, but rather what "has only one isndd {laysa lahu ilia isndd wdhidj"; 
al-Hakim, Ma'rifat 'ulum al-hadith, 148. 

™ Ibn Hajar, al-Nukat 'aid kitdb Ibn al-Saldh, 40. For more on this debate, see Ibn 
Rajab, Sharh 'Hal al-Tirmidhi, 1:450-62. Ibn Rajab maintains that al-Bukharl, Muslim 
and others like al-Shafi'l defined shddhdh and munkar differently than al-Hakim and 
al-Khallll. 

"" Ibn Hajar, al-Mukat 'aid kitdb Ibn al-Saldh, 134. 



250 CHAPTER SIX 

Al-Bukhari and Muslim were also frequently invoked as the exem- 
plum that set the rules of selecting acceptable hadith transmitters. In 
his Kifayafi 'Urn al-riwaya, al-Khatib al-Baghdadi states that the general 
practice among hadith scholars is not to accept any criticism of a nar- 
rator unless the critic has explained the reasons for his objection. He 
proves this point by explaining that "this was the practice of the imams 
from among the masters of hadith and critics such as Muhammad b. 
Isma'll al-Bukharl and Muslim b. al-Hajjaj al-Naysaburl.""^ Ibn al-Salah 
follows al-Khatib, invoking Muslim's use of impugned transmitters, such 
as Suwayd b. Sa'ld, and al-Bukharl's reliance on 'Ikrima, Ibn 'Abbas's 
pro-Kharijite client.'*" 

The Sahihayn canon, however, was a double-edged sword that could 
be wielded by parties at odds with one another on the proper rules of 
hadith criticism. The case of accepting reports from heretics {mubtadi') 
clearly illustrates this. Some early scholars like al-Shafi'l generally per- 
mitted narrating from them, while more strict critics condemned it. A 
middle ground formed with scholars like Malik and Ibn Hanbal who 
accepted hadlths transmitted from heretics provided they were neither 
extremists nor proselytizers.'*' The Shafi'l legal theorist of Baghdad, 
Ahmad b. 'All Ibn Barhan (d. 518/1 124), defended the Shafi'l school's 
stance on the issue. He states that one can accept reports from all her- 
etics except the extremist Shiite group, the Khattabiyya, and Shiites who 
rejected the first two caliphs (Rafida).'*'^ As proof, Ibn Barhan invokes 
the umma's consensus on the authenticity of the Sahihayn: al-Bukharl 
and Muslim included hadlths narrated from Qadarites like Qatada b. 
Di'ama and the Kharijite 'Imran b. Hittan, so it must be permissible 
for others to imitate them.'*^ 

Ibn al-Salah, however, employs the Sahihayn canon to espouse what 
became the more strict mainstream opinion. Like Ibn Barhan, he states 
that rejecting the narrations of all heretics [mubtadi'un) is untenable 
because al-Bukharl and Muslim rely on them in both their primary iusul) 



'™ Al-Khatib, al-Kifqya, 1:338. 

"° Ibn al-Salah, Muqaddima, 221. 

'" For a summary of this, see al-Khatib, al-Kifaya, 1:384 ff.; Ibn Rajab, Shark 'Hal 
al-Tirmidhi, 1:53-56. 

"^ For a discussion of the Khattabiyya, see W. Madelung, "Khattabiyya," EP. Al- 
Dhahabl explains that al-Shafi'l had not allowed narration from these groups because 
they allowed lying; al-Dhahabi, al-Muqi^a, 85. 

'*' Ahmad b. 'All Ibn Barhan, al-Wusul ild al-usul, ed. 'Abd al-Hamid 'All Abu Zayd, 
2 vols. (Riyadh: Maktabat al-Ma'arif, 1404/1984), 2:184-5. 



THE CANON AND THE NEEDS OF THE COMMUNITY 251 

and auxiliary [shawdhid) hadlths. He adds, however, that the Sahihayn 
do not include proselytizing heretics, from whom transmission would 
be forbidden.'** 

The Sahihayn canon did not only serve as an exemplum that could 
be employed to set the rules of hadlth criticism. The two works could 
also be referred to in order to elaborate how Prophetic hadlths should 
be employed in deriving law. In his al-Wusul ild usul, for example, Ibn 
Barhan describes the case advanced by some Hanafi scholars for the 
broad acceptance of mursal hadlths in deriving law. Arguing against 
transmission-based scholars who generally considered a mursal hadith 
to be flawed due to the break in its isndd, these Hanafis had suppos- 
edly claimed that the ashdb al-hadith had in fact accepted mursal reports. 
Al-BukharT and Muslim, they argued, had even included many mursal 
hadlths in their Sahihs,.^*'' This claim was, of course, highly erroneous. 
The Sahihayn are certainly not replete with mursal hadlths, and Muslim 
himself specified that mursal hadlths were not acceptable proofs {hujja) 
in the introduction to his collection.'*'' 



The Limits of the Canon's Authority: The Dialogic Power of the Sahihayn 

The power of the SaMhoyn canon stemmed from the assertion that the 
absolute authenticity of the hadlths they contained would validate one's 
stance in argument or exposition. Although Abu Ishaq al-Isfarayinl's 
statement obliging scholars to rule on issues according to the contents 
of Sahthayn had allowed for the possibility of interpreting a hadith in a 
manner that could neutralize its legal import, this did not obscure the 
thrust of his declaration: ruling against a hadith from the two books was 
tantamount to breaking consensus. Abu Nasr al-Wa'Ul and al-Juwayni 
reinforced this claim by affirming the absolute authenticity of the two 
collections. Al-Ghazali's remark that a jurist must rule according to the 
Sahihayn or break with ijmd' merely represented the crystallization of 
this edifice of authority built around the Sahihayn in the first half of 
the fifth/ eleventh century. 



'" Ibn al-Salah, Muqaddima, 299-300. 

"^ Ibn Barhan, al-Wusulild al-usul, 2:179. 

"*' Muslim, Sahih, 1 :24. This claim is so ludicrous that it is difficult to believe that any 
educated Hanafi would make it. It may be that Ibn Barhan was unwittingly engaging 
in a 'straw man' argument. 



252 CHAPTER SIX 

The power of the canon, however, was a fagade that could only 
intimidate or convince those confronted with it from outside. It was 
an illusion conjured and maintained in the relative space between 
adversaries in the arena of debate, or between author and intended 
reader in expository writing. An individual Hanafi jurist or Ash'ari 
theologian felt no compunction about ignoring or rejecting a hadlth 
from al-Bukharl's or Muslim's books if it clashed with his own position. 
As the great Hanafi legal theorist Abu al-Hasan al-Karkhi (d. 340/952) 
proclaimed, his default position [ad) is that any Qur'anic verse or hadith 
that "contradicts the stance of our school [ashabina) is assumed to have 
been either abrogated or set aside in favor of another [tuhmalu 'ala 
al-naskh aw 'aid al-tarjih)."^*^ Such policies led the Damascene scholar 
Tahir al-Jaza'iri (d. 1920) to note incisively, "The jurists interpret away 
(ju'awwilun) any hadith that disagrees with their madhhab, or oppose it 
with another hadith even if it is not well-known, even if that [first] 
hadith is found in the Sahihayn."^*^' 

In general, it was not uncommon for Muslim scholars engaged in 
debate to insist on a rule in one context, then invert it in order to defend 
their school's stance in another Ibn al-JawzT, for example, adhered to 
the Hanball school that had led the campaign for the admission of dhdd 
hadlths in elaborating dogma as well as law. When responding to the 
Shiite claim that 'A'isha was guilty of unbelief {kujr) for fighting 'All, 
however, Ibn al-Jawzi changed positions. He argued that the hadith 
cited by Shiites as evidence for this, "You will fight him (i.e., 'All) and 
you will be wrong [satuqdtilinahu wa anti zdlimd)" "is all by reports of 
limited attestation [dhdd)" and "is thus not epistemologically certain by 
this means [Idyuqta'u bi-mithlihi)."^*^ 

Treatises on the legal theory reveal the Sahihayn canon's limited exis- 
tence in relative space. In general, usul books from both the Hanafis 
and the 'Majority' [al-jumhur) school espoused by Shafi'ls, Malikis and 
most Hanbalis offer nothing but silence about the place of the Sahihayn 
in Islamic epistemology Even al-Khatib al-Baghdadl, a Shafi'T/Ash'arT 
hadith scholar who was very aware of the rhetorical power of the 
Sahihayn canon, reserves no place for it in his Kitdb al-faqih wa al-mutafaqqih 
(Book of the Jurist and Law Student), a work designed to familiarize 



'*' Al-Karkhl, al-Usul allaS 'alayhd madar juru' al-hanajijya, 84-5. 
'*' Al-Jaza'iri, Tawjih al-nazar ild usul al-athar, 1:320. Khalll MuUa Khatir agrees; 
MuUa Khatir, Makdnat al-Sahihayn, 1 54. 
'*" Ibn al-jawzl, al-Munta^am, 15:296. 



THE CANON AND THE NEEDS OF THE COMMUNITY 253 

hadith scholars with usul al-jiqh. Although he notes that dhad hadlths 
agreed upon by the umma yield certainty {'Urn), he dismissingly relegates 
"the sunan and the sahih books (sikdh)" to the category of reports that 
convey only probability [zann)}^^' 

One of the few instances in which the epistemological standing of 
the Sahihayn is mentioned at all in an usul work is a denial of any special 
status. Discussing the well-established fact that dhdd hadlths yield only 
probability, the Shafi'i legal theorist Ibn Barhan (d. 518/1124) rejects 
the opinion of "some ashdb al-hadith" who say that the authenticity of 
what is narrated in the Sahihayn is absolutely certain (maqtu' bi-sihhatihi)}^^ 
He explains that al-Bukhari and Muslim were not infallible {ma 'sum 'an 
al-khata'), since hadith scholars have criticized their work and found 
errors (awhdm). If their works were epistemologically certain, this would 
be impossible. Ibn Barhan further rejects any exceptional status for the 
Sahihayn by arguing that the only evidence supporting this claim, the 
acceptance of their hadlths by consensus, does not prove their absolute 
authenticity. The Muslim community accepted the two books because 
they felt that their contents were legally compelling; but not all that 
is legally compelling is absolutely authentic. ''^^ Although Ibn Barhan 
attributes this opinion to more extreme transmission-based scholars, he 
is in effect demolishing the argument made by his fellow Shafi'l/Ash'arls 
Abu Ishaq al-Isfarayini and al-Juwayni. The irony of this situation lies, 
of course, in Ibn Barhan's above-mentioned claim about narrating 
from heretics, where he invokes the umma's agreement on the Sahihayn 
to prove his point. The power of the canon thus appears only in the 
dialogic space of debate and exposition. Even within the scope of 
one book like Ibn Barhan's al-Wusul, a scholar can wield the canon's 
authority against opponents in one instance and then circumscribe it 
in other, less combative settings. 

Although ignored or contested in usul works, the source and degree 
of the Scihihciyn canon's authority as originally declared by Abu Ishaq 
al-Isfarayinl was finally properly acknowledged by Ibn al-Salah in 
the seventh/thirteenth century. In several of his hadith works, he 
states that the authenticity of al-Bukhari's and Muslim's hadlths "is 
absolutely certain, and epistemologically certain discursive knowledge 



Al-Khatib, Kitdb al-faqih wa al-mutafaqqih, 1:278. 
Ibn Barhan, al-Wusul ild al-usul, 2:172-3. 
Ibn Barhan, al-Wusul ild al-usul, 2:174, 



254 CHAPTER SIX 

{'ilmyaqini nazan) occurs with [them].'"" He exempts from this claim, 
however, that "small amount of material [ahruf yasird)" criticized by 
major scholars like al-Daraqutni, since one could not claim consensus 
on its authenticity.'''* 

Ibn al-Salah's claim proved a tempting foil for later hadith scholars, 
who have devoted a great deal of energy to arguing for or against its 
validity. Those who have supported the notion that the contents of the 
Sahihayn yield certain discursive knowledge include prominent figures 
such as Ibn Taymiyya, Ibn Kathir, al-'Iraqi, al-BulqInl, and the major 
formulators of the late Sunni tradition: Ibn Hajar al-'Asqalani, al- 
Sakhawl, Zakariyya al-Ansari (d. 926/1520) and Ibn Hajar al-Haythami 
(d. 974/1597).'"''^ More recently, modern scholars such as Khalil MuUa 
Khatir have joined these ranks. Those who have disagreed with his 
claim have been far fewer in number: Ibn al-Salah's virtual disciple, 
al-NawawI, his opponent al-'Izz b. 'Abd al-Salam, Badr al-Din Ibn 
Jama'a (d. 733/1333), and the Salafi maverick Muhammad b. Isma'll 
al-AmIr al-San'anI (d. 1768).'-"' 

Ibn al-Salah's claim, however, has done little to earn the Sahihayn 
any special absolute status in Sunni epistemology Although this dis- 
cussion has attracted the attention of generations of hadith scholars, 
it has not spread beyond the limited genre of the technical study of 



'^' Ibn al-Salah went through several phases in his opinion on this issue. He states 
in his Muqaddima that he had originally believed that the hadlths of the Sahihayn, like 
all dhdd reports, yield only probability {^anri). Later he realized that the infallible con- 
sensus of the umma on the two works meant that what seemed like probability was 
in fact certainty. In this work and in his Sijdnat SaMh Muslim, Ibn al-Salah asserts this 
for the contents of both al-Bukharl and Muslim, not just the hadlths that they both 
agreed on. His follower, al-NawawI, tells us that in another (earlier?) work (juz') Ibn 
al-Salah stated that the truthfulness of what al-Bukharl and Muslim both included is 
absolutely guranteed. Ibn Hajar quotes this from Ibn al-Salah's lost sharh of Muslim; 
al-NawawI, Sharh Sahih Muslim, 1:128; Ibn Hajar, al-Nukat 'aid kitdh Ibn al-Saldh, 112; 
see n. 154 below. 

'^* Ibn al-Salah, Sijdnat SaMh Muslim, 85; idem, Muqaddima, 170-1. 

'" Ibn Taymiyya, Majmu'fatdwd; 1:25; 618:20; idem, llm al-hadith, ed. Musa 
Muhammad 'All ([Cairo]: Dar al-Kutub al-Islamiyya, 1404/1984), 100; Ibn Kathir, 
al-Bd'ith al-hathith, 30; al-BulqInl, 172; Ibn Hajar, Mizhat al-na^ar, 29 (Ibn Hajar adds 
another qualification to this claim, namely that it only applies to what is in the Sahihayn 
but does not contradict their other contents);al-SakhawI, Fath al-mughith, 1:74 (he fol- 
lows Ibn Hajar); al-Ansarl, Path al-bdqi, 83-4 (he also follows Ibn Hajar); Ahmad b. 
Muhammad Ibn Hajar al-HaythamI, al-Fatdwd al-hadithiyya, 2nd ed. (Cairo: Matba'at 
Mustafa al-Babi al-Halabi, 1390/1970), 92. 

''''' Al-'IraqI, al-Taqyid wa al-iddh, 38; al-NawawI, al-Taqrib, 6; Ibn Jama'a, 128-9; 
al-San'anI, Thamardt al-na^ar fi 'ilm al~athar, ed. Ra'id b. SabrI b. Abl 'Alafa (Riyadh: 
Dar al-'Asima, 1417/1996), 131, 137. 



THE CANON AND THE NEEDS OF THE COMMUNITY 255 

hadith science [mustalahat al-hadith). Usui texts, treatises on madhhab law, 
theology or hadlth-based legal derivation (what is referred to asfiqk al- 
sunnd) rarely go beyond the established references to dhdd or mutawdtir as 
epistemological classes for reports. The general inconsequence of the 
discussion surrounding Ibn al-Salah's statement is further revealed by the 
argument of his opponents. Far from constituting any massive assault 
on the canon, al-Nawawl's rebuttal of Ibn al-Salah actually affirms 
the canonical role of the Sahihayn. Like Ibn Barhan, al-NawawI (who 
is followed by Ibn Jama' a) only rejects the notion that the community's 
collective acceptance of the Sahihayn renders their contents epistemo- 
logically certain. The fact of this consensus on the two works stands 
uncontested, as does their compelling power in debate. Al-NawawT 
affirms this; the special status of al-Bukharl's and Muslim's collections 
resides in the fact that their contents have been lifted above the need 
for critical examination.'" 

The undeniable proof of the relative nature of the canon's authority, 
however, lies in the willingness of legal or theological schools to unhesi- 
tatingly ignore or criticize a hadith from the SaMhoyn if it counters their 
positions. When this stems from a disagreement over the interpretation 
of a hadith, it entails no transgression of the canon's authority. The 
Hanafis al-SarakhsI and al-Nasafi had, after all, asserted that muhaddiths, 
were not qualified to appreciate the true legal implications of their 
hadlths. On the question of tasriya, or tying the udders of a milk-ani- 
mal-for-sale in order to temporarily increase its milk and attract buyers, 
Hanafis rejected explicit reports from al-Bukharl's Sahih discouraging 
the practice. While both al-Bukharl and the Shafi'l school followed a 
hadith that granted a buyer deceived by such a scheme the right to a 
refund and an amount of dates in compensation, Hanafis held that the 
original sale was valid. 'Abd al-'AzIz al-Bukharl explicitly states that 
this hadith is authentic and found in the Sahihayn. Yet it contradicts 
juridical reasoning based on the Qur'an and sunna and thus cannot be 
acted on. According to Hanafi jurisprudence, the Qur'an and juridical 
reasoning dictated that a transaction only requires the health or good 
quality of the item sold [saldmat al-mabi'). A paucity of milk does not 
compromise this.''^** 



''" Al-Nawawl, Shark Sahih Muslim, 1:128. 

'^" A sizable minority opinion within the Hanafi school, following the work of Ibn 
Aban, requires a narrator to have sufficient legal mastery of the material he transmits 
in order for his hadith to supersede qiyds. Abu Hurayra, who is the Companion who 



256 CHAPTER SIX 

The Maliki Abu al-Walid al-Baji also asserted the jurists' right to 
disagree with the legal implications of hadlths from the Sahihayn or 
their authors' legal assumptions. He states that "al-Bukharl is deferred 

to in the science of hadlth, but not in jurisprudence {'ilm al-fiqh) " 

Al-Baji then refers to some of al-Bukhari's chapter titles to show how 
he did not derive the correct rulings from his hadlths and that he might 
even have sometimes hunted for proof texts to support his own legal 
opinions.''^ 

Not all rejections of hadlths from the Sahihayn, however, stemmed 
from differences in interpretation. Adherents of legal and theological 
schools sometimes actually criticized their authenticity. The Hanaft 
school, for example, rejected material from both Sahths if their nar- 
rations proved too problematic. Hadlths dealing with the issue of the 
Prophet's prayer in the event of an eclipse [hadith al-kusuf), for example, 
proved exceptionally difficult to reconcile with one another When an 
eclipse surprised the Muslim community, the Prophet left his house and 
convened a public prayer. The hadlths detailing his prayer, however, 
disagree on the number of times the Prophet bowed [ruku'). The Hanafi 
hadlth scholar al-Zayla'l attempts to navigate the impossibly confused 
web of conflicting matn% for these hadlths in his Nasb al-raya, where 
he presents the contradictory reports from within the Sahihoyn and 
the other Six Books. The most reliable version according to al-Zayla'T 
is that narrated by 'A'isha describing only one bow, while the others 
have two, three, four or five bows.""" As a result, the Indian Hanafi 
Muhammad 'Abd al-Hayy al-Laknawi (d. 1886—7) concludes that his 



transmits this hadith, is not considered so qualified. See, for example, al-Shashi, Usui 
al-Shdshl, 272; 'Abd al-'AzIz b. Ahmad al-Bukharl, Kashf al-asrdr, 2:381. For discussions 
of tasriya, see Ibn Hajar, Fath al-bdri, 4:458-60; al-LaknawI, ^o/a?' al-amdm, 66. For this 
hadlth, known as hadith al-musarrdt; see Fath # 2148; Sahih al-Bukhdri: kitdb al-buyu, bdh 
al-nahy ti't-bd'i' an tdyuhqffila al-ibil. 

'^' Nasir al-Din Ahmad b. Muhammad b. al-Munayyir al-Maliki, al-Mutawdn 'aid 
abwdb al-Bukhdn, ed. 'All Hasan 'All 'Abd al-Hamid (Beirut: al-Maktab al-Islami, 
1411/1990), 36. See also, al-Kirmanl, Shark Sahih al-Bukhdri, 1:5 for the author's 
opinion, 

">" Al-Zayla'l, Nasb al-rdya, 2:225-3 1 . 'A'isha's narration can be found in Fath # 1 058, 
Sahih al-Bukhdri: kitdb al-kusdf, bdb Id tankasiju al-shams li-mawt ahad. For a brief sample 
of the conflicting narrations of this tradition, see: Sahih al-Bukhari: kitdb al-kusuf, 
bdb tul al-sujadji al-kusuf, bdb al-saldt f kusiif al-qamar, bdb al-rak'a al-uldji al-kusif a'wal; 
Sahih Muslim: kitdb al-kusdf bdb saldt al-kusuf bdb md 'urida said al-Nabi (s) min amr al- 
janna wa al-ndr, bdb dhikr al-nidd' bi-saldt al-kusuf saldt jdmi'a; Sunan Abi Datviid: kitdb 
al-istisqd', bdb man qdla arba' raka'dt. 



THE CANON AND THE NEEDS OF THE COMMUNITY 257 

school had abandoned the Sahihayn'?, hadiths on this issue, since they 
had "become grossly problematic [idtaraba idtirab"" fihish"").'"^^'^ 

Perhaps the most starkly partisan criticism of a hadith in the Sahihayn, 
however, occurs at the hands of the Shafi'i school that had played 
such an important role in canonizing the two works. Muslim includes 
a narration by the Companion Anas b. Malik in which he states that 
he had prayed behind the Prophet and the first three Caliphs but had 
heard none of them say the basmala out loud. Shafi'is from the time of 
al-Daraqutni and al-BayhaqI criticized this narration from Sahih Muslim, 
which explicitly contradicted the madhhab's stance on the basmala. After 
a lengthy chapter in his al-Sunan al-kubrd featuring hadiths showing 
that one should say the basmala aloud during prayer, al-BayhaqI has 
a chapter on hadiths arguing the opposite. For each tradition (cluster 
of narrations) opposing his school's stance, he finds some problem 
undermining its reliability. Al-Bayhaqi notes that the hadith of Anas 
(narrated via al-Awza'i •<— Q_atada b. Di'ama) is featured in Sahih Muslim, 
and he mentions that this and several other narrations through Qatada 
all have sections specifically saying that "I did not hear any of them 
say Bismilldh al-Rahmdn al-Rahim . . ." or "and they did not say [it] . . . 
out loud." Al-Bayhaqi rebuts these narrations, however, by arguing 
that others had narrated this hadith from Shu'ba •<— Qatada •<— Anas 
without the explicit negation of the basmala. Relying on al-Daraqutni's 
opinion, al-BayhaqI favors this latter version of the hadith, which al- 
Bukharl includes in his Sahih.^^''^ 

Oddly, Ibn al-Salah literally uses Muslim's narration through Anas as 
a textbook example of a flaw {'ilia) occurring in the text of a hadith, an 
example that became enshrined in the pedagogical Aljiyya poem that al- 
'Iraql composed for hadith students based on Ibn al-Salah's Muqaddima. 
Following the takhrij ranking system, Ibn al-Salah favored the version 
of the hadith agreed upon by both al-Bukharl and Muslim, without 



""' Al-Laknawl, Z'^fa^ al-amdni, 400; al-Qanubl, al-Sayf al-hddd, 111. The Hanafts 
stuck with the "default in prayer (al-asl fi al-salaif namely that raAw' occurs only once 
{al-tawahhud Ji al-ruku'). 

'*^ Al-BayhaqI, al-Sunan al-kubrd, 2:73-76, kitdb al-saldt / bdb man qdla Idyajharu bihd; 
Sahih al-Bukhdn: kitdb al-saldt / bdb 240 / hadith #1; al-BayhaqI, Ma'rifat al-sunan wa 
al-dthdr, 1:524; al-Daraqutnl, Sunan al-Ddraqutm, ed. 'Abdallah Hashim al-Madanl, 4 
vols, in 2 (Cairo: Dar al-Mahasin li'1-Tiba'a, 1386/1966), 1:316. Al-Daraqutnl does 
not note that any of these narrations appear in Sahih Muslim, nor does he include this 
criticism in his Kitdb al-tatabbu'. 



258 CHAPTER SIX 

Anas's addition of "not one of them said [the basmala] out loud." He 
further undermines Anas's narration by citing one Sa'ld b. Yazld asking 
Anas about the basmala, to which Anas replies, "indeed you have asked 
me about something on which I have memorized no [hadlths] , nor has 
anyone before you asked."""^ Later, prominent Shafi'ls such as al-'Iraqi, 
Ibn Hajar and al-Ansarl followed Ibn al-Salah's argument."'* 

Scholars like Ibn Hajar could not conceal the clear partisan moti- 
vations for criticizing Muslim's report and noted that opinions on its 
authenticity break down along madhhab lines between those who affirm 
saying the basmala out loud and those, like the Hanafis, who do not. 
As a Shafi'i, Ibn Hajar ultimately sided with Ibn al-Salah's criticism 
of Muslim. Consequently, his Hanafi nemesis in Cairo, Badr al-Din 
al-'Ayni (d. 855/1451), mocked him for rejecting a perfectly valid nar- 
ration he otherwise would have considered authentic."" 

Leading Ash'ari theologians such as al-Baqillani, al-Juwaynl and 
al-Ghazali also severely criticized a hadlth appearing in both the 
Sahihayn in which the Prophet prays for the forgiveness of the most 
flamboyant hypocrite (jnundfiq) in Medina, the Khazraj leader Abdallah 
b. Ubayy."''' Ibn 'Umar narrates that when the Prophet went to pray 
over the deceased 'Abdallah 's grave, 'Umar b. al-Khattab objected. He 
reminded the Prophet that God had forbidden Muslims from praying 
for the forgiveness of hypocrites, referring to the Qur'anic verse, "Pray 
for their forgiveness or do not pray, even if you pray seventy times God 
will not forgive them (Qur'an: 9:80).'"*'' The Prophet replies that in 
the verse God had "given [him] a choice (khayjaram Allah) ," and that 
he "will exceed seventy [times]." 



"'^ Ibn al-Salah, Muqaddima, 261. Al-'Iraql remarks how bizarre it is for Ibn al-Salah 
to use a hadlth from Muslim as an example of a flawed narration after asserting that 
everything in the Sahihayn is absolutely certain. He justifiably explains this, however, by 
adding that Ibn al-Salah had exempted from this claim material that had been criticized 
by great critics like al-Daraqutnl; al-'Iraql, al-Taqyid wa al-iddh, 98. 

"'' Al-'IraqI, al-Taqyidwa al-iddh, 98, 100; Ibn Hajar, Fath al-bdri, 2:289-91; al-Ansarl, 
Path al-bdqi, 198-200; cf Ibn 'Abd al-Barr, al-Tamhid, 2:228-31. 

"'' Ibn Hajar, Intiqdd al-i'tirddfi al-radd 'aid al-'Aymfi sharh al-Bukhdn, ed. Hamdl b. 
'Abd al-Majld al-Salafi and Subhl b. Jasim al-SamarraX 2 vols. (Riyadh: Maktabat 
al-Rushd, 1413/1993), 1:369. For a discussion of Ibn Hajar's astonishingly 'academic' 
rivalry with al-'Aynl, see Anne F. Broadbridge, "Academic Rivalry and the Patronage 
System in Fifteenth-Century Egypt: al-'Aynl, al-MaqrIzI and Ibn Hajar al-'AsqalanI," 
Mamluk Studies Review 3 (1999): 85-108. 

'^^ See SMh al-Bukhdn: kitdb tafsir, sura 9, bdb 13; Sihih Muslim: kitdb al-tafsir/ sural 
al-Tawba/Bdb 13. 

"" "Istaghfir lahum aw Id tastaghjir lahum, in tastaghfir lahum sab'ina marraf" fa-lan yaghfira 
Alldhu lahum." 



THE CANON AND THE NEEDS OF THE COMMUNITY 259 

This hadith caused a great uproar amongst Ash'ari theologians and 
legal theorists, because it implied that the Prophet felt that he could 
circumvent the command implicit in the verse, namely not to pray 
for hypocrites. Ibn Hajar explains that a number of prominent schol- 
ars had therefore attacked the authenticity of the hadith despite its 
widespread narrations and the Shaykhayn'% agreement on it. He quotes 
Nasir al-Din Ahmad Ibn al-Munayyir (d. 683/1284), who states that 
Abu Bakr al-Baqillani said, "It is not possible to accept the hadith or 
that the Prophet said it." In his Taqnb, al-BaqUlanl supposedly said that 
"this hadith is one of the ahad reports whose soundness [thubutuha) is not 
known."""" Al-Juwaynl says in his Burhdn that "the ahl al-hadith have not 
deemed this sound.""'^ Al-Ghazall agrees in his Mustasja, asserting that 
"this is an dhdd report [khabar wdhid) that cannot be used to establish 
proof {hujjd) for the implications of speech [Ji ithbdt al-lughd); besides, it 
is more probably [zahard) not sahih.'""^^ 

Ironically, al-Ghazali's objection to this hadith demonstrates the 
paradox of the Sahihayn canon and its restriction to relative space. 
Although he undeniably questions the authenticity of this hadith in 
his Mustasja, earlier in his Mankhul he had defended it. There he insists 
that the Prophet's actions in the hadith neither compromised the truth 
of the Qjur'anic verse nor the reliability of the report. God had given 
him the choice to ask for forgiveness or not.''' Al-Ghazall wrote his 
Mustasja many years after the Mankhul, and it is possible that he simply 
changed his opinion on the hadith. Context, however, provides a more 
convincing explanation. The Mankhul is generally a polemical work 
directed at the Hanafi school. In it, the hadith about the Prophet pray- 
ing for 'Abdallah's forgiveness plays a role in the author's defense of the 
Shafi'l/Ash'arl notion of ^^majhum al-kalam^^ or methods for deriving the 
indirect legal implications of a divine injunction. Specifically, al-Ghazall 
is defending this notion against Hanafi critics who reject the authenticity 
of the hadith and thus its applicability as evidence for majhum al-kaldm, 
a type of proof considered invalid among Hanafis."^ In his Mustasja, a 



">' I was unable to find the statement quoted by Ibn Hajar in al-Baqillanl's Kitdb 
al-tamhid or the 1413/1993 Mu'assasat al-Risala edition of his al-Taqrib iva al-irshdd; 
Ibn Hajar, Path al-bdn, 8:430-1. 

">' Al-Juwaynl, al-Burhdn, 1:458. 

"" Al-Ghazall, al-Mustasfa, 2:87. For my rendering of majhum and lugha, see Bernard 
Weiss, The Search for God's Law, 1 17; Hallaq, A History of Islamic Legal Theories, 58. 

'" Al-Ghazall, al-Mankhul, 212. 

"^ For a discussion of a Hanafi perspective on one of the dimensions of majhum 
al~kaldm, dalil al-khitdh (i.e., the indirect implication from an injunction, so that if the 



260 CHAPTER SIX 

pedagogical tool written many years later after al-Ghazall had sworn 
off debate and returned to teaching at the Shafi'i/Ash'ari-dominated 
Naysabur Nizamiyya, he could comfortably question material that 
seemed to contradict the tenets of Ash'ari theory''^ As a young fire- 
brand polemicist in Baghdad, however, the writer of the Mankhul had 
to defend his Shafi'i school against its Hanafi opponents.''* 



Conclusion 

In the mid-fifth/ eleventh century, the Sahihayn canon stood ready to 
fulfill important functions for Muslim scholars in cities like Baghdad 
and Naysabur Studied extensively by the Sahihayn Network, focused 
by al-Hakim al-Naysaburi into a measure of authenticity and autho- 
rized by scholars like Abu Ishaq al-Isfarayini, Abu Nasr al-Wa'ili and 
al-Juwaynl, the Sahihayn provided an important convention for schol- 
arly debate and exposition. In a time when the legal discourse of the 
madrasa was drifting farther and farther from the specialized study of 
hadith, the two works became the most authoritative hadlth references 
for jurists more narrowly focused on law. Whether used in polemics 
or to buttress the proof texts relied on by a particular school in the 
language of a common convention, the S^hihoyn served as the measure 
of authenticity for prominent Shafi'is, Hanbalis and Malikis from the 
mid-fifth/eleventh century on. In the eighth/fourteenth century even 
the hadith-wary Hanafi school found itself grudgingly forced to adopt 
the common measure of authenticity. The canon's authority, however, 
was not absolute. It was a collaborative illusion summoned to provide 
common ground among rivals. Alone, within a particular legal or theo- 
logical school, the authoritative edifice of al-Bukharl's and Muslim's 
hadlths collapsed before interpretive differences or partisan agendas. 



Prophet says pay tithe on a certain kind of sheep one need not pay it on others), 
see Marie Bernand, "Hanafi Usui al-fiqh through a Manuscript of al-Gassas," 628; 
Ahmad b. 'All Ibn al-Sa'atI (d. 694/1294-5), Nihdyat al-wuml ild 'Urn al-uml, ed. Sa'd b. 
Gharlr b. Mahdl al-SulamI, 2 vols. (Mecca: Jami'at Umm al-Qura, 1418-19/1997-99), 
2:560 ff. 

'" For al-Ghazall's oath never to engage in debate again, see J. Brown, "The Last 
Days of al-Ghazzall," 95. 

"* 'Abbas Eqbal, ed., Makatib-e Jarsi-ye Ghazzali beh ndm-e fadd'el al-andm min rasd'el 
hojjet al-esldm (Tehran: Ketabforushl-ye Ibn Sina, 1333/[1954]), 12; George F. Hourani, 
'A Revised Chronology of Ghazali's Writings," Journal of the American Oriental Society 
104, no. 2 (1984): 290-1, 301. 



THE CANON AND THE NEEDS OF THE COMMUNITY 261 

The vaunted station of the two books, however, was not simply due 
to the declarations of scholars like al-Isfarayinl or al-Wa'ili. Al-Bukharl's 
and Muslim's works consistently bested other respected collections 
used for takhnj by meeting the highest levels of excellence established 
by the Sunni hadlth tradition as it reached its full maturity between 
the fifth/eleventh and seventh/thirteenth centuries. Implicit in this 
success, however, lay the potential for serious tension surrounding the 
place and role of the Sahihayn canon. Although scholars attempting to 
systematize the Sunni study of hadith like al-Khatib al-Baghdadi and 
Ibn al-Salah often employed the Sahihayn as the exemplum that set the 
rule, the Sunni hadith tradition operated according to rules external 
to the two books. As exemplified by the reaction to Ibn al-Salah's 
attempt to replace the living isnad with the Sahihayn, here lay the seeds 
of tension between the continuing practice of hadith critics and the 
institution of al-Bukhari and Muslim. If the canon was to maintain 
its air of compelling authority in the arena of discourse, a canonical 
culture would have to be forged to extend the two books the charity 
required to reconcile this tension. 



CHAPTER SEVEN 

THE PRINCIPLE OF CHARITY AND THE CREATION OF 
CANONICAL CULTURE 



Introduction 

By the end of the fifth/eleventh century, the Sahihayn had become 
synonymous with authenticity in Sunni discussions of the Prophet's 
legacy as well as an exemplum of excellence in hadlth scholarship. 
The institution of the canon, however, faced potent challenges from 
two different fronts. First, the pre-canonical past of the two works 
was fraught with fissures. The initial negative reactions to the sahih 
movement, al-Bukhari's checkered career and the fact that Naysabur 
scholars had ranked Muslim's collection above that of al-Bukharl all 
threatened the stability of the canon. Second, there existed inconsis- 
tencies between al-Bukhari's and Muslim's work on the one hand and 
the conventions of hadith criticism on the other In the post-canonical 
world, these inconsistencies created a tension between the institution 
of the canon and the Sunni hadlth tradition as it matured fully in the 
early seventh/thirteenth century. 

To protect and maintain the canonicity of the Sahihayn would require 
reconciling the canonical vision of the two works and the personas 
of their authors with both their pre-canonical past and the external 
rules of hadlth scholarship. This would entail reading the texts of al- 
Bukhari and Muslim according to the Principle of Charity, which calls 
for interpreting a text in the best possible light in order to bring into 
harmony external notions of truth and those presupposed within the 
text. Just as Davidson described the Principle of Charity's function in 
speech communities, so would participants in elaborating Sunni schol- 
arly culture treat the texts of al-Bukhari and Muslim with charity "in 
order to preserve a reasonable theory of belief" in the canon.' 

The worldview that demands the extension of charity to canonical 
texts can be termed the books' canonical culture. It is the environment 



Davidson, 196. 



THE PRINCIPLE OF CHARITY 263 

created and cultivated by the community to which the canon is bound, 
by an audience that recognizes that "canonizing a text . . . requires a 
commitment to make the best of it."^ Canonical culture rereads his- 
tory and text to reconcile them with canonical authority. The saga of 
al-Bukharl and Muslim can thus be viewed as a process of creating 
and maintaining the Sahihayn canonical culture, which emerged with 
the canonization of the two works in the late fourth/tenth and early 
fifth/ eleventh centuries. The earliest surviving elaboration of the canoni- 
cal culture consists of the image of al-Bukhari and Muslim forged by 
al-Khatib al-Baghdadi (d. 463/1071). The personas of the two scholars 
that he crafts in his Tarikh Baghdad established the dominant themes 
of the Sahihayn canonical culture: the place of al-Bukhari, Muslim and 
their works at the pinnacle of hadlth scholarship; the vindication of 
al-Bukhari from the scandal of the created lafz; al-Bukharfs superiority 
to Muslim; and the simultaneous complementary relationship between 
the two. Even after constructing the Sahihayn canonical culture, however, 
generations of scholars would resort to interpretive gymnastics and 
editorial revisions of history in order to maintain it. 

Mirroring the canonical culture established around the personas 
of al-Bukhari and Muslim was the extension of charity to the texts 
of the S^ihihayn themselves. Both before and after their canonization, 
the collection and criticism of hadith functioned according to rules 
that were external to al-Bukhari's and Muslim's works. As the Sunni 
hadlth tradition became increasingly systematized with the writings of 
al-Khatib al-Baghdadi, and even more so with those of Ibn al-Salah 
(d. 643/1245), the conventions of hadith scholarship emerged as an 
institution with which the canon stood in potential tension. Examining 
the issues of obfuscation in transmission [tadlis) and the criticism of 
transmitters, we shall see that the Sahihcyn sometimes fell short of the 
established standards of hadith scholarship. Preserving the authority 
of the canon thus depended on charitable interpretations of the works 
that exempted them from these rules. 

Divergences between the methods of the Shaykhayn and other hadlth 
critics had manifested themselves concretely in critiques of the Sahihayn, 
such as that of al-Daraqutni. Protecting the canonical culture would 
thus require three of its great proponents, Ibn al-Salah, al-NawawI and 
Ibn Hajar, to employ the Principle of Charity and their mastery of the 
hadlth tradition to resolve these outstanding criticisms of the canon. 

2 Halbertal, 28. 



264 CHAPTER SEVEN 

The Beginnings of Canonical Culture: Between 390-460/1000-1070 

From the evidence available, the canonical culture surrounding the 
Sahihayn seems to have emerged in Baghdad in the period between 
al-Daraqutni's career in the mid- to late fourth/tenth century and 
that of al-Khatib al-Baghdadi in the mid-fifth/eleventh. Considering 
the direct relationship that Halbertal posits between the canonicity 
of texts and the charity with which they are treated, it is no surprise 
that the construction of a canonical culture surrounding the Sahihayn 
began at the same time as the emergence of the canon itself Between 
approximately 390/1000 and 460/1070 the hadith-scholar environ- 
ment in Baghdad transformed from an openess toward criticism of 
the Sahihayn to a canonical culture that demanded the extension of 
charity to al-Bukhari and Muslim. Although Ibn 'Ammar al-Shahid, 
al-Isma'ili and al-Daraqutni had all exhibited profound interest in al- 
Bukharl's and Muslim's collections, they had no compunction about 
criticizing the Sahihcyn if they felt their authors had erred. Neither did 
these fourth /tenth- century scholars feel obliged to qualify or apologize 
for such critiques. Their evaluations merely represented an aspect of 
scholarly interest in the Sahihayn, two works that did not differ ontologi- 
cally from any other hadith book. Only after their canonization had 
endowed al-Bukhari's and Muslim's collections with an authoritative 
role and significance for communal identification did criticizing the 
works or their authors pose any threat. 

The construction of the SaMhoyn canonical culture first becomes 
evident in the work of al-Daraqutnl's student Abu Mas'ud al-Dimashqi 
(d. 401/1010-1 1), a member of the Baghdad knot who penned a work 
defending Sihih Muslim against some of al-Daraqutni's criticisms. His 
Kitdb al-ajwiba (Book of Responses) might have been nothing more 
than an exercise in objective scholarship: al-Daraqutnl had made cer- 
tain criticisms that Abti Mas'ud believed were incorrect. In the work, 
however, it becomes immediately clear that Abu Mas'ud's agenda 
bears far more significance: he aims primarily at exonerating Muslim's 
scholarly legacy from any sort of blame. Even when he admits that 
al-Daraqutni's critiques are correct, for example, he tries to deflect the 
blame from Muslim to transmitters in the isnad. "And as for attribut- 
ing the oversight to Muslim among the others, no . . .," he states in one 
case.-' In two instances of inappropriate Addition, Abu Mas'ud admits 

' Abu Mas'ud al-Dimashql, Kitdb al-ajwiba; 152, 321. 



THE PRINCIPLE OF CHARITY 265 

that al-Daraqutni was correct in objecting to Muslim's inclusion of the 
narration. He defends Muslim, however, by saying that he did not have 
the correct version at his disposal. If he did, he would have taken it 
instead.* In three instances he argues charitably that Muslim included 
the problematic version only to demonstrate its flaw^ 

Abu Mas'ud's defensiveness about Muslim's work stands in stark 
contrast to al-Daraqutnl's impartial study'' At one point in the Kitdb 
al-ajwiba, al-Daraqutnl criticizes a narration noted by Muslim but 
acknowledges that the scholar ultimately decided to leave it out of his 
Sahth. For al-Daraqutni, whose scholarly interest lay in identifying flawed 
narrations regardless of where he found them, this was still worthy of 
note. Abu Mas'ud, however, objects angrily, "So if he left it out, what 
is the meaning of attributing error to him [Muslim] in this!?"' 

Within a few decades of al-Daraqutnl's death the charity called 
for by Abu Mas'ud had become expected. In Baghdad, the canonical 
culture surrounding al-Bukharl in particular seems to have gelled by 
approximately 450/1060. The writings of al-Khatib al-Baghdadi indi- 
cate a prevailing expectation of charity in discussing al-Bukhari's works 
among hadlth scholars. Al-Khatib composed a book dealing with the 
overall problem of mistaken identities in biographical dictionaries of 
hadlth transmitters, titling it Kitdb mudih awhdm al-jam ' wa al-tafnq (The 
Book of Clarifying Errors of Conflation or Distinction). Although this 
work criticizes a whole slew of hadlth scholars, al-Khatib opens the 
book with a mistake made by al-Bukhari in his sd-Tdnkh al-kabir. He 
follows this with a fascinating statement: 



' Abu Mas'ud al-Dimashqi, Kitdb al-ajwiba; 168, 212. 

^ Abu Mas'ud al-Dimashql, Kitab al-ajwiba; 159, 180, 188. 

'' Yet we know that Abu Mas'ud also criticized some narrations in Sahih Muslim in his 
Atrdf al-Sahihayn. These criticisms, however, seem to have been restricted to Muslim's 
auxiliary narrations (niutdbi'dt/shawdhid) or to have been citations of earlier criticisms 
such as those of al-Daraqutnl. On one such occasion, Abu Mas'ud vaguely notes a 
"disagreement" on one of five auxiliary narrations Muslim provides for his two principal 
narrations of a hadlth in which the Prophet tells his followers not to kill an enemy if 
they have professed faith in Islam. In another case Abu Mas'ud follows al-Daraqutnl 
in criticizing one of Muslim's narrations for omitting a transmitter These criticisms 
are preserved in the surviving elements of al-Dimashqi's Atrdf and also in Abu 'All al- 
Ghassanl al-Jayyanl's al-Tanbih 'aid al-awhdm al-wdqi'aji Sahth al-imdm Muslim. See, al- 
Jayyanl, al-Tanbih 'aid al-awhdm al-wdqi'aji Sahih al-imdm Muslim, ed. Muhammad AbO 
al-Fadl ([Rabat]: Wizarat al-Awqaf, 1421/2000), 69 {SaMh Muslim: kitdb al-imdn, bdb 
tahnm qatl al-kdfir ba'da an qdla Id ildh ilia AlldK), 76. See also, Abu Mas'ud al-Dimashql, 
Atrdf al-Sahihayn; 3b, 26b. 

' Abu Mas'ud al-Dimashqi, Kitdb al-ajwiba, 264. 



266 CHAPTER SEVEN 

It may be that some people who read these hnes will assume the worst 
of us, believing that we intend to impugn our predecessors, exposing the 
faults of our venerable shaykhs and the scholars of yesteryear Far from 
it, for by the beams of their light do we see, and by following in their 
clear footsteps do we distinguish [truth from falsehood] . Indeed, it is by 
their well-worn path that we circumvent error Our relationship to them 
is nothing more that what Abu 'Amr b. 'Ala' (d. 154/771 or 157/774) 
said (he gives an isnad): 'Compared to those who have come before us, 
we are nothing but a tiny root on the base of a great date palm.' Indeed, 
when God creates luminaries among men and raises up a leader for each 
community, he requires those whom they guide to adhere to the truth 
that they illuminate. [Yet] God obliges those who stand by the truth and 
follow in their footsteps and are blessed with understanding to illuminate 
what [earlier scholars] neglected and to correct their oversights. This, 
because [these earlier scholars] were not immune to mistakes and were 
not totally protected from the ugly face of error This is the right of the 
learned scholar over the student, and the obligation of those who follow to 
those who precede. We hope that this apology will be clear to whomever 
comes upon our book, the History of the City of Peace [Tarikh Baghdad) . . ., 
for in it we have presented, from among the virtues of al-Bukharl, mate- 
rial sufficient to clear away any suspicion of our opinion of him as well 
as any accusations concerning our correcting his errors. . . .'"' 

Al-Khatib continues with a quote from al-Muzani, saying, "If a book 
v\?ere looked over seventy times there would still be a mistake in it, for 
God has not permitted that any book be sahih except His Book (i.e., the 
Qur'an)." He quotes Ibn Hanbal's son 'Abdallah as saying, "I read a 
book to my father [for checking] thirteen times, and on the fourteenth 
time he came up with a mistake, so he put the book down and said, 
'Indeed I have denied that any book could be perfectly correct [jasihhd) 
except the Book of God most high.'"^ 

Al-Khatlb's tortured apology for even minor criticisms of al-Bukhari's 
identification of hadlth transmitters reflects an intense anxiety over 
reactions to his work and the powerful canonical culture that evidently 
surrounded the scholarly persona of al-Bukharl by that time. Al-Khatlb's 
homily invoking the sacred duty of scholarly vigilance, phrased in the 
idiom of the hadith student's pietistic reverence for his teachers, rep- 
resents an effort to counterbalance the charity the author feels he is 
expected to show al-Bukhari. By referring his readers to the formidable 
accolades he grants al-Bukhari in his Tarikh Baghdad (whose biography 



'' Al-Khatib, Kitab mudih aivham al-jam' wa al-tafnq, f:5-6. 
' Al-Khatib, Kitdh mudih aivham al-jam' wa al-tafnq, f:6. 



THE PRINCIPLE OF CHARITY 267 

is perhaps the longest of any figure in the work), al-Khatib seeks to 
placate potential critics by calling their attention to his contribution 
and obedience to the canonical culture. Read against the grain, al- 
Khatib's agonized preemptive defense suggests a scholarly atmosphere 
totally different from the one in which al-Daraqutnl, a fellow Shafi'l 
of Baghdad, had freely criticized al-Bukharl less than a century earlier. 
When students asked him about several dozen transmitters from the 
Sahihayn that al-Nasa'l (d. 303/915) had criticized, al-Daraqutni bluntly 
seconded most of al-Nasa'i's evaluations.'" Although al-Daraqutni's Mtdb 
al-tatabbu ' contains serious and substantive criticisms of the Sahihayn, its 
author felt no need to justify or apologize for his critique. 

We cannot be sure of exactly whom al-Khatib was so wary in his 
minor criticisms of al-Bukhari. We know that he faced consistent 
intimidation from the Hanballs, from whose ranks he had defected 
and who publicly questioned his transmission-based Sunni allegiance." 
Considering the ferocity with which the Shafi'i Abu Mas'ud al-Dimashqi 
had defended Muslim's Sahth, however, we can easily imagine that al- 
Khatib's fellow Shafi'i hadith scholars in Baghdad may have aroused 
his concern just as much as the Hanballs. We do not know when al- 
Khatib wrote the Kitab mudih al-awham, so we cannot know precisely 
what forces were affecting him at that point in his career Based on 
the absence of any apologies in al-Daraqutni's critique of the Sahihayn, 
the vehemence of Abu Mas'ud's eventual rebuttal of his teacher and 
finally al-Khatib's writing, we can conclude that in Baghdad a canoni- 
cal culture arose around the Stth^hayn between 390/400 and al-Khatib's 
death in 463/1071. 



The Character of the Canonical Culture: 
Al-Khatib al-Baghdadi and Defining the Personas of al-Bukhdri and Muslim 

The canonization of al-Bukhari and Muslim is a question of how 
the Muslim community has viewed these two scholars' legacies. Their 
historiographical personas thus form as much a part of the text of the 
canon as their actual books. The extent to which Islamic civilization 



'" See al-Daraqutni, "Dhikr aqwam akhraja lahum al-Bukhari wa Muslim ft 
kitabayhima wa da"afahum al-Nasa'l," MS Ahmet III 624, Topkapi Sarayi, Istanbul: 
fols. 253a-254b. ' 

" Al-Dhahabi, Tadhkirat al-huJfSs, 3:225. 



268 CHAPTER SEVEN 

has identified tlie Sahihayn witli tlieir respective authors is illustrated by 
their agency in the formulaic statement "al-Bukharl/Muslim included 
it..." or equating the works with their compilers in common phrases 
such as "the hadith is in Muslim." Indeed, the skill, piety and critical 
rigor of the two scholars served as the basis on which their authority 
was founded. Questioning al-Bukharl's or Muslim's judgment or devo- 
tion to the Prophet's legacy thus constituted a threat to the Sahihayn 
canon itself Although al-Khatib al-Baghdadl's apology did not even 
involve al-Bukharl's SaMh per se, the idea of criticizing that expert's 
judgment in his al-Tdnkh al-kabir proved sufficiently alarming to prompt 
an apology. 

Al-Khatlb's biographies of al-Bukhari and Muslim provide our 
earliest extant expressions of the canonical culture surrounding the 
Shaykhayn. As al-Khatlb himself informs us, he intended his biography 
of al-Bukhari in the Tdrikh Baghdad to describe the scholar with the 
proper reverence. Although al-Khallll's brief biographies of al-Bukhari 
and Muslim as well as fragments of al-Hakim's entries have survived, 
the Tdnkh Baghddd offers us the earliest complete and, indeed, self- 
conscious expression of the Sahihayn canonical culture. The majority of 
biographies in the Tdnkh Baghddd consist only of reports from earlier 
sources that al-Khatib presents through their isndds. As a result, his role 
in crafting al-Bukharl's and Muslim's biographies is that of an editor 
who constructs an image of the two scholars by choosing selectively 
from the vast pool of historiographical raw material about them. 

Like all later Sunni biographers, al-Khatlb freely ladled out hyper- 
bolic descriptions of al-Bukhari's and Muslim's virtues, as well as those 
of other great scholars such as Ibn Hanbal. There was never a dearth 
of praise for the guardians of the faith. Al-Khatlb therefore leaves 
the reader with no doubt as to al-Bukharl's or Muslim's prodigious 
memories, piety or mastery of hadith. What concerns us here is not 
the mere quantity of positive evaluations, however, but rather the pic- 
ture that such praise paints, the contours of the personas it shapes or 
the unspoken problems it intends to address. A canonical culture must 
reconcile the history that was with the history that should have been. 
The culture that al-Khatlb elaborates thus directly addresses the most 
prominent issues in the saga of the Sahthcyn: the proper relationship 
between the Shaykhayn and the greatest generation of their teachers, 
appropriately acknowledging the accomplishment represented by the 
Sahihayn, al-Bukharl's scandal of the lafz of the Qur'an, and the proper 
ranking of al-Bukhari and Muslim. 



THE PRINCIPLE OF CHARITY 269 

We have seen the problem that al-Bukharl's and Muslim's superla- 
tive scholarship presented for the atavistic logic of the hadith-scholar 
community in the tale of al-Bukhari plagiarizing his Sahih from his 
teacher. Scholars such as Ibn Abi Hatim and al-RamahurmuzI did 
not perceive the Sahihayn or their authors as superseding the greatest 
generation of Ibn Hanbal, Ibn Ma'in and 'All b. al-Madinl. It was 
not until the writings of Ibn Manda (d. 395/1004-5) that al-BukharT, 
Muslim and the sahih movement as a whole began to be seen as the 
pinnacle of the hadlth tradition. The Sahihayn canonical culture would 
have to correct this imbalance. 

Al-Khatlb's treatment of al-Bukhari and Muslim thus leaves little 
doubt about their superiority over their teachers. He cites one Ahmad 
b. Abl Bakr al-Madlnl as asserting that al-Bukharl possessed better legal 
acumen [afqali) and was more perceptive (absar) than Ibn Hanbal. When 
someone objects to this provocative statement (as al-Khatib's reader 
might), al-Madani replies, "If you looked at al-Bukhari and Malik 
you would see they were the same in juristic knowledge and hadith.'"^ 
Ahmad b. Nasr al-Khaffaf is quoted as saying that al-Bukhari is more 
knowledgeable than Ishaq b. Rahawayh and Ibn Hanbal by twenty 
degrees.'^ Ahmad b. 'Abdallah b. al-Bukharl, the great scholar's grand- 
son, heard his grandfather say that he did not humble himself (istasghara) 
in the presence of anyone except 'All b. al-Madini, but admitted that 
"perhaps I still mentioned hadiths he did not know (ugharribu 'alqjihi)."^* 
Al-Khatib relies on a narration through al-Hakim al-Naysaburl from 
Muslim's colleague Ahmad b. Salama, who saw "Abu Zur'a and Abu 
Hatim al-Razi place Muslim before the shaykhs of their time in the 
knowledge of authentic hadiths.'"'^ 



'2 Al-Khatib, Tdnkh Baghdad, 2:19; Ibn 'Asakir, Tdnkh madinat Dimashq, 52:86; al- 
Dhahabl, Tdnkh al-isldm, 19:256; Ibn Hajar, Hady al-sdn, 667. 

" Al-Khatib, Tdnkh Baghddd, 2:27; Ibn 'Asakir, Tdnkh madinat Dimashq, 52:78; al- 
Subkl, Tabaqdt, 2:221, 225 (this includes an additional description of al-Bukharl as 
'^al-taqi al-naqi al-'dlim alladhi lam am mithlahu^'); cf. Ibn Kathir, al-Biddya iva al-nihdya, ed. 
Fu'ad Sayyid et al. (Beirut: Dar al-Kutub al-'Ilmiyya, 1405/1985), 11:29; Ibn Hajar, 
Hady al-sdn, 671. 

" Al-Khatib, Tdnkh Baghddd, 2: 1 7; Ibn Adi, Asdmi, 125 (without the comment about 
knowing more hadiths); Ibn Abl Ya'la, Tabaqdt al-handbila, 1:311; Ibn 'Asakir, Tdnkh 
madinat Dimashq, 52:81-2; al-Hasan b. Muhammad al-Saghani (d. 650/1252), Asdmi 
shuyukh Abl Abdalldh Muhammad b. Ismd'il b. Ibrdhim b. al-Mughira al-Bukhari, ed. 'All b. 
Muhammad al-'Imran ([Mecca]: Dar 'Alam al-Fawa'id, 1419/[1998]), 2; al-Dhahabi, 
Tdnkh al-isldm, 19:252; al-Safadi, al-Wdfi bi'l-wafaydt, vol. 2, ed. S. Dedering (Istanbul: 
Matba'at Wizarat al-Ma'arif, 1949), 208; Ibn Hajar, Hady al-sdn, 669. 

'■■ Al-Khatib, Tdnkh Baghddd, 13:102; al-Ghassam, al-Tanbih, 27; al-Qadi 'lyad. 



270 CHAPTER SEVEN 

In the case of al-Bukhari, his disgrace at the hands of the iiber- 
Sunnis in the lafz scandal had tarnished his name in the eyes of promi- 
nent architects of the hadlth tradition, such as Ibn Abi Hatim al-RazI. 
The narrative constructed by al-Khatib, however, is one of vindica- 
tion in which al-Bukharl righteously stood by what would become the 
orthodox position on the Qur'an."' As the Shafi'i/Ash'ari al-Subkl later 
explains, "Every reasonable person knows that our wordings are from 
among our deeds, and that our deeds are created, and that thus our 
wordings are created."" Al-Bukhari's contemporary Muhammad b. 
Khushnam is invoked as a witness that al-Bukhari denied the accusa- 
tion that he believed the Qur'an itself was created, insisting instead 
that the acts of men are created. He states that he will not change 
his position until proven wrong.'" For al-Bukhari, certain of the truth 
of his position, "the complimenter and the detractor are the same."'^ 
Al-Khatib relies on al-Hakim for the comeuppance of the amir of 
Bukhara, who had used al-Bukhari's stance on the lafz of the Qur'an 
to expel him from the city: he was imprisoned less than a month later 
by the Abbasid caliph in Baghdad. As for Hurayth b. Abi al-Waraqa', 
the Hanafi scholar whose assistance the amir had enlisted in con- 
demning al-Bukhari, members of his family were afflicted by suffer- 
ing too terrible to describe.^" To further assure al-Bukharl's orthodox 
standing, al-Khatib narrates a report through al-Hakim that invokes 



Ikmdl al~mu'iim, 1:79; Ibn 'Asakir, Tdrikh madinat Dimashq, 58:89—90; Ibn al-Salah, 
Siydnat Sahih Muslim, 61; al-Dhahabi, Tdrikh al~isldm, 20:184; Ibn Kathir, al-Biddya iva 
al-nihdya, 1 1:37. 

"" For the Ash'arl exposition of this stance, see al-BayhaqI, Kitdb al-asmd' wa al-sijdt, 
2:17 IF.; al-Juwaynl, Textes apologetiques de Guwaini, ed. and trans. Michel Allard (Beirut: 
Dar al-Machreq, 1968), 146. By the mid-fifth/eleventh century even moderate HanbalTs, 
such as Ibn al-Farra', acknowledged that the wording of the Qiar'an was created; Ibn 
al-Farra', al-Masd'il al-'aqdiyya, 11 IF. Ibn Abi Ya'la's biography of al-Bukharl includes 
a report that does not uphold this image, but rather has al-Bukharl telling Ibn Hanbal 
that anyone who says that the lafzi of the Qiar'an is created is a ''Jahmi kdfir." This is 
almost certainly an early Hanball attempt to exonerate al-Bukharl, since his Khalq af'dl 
al- 'ibdd leaves no doubt that he did in fact believe that the wording of the Qur'an was 
created; Ibn Abi Yala, Tabaqdt al-handbila, 1:259. 

" Al-Subkl, Tabaqdt, 2:230. 

'* Al-Khapb, Tdrikh Baghddd, 2:29; Ibn Asakir, Tdrikh madinat Dimashq, 52:94. 

'5 Al-Khatib, Tdrikh Baghddd, 2:29. 

2" Al-Khatib, Tdrikh Baghddd, 2:32; Ibn Asakir, Tdrikh madinat Dimashq, 52:97; Ibn 
Khallikan, Wajdydt al-a'ydn, 4:190 (Ibn Khallikan provides the most copious informa- 
tion about the amir's fate in Baghdad); al-Dhahabi, Tdrikh al-isldm, 19:271-2; al-Subkl, 
Tabaqdt, 2:233; cf Ibn Kathir, al-Biddya wa al-nihdya, 11:30; Ibn Hajar, Hady al-sdri, 
680; cf MuUa All Qarl, Mirqdt al-majdtih sharh Mishkdt al-masdhih (Cairo: al-Matba'a 
al-Maymuniyya, 1891), 1:14. 



THE PRINCIPLE OF CHARITY 271 

the authority of a vehement opponent of the created Qjur'an, Ibn 
Khuzayma, saying that "there is no one under the heavens more 
knowledgeable in hadith than al-Bukhari."^' 

Furthermore, al-Khatib portrays al-Bukhari's accuser, the great 
muhaddith Muhammad b. Yahya al-Dhuhll, as both inferior to al-Bukhari 
in the science of hadith and motivated by petty jealousy Al-Khatib cites 
al-Husayn al-'Ijll as describing Abu Zur'a and Abu Hatim al-Razi listen- 
ing to al-Bukhari attentively, adding that he was "more knowledgeable 
than al-Dhuhll in this and that."^^ Another contemporary of al-Bukharl 
reports that he saw him and al-Dhuhli walking together in a funeral 
procession. Al-Dhuhll was asking al-Bukhari questions, to which he 
replied with such ease it was as if he were reading one of the shortest 
sura'i of the Qur'an (no. 112, surat al-Ikhlds)P Al-Khatib then includes 
two separate reports that al-DhuhlT began attacking al-Bukhari for his 
stance on the wording of the Qur'an only after al-Dhuhli's students 
began deserting him and flocking to al-Bukharl's study circle.^* 

The canonical culture as depicted by al-Khatib also emphasizes what 
a momentous feat the compilation of the Sahihayn represented as well 
as their authors' critical stringency. He provides several reports telling 
us that al-Bukhari selected his Sahih from over 600,000 hadlths and 
spent ten years compiling it, intending it as "a testament [hujjd) between 
[himself] and God."^' A report from al-Firabri tells us that al-Bukharl 



^' Al-Khatib, Tdnkh Baghdad, 2:26; al-Hakim, Ma'rifat 'ulum al-hadith, 93; Ibn 'Asakir, 
Tdnkh madinat Dimashq, 52:65; al-Saghanl, Asdmi, 2; al-Nawawl, Tahdhib al-asmd', 1:70; 
al-Dhahabi, Tdnkh al-isldm, 19:256; al-Subkl, Tabaqdt, 2:218; Ibn Kathir, al-Biddya wa 
al-nihdya, 11:29; Ibn Rajab, Sharh Hal al-Tirmidhi, 1:225; Ibn Hajar, Hadji al-sdn, 671; 
MuUa 'All Qarl, Mirqdt al-majatih, 1:14. 

^^ Al-Khatib, Tdnkh Baghddd, 2:29; Ibn 'Asakir, Tdnkh madinat Dimashq, 52:85; Ibn 
Kathir, al-Biddya wa al-nihdya, 1 1:29; Ibn Hajar, Hady al-sdn, 670. 

^' Al-Khapb, Tdnkh Baghddd, 2:30; Ibn 'Asakir, Tdnkh madinat Dimashq, 52:95; al- 
Nawawl, Tahdhib al-asmd', 1:68; al-Subkl, Tabaqdt, 2:229; Ibn Kathir, al-Biddya wa 
al-nihdya, 11:29; Ibn Rajab, Sharh 'Hal al-Tirmidhi, 1:225; Ibn Hajar, Hady al-sdn, 674; 
Ibn al-'Imad, Shadhardt al-dhahab, 2:134-5. 

^' Al-Khatib, Tdnkh Baghddd, 2:29, 30; Ibn 'Asakir, Tdnkh madinat Dimashq, 52:91; 
al-Subkl, Tabaqdt, 2:228. 

^^ Al-Khatib, Tdnkh Baghddd, 2:9, 14; Ibn AblYa'la, Tabaqdt al-handbila, 1:256, 7; Ibn 
'Asakir, Tdnkh madinat Dimashq, 52:72; cf. Ibn al-jawzl, al-Muntaiam, 12:1 15; al-SaghanI, 
Asdmi, 2; Ibn Khallikan, Wajdydt al-a'ydn, 4:190; al-KirmanI, al-Kawdkib al-dardri, 1:11; 
al-Safadi, al-Wdfi bi'l-wafaydt, 2:208; al-Dhahabi, Tdnkh al-isldm, 19:249; al-Subkl, 
Tabaqdt, 2:221; Muhammad b. Abl Bakr Ibn Nasir al-Din al-Dimashql, Majmd' Jihi 
rasd'il li'l-hqfig. Ibn Ndsir al-Din al-Dimashqi, ed. Abu 'Abdallah Mish'al al-MutayrI (Beirut: 
Dar Ibn Hazm, 1422/2001), 344; Ibn Hajar, Hady al-sdri, 675; Ibn al-'Imad, Shadhardt 
al-dhahab, 2:134; MuUa 'All Qarl, Mirqdt al-majdtih, 1:13. 



272 CHAPTER SEVEN 

included only the most authentic hadiths, and that he performed ablu- 
tions and prayed two rak'a'i, before inserting any hadith in the book.^'' 
Again relying on a report from al-Hakim, al-Khatib includes a report 
that Muslim compiled his Sahih from a selection of 300,000 hadiths.^' 
We then find the famous statement of Abu 'All al-Naysaburl that "there 
is no book under the heavens more authentic than Sahih Muslim in the 
science of hadith."^" 

The canonical culture also reflects the nature of the Sahihayn canon 
itself Muslim is thus clearly ranked below al-Bukharl. Al-Khatib includes 
a report narrated through al-Hakim in which a scholar says that he 
once saw Muslim asking al-Bukhari questions like a youth before his 
teacher.^^ In one instance, Muslim was so impressed with al-Bukharl's 
knowledge of hadith that he almost cried.'" On the same occasion, 
Muslim professes to al-Bukharl, "I testify that only the jealous could 
hate you, and that there is none like you."^' In a report narrated 
through al-Hakim, Muslim comes to al-Bukhari seeking his expertise, 
then kisses his forehead and calls him doctor {tabib) of hadith and its 
ills /flaws (literally, 'Hal)}''- 



^*' Al-Khapb, Tdnkh Baghdad, 2:9. Ibn 'Adi includes a report that describes al-Bukharl 
praying two rak'ai before writing the chapter titles (tardjim) of his book; Ibn 'Adi, Asdmi, 
6 1 ; Ibn Abl Ya'la, Tabaqdt al-handbila, 1:256; Ibn 'Asakir, Tdrikh madinat Dimashq, 52:72; 
Ibn al-jawzl, al-Muntazam, 12:115; cf. al-SaghanI, Jjami", 2; Ibn Khallikan, Wafiydt al- 
a'ydn, 4:190; al-KirmanI, al-Kawdkib al-dardn, 1:11; al-Dhahabi, Tdrikh al-isldm, 19:248 
(al-Dhahabi notes that this meant before sitting down to work on his book); al-Safadi, 
al-Wdfi bi'l-wafaydt, 2:208; al-Subkl, Tabaqdt, 2:220; Ibn Nasir al-Din, Majmufihi rasd'il, 
344; Ibn Hajar, Hadj al-sdn, 675; Ibn al-'Imad, Shadhardt al-dhahab, 2:136; MuUa 'All 
Qzn, Mirqdt al-majdtih, 1:13. 

2' Al-Khatib, Tdnkh Baghddd, 13:102; al-Ghassani, al-Tanbih, 28; Ibn Abl Ya'la, Tabaqdt 
al-handbila, 1:311; Ibn al-Salah, Siydna SaKh Muslim, 67; Ibn Khallikan, Wafdydt al-a'ydn, 
5:194; al-Dhahabi, Tdnkh al-isldm, 20:185; Ibn Kathir, al-Biddya wa al-nihdya, 1 1:37; Ibn 
al-'Imad, Shadhardt al-dhahab, 2:144. 

^^ Al-Khatib, Tdnkh Baghddd, 13:102; al-Ghassani, al-Tanbih, 29; al-Qadi 'lyad, 
Ikmdl al-mu'lim, 1:80; Ibn al-Salah, Siydnat Sahih Muslim, 68-9; Ibn Khallikan, Wajdydt 
al-a'ydn, 5:194; al-Dhahabi, Tdrikh al-isldm, 20:186; Ibn Kathir, al-Biddya wa al-nihdya, 
11:37; Ibn Nasir al-Din al-Dimashql, Majmu jihi rasd'il, 330; Ibn al-'Imad, Shadhardt 
al-dhahab, 2:144. 

^' Al-Khatib, Tdrikh Baghddd, 2:29; Ibn 'Asakir, Tdrikh madinat Dimashq, 52:89; al- 
Nawawl, Tahdhib al-asmd', 1:70. 

'" Al-Khatib, Tdrikh Baghddd, 2:28; Ibn 'Asakir, Tdrikh madinat Dimashq, 52:69-70; 
Ibn Hajar, Hady al-sdri, 675. 

" Al-Khatib, Tdrikh Baghddd, 2:28; al-Khallll, al-Irshdd, 380; Ibn 'Asakir, Tdrikh madinat 
Dimashq, 52:70; al-NawawI, Tahdhib al-asmd', 1:70; Ibn Rajab, Sharh 'Hal al-Tirmidhi, 
1:225; Ibn Hajar, Hady al-sdri, 675; Ibn al-'Imad, Shadhardt al-dhahab, 2:134. 

'^ Al-Khatib, Tdrikh Baghddd, 13:103; al-Hakim, Ma'rifat 'ulum al-hadith, 141; Ibn 
Abl Ya'la, Tabaqdt al-handbila, 1:255; Ibn 'Asakir, Tdrikh madinat Dimashq, 52:68, 58:91; 



THE PRINCIPLE OF CHARITY 273 

As part of the accolades he includes for Muslim, al-Khatib provides 
the report of Ibn 'Uqda saying that Muslim made fewer errors than 
al-Bukharl because he included fewer hadiths with incomplete isnadsP 
In a rare instance of personal commentary, however, al-Khatib restores 
the proper relationship between the two books by adding that "Muslim 
followed in Bukhari's footsteps and gained from his knowledge (nazara 
fi 'ilmihi) . . . and when al-BukharT came to Naysabur near the end of 
his life, Muslim followed him around constantly"^* To further counter 
expert opinions ranking Muslim above al-Bukharl, al-Khatib quotes the 
great al-Daraqutni as stating, "If not for al-Bukhari, Muslim would not 
have come or gone."^' The authors of other prominent sahth collec- 
tions are also featured complimenting al-Bukhari in particular In one 
report, al-Nasa'l says that al-Bukharl's Sahth is the best book available.'*" 
Al-Tirmidhi is quoted as calling al-Bukharl "the ornament {zayn) of 
the umma."'' 

In al-Khatlb's treatment of al-Bukhari and Muslim, we also notice 
that the two scholars, like their works, present a unified and comple- 
mentary pair. Al-Khatib makes another personal addendum to a report 
of Muslim venerating al-Bukharl, explaining that "Muslim used to 
defend [nadala 'an) al-Bukharl to the point that what happened between 
[Muslim] and Muhammad b. Yahya al-Dhuhli got worse {hattd awhasha) 
because of him."'"' Al-Khatib includes Ibn al-Akhram's famous comment 



al-SaghanI, Asdmi, 2 (here the author conflates the above three reports about MusUm); 
al-NawawI, Tahdhib al-asma, 1:70; al-KirmanI, al-Kawdkib al-dardn, 1:11; al-Dhahabi, 
Tdnkh al-isldm, 19:257; al-Subkl, Tabaqdt, 2:223; Ibn Kathir, al-Biddya iva al-nihdya, 1 1:29; 
Ibn Hajar, Hady al-sdn, 675; MuUa 'All Qarl, Mirqdt al-mqfdtih, 1:13. 

'^ Al-Khatib, Tdnkh Baghdad, 13:103; Ibn 'Asakir, Tdrikh madmat Dimashq, 58:90; al- 
Dhahabl, Tdnkh al-isldm, 20:185; al-Safadi, al-Wdfi bi'l-wafaydt, vol. 25, ed. Muhammad 
al-Hujayn (Beirut, 1420/1999), 25:552; Ibn Kathir, al-Biddya wa al-nihdya, 11:37. 

" Al-Khatib, Tdnkh Baghddd, 13:103; Ibn al-'Imad, Shadhardt al-dhahab, 2:144. 

'^ Al-Khatib, Tdnkh Baghddd, 13:103; al-Ghassani, al-Tanbih, 29; Ibn 'Asakir, Tdnkh 
madinat Dimashq, 58:90; Ibn al-Jawzi, al-Munta^am, 12:1 17; al-Safadl, al-Wdfi bi'l-wafaydt, 
25:552; al-Dhahabi, Tdnkh al-isldm, 20:187; Ibn Kathir, al-Biddya wa al-nihdya, 11:37; 
Ibn Hajar, Hady al-sdn, 676; MuUa 'All Qarl, Mirqdt al-mqfdtih, 1:16. 

"■ Al-Khatib, Tdnkh Baghddd, 2:9; Ibn 'Asakir, Tdnkh madinat Dimashq, 52:74; al- 
Nawawl, Tahdhib al-asmd', 1:74; Ibn Nasir al-Din, MajmU'fihi rasd'il, 329; Ibn al-'Imad, 
Shadhardt al-dhahab, 2:135. 

'' Al-Khatib, Tdrikh Baghddd, 2:26; Ibn 'Asakir, Tdrikh madinat Dimashq, 52:79; cf. al- 
Kirmanl, al-Kawdkib al-dardri, 1:11; al-Subkl, Tabaqdt, 2:221; Ibn Kathir, al-Biddya wa 
al-nihdya, 11:29; Ibn Hajar, Hady al-sdri, 671. 

'" Al-Khatib, Tdrikh Baghddd, 13:103. It is not obvious from the text of al-Khatlb's 
work that he himself made this addition, but al-Ghassanl, who had both Tdrikh Baghddd 
and al-Hakim's work, from which the report is cited, at his disposal, notes that al-Khatib 
made this addition; Ghassanl, al-Tanbih, 30; Ibn Khallikan, Wajdydt al-a'ydn, 5:194; 



274 CHAPTER SEVEN 

that, together, al-Bukhari and Muslim missed very few authentic hadiths 
[qallama yafitu al-Bukhdn wa Muslim mayathbutu min al-hadtth)/''' 

The personas of al-Bukharl and Muslim in the Tdnkh Baghdad formed 
the basis for all later biographies of the two scholars. Particularly in 
the case of al-Bukhari, al-Khatlb's work actually provided one of the 
two largest sources for later historians. Material from the Tdnkh Baghddd 
makes up approximately 47% (52/110 reports) of al-Dhahabl's com- 
prehensive biography of al-Bukharl in the Tdnkh al-isldm, and 41% 
(11/27) of his entry on Muslim. 

The second major source on which later biographers such as al- 
Dhahabl and al-Subki drew was al-Hakim's lost Tdnkh Naysdbur. Al- 
Hakim served as the premier source for information about Muslim in 
particular, since he had been a veritable Naysabur institution. Even 
al-Khatib, who relies on al-Hakim for only half a dozen reports in the 
Tdnkh Baghddd's, massive biography of al-Bukhari, refers to al-Hakim for 
50% (7/14) of the reports he includes in his much shorter biography 
of Muslim. 

The Tdnkh Bukhdrd (now lost) of Abu 'Abdallah Muhammad b. 
Ahmad Ghunjar al-Bukhari (d. 412/1021) was one of the earliest 
sources on al-Bukhari, but al-Khatib seems to have incorporated much 
of its material in the Tdnkh Baghddd through a transmission of the book 
from its author.*" The other early source of original material on al- 
Bukharl of which neither al-Khatib nor al-Hakim seem to have made 
any use is the Tdnkh Samarqand (now lost) of Abu Sa'd 'Abd al-Rahman 
b. Muhammad al-Astarabadhi (d. 405/1015). Later scholars like al- 
Dhahabi relied on the Tdnkh Samarqand for reports about al-Bukhari's 
grave, which was in the vicinity of Samarqand. These include stories 



al-Dhahabl, Tdrikh al-isldm, 20:188; Ibn Kathir, al-Biddya wa al-nihdya, 1 1:37; al-Safadi, 
al-WdJi bi'l-wafaydt, 25:553; Ibn al-'Imad, Shadhardt al-dhahab, 2:144. 

'^ Al-Kliapb, Tdnkh Baghddd, 13:102; al-Ghassanl, al-Tanhih, 29; Ibn 'Asakir, Tdrikh 
madinat Dimashq, 58:91. 

'" See, for example, the report in which al-Bukharl's having memorized 200,000 
reports is contrasted with Isliaq b. Rahawayh memorizing only 70,000; al-Khatib, Tdrikh 
Baghddd, 2:24-5; Ibn 'Asakir, Tdrikh madinat Dimashq, 52:63-4; al-Dhahabi, Tdrikh al-isldm, 
19:245; al-Subkl, Tabaqdt, 2:218; Ibn Hajar, Hady al-sdri, 674. Also, see the report about 
al-Bukharl knowing the hadlth of Basra better than Basrans; al-Khatib, Tdrikh Baghddd, 
2:15-6; Ibn Hajar, Hady al-sdri, 672-3. Al-Khatib did not replicate Ghunjar's biography 
of al-Bukharl in its entirety, however, since some reports appear in Ibn 'Asakir's Tdrikh 
madinat Dimashq from Ghunjar that do not appear in Tdrikh Baghddd. See, for example, 
Ibn Asakir, Tdrikh madinat Dimashq, 52:90. Al-Khatib mentions Ghunjar's Tdrikh Bukhdrd 
by name in the Tdrikh Baghddd as well; al-Khatib, Tdrikh Baghddd, 10:29. 



THE PRINCIPLE OF CHARITY 275 

of al-Bukhari's enemies visiting his grave to offer repentance, and the 
many miraculous phenomena that transpired around his tomb (his 
grave, for example, emitting a perfumed scent and eventually attracting 
pilgrims from far and wide).*' 

Although we do not know exactly how al-Hakim portrayed al-Bukhari 
and Muslim, the surviving elements of his Tankh Naysabur emphasize 
the same themes as al-Khatib. In fact, al-Khatib relied on narrations 
through al-Hakim in a number of the above-mentioned reports illus- 
trating the feat involved in producing the Sahihayn, al-Bukhari's and 
Muslim's preeminence in the pantheon of hadlth scholars, and al- 
Bukhari's vindication against his accusers. 



Chanty and the Maintenance of Canonical Culture 

The themes that al-Khatib al-Baghdadi emphasized — the Sahihayn as 
the pinnacle of hadith scholarship, al-Bukhari's vindication, his supe- 
riority to Muslim, and the unified front of the Sahihcyn — would define 
the contours of the Sahihayn canonical culture from the fifth/ eleventh 
century on. By selecting which reports to provide his readers, al-Khatlb's 
recension of al-Bukhari's and Muslim's biographies sought to bring the 
vagaries of history and the problematic origins of the S^hihoyn into 
accord with their authoritative station in the Sunni community. 

Yet several of these reports inherently challenged the canonical 
culture surrounding the two works. Through applying three levels of 
interpretive or editorial processes to them, however, the Sunni schol- 
arly tradition was able to maintain and protect the S^hihoyn canonical 
culture. First, the canonical culture itself exerted a subtle influence 
on the transmission and copying of historical works. Second, scholars 
resorted to interpretive gymnastics in order to reconcile the data of his- 
tory with canonical culture. Finally, scholars actually edited problematic 
reports to fit expectations of how the Muslim community should view 
al-Bukhari and Muslim. 



*' Quoted from al-Sagham, Asdmi, 1-2; al-Dhahabi, Tdnkh al-isldm, 19:282; al- 
Kirmanl, al-Kawdkih al-dardn, 1:12; al-Subkl, Tabaqdt, 2:234; cf. Ibii Katlilr, al-Biddya 
wa al-nihdya, 11:30; al-Qastallam, Irshdd al-sdri, 1:39; cf. MuUa 'All Qarl, Mirqdt al- 
mafdtih, 15. 



276 CHAPTER SEVEN 

a. Reinventing the Etiology: Charity and Legitimizing al-Bukhan's Sahlh 

Compiling hadith collections devoted solely to sahih reports had been a 
revolutionary act, and venerable hadith scholars like AbQ Zur'a al-Razi 
had protested it. This posed a challenge to the authoritative status of the 
Sahihayn, for how could the compilation of the two most authoritative 
collections have met with disapproval from leaders in the hadlth-scholar 
community? By the early sixth/twelfth century, 'AbdaUah b. Muhammad 
al-Batalyawsi of Andalusia (d. 521/1127) had reinterpreted the initial 
reception of the SaMhcyn in a manner that shifted the blame from trans- 
mission-based legal scholars like Abu Zur'a to the more reason-based 
'jurists [fuqahd').' Al-Bukharl and Muslim, he explains, had battled the 
forgery of hadlths until the people of their age persecuted them for 
it. It was this critical stringency in hadith that "stirred up anger in the 
hearts of the jurists [fuqahd') against al-Bukhari."*^ By the time of al- 
Nawawl, however, the urge to cast the origins of the sahih movement 
in a better light had moved beyond reinterpreting history to revising 
historical reports themselves. 

The impetus for the sahih movement as described in al-Khatlb's 
account of al-Bukhari's life is not completely clear. The great scholar's 
decision to begin compiling his Sahih is explained in a report nar- 
rated through al-Hakim from one of al-Bukharl's students, Ibrahim 

b. Ma'qil al-Nasafl. Al-Bukharl recounts that, "We were with Ishaq b. 
Rahawayh, and one of our companions said to us, 'If only you (plural) 
would compile an abridged book on the sunan of the Prophet (s) (kitdb"" 
mukhtasaf li-sunan al-Nabi).' That stuck in my heart, and I undertook 
collecting this book — namely, the Jdmi' (i.e., the Sahih)"*'' Here we 
see that there is, in fact, no mention of that characteristic that would 
distinguish al-Bukharl's collection from previous works: its sole focus 
on authentic reports. 

In al-Nawawl's succinct lexical reference and biographical dictionary 
of the Shafi'l school, the Tahdhib al-asmd' wa al-lughdt, however, we find 
that the report has been transformed. Al-Nawawl also cites Ibrahim b. 
Ma'qil al-Nasafl's quotation from al-Bukharl. In this version, however, 
a scholar says, " 'If only you (plural) would collect an abridged book 



*^ Abu Muhammad 'Abdallah b. al-Sayyid al-Batalyawsi, Kitab al-tanbih 'ala al-asbab 
allati awjabat al-ikhtildf bayn al-muslimin, ed. Ahmad Hasan Kahll and Hamza 'Abdallah 
Nasharti (Cairo: Dar al-I'tisam, 1398/1978), 173. 

« Al-Khatib, Tdrikh Baghdad, 2:8. 



THE PRINCIPLE OF CHARITY 277 

{kitdb"" mukhtasar"") of the authentic sunan of the Messenger of God (s) 
{al-sahih li-sunan al-rasul)' and that became stuck in my heart and I 
undertook collecting that book."** This addition of "authentic" also 
appears in the versions of this report found in major later biographies 
of al-Bukhari, such as Ibn Nasir al-Din al-Dimashqi's (d. 846/1438) 
introduction to his commentary on al-Bukharl, the Iftitdh al-qdn li-Sahih 
al-Bukhdn.*-^ Although he narrates the same report through al-Khatib, 
in his Hady al-sdn Ibn Hajar makes Ishaq b. Rahawayh himself the one 
who suggests collecting the authentic reports of the Prophet.*^ 

In al-Nawawl's recension of the quote, we are thus led to believe 
that al-Bukharl's decision to compile a collection of authentic hadiths 
was no longer a radical departure from tradition. Rather it was recast 
as a response to a need expressed by fellow scholars in the company of 
a senior hadith master. In Ibn Hajar's recension, the suggestion comes 
from Ibn Rahawayh himself, a member of the greatest generation of 
al-Bukharl's and Muslim's teachers. 

Al-Nawawl also includes another etiology for al-Bukharl's Sahth. He 
provides a report with no isndd in which al-Bukhari states, "I saw the 
Prophet in a dream, and it was as if I were standing before him with 
a fan in my hand swatting the flies away from him {adhubbu 'anhu), so I 
asked a dream interpreter and he told me, 'You are swatting lies away 
from him [tadhubbu 'anhu al-kadhib),' and this is what led me to produce 
the Sahih-^^''^ In his comprehensive biographical survey of Islam's first 
millennium, Shadhardt al-dhahab, this is the only etiology for the Sahth 
that Ibn al-'Imad (d. 1089/1679) presents.*** The great Meccan hadith 
scholar, MuUa 'All Qari (d. 1014/1606), also notes that this dream 
propelled al-Bukhari to compile his collection.*^ The twentieth-century 



** Al-Nawawl, Tahdhib al-asmd' wa al-lughdt, 1:74. This version of the report seems to 
have circulated before al-NawawI, however, alongside the other version. Abu al-Walld 
al-Bajl mentions a permutation of this version in the mid-fifth/eleventh century, citing 
it through al-Hakim al-Naysabilrl. Al-Nawawl, however, seems to have been the first 
to have made this version of the quote the official one; al-Bajl, Abu al-Walid Sulaymdn 
b. Khalaf al-Bdji wa kitdbuku, 1:309. 

*^ Ibn Nasir al-Din al-Dimashql, Majmu'Jihi rasd'it, 346. Like al-Khatib, Ibn Nasir 
al-Dln al-Dimashql cites al-Hakim (although here it is specifically al-Hakim's al-Madkhal 
ltd ma'rifat rijdl al-Sahihayn). Interestingly, Ibn Nasir al-Din cites both versions of the 
report side by side. 

'"' Ibn Hajar, Hady al-sdn, 1 . 

■" Al-Nawawl, Tahdhib al-asmd', 1:74; Ibn Hajar, Hady al-sdri, 1. 

'"' Ibn al-'Imad, Shadhardt al-dhahah, 2:134. 

" MuUa 'All Qarl, Mirqdt al-majdtih, 13. 



278 CHAPTER SEVEN 

Moroccan scholar Fath Allah b. Abi Bakr al-Bannani (d. 1934—5) con- 
curs in his commentary on al-Bukhari's work.'" In this dream etiology 
the impetus for initiating the sahih movement comes through direct 
inspiration from the Prophet himself, phrased as the hadith scholars' 
commendable duty to preserve his authentic legacy. 

It is important to note, however, that there was no categorical attempt 
to doctor the historical record. Encyclopedic and fastidious historians 
like Ibn 'Asakir, al-Dhahabi and Shah 'Abd al-'AzIz al-DihlawI (d. 1824) 
preserved the original wording of al-Khatib's report and excluded the 
isnad-less, account of al-Bukharl's dream.'' Nor should we assume that 
scholars like al-NawawI consciously altered the report originally found 
in Tarikh Baghdad. In the canonical culture of the Sahihayn, authenticity 
was the defining characteristic of al-Bukharl's work. For the scholars 
who copied al-Khatlb's history, it would have been an understandable 
oversight to interpolate the adjective ''saMf into al-Bukhari's account. 
As in language, the application of the Principle of Charity means gloss- 
ing over or reinterpreting momentary inconsistencies in the grammar 
of canonical culture. Working in the midst of the Sahihayn canonical 
culture, a copyist could not be faulted for subconsciously correcting 
this 'oversight.' 

b. Charity and Maintaining the Superiority of al-Bukhdri to Muslim 

The primacy of the SaMhayn in the Sunni vision of the Prophet's legacy 
represented both an act of communal consensus and the priorities 
that the Sunni tradition had set in elaborating the hadith sciences. 
The Sunni tradition was thus heavily invested in defending the posi- 
tion of the two books as the acme of hadith scholarship. Al-Shafi'l's 
statement that the Muwatta' was the most authentic (or useful) book 
after the Qur'an thus attracted a great deal of interpretive concern. 
Ibn Jama'a and Ibn Taymiyya explain that this opinion, trumpeted by 
Malikis like Ibn 'Abd al-Barr and al-Qadi 'lyad, in no way proves the 



''" Fath Allah b. Abi Bakr al-Bannam, Rafd al-qan bi-muqaddimat iftitah Sahih al~Bukhan 
(Rabat: al-Matba'a al-Maghribiyya al-Ahliyya, 1347/[1928-9]), 7. 

^' Ibn 'Asakir, Tarikh madinat Dimashq, 52:72; al-Dhahahi, Juz'Jihi tarjamat al-Bukhdri, 
ed. Hashim Ibrahim b. Mansur al-HashimI al-AmIr (Beirut: Mu'assasat al-Rayyan, 
1423/2002), 39; al-Subkl, fabaqdt, 2:221; Shah 'Abd al-'Aziz al-Dihlawl, Bustdn 
al-muhaddithin fi baydn kutub al-hadith wa ashdbihd at- 'uzz al-maydmin, ed. and trans. 
Muhammad Akram al-NadwI (Beirut: Dar al-Gharb al-lslami, 2002), 73-4. 



THE PRINCIPLE OF CHARITY 279 

superiority of the Muwatta ' to the Sahihayn or undermines the umma's 
consensus on the primacy of the two books. When al-Shafi'T made his 
evaluation, they explain, al-Bukhari and Muslim had not yet compiled 
their collections.''^ 

More difficult was maintaining the proper relationship between 
the Sahihayn themselves, which proved a persistent concern for Sunni 
guardians of the canonical culture. Ignoring al-Bukharl's superiority to 
Muslim in matters of critical methodology threatened the received opin- 
ion and practice among hadlth scholars on issues like the acceptability 
of narrations communicated by the phrase "from/on the authority of 
Can)." Although the vast majority of hadlth scholars recognized that 
al-Bukhari had produced a more thorough and demanding work, the 
opinions of several respected figures broke with this consensus. Abu 'All 
al-Naysaburi had said that Muslim's book was the most authentic work 
available."'^ Al-Qa.di 'lyad adds that a Maghribi scholar, Abu Marwan 
'Abd al-Malik al-Tubni (d. 456/1064)'^* mentioned that at least one 
of his teachers preferred Muslim's SaMh to that of al-Bukharl. Ibn 
Hajar and others mention that Ibn Hazm had also favored Muslim's 
work."'"' 

Although al-Khatib had indirectly undermined this minority opinion 
by mustering contrary evidence from towering sages like al-Daraqutnl, 
it was Ibn al-Salah who first actively attempted to disarm this threat to 
the Sahihayn canonical culture. He explains that if Abu 'All al-Naysaburi 
had meant that Muslim's work was superior only in that it did not 
include hadlths with incomplete isndd?, as legal commentary, this would 
be correct. If those scholars in the Maghrib mentioned by al-Q_adi 
'lyad preferred Muslim's SaMh because all the narrations of one Pro- 
phetic tradition are found in one place as opposed to being scattered 
throughout the work, this would also be a valid point. Asserting that 
Muslim surpassed al-Bukharl in methodology and judging authentic 
hadlths, however, was categorically incorrect.'^'' 



^^ Ibn Jama'a, al-Manhal al-rawi, 116-7; Ibn Taymiyya, Sihhat usul madhhab ahl al- 
Madina, ed. Zakariyya 'All Yusuf (Cairo: Matba'at al-Imam, [1964]), 34; al-Harawl, 
Jawdhir al-usul, 18. 

^' Al-QSi<ii 'lyS.d, Ikmdl at~mu'lim, 1:80. 

^^ Al-Safadi has his death as 456 AH; al-Safadi, al-Wdfi bi'l-wafaydt, 19:163. 

^^ Ibn Hajar, Hady al-sdri, 13. 

^^' Ibn al-Salah, Sijdnat Sahih Muslim, 69; al-Navvawi, Shark Sahih Muslim, 1:121. 



280 CHAPTER SEVEN 

This explanation became commonplace among later defenders of 
the canonical culture such as al-NawawI and Ibn Hajar'^' Al-Sakhawi's 
student 'Abd al-Rahman b. 'All Ibn al-Dayba' (fl. 900/1500) composed 
a verse: 

People have disputed before me concerning al-Bukharl and Muslim, 
which should we favor? 

I said, "Indeed al-Bukharl has excelled in authenticity, as Muslim excelled 
in finely crafting [his book].'"" 

Ibn Hajar further attempted to neutralize Abu 'All al-Naysaburi's 
comment by suggesting that no evidence existed that the scholar had 
ever seen al-Bukharl's book.'^" The fact that certain Maghribi scholars 
preferred Muslim's Sahih to that of al-Bukhan, he continued, does not 
entail that Muslim's work was more reliable. Ascribing "preference 
{afdaliyyaj" to a work is not equivalent to ascribing it "greater authentic- 
ity (asahhiyjia).'"''" Al-Subkl's defense of the canonical culture was more 
blunt; he stated simply that "there is no weight to the opinion of those 
who favor Sahth Muslim to it [Sahth al-Bukhart], since that opinion is 
irregular {shddhdhd) and is thus not to be depended on.'""' 

c. Chanty and Muslim's Meeting with AbU ^ur'a al-RdzT 

In all accounts of Muslim's encounters with Abu Zur'a al-RazT, the 
tension surrounding the notion of limiting the collection of authentic 
reports is palpable. When one of Abu Zur'a's colleagues introduces 
Muslim as the man who had collected a book of four thousand authen- 
tic traditions, numerous reports describe Abu Zur'a as objecting, "To 
whom [li-man)/w\\y (li-md) did he leave the rest?" This comment fore- 
shadows the efforts of al-Hakim al-Naysaburl to increase the number 
of authentic hadlths in circulation and reinforces the mainstream stance 
that al-Bukharl's and Muslim's works did not exhaust the corpus of 
authentic hadlths. Although Abu Zur'a's remark seems slightly criti- 
cal of Muslim, in actuality it implicitly legitimizes the actions of later 
scholars who would use the "standards of al-Bukhari and Muslim" to 



" See also, MuUa 'All Qarl, Mirqdt al-mafatih, 1:16, where the author replicates Ibn 
Hajar's discussion. 

^^ Shah 'Abd al-'AzIz al-DihlawI, Bustdn al-muhaddithin, 78. 

^^ Ibn Hajar, Hady al-sdn, 13; cf. idem, al-Nutmt 'aid kitdb Ibn al-Saldh, 62-3. 

™ Ibn Hajar, Hady al-sdri, 13. 

" Al-Subkl, Tabaqdt, 2:215. 



THE PRINCIPLE OF CHARITY 281 

extend the authority of the canon to new material. This report thus 
frequently appears in later work on Muslim's Sahih. 

The most complete versions of this encounter, however, include a far 
more critical remark by Abu Zur'a. Ibn 'Asakir and al-Dhahabi preserve 
an additional section in which Abu Zur'a further berates Muslim in his 
absence for not properly respecting al-Dhuhli. It reads: 

Abu Quraysh said: We were with Abu Zur'a, and Muslim came and 
greeted him. He sat down for a while and they [two] discussed [tadhdkara) 
hadlths. When Muslim left I said to Abu Zur'a, "He has collected 4,000 
hadlths in 'the Sahih,' " and Abu Zur'a said "Why did he leave the rest 
{li-md taraka al-bdqi)?" Then [Abu Zur'a] said, "He doesn't have any sense 
{laysa li-hddhd 'aql); if he'd tended properly to [ddra) Muhammad b. Yahya 
[al-Dhuhli] he'd have become a man.'"'^ 

Ibn al-Salah's rendition of this report in his Siyanat Sahih Muslim min 
al-ikhldl wa al-ghalat (Preserving Sahih Muslim from Ruin and Error), 
however, excludes Abu Zur'a's critical remark about al-Dhuhli.''^ This 
truncated version is repeated in al-Nawawi's famous commentary on 
Muslim's Sahih and in Ibn Nasir al-Din al-Dimashqi's Ifiitah al-qari li- 
Sahih al-Bukhan^^ These scholars' decision to omit the second part of 
Abu Zur'a's statement represents a defense of the canonical culture 
surrounding the Sahihayn. Not only does Abti Zur'a's comment belittle 
Muslim, accusing him of poor judgment as well as subordinating him 
to al-Dhuhli, it also threatens the canonical version of the quarrel 
between al-Bukhari, Muslim and al-Dhuhll. 

As we saw in Chapter Three, although al-Dhuhli's attack on al- 
Bukhari certainly inflamed his quarrel with Muslim, the falling out 
between al-Dhuhli and Muslim was the culmination of a series of 
disagreements between the two. In al-Khatlb's personal commentary, 
however, Muslim's alienation from al-Dhuhll centers on the former's 
stalwart and loyal defense of al-Bukharl. In his Tdnkh Naysdbur, al-Hakim 
seconded this by reporting that only Muslim and Ahmad b. Salama 
had stayed with al-Bukharl when al-Dhuhli denounced him.''"' This 



''^ Al-Dhahabi, Tdnkh al-isldm, 12:187; Ibn 'Asakir, Tdnkh madinat Dimashq, 58:93. 
These two versions feature the initial wording "why did he leave the rest?" Cf. al- 
Dhahabl, Tdnkh al-isldm, 19:341 (this version includes the wording "to whom did he 
leave the rest?"). 

'■' Ibn al-Salah, Siydnat Sahih Muslim, 101. 

'"' Al-Nawawl, Sharh Sahih Muslim, 1:129; Ibn Nasir al-Din al-Dimashql, Majmu'Jihi 
rasd'il, 336. 

"^ Ibn Hajar, Hady al-sdri, &11 . 



282 CHAPTER SEVEN 

theme matured more fully in the work of Abu 'Abdallah Muhammad 
b. 'All al-Mazari (d. 536/1141), who asserted that Muslim was in fact 
the only person who stood by al-Bukhari when the scholars of Naysabur 
turned against him/''' 

Abu Zur'a's comment challenges this narrative. Indeed, it is far more 
congruent with the pre-canonical notion that Muslim and al-Dhuhli 
were involved in a private drama between student and teacher Abu 
Zur'a clearly sides with al-Dhuhli, faulting Muslim for neither showing 
his teacher the proper respect nor finishing his education with him. 
To retain the additional section would be to undermine the scenario 
of al-Bukhari and Muslim standing against a jealous and fickle mob 
driven by al-Dhuhli, threatening al-Bukhari's vindication and the united 
front of the Shaykhayn. 



Reconciling the Canon with Convention: The Sahihayn and the Rules of Hadith 

Although al-Hakim al-Naysaburi and al-Khatib al-Baghdadi had often 
invoked al-Bukhari and Muslim as models of excellence to be followed 
in the collection and criticism of Prophetic hadiths, these sciences func- 
tioned according to rules external to the SaMhcyn. Before al-Bukhari 
and Muslim, generations of great critics such as Malik b. Anas, 'Abd 
al-Rahman b. Mahdi and 'All b. al-Madlnl had sifted through thousands 
of hadith notebooks, sorting the strong from the weak according to their 
own criteria. Even in the wake of al-Bukhari's and Muslim's illustrious 
careers, scholars like Ibn 'Ammar al-Shahid and al-Daraqutni flour- 
ished according to their own idiosyncratic methodologies. Al-Daraqutnl 
maintained standards for transmitters that sometimes proved stricter 
than those of al-Bukharl, while Ibn 'Ammar al-Shahid could require 
a stronger reliance on written sources than Muslim. Both upheld more 
stringent standards for the acceptance of Addition that those employed 
in the SaMhcyn. 

Even after the canonization of al-Bukhari and Muslim, some schol- 
ars espoused standards for the evaluation of hadiths that far exceeded 
those of the Shaykhayn. The Shafi'i legal theorist and hadith scholar Abu 



'''' Abu 'Abdallah Muhammad b. 'Ah al-Mazan, al-Mu'lim bi-fawa'id Muslim, ed. 
Muhammad al-Shadhih al-Nayfar, 2nd ed., 3 vols. (Beirut: Dar al-Gharb al-Lslaml, 
1992), 1:182. 



THE PRINCIPLE OF CHARITY 283 

al-MuzafFar Mansur al-Sam'ani of Khurasan (d. 489/1096), for exam- 
ple, proved even more rigorous than al-Bukharl in his requirements for 
using "from/ on the authority of ( (zn)" in transmission. Beyond the mere 
requirement of having met at least once, he demanded that the trans- 
mitter have studied extensively with his teacher {0 al-suhba)S'^ 'Uthman 
b. Sa'id al-Dani of Andalusia (d. 444/1053) required the scholar nar- 
rating via " 'an" to be well-known as a narrator from that source.''" 

In addition to the personal methodologies of individual scholars, the 
Sahihayn canon might also stand in tension with the general conventions 
of Sunni hadith scholarship. This tradition reached maturity in the 
writings of Ibn al-Salah, whose monumental treatise on the sciences of 
hadith transmission and criticism became the basis for later studies in 
the field. ^^ With the systematization of the hadith tradition that began 
with al-Hakim and solidified with Ibn al-Salah, hadith scholarship 
acquired a unified and refined authority that could present a serious 
challenge to the Sahihayn canon. The conventions of the hadith tradition 
comprised a body of rules that the Sahihayn might occasionally fail to 
follow. The canon fulfilled important functions in the scholarly and lay 
community, so how could hadith experts address instances in which the 
two books fell short of the standards established by the hadith tradi- 
tion? This potential tension between the practice of hadith scholars 
and the authoritative institution of the Sahihayn canon would have to 
be resolved by recourse to the Principle of Charity. 

a. Chanty and Tadlls 

One of the most glaring areas in which the Sahihayn occasionally ran 
afoul of the accepted practice of Sunni hadith scholarship was tadlis, 
or obfuscation, a phenomenon that occurred in two contexts. First, 
tadlis could entail a student narrating something from a teacher with 
whom he had studied but from whom he had not actually heard that 
particular report (generally termed tadlis al-isndd). Secondly, tadlis could 
involve a student obfuscating the identity of his source (termed tadlis 
al-shaykh). In both cases, tadlis consisted of misleading others about the 
true immediate source of one's hadlths. The first type of tadlis occurred 



''' Al-Sam'am, Qawdti' al-adilla, 2:456-7. 

1^" Ibn Kathir, al-Bd'ith al-hathith, 45. 

'"' See J. Robson, "Hadith: the Study and Transmission of Tradition," EP. 



284 CHAPTER SEVEN 

commonly, and often not due to any deceptive intent. If a student 
attending the dictation sessions of a certain teacher excused himself 
to answer nature's call and later heard the material he had missed 
from another student, he might omit his colleague from the chain of 

transmission and simply state "the teacher said " The second type of 

tadlTs could also be innocuous, often resulting from a transmitter assum- 
ing that his audience understood who his sources were without giving 
their full names. It could also, however, serve to disguise an impugned 
or discredited source. If a transmitter said "a notable scholar told me," 
he might be trying to employ a hadith that he had actually heard from 
a person others considered unreliable or heretical. 

In the wake of al-Khatib al-Baghdadi's writings, what emerged as 
the regnant policy among Sunni hadith scholars for evaluating the first 
type of tadlis was that one could accept a report from someone known 
to commit tadlis (called a mudallis) provided that he explicitly stated 
that he had heard the report directly {samd') from his source.'" This he 
could accomplish by using technical terms known to denote face-to-face 
transmission, such as "he narrated to us {haddathand)," "I heard from 
him {sami'tu)" or "he reported to us (akhbarand)." If the mudallis used 
a vaguer phrase, such as "from/on the authority of ('an)" or "so and 
so said iqdla)," the hadith could not be accepted as authentic due to 
a presumed break in the chain of transmission. Ibn al-Salah affirmed 
this position in his classic manual on the hadith sciences, and no sig- 
nificant objection to this policy appeared. Employing the Sahihayn as 
an exemplum, he stated that al-Bukhari's and Muslim's collections, as 
well as other relied-upon books, often depended on the transmission 
of a mudallis if it was phrased in wording that eliminated any doubt 
about the continuity of transmission.'' 

As Ibn Hajar later noted, however, the Sahihayn also contain numer- 
ous hadlths in which a mudallis narrates from his source via the prob- 
lematic phrase "from/on the authority of {'an)." Here it seemed that 
al-Bukhari's and Muslim's collections could not provide the evidence of 
continuous transmission required by convention among hadith scholars. 
Only reading the Sahihayn in the most favorable light could resolve the 
inconsistency between the canon and the rules of hadith scholarship. 



'" Ibn Hibban, Sahih, 1:122; al-Khatib al-Baghdadi, al-Kifaya, 2:385-6; cf. al-Sam'anI, 
Qawdti ' al-adilla, 2:312. 

" Ibn al-Salah, al-Muqaddima, 235; al-' Ala i, Jdmi' al-tahsil, 1 1 1—12; al-Sakhawi, Fath 
al-mughith, 1:227 IF. 



THE PRINCIPLE OF CHARITY 285 

Ibn al-Salah's follower, al-Nawawi, recognized this and authoritatively 
declared, "Know that what is in the Sahihayn [narrated] from mudallisa 
via [the phrase] "an' or something like it is to be interpreted [mahmul) 
as having been established as direct transmission [samd') via some other 
narration [of the hadith] '"^ 

Important hadith scholars accepted al-Nawawl's extension of charity 
to all instances of tadlis in the Sahihayn. The Levantine Mamluk-period 
scholar Khalil b. Kaykaldl al-'Ala'l (d. 761/1359) treated both al-Bukhari 
and Muslim with extreme charity in his definitive monograph on the 
issue of broken transmissions. He explains, for example, that in the case 
of the famous mudallis, the Successor Abu al-Zubayr Muhammad b. 
Muslim al-Makki (d. 126/743-4), many senior hadith scholars refused 
to use reports he narrated from the Companion Jabir b. 'Abdallah as 
proof texts. Such critics only accepted what the great Egyptian scholar 
al-Layth b. Sa'd (d. 175/791) had vetted from al-Makki. Al-'Ala'i, how- 
ever, notes that Muslim's Sahih contains numerous hadiths from Jabir -^ 
al-Makkl that al-Layth did not narrate though this isndd. Yet he adds 
that it was "as if Muslim, may God bless him, was aware that these 
[hadiths] were from material that al-Layth narrated from [Jabir] even 

if he did not narrate them through his path [of Jabir -^ al-Makki] " 

Al-'Ala'i thus assumes Muslim knew that al-Layth had approved of this 
material even though it did not meet the standards scholars generally 
employed when evaluating al-Makkl's hadiths." 

After providing a long list of notorious mudallises, al-'Ala'l admits that 
"there are many hadiths from these [transmitters] in the Sahihayn" that 
lack explicit evidence for direct transmission. Referring to al-NawawI, 
he adds, "One imam has interpreted (hamala) this as that the Shaykhayn 
were aware of the direct transmission [samd') of the individual for that 
hadith . . . but this is a lengthy matter [wafihi tatwil)." Although al-'Ala'l 
feels that al-Nawawl's argument is slightly tenuous, he nonetheless 
states that al-Bukharl and Muslim included such reports because they 
had reliable evidence that their transmitters could be trusted and an 
uninterrupted chain of transmission guaranteed.'* 



'^ Al-Nawawi, Shark Sahth Muslim, 1:146. 

" A1-' Ma'i, Jami' al-tahsil, 126. For his biography, see Ibn Hajar, al-Durar al-kdmina, 
2:52. 

'* M-' MaT, Jdmi' al-tahsil, 130. 



286 CHAPTER SEVEN 

Ibn Hajar categorically supports al-NawawT's charitable treatment 
of the Sahihayn. He states that any instance of tadlis via "from/on the 
authority of ((zn)" occurring in the primary [usul) narrations of the 
Sahihayn is assumed to be a locus of direct transmission. If al-Bukhari or 
Muslim included the report of a mudallis using 'from/ on the authority 
of {'ari)' in the isndd among their auxiliary (mutdba'a/shawdhid) narrations, 
this presented no problem since the two scholars did not uphold their 
rigid criteria in these cases. '^ Qutb al-Din 'Abd al-Karim al-Halabi 
(d. 735/1335) stated that all these instances of tadlis though the phrase 
"'an" should be treated as direct transmission since "the instances of 
'an in the Sahihayn have the status of direct transmission.""" Al-Dhahabi 
even exempted "what is in Sahih al-Bukhdn and similar books" from the 
second type of tadlis, the obfuscation of one's teacher's identity. He 
explains that whomever al-Bukharl uses as a source is reliable.'' 

Several hadith scholars who exempted the Scihihcyn from the stan- 
dard rules governing the evaluation of tadlis seemed very conscious 
of the charity they had extended the two books. Taqi al-Din al-Subki 
once asked Jamal al-Din al-MizzI (d. 742/1341), the compiler of the 
most comprehensive biographical dictionary of hadith transmitters, if 
al-Bukhari and Muslim had really made certain that all instances in 
their collections in which tadlis had occurred were guaranteed by direct 
transmission. Al-Mizzi replied, "So it is said, but that is only out of 
giving the benefit of doubt {tahsin al-zann) to these two, since otherwise 
there are hadiths narrated by mudallises that exist only via that narration 
found in the SahihfcynJ."'^ Al-'IraqI echoes this when he explains that 
the umma's consensus on the Sahihcyn demands that Muslims extend 
"the benefit of doubt (tahsin al-zann)" to the two works. '^ 



b. Charity and Transmitters 

Al-Khatib al-Baghdadi had stated that al-Bukhari and Muslim occasion- 
ally relied on transmitters who had been previously impugned as part 
of his argument that such criticisms were only valid if accompanied 



'' Ibn Hajar, al-Nukat 'aid kitdb Ibn al-Saldh, 255-6. 

'*' Al-Sakhawl, Fath al-mughith, 1:233. For al-Halabl's biography, see Ibn Hajar, al- 
Durar al-kdmina, 2:243-4, 

" AI-Dhahabi, al-Muqiia, 50. 

'" Ibn Hajar, al-Nukat 'aid kitdb Ibn al-Saldh, 256. 

'" Al-'IraqI, al-Taqyid wa al-iddh, 366. 



THE PRINCIPLE OF CHARITY 287 

by some explanation. Al-Khatib was only invoking al-Bukhari and 
Muslim as part of this larger argument, and he was wise not to claim 
that none of the transmitters featured in the Sahihayn had been criticized 
without good reason. Al-Bukharl and Muslim relied on Ayyub b. 'A'idh 
al-Ta'i, for example, whom al-Bukharl himself had accused of being 
a Murji'ite.^" We have already seen the example of the arch-Kharijite 
'Imran b. Hittan, through whom al-Bukharl transmitted a hadith. As 
the fifth /eleventh century drew to a close, however, and the Sahihayn'% 
role as an authoritative reference and a measure of authenticity became 
better established, the questionable status of some of al-Bukhari's and 
Muslim's transmitters emerged as a problem. If, as al-Nawawi replied 
in his fatwd, the Sahihayn contained only authentic hadiths, how should 
scholars handle the presence of impugned transmitters in the two 
collections? 

One of al-Khatlb's students, Muhammad b. Futuh al-Humaydi 
(d. 488/1095), an Andalusian who settled in Baghdad and composed 
his famous combined edition of the Sahihayn,^^ proffered the Sahihayn 
as an exemplum to be imitated in evaluating hadith transmitters. The 
two works, in fact, provided veritable dictionaries of reliable, upstanding 
narrators. He asserted that the most important result of al-Bukhari's 
and Muslim's work was their declaration of the uprightness {'adala) 
of all the narrators of the principal hadiths (usul) included in the two 
books. Al-Humaydi's claim was built on the canonical authority of al- 
Bukhari and Muslim, for: 

The testimony of those two imams, or one of them, to that effect, and 
their declaring [that narrator] as sahih is an assessment [hukm] that requires 
foffowing, a message designed to be heeded {jata'ayjanu al-inqiydd lahu), and 
a cautioning inidhdra) the disobedience of which is to be feared "^ 

The authoritative station of al-Bukhari and Muslim therefore demanded 
a charitable view of their transmitters. Al-Humaydl's younger contem- 
porary, Muhammad b. Tahir al-MaqdisI, echoed this, stating that even 
if some of al-Bukhari's and Muslim's transmitters had been criticized, 
inclusion in the Sahihayn trumps this. The Shaykhayn, he explained, 
only narrated from "trustworthy, upright masters [thiqa 'adl hdfiz) with 



"" Al-Bukharl, al-Tdnkh al-kabir, 1:420. 
"' See al-Dhahabi, Tadhkirat al-hujfai, 4:13-14. 

"^ Abu ' Abdallah Muhammad b. Futuh al-Humaydl, al-Jam ' bayn al-Sahihayn, ed. 'All 
Husayn al-Bawwab, 4 vols. (Beirut: Dar Ibn Hazm, 1419/1998), 1:76. 



288 CHAPTER SEVEN 

a Strong probability of having heard from the preceding person in the 
isnad, except for a very few instances [ahruf'"').'"^'^' 

It was the Maliki hadith scholar AbQ al-Hasan 'All b. al-Mufaddal al- 
Maqdisi (d. 611/1214) who demanded total charity towards al-Bukhari's 
and Muslim's transmitters by declaring famously that all those included 
in the Sahihayn "have passed the test {jdza al-gantard)."^* This principle 
proved axiomatic for Ibn al-Salah a few decades later In his Muqad- 
dima he says that hadith scholars should not pay heed to criticism of 
those whom al-Bukhari and Muslim included in the Sahihayn.^''' In his 
defense of Muslim's Sahih, Siydnat Sahih Muslim, Ibn al-Salah specifi- 
cally exonerates Muslim from any criticism for using weak transmit- 
ters. All such criticisms of Muslim, he argues, can be rebutted by one 
of four points. First, if Muslim used narrators that other experts had 
criticized, it is assumed {mahmul) that the criticism was not adequately 
established. He adds, "And it is also probable that these are instances 
in which, even if the critic {jdrih) did clarify his reason [for criticizing 
one of Muslim's men], Muslim demonstrated its falsity." Second, the 
weak narration may not be one of Muslim's primary hadiths, but rather 
one of his less rigorous auxiliary narrations {shawdhid, mutdbi'dt). Third, 
the narrator in question may have lost his reliability only after Muslim 
had taken hadiths from him. Finally, referring to Muslim's explanation 
to Ibn Wara, he might have used a narration with a weak transmitter 
because its isndd was shorter than a more reliable version.'*'' 

Ibn al-Salah's follower, al-NawawI, repeated these reasons for exon- 
erating Muslim. He concluded that although a number (jamd'd) of nar- 
rators from the Sahihayn have been criticized, it emerges upon reflection 
that trust (thiqa) is conferred upon them and that one must accept their 
hadiths."^ Moreover, al-Nawawi cunningly reinterpreted al-Khatib al- 
Baghdadi's aforementioned argument to provide an earlier historical 
precedent for treating al-Bukhari's and Muslim's transmitters with total 
charity. Arguing that "criticism [of narrators] is not accepted unless it 



"' Al-Maqdisi, Kitab al-jam' bayn kitahay Abi Masr al-Kalahadhi iva Abi Bakr al-hbaham, 
1:3. 

"' Ibn Daqlq al-'Id, al-Iqtirdh, 327. Ibn Daqlq does not identify al-MaqdisI beyond 
the fact that he is his teacher's teacher and that his name is AbU al-Hasan. See al- 
Dhahabl, Sijar, 22:66-9. 

"' Ibn al-Salah, Muqaddimat, 292. 

"'' Ibn al-Salah, Siydnat Sahih Muslim, 96 ff. 

'" Al-Nawawl, al-Taqnh, 17; idem, Sharh SaKh Muslim, 1:134. 



THE PRINCIPLE OF CHARITY 289 

is explained," al-Khatib had added, "for indeed al-BukharT relied on 
{ihtajja) a number [of transmitters] who had been previously criticized 
by others..., as did Muslim b. al-Hajjaj..., Abu Dawtid al-Sijistani, 
and more than one other. . . .'""' Paraphrasing al-Khatib, al-NawawI 
interpreted this as the extension of complete charity to al-Bukhari's and 
Muslim's transmitters. He states, "Al-Khatib al-Baghdadi and others 
have said, 'What al-Bukhari, Muslim and Abu Dawud used as proof 
{ihtajja bihi) from among a number [of transmitters] who had been 
criticized before by others, is to be treated [mahmul) as if no effective, 
explained criticism had been established.""'^ 

What al-Khatib had intended as evidence that criticisms of transmit- 
ters were not valid unless accompanied by some explanation, al-Nawawi 
thus transformed into an exemption of al-Bukhari's, Muslim's and Abu 
Dawud's transmitters from any criticism. The charitable premise on 
which al-NawawI bases this act of legerdemain, however, lacks cred- 
ibility. As discussed above, some transmitters used in the Sahihayn were 
indeed criticized with valid explanations.^" 

Al-Khatib al-Baghdadi was a foundational figure in the systematiza- 
tion of the Sunni hadith tradition — Abu Bakr b. Nuqta (d. 629/1231) 
elegized him by stating that "no one of sound thought can doubt that 
the later scholars of hadith are utterly dependent on ['iydl 'ala) Abu Bakr 
al-Khatib."^' But al-Khatib's works provided no extension of charity 
to the Sahihayn comparable to the statements made by al-Humaydl, 
al-Maqdisi, Ibn al-Salah or al-NawawT. Al-Nawawi's interpretive leap, 
however, grounded his exemption of al-Bukharl's and Muslim's trans- 
mitters from the conventional rules of hadith criticism as articulated by 
al-Khatlb. Moreover, generations of later hadith scholars have treated 
al-Nawawi's paraphrase as the words of al-Khatib himself!^^ In his 
book on al-Bukhari, the modern scholar 'Abd al-Ghani 'Abd al-Khaliq 
attributes the statement directly to al-Khatib, even omitting mention 



™ Al-Khatib, al-MJuya, 1:339. 

" Nawawl, Shark Sahih Muslim, 1:134. 

'"' Al-San'anI points this out; al-San'anl, Tawdih al-qfkdr, 1:99. 

" Abu Bakr Muhammad b. 'Abd al-Ghanl Ibn Nuqta al-Baghdadi, Kitab al-Taqyid 
li-ma'nfat ruwdt al-sunan iva al-masdnid, ed. Kamal Yusuf al-Hut (Beirut: Dar al-Kutub 
al-'Ilmiyya, 1408/1988), 154. 

'^ See, for example, Badr al-Dln al-'Aynl, 'Umdat al-qdri, ed. Idarat al-Tiba'a al- 
Mumriyya et al., 25 vols, in 12 (Beirut: Muhammad Amin Damaj, [1970], reprint of 
the 1891 Cairo edition, citations are to the Beirut edition), 1:8; MuUa Khatir, Makdnat 
al'Salnhajn, 238. 



290 CHAPTER SEVEN 

of Abu Dawud.^' Another present-day scholar, 'Abd al-Mu'ti Amin 
Q_arajl, has done the same.^* 

In the wake of al-Nawawi's statement, many later pillars of the 
hadith tradition exempted al-Bukhari's and Muslim's transmitters from 
criticism. In his abridgment of Ibn al-Salah's work, the Egyptian Ibn 
Daqlq al-'Id (d. 702/1302) acknowledges that some of al-Bukharl's and 
Muslim's transmitters have been criticized. Explaining Abu al-Hasan 
al-Maqdisl's famous declaration that the SahThayn'i, transmitters "passed 
the test," Ibn Daqiq states that he meant, "He pays no heed to what 
is said [critically] about them; this is what he believes, and this is our 
opinion." Ibn Daqiq thus instructs those seeking to determine whether 
or not a narrator is reliable to consult the Sahihayn as a dictionary of 
accepted transmitters. The Muslim community's consensus on the 
two books, its collective decision to dub them "the two Sahihs," and its 
referral to them for rulings on authenticity make the two works the 
most reliable source. ^'^ 

Ibn Daqiq's student al-Dhahabl takes the same course in his even 
more succinct reference work on the technical terms of hadith criti- 
cism. If someone is included in the Sahihayn, he is automatically deemed 
reliable {thiqa) by that fact alone. If this transmitter appears only in 
al-Tirmidhl's or Ibn Khuzayma's collections, however, he merits the 
less lustrous rating of "good (jcyyid)."^'' Al-Dhahabi further echoes his 
teacher: "All those included in the Sahihcyn have passed the test [qafaza 
al-qantard), and one cannot turn away from them [la ma'dil 'anhu) except 
by some clear evidence (burhan)."^^ Al-DhahabT even urges readers to 
ignore criticism of those transmitters from the SaMhoyn that he had 
included in his own dictionary of impugned narrators, the Mizdn al-i'tiddl 
(The Scale of Judgment). He states that these criticisms "should not be 
heeded," and adds that "if we open that door to ourselves, a number 
of the Companions, Successors and imams would enter it."^** 

Al-Dhahabl's analogy between the transmitters of the Sahihcyn and 
the Companions of the Prophet is apt, for both groups received the 
blanket approval of the umma. Al-'Iraqi recognized the comparable 



'' 'Abd al-Khaliq, al-Imdm al-Bukhdn wa Sahihuhu, 221 . 

" See al-'Uqayll, Kitdb al-du'ajd', 1:54 (editor's introduction). 

'' Ibn Daqiq al-'Id, al-Iqtirdh, 326-8. 

'« Al-Dhahabi, al-Muqiza, 78. 

" Al-Dhahabi, al-Muqi^a, 80. Ibn Hajar repeats this argument; Ibn Hajar, Hady 
al-sdn, 543. 

'" AI-Dhahabl, Ma'rifat al-ruwdt al-mutakallam fihini, 45. 



THE PRINCIPLE OF CHARITY 291 

charity extended to these two groups when he noted that the only two 
classes of hadlth transmitters whose status is not affected by only having 
one narrator from them, which would normally render them majhul, are 
the Companions and the men of al-Bukharl and Muslim. ^^ 



Rebutting Earlier Criticisms 

The most compromising consequence of the inconsistencies between the 
methods employed by al-Bukhari and Muslim in their works and those 
of other prominent hadith scholars was the criticisms that venerated 
critics made of the Sahihayn. The critique of the great fourth/tenth- 
century hadlth scholar, al-Daraqutni, as well as those of the Andalusian 
muhaddith Abu 'All al-jayyani al-Ghassani (d. 498/ 1 1 05) and the North 
African Maliki 'Abdallah Muhammad b. 'All al-Mazari (d. 536/1141) 
proved the most problematic for the maintenance of the Sahihayn 
canonical culture. It was to these criticisms that the canonical culture's 
greatest advocates, Ibn al-Salah, al-NawawI and Ibn Hajar thus turned 
their attention. Although these three masters' inimitable command of 
the hadlth tradition allowed them to effectively overturn many of these 
earlier criticisms, their defenses also relied on charitable assumptions 
about al-Bukhari's and Muslim's work. Indeed the Principle of Charity 
imbued the notion that the Sahihayn'% auxiliary narrations were not to 
be held to the same standard as their primary hadiths, as well as the 
claim that al-BukharT and Muslim included problematic narrations 
only because they assumed their audience would know more reliable 
versions. 

It is important to note that the canonization of the Sahihayn did not 
end criticism of the two works. As we saw in Chapter Six, the very 
Ulusory nature of the SaMhcyn canon enabled criticism of its contents 
even as scholars wielded it against opponents. Even scholars who actively 
employed the Sahihayn canon occasionally criticized a hadlth from the 
two books if it contradicted the doctrines of their school of law or 
theology. The arch-Shafi'i al-Bayhaqi thus criticized Muslim's report 
that ordered that one should not say the basmala out loud. 



Al-'Iraqi, al-Taqyid wa al-idah, 123. Al-'Iraqi even wrote a book on these men. 



292 CHAPTER SEVEN 

Hadith scholars also continued to criticize items from the Sahihayn 
not for partisan purposes, but as part of their unabated critical review 
of transmissions from the Prophet.'"" As al-'iraqi had said, evaluating 
reports was "the muhaddith%' job." Like earlier 'Hal studies, most such 
criticisms involved problems in the chains of transmission of certain 
hadiths, such as breaks in isnddi or inappropriate Addition. Al-Mazari 
thus singled out fourteen instances of broken isndd's, in Muslim's Sahih. 
Abu al-Husayn Hibatallah Ibn 'Asakir (d. 563/1167—8) appended five 
original criticized narrations he had culled from Muslim's Sahih to 



100 'Yhh critical review of the Sahihayn also stemmed from the very nature of manu- 
script transmission in the pre-print world. A constant reexamination of a text was 
required in order to prevent errors from creeping in as students copied their teach- 
ers' books. Abu 'All al-Jayyanl's criticisms of al-Bukharl and Muslim thus originated 
from his efforts to synchronize the variant transmissions of the two texts. Although he 
never left Andalusia, al-jayyani had access to all the major recensions of the works, 
and produced a book on the inconsistencies and ambiguities in the SaMhayrii transmis- 
sion. His criticisms of al-Bukharl's and Muslim's texts therefore often involve errors 
that had materiahzed during the transmission process, though he also notes mistakes 
made by the authors themselves. In the case of Mushm's work, he has a section on 
'Hal not mentioned by al-Daraqutnl in his Kitab al-tatabbu'. There, for example, he 
criticizes Muslim for erring in the identity of a certain transmitter and inappropri- 
ate isndd Addition; al-GhassanI, Kitdb al-tanbih 'aid al-awhdm al-wdqi'aji Sahih al-imdm 
Muslim, 51, 55. It is important to note that many of the errors that al-jayyani notes 
occur only in Ibn Mahan's recension of the Sahih; see ibid., 73. For al-Bukhari, he also 
has a short section on 'Hal in what is otherwise also a book designed to compare and 
correlate transmissions of his Sah^h', al-GhassanI, Kitdb al-tanbih 'aid al-awhdm al-wdqi'a 
Ji al-musnad al-sahih li'l-Bukhdri, 111-2. For studies by Muslim scholars on the transmis- 
sion of al-Bukharl's Sahih and the scholars who played a prominent role in editing it 
at different stages, see Ibn Rushayd, IJddat al-nasih fi al-ta'rif bi-sanad al-Jdmi' al-sahih, 
ed. Muhammad al-Hablb Ibn al-Khawja (Tunis: al-Dar al-Tunisiyya, [1973]); Yusuf 
b. 'Abd al-Hadi Ibn al-Mubrid (d. 909/1503-4), al-Ikhtildf bayn ruwdt al-Bukhdri 'an 
al-Firabri wa riwdydt 'an Ibrahim b. Ma'qil al-Nasafi, ed. Salah Fathi Halal (Riyadh: Dar 
al-Watan, 1420/1999). For modern studies on scholars who edited the authoritative 
versions of Sahih al-Bukhdri, such as the Indian who settied in Baghdad, al-SaghanI 
(d. 650/1252), and the Egyptian Hanbah al-Yunlnl (d. 658/1260), see Alphonse 
Mingana, An Important Manuscript of the Traditions of al-Bukhdri (Cambridge: W. Heffer 
and Sons, 1936); Rosemarie Qiairing-Zoche, "How al-Bukharl's Sahih was edited in 
the middle ages: All al-Yunlnl and his Rumdz," Bulletin d'Etudes Orientates 50 (1998): 
191-222; and Johann Fiick, "Beitrage zur Uberlieferungsgeschicte von Buharl's 
Traditionssammlung," ^eitschrift der Deutschen Morgenldndischen Gesellschaft 92 (1938): 
60-82 (this article has several detailed charts of the transmission of the Sahih)- For 
a discussion of the transmission of Muslim's Sahih, see Diya' al-Din al-MaqdisI, "al- 
Ruwat 'an Muslim," in Juz'dn 'an al-imdm Muslim b. al-Hajjdj, ed. Abu Yahya 'Abdallah 
al-Kandarl and Abu Ahmad Hadi al-MurrI (Beirut: Dar Ibn Hazm, 1416/1996);James 
Robson, "The Transmission of Muslim's Sahih," Journal of the Royal Asiatic Society (1 949): 
46—6 1 . For a discussion of the textual authenticity and attribution of al-Bukharl's and 
Muslim's works, see Appendix II. 



THE PRINCIPLE OF CHARITY 293 

the end of his copy of Ibn 'Ammar's 'Hal work.'"' A later copyist of 
the same manuscript, one Muhammad b. al-Hasan b. Abl al-Fadl of 
Damascus (d. 630/1232-3), added one more narration he had found 
in his reading of Muslim for Normative Matn Addition.'"^ The bold- 
est isnad criticisms of the Sahihayn came from the great HanbalT jurist, 
preacher and pious activist of Baghdad, Ibn al-jawzl (d. 597/1200). In 
his famous Kitdb al-mawdu'dt (Book of Forgeries), Ibn al-jawzl includes 
at least two narrations from Sahth al-Bukhdn and one from Muslim's 
collection due to various flaws in their isnddsJ"-' 

Ibn al-Salah represents the first holistic champion of the Sahihayn 
against earlier criticisms. His commentary on Muslim's work has been 
lost, but much of his efforts at defending the Sahihayn have survived in 
his Siydnat Suhih Muslim. Although Ibn al-Salah tries to overturn a criti- 
cism whenever possible, his main strategy centers on invoking charity: 
he claims that any problematic narration of a hadith either comes from 
al-Bukharl's or Muslim's less demanding auxiliary narrations or that a 
correct version appears in authentic forms elsewhere. Although he is 
able to find evidence from other major hadith collections to disprove 
one of al-Jayyanl's criticisms, he must resort to the Principle of Char- 
ity for rebutting al-Daraqutnl and al-Mazarl. '"* He objects to Mazarl's 
statement that Muslim's SaMh has fourteen narrations with breaks in 
their chains of transmission [inqitd'), arguing: 

This falsely conveys an impression of disarray [yuhim khalal""), and that 
is not the case. For there is nothing of that sort, praise be to God, for 
he [Muslim] included these [problematic narrations], especially what has 
been mentioned here, as auxiliary narrations [mutdba'a) and included a 



'"' Ibn 'Ammar, 143—9. The author criticized these narrations for being uncorrobo- 
rated from specific transmitters Muslim had cited [tafarrud). These impugned narrations 
are not found among al-Daraqutnl's criticisms. 

'"^ Ibn 'Ammar, 150—1. Here the critic was unwittingly parroting an earlier criticism 
made by al-Daraqutnl. 

103 Yor the first criticism, see Jalal al-Dm al-SuyutI, al-Mukat al-badi'dt 'aid al-Mawdudt, 
ed. 'Amir Ahmad Haydar ([Beirut]: Dar al-Janan, 1411/1991), 47; Sahih al-Bukhdn: 
kitdb al-tibb, bdb shurut al-ruqyd bi-Fdtihat al-kitdb; cf ibid., kitdb al-ijdra, bdb 16, for another 
narration. For the second criticism, see al-SuyutI, al-Nukat al-badTdt, 212. Here al-SuyutI 
states that al-'iraqi had found an authentic counterpart narration for this report. This 
narration does not appear in any extant recensions of al-Bukharl's collection, but Ibn 
al-jawzl found it in Hammad b. Shakir's lost recension. For the third, see al-SuyutI, 
al-Nukat al-badTdt, 262; Sahih Muslim: kitdb al-janna iva sijat na'imihd, bdb 13; cf Ibn Hajar, 
Tahdhib al-tahdhib, 1:333—4. 

"" Ibn al-Salah, Siydnat Sahih Muslim, 159—60. For an example of al-Mazarl's noting 
broken narrations, see al-Mazari, 1:283. 



294 CHAPTER SEVEN 

complete version in the same book. He felt that this was sufficiently well 
known among the ahl al-hadith, just as he narrated from a group of weak 
transmitters relying on the fact that these hadlths were known through 
reliable transmitters '"" 

Here he thus relies on the argument that, although certain narrations 
of hadlths are problematic, Muslim allowed them as auxiliary reports 
only because he assumed his readers knew that correct versions existed 
elsewhere. Ibn al-Salah makes the same case for the incomplete isndds 
found in al-Bukharl's Sahih}^^^' He further defends al-Bukhari and Mus- 
lim against one of Daraqutnl's criticisms, noting that, like almost all 
of al-Daraqutni's critiques, "it is a criticism of their [al-Bukhari's and 
Muslim's] isnad's, and does not remove the texts [matn) of their hadlths 
from the realm of authenticity [hayyiz al-sihhd)."^'^^ One narration of a 
Prophetic tradition might be flawed, but sound ones existed elsewhere 
that established the reliability of the Prophet's statement. 

Ibn al-Salah 's Egyptian contemporary, Rashid al-Din al- 'Attar (d. 662/ 
1264), also mounted a defense of Muslim against al-Mazari's criti- 
cisms. His Kitdb ghurar al-fawd'id al-majmu'a fi bay an ma waqa'a fi Sahih 
Muslim min al-ahddith al-maqtu'a deals with seventy criticized narrations 
from Muslim's work, which he calls "exceptions to [Muslim's] standard 
method {rasm}." The author's chief concern is that such criticisms pose 
a threat to the function of Muslim's book as a measure of authenticity 
and authoritative reference. He states: 

Perhaps someone looking at [al-Mazarl's] book who does not have a great 
concern for hadlth nor any knowledge of how to collect their different 
narrations, might think that [these criticized hadlths] were among those 
hadlths that lack unbroken chains back to the Prophet, and that one can 
thus not use them as proof texts. 

He has seen many people with this impression, which he hopes to coun- 
ter by proving that all these hadlths in fact possess complete unarfs.'"" 
The most categorical defense of Muslim's Sahth against al-Daraqutnl 
came at the hands of Ibn al-Salah's follower, al-Nawawi, whose com- 
mentary on Muslim's work includes detailed responses to all the 



'"^ Ibn al-Salah, Siydnat Sahih Muslim, 82; al-Nawawi, Shark Sahih Muslim, 1:125. 

'"*' Ibn al-Salah, Siydnat Sahih Muslim, 83. 

'"' Ihn al-SalsLh, Siydnat Sahih Muslim, 177. 

'"" Rashid al-Dlii Yahya b. 'All al-MisrI al-'Attar, Kitdb Ghurar al-fawd'id al-mqjmu'a 
fi baydn md waqa'a ft Sahih Muslim min al-ahddith al-maqtu'a, ed. Salah al-AmIn Ballal 
(Riyadh: Maktabat al-Rushd, 1421/2000), 140-1. 



THE PRINCIPLE OF CHARITY 295 

impugned narrations. While he and Ibn al-Salah had labored to exempt 
al-Bukharl and Muslim from conventions of hadlth criticism that occa- 
sionally proved too demanding for the Sahihayn, al-Nawawi also knew 
how to use these rules to the canon's advantage. He defends Muslim 
against the most frequent flaw identified by al-Daraqutnl, inappropri- 
ate Addition, by referring to the consensus arrived at by al-Khatib al- 
BaghdadT and the majority of legal theorists (but not by most hadith 
scholars): any Addition by a trustworthy transmitter is acceptable.'"^ 
Al-Nawawl thus neutralizes al-Daraqutnl's criticisms by demonstrat- 
ing that his methods were far harsher than the accepted norm. He 
therefore warns his readers that al-Daraqutni's methods are "the defi- 
cient principles of some hadith scholars, contrary to the vast majority 
{al-jumhur) of legal scholars and theorists [ahl al-jiqh wa al-usul), so do 
not be swayed [by them]!""" Throughout the text of his commentary 
on Muslim's work, al-NawawI undertakes a case-by-case rebuttal of 
al-Daraqutni's criticisms.'" 

Ibn Hajar mirrored al-Nawawi's defense of Muslim in the sizable 
introductory volume to his mammoth commentary on Sahih al-Bukhari, 
the Fath al-ban. There Ibn Hajar includes a massive chapter entitled 
"Putting forth the hadiths that the hadith master of his age, Abu al- 
Hasan al-Daraqutni, and others, criticized . . . and furnishing what is 
available as a rebuttal." This section includes a case-by-case response 
to al-Daraqutnl's criticisms. Like Ibn al-Salah and al-NawawI, he argues 
that many of the problematic narrations in al-Bukhari's collection 
come from his laxer auxiliary narrations. But while al-NawawT excuses 
Muslim's inclusion of reports with inappropriate Addition by referring to 
the conventions of legal theorists, Ibn Hajar relies more on al-Bukharl's 
peerless expertise. Al-Bukhari possessed an unrivaled mastery of the 
hadith sciences, Ibn Hajar argues, and judged the reliability of each 
hadith based on the circumstances {qard'in) of that case. One can thus 
not hold him accountable to the judgment of lesser scholars or the 
rigid rules they employed."^ 



'™ Al-Nawawl, Sharh Sahih Muslim, 1:145; cf. al-Khatib, al-Kifiya, 2:516, 538. 

"" Ibn Hajar, Hady al-sdri, 501 (quoted from al-Nawawl's lost commentary on al- 
Bukharl). 

'" See for example, al-NawawI, Sharh Sahih Muslim, 1:190; 2:334 ff The Dar al-Qalam 
edition of al-Nawawl's Sharh contains an appendix with all al-Daraqutnl's criticisms 
and al-Nawawl's responses. 

"2 Ibn Hajar, Hady al-sdri; 503, 543. 



296 CHAPTER SEVEN 

Yet Ibn al-Salah, al-Nawawi and Ibn Hajar all found themselves 
forced to admit that several of al-Daraqutni's criticisms were undeniably 
correct."^ Because al-Daraqutni was such a hugely respected figure in 
the pantheon of hadith scholars, and because he played such a forma- 
tive role in the early study of the Sahihayn, Ibn al-Salah, al-NawawI and 
Ibn Hajar exempted the material that he criticized from the claim of 
consensus on the two works' absolute authenticity. Even if one could 
successfully rebut some of al-Daraqutni's criticisms, one could hardly 
claim consensus on those elements of the Sahihayn rejected by a scholar 
of his caliber. These exceptions fell outside the pale of ijmd' and thus 
did not yield epistemological certainty"* 

Interestingly, Ibn al-Salah 's exemption of material criticized by master 
hadith scholars from the umma's consensus actually provided a window 
for selectively admitting the existence of problems in the Sahihayn}^'' 
Because earlier pillars of the hadith tradition such as al-Daraqutnl and 
Ibn 'Abd al-Barr had criticized Muslim's narration negating the voiced 
basmala, Ibn al-Salah, al-'iraqi and other later Shafi'ls were able to 
champion their madhhab's stance on this issue by openly discussing the 
report as a textbook example of a flaw {'ilia) in the text of a hadith. 

Other reports also contained errors beyond defense, sometimes in the 
content of the hadith. Al-Nawawl therefore acknowledged that one of 
Muslim's hadlths saying that the first chapter of the Qur'an revealed to 
the Prophet was sural al-Mudalhlhir (no. 74) is "weak, even false (bdtil), and 
the correct [position] is that the absolute first to be revealed was 'Read, 
in the name of your Lord who created ... (sural al-'Alaq, no. 96).'""' In 
the case of al-Bukharl's hadith that describes Adam incredulously as 
having been "sixty arms tall," Ibn Hajar admitted that "nothing has 



"' Al-Nawawi, Shark Sahih Muslim, 1:128; Ibn Hajar, al-Mukat 'ala kitab Ibn al-Salah, 
118. 

"' Ibn al-Salah, Siydnat Sahih Muslim, 87; Ibn Hajar, Hady al-sdri, 501; idem, al-Mukat 
'aid kitdb Ibn al-Salah, 1 16; Brown, "Criticism of the Proto-Hadith Canon," 2. 

"' Ibn al-Salah, Siydnat Sahih Muslim, 87. 

"*' Al-Nawawl, Sharh Sahih Muslim, 2:565-6; Sahih Muslim: kitdb al-imdn, bdb bad' al- 
wahy, hadith of AbO Salama. This criticized narration comes after numerous other 
narrations that confirm that the beginning of sdrat al-'Alaq was indeed the first part of 
the Qur an revealed. Muslim's inclusion of the minority report stems from the impartial 
methodology he followed in compiling his Sahih. Just as he often included reports with 
conflicting legal implications provided that all their isndds were sound, so here does he 
include a historical report differing from other hadlths. 



THE PRINCIPLE OF CHARITY 297 

yet appeared to me that removes this problematic issue (ishkal)."^" 
Such criticisms, however, were few among staunch proponents of the 
canon and occurred against the backdrop of these scholars' devotion 
to defending the Sahihayn canonical culture. 

In the wake of Ibn al-Salah's and al-Nawawl's campaign for strength- 
ening the Sahihayn canonical culture, many hadith scholars devoted 
works to defending al-Bukharl and Muslim from criticism or trying to 
clarify problematic material in their works. Ibn Kathir wrote a whole 
book refuting the two hadiths, al-Bukharl's story of the Prophet seem- 
ingly making his miraculous voyage to Jerusalem before the start of his 
prophetic career and Muslim's report of the Prophet marrying Umm 
Hablba (see Chapter Eight for more discussion), that Ibn Hazm had 
criticized as incontrovertibly forged."" Al-'Iraqi finished the rough draft 
of a small book detailing all the impugned narrations in the Sahihayn and 
providing defenses for them, but he never completed the work."^ His 
son. Wall al-Din Abu Zur'a Ahmad b. 'Abd al-Rahman (d. 826/1423) 
also wrote a book called al-Bayan wa al-tawdih li-man khurrija lahufi al- 
Sahih wa qad mussa bi-darb min al-tajnh (Elucidation and Clarification of 
those who Appear in the Sahih and had been Tainted by Some Sort 
of Criticism).'^" Jalal al-Din 'Abd al-Rahman b. 'Umar al-BulqIni (d. 
824/1421), the son of Shafi'i hadith scholar of Cairo, al-BulqInl, also 
wrote a book called al-Ijhdm li-mdfi al-Bukhdn min al-awhdm (Explicating 
the Errors found in al-Bukhari).'^' Ahmad b. Ibrahim Sibt al-'AjamI 
al-Halabi (d. 884/1479-80), another Shafi'l, composed a book based on 
Ibn Hajar's Fath called al-Tawdih li'l-awhdm al-wdqi'afi al-Sahih (Clarifying 
the Errors Occurring in the Sahih). He also had a book on ambiguities 



m 



Sahih Muslim {Mubhamat Muslim). 



122 



'" Ibn Hajar, Fath al-bdn, 6:452-3. Sahih al-Bukhdn: kitab ahddith al-anbiyd', bab 1; 
Fath # 3326; khalaqa Alldh Adam wa tuluhu situna dhird'"" . . .fa kull man yadkhulu al-janna 
'ala surat Adam, fa-lam yazal al-khalq yanqusu hattd al-dn." 

"" Ibn al-WazIr, Tanqih al-an^dr, 54; cf. Ibn Hazm, [Two Hadiths from the Sahihayn], 
28b-29a. 

'" Ibn Hajar, al-Mukat 'ala kitab Ibn al-Saldh, 116. 

"" Al-Makkl, Lahz al-lihdi, 5:186. This book has been published as Abu Zur'a Ahmad 
al-'iraqi, al-Baydn wa al-tawdih li-man ukhrija lahufi al-sahih wa mussa bi-darb min al-tajrih, 
ed. Kamal Yusuf Hut (Beirut: Dar al-Jinan, 1410/1990). 

'^' Al-Sakhawl, al-Daw' al-ldmi' li-ahl al-qarn al-tasi', 12 vols, in 6 (Beirut: Dar Maktabat 
al-Hayat, [1966]), 4:109. This book has survived in manuscript form, see Qa'imat al- 
makhtdtdt al-'arabiyya al-musawwara bi-mikrdjilm min al-jumhuriyya al-'arabiyya al-yamaniyya 
(Cairo: Matba'at Dar al-Kutub, 1967), # 86. 

'^^ Al-SakhawI, al-Daw' al-ldmi', 1:199. This book on al-Bukharl may be the work 



298 CHAPTER SEVEN 

Conclusion 

The pre-canonical history of al-Bukharl, Muslim and their masterpieces 
contained elements that did not accord with the shape and station of the 
Sahihayn canon. As the canon emerged at the dawn of the fifth /eleventh 
century, the environment of hadith study in Baghdad transformed into 
a canonical culture that required a charitable reading of the text of 
the canon. With al-Khatib al-Baghdadi's biographies of al-Bukhari and 
Muslim, we see the contours of this culture take shape and emphasize 
themes that reconcile the canon with history. Al-Bukhari, Muslim and 
their Sahths are placed at the acme of the hadith tradition, erasing 
initial objections to the sahih movement. The Sahihayn are shown as the 
products of almost superhuman scholarly and pietistic effort. Al-Bukhari 
is vindicated in the scandal of the Qur'anic lafz, an early advocate of 
orthodoxy against a jealous adversary. As both a persona and a book, 
al-Bukhari is ranked above Muslim. Nonetheless, the twin components 
of the Sahihayn form a complimentary and conjoined pair The con- 
struction of this canonical culture, however, did not suffice. Further 
interpretive and editorial efforts were required to defend the Sahihoyn 
canon against the enduring dangers of its pre-canonical past. 

The personas of al-Bukhari and Muslim were not the only element 
of the canon that required charity. Al-Bukhari and Muslim were only 
two figures in the wider world of Sunni hadith scholarship, a tradition 
characterized by a relative diversity of methodologies both before and 
after the formation of the canon. With the systemization of the Sunni 
hadith sciences between the writings of al-Hakim, al-Khatib and Ibn 
al-Salah, the potential for inconsistency between this tradition and 
the methods of al-Bukhari and Muslim became pronounced. On two 
specific topics, tadlis and the criticism of transmitters, defenders of the 
canonical culture would have to extend full charity to the Sahihayn in 
order to reconcile the institution of the canon and the conventions of 
hadith study. Proponents of the canonical culture also found it neces- 
sary to address earlier criticisms that had resulted from inconsistencies 
between al-Bukharl's and Muslim's methods and those of other major 
hadith scholars. Again, the Principle of Charity constituted an important 
tool in the arsenals of Ibn al-Salah, al-NawawI and Ibn Hajar 



of the author published as al-Tawdih li-mubhamdt al-Jdmi' al-saluh, ed. Abu al-Mundhir 
al-Naqqash Ashraf Salah 'All (Beirut: Dar al-Kutub al-'Ilmiyya, 1422/2001), which 
does not deal with supposed errors occurring in the Sahih. 



THE PRINCIPLE OF CHARITY 299 

In the maintenance of the Sahihayn canonical culture, we see a direct 
correspondence between the canonicity of these texts and the amount 
of charity they are afforded. '^^ In all aspects of the Sahihayn canonical 
culture, it was Ibn al-Salah and his follower al-NawawI who played the 
most prominent and creative roles. This should come as no surprise, for 
Ibn al-Salah had proven the most fervent proponent of their canonical 
functions. He had taken dramatic steps in declaring the infallibility of 
the Sahihayn, and produced the boldest and most influential argument 
for institutionalizing al-Bukharl's and Muslim's collections as authorita- 
tive references that could replace the arcane critical methodology of 
hadith scholars. Al-Nawawi inherited his master's agenda, replicating 
his arguments and reinforcing the canonical edifice. 



Halbertal, 29. 



CHAPTER EIGHT 

THE CANON AND CRITICISM: ICONOCLASM AND 

THE REJECTION OF CANONICAL CULTURE FROM 

IBN AL-SALAH to the MODERN SALAFi MOVEMENT 



Introduction 

Discussing the standing of the Sahihayn, Goldziher concluded that ven- 
eration for them "never went so far as to cause free criticism of the 
sayings and remarks incorporated in these collections to be considered 
impermissible or unseemly. . . .'" He insightfully observed that "venera- 
tion was directed at this canonical work [of al-Bukharl] as a whole but 
not to its individual lines and paragraphs."^ In his Rethinking Tradition 
in Modern Islamic Thought, Daniel Brown concurs. He states that in the 
"classical" period there was a great deal of leeway for the criticism of 
the canonical collections.^ As we have seen, Goldziher 's and Brown's 
assessments accurately describe the pre-canonical period as well as the 
continued criticism of the two books even after their canonization. 
They do not, however, recognize the important change that occurred 
in the dynamic of the canon and criticism in the early modern and 
modern periods. 

Especially in recent times, criticisms of the SaMhoyn canon have met 
with remarkable hostility. Mohammad Abd al-Rauf has recognized the 
dramatic change in the reaction to criticism, but identifies it as the result 
of Ibn al-Salah's buttressing the canonical culture in the seventh/thir- 
teenth century. He asserts that in the wake of Ibn al-Salah's writings, 

"no more criticism could be tolerated "* Although Ibn al-Salah and 

al-NawawI certainly did demand a charitable reading of the Sahihayn, 
their contributions to the canonical culture marked neither a morato- 
rium on criticism nor an actual end to it. 



' Goldziher, 236-7. 
2 Goldziher, 247. 

' Daniel Brown, Rethinking Tradition in Modern Islamic Thought (Cambridge: Cambridge 
University Press, 1996), 111. 

' Abd al-Rauf, "Hadtth Literature," 285. 



THE CANON AND CRITICISM 301 

Indeed, criticism of the Sahihayn continued in force well after Ibn 
al-Salah's and al-Nawawi's seminal careers. In the century after their 
deaths, a number of hadith scholars rejected the canonical culture 
built around al-Bukhari and Muslim. These objections gave voice to 
the long-standing tension between the drive for institutional security 
that had transformed the Sahihayn into authoritative references and 
the iconoclastic strain in hadith scholarship that remained steadfastly 
focused on the critical evaluation of individual reports. 

It was the emergence of the Salafi reform movement in the eighteenth 
century that brought this simmering tension to a boU. Its revitalized 
focus on the critical study of hadith, its prioritization of hadith above 
the hermeneutic traditions of the madhhabs and its willingness to ques- 
tion ijmd' attacked the very foundation of the hadith canon. Two of 
its premier hadith scholars, Muhammad b. Isma'll al-AmIr al-San'anI 
(d. 1768) and Muhammad Nasir al-Din al-AlbanI (d. 1999), exemplified 
this critical rejection of the Sahihayn canonical culture. For early modern 
and modern advocates of the traditional schools of law or reformists 
concerned with defending an increasingly beleaguered Islamic civiliza- 
tion, these criticisms of the Sahihayn came to represent a rejection of 
the institutions that had authorized the canon and that it served. The 
ferocity with which proponents of the madhhab?, have attacked al-Albanl's 
criticism of the Scihihoyn in particular reflects both the canon's role as 
a symbol of the classical Islamic institutional tradition and the canon's 
important function in scholarly culture. 



Rejection of the Canonical Culture: Criticism after Ibn al-Salah 

The Sahihayn canonical culture existed to safeguard the institution of 
the canon and the important functions it served in the Sunni scholarly 
tradition. The charity extended to the two works in order to overcome 
the tension between the methods of their authors and the independent 
rules of hadith criticism reflected the needs of non-hadlth specialists, 
who relied on the S^hihoyn as a measure of authenticity and authorita- 
tive reference. The Sahihayn canon was supposed to provide these jurists 
with the authority of the Prophet's authentic sunna in a manageable 
form, sifted by those two scholars who had come to epitomize the criti- 
cal rigor of the hadith tradition and approved by the umma's infallible 
consensus. 



302 CHAPTER EIGHT 

The authoritative edifice of the canon, however, was a construct. It 
was the creation of scholars struggling to provide the Islamic intellec- 
tual tradition with the secure institutions it required to meet the needs 
of the wider Sunni community. Major late architects of the Sunni 
hadith tradition, such as Ibn Hajar, embraced the canonical culture 
shaped by al-Khatib al-Baghdadi and elaborated by Ibn al-Salah and 
al-Nawawi. Yet at its heart, the hadith scholar's study of the Prophet's 
legacy remained an austere cult of authenticity that acknowledged no 
source of authority beyond the chain of transmission that connected 
Muslims to the charisma of their Prophet. The culture of the hadith 
scholar thus nurtured an iconoclastic strain that did not easily suffer 
the elaboration of authoritative institutions above and beyond the isnad. 
Just as many hadith scholars had rejected Ibn al-Salah's perceived call 
to rely on sahih books and end the critical evaluation of hadlths, so did 
many refuse the demand to grant the Sahihayn an iconic status above 
the conventions of hadith criticism. While scholars like al-Dhahabi 
and Ibn Hajar generally accepted the cases for charity advanced by 
Ibn al-Salah and al-NawawI, other hadith scholars considered them 
baseless assertions with no grounding in the principles of the hadith 
sciences. Criticism thus continued despite the strength of the Sahihayn 
canonical culture. 

Although the great Syro-Egyptian hadith master Ibn Daqlq al-'Id 
(d. 702/1302) had embraced the Sahihayn canonical culture on the 
issue of exempting al-Bukharl's and Muslim's transmitters from criti- 
cism, he exhibited skepticism over al-Nawawl's argument on tadlis. The 
notion of distinguishing the Scihihcyn from other books in this case, he 
explained, was baseless. Such a charitable distinction must entail one of 
two untenable claims. Either we are sure that al-Bukharl and Muslim 
made certain that every instance of possible tadlis was actually a direct 
transmission isamd') — which we cannot know — or the consensus {ymd') 
of the umma guarantees that no such error occurred. Yet this again 
depends on the impossible task of scholars having ascertained that 
al-Bukharl and Muslim were entirely thorough in eliminating breaks 
in their isndds." 



'' Ibn Hajar, al-Nukat 'ala kitab Ibn al-Salah, 255. 



THE CANON AND CRITICISM 303 

Another Shafi'i contemporary of Ibn Daqiq in Cairo, Sadr al-Din 
Abu 'Abdallah Muhammad Ibn al-Murahhal (d. 716/1317)/' seconded 
this skepticism towards Ibn al-Salah's and al-Nawawi's exemption of 
al-Bukharl and Muslim from the rules governing tadlis. In his Kitdb al- 
tK5a/" (apparently lost) he explained: 

Indeed, in this exemption {istithnd') something makes my soul uneasy. For 
it is a claim without proof, especially since we have found that many of 
the hadlth masters [huffaz) have criticized hadlths found in the Sahihayn 
or one of them for the tadlis of their narrators.' 

The Cairene Hanafi Ibn Abl al-Wafa"s rejection of the Sahihayn canoni- 
cal culture moves beyond such skepticism, however, entering the realm 
of unmitigated contempt. He argues that the notion of al-Bukhari's and 
Muslim's transmitters having "passed the test" is preposterous. Muslim, 
he explains, had narrated from demonstrably weak transmitters. Ibn 
Abl al-Wafa' also rejects Ibn al-Salah's argument that one should not 
hold al-Bukharl's and Muslim's auxiliary narrations to the same stan- 
dard as their primary ones. Such narrations are supposed to explain 
the status [hdl) of a hadlth, and if Muslim's collection was supposed 
to include only authentic reports, what do weak auxiliary reports say 
about the condition of his main hadiths?'' Accepting all instances of a 
mudallis narrating via "from/ on the authority of ( (zn)" if they occur in 
the Sahihayn but not in other works is similarly baseless and represents 
nothing more than vain posturing [tajawwuh)? 

Ibn Abl al-Wafa' then administers his coup de grace to the canonical 
culture, detailing a number of hadlths from the SdhWciyn whose contents 
render them unquestionably false. He mentions Muslim's hadlth that 
"God most great created the earth (al-turba) on Saturday...," which 
contradicts the Qur'anic statement that the world had been created 
in six days (Saturday being the seventh).'" He brings up a hadlth from 
Sahth al-Bukhdn that seems to recount the Prophet making his miracu- 
lous night journey to Jerusalem before he had even received his first 



'' Mahdl SalmasI, "Ibn al-Murahhaf" Daerat al-ma'dref-e bozorg~e esldmi, 4:200—1. 
' Ibn Hajar, al-Mukat 'aid kitdb Ibn al-Saldh, 255. 
" Ibn Abl al-Wafa', al-Jawdhir al-mudiyya, 4:566. 
' Ibn Abl al-Wafa', al-Jawdhir al-mudiyya, 4:566 S. 

'" Sahih Muslim: kitdb si/dt al-mundfiqin wa ahkdmihim, bdb ibtidd' al-khalq wa khalq Adam 
'alayhi al-saldm (1). 



304 CHAPTER EIGHT 

revelation." Finally, he notes Muslim's report of the Prophet promis- 
ing the newly converted AbQ Sufyan that he will marry his daughter, 
Umm Habiba, in the wake of the Muslim conquest of Mecca. '^ Ibn 
Abl al-Wafa' points out that scholars had agreed that the Prophet had 
already married her years earlier. The Hanafi dismisses the various 
efforts to explain this evident contradiction as vain posturing (tajawwuh) 
and "futile responses [ajwiba ghayr ta'ila)."^^ 



Iconodasm and Institutional Security in Islamic Civilization: 
The Salafi Tradition 

Ibn al-Murahhal and Ibn Abi al-Wafa' rejected the Sahihayn canonical 
culture and instead evaluated material from the two books according 
to the critical conventions of the hadith tradition. Yet their criticisms 
met with no obvious reprimand. The only condemnation of criticizing 
the Sahihayn came from Yusuf b. Musa al-Malati (d. 803/1400-1), a 
controversial Hanafi student of al-Mughultay His unusual and little- 
known statement that "anyone who looks critically [nazarafi) at \Sahih^ 
al-Bukhdri has become a heretic (tazandaqa)," however, was perceived as 
patently bizarre by contemporaries and later Muslim biographers. Ibn 
al-'Imad (d. 1089/1679) even listed it along with allowing the consump- 
tion of hashish as an example of al-Malati's deviant opinions.'* 

In the early modern period, the iconoclastic strain of hadith study 
evident in scholars like Ibn Abl al-Wafa' would resurface in the Salafi 
movement, with muhaddiths, like Muhammad b. Isma'll al-San'ani and 
later Muhammad Nasir al-Din al-Albani. In the turbulent struggle 
over defining Islam in the modern era, however, their rejections of 
the Sahihayn canonical culture would meet with fierce criticism from 
defenders of the classical Islamic institutions bound closely to the canon. 
For the first time, criticizing the So^luhoyn would become anathema for 
many scholars. 



" See Fath al-bdri, #'s 349, 3886, 7517; Sahih al~Bukhdri: kitdb al-salat, bdb 1, kitdh 
mandqib al-ansdr, bdb 41 and 42, kitdb al-tawhid, bdb 37. 

'^ Sahih Muslim: kitdb fadd'il al-sahdba, bdb fadd'il Abi Sujydn b. Harb (40). 
" Ibn Abl al-Wafa', al-Jawdhir al-mudiyya, 4:568-69. 
'* Ibn al-'Imad, Shadhardt al-dhahab, 7:40. 



THE CANON AND CRITICISM 305 

a. Revival and Reform in the Early Modern and Modern Periods 

Since the eighteenth century, movements of revival and reform arising 
as responses to both internal stimuli and the pervasive influence of 
Western civilization have dominated Islamic intellectual history. These 
movements have all faced the problem of determining the proper role of 
hadlth in defining Islamic law, ritual and worldview in ongoing debates 
about the shape that Islam should take in the modern world. Islamic 
Modernists such as the Indian Sir Sayyid Ahmad Khan (d. 1898) 
have dismissed the classical tradition of hadith study as incapable of 
guaranteeing an authentic vision of the Prophet's sunna. They have thus 
rejected the role of Prophetic traditions as a central tool for interpreting 
Islam. Diametrically opposed to these modernists are those scholars one 
might refer to as Madhhab Traditionalists, who believe that the 
classical Islamic institutions of the schools of law, theology and Sufi 
guilds offer the only correct path for understanding Islam. 

Lying in between these two camps on the spectrum of embracing or 
casting off the classical institutions of Islamic civilization are the diverse 
movements loosely grouped under the term 'Salafi,' or those willing 
to reevaluate the institutions of medieval Islam in order to revive the 
pure Islam of the Prophet and the first righteous generations (salaf) 
of Muslims. Modernist Salafis such as the Muhammad 'Abduh 
(d. 1905), Rashid Rida (d. 1935) and Shaykh Muhammad al-Ghazall 
(d. 1996) have eclectically utilized elements of the classical Islamic 
tradition that they felt could aid in reviving this original greatness. 
'Abduh thus attempted to revive the rationalism of the Mu'tazila, and 
al-Ghazall mined the various interpretive methods of the different 
Sunni madhhabs to produce a vision of Islam that was traditionally 
authentic but more compatible with modernity. Both tried to curb 
those parts of the hadith tradition that clashed with modernity by 
making hadith more subservient to the over-arching principles of the 
Qjur'an and the methods of Muslim legal theorists.''^ Tied to this group 
are the Traditionalist Salafis, who invert this equation: like other 
reformists, they seek to rejuvenate the Muslim community by reviving 
the primordial greatness of Islam, yet they have sought to recreate the 
Prophet's sunna by making the classical study of hadith and the ways 
of the early community paramount. 



''' See Muhammad al-Ghazali, al-Sunna al-nabawiyya bayn ahl al-fiqh wa ahl al-hadith, 
11th ed, {Cairo: Dar al-Shuruq, 1996). 



306 CHAPTER EIGHT 

For all these reformist strains, the Sahihayn have served as a power- 
ful symbol in debates over the proper role of hadith in modern times. 
Islamic Modernists like the Egyptian Mahmtid Abu Rayya have used 
al-Bukharl's and Muslim's venerated status to severely criticize the 
classical hadith tradition by demonstrating how even the Sahihayn con- 
tain forged reports."' Daniel Brown describes how Modernist "deniers 
of hadith have especially delighted in exposing traditions in the sahih 
collections, especially Bukharl and Muslim, which they take to be 
vulgar, absurd, theologically objectionable, or morally repugnant."" 
Conversely, Muhammad al-Ghazall employed the canon to assist him 
in boldly reinterpreting the classical Islamic tradition to prove that 
women can hold high public office and to affirm matters of dogma 
such as the punishment of the grave. Unlike Abu Rayya, he vener- 
ated al-Bukharl and Muslim and so used their decisions not to include 
certain problematic hadlths on these issues to neutralize the reports' 
efficacy as proof texts.'" 

Because we are concerned with the tension between the Sahihayn 
canon and the methods of hadith criticism indigenous to the Islamic 
tradition, we will focus only on the treatment of the canon by Tradi- 
tionalist Salafis and Madhhab Traditionalists. The other two reformist 
strains, the Islamic Modernists and Modernist Salafis, have been pri- 
marily concerned with reacting to the West. Sir Sayyid Ahmad Khan's 
dismissal of the classical hadith tradition resulted from his encounters 
with the Orientalist William Muir, who questioned the authenticity 
of the hadith corpus.'^ Muhammad 'Abduh's and Jam al al-Din al- 
Afghanl's intellectual output and political activism were responses to 
European political and cultural encroachment. Muhammad al-Ghazall's 
reevaluation of the proper role of women in Islamic society stemmed 
in part from witnessing the effective leadership of Margaret Thatcher^" 



"' Daniel Brown, Rethinking Tradition, 89. 

" Daniel Brown, Rethinking Tradition, 95. 

'" Muhammad al-Gliazall, Turdthund al-fikri, 6th edition (Cairo: Dar al-Shuruq, 2003), 
180-2; idem, al-Sunna al-nabawiyya bayn ahl al-fiqh wa ahl al-hadith, 64, 

'" Daniel Brown, Rethinking Tradition, 33—6. 

™ Haifaa G. Khalafallah, "Rethinking Islamic Law: Genesis and Evolution in the 
Islamic Legal Method and Structures. The Case of a 20th Century Alim's Journey into 
his Legal Traditions: Muhammad al-Ghazali (1917-1996)," (Ph.D. diss., Georgetown 
University, 2000), 89; idem, "Muslim Women: Public Authority, Scriptures and 'Islamic 
Law,' " in Beyond the Exotic: Women's Histories in Islamic Societies, ed. Amira Sonbol (Syracuse: 
Syracuse University Press, 2005), 41-2. 



THE CANON AND CRITICISM 307 

Skeptical of Prophetic reports that clash with rationalism or the 
expectations of modernity, but simultaneously eager to defend the 
hadlth as the repository of the Prophet's golden age, the apologetic 
thought of the Modernist Salafis has yielded no systematic approach 
to classical methods of authenticating hadlths.^' 

Although Western cultural, intellectual and political domination has 
cast its shadow over almost every corner of Muslim discourse in the 
modern period, the Traditionalist Salafis and the Madhhab Traditional- 
ists have been more concerned with each other's rhetoric than with the 
West. For Traditionalist Salafis, the umma's immediate challenge is 
the corruption of the Prophet's sunna wrought by excessive loyalty to 
the madhhab's, and the practices of popular religion. For the adherents of 
these traditions, the Salafi threat to classical Islamic institutions looms 
larger than Western encroachment. For both groups. Westernization 
and any Muslim contaminated by it are evils beyond the scope of 
dialogue. That they both dismiss any Muslim thinker who does not 
approach questions of Islam through the classical methodologies of 
jiqh or hadlth as "Occidentalists (mustaghribun)" or "imitators of the 
Orientalists" testifies to their shared indigenous focus. ^^ 

The varied strands that would make up the Traditionalist Salafi 
movement emerged from the various revival and reform movements 
that began dominating the intellectual landscape of Islamdom in the 
eighteenth century. The rise of the Wahhabi movement in Arabia, the 
Sokoto caliphate in West Africa and later the ahl-e hadith movement 
in India formed part of a broader network of Islamic movements. 
At their core lay the objective of renewing the bond with the pure 
origins of Islam though a rejuvenated interest in Prophetic hadlth. 
These reformists sought to break free from the historical accretions of 
Islamicate civilization, condemned as bid'a, and return Muslim societies 
to the radical monotheism itawhid) of the Prophet's original message. 
They often embraced the study of hadith as the most direct means to 



^' See Daniel Brown, Rethinking Tradition, 37; cf. Albert Hourani, Arabic Thought in the 
Liberal Age 1798-1939 (Cambridge: Cambridge University Press, 1983), 146 ff. 

^' Al-AlbanI, Mukhtasar Sahih al-Bukhari, 4 vols. (Riyadh: Maktabat al-Ma'arif, 
1422/2002), 2:8-9. Here al-AlbanI uses AbQ Rayya and Muhammad al-Ghazall as 
examples. Madhhab Traditionalists, however, generally use the term "imitator imuqallid)" 
only for Muslim scholars who do not follow the classical methodologies at all. Azhar 
shaykhi like al-Ghazall would probably fall outside this category. Instead, they would 
be dismissed as "preachers {dd'ija pi. du'dt)." 



308 CHAPTER EIGHT 

replicating the Prophet's ideal Medinan community and turning away 
from both the excesses of popular religion and the strict allegiance to 
specific schools of law.^^ 

As John VoU has identified, the shrine cities of Mecca and Medina 
served as a central junction in this massive revival phenomenon. With 
the move of prominent muhaddiths, such as the Cairene Ibn Hajar al- 
Haythaml (d. 974/1597) and MuUa 'All Qarl of Herat (d. 1014/1606) 
to the shrine cities, the Hijaz played host to a cadre of hadith- 
oriented scholars such as Ibrahim b. Hasan al-KuranI (d. 1101/1689), 
Muhammad Hayat al-Sindi (d. 1751) and 'Abdallah b. Salim al-Basri 
(d. 1722), who would exercise a tremendous influence on students from 
as far away as Malaysia.^* These circles produced preeminent activist 
scholars like Muhammad Ibn 'Abd al-Wahhab (d. 1792) and Shah Wall 
Allah al-DihlawI (d. 1762). While the thought and programs of Ibn 
'Abd al-Wahhab and Shah Wall Allah differed dramatically, they both 
exemplified a willingness to reconsider and break with the mainstream 
traditions of Sunni thought as it existed in the late medieval period.^' 
To different extents, both questioned taqltd, or the practice of following 
an existing madhhab without questioning its proofs, and made a direct 
consultation of Prophetic hadiths the ultimate determinant in interpret- 
ing the message of the Qjur'an.-'' 



'^'' Barbara Daly Metcalf, Islamic Revival in British India: Deoband, 1860—1900 (Princeton: 
Princeton University Press, 1982), 6; Basheer M. Nafi, "Tasawwuf and Reform in 
Pre-Modern Islamic Culture: in Search of Ibrahim al-KuranI," Die Welt des Islams 42, 
no. 3 (2002): 313. 

^* See John VoU, '"AbdaUah b. Salim al-Basri and 18th Century Hadith Scholarship," 
Die Welt des Islams 43, no. 3 (2002): 356-72; idem, "Foundations for Renewal and 
Reform: Islamic Movements in the Eighteenth and Nineteenth Centuries," in The 
Oxford History of Islam, ed. John Esposito (Oxford: Oxford University Press, 1999), 
509-47; idem, "Hadith Scholars and Tariqahs: an Ulama Group in the 18th century 
Haramayn and their Impact in the Islamic World," Journal of African and Asian Studies 
15 (1980): 264—73; Metcalf, Islamic Revival, 19; Muhammad Ishaq, India's Contribution to 
Hadith Literature (Dhaka: University of Dacca, 1955), 152 IF.; Daniel Brown, Rethinking 
Tradition, 23. 

^' See Ahmad Dallal, "The Origins and Objectives of Islamic Revivalist Thought: 
1750-1850," Journal of the American Oriental Society 113, no. 3 (1993): 341-59. 

^'^ DeLong Bas, Wahhabi Islam, 10-13. See Muhammad Ibn 'Abd al-Wahhab, 
"Fatawa wa masa'il al-imam al-shaykh Muhammad b. Abd al-Wahhab," in Mu'allafat 
al-shaykh al-imdm Muhammad b. Abd al-Wahhab, ed. Salih b. 'Abd al-Rahman al-Atram 
and Muhammad b. 'Abd al-Razzaq al-Duwaysh, vol. 3 (Riyadh: Jami'at Muhammad 
b. Su'ud al-Islamiyya, 1398/[1977]), 32. 



THE CANON AND CRITICISM 309 

This common interest in reviving the study of Prophetic hadiths 
and condemning excessive or blind adherence to an established school 
of law ran like a common thread through most of the eighteenth- 
century movements of revival and reform. To varying degrees, they 
all championed the practice of ijtihad, or turning anew to the Qur'an, 
the Prophet's sunna and the practices of the early community in order 
to find new answers to the legal or religious problems of the day In 
their focus on the early Muslim community and a return to its legacy 
at the expense of the later developments of Islamic orthodoxy, these 
movements were fundamentalist in character They telescoped religious 
history, demonstrating a willingness to sacrifice the elaborate develop- 
ments of classical Islamicate civilization in order to recapture the unity, 
purity and authenticity of the early community."' After the Prophet's 
life and the first few generations of his followers there were no more 
qualitative distinctions in history. In this, scholars like Shah Wall Allah 
and Ibn 'Abd al-Wahhab subverted the atavistic conservatism of the 
Sunni intellectual tradition, asserting that devout and competent modern 
Muslims were every bit as capable of understanding the message of 
Islam as the founders of the madhhabs had been.^" 



b. Traditionalist Salafis in the Middle East 

The loosely grouped Traditionalist Salafi movement in the Middle 
East developed in four dispensations. The earliest, most persistent and 
most politically active was founded by Ibn 'Abd al-Wahhab in the mid- 
eighteenth century in central Arabia, expanding through its alliance 
with the Saud family of Najd and eventually becoming the dominant 
religious movement on the Arabian Peninsula. A second Salafi strain 
appeared in the Yemeni city of San'a', with the iconoclastic hadlth 
scholar Muhammad b. Isma'il al-San'ani (d. 1768) and two generations 
later with the reformist thinker and hadlth scholar Muhammad b. 'All 
al-Shawkani (d. 1839).^^ A third school developed in Damascus in the 
second half of the nineteenth century around revivalist scholars such 



" Rudolph Peters, "Idjtihad and Taqlld in 18th and 19th Century Islam," Die Welt 
des hlams 20, no. 3-4 (1980): 131-2. 

^° Dallal, "The Origins and Objectives of Islamic Revivalist Thought," 347; Peters, 
"Idjtihad and Taqlld," 139; Daniel Brown, Rethinking Tradition, 23. 

^' Nafi, "Tasawwuf and Reform in Pre-Modern Islamic Culture," 351. 



310 C HAPTER EIGHT 

as 'Abd al-Razzaq al-Baytar (d. 1917) and his students, Jamal al-Din 
al-Q_asimi (d. 1914) and Tahir al-Jaza'iri (d. 1920).^° Finally, an influ- 
ential Salafi school formed in Baghdad through the Hanbali revival 
led by the Alusi family: Mahmud al-Alusi (d. 1853), Nu'man al-AlusT 
(d. 1899) and Mahmud Shukri al-AlusI (d. 1924).^' 

These three schools were distinct from the Wahhabl movement, with 
both the Baghdad and Damascene schools espousing a more tolerant 
approach to classical Sufism. Indeed, their ideological fraternity with 
the Wahhabis often proved dangerous for Salafis in Damascus and 
Baghdad. Their opponents would often accuse them of being Wahhabis, 
and the Ottoman state held them under suspicion of being a Wahhabl 
fifth column within the empire.'^ Al-San'anI was a contemporary of 
Ibn 'Abd al-Wahhab himself, and despite their similar Salafi leanings, 
the Wahhabl proclivity towards declaring other Muslims unbelievers 
[takfir) detracted from al-San'ani's initial positive impression of the 
movement. He wrote in verse: 

I recant that which I said about the Najdl (Ibn 'Abd al-Wahhab), 

for things have come to me from him on which I differ 

I thought well of him and said, 'Could it be, could it be, 

'That we have found someone to seek God's path and His slaves 

deliver?' 

But some of his letters have come to me from his own hand. 

Declaring all the world's peoples disbelievers intentionally. 

In this he has contrived all his proofs and. 

You see them weak as a spider's web when examined critically. ^^ 

Nonetheless, the Damascene, Baghdadi, Yemeni and Wahhabl dispen- 
sations of the Salafi phenomenon influenced one another Scholars 
like al-QasimI and Mahmud Shukri al-Alusi corresponded, and, more 



'" See David Dean Commins, "The Salafi Islamic Reform Movement in Damascus, 
1885—1914: Religious Intellectuals, Politics and Social Change in Late Ottoman 
Syria," (PhD diss., University of Michigan, 1985); Itzchak Weisman, "Between Sufi 
Reformism and Modernist Rationalism: A Reappraisal of the Origins of the Salafiyya 
from the Damascene Angle," Die Welt des Mams 41, no. 2 (2001): 206-236; W. Ende, 
"Salafiyya," EF . 

" Daniel Brown, Rethinking Tradition, 30. 

'- Halah Fattah, " 'Wahhabi' Influences, Salafi Responses: Shaykh Mahmud Shukri 
and the Iraqi Salafi Movement, 1745-1930," JoMraa/ of Islamic Studies 14, no. 2 (2003): 
138-9, 146. 

'' Al-Qanubl, al-Sayf al-hddd, 40. Supporters of Ibn 'Abd al-Wahhab dispute al- 
San'anl's authorship. 



THE CANON AND CRITICISM 311 

recently, al-Albani used Ibn 'Abd al-Wahhab's books in his lessons.'* 
Although the Traditionalist Salafi school differed significantly from the 
apologetics and Euro-centered political activism of Modernist Salafis 
like 'Abduh, the trends nonetheless informed one another^' 'Abduh's 
disciple, Rashid Rida, considered al-San'anI to be the renewer [mujaddid) 
of the twelfth Islamic century ''' Al-AlbanI, in turn, started down the 
path of reformist thinking when he came across an article by Rida in 
an issue of 'Abduh and Rida's al-Mandr journal/'' 

Like the other reform movements, the Traditionalist Salafis have 
aimed at reviving Islam's original purity and greatness by clearing away 
the dross of later cultural accretions. Unlike Modernists, however, they 
have focused literally on reviving the Prophet's sunna as expressed in 
the hadlth corpus. The primary culprits in distancing the Muslim com- 
munity from the authentic sunna have been "excessive loyalty to the 
madhhabs (al-ta'assub al-madhhabi)," an over-involvement in the science 
of speculative theology (kaldm), and popular religious practices such as 
those found among Sufi brotherhoods. What al-San'ani charmingly 
calls "the bid'a of madhhabism (al-tamadhhub)" causes Muslims to take 
the rulings of later scholars over the direct injunctions of the infallible 
Prophet.'" The speculative sciences have led Muslims away from the 
textual authenticity that gives Islam its purity. Popular religion and 
indulging in cultural accretions have led them to engage in bid'a that 
threatens Islam's essential monotheism (tawhid), such as visiting graves 
and seeking the miracle-working of local saints. 

To cure these ills. Traditionalist Salafis have not merely engaged in 
the study of hadlth, they have tried to cultivate its most critically rigor- 
ous spirit. Jamal al-Din al-Qasimi's Qawd'id al-tahdith minfunun mustalah 
al-hadith (The Principles of Regeneration from the Technical Science 
of Hadlth Study) and Tahir al-Jaza'iri's Tawjih al-nazar ild usul al-athar 
(Examining the Principles of Transmitted Reports) resemble classical 



'* See Jamal al-Din al-QasimI and Mahmnd Shulcrl al-AlusI, al-Rasd'il al-mutabddala 
baynjamdl al-Din al-Qasimi wa MahmUd Shukn al-Alusi, ed. Muhammad b. Nasir al-'AjamI 
(Beirut: Dar al-Basha'ir al-Islamiyya, 2001). For a sample of al-Albanl's curriculum, 
see Ibrahim Muhammad 'All, Muhammad Ndsir al-Din al-Albdni: muhaddith al-'asr wa ndsir 
al-sunna (Damascus: Dar al-Qalam, 1422/2001), 24. 

'^ Weisman, "Between Sufi Reformism and Modernist Rationalism," 235. 

"" J.J.G. Jansen, "Shawkani," EF. 

^' Al-AlbanI, "Tarjamat al-Shaykh al-Albdni — Nash'at al-ShaykhJi Dimashq," lecture by 
al-Albani from www.islamway.com, last accessed 6/3/2004. 

^^ Al-San'anl, Kitdb iqd^ al-fikra li-murdja'at al-fitra, ed. Muhammad Subhl b. Hasan 
al-Hallaq (Beirut: Dar Ibn Hazm, 1420/1999), 52. 



312 C HAPTER EIGHT 

manuals on the science of hadith such as Ibn al-Salah's Muqaddima, 
but urge Muslims to move beyond the simple acceptance of earlier 
opinions when evaluating the authenticity of a hadith.'^ Reviving the 
stringent spirit of al-Bukhari and Muslim, Salafis reject the lax use of 
weak hadlths in defining a Muslim's worldview. Al-Albani asks rhetori- 
cally: if we do not treat weak hadiths as such, what is the point of the 
science of hadith criticism? "For the heart of the issue," he explains, "is 
that it be highly probable, without serious doubt, that the Prophet (s) 
actually said that hadith so that we can depend on him in the Sharia, 
and attribute rulings to him."*" 

Their work is reminiscent of al-Bukharl's and Muslim's innovative 
pioneering of the sahih movement a millennium earlier, with their rejec- 
tion of weak hadlths and willingness to break with the laxer standards 
of Ibn Hanbal's greatest generation. It is thus no surprise that one of 
al-Albanl's students, the Yemeni Muqbil b. Hadi al-Wadi'l (d. 2001), 
compiled the first comprehensive sahih collection in almost a thousand 
years, a work designed to provide Muslims with all the authentic hadlths 
not included in the Sahihayn.*^ 

Salafis thus cast aside the institutions of classical Islam, relying on 
hadlths from the Prophet as the ultimate authoritative medium for 
transmitting the proper interpretation of the faith. According to the 
Salafi school, this obviates the chains of mystical and legal authority 
that allowed new practices such as Sufi rituals or fixed legal codes to 
enter Islam, merely masking departures from the authentic teachings 
of the Prophet. These were preserved in the authentic hadlths, which 
are accessible to any Muslim who can correctly navigate the volumes 
in which they were collected. The Qjur'an and the Prophet's sunna are 
the only criteria forjudging right from wrong. Partisanship or loyalty to 
a certain scholar or school should not blind Muslims from the ultimate 
authority of these two sources. 

The Traditionalist Salafi focus on hadith, reviving the ways of the 
early Muslim community and questioning the institutions of classical 
Islam that had arisen since, stemmed from the same iconoclastic strain 
as the Hanball reformer Ibn Taymiyya (d. 728/1328). Indeed, the 



'^ Daniel Brown, Rethinking Tradition, 32. 

'"' Al-Albanl, Sahih al-Targhib wa al-tarhib, 3 vols. (Riyadh: Maktabat al-Ma'arif, 
2000), 1:60. 

*' Abu 'Abd al-Rahnian Muqbil b. Hadi al-Wadi'l, al~Jdmi' al~sahih minima laysa fi 
al-Sahihayn, 6 vols. (Cairo: Dar al-Haramayn, 1416/1995). 



THE CANON AND CRITICISM 313 

Wahhabl, Baghdadi and Damascene schools originated in part from 
a renewed interest in Ibn Taymiyya's writings.*^ As Marshall Hodg- 
son explains, this iconoclastic strain was inherent in the hadlth-based 
Hanbali tradition: 

Hanbalism had never really been primarily a school of fiqh at all. It 
remained a comprehensive and essentially radical movement, which had 
elaborated its own fiqh in accordance with its own principles, but whose 
leaders were often unwilling to acknowledge the same kind of taqlid as 
provided the institutional security of the other schools and rejected the 
ijma' tradition of the living community on principle. "^^ 

As we shall see, the manner in which Ibn Taymiyya and his student 
Ibn Qayyim al-Jawziyya utilized the Sahihayn resurfaces in the Salafi 
approach to the canon. As we saw in Chapter Six, Ibn Taymiyya and 
Ibn al-Qayyim cunningly employed the Sahihayn as a rhetorical foil 
against their Ash'ari opponents. Ibn Taymiyya dramatically supported 
Ibn al-Salah's claim about the authenticity of the two works, asserting 
that "[Al-Bukharl and Muslim] do not agree on a hadlth except that 
it is authentic without a doubt" and compiling the most comprehen- 
sive list of scholars whom he claimed seconded this opinion.** For Ibn 
Taymiyya, the canon proved very useful, for al-Bukhari and Muslim 
provided the centerpiece for his efforts to shift the ultimate authority 
in determining the Prophet's true legacy towards hadith scholars as 
opposed to the later substantive law of the jurists.*'' 

Yet, just as he treated other aspects of Sunni scholarly production, 
Ibn Taymiyya refused to admit any iconic status for the Sahihayn. His 
subtle qualification that only material found in both al-Bukharl's and 
Muslim's works is without a doubt authentic allowed him to criticize 
freely reports found in only one. Unlike al-NawawI, his public ^toas 
announced that numerous reports in al-Bukharl's or Muslim's work 
were flawed. He openly criticized Muslim for approving the hadith of 
the earth being created on Saturday and the report about the Prophet 
marrying Abii Sufyan's daughter*'' He noted that al-Bukharl's work 
includes at least three impugned traditions, such as the hadith of the 



■'^ Weisman, "Between Sufi Reformism and Modernist Rationalism," 210-13; Daniel 
Brown, Rethinking Tradition, 30. 

" Hodgson, Tfie Venture of Mam, 3:160. 

" Ibn Taymiyya, Afo/'ma' a/^iaroa, 18:20. 

*^ Ibn Taymiyya, Itm al-haditfi, 112; idem, Majmu at-fatdwd, 13:352. 

■"> Ibn Ta^ymiyya., Majmu at-fatdwd, 17:235-7. 



314 C HAPTER EIGHT 

Prophet marrying Maymuna while in a state of pilgrimage {muhrim). 
Ibn Taymiyya exceeded even his own boundaries by criticizing the 
hadith of the Prophet praying after the eclipse, which appears in both 
the Sahihayn.^'' This seemingly contradictory approach to the canon, 
wielding its authority as the acme of critical hadith scholarship but 
simultaneously denying it iconic status, would reappear with the mod- 
ern Salafi movement. 



Muhammad b. Isma 'il al-San 'am: A Yemeni Salafi 

The Zaydi Shiite center of San'a' was an unusual setting for a revival 
of the Sunni hadith tradition. This environment, however, produced 
a succession of hadith scholars of singular dynamism and devotion to 
the study of the Prophet's sunna through the medium of hadith. An 
early progenitor was the ninth/fifteenth-century scholar Muhammad 
b. Ibrahim Ibn al-Wazir (d. 840/1436). Although he sprang from 
Zaydi origins, Ibn al-Wazir wrote a rebuttal of this Shiite school and 
then penned a massive defense of the Prophet's sunna as understood 
through the Sunni prism of Prophetic hadith.*" Ibn al-WazIr's intel- 
lectual interests lay in interacting with the Sunni hadith tradition, and 
he thus composed a commentary on Ibn al-Salah's Muqaddima. In this 
work, the Tanqih al-anzdr, he demonstrates an intellectual creativity 
unparalleled by his contemporaries in Cairo. Far from blindly following 
Ibn al-Salah's chapter structure like al- 'Iraqi and others, he addresses 
neglected issues such as the reliability of Ibn Majah's Sunan topically. 
He foreshadows the Salafi movement's anti-madhhab stance by stating 
that, in matters of law, it is not permitted to ignore a hadith declared 
sahih unless one can demonstrate a damning flaw in the report.*^ 

Although he lived over three centuries later, Muhammad b. Isma'll 
al-San'anI (b. 1099/1688, d. 1768) inherited Ibn al-WazIr's Salafi spirit, 
devoting a large commentary to his Tanqih al-anzdr and frequently citing 



" Ibn Taymiyya, 'Ilm al-hadith, 160; idem, Majmii' al-fatdwd, 18:22. 

'"' Al-Sakhawi, al-Daw' al-ldmi', 6:282. This second work has been published as 
al-'Awdsim min al-qawdsimji al-dhabb dn sunnat Abi al-Qdsim, ed. Shu'ayb Arna'ut, 2nd 
ed., 9 vols. (Beirut: Mu'assasat al-Risala, 1992). For a brief discussion of Ibn al-WazIr 
and his place in Yemeni intellectual history, see Bernard Haykel, "Reforming Islam 
by Dissolving the Madhdhib: ShawkanI and His Zaydi Detractors in Yemen," in Studies 
in Islamic Legal Theory, 338. 

*^ Ibn al-WazIr, Tanqih al-anidr, 48. 



THE CANON AND CRITICISM 315 

his predecessor with great affection. ■'" Like Ibn al-Wazir, he hailed from a 
Zaydi background but remained steadfastly focused on the Sunni hadlth 
tradition. His oeuvre also consisted almost entirely of commentaries on 
the works of major Sunni muhaddiths: Ibn Daqiq's Ihkdm al-ahkdm, Ibn 
Hajar's Bulugh al-mardm and al-Suyutl's al-Jdmi' al-saghir. Al-San'anl's 
Kitdb Tqdz al-jikra li-murdjadt al-jitra (The Awakening of Thought for a 
Return to the Pure Nature [of Islam]) represents an attempt to break 
theological discussion out of what he sees is the stupor of taqlid and 
senseless speculation [khawd), returning it to the ways of the Salaf He 
declares that blind imitation has always been mankind's pitfall, but fur- 
ther lambastes decadent Muslim scholars for their laziness, divisiveness, 
and obsequiousness. He accuses participants in speculative theology of 
constructing straw-man arguments for their opponents and then fail- 
ing to reevaluate such useless assertions. Furthermore, if a hadlth or 
Qur'anic verse contradicts these scholars' stance or school of thought, 
they try to interpret it away even if the interpretation is impossible in 
that context.'^' 

Al-San'anI studied in Mecca and Medina with Salim b. 'Abdallah 
al-Basrl and others, then returned to San'a' to serve as the preacher 
in the city's main mosque. He frequently provoked the ire of Zaydi 
scholars and the community's leaders, however, with his preoccupation 
with studying and teaching the "classic [ummahdtj" Sunni hadlth books. 
He also broke with the rest of the community in his insistence on fol- 
lowing hadiths instead of the Zaydi school in matters of ritual. Like 
al-Bukhari before him and later the ahl-e hadith in India, he insisted on 
raising his hands in prayer and holding them by his chest instead of by 
his side like other Shiites." Al-ShawkanI, al-San'ani's principal biogra- 
pher, held him in great personal admiration and saw him as an ideal 
Salafi hadlth scholar unafraid of breaking with social convention. He 
described al-San'anI as one who "fled from taqlTd and the spuriousness 
of those opinions of the jurists that lacked any proof '"'^ 



^" See, for example, al-San'ani, Hadlth iftiraq al-umma ila nayyif wa sab'injirqa, ed. Sa'd 
b. 'Abdallah al-Sa'dan (Riyadh: Dar al-'Asima, 1415/ [1994]), 95-7. 

^' Al-San'anI, Kitdb iqdz: al-fikra li-murdja'at al-fitra, 48-50. 

^^ Muhammad b. 'All al-Shawkanl, al-Badr al-tdli' bi-mahdsin man ha'd al-qarn al-sdbi', 
ed. Khalll Mansur, 2 vols. (Beirut: Dar al-Kutub al-'Ilmiyya, 1418/1998), 2:53-5; 
Metcalf, Islamic Revival in British India, 275. 

^' Muhammad b. 'All al-ShawkanI, al-Badr al-tdli', 2:53. 



316 C HAPTER EIGHT 

Indeed, al-San'ani stands out as one of the most fearlessly iconoclastic 
hadlth scholars in Islamic history. Five centuries after Sunni consensus 
had solidified on the complex question of defining the uprightness 
{'addld) of a hadith transmitter in the work of Ibn al-Salah, al-San'am 
proposed a total reconsideration. Whereas Sunni hadith scholars had 
accepted Ibn al-Salah's definition that an upstanding transmitter be 
"an adult Muslim of sound mind, free of the paths of sin and defects 
in honor [muru'd)" al-San'anl's Thamardt al-nazar Jt 'ilm al-athar (The 
Fruits of Reasoning in the Science of Traditions, written 1758) argues 
that this elaborate definition is pointless. Rather, 'addla is simply the 
state of "the likelihood of truthfulness [mazannat al-sidq)." The existing 
standards of uprightness, al-San'ani continues, are too lofty for the 
material they supposedly govern. Muhaddiths, like scholars in the other 
Islamic sciences, had become distracted in setting up principles (usul) 
that do not hold up in actual application (furu')."* 

Al-San'ani's iconoclasm appears most clearly in his treatment of al- 
Bukharl's and Muslim's works. Although he greatly respected the two 
masters, this maverick rejected almost every feature of the Sahihayn 
canonical culture as constructed by al-Khatib, Ibn al-Salah, al-Nawawi 
and Ibn Hajar He states quite simply that "we respect the Sahihayn, but 
we do not give them more station than they deserve."^' 

Most dramatically, he rejects the claim of the umma's consensus 
on the two books. Although al-Nawawi had earlier refused the notion 
that this consensus meant that the contents of the Sahihayn yielded 
epistemological certainty, he never questioned that ijmd' on the books' 
authenticity had in fact occurred. Al-San'ani, on the other hand, refutes 
this, citing the improbability of all the Muslim scholars agreeing on 
the authenticity of al-Bukhari's and Muslim's hadiths. Are we also to 
assume, he asks, that everyone who had in fact approved the two books 
was truly famUiar with their contents? Even before the SdhWcyn were 
written, he concludes, such practical difficulties in evaluating consensus 
had led Ibn Hanbal to pronounce that anyone who claimed ijmd' had 
occurred on an issue was a liar"'*' The main hadith providing justifica- 
tion for the infallibility of the umma's consensus, he continues, would 
not even apply to the intricacies of hadith criticism. The Prophet had 



Al-San'anl, Thamardt al-na^ar, 125. 
Al-San'anl, Thamardt al-na^ar, 137. 
Al-San'anl, Tawdih al-ajkdr, 1:93. 



THE CANON AND CRITICISM 317 

Stated that his community would not agree on "going astray (dalala)," 
while a minor flaw in a narration can hardly merit such a title. The 
umma is immune to error writ large, not small oversights {khata ') such 
as making a mistake in evaluating the isndd of an dhdd hadith." 

Al-San'ani also attacked the canonical ranking of al-Bukhari above 
Muslim. He argued that the feature that had most clearly distinguished 
al-Bukhari above Muslim, his requirement for at least one meeting 
between transmitters in narrations via "from/on the authority of 
{'an)," had little practical value and provided no real guarantee of 
direct transmission. How could a transmitter who may have narrated 
hundreds of hadlths from a particular teacher hear all these reports in 
one sitting? Considering this, what use is al-Bukhari's requirement for 
one meeting in guaranteeing the direct transmission of all the hadlths 
passed through this link? There still remains the possibility of a break 
{irsdl) in the isndd/'^ Just as al-San'ani deflates al-Bukhari's requirement, 
he gives a more positive evaluation of Muslim's. Muslim's requirement 
for contemporaneity in dn transmissions was not a naive assumption 
that two people who lived at the same time had heard their hadlths 
from one another; Muslim simply required the high probability that 
the two had met for direct transmission. In reality, this was the same 
level of assurance provided by al-Bukhari's theoretically more rigorous 
conditions."''' 

Al-San'ani also rejects attempts to disarm the opinions of scholars 
who had favored Muslim's Sahih over al-Bukhari's. Unlike the standard 
line that "some" scholars from the Maghrib had preferred Muslim's 
collection, he feels that a large number of prominent hadith experts 
had in fact favored Muslim. Furthermore, they did so for reasons more 
significant than Muslim's exclusion of incomplete legal-commentary 
reports {ta'liqdt) and his convenient grouping of all the narrations of 
a tradition in one place. Al-San'anI claims that he saw in the writings 
of al-NawawI, Ibn Jama'a and Taj al-Din al-TabrizI indications that 
these scholars felt Sahth Muslim was more authentic than Sahih al- 
Bukhdn. He also rejects Ibn Hajar's attempts to explain away Abu 'All 
al-Naysaburi's proclamation that Muslim's work was the most authentic 
book available.'''' 



Al-San'anI, Tawdih al-afkur, 1:94. 

Al-San'anl, Tawdih al-afkur, 1:302-3. 

Al-San'anl, Tawdih al-ajkdr, 1:47-8. 

Al-San'anI, Tawdih al-ajkar, 1:50-1. 



318 C HAPTER EIGHT 

Ibn al-Salah's and al-Nawawi's demands for charity on the issues of 
tadlis and the criticism of transmitters did not convince al-San'ani. He 
reminds us that many of al-Bukhari's and Muslim's transmitters were 
criticized with good reason and clear explanations.*'' In response to al- 
Nawawi's claim that instances of a mudallis's transmitting through 'an 
in the Sahihayn should be treated as direct transmission, al-San'anI cites 
Ibn Daqiq's and Ibn al-Murahhal's skeptical objections/'^ He comments 
that "this is a claim, but where is the proof?" Here he even breaks 
with Ibn al-Wazir, who had acceded to the notion that al-Bukhari and 
Muslim would not have included a mudallis's narration via 'an unless 
they knew it occurred through another reliable isndd. Again, al-San'ani 
objects that there is no proof for such a claim. ''^ 



Shah Wall Allah and the First Condemnation of Criticizing the Canon 

Like Ibn 'Abd al-Wahhab and al-San'am, the great Indian scholar 
Shah Wall Allah voyaged as a young man to the HijazI crucible of 
reformist hadith scholarship and returned to his native Delhi with a 
heightened appreciation for the authority of the hadith tradition. In 
terms of fluency with the labyrinth of Islamic sciences, however, he 
proved far more advanced than the stark hadith-based Hanbalism of 
Ibn 'Abd al-Wahhab. Even al-San'ani, who grasped and engaged the 
Ash'ari and Mu'tazilite traditions of dialectical theology, did not match 
Shah Wall Allah's innovative mixture of hadith scholarship, reformed 
Sufism, social and political activism, and even Neo-Platonism. 

Unlike Ibn 'Abd al-Wahhab's and al-San'ani's preoccupation with 
matters of creed and ritual observation. Shah Wall Allah's career 
tackled the troubling political realities of India in his time. The sud- 
den failure of Moghul imperial power after the death of the emperor 
Aurangzeb in 1707 marked the end of unified and effective Moghul 
rule in the subcontinent. Shah Wall Allah was eyewitness to the terrible 
destruction wrought on the unprotected Moghul realm in the wake of 
the empire's decay. In 1739, the Afghan conqueror Nadir Shah sacked 
Delhi and caused tremendous bloodshed. Combined with a series of 



Al-San'anl, Tawdih al-ajkar, 1:99. 

Al-San'anl, Tawdih al-qfkdr, 1:320 IF. 

Al-San'anl, Tawdih al-ajkdr, 1:323; cf. Ibn al-VVazIr, Tanqih al-anidr, 144. 



THE CANON AND CRITICISM 319 

disastrous Afghan invasions in 1748, 1757 and 1760, these events 
traumatized the psyches of men like Shah Wall Allah.''* For scholars, 
it represented the fragmentation of Islamic society in India. As Ahmad 
Dallal writes, "Disunity is a central theme that occupied [Shah Wall 
Allah] throughout his life.'"'-' 

In his role as a scholar, teacher, and social activist and in his rela- 
tions with local Indian rulers. Shah Wall Allah sought to regain a lost 
unity. He believed that political power was an essential component of 
a rejuvenated Islamic civilization in India. In the wake of the Moghul 
failure, he wrote to several leaders such as the Nizam of Hyderabad 
asking them to take on the role of Islam's patron and leader in the 
subcontinent. *'*' This desire to protect communal cohesion resulted in 
an attitude towards religious disagreement and popular practices that 
was more pluralistic than those of Ibn 'Abd al-Wahhab, al-San'anI or 
the founder of the West African Sokoto Caliphate, Usman dan Fodio 
(d. 1817). Unlike the Wahhabis, he proved very conservative about 
excommunication, limiting it to cases for which the Qur'an or hadith 
provided direct evidence and not extending it to acts of associationism 
{shirix) such as prostrating to trees. He allowed people to visit tombs for 
mourning and to seek the intercession of pious people, provided one 
did not glorify them.'"' 

Shah Wall Allah agreed with the other reformists that excessive 
loyalty to the madhhabi, had seriously hobbled the Islamic intellectual 
tradition and led it away from the Prophet's true message. Yet he also 
recognized the tremendous utility of these institutions. He personally 
treated all four Sunni madhhabs equally, and urged scholars to use them 
eclectically as reservoirs of expert opinions. The ultimate determinant in 
selecting which school's ruling to take, however, was the direct sayings 
of the Prophet. Since all the schools of law had theoretically derived 
their authoritative rulings from the Prophet's sunna, the hadlths retained 
an inherent and constant superiority to these bodies of substantive law. 
Each generation of scholars should thus consult them anew^" For the 
masses of Sunni Muslims, however, following one of the four established 



''■' Metcalf, Islamic Revival in British India, 25. 

'" Dallal, "The Origins and Objectives of Islamic Revivalist Thought," 343; Hodgson, 
The Venture of Islam, 3:148. 

'"'' Metcalf, Islamic Revival in British India, 35. 

'"' Dallal, "The Origins and Objectives of Islamic Revivalist Thought," 346. 

''^ Metcalf, Islamic Revival in British India, 37; Dallal, "The Origins and Objectives of 
Islamic Revivalist Thought," 347-8. 



320 CHAPTER EIGHT 

madhhab's, was essential. In India, they should adhere to the rulings of 
their traditional Hanafi school/'^ 

Shah Wall Allah's commitment to communal cohesion governed his 
attitude towards the Sahihayn canon. Despite the reformist tendencies 
he shared with his fellow student in the Hijaz, al-San'ani, Shah Wall 
Allah was no harsh iconoclast. He staunchly defended the canon. 
Like the schools of law, they provided indispensable institutions for 
the preservation of unity in Islamic thought. He states at the begin- 
ning of his discussion of hadith in his magnum opus, the Hujjat Allah al- 
bdligha (God's Conclusive Argument), "Know that there is no path for 
us to know the precepts of the Sharia or its rulings except though the 

reports of the Prophet (s) " Reliable books of hadith, foremost the 

Sahihayn and Malik's Muwatta', are essential for this, since "there does 
not exist today any non- written, reliable transmission [riwaya . . . ghayr 
mudawwana) [back to the Prophet]."'" He then lists the various levels 
of hadith collections, beginning with the top level of the Muwatta ' and 
the Sahihayn. Alluding to a Qur'anic verse (Qur'an 4:115) used since 
the time of al-Shafi'l (d. 204/819-20) to emphasize the importance of 
consensus {ymd'), he states: 

As for the Sahihayn, the hadith scholars have come to a consensus that 
everything in them with an isndd back to the Prophet is absolutely authen- 
tic, that [the two books] are attested by massive transmission back to their 
authors, and that anyone who detracts from their standing is a heretic 
[mubtadi') not following the path of the believers." 

This represents the first moratorium on criticism of the Sahihayn. 
Although Abu Mas ud al-Dimashql, Ibn al-Salah, al-NawawI and Ibn 
Hajar had all rallied to the defense of al-Bukharl and Muslim, they had 
never condemned criticism of the S^hihoyn as inherently unacceptable. 
Even after the consolidation of the canonical culture in the seventh/ 
thirteenth century, no one attacked the critiques of Ibn Taymiyya or 
the virulent criticisms of Ibn Abl al-Wafa' as violations of the canonical 
orthodoxy. Ibn al-Salah and al-NawawI had struggled to protect the 



^'^ Metcalf, Islamic Revival in British India, 39; Peters, "Idjtihad and Taqlld in 18th and 
19th Century Islam," 143; Marcia K. Hermansen, trans., The Conclusive Argument from 
God (Islamabad: Islamic Research Institute, 2003), xxx, 

™ ShiihWAliAMh, Hujjat Allah al-bdligha, 1:132-3. 

" Shah Wall Allah, Hujjat Allah al-bdligha, 1:134. For a discussion of the use of this 
verse as a proof text for ijma, see Abu Zahra, Ibn Taymiyya, 469 IF. 



THE CANON AND CRITICISM 321 

Sahihayn because the books had become crucial institutions in Sunni 
scholarly culture. Yet in the relative stability of Mamluk Cairo, attacks 
by critics like Ibn Abl al-Wafa' held little consequence for the sturdy 
and blossoming Sunni religious culture of the period. 

For Shah Wall Allah, the stakes had become much higher indeed. 
Although we do not know exactly to whom he directed his warn- 
ing about criticizing the Sahihayn, only a merchantman's ride away 
across the Indian Ocean in Yemen his contemporary al-San'ani was 
flagrantly dismissing the canonical culture that had been constructed 
to protect the institution of the Sahihayn. Although Shah Wall Allah 
was a hadlth-oriented reformist who sought to limit the divisive effects 
of the madhhabi,, he appreciated the roles of such institutions in main- 
taining social, intellectual and political order in a beleaguered umma. 
It is not difficult to imagine that he had come across the iconoclastic 
thought of the young San'ani while in the Hijaz, perhaps in the classes 
of their common teacher Abu Tahir b. Ibrahim al-Kurdi (d. 1732—3), 
and later sensed the danger it posed for his reformist agenda. While 
we can hardly contend that Shah Wall Allah's harsh condemnation of 
criticizing al-Bukhari and Muslim was an actual response to al-San'anl's 
writings, it might as well have been. What al-San'ani reviled as "the 
heresy of madhhabism," and the baseless premises of the Scihihcyn 
canonical culture. Shah Wall Allah saw as essential institutions for the 
Islamic revival. 



Muhammad JVasir al-Din al-Albani: Iconoclast Extraordinaire 

Muhammad Nasir al-Din al-Albani was born in 1914 in Shkoder, 
Albania, to a family of staunchly Hanafi scholars. When he was nine 
years old, however, his family emigrated to Syria. There the young 
Albanl followed in his father's footsteps and studied Hanafi jurispru- 
dence with other Albanian students in Damascus. As a young man, he 
entered a bookstore near the Umayyad Mosque one day and found a 
copy of Rashid Rida's and Muhammad 'Abduh's reformist journal al- 
Mandr. An article written by Rida in particular struck al-Albani. Rida 
was criticizing the great champion of classical Sufism, Abu Hamid 
al-Ghazall, for his Sufi teachings and his use of unreliable hadiths to 
justify them. Al-Albani also found the hadith scholar Zayn al-Din al- 
'Iraqi's (d. 806/1404) book detailing those weak hadiths that al-Ghazali 



322 CHAPTER EIGHT 

had included in his classic Ihyd' 'ulum al-din (Revival of the Religious 
Sciences).'^ These works sowed the seeds of mistrust in al-Albani's heart 
for Sufism and weak hadiths; for him they were loopholes through which 
'inauthentic' practices could enter Islam. Attracted by al-Manar's, call 
for the purified, Arab Islam of the Prophet's time, he began studying 
the hadith sciences independently 

Like Ibn Taymiyya and Ibn 'Abd al-Wahhab before him, al-Albani 
turned against the practices of popular Sufism and the strict adher- 
ence to one school of law in the face of contradicting hadiths. He 
read through all of Ibn 'Asakir's mammoth Tdrikh madinat Dimashq 
and, discovering that the Umayyad Mosque had formerly been the 
Church of St. John built on his tomb, refused to pray there." Like other 
Salafis, al-Albani regarded the act of incorporating graves into worship 
as bid'a.'* These non-conformist ways eventually angered al-Albanl's 
father, who told his son that he needed to choose between "disbelief 
and monotheism [al-kufr wa al-tawhid)." Al-AlbanI replied that equally he 
must choose between "the sunna [of the Prophet] and taqlid." Cast out 
penniless by his father, al-AlbanI became a watch repairer and began 
spending long hours in the Zahiriyya Library in Damascus (founded 
by Tahir al-Jaza'irl) poring over hadith manuscripts.''' 

Al-AlbanI devoted himself to hadith scholarship in the Salafi idiom. 
He undertook what became an extensive project that he would later 
dub "bringing the sunna within reach of the umma {taqnb al-sunna bayn 
yaday al-ummd)," the principal aim of which was to remove what he 
deemed weak hadiths from important classical Islamic texts. It was the 
deleterious effects of these weak hadiths that had allowed the Muslim 
community to stray so far from the authentic legacy of the Prophet. 
This Salafi philosophy is best glimpsed in al-Albanl's massive, thirteen- 
volume work identifying weak hadiths entitled Sihilat al-ahddith al-da 'ifa 
wa al-mawdu'a wa ta'thirihd al-sayyi' fi al-umma (The Series of Weak and 
Forged Hadiths and Their Negative Effect on the Umma). He also 
composed books identifying the weak hadiths found in famous works 
such as al-Mundhirl's (d. 656/1258) al-Targhib wa al-tarhib, al-Bukharl's 



'^ Al-Albanl, ^^Tarjamat al-shaykh al-Albani — Nash'at al-Shaykhfi Dimashq" lecture from 
www.islamway.com, last accessed 6/3/2004. 

" Al-AlbanI, "Tarjamat al-Shaykh al-Alhdm — 2," lecture from www.islamway.com, last 
accessed 6/3/2004. 

'* 'All, Muhammad Mdsir al~Din al-Albani, 23. 

'^ Al-AlbanI compares his breaking with his father's legal school with Abraham's 
leaving his father's idolatrous ways; see al-Albanl, "Tarjamat al-Shaykh al-Alhdni — 2." 



THE CANON AND CRITICISM 323 

al-Adab al-mufiad and finally the famous Four Sunans of Abu Dawud, 
al-Tirmidhi, al-Nasa'T and Ibn Majah.'*' 

Al-AlbanT combined such focused hadith scholarship with intensive 
scholarly activism. Through his books and preaching, he sought to 
reform the community around him by calling them to heed the Qur'an 
and the Prophet's sunna above all things. He traveled from city to city, 
attacking in speeches and writings what he called "corrupting morals, 
illegitimate forms of worship and false beliefs."" He called on the pre- 
dominantly Hanafi scholars around him to ensure that their school's 
rulings accorded with the sunna of the Prophet as expressed in the 
hadith corpus. A mufti might advocate his school's position on a ques- 
tion, but he should always provide direct evidence from the Qur'an and 
the hadith before doing so.'" His books attacked innovative religious 
practices (bid'a) and sought to eradicate them from social institutions such 
as funerals, wedding ceremonies, and the annual pilgrimage. His criti- 
cisms extended to state interference in religious affairs, for he rejected 
the Syrian government's support for the Hanafi legal code as embodied 
in the Ottoman Majelle as well as the position of scholars who allowed 
interest for the sake of facilitating modern finance.'^ Eventually he was 
imprisoned in Syria, where he wrote a major work on al-Bukhari's Sahih, 
and was forced to emigrate to Jordan in 1980. 

Al-Albani, like Ibn 'Abd al-Wahhab and Shah Wall Allah, telescoped 
the normative dimension of time in Islamic religious history. He rejected 
the atavistic logic of the Islamic intellectual tradition and considered 
himself qualified to review the work of the classical scholars of Islam."" 
Al-Albani was not calling for intellectual anarchy or the neglect of 
scholars; like all Muslim scholars, he clearly identified a certain group 



"^' See al-Albanl, Da'iJ Sunan Abi Ddwud (Beirut: al-Maktab al-lslami, 1408/1988); 
idem, Da'if Su?ian al-Tirmidhi (Beivui: al-Maktab al-lslami, 1411/1991); idem, Da'if Sunan 
al-Nasd'i (Beirut: al-Maktab al-lslami, 1411/1990); idem, Da'lf al-Targhib wa al-tarhib, 
2 vols. (Riyadh: Maktabat al-Ma'arif, 1421/2000); idem, Sahth al-Targhib wa al-tarhib 
(Riyadh: Maktabat al-Ma'arif, 2000). 

" "akhldq Jusida, 'ibdddt mubtadi'a wa 'aqiddt bdtila...," Muhammad Nasir al-Din al- 
Albanl, "Silsitat as'ilat Abi Ishdq al-Huwayni li'l-shaykh Muhammad Mdsir al-Din al-Albdni," 
lecture from www.islamway.com, last accessed 2/13/2002. 

'" Al-Albanl, "al-Taqlid," two-part lecture from www.islamway.com, last accessed 
2/12/2002. 

" Al-AlbanI, "al-Taqlid," and "Silsilat as'ilat Abi Ishdq al-Huwayni." 

"'' See al-AlbanI, Fatdwd al-shaykh al-Albdni, ed. 'Akasha 'Abd al-Mannan al-Tayyibi 
(Cairo: Maktabat al-Turath al-lslami, 1414/1994), 162. Here the author states that one 
scholar's position cannot be taken over another's simply because he lived earlier 



324 CHAPTER EIGHT 

known as "the people of knowledge [ahl al-'ilm)" to whom everyday 
Muslims should turn for religious expertise. Nor was he rejecting the 
work of classical Muslim scholars; indeed al-Albani relied entirely on 
earlier criticisms of hadlths and their transmitters in his reevaluation of 
the contents of famous works. Although he considered himself qualified 
enough to reexamine classical texts, he could not recreate the intimate 
access that classical scholars had to the minutiae of hadlth criticism. 
Al-Albani's books, such as the Silsilat al-ahadith al-da 'ifa, thus apply the 
opinions of classical hadlth masters and later critics such as Ibn Abi 
al-Wafa' to classical texts. They are thus replete with citations from 
the whole range of Sunni authorities, including al-Shafi'i, Ibn Hajar 
and Ibn Hazm."' 

This telescoped vision of religious history centered on the study 
of hadlth as a continuous and living tradition in a constant state 
of reevaluation. When asked about his controversial criticism of a 
famous hadlth transmitter from the early Islamic period, al-Albani 
replied that the science of hadlth criticism "is not simply consigned to 
books [mastur fi al-kutub),"^"^ it is a dynamic process of critical review. 
Al-Albani explained that one of the principles of Islamic scholarship 
is that "religious knowledge {'ilm) cannot fall into rigidity [Idyaqbalu al- 
jumud)."^^ It is thus not surprising that al-Albani and his students are 
the first Muslim scholars in centuries to produce massive collections 
evaluating Prophetic traditions. 

Al-Albani's career has certainly been one of the most controversial 
in modern Islamic intellectual history. In both his legal rulings and 
hadlth evaluations, al-Albani broke with the communal consensus of 
the madhhab traditions. Like Ibn 'Abd al-Wahhab, he was thus attacked 
for breaking with the infallible ijmd' of the umma."* Although he drew 
almost entirely on the work of classical scholars, his reevaluation of 
hadlths long considered authentic or relied on by elements of the Mus- 
lim community provoked controversy. Madhhab Traditionalists recoiled 



"' See, for example, al-Albani, Silsilat al-ahadith al-da'ifa wa al-mawdua, 13 vols. 
(Riyadh: Maktabat al-Maarif, 1422/2002), 1:141, where he draws from Ibn Hazm's 
al-lhkdmji uml al-ahkdm. 

"^ Al-AlbanI, "Silsilat as'ilat Abi Ishdq al-Huwayni li'l-shajkh Muhammad Mdsir al-Din 
al-Albdni." 

"' Al-AlbanI, Sahih al-Targhib wa al-tarhib, 1:4. 

'* For this criticism of Ibn 'Abd al-Wahhab, see Samer Traboulsi, "An Early 
Refutation of Muhammad Ibn 'Abd al-Wahhab's Reformist Views," Die Welt des Islams 
42, no. 3 (2002): 393. 



THE CANON AND CRITICISM 325 

at his influential and barbed criticisms of the traditional schools of 
jurisprudence, broad rejection of Sufism and controversial legal rulings. 
His prohibition on women wearing gold bracelets, otherwise considered 
a female prerogative, angered traditionalists, while his statement that 
women need not cover their faces drew the ire of conservatives who 
might otherwise embrace his fundamentalist calling. ''^ According to even 
his own students, al-Albanl's personality could be caustic. 

A plethora of books have thus appeared attacking al-Albani and 
refuting his positions, most of them from the pens of Madhhab Tradi- 
tionalists. The Jordanian Ash'ari theologian, Hasan b. 'All Saqqaf, for 
example, composed a book entitled Qamus shatd'im al-Albdni (Diction- 
ary of al-Albanl's Slanderings). Other scholars have more specifically 
criticized al-Albanl's rulings on the authenticity of hadiths in his Silsilat 
al-ahddith al-da'ifa, his Silsilat al-ahddith al-sahiha, and his listing of weak 
reports from the Four Sunani,.''^' 

Al-Albanl's sometimes autodidactic education was a further affront to 
many Muslim scholars, who absolutely required a student to read texts 
at the hands of a scholar trained within an interpretive school and to 
eventually receive license (jjdzd) for his understanding of that book. Just 
as Taqi al-Din al-Subki (d. 756/1356) had accused Ibn Taymiyya of 
not learning the proper interpretation of classical texts from qualified 
transmitters, so too many scholars have attributed al-Albani's unaccept- 
able positions to his lack of ijdza's,.'''' 



Against the Canon: Al-Albani's Criticism of the Sahihayn and His Detractors 

Al-Albani used the Sahihayn canon for the same dialectical purposes as 
generations of Muslim scholars before him: they provided him a trump 
card in debates over the authenticity of hadiths. He acknowledged the 
rhetorical power of the two books, saying that "it has become like a 



«■' Al-Albanl, Fatdwd, 593 ff. 

°^ For example, see Salah al-Din al-Idilbl, Kashf al-ma'lul mimmd summiya bi-Sildlat 
al-ahddith al-sahiha (Amman: Dar al-Bayariq, 1421/2001); Mahmud Sa'ld Mamduh, 
al-Ta'nf bi-awhdm man qassama al-sunan ild sahih wa da'ij, 6 vols. (Dubai: Dar al-Buhuth 
li'1-Dirasat al-Islamiyya wa Iliya' al-Turath, 1421/2000); Hasan b. 'All Saqqaf, Qamus 
shatd'im al-Albdm (Amman: Dar al-Imam al-NawawI, 1993). 

" Al-Subkl, al-Sayf al-saqil, 63. Muhammad Abu Zahra has convincingly argued 
against this accusation leveled at Ibn Taymiyya. See Abu Zahra, Ibn Taymiyya; 111 S., 
118. 



326 CHAPTER EIGHT 

general convention {'u7f ""'dmm'"')" among Sunni scholars that anything 
included in the Sahihayn is without a doubt authentic."" When asked 
about several pro-Shiite hadiths asserting 'All's rightful place as the 
Prophet's successor, al-Albani replied that if someone really believes 
these reports, he should "lay out the Sahihayn before him" and find the 
hadiths in one of them as proof "^ 

Yet like the Damascene firebrand Ibn Taymiyya, al-Albani openly 
undermined any iconic status for the two works beyond their conve- 
nience as authoritative references in debate. He rejected the practice 
of some less thorough jurists who, like al-Hakim al-Naysaburi, would 
manipulate the legitimizing power of the "standards of al-BukharT 
and Muslim" by claiming that a hadlth met these criteria simply if 
the transmitters in its isnad were found in the Sahihayn.'^^ As his Egyp- 
tian student Abu Ishaq al-Huwayni explained, jurists cannot simply 
look up the narrators found in an isnad in a dictionary of transmitter 
criticism and declare the hadith authentic if none of them have been 
impugned. The science of hadith evaluation requires that one explore 
any corroborating or contrasting narrations of the hadith to determine 
its reliability'" 

In March 1969, al-Albanl published an edition of 'Abd al-'Azim al- 
Mundhirl's Mukhtasar Sahih Muslim as part of his efforts to provide the 
Muslim community with accessible versions of classical hadith works 
expunged of all weak material. His extreme respect for al-Bukhari 
and Muslim is evident, for he adds, "That is with the exception of 
the Sahihayn, due to the scholars' approval of these collections and 
their being free of weak or unacceptable reports (al-ahddith al-da'ifa wa 

al-munkard) "^^ This statement, however, clearly did not accurately 

represent the author's stance on the Sahihcyn. Drawing on well-known 
earlier criticisms, such as the problem of Abu al-Zubayr al-Makki's tadlTs, 
al-Albani notes in brief footnotes that about two dozen narrations in 
Muslim's collection contained flaws due to vagaries in their chains of 



"" Al-Albam, ed., Shark al-'Aqida al-Tahawiyya (Amman: al-Dar al-Islami, 1419/ 
1998), 22. 

»9 Al-Albanl, "al-Taqlid." 

'" Al-Albanl, Sahih al-Targhib wa al-tarhib, 1:70. 

" Abu Ishaq al-Huwaym, "Shurut al-Bukhdri wa Muslim" lecture from www.islamway. 
com last accessed 2/03/2004. 

"^ 'Abd al-'AzIm Zaki al-Din al-Mundhirl, Mukhtasar Sahih Muslim, ed. Muhammad 
Nasir al-Dln al-Albanl (Riyadh: Maktabat al-Ma'arif, 1416/1996), 23. 



THE CANON AND CRITICISM 327 



developed, he also criticized, in his lectures and writings throughout the 
1970s, 80s and 90s, hadiths from al-Bukhari's collection for isnad and 
content reasons, such as the report of the Prophet marrying Maymuna 
while in a state of pilgrimage."* 

Al-Albanl's empty homage to the consensus on the Sahihayn and his 
use of the two books as measures of authenticity in polemics, despite his 
many criticisms of the works, mirror the rhetorical duplicity with which 
the canon was employed in the classical period. Al-Albani's reliance 
on well-established criticisms of the Sahihayn does, however, clarify the 
seeming contradiction between such critiques and his condemnation of 
"Westernized" Modernist scholars who reject hadiths that "the umma 
has accepted with consensus": he did not feel that he himself was 
actually criticizing any of al-Bukhari's or Muslim's hadiths.''"' Rather, 
he was simply noting existing critiques made by the historical giants 
of hadlth scholarship. As he stated in defense of his noting a flaw in 
one of al-Bukharl's isndds, earlier critiqued by al-Dhahabi, "I am not 
the innovator {mubtadi') of this criticism ""^ 

Nonetheless, the outcry from the Madhhab Traditionalists over al- 
Albanl's perceived attack on the SaMhoyn was ferocious. In the early 
1970s, the Syrian Hanafi hadith scholar 'Abd al-Fattah Abu Ghudda 
(d. 1997) published a tract against al-Albani's reevaluation of the 
Sahihayn. In 1987 the Egyptian hadlth scholar Mahmud Sa'id Mamduh 
published a work entitled Tanbih al-muslim ild ta 'addi al-Albdni 'aid Sahih 
Muslim (Alerting the Muslim to al-Albani's Transgression upon S^hih 
Muslim):''' The Lebanese scholar and staunch defender of the traditional 
Islamic schools of law, Gibril Fouad Haddad, has dubbed al-Albani "the 
chief innovator of our time" and accused him of bid'a for publishing 
" 'corrected' editions of the two Sahihs of al-Bukhari and Muslim ... in 
violation of the integrity of these motherbooks."'"' 



'' See, for examples, al-Albanl, ed., Mukhtasar Sahih Muslim; 49 (#153 for the Jabir — > 
Abu al-Zubayr al-Makkl flaw), 1 2 1 (#'s 446 and 448, which al-AlbanI deems "weak"), 
210 (#831, criticized for a lackluster transmitter, 'Umar b. Hamza), 343 (#1293, again 
for 'Umar b. Hamza), 272 (#1039 for Literal Matn Addition). 

'" Al-AlbanI, ed., Shark al-'Aqida al-Tahdwijya, 23. 

»^ Al-Albanl, Mukhtasar Sahih a/-&ttan (Riyadh: Maktabat al-Ma'arif, 1422/2002), 
2:8-9. 

"" Al-Albanl, ed., Sharh al-'Aqida al-Tahdwijya, "il . 

'' Mahmud Sa'ld Mamduh, Tanbih al-muslim ild ta'addi al-Albdni 'aid Sahih Muslim 
([Cairo]: [n.p.], 1408/1987). 

" See www.sunnah.org/history/Innovators/al_albam.htm, last accessed 5/31/04. 



328 CHAPTER EIGHT 

The works of two of al-Albani's critics are particularly instructive 
in examining the dynamic between the canon and criticism. The 
most persistent detractor of al-Albani's hadith scholarship has been 
Mahmud Sa'id Mamduh, who studied with two of the scholar's bit- 
terest adversaries, 'Abd al-Fattah Abu Ghudda and the Moroccan Sufi 
'Abdallah b. al-Siddlq al-Ghumarl (d. 1993). Mamduh has written at 
least four rebuttals of al-Albanl's work on different subjects, but al- 
Albanl's impudence in criticizing the Sahihayn has proven the lodestone 
for Mamduh's attacks.^'' The most incisive and comprehensive defense 
of the Sahihayn canon, which perforce addresses al-Albani's criticisms, 
is the monumental Makanat al-Sahihoyn (The Place of the Sahihoyn) of 
the Medinan scholar Khalll MuUa Khatir 

For Madhhab Traditionalists, al-Albanl's criticism poses two main 
challenges. First, it threatens the important role of the Sahihayn canon 
in scholarly culture. Second, it undermines the institutions of consensus, 
scholarly hierarchy and the vision of history on which the canon rests. 
At the root of the Traditionalists' refutations of al-Albani's scholarship 
in general is his willingness to question the established practices and 
presuppositions of the Sunni scholarly tradition. Rejecting al-Albani's 
condemnation of using weak hadlths in Islamic law and ritual, Mamduh 
declares: 

Indeed, I have concluded that his methods disagree with those of the 
jurists and hadith scholars, and that he is creating [yuhdithu] great disar- 
ray and evident disruption in the proofs of jurisprudence both generally 
and specifically. He lacks trust in the imams of law and hadith, as well 
as in the rich hadith and law tradition handed down to us, in which the 
umma has taken great pride.'"" 

In contrast, MuUa Khatir reiterates the predominant non-Salafi view 
of Islamic religious history, according to which later generations are 
only worthy of imitating the great scholars of yore. "Al-Bukhari is a 
migtahid" he explains, "and contemporary people are imitators (muqallid), 
walking according to his principles and constraints, as well as those of 
others like him from among the people of knowledge.""" In his rebuttal 



'" An additional example of Mamduh's rebuttals of al-Albanl is his Wusul al-tahdni 
bi-ithbdt mnniyyat al-sibha wa al-radd 'aid al-Albdm. For a tangential discussion of al-Albani's 
inappropriate criticism of al-Bukharl, see Mamduh, al-Naqd al-sahih li-md u 'turida 'alayhi 
min ahddith al-Masdbih, 16-7 (see Ibn Hajar, Fath #'s 843 and 6329). 

'™ Mumduh, al-Ta'rif bi-awhdm, 1:14. 

"" MuUa Khatir, Makdnat al-Sahihayn, 494. 



THE CANON AND CRITICISM 329 

of al-Albanl's removing weak hadiths from the Four Sunans, Mamduh 
derides him for assuming that in the bygone ages Islam had been in 
error but that now, when the umma has devolved into the terminal 
and pervasive ignorance of endtime, he could return the community 
to the straight path. "As if the umma," he mocks, "was in error in the 
ages of light. . .l"'"^ Concerning al-Albani's removal of weak hadiths 
from al-Bukhari's work al-Adab al-mufrad, Mamduh asks rhetorically, "I 
wonder, was al-Bukhari, God bless him, unable to select the hadiths 
of al-Adab al-mufrad as he did with his Sahifi?"^^''' MuUa Khatir, who is 
too polite to name al-AlbanI specifically, merely talks of an "upstart at 
the end of time (ghirrji dkhir al-zamdn)" who impudently challenges the 
umma's consensus on the Sahthayn's absolute authenticity'"* 

The practical manifestation of the authority of tradition in Sunni 
scholarship is the notion of consensus, which transforms received opin- 
ion among scholars into a direct manifestation of God's authority as 
deposited in His chosen umma. One of the primary faults that Madhhab 
Traditionalists find in al-Albanl's criticism of the Sahihayn is thus his 
rejection of the consensus established with regard to the two works' 
authenticity. Mamduh states unequivocally in his Tanbih that al-Albanl's 
deigning to "examine critically (al-nazar Ji)" the Sahihayn constitutes an 
affront to the umma's acceptance of the two works and attacks the 
ymd' that hadlth scholars since the early 400s/ 1000s have declared on 
the two works. Even considering the possibility that some of the isndds 
in the Sahihayn contain flaws is to doubt the defining characteristic of 
the two books: all the material they contain is sahih by very dint of its 
inclusion.'"' The absolving power of i/'ma' provides the answers to any 
criticisms al-AlbanI might raise about the SdhWciyn, such as the ques- 
tion of tadlis in the two works. Invoking the charitable declarations 
made by Ibn al-Salah and al-NawawI, Mamduh explains, "The rules 
of hadlth have determined that al-Bukharl and Muslim were correct, 
and the umma has agreed on this.'""'' He adds that al-AlbanI "throws 
out the ijmd' of the umma and the craft of its hadlth masters, entering 



"'^ Mamduh, al-Ta'rif bi-awhdm, 1:11. This rebuttal duplicates early rebuttals 
of Wahhabism, such as that of 'Abd al-Wahhab b. Ahmad al-Shafi'l al-Azharl al- 
Tandatawl's rebuttal of Ibn abd al-Wahhab. See Traboulsi, "An Early Refutation of 
Muhammad Ibn Abd al-Wahhab's Reformist Views," 395. 

'"' yidi.m6-\Xh, al-Tanf bi-awhdm, 1:31. 

'"■' MuM KhSiin, Makdnat al- Sahihayn, 127. 

'"■^ MamdQh, Tanbih al-muslim, 13-14. 

""" Mamdnh, Tanbih al-muslim; 24, 53. 



330 CHAPTER EIGHT 

into a matter settled long ago and whose authenticity was agreed on 
centuries ago.""" 

Al-Bukharl's and Muslim's canonical function as the exemplum of 
excellence in hadith scholarship also serves as an exhibit in the canon's 
defense. Their work defines the rules of hadith scholarship, so who is 
al-Albani to question their judgment? MuUa Khatir states: 

Al-Bukharl and Muslim, may God bless them, they are the imam?, of this 
science, the stallions of its arena, without peer in their time, the heroes of 
their age, in mastery, criticism, research, examination and in encompassing 
knowledge . . . there can be no objection to the Shaykhayn.^"'^ 

In addition to breaking with consensus, critics of al-Bukhari and Muslim 
thus face the impossible task of superseding their ultimate expertise in 
hadith.'"^ MuUa Khatir correctly adds that nowadays hadith scholars 
cannot access all the material that al-Bukharl and Muslim had at their 
disposal but has since vanished."" How can al-AlbanI thus dare to cor- 
rect these vaunted masters? 

Like Shah Wall Allah's defense of the Sahihayn canon, Mamdtih and 
MuUa Khatir also reject al-Albanl's criticisms because they threaten 
the canon's well-established utility. MuUa Khatir notes that one of 
the properties of the two works is that one can act on their hadlths 
without any need to prove their authenticity.'" Perhaps his greatest 
objection to al-Albanl's scholarship is the very notion of "correcting 
the Sahihayn {tashih al-Sahthayn)," to which MuUa Khatir devotes an 
entire chapter in his book. For him the very notion of qualifying the 
phrase "al-Bukhari/Muslim included it" with the comment "and it 
is authentic" represents unmitigated effrontery to the purpose of the 



{hiiffa^)," he states, "if they cite a hadith from one of the Sahihayn, that 
was sufficient to rule that the hadith was authentic, so you do not see 
them researching the unarfs.""^ 



Mamduh, Tanbih al-muslim, 7. 
MuUa Khatir, Makdnat al-Sahihayn; 246, 256. 
MuUa Khatir, Makdnat al-Sahihayn, 318. 
MuUa Khatir, Makdnat al-Sahihayn, 488. 
MuUa Khatir, Makdnat al-Sahihayn, 80. 
MuUa Khatir, Makdnat al-SaMhayn, 474-6. 
Mamduh, Tanbih al-muslim, 7. 



THE CANON AND CRITICISM 331 

Al-Albanl's criticism of the Sahihayn also manifests the Salafi threat 
to the principles of following an established madhhab [taqlid) and the 
hierarchy of scholars so valuated among Madhhab Traditionalists. 
MamdQh asserts that al-Albanl's criticizing the SaMhoyn invites further 
criticism of the two works and is a call for unconstrained independent 
reasoning (ytihad) instead of the proper reliance on qualified scholars 
{taqlid). Criticizing these established institutions of Islamic scholarship 
"opens a door we cannot easily shut.""* Furthermore, it represents a 
challenge to the hermeneutic hierarchy of the madhhab?, and their system 
of authorized interpretation of texts. Mamduh states that al-Albani's 
opinions contain "great dangers" since he has given "to any claimant 
the right to judge the hadlths of the Sahihayn by what he sees as within 
the bounds of the scientific principles of hadlth.""'' MuUa Khatir's final 
evaluation of correcting the Sahihoyn is thus that criticizing "what the 
umma has agreed on is pure calumny and misguidance, the greatest of 
losses [al-khusrdn al-mubin) and the fatal blow [qdsimat al-iahr)."^^'' 



Conclusion: 
Al-Albdni's Reply and the Continuity of Iconoclastic Hadith Criticism 

Al-Albani was defiant in the face of his critics. He responded to 
Mamduh's condemnation of his reevaluation of some of Muslim's 
narrations by exclaiming, "As if, by Muslim's inclusion of these hadiths, 
they acquired some immunity (him"") from criticism. That is without a 
doubt a mistake.'"" In the last edition of his Mukhtasar Sahih al-Bukhdn, 
al-Albani states: 

It is essential that I put forth a word of truth for the sake of scholarly 
integrity (li'l-amdna al-'ilmiyya) and exoneration from blame (tabri'a li'l- 
dhamma, sic): a scholar must admit an intellectual truth expressed by Imam 
al-Shafi'l in a narration attributed to him: God has forbidden that any 
except His Book attain completion (aha Allah anyatimma ilia kitdbuhu)."'' 



"■' Mamduh, Tanbih al-muslim, 13-14. 

"^ Mamduh, Tanbih al-muslim, 24. 

'"^ MuUa Khatir, Makdnat al-Sahihayn, 488. 

'" Al-AlbanI, ed., Mukhtasar Sahih Muslim, 17. Here al-Albam seems to be direcdy 
quoting the seventh/thirteenth-century scholar of Marrakesh, Ibn al-Qattan al-FasI 
(d. 628/1231) in his massive hadlth work Baydn al-wahm wa al-ihdm. See Ibn al-Qattan 
al-FasI, Baydn al-wahm wa al-ihdm, 4:298. 

"' Al-Albanl, ed., Sharh al-'Aqida al-Tahdwiyya, 23; idem, Mukhtasar Sahih al-Bukhdri, 
2:5-6. 



332 CHAPTER EIGHT 

After describing a problematic hadith in al-Bukharl's collection, he 
adds that this is but one of dozens of examples that demonstrate the 
ignorance "of those impudent ones who chauvinistically acclaim al- 
Bukhari's Sahih, as well as that of Muslim, with blind loyalty and say 
with complete certainty that everything included in those two books 
is authentic.""^ 

Here we see al-Albani repeating essentially the same quote cited by 
al-Khatib al-Baghdadi nine centuries earlier as he defended his right to 
criticize al-Bukhari's identification of transmitters (although al-Khatib 
cites al-Shafi'i's student al-MuzanI as the source). Both deny that any 
book other than the Qjur'an can be free from error or attain immu- 
nity from criticism. Al-Khatib played a crucial role in constructing the 
SahihayrC% canonical culture, but he reserved the scholar's right to cor- 
rect his predecessors. No work can achieve an impervious iconic status, 
for scholars always reserve the right to scrutinize it critically. Al-AlbanI 
thus explains that '\SahTh al-Bukhdn, despite its glory and the scholars' 
acceptance of it..., has not been totally free of criticism from some 
scholars.'"^" Responding to the attacks of the Hanafi Abu Ghudda, 
al-Albani correctly points out that the Hanafi school has a long and 
persistent history of criticizing the Sahihayn.™ 

Al-Albani clarifies that his intention is not to reduce the utility of 
hadith collections or question the authority of Prophetic reports. He 
is merely noting existing criticisms of hadlths found in the Sahihayn 
for the benefit of the reader. Many such criticisms pertain only to one 
narration of the hadith and not to the Prophetic tradition itself '^^ In 
fact, he says that by showing that some hadlths criticized in works like 
Ibn Majah's Sunan actually have authentic and reliable versions, he 
"has saved hundreds of hadlths from the weakness that some of their 
isndd's, entail." '^^ 

For al-AlbanI, exempting the Sahihayn from critical review constitutes 
a betrayal of "scholarly integrity." Embracing a canonical culture that 
sacrifices critical honesty for the security of scholarly institutions violates 
a Muslim scholar's responsibility. The acceptability of criticizing the 



"5 Al-Albani, Silsilat al-ahadith al-sahiha (Riyadh: Maktabat al-Ma'arif, 1416/1996). 
6:2:93. 

"" Al-Albanl, Mukhtasar Sahih al-Bukhdn, 2:7. 

'^' Al-Albanl, ed., Shark al-'Aqida al-Tahdwiyya, 38-42. 

'^^ Al-Albanl, Mukhtasar Sahih al-Bukhdri, 2:4. 

'^' Al-Albanl, Mukhtasar Sahih al-Bukhdri, 2:5. 



THE CANON AND CRITICISM 333 

Sahihayn enunciates the contrast between this Salafi attitude towards 
the canonical culture and that of its staunch supporters. When Ibn 
al-Jawzi declared some hadiths from Ibn Hanbal's Musnad forgeries 
because their contents seemed to contradict tenets of the faith, the 
great champion of the Sahihayn canon, Ibn Hajar, wrote that we must 
try to reconcile this material and not dismiss it. "For if people open 
that door to rejecting hadiths," he wrote, "it would be claimed that 
many hadiths from the SahThayn were false, but God most high and 
the believers have refused to let this happen."'^* In contrast, the Salafi 
hadith scholar Tahir al-Jaza'iri argues that Ibn Taymiyya justifiably 
criticized a hadith from al-Bukharl's collection for unacceptable con- 
tent. Al-Jaza'irl expresses surprise and concern over scholars who try to 
suppress discussion of mistakes in the Sahihayn because they think that 
allowing criticism of the matn will open the door to the "people with 
agendas [ahl al-ahwd').'" He disagrees, saying that proper criticism is a 
worthy practice.'^' Al-AlbanI echoes this sentiment, saying that proper 
criticism based on the principles of hadith scholarship is never inap- 
propriate. He quotes Malik as saying that "there is not one among us 
who has not rebutted or been rebutted except the master of that grave 
[i.e., the Prophet] (s)."'26 

Between al-Khatib al-Baghdadl's invocation of the notion that no 
book except the Qur'an is above criticism and al-Albanl's repetition of 
this mantra almost a thousand years later, we see a continuous strain of 
iconoclastic hadith scholarship that survived alongside the burgeoning 
canonical culture of the Sahihayn. The work of al-Daraqutnl before 
the canonization of the Sahihayn, and of al-Mazarl, al-jayyani and 
Ibn Abl al-Wafa' after it, represents the continued application of the 
critical methods of hadith scholarship despite the protective culture 
constructed around the icons of al-Bukharl and Muslim. Those scholars 
who elaborated and defended the canonical culture did so because they 
believed that the canon fulfilled certain crucial purposes in the scholarly 
community. Iconoclastic hadith scholars like Ibn al-Murahhal and Ibn 
Abl al-Wafa' did not concede to prioritizing the canonical culture above 
the critical standards of hadith criticism. 



Ibn Hajar, al-Nukat 'ala kitdb Ibn al-Saldh, 158. 
Al-Jaza'irl, Tawjih al-na^ar ild usul al-athar, 1:331-2. 
Al-Albanl, Sahih al-Targhib wa al-tarhib, 1:25. 



334 CHAPTER EIGHT 

Yet, if criticism of the Sahihayn canon was not novel, why do vocifer- 
ous condemnations of these critiques only begin in the early modern 
period? In the case of Shah Wall Allah, defending the canon was an 
act of protecting and consolidating the truly unifying institutions of 
Islam in the besieged and beleaguered Indian subcontinent. Possibly 
in the work of Shah Wall Allah, and certainly in the case of the Mad- 
hhab Traditionalists, we see that the Sahihayn serve as proxies for the 
institutions of classical Islamic scholarship. The Sahihayn canon was 
both a product of and a response to the needs of the Sunni legal and 
theological schools as they solidified in the fifth/ eleventh century. The 
authority of al-Bukharl and Muslim rested on the power of ijma\ The 
Madhhab Traditionalists' categorical rejection of criticizing al-Bukhari 
and Muslim stemmed from their perception that an attack on the two 
books was a manifestation of the Salafl attack on consensus, scholarly 
hierarchy and even the valuated notion of time itself This dimension 
of criticizing the canon only appeared with the tremendous wave of 
revival and reform movements in the eighteenth century and the con- 
comitant reemergence enforce of the iconoclastic Salafl strain of hadlth 
scholarship with men like al-San'anI and al-Albam. Only in response 
to the unprecedented threats they posed to the unifying institutions of 
classical Islamic religious culture did these increasingly beleaguered 
institutions find it necessary to defend themselves. 



CHAPTER NINE 

CANON AND SYNECDOCHE: 
THE SAHlHATNm NARRATIVE AND RITUAL 



Introduction 

So far, we have discussed the Sahihayn canon as a practical and power- 
ful tool of scholarly debate and exposition. It is the kanon of truth, the 
measure of authenticity through which the redemptive media of the 
Prophet's legacy can be applied decisively It is the authoritative refer- 
ence and exemplum that can be invoked to set the rule of a genre. 
Yet to remain focused solely on jurisprudence or the study of hadlth 
inexcusably limits the role of the Prophet's sunna in Muslim life. It 
ignores important dimensions of how text, authority and communal 
identification can interact through the medium of the Prophet's charis- 
matic legacy. Our view has also been limited to the form of canonicity 
that Sheppard and Folkert conceived of as a criterion of distinction 
(Canon 1). As we widen our lens beyond the scholarly world, we must 
examine what functions al-Bukhari and Muslim fulfilled in their capacity 
as Canon 2: a fixed collection and delimited set of texts.' 

The Prophet's persona has cast a commanding shadow in Islamic 
civilization, but it has often remained intangible. In the centuries 
after their canonization, the Sahihayn would thus meet a pressing need 
beyond their strictly scholarly functions: that of a trope representing 
the Prophet's legacy in the broader Sunni community. In both the 
realms of ritual and the construction of historical narrative in Islamic 
civilization, al-Bukhari and Muslim would symbolize the Prophet's role 
as the pure wellspring of the faith and the liminal point through which 
his community could access God's blessings. The two works would be 
the part that symbolized and essentialized the whole, a synecdoche for 
Muhammad himself 

As a literary trope, synecdoche closely resembles metonymy, or the 
replacement of one word with another because of some common 



Sheppard, "Canon," 66; Folkert, "The 'Canons' of 'Scripture,'" 173. 



336 CHAPTER NINE 

association between them. Scholars like Hayden White, however, have 
distinguished between metonymy's function as a part representing the 
whole and synecdoche's function as a part essentializing it.^ 'Fifty sails' 
indicates fifty ships metonymically but the synecdoche of 'the English 
Crown' is the part of the royal person that essentializes the power and 
sovereignty of the British state. Due to the tremendous veneration that 
the Sahihayn had earned in Sunni Islam as the most authentic reservoirs 
of the Prophet's legacy, they were ideally suited to essentialize it. 



Delimiting the Infinite: Managing the Sunna through the Hadith Canon 

As Norman C alder observed, "One feature of Muslim tradition is that 
it acknowledges an indeterminately large body of hadith literature."' 
The Prophet's oral legacy within his community is amorphous and 
boundless, subsuming a seemingly infinite number of reports rang- 
ing from the most well-authenticated hadiths to common household 
sayings popularly attributed to the Prophet. As al-Shafi'l noted in the 
second/eighth century and Ibn Taymiyya emphasized at the turn of 
the seventh/thirteenth, any claim to have encompassed all the extant 
hadiths attributed to the Prophet was absurd.* In order to fulfill its 
important role in society, ritual and law, the Prophet's sunna thus 
needed to be contained in a manageable form. It is in this capacity 
that the Sahihayn canon, and the Sunni hadith canon as a whole, has 
served admirably. 

To the extent that there existed a simple need for some sort of 
synecdochic delimitation, the Sunni hadith canon has been relatively 



^ Hayden V White, Metahistory: the Historical Imagination in Mineteenth-Century Europe 
(Baltimore: Johns Hopkins University Press, 1973), 31-34. 

' Norman Calder, "The Limits of Islamic Orthodoxy," in Intellectual Traditions in 
Islam, ed. Farhad Daftary (London: LB. Tauris, 2000), 75. See also, Weiss, The Search 
for God's law; 260, 266; Wheeler, Applying the Canon in Islam, 59. 

' Muhammad b. Idrls al-Shafi'l, al-Risdla, ed. Ahmad Shakir (Beirut: al-Maktaba 
al-'Ilmiyya, [n.d.], a reprint of the 1940 Cairo edition), 42-3; Ibn Taymiyya, Rqf'al- 
maldm 'an al-a'imma al-a'ldm, ed. Muhyl al-Din Khatib (Cairo: al-Maktaba al-Salafiyya, 
1387/[1967]), 4. Al-SuyutI (d. 911/1505) claimed that all the hadiths of the Prophet 
could be encompassed by amalgamating all the collections of hadith — a task he 
attempted in his massive al-Jdmi' al-kabir. Later scholars, however, such as Abu 'Ala' 
al-'iraqi al-FasI (d. 1770-1) added over 5,000 hadiths that al-Suyup had missed in 
his mega-collection; al-Sujuii, Jam' al-jawdmi', 1:1-2; 'Abd al-Hayy b. al-Siddlq al- 
Ghumarl, Iqdmat al-hujja 'aid 'adam ihdtat ahad min al-a 'imma al-arba 'a bi'l-sunna, (unpub- 
lished manuscript), 15. 



CANON AND SYNECDOCHE 337 

elastic. Beyond the Sahihayn, we thus find common references to the 
canonical units of the Five or Six Books. Any delimited unit could 
theoretically stand in for the Prophet's sunna as a whole. When the 
great Ilkhanid vizier and historian Rashid al-Din (d. 718/1318) sought 
to properly honor God's revelation and the sunna of the Prophet in one 
of his pious endowments, he ordered the custodians of his mosque to 
produce one copy of the Qur'an and one copy of Ibn al-Athlr's Jam' 
al-usulfi ahadith al-rasul (Compendium of the Texts of the Prophet's 
Hadlths) every year'^ Rashid al-Dln's reason for choosing the Qjur'an 
for this purpose is obvious, but why did he select Ibn al-Athir's Jamr 
al-usul? The minister must have felt that the work, which condenses the 
hadlths from the Sahihayn, the collections of al-Tirmidhi, al-Nasa'T, Abu 
Dawud and Malik, effectively symbolized the Prophet's legacy and was 
the proper counterpart to God's revealed word. Earlier, the Alexandrian 
hadith scholar Abu Tahir Ahmad b. Muhammad al-Silafi (d. 576/ 1 180) 
had equated the Prophet's legacy synecdochically with the Five Books 
of al-Bukhari, Muslim, Abu Dawud, al-Nasa'l and al-Tirmidhi. He 
stated that those who opposed [mukhalif) these five books on which the 
umma had agreed opposed the Prophet himself and are like Islam's 
adversaries in Christian and pagan lands [ddr al-harb)f' 

For al-Silafi, these five books symbolized the Prophet's very words 
and the normative legacy that bound the Sunni community together To 
disagree with their status was thus to forgo membership in the Prophet's 
umma. In al-Silafi's statement, we can clearly perceive the unambiguous 
role that this set of authoritative texts played in defining the boundar- 
ies of the orthodox community. Like Moshe Halbertal's "text centered 
communities," the borders of al-Silafi's Abode of Islam (Ddr al-Isldm)' 
"are shaped in relation to loyalty to a shared canon.'" 



^ Rashid al-Din stipulated that the two books then be placed between the pulpit 
and the prayer niche {mihrdb) and that an invocation be said for him, so that he might 
receive blessings for all those who benefited fi'om them; Rashid al-Din, Vaqfhdme-ye rob '-e 
rashidi: al-vaqfiyya al-rashidiyya be-khatt al-vdqej fi baydn shard'et omur al-vaqf wa al-masdref 
(Tehran: Ketab-khane-ye Melll, 1350/ [1972]), 167. 

^ Abu Tahir Ahmad b. Muhammad al-Silaft, "Muqaddimat al-hdfig, al-kabir Abi Tdhir 
al-Silafi" in al-Khattabi, Ma'dlim al-sunan, 4:362. 

' Halbertal, 129. We should note that this synecdochic use of a hadith collection 
to represent the Prophet himself was not strictly limited to the Sahihayn or canons in 
which the two books formed the core. Abu 'Isa al-Tirmidhi (d. 279/892), for example, 
is reported to have said that if you had his Jdmi' in your house, it is as if the Prophet 
himself was speaking in your home. Such claims, however, have been rare; the vast 
majority of synecdochic representations of the Prophet's sunna have centered on the 
Sahihayn or one of the two books; al-Dhahabi, Tadhkirat al-hujfaz, 2:155. 



338 CHAPTER NINE 

Synecdoche in Ritual: Usage of the Sahihayn Canon in Ritual Contexts 

Having been endowed with a substantial religious authority in the fifth/ 
eleventh century, al-Bukhari's and Muslim's collections were well situ- 
ated to dramatize religious meaning. The SaMhoyn canon has thus found 
plentiful usage in the realms of political, calendrical and supplicatory 
rituals. How would these two voluminous hadlth books, however, be 
employed in a ritual setting? Kendall Folkert insightfully identified 
the two manners in which a canonical text can serve as a vehicle for 
meaning in ritual. First, a canonical text can function as a collection 
of scriptures accessed during the ritual. Second, the physical text of 
the canon can function as an actual participant in ritual. In this case, 
rather than just being a storehouse of authoritative writings, the canon 
can actually serve as a carrier of that authority in physical space. In 
addition to the contents of the books per se, the book itself can wield 
power as a symbol or icon." Reading al-Bukhari's SaMh over a sick per- 
son to heal him involves the first function of the canon; the contents 
of the book provide some communion with a higher power and access 
to God's blessings. An army carrying al-Bukhari's collection before it 
like an ark, however, utilizes the second mode of canonical function; 
the physical book is a central participant in the ritual. 

When used in the first mode, the Sahihayn have served as scripture 
in public or private readings. Reading a book in public has long been 
the centerpiece of the Islamicate educational and collective religious 
experience. Just as Halbertal describes the Jewish text-centered com- 
munity. Islamic religious books have been "a locus of religious experi- 
ence" whose readings have constituted "a religious drama in and of 
itself"" As Michael Chamberlain and Jonathan Berkey have shown 
in their studies on knowledge and society in medieval Damascus and 
Cairo respectively, the public reading of books was one of the main 
forms of cultural production in the Islamicate world.'" Even today 
in madrasah from Morocco to Indonesia, students gather to hear their 
teacher read a text or comment on a senior disciple's [sdrid) reading." 



" Folkert, "The 'Canons' of 'Scripture,'" 178. 

^ Halbertal, People of the Book, 7-8. 

'" Michael Chamberlain, Knowledge and Social Practice in Medieval Damascus, 1 190-1350 
(Cambridge: Cambridge University Press, 1994), 136;Jonathan Berkey, The Transmission 
of Knowledge in Medieval Cairo (Princeton: Princeton University Press, 1992), 210 IF, 

" See, for example. Dale E Eickelman, "The Art of Memory: Islamic Education 
and its Social Reproduction," in Comparing Muslim Societies, ed. Juan R.I. Cole (Ann 
Arbor: University of Michigan Press, 1992). 



CANON AND SYNECDOCHE 339 

At Friday prayers or lessons convened in the mosque for the general 
public, a professional reading of the Qur'an, hadlth or pietistic texts 
serves as the crux of the performance or lesson. Books could also be 
read in private settings, either by individuals, in the households of 
notables or in the palaces of rulers for the sake of private appreciation 
or exclusive access to blessings. 

But the Sahihayn are not works of creative scripture, narrative or 
liturgical prose. They are essentially synecdochic segments cut out of 
the endless continuum of the Prophet's sunna, discrete instances of 
his normative legacy selected and arranged by al-Bukhari or Muslim. 
Consisting of page after page of Prophetic hadiths with rare com- 
mentary, there is little beyond the editorial choices of the two scholars 
to provide any tangible notion of authorship. To read the Sahihayn is 
to read a synecdoche of the Prophet's legacy, the value of which has 
been assured by the two great canonical figures of the Sunni hadith 
tradition. 

Although the Sahihayn could represent the sunna in a manageable 
form, the two works are nonetheless massive. Even professional hadith 
scholars like al-Khatib al-Baghdadi who devoted themselves to cease- 
less study sessions of al-Bukharl's work required at least several days 
to complete hearing the collection from a teacher'^ As a result, public 
readings of al-Bukhari's or Muslim's works could take a more accessible 
private-public form, with a select group of religious devotees gathering 
in a mosque or Sufi lodge to read the bulk of the text and the general 
public participating only in the culmination (khatm) of the book. '^ Just 
as the congregation attending the nightly reading of the Qjur'an during 
Ramadan swells at the khatm of the holy book on the twenty- seventh 
night of the month, the putative Night of Power, so too the khatm of 
a Sahih was the public ritual focus of its reading. As a result, from the 
late 800s/ 1400s we see a proliferation of books on performing the 
khatm of the Sahihayn and other major hadith works as well as providing 
vignettes about the lives of their authors, such as that of 'Abd al-Salam 
b. Mahmud al-'AdawI (d. 1033/1623) on al-Bukharl's collection and 
that of al-Sakhawi on Abu Dawud's or al-Nasa'l's Sunans.'* 



'2 Al-Dhahabi, Tadhkirat al-huffaz, 3:222. 

" See, for example, Yusuf al-KattanI, Madrasat al-BukhdnJi al-Maghrib, 2 vols. (Beirut: 
Dar Lisan al-'Arab, [198-]), 2:549. 

'^ Sezgin, Geschichte des arabischen Schnfttums, 1 : 1 30; al-SakliawI, Badhl al-majhudfi khatm 
al-Sunan li-Abi Ddwud, ed. 'Abd al-Latif al-Jllani (Riyadh: Adwa' al-Salaf, 2003); idem, 
Bughyat al~rdghib al-mutamanni ft khatm al-Nasd'i. 



340 CHAPTER NINE 

Let US now examine the three main vectors of ritual activity that have 
employed the Sahihayn: supplicatory, calendrical and political. In all three 
cases, ritual use of the Sahihayn seems to have begun in force during 
the seventh/thirteenth and eighth/fourteenth centuries, approximately 
two to three centuries after their canonization. There is scant evidence 
of ritual usage for the two books in sources covering the earlier period 
between the careers of al-Bukhari and Muslim and the late sixth/twelfth 
century, like al-Khatib's Tarikh Baghdad, 'Abd al-Ghafir al-Farisi's Tdnkh 
Naysdbur, Ibn al-Jawzi's al-Munta^am or 'Abd al-Karim b. Muhammad 
al-RafiYs (d. 623/1226) al-Tadwinfi akhbdr Qazwtn. It is not completely 
clear why ritual use of the Sahihayn began in this period, but exploring 
the nature of their usages may offer explanations. 

a. Supplicatory and Medicinal Rituals 

Supplicatory rituals are rites through which people call on the super- 
natural for assistance. This genre of ritual activity overlaps with rituals 
of exchange and communion, in which humans undertake an act in the 
hope or expectation that the supernatural will reciprocate.'^ Employing 
the S^ihihoyn canon in supplicatory or medicinal rituals seems to be the 
earliest ritual usage of the two books. This role of the books followed on 
the heels of the ritual attention paid in particular to al-Bukhari's grave, 
which became a locus for intercession and miracles within a century of 
his death, as the Tdnkh Samarqand of 'Abd al-Rahman b. Muhammad 
al-Astarabadhl (d. 405/1015) informs us."' The Andalusian muhaddith 
Abu 'All al-jayyani (d. 498/1105) recounts that one Abu al-Fath Nasr 

b. al-Hasan al-Samarqandl (fl. 470/1080) visited him in Valencia in 
464/1071-2 and described how the people of Samarqand had been 
afflicted by a terrible drought. This was alleviated only when the people 
of the city went to al-Bukhari's grave and invoked God's mercy. ' ' 

An unusual ritual usage seems to have appeared for Muslim's &M 
in the early sixth/twelfth century, when it became the vehicle for an 
apparently isolated ordeal of mourning. When the son of the scholar 
Abu al-Qasim Isma'll b. Muhammad al-Taymi (d. 535/1140-1) died, 
he buried him and then read Sahih Muslim by his grave in Hamadhan. 



'^ Catherine Bell, Ritual: Perspectives and Dimensions (New York: Oxford University 
Press, 1997), 108. 

'^ Cited from al-SaghanI, ^^amz, 1—2. See Chapter 7, n, 41. 

" Al-Dhahabi, Tdnkh al-isldm, 19:273-4; cf al-QastallanI, Irshdd al-sdn, 1:29. 



CANON AND SYNECDOCHE 341 

In an act reminiscent of a ritual rejoining of the community after a 
transitional ordeal, the day al-Tayml finished his reading he set up a 
large table with sweets and food and invited all his friends to join him 
in a feast.'" We have no other evidence, however, of the Sahihayn being 
used in this manner 

By the 700s/ 1300s al-Bukharl's Sahih had become a well-known tool 
for people seeking God's intervention in times of illness and hardship 
within the cultural orbit of Mamluk Egypt and Syria. The Damascene 
Taj al-Din al-Subki (d. 771/1370) notes that the book was "a refuge 
from predicaments (mu'dilat) and well-tried for responding to needs," 
adding that "this is a well-known matter, and if we were pushed to 
mention all this and what occurred with it, the explanation would 
be too lengthy.""' In 790/1388, one of the many instances in which 
the bubonic plague struck Cairo, the Shafi'i chief judge ordered al- 
Bukharl's work read in the Azhar Mosque as a plea for relief When 
the plague continued, he ordered it read again two weeks later in the 
Mosque of al-Hakim. In a final, desperate petition for divine succor, the 
judge convened a reading three days later in the Azhar Mosque with 
orphaned children in attendance.^" Muhammad b. Yusuf al-Kirmani 
(d. 786/1384) explains that he decided to write his onamasticaUy focused 
commentary on al-Bukhari because "a certain sultan from an impor- 
tant Muslim land (ba'd ummahdt bildd al-Isldmj" (probably the Mamluk 
sultan) fell ill and wanted al-Bukhari's work read over him so that its 
blessing [barakd) might cure him. The scholars charged with the reading, 
however, could not confidently read the isndd'i without stumbling over 
the unvoweled names of the transmitters.^' The Cairene Ibn Hajar 
al-'Asqalani reported that his teacher Abu Muhammad 'Abdallah b. 
Abl Hamza was told by a "mystic {'drif)" that "Sahth al-Bukhdn has not 
been read in a time of severity except that this has been relieved, nor 
[has anyone who read it] when embarking a ship [had that] ship sink." 
He adds that Ibn Kathir says that al-Bukharl's collection can be read 
as an invocation for rain {istisqd')P 



'" Al-Dhahabi, Tadhkirat al-hujfaz, 4:51. 

'' Al-Subkl, Tabaqdt, 2:234. 

^" TaqI al-Din Ahmad b. 'All al-Maqrlzi (d. 845/1441), Kitdbal-suluk li-ma'rifat duwal 
al-muluk, ed. Sa'ld Abd al-Fattah 'A.shur, 1 1 vols, in 4 (Cairo: Matba'at Dar al-Kutub, 
1970), 3:2:577. 

^' Al-KirmanI, al~Kawdkib al-dardn, 1:5. 

^^ Ibn Hajar, Hadj al-sdri, 14; al-QastallanI, Irshdd al-sdn, 1:29. 



342 CHAPTER NINE 

In the Ottoman Hijaz, the Hanafi emigre from Herat, MuUa 'All 
Q_ari (d. 1014/1606), tells us that al-Bukharl's Sahih had been dubbed 
"the well-tried antidote [al-tiryaq al-mujarrab)." He quotes one Sayyid 
Asil al-Din as saying, "I have read al-Bukhdn one hundred and twenty 
times for events [waqaV) and important tasks [muhimmat) of mine and 
of others, and the desired result occurred and the needs were met. . . ."^' 
The reputation of al-Bukharl's Sahih had spread as far as India in the 
eighteenth and nineteenth centuries. Shah Wall Allah's son. Shah 'Abd 
al-'Aziz (d. 1824), says that reading the work in times of severity, fear, 
illness, famine or drought "is a tried and tested cure."^* 

There is much less evidence for widespread use of Muslim's book 
in medicinal or supplicatory rituals. Nonetheless, the collection did 
attain at least a portion of the fame of its more illustrious counterpart. 
The famous Central Asian hadith and Qjur'an scholar Muhammad b. 
Muhammad Ibn al-Jazari (d. 833/1429), for example, read part of 
Muslim's Sahih at Muslim's grave for barakaP 

b. Calendrical Rituals 

Calendrical rituals impose a framework of human significance on the 
abstract dimension of time or the endless cycles of nature. In general, 
such rituals are either based on the seasons or on commemorating 
important moments in a community's collective experience. In the 
Islamic calendrical system, where the calendar year has been deliberately 
severed from the solar year and planting seasons, religious holidays serve 
as anchors in the Muslim sense of time. The month of Ramadan and 
the Night of Power are thus two markers of the Islamic year^'' As we 
shall see, a three-month reading of the Sahihayn would also effectively 
create a ritual 'season.' 

The use of the Sahihayn in calendrical rituals seems to have begun 
slightly later than the books' supplicatory role. From the available 
evidence, it seems that around the early eighth/fourteenth century 



2' MuUa 'All Qarl, Mirqat al-majatih, 1:13. 

^* Shah 'Abd al-'AzIz al-DihlawI, Bustdn al-muhaddithin, lb. 

^^ Abu Muhammad 'Abdallah b. Muhammad Yusuf Efendizade, " 'Inayat al- 
malik al-mun'im li-sharh Sahih Muslim," MS 343—5 Hamidiye, Siileymaniye Library, 
Istanbul: 1:3b. 

2« Bell, Ritual, 103. 



CANON AND SYNECDOCHE 343 

al-Bukhari's book, and to a lesser extent Muslim's, was being read 
in mosques to mark the consecutive months of Rajab, Sha'ban and 
Ramadan, climaxing with the celebration at the end of the holy month. 
In Cairo, the Mamluk sultan al-Zahir Barquq (d. 801/1399) hired a 
scholar to read the Sahihayn in his newly founded Zahiriyya Mosque 
during Sha'ban and Ramadan.^' In 1515 CE, the madrasa of al-Sayfl 
Baybars was founded in Cairo and a scholar was hired specifically to 
read Sahih al-Bukhdri during Rajab, Sha'ban and Ramadan. ^^ 

Even in the far-flung Songhay empire of Mali, with its grand mud- 
built capital at Timbuktu, 'Abd al-Rahman b. 'AbdaUah al-Sa'di (d. after 
1065/1655-6), an imam in Jenne and administrator in Timbuktu, tells 
us that the Sahihayn were read in mosques during these three months. 
This is not surprising, since Mali's scholars traveled and studied in 
the Maghrib, Egypt and the Hijaz, bringing ritual practices back with 
them. Ahmad b. Ahmad Aqit of Timbuktu (d. 991/1583) recited 
the SaMhoyn during Rajab, Sha'ban, and Ramadan annually for over 
twenty years. ^^ His contemporary, the hadith scholar Ahmad b. al-Hajj 
Ahmad b. 'Umar, was also known as "the reciter of the two Sahih's, in 
the Sankore mosque."^" Across the vast dune sea to the northwest, an 
anonymous mid-ninth/fifteenth-century scholar in Marrakesh would 
read al-Bukhari's Sahih to the descendents of the Prophet in the city 
during Ramadan.^' 

Even in Syria in the late 1800s, al-Bukhari's SaMh was read in the 
Nasr Dome of the Umayyad Mosque in Rajab, Sha'ban and Rama- 
dan with great attendance and fanfare. ^^ In Morocco during the same 
period, main mosques and Sufi lodges began reading the SaMh in Rajab, 
continued through Sha'ban and finished on the Night of Power in 
Ramadan.''^ Al-Bukhari's collection was also read on other important 
religious occasions. In 1119/1707—8, for example, 'Abdallah b. Salim 



^' Berkey, The Transmission of Knowledge in Medieval Cairo, 213. 

^° Berkey, The Transmission of Knowledge in Medieval Cairo, 17, 75. 

^^ John O. Hunwick, Timbuktu and the Songhay Empire: al-Sa 'di's Ta nkh al-suddn down to 
1613 and other Contemporary Documents (Leiden: Brill, 1999), 61. For more on scholars in 
Timbuktu, see Elias N. Saad, Social History of Timbuktu: the Role of Muslim Scholars and 
Notables 1400-1900 (Cambridge: Cambridge University Press, 1983), 58-126. 

^" Hunwick, Timbuktu and the Songhay Empire, 46. 

" Hunwick, Timbuktu and the Songhay Empire, 69-70. 

'^ Commins, The Salafi Reform Movement in Damascus, 57-8. 

'' M-Knttam, Madrasat al-Bukhdrif al-Maghrib, 2:544-5. 



344 CHAPTER NINE 

al-Basri (d. 1722) was assigned to read the work at the Grand Mosque 
in Mecca upon its renovation by the orders of the Ottoman Sultan 
Ahmad III.'* 

c. Political Rituals 

One of the most dramatic usages of the Sahihayn canon has been in 
the realm of political ritual, which generally serves two primary func- 
tions. First, rites of political ritual create a sense of coherence and 
common order among a collectivity of people. Second, they legiti- 
mize this sense of political community by establishing a link between 
it and the higher orders of the cosmos.'^'' The usage of the Sahihayn 
in political ritual seems to have begun in the seventh/thirteenth and 
eighth/fourteenth centuries in Mamluk Egypt and Syria. The Mamluk 
army that marched out of Cairo against the Ilkhanid Mongols at the 
beginning of the eighth/thirteenth century was led by a person carry- 
ing Sahih al-BukhdnJ"^' Ibn Kathir says that in Sha'ban 766/1365, when 
the amir Sayf al-Din Baydar (the Mamluk sultan's erstwhile deputy in 
Syria) returned to Damascus to take up the governorship of the city, 
prominent citizens received him with a large public celebration. These 
festivities involved public readings of the final sections of al-Bukharl's 
Sahih (khatmat al-Bukhariyyat) in the Umayyad Mosque and other loca- 
tions in succession at different mosques all day. MeanwhUe Sahih Muslim 
was being read at the Hanball mihrdb at the Nuriyya madrasa near the 
Umayyad Mosque. Ibn Kathir, who was responsible for arranging all 
this, said that such an event had not taken place at any other time 
in recent years.'' When the army of the Moroccan Sa'dian dynasty 
marched out of their ochre-colored southern capital of Marrakesh to 
fight the invading Portuguese in 998/1589-90, scholars performed a 
public khatm of al-Bukharl's Sahih as the army left the gates.''' 

Perhaps the most consistently cunning exploiter of the Scihihcyn canon 
for political ritual has been the reigning 'Alawid dynasty of Morocco. 
Deriving their political legitimacy from their descent from the Prophet, 



" Voll, '"Abdallah b. Salim al-Basri and 18th Century Hadith Scholarship," 360. 
'' Bell, ffito/, 129. 

^^' J. De. Somogyi, "Adh-Dhahabi's record of the destruction of Damascus by the 
Mongols in 699-700/1299-1301," Goldziher Memorial 1 (1948): 361. 
" Ibn Kathir, al-Biddja iva al-nihdya, 14:326-7. 
'" M-KaXtani, Madrasat al-BukhdnJi al-Maghrib, 2:549. 



CANON AND SYNECDOCHE 345 

'Alawid rulers have turned to al-Bukhari's Sahih as a physical manifes- 
tation of Muhammad's legacy. The true founder of the dynasty, the 
conqueror and statesman Mawla Isma'il (d. 1727), sought to transform 
his patrimony from a family of raiders dependent on the ephemeral 
loyalties of local Berber tribes into a true state with a dependable 
standing army. He thus built up a core unit of African slave soldiers, 
originally captured in the conquest of gold-laden Timbuktu, to serve 
as the centerpiece of his army. This unit grew in size, as Mawla Isma'll 
had their sons trained by artisans and then enlisted in the ranks upon 
reaching the age of ten, until it reached the awesome size of 150,000 
men.^" Mawla Isma'il dubbed these soldiers "The Slaves of al-Bukhari 
{'Abid al-Bukhari)" for it was upon the Sahih and its representation of 
the Prophet's sunna that their loyalty to their ruler was based. The 
Moroccan archivist and historian Abu al-'Abbas Ahmad al-Nasiri 
(d. 1897) explains that in his efforts to free himself of reliance on the 
fickle loyalties of tribal forces, Mawla Isma'il gathered the leaders of 
his slave regiment around a copy of al-Bukhari's Sahih. He said: 

I and you are slaves to the sunna of the Messenger of God (s) and his 
sacred law as collected in his book (i.e., the Sahih), so all that he has com- 
manded we will do, and all that he has forbidden we will forsake, and by 
it we will fight (wa 'day hi nuqdtil). 

He then took their oaths by al-Bukharl's book. At one end of the great 
parade ground that the ruler built for his praetorian at his hilltop impe- 
rial palace in Meknes, Mawla Isma'il constructed a madrasa named after 
al-Bukharl. He ordered that copy of the Sahih on which the soldiers' 
oaths had been taken preserved there and that they carry it "like the 
Ark of the Children of Israel (tdbut bani Isrd'il)" when they went out 
on campaign.*" 

The 'Alawid dynasty has maintained the prominent place of Sahih 
al-Bukhari in political rituals. When King Hasan I came to Rabat on 
Eid al-Fitr in 1873, he ordered festivities including the reading of the 
Sahih and culminating in a large public gathering with all the city's 



'' Abu al-Qasim al-Zayyanl, al-Bustdn al-iarif fi dawlat awlad mawldya al-sharif, ed. 
Rashid al-Zawiya (Rabat: Matba'at al-Ma'arif al-Jadlda, [1992]), 1:171; Maurice 
Delafosse, "Les debuts des troupes noires du Maroc," Hesperis 3 (1923): 7-8. 

'" Abu al- Abbas Ahmad b. Khalid al-Nasirl, Mtdb al-istisqd li-akhbdr duwal al-Maghrib 
al-aqsd, ed. Ja'far al-Nasirl and Muhammad al-Nasirl, 9 vols. (Casablanca: Dar al-Kitab, 
1956), 7:58. 



346 CHAPTER NINE 

notables. The king also did this upon the completion of his royal palace 
in Rabat.*' 

The 'Alawid dynasty has relied on its claim of descent from the 
Prophet as the central pillar of its political legitimacy in Morocco. Basing 
the esprit de corps of his praetorian on al-Bukharl's Sahih and maintaining 
the collection as the unit's mascot reinforced Mawla Isma'll's chosen 
role as heir to the Prophet's political authority. The Sahth's ability to 
stand in for the Prophet's persona in ritual, literally carried before the 
king's advancing army, was central to the logic of this political ritual. 
Similarly, the esteemed station of the Sahihayn allowed Ibn Kathir to 
help transform the arrival of the Bahrl Mamluk governor in Damascus 
into an evocation of religious significance. 



The Ritual Power of the Sahihayn: The Muhammadan Blessing 

In Islam, God is the source of all baraka, or what Josef Meri calls "the 
stuff of faith. "*^ It is the blessing by which men's felicity is ensured in 
both the earthly life and the hereafter Proximity to God through either 
piety or some link to a liminal figure entails greater access to His baraka.*^ 
As the receptacle of revelation and the bridge between the divine and 
the temporal, the Prophet is the ultimate liminal figure in Islam. As the 
perfect human, possessed of "tremendous character (Qur'an 68:4)," and 
on whom God and the angels "shower their prayers (Qur'an 33:56)," 
the figure of Muhammad has enjoyed the greatest access to baraka. His 
persona is the most completely endowed with "the capacity to mediate 
between humanity and the Deity"** Imitating his lifestyle and obeying 
his commands as embodied in the Sharia enables Muslims to approach 
this locus of God's blessings. Gaining physical or aural proximity to the 
Prophet's words, his relics or members of his family provides extended 
access to his liminality*'' Similarly, pious individuals who have themselves 



*' Ai-KattiLTu, Madrasat at-BukhanJi al-Maghrib, 2:547. 

■'^ Josef W. Meri, The Cult of Saints among Muslims and Jews in Medieval Syria (Oxford: 
Oxford University Press, 2002), 17. 

" See G.S. Colin, "Baraka," EF\ Ernest Gellner, Saints of the Atlas (London: Wei- 
denfeld and Nicolson, 1969), 12. 

■''' Gellner, Saints of the Atlas, 70. 

*" For a discussion of the salvational role of the Prophet and his family in Egyptian 
popular Sufism, see Valerie J. Hoffman-Ladd, "Devotion to the Prophet and His Family 
in Egyptian Sufism," International Journal of Middle East Studies 24 (1992): 617. For a 



CANON AND SYNECDOCHE 347 

earned a station close to God and His blessing themselves become loci 
of liminality and baraka for others. 

Like saints, who wield extraordinary powers through their proximity 
to God, books enjoying such proximity are also a "nexus of baraka, 

miracles and mediation "*'' Michael Chamberlain describes religious 

knowledge {'ilm) as a source of blessing {baraka) that Muslims of all 
social standings tried to acquire.*' The pursuit and study of 'ilm was 
thus a ritual practice, equated with forms of worship such as ritual 
remembrance of God {dhikr) and canonical prayer, and thus requiring 
the same levels of ritual purity. Acquiring knowledge was a "collective 
liminal experience" in which the attempt to grasp and appreciate God's 
will brought the audience closer to Him.*" 

Reading or listening to a performance of a hadith collection was 
thus to increase one's proximity to God's blessings as deposited and 
dispensed through His Prophet. AsJ.Z. Smith states, "Ritual is, first and 
foremost, a mode of paying attention. It is a process for marking inter- 
est."*^ In the ritual logic of the audience, reading Muhammad's words 
is to give his person and legacy attention. To consider his example is 
to please God as the Prophet had pleased Him and incur that blessing 
that God showered upon him. It is to walk that path of liminality. The 
ritual of listening to or acting on a hadith becomes a metaphoric act 
of accessing the blessings the Prophet enjoyed."'" 

The conspicuous Muslim habit of calling God's peace and blessings 
down upon the Prophet after every mention of his name in either writ- 
ten or oral expression emphasizes the role of the Prophet as a channel 
for access to God's baraka. One widely cited hadith states that "whoever 
prays upon me once, God prays upon him ten times."'"' In activities such 



discussion of the role of the descendents of the Prophet ligurram) among Berbers in 
Morocco, see Gellner, Saints of the Atlas, 70-80. 

"• Taylor, In the Vicinity of the Righteous, 121 . 

" Chstmherlain, Knowledge and Social Practice, 122. 

*° Chstmherlain, Knowledge and Social Practice, 127-9. 

" Jonathan Z. Smith, To Take Place: Towards Theory in Ritual (Chicago: University of 
Chicago Press, 1987), 103. 

^^ See Edmund Leach, Culture and Communication (Cambridge: Cambridge University 
Press, 1976), 30 ff. 

^' Sahih Muslim: kitdb al-saldt, hdb 17; Sunan Abi Ddwud: kitdb al-witr, bah 26; Sunan 
al-Masd'i: kitdb al-sahw, bdb 55; Jdmi' al-Tirmidhi: kitdb al-saldt, bdb 240; Sunan al-Ddrimi: 
kitdb al-raqd'iq, bdb 58. Al-Tirmidhl's citation of the hadith is followed by the earhest 
occurrence I have found of the explanation, here attributed to Sufyan al-Thawrl, that 
God's 'prayer' upon mankind is mercy {rahmdj, while that of the angels is 'seeking 
forgiveness [for mankind] (istigkfdr).^ Other, more unusual reports on this issue include 



348 CHAPTER NINE 

as the Sunni canonical prayer, in fact, invocations for the Prophet's sake 
equal or supersede the performer's set prayers for himself or herself 
Here Muhammad becomes a proxy for the believer's own personal 
invocations. The Egyptian Shafi'i al-Sakhawi (d. 902/1497) notes that 
the purpose of such intense prayer on the Prophet is "growing close 
to God most high by imitating His act [of blessing the Prophet] and 
fulfilling the right due the Prophet (s)." Al-Sakhawl quotes one Abu 
Muhammad al-Marjani as saying, "In calling your prayers on him [the 
Prophet], you are, in truth, because of the benefits that these prayers 
return to you, praying for yourself"'^ The benefits of calling God's 
peace and blessings down upon the Prophet extend to the scholarly 
realm of those who write books in addition to their audiences. Abu 
Tahir al-Silafi mentions a hadith that guarantees baraka for an author 
who writes "may the peace and blessings of God be upon him" after 
the Prophet's name. The hadith states that "whoever prays [salla 'alayyd) 
for me in a book, angels will continue to pray for him as long as my 
name is in that book."'^^ 

In ritual, the Sahihayn thus act synecdochically as a channel for God's 
blessings as transmitted through the Prophet. The Mamluk sultan whom 
al-Kirmani mentioned as having fallen ill hoped the baraka of Sahih al- 
Bukhdn would cure him."'* We find in the letter of the Moroccan scholar 
'Abd al-Kabir b. Muhammad al-Kattani (d. 1914-5) instructions to 
read through al-Bukharl's Sahth in mosques and houses in order to get 
the "Muhammadan intercession {al-shafa'a al-muhammadiyyd).^^ MuUa 
'All Qari quotes Sayyid AsTl al-Din as crediting the miraculous pow- 
ers of the Sahih "to the barakat of the most noble of the nobles (the 
Prophet) and the source of felicity, may the most favored prayers and 
most perfect greeting be upon him."-^'' 



one attributed to Abtl Bakr that "prayer upon the Prophet eliminates sins more than 
water does fire...;" al-Khatlb, Tdnkh Baghdad, 7:172. 

^^ Al-Sakhawl, al-Qawl al-badi' fi al-saldt 'aid al-habih al-shaji' (Beirut: Matba'at al- 
Insaf, 1383/1963), 25. "Indeed God and His angels pray upon the Prophet; O you 
who believe shower prayers and blessings upon him (Qur'an 33:56)." 

^' Al-Silaft, al-Wajiz fi dhikr al-majdz iva al-mujiz, ed. Muhammad Khayr al-Biqa'l 
(Beirut: Dar al-Gharb al-Islami, 1411/1991), 95. 

^* Al-Kirmani, al-Kawdkib al-dardri, 1:5. 

^^ Al-KdiXam, Madrasat al-BukhdnJi al-Maghrib, 2:545-6. 

* Mulla 'All Qarl, Mirqdt al-mqfdtih, 1:13. 



CANON AND SYNECDOCHE 349 

The synecdochic function of the Sahihayn in these rites provides the 
best explanation for why ritual usage of the canon began on any appre- 
ciable scale only in the seventh/thirteenth century. Marshall Hodgson 
notes that at this time Islamicate civilization in the Nile-Oxus region 
had reached some critical distance from the faith's epicenter in the 
person of the Prophet. Society required new vehicles for bridging this 
divide and accessing the Prophet's baraka, and the seventh/thirteenth 
and eighth/fourteenth centuries thus witnessed an intensified interest in 
pilgrimages to Muhammad's grave in Medina, those of his purported 
descendents throughout the Islamic world and other local saints." The 
Sahihayn provided a textual alternative. 

The popularization of the Sahihayn in public rituals such as readings 
during Ramadan mirrors the wider popularization of communal ritual 
such as those practiced by Sufi brotherhoods, which began flourish- 
ing in their institutional tariqa form in the 600s/ 1200s."''' Similarly, 
the initiative that the Mamluk rulers took in organizing and funding 
public readers of the Sahihayn dovetails with their general sponsorship 
of popular religious practices, such as building major Sufi lodges in 
Cairo and Damascus.'^ 



The Canon and Synecdoche in Narrative: 
A Salvational Trope in a Narrative of Decline and Salvation 

Just as the Sahihayn represented the Prophet's liminality and charisma, 
granting access to the baraka to which he was the key, al-Bukhari and 
Muslim also became a synecdochic trope for scholars constructing 
narrative in Islamic history. Hadith literature is not limited to the dry 
compilation and criticism of Prophetic reports. It encompasses a network 
of genres that either orbit the collection and evaluation of reports or 
mold these activities into forms that address specific needs. Hadlth- 
oriented biographical dictionaries like Tdrikh Baghdad, works on 'Hal 
and the technical terms of hadith evaluation fit into the first category. 
The second category includes specific types of hadith collections that 



■" Hodgson, The Venture of Islam, 2:453; Taylor, In the Vicinit)) of the Righteous, 14. 
^^ J. Spencer Trimingliam, The Sufi Orders in Islam (Oxford: Oxford University Press, 
1998), 9-10; J.O. Hunwick et al, "Tasawwuf," EF. 
■^' Taylor, In the Vicinity of the Righteous, 1 2 ff. 



350 CHAPTER NINE 

could channel the Prophet's charisma through an individual scholar's 
personal religious expression. Mustakhrcy's, personal mu'jams document- 
ing all the lands to which a collector had traveled {rihld) and all the 
teachers from whom he had heard [mashyakha, barnamaj), as well as the 
great hadlth collections themselves fall into the second. Together, all 
these genres weave a meta-narrative that serves as the shared culture 
of hadlth scholars or those other Muslim sages or laity who trade on 
their domain. 

This is a romantic narrative of decline and salvation. It constantly 
replays what Marshall Hodgson called "the old man's view of history," 
in which the community seems bound inevitably towards religious and 
moral entropy but clings to a lingering hope for the survival of the 
true faith through the uniquely pious efforts of the scholar "The best 
of generations is the one in which I was sent, then that which comes 
after it, then that which follows"; this Prophetic tradition embodies the 
Sunni vision of religious history, as the Muslim community drifts farther 
and farther in time from the epicenter of the Prophet's mission. Each 
successive age after that greatest community has a more tenuous grasp 
of the Prophet's salvational message.''" 

Ibn Hibban (d. 354/965) thus complains that his surroundings were 
flooded with ever-multiplying attributions to the Prophet and dilettantes 
who could not tell authentic hadiths from forged ones.*"' His student 
al-Hakim writes in the beginning of his Ma 'rifat 'ulum al-hadtth: 

Indeed, when I saw heretical innovations in religion {bida') increasing 
in our time, and the people's knowledge of the fundamentals of the 
sunna decreasing . . . this called me to compose a small book including 
all the branches of the sciences of hadlth that students of reports might 
need ^^ 



^ See Hodgson, The Venture of Islam, 1:381; see also, Tarif Khalidi, Arabic Historical 
Thought in the Classical Period, 25; idem, "The Idea of Progress in Classical Islam," 
Journal of Near Eastern Studies 40, no. 4 (1981): 277-89. Other examples of hadiths or 
statements expressing this historical entropy include the hadlth "there will not come 
upon you a time except that the era after it will be worse {lay a 'ti 'alaykum zamdn ilia wa 
alladhi ba'dahu sharr minhu), and the statement attributed to al-Hasan al-Basri, "every 
year you (pi.) will worsen (hull 'dm tardhuluri)"; SaMh al-Bukhdn: kitdb al-fitan, bdb Idya'ti 
'alaykum zamdn ilia wa alladhi ba'dahu sharr minhu; MuUa 'All Qarl, al-Masnu' fi ma'rifat 
al-hadtth al-mawdu', ed. 'Abd al-Fattah Abu Ghudda, 6th ed. (Beirut: Dar al-Basha'ir 
al-Islamiyya, 1426/2005), 136. 

'■' Ibn Hibban, Sahih Ibn Hibbdn, 1:58. 

'•^ Al-Hakim, Ma'rifat 'ulum al-hadith, 2. 



CANON AND SYNECDOCHE 351 

In the introduction to his commentary on Sahih al-Bukhan, al-Khattabi 
says: 

I contemplated the recourse for the affairs of our time, such as the scar- 
city of 'ilm, the prevalence of ignorance [jahl), and the dominance of 
the people of religious heresies (bida'), that many of the people's affairs 
have deviated towards their different schools of thought [madhahib) and 
turned away from the holy book and the sunna. I feared that this mat- 
ter would become more severe in days to come, that knowledge will 
be more preciously rare [a'azz) due to the paucity of those whom I see 
today... attending faithfully to [hadith] and attaining a sound {sdlih) level 
of knowledge in it.'''' 

Writing over a century later in Khurasan, al-Baghawi (d. 516/1122) 
describes the crises of heresy and ignorance on his environment: "Noth- 
ing remains of the religion except its outlines {rasm), nor of knowl- 
edge except its name, to the point that falsity is considered to be the 
truth among most people in our time, and ignorance is confused with 
knowledge."" 

In the face of this decline, the struggle of the 'true Sunni schol- 
ars' to preserve the legacy of the Prophet represents the only hope 
for personal and communal salvation. One of the most frequently 
quoted hadiths in the introductions to works of hadith literature thus 
prophesies, "One party from among my umma will always stand by 
the truth unharmed by those who forsake them, until the command 
of God comes. '"■'' Ibn Hanbal is frequently quoted as identifying this 
sect with the ahl al-hadith, whom al-Hakim describes as "trumping the 
people of heresy with the sunna of God's messenger'"'*' Only by stub- 
bornly clinging to the continuous study and repetition of the Prophet's 
legacy can the hadith tradition fulfill its destiny as the sole guardians 
of Islam's pure origins. 

Moreover, it is always the author's own immediate efforts that embody 
this hope of salvation. Al-Baghawi thus offers his huge legal compen- 
dium of hadith [Shark al-sunnd) as an attempt to revive the path of the 



f'^ M-Kh&xX3b\, A'ldm al-hadith, f:f02-3. 

''' Af-Baghawl, Shark al-sunna, ed. Shu'ayb Arna'ut and Zuhayr af-ShavvIsh, f4 vofs. 
([Beirut]: af-Maktab af-Isfami, 1390/f97f), 1:3-4. 

''^ "La tazdlu td'ifa min ummati zahirin 'aid al-haqq Id yadurruhum man khadhalahum hattd 

jia'tiya amr Allah"; Sahih Muslim: kitdb al-imdra, bdb qawlihi (s) la tazdlu td'ifa For another 

version, see al-Hakim, Ma'rifat 'ulum al-hadith, 2. 

'''' Yahya b. Meinda., Juz'Jihi mandqib al-shaykh al-Tabardni, 5b (quoted from al-Hakim 
al-Naysaburl's lost Mandqib ashdb al-hadith). 



352 CHAPTER NINE 

righteous forbears who established the religion, acting as "one striving to 
light a lamp in the encompassing darkness, [so that] the perplexed can 
be guided by it or someone seeking guidance can find the path.'"'' 

The notion of the sahih movement as the pinnacle of hadith schol- 
arship, evident after the writings of Ibn Manda (d. 395/1004-5), 
provided a convenient trope in this narrative. Al-Bukhari and Muslim 
in particular came to represent the acme of critical rigor in hadith 
study. Majd al-Din Ibn al-Athir (d. 606/1210) describes how, while the 
number of hadith collections blossomed in the wake of al-Bukhari's 
and Muslim's careers, their authors were pursuing all sorts of agendas 
(aghrad, maqdsid) and the glorious age of the Shaykhayn had vanished 
{inqaradd). Even with the continued work of Abu Dawud, al-Tirmidhi 
and al-Nasa'l, it was as if the age of al-Bukhari and Muslim "was the 
sum of all ages in terms of the acquisition of that science {'Urn), and 
it ended with it. Afterwards that quest waned.'"'" 

Because they represented the pinnacle of achievement in the hadith 
tradition, the Sahihayn could serve as the perfect symbol for the Prophet's 
legacy in the narratives that scholars spun around the tension between 
the 'authentic teachings of the Prophet (sunna)' and 'heretical innova- 
tion {bid'af in Islamic religious culture. Writing within a Sunni com- 
munity that acknowledged the two works' unparalleled status, scholars 
could wield them as representations of the salvation that results from 
embracing the Prophet's authentic legacy.^" 

a. IQfdje 'Abdalldh al-Ansdn and the Beginning of Synecdoche in Narrative 

The earliest extant example of Muslim scholars utilizing the Sahihayn 
as a synecdoche for the Prophet's legacy in narrative comes from the 
fifth/ eleventh century writing of Abu al-Fadl al-Maqdisi (d. 507/1 1 13). 
His teacher in the Khurasan! city of Herat, the fierce iiber-Sunni 
Kh"aje 'Abdallah al-Ansari (d. 481/1089), cuts an interesting figure 
in Islamic intellectual history. A staunch Hanball who condemned the 
cultivation of speculative theology in a massive multivolume book, he 



''' Al-Baghawi, Shark al~sunna, 1:3—4. 

i^'" Ihn a\-AihTr, Jdmi' al-usul, 1:42. 

69 \Ye must note that al-Bukharl's and Muslim's function as a synecdoche in this 
context in no way resembles Hayden White's analysis of tropology in modernist 
European historical writing, where synecdoche describes a manner in which a historian 
can manipulate and transition between ideas. Rather, the Sahihayn were quite literally a 
synecdoche for the Prophet's authentic legacy as valuated by Sunni Muslim scholars. 



CANON AND SYNECDOCHE 353 

was also a committed Sufi wlio penned a complex work on the technical 
terminology of mysticism and the progressive stages toward complete 
consciousness of God.'" Al-Dhahabi cites an apparently lost text from 
al-MaqdisI describing the famous Seljuq vizier Nizam al-Mulk sum- 
moning Kh^aje 'Abdallah to a debate in Herat. Both the vizier and his 
master, the Seljuq sultan Alp Arslan, had arrived in Herat on a visit and 
had heard complaints from Shafi'i and Hanafi scholars about Kh™aje 
'Abdallah's intolerant iiber-Sunnism. He had stated, for example, that 
he would curse anyone who denied that God was physically above the 
earth. Nizam al-Mulk demanded that Kh™aje 'Abdallah respond to his 
detractors in a debate, and the scholar agreed on one condition: that he 
be allowed to debate his opponents only with what he had in his two 
sleeve pockets {kumm). Nizam al-Mulk asked what the pockets contained, 
and Kh™aje 'AbdaUah replied, "The Book of God," pointing to his right 
sleeve (kumm), "and the sunna of the Messenger of God," pointing to 
his left. From his right sleeve Kh™aje 'Abdallah then produced a copy 
of the Qjur'an, and from his left the Sahihayn. Al-Maqdisi continues, 
"So the vizier looked at [Kh*aje 'Abdallah's opponents], seeking a 
response, and there was no one from among them who would debate 
him in this manner."'' 

Al-Maqdisi's story makes clear use of the Sahihayn as a synecdoche 
for the Prophet's sunna. Almost a century after their canonization. 



'" See 'Abdallah al-Ansarl al-HarawI, Mandzil al-sd'irin, ed. Ibrahim 'Atvvl 'Awad 
([Cairo]: Maktabat Ja'far al-Hadltha, [1977]) and idem, Dhamm al-kaldm iva ahlihi, 
ed. 'Abd al-Rahman b. al-'AzIz al-Shibl, 5 vols. (Medina: Maktabat al-'Ulum wa al- 
Hikam, 1995). 

" Al-Dhaliabi, Tadhkirat al-huffa^, 3:250-1. It seems bizarre that someone could fit 
books as massive as the Sahihayn in their sleeve, but scholars routinely wrote out such 
books in print so small that they could fit into one volume. Even a much later hadlth 
scholar like Abu al-Hasan al-Sindl (d. 1773) used to produce one copy ofSahih al-Bukhdn 
every year in one small volume; al-Mizjajl, Nuzjiat riydd al-ijdza al-mustatdba, 262. We 
can reliably date al-Maqdisl's dramatic story to the late fifth/eleventh century when 
al-MaqdisI was writing. We should certainly not treat it as a reliable transcript of an 
historical event, however, for the iiber-Sunni al-MaqdisI shared his teacher's leanings 
and furnished a highly partisan account of the debate. Moreover, although al-MaqdisI 
himself studied with Kh"aje 'Abdallah, he reports this story second-hand through "one 
of our colleagues (ashdbind)." There is no reason to suspect that al-Dhahabi was citing 
a forged source from a later period, however, since most of al-Maqdisl's prolific oeuvre 
has not survived for our examination. This absence of evidence should therefore not 
lead us to doubt al-Maqdisl's authorship. Even if al-MaqdisI himself creatively altered 
the report of his teacher's debate, we can nonetheless still date it to his career in the 
late fifth/eleventh century. For the most comprehensive list of al-Maqdisl's works, see 
al-MaqrIzI, Kitdb al-muqaffd al-kabtr, ed. Muhammad al-Ya'lawI, 8 vols. (Beirut: Dar 
al-Gharb al-lslami, 1411/1991), 5:735-8. 



354 CHAPTER NINE 

al-Maqdisi and perhaps even Kh"aje 'Abdallah himself understood 
the symbolic power of al-Bukhari and Muslim within the wider Sunni 
community. In the face of the Hanafi and Shafi'l schools' 'heretical' use 
of reason and indulgence in speculative theology, al-Maqdisi portrays 
Kh"aje 'AbdaUah as standing by the two pure sources of the faith: God's 
revelation and its authoritative interpretation as transmitted through 
the Prophet's hadlths. The canonical text of the Qjur'an is small and 
easUy manageable. The Prophet's sunna, however, is not. Al-Bukhari's 
and Muslim's books thus serve as its commonly acknowledged physi- 
cal manifestation in the arena of debate. Just as they functioned as an 
authoritative reference and measure of authenticity, so did the Sahihayn 
serve as a symbolic convention as well. 

b. Al-Ghazalfs Return to the Straight Path: The Sahihayn as Synecdoche 

The seminal Shafi'i/Ash'arl jurist, theologian and mystic Abu Hamid 
al-Ghazall (d. 505/1111) has proven one of the most powerful and 
controversial figures in Islamic intellectual history. He became a cen- 
tral pillar of the Shafi'i/Ash'ari orthodoxy, and has been honored as 
"the Proof of Islam {hujjat al-Isldnij" by the multitude of later scholars 
who have shared his doctrinal leanings. Scholars from a wide range of 
temperaments, however, have also criticized him heavily for his laxity 
in using hadlths, his excessive mystical bent and his wholesale adoption 
of logic as a tool in Islamic thought. Al-Mazarl took al-Ghazall to task 
for attributing to saints miracles that befitted the Prophet alone. The 
Maliki Abu al-Walld al-Turtushi, who said he had met al-Ghazali, 
described him as a great scholar who had foolishly "become a Sufi, 
departing from the sciences and the scholars, entering the sciences of 
inspiration (al-khawatir), the mystics {arbdb al-qulub), and the murmurings 
of the Devil. '"^ Ibn al-Jawzi (d. 597/1200) criticized him for ignorance 
in the science of narrating hadlths and for including forged reports in 
his Ihyd' 'ulum al-dinJ'^' Ibn al-Salah (d. 643/1245) faulted al-Ghazall 
for placing logic at the forefront of the Islamic sciences as the com- 
mon language of scholarly discussion. Al-Dhahabi, who was one of 



'^ Al-Subkl, Tabaqdt, 6:243. For an excellent discussion of the controversy surround- 
ing al-Gliazall's career, see Kenneth Garden, "AI-Ghazall's Contested Revival: Ihyd' 
'ulum al-din and its Critics in Khorasan and the Maghrib," (PhD diss.. University of 
Chicago, 2005). 

" Ibn al-jawzl, al-Munta^am, 17:126. 



CANON AND SYNECDOCHE 355 

al-Ghazall's most outspoken critics, argued that his penchant for sciences 
originally foreign to Islam and straying into the realm of philosophical 
speculation plagued the scholar throughout his career'* 

In efforts to salvage al-Ghazall's image from these serious critiques, 
narrative about the scholar's life became a microcosm of the Sunni 
romance of decline and salvation. One of the earliest attempts to 
repair al-Ghazall's reputation and draw it closer to the conservative 
Sunni tradition as embodied in the study of hadith is 'Abd al-Ghafir al- 
Farisl's (d. 529/1134-5) biography of the scholar'"' A hadlth-oriented 
Shafi'T who fondly and frequently identifies with the ahl al-hadith, al- 
FarisT nonetheless evinces profound admiration for al-Ghazall. Yet his 
treatment of the great scholar, whom he had met more than once, 
focuses more on his concern for al-Ghazali's failings.'*" Struggling to 
salvage al-Ghazali's valuable works in fields such as jurisprudence and 
dogma, al-Farisi limits his critique to al-Ghazali's mystical and esoteric 
works. He states that al-Ghazali went astray from the bases of Islam in 
books like his Persian ethical treatise Kemyd-ye sa 'ddat (The Alchemy of 
Felicity)." Al-FarisI argues that he should never have entered into such 
esoteric matters because they might confuse the masses of Muslims and 
negatively affect their conception of proper belief'^ 

The chief thrust in rehabilitating al-Ghazall, however, comes at the 
end of al-Farisi's biography, where he portrays al-Ghazali as returning 
to the sound path of Sunnism and affirms his own hadlth-oriented, 
Sunni identity. Al-FarisI states that in the last years of his life, al-Ghazali 
occupied himself with study of hadith and poring over the Sahihayn. 
Had he lived longer, al-FarisI opines, al-Ghazali would have become 
the master of this noble science. Playing on al-Ghazali's honorary title, 
he adds,"It is these two [books, the Sahihayn^ that are the Proof of 
Islam [hujjat al-Isldm)."^^ 

Establishing al-Ghazali's repentance from his heretical musings in 
philosophy and Sufism by associating him with the Sahihayn became 
a central tool for rehabilitating his reputation. The Shafi'l/Ash'ari 



" Al-Dhahabl, Siyar; 19:330-1, 327-9. 

'"■ This has survived in part in an abridgement of his history of Naysabur and more 
fully in the works of Ibn 'Asakir, al-Dhahabi and al-Subkl. 

"' Ibn 'Asakir, Tdrikh madinat Dimashq, 55:202. 

' ' 'Abd al-Ghafir al-FarisI, Tdrikh Naysabur al-muntakhab min al-Siydq, 84. 

™ Al-Dhahabi, Siyar, 19:326-7. 

" 'Abd al-Ghafir al-FarisI, Tdrikh Naysabur al-muntakhab min al-Siydq, 84; al-Subkl, 
Tabaqdt, 6:210-11; Ibn 'Asakir, Tdrikh madinat Dimashq, 55:204, 



356 CHAPTER NINE 

Ibn 'Asakir of Damascus (d. 571/1176) opens his biography of al- 
Ghazall with the statement that he had heard Sahih al-Bukhdn from 
one Muhammad b. 'Ubaydallah al-Hafsi."" The Shafi'l biographer 
Abu Sa'd 'Abd al-KarIm al-Sam'ani (d. 562/1 166) of Merv included a 
report in his entry on al-Ghazall that portrays him inviting one 'Umar 

b. 'Abd al-KarIm al-RawwasI (d. 503/1 109) to stay at his house in Tus 
in order to provide extended private lessons on the Sahihayn. But even 
avid defenders of al-Ghazall, such as al-Subkl, considered this report 
to be a blatant forgery"' Al-Sam'anI most probably included it in his 
zealous efforts to affirm al-Ghazali's devotion to the hadith tradition. 
Although the Hanbali Ibn al-Jawzi is extremely critical of al-Ghazali, 
he also notes that late in life he occupied himself with learning the 
''sahih collections {al-sihdh)."^'^ The great apologist for the Shafi'i/Ash'ari 
tradition, al-Subkl (d. 771/1370), leaves us the most exhaustive defense 
of al-Ghazali's legacy in his two-hundred page biography of the scholar 
in the Tabaqdt al-shafi'iyja al-kubrd. Al-Subki's defense of al-Ghazali 
centers on the same theme advanced by al-FarisI: al-Ghazali's evident 
recantation from the unrestricted use of speculative theology in the 
last years of his life and simultaneous decision to devote himself to the 
study of the Sahihayn. The Hanafi hadith scholar and theologian MuUa 
'All Qarl provides an even more dramatic depiction of al-Ghazali's 
final return to the straight path: al-Ghazali died with copy of Sahih 
al-Bukhdn on his chest."^ 

c. Al-DhahabT's Narrative of Islamic History: The Sahihayn as Synecdoche 

The Salafi-oriented Shafi'i scholar Shams al-Din Muhammad al- 
Dhahabl (d. 748/1348) shines as one of the most intelligent and 
influential figures in Islamic intellectual history. A member of the 
remarkable Damascus circle of Ibn Taymiyya, Jamal al-Din al-Mizzi 
and Ibn Kathir, his works and those of his associates have exercised an 
inordinately powerful effect on the course of Sunni thought. Through 
his many studies on the hadith sciences and remarkable biographical 
dictionaries, al-Dhahabi elaborated an independent hadlth-oriented 
vision of Islamic history that angered more staunch devotees of the 



Ibn 'Asakir, Tankh madinat Dimashq, 55:200. 
Al-Subkl, Tabaqdt, 6:215. 
Ibn al-Jawzl, al-Muntaiam, 17:126. 
MuUa 'All Qarl, Shark al-Fiqh al-akbar, 30. 



CANON AND SYNECDOCHE 357 

legal and theological schools as much as it provided them indispensable 
benefit."* Al-Dhahabi rejected the tradition of speculative theology as 
well as what he perceived as the over-involved and self-indulgent com- 
plexities of the Sunni scholarly edifice. In his biography of al-Ghazali 
he urges a hadlth and piety-based minimalism, telling the reader that 
all a Muslim requires to attain success and salvation are the Qur'an, 
the Sahihayn, al-Nasa'i's Sunan and al-Nawawfs two pietistic works, 
Riydd al-sdlihin (The Gardens of the Righteous) and the Kitdb al-adhkdr 
(Book of Prayers)."' 

Al-Dhahabl's Tadhkirat al-hujfdz (Aide-Memoire of the Hadith Mas- 
ters) provides a concise glimpse into the scholar's conception of Islamic 
civilization's historical course. Unlike his gigantic Tdnkh al-isldm (History 
of Islam) or his expansive Siyar a'ldm al-nubald' (The Lives of the Noble 
Figures), the Tadhkira consists of only a few volumes devoted solely to 
a chronological treatment of those figures who emerged as prominent 
participants in the Sunni hadith tradition. In rare comments at the end 
of some outstanding generations, al-Dhahabi includes his own evalua- 
tions of the umma's unfolding history. At the end of the first generation 
to succeed the Companions, for example, he describes how at this time 
Islam had become powerful and glorious, "having conquered the lands 
of the Turks in the east and Andalusia in the west.""'' 

After the fifth generation, consisting of scholars like Ibn Jurayj and 
Abu Hanlfa who died between 140 and 150 AH, al-Dhahabi writes, 
"Islam and its peoples were endowed with total might and profuse 
knowledge, the standards of jihad spread wide and the sunna [sunan) 
widespread." He adds that "heresy [bid'a) was suppressed, and those 
constantly speaking the truth were many. The servants [of God] were 
plentiful in number and the people were living at the height of pros- 
perity with security ""' But after the civil war between al-Amin and 

al-Ma'mun, the two sons of the Abbasid caliph Harun al-Rashid, the 
strength of the state waned. Accompanying this political division, the 



"' For a harsh criticism of al-Dhahabi by one of his students, Taj al-Din al-Subkl, 
who also relied on him heavily in his Tabaqdt al-shdfi'ijja, see al-Subkl, "Qa'ida ft 
al-jarh wa al-ta'dll," in Arba' rasailji 'ulum al-hadith, ed. 'Abd al-Fattah Abu Ghudda, 
6th edition (Beirut: Dar al-Basha'ir al-lslamiyya, 1419/1999), 37 IT. For praise of 
al-Dhahabi from Indian Hanafts, see al-LaknawI, al-Raf wa al-takmil, 286. See also, 
Makdisi, "Hanbalite Islam," 240. 

"^ Al-Dhahabi, Sijar, 19:340, 

"'' Al-Dhahabi, Tadhkirat al-hujjai, 1 :56. 

^'' Al-Dhahabl, Tadhkirat al-hujjai, 1:179. 



358 CHAPTER NINE 

State of the faith deteriorated. The power of the Shiites and Mu'tazilites 
increased and the Baghdad Inquisition occurred. 

The star of Shiism rose and revealed its enmity [abdd sajhatahu), the dawn 
of speculative theology broke, the philosophy (hikmd) of the ancients, the 
logic of the Greeks and astrology were all translated into Arabic. A new 
science thus emerged for the people, abhorrent, destructive, incongruous 
with the knowledge of Prophecy and not in accordance with the unity 
of the believers that had held the umma in well-being."" 

With the narrative of entropy and decline into religious ruin set, al- 
Dhahabi bemoans the weakening of scholarship since the heady days 
of Ibn Hanbal's and 'All b. al-Madlnl's greatest generation. Al-Dhahabi 
specifically complains about the state of Islamic knowledge in his own 
time, condemning blind imitation [taqlid) in law and the obsession 
with empty speculative theology (kaldm). In such times, he concludes, 
"may God bless that individual who devotes himself to his task, who 
shortens his tongue, draws near to reading his Qur'an, cries over his 
time [zamanihi) and pores over the Sahihayn."^-' 

In his grief over the deterioration of scholarship and piety, al-Dhahabi 
thus calls for a return to the twin roots of Islam: the Qur'an and the 
sunna of the Prophet. The route to salvation, if only on the individual 
level, is to embrace the holy book and those volumes that had come 
to represent synecdochically the Prophet's true legacy, the Sahihayn of 
al-Bukhari and Muslim. 



Conclusion 

In its roles as a measure of authenticity, authoritative reference for 
non-specialists and exemplum, the Sahihayn canon functioned as Canon 
1: a criterion between truth and falsehood. Al-Bukhari's and Muslim's 
books, however, played another crucial role beyond the limited circles of 
jurists and hadlth scholars. The two collections came to synecdochically 
represent the Prophet's legacy itself within the wider Sunni community. 
Ironically, in their denial of the existence of a hadith canon, both 
Wheeler and Weiss alluded to the important function that the major 
Sunni collections served in their capacity as Canon 2: they delimited 



Al-Dhahabi, Tadhkirat al-huffa^, 1:240. 
AI-Dhahabi, Tadhkirat al-hujfai, 2:86. 



CANON AND SYNECDOCHE 359 

the vast expanse of the Prophet's sunna and embodied it in a manage- 
able form.'"' Whether the canonical unit of the Five Books or just the 
Sahihayn, this circumscription drew the boundaries of the greater Sunni 
community. Loyalty to the canon meant loyalty to the umma. 

The SahihayrC% synecdochic representation of the Prophet rendered 
the books invaluable in both scholarly and lay interaction with the 
heritage of Muhammad. In the narratives that hadith-oriented Sunni 
scholars developed to describe the historical course of Islamicate civi- 
lization, al-Bukhari and Muslim became a trope for the straight path 
of adherence to the Prophet's sunna in the face of the ever-multiply- 
ing threats of heresy and iniquity. In the Sunni narrative of decline 
from the halcyon days of the righteous early community, the SaMiayn 
represented salvation through a return to their teachings. More impor- 
tantly, by the seventh/thirteenth century al-Bukharl's and Muslim's 
collections had taken on prominent roles in political, calendrical and 
supplicatory rituals. Again, the two works symbolized the Prophet's 
legacy. For Mawla Isma'il they symbolized loyalty to the Prophet and 
the 'Alawid state that governed in his name. For the scholars who read 
the Sahihayn during Rajab, Sha'ban and Ramadan in Timbuktu, Cairo, 
Mecca or Damascus, the Sahihayn imbued a set period of the year with 
the religious significance of the Prophet's persona. In all these instances 
of ritual use, but perhaps most palpably in their roles as tools of sup- 
plication, the Sahihayn synecdochically represented the Prophet's access 
to divine blessing. Like relics or Muhammad's descendents, the hadlth 
collections personified the Prophet's role as the intercessor between 
humanity and the divine. 



Wheeler, 59; Weiss, The Search for God's Law, 260, cf. 266. 



CHAPTER TEN 
CONCLUSION 



Instead of summarizing the results of this study in abstract form (see 
the Thesis section in the Introduction), we conclude in a manner more 
useful to students of Islamic civilization and its magnificent tradi- 
tion of hadith scholarship. As the present study proceeded, teachers, 
scholars and students consistently posed the same questions about the 
Sahihayn canon and its historical development. I have thus attempted 
to use these questions as a framework for summarizing the conclusion 
of this study. 



I. Why the Sahihayn and Not Other Books? 

Asking why one text achieves membership in the canon and another 
does not poses trenchant questions about the forces that drive intellectual 
history and about the possibility of objective scholarly evaluation. Can 
historians always explain choices made in the past through a material- 
ist lens, or can historical actors establish and act on sets of aesthetics 
independent of material surroundings? One might contend that there 
is nothing in the writings of Shakespeare that makes them intrinsically 
better than the works of other playwrights or poets. The canonical status 
of Romeo and Juliet might ultimately hinge on the number of copies of 
the text that were produced at some crucial point in time, the nature 
of the network that distributed and performed the play, the charisma 
of those scholars who promoted its study or its resonance with some 
great social issue of the day. Another, better play written by a now- 
unknown litterateur may have disappeared into history for similar rea- 
sons. Canonicity from this perspective, is the product of material forces 
and the accidents of history. It is not a matter of objective quality. 

This perspective robs the critic or the scholar of his right to aesthetic 
evaluation; eminently a creature of the material world around him, he 
is no more able to escape these constraints than the texts he purports 
to judge. Is this perspective accurate, or must we allow for the seren- 
dipitous variable of scholarly preference? Should we acknowledge that 



CONCLUSION 361 

a well-respected critic or sincere scholar could rise above the material 
constraints of his day and pronounce an influential verdict on a book 
based on purely aesthetic grounds? It seems that the Sahihayn canon was 
the product of both the material accidents of history and the explicit 
judgments of influential Muslim scholars as to which hadith collections 
provided the best understanding of the Prophet's charismatic legacy. 

To isolate the factors that shaped the Sahihayn canon, let us review the 
fate of four hadith collections written by prominent transmission-based 
Sunni scholars of the sakih movement between 250/865 and 350/960 in 
the Khurasan region: the SaMhoyn, the SaMh of Ibn Khuzayma (d. 311/ 
923) and the Sahih of Ibn Hibban (d. 354/965). All four of these hadith 
scholars were Sunnis who compiled comprehensive legal and doctrinal 
references on hadith restricted to only what they considered authentic 
reports. All four had comparable visions of what Islam and the sunna 
of the Prophet 'should' be. By the eighth/fourteenth century, all four 
collections had won approval from the Sunni scholarly community. As 
our judge of canonicity let us turn to al-Hakim al-Naysaburl, whose 
seminal study of the Sahihayn in fact sparked their canonization. While 
al-Hakim viewed al-Bukharl's and Muslim's collections as the pinnacle 
of critical stringency and excellence in hadith evaluation, he dismissed 
both the Scihih of his teacher Ibn Hibban and that of his exemplar Ibn 
Khuzayma. 

Ibn Hibban's work seems to have been the victim of the accidents 
of history. Al-Hakim condemned the work of his teacher, a belated 
participant in the sahih movement, due to the presence of unknown 
transmitters in its isnads. As we know, however, early members of 
the Sahihayn Network had also been unable to identify some of al- 
Bukhari's transmitters. Only after several generations of study were 
these 'unknown' narrators identified. For al-Hakim, the absence of 
unknown transmitters in the Sahihayn proved central to his claims on the 
books' authority. Had Ibn Hibban lived a century earlier and produced 
his Sahih at the same time as al-Bukhari, perhaps scholars could have 
identified his unknown transmitters as well. 

In the case of Ibn Khuzayma's S^hih, however, we cannot explain its 
exclusion from the canon as the result of material forces or ideological 
pressures. Influential scholars who evaluated Ibn Khuzayma's Sahih 
simply did not approve of his quality selections. Ibn Khuzayma was the 
axis of transmission-based jurisprudence, theology and hadith study in 
Khurasan during the late third/ninth and early fourth/tenth centuries. 
Our earliest sources on the period accord him accolades that dwarf 



362 CHAPTER TEN 

those of al-Bukharl and Muslim.' Yet when al-Hakim was asked about 
whether or not Ibn Khuzayma was a reliable judge of the authenticity 
of Prophetic reports, he replied, "That I do not say"^ Al-Isma'llT had 
preferred al-Bukharl's legal analysis to Muslim's relative impartiality, 
and Ibn 'Uqda had favored Muslim's focus on Prophetic hadiths to al- 
Bukharl's insistence on providing incomplete reports as legal commen- 
tary Yet both these critics explicitly stated that al-Bukharl and Muslim 
provided the community with eminently reliable representations of the 
Prophet's sunna. Ibn Khuzayma's Sahih never attracted the scholarly 
interest heaped on the Sahihayn, and its exclusion from the Six Book 
canon seems to be the result of his failure to inspire the same confidence 
in the community that canonized al-Bukharl and Muslim. 

The reason why the Sahihayn, not other canonical hadith books, played 
such a salient role in ritual and narrative grew out of the unique status 
they had achieved by the dawn of the fifth /eleventh century. In Islam, 
an object becomes religious through a perceived link to God and His 
Prophet. As the community of God's final messenger, guarded against 
communal error by God Himself, the umma can further enunciate 
His will through claims of consensus {ipna'). Goldziher thus astutely 
recognized that ijma' was the bedrock on which Sunnism was founded.' 
Claims based on the umma's consensus underpinned the Sahthayn canon, 
and no other book after the Qur'an could boast such recognition. As 
objects endowed with religious significance, the Sahihcyn were ideally 
suited to dramatize religious meaning in acts of ritual or represent it 
in historical narrative. 



II. What Forces Led to the Canonization of the Sahihayn? 

We have asserted that canons form at the nexus of text, authority and 
communal identification. By authorizing texts, communities express, 
delineate and affirm their identities or boundaries. The creation of a 
canon thus stems from a twofold need to embody authority in text and 
delineate community through text. We have also contended that the 



' Al-Hakim, Tankh Nishabur, 120. 

2 Al-Khallll, al-lrshdd, 313. 

' Berkey, Formation of Islam, 189—90; Goldziher quoted in Makdisi, "Hanbalite 
Islam," 253. This observation is reminiscent of the Azhar adage that ijma is 'al-rukn 
al-rakin yastanidu ilayhi al-din.' 



CONCLUSION 363 

communal drama in which the canonization of the Sahihayn played a 
salient role was the articulation of Sunnism. Scott C. Lucas has sug- 
gested that discovering how such initially controversial figures (from 
a Sunni perspective) as al-Bukhari and Abu Hanlfa achieved 'Sunni' 
status remains an important but unanswered question in the study of 
this community's history* We might rephrase the question to ask how 
Sunnism adapted to adopt these figures into its fold. 

Sunnism began as the exclusive worldview of the transmission-based 
scholars, whose fixation with hadiths and their literal interpretation 
was intractably rigid. The iiber-Sunni credo of Ibn Hanbal, Abu 
Zur'a al-RazI or Abu Nasr al-Wa'ill brooked no school of thought 
that had either elaborated a more varied set of interpretive tools for 
understanding the cosmos, like the Mu'tazilites and Ash'aris, or defined 
the Prophet's sunna by means other than a stubborn obsession with 
hadiths, like the Hanafis. 

To explain how the conservative ethos of these 'people of the sunna 
and community [ahl al-sunna wa al-jamd'a)' expanded to include the 
relatively diverse four schools of Sunni law as well as the Ash'arT and 
Maturldi schools of theology, it may be useful to conceive of Sunnism 
more as a rhetorical mantra than a rigid doctrine. As it solidified in 
the fourth/tenth and early fifth/ eleventh centuries, Sunnism certainly 
required the espousal of certain specific beliefs: the proper ranking of 
the Four Rightly Guided caliphs (Abu Bakr, 'Umar, 'Uthman then 'All) 
and the belief that the Qur'an was uncreated, for example. Beyond 
such limited dogmatic tenets, however, we can envision Sunnism as an 
austere rhetorical call to stand fast by the Qur'an, the Prophet's sunna 
and the ways of the early community in the face of foreign innovations 
in faith, thought and practice. 

As a rhetorical mantra, Sunnism eventually proved charismatic and 
flexible enough that differing schools of law or theology were able to take 
it up in order to affirm their identification with a perceived traditionalist 
orthodoxy — even though their own doctrines or practices might at times 
differ significantly from it. The theological and epistemological school 
of Abu al-Hasan al-Ash'ari (d. 324/935-6) epitomizes this rhetorical 
flexibility. Although this scholar publicly repented his Mu'tazilite ratio- 
nalist ways and embraced the traditionalist beliefs of Ibn Hanbal and 
the ahl al-sunna wa al-jamd'a, the school that developed from his writings 



Personal communication. 



364 CHAPTER TEN 

(and perhaps his writings themselves) continued to delve deeper into 
speculative theology and Hellenistic epistemology.'' 

While the iiber-Sunni strain of the transmission-based school was 
parochially limited, the legal and theological tradition that coalesced 
around the teachings of al-Shafi'l was more open to methods of ana- 
logical reasoning and eventually Hellenistic logic and speculative 
thought. Just as al-Shafi'T himself had accommodated analogical legal 
reasoning (qiyas) in the transmission-based methodology, so too later 
Shafi'i/Ash'aris like Abu Ishaq al-Isfarayini or al-Juwayni were able 
to elaborate systems of legal theory or theology derived significantly 
from Mu'tazilite rationalism while making convincing arguments for 
their loyalty to the hadith-centric Sunni worldview. An Ash'arl who 
had written extensively on speculative theology, al-Juwaynl could when 
necessary also avow his membership in the ahl al-sunna by trumpeting the 
mantra that "the foremost [calling] is following the Salaf and rejecting 
religious innovation [bid'd) . . . .'"^' 

Eventually, the Hanafi school could also imitate the Shafi'T/Ash'ari 
orthodoxy and take up this elastic Sunni mantra. The Hanafi inter- 
pretive tradition had initially been anathema to the ahl al-sunna wa 
al-jama'a. Original 'Sunni' scholars had in fact reviled early pivots of 
the school like Abu Hanlfa and Muhammad b. al-Hasan al-ShaybanI 
(d. 189/805) as heretical Jahmi rationalists.' When a mid-third/ninth 
century Hanafi scholar named Ibn al-Thalji (d. 265/879) dared to use 
Prophetic reports to buttress the position of his school against that of 
its ahl al-sunna opponents, Ibn Hanbal and his followers devastatingly 
dismissed him as an 'unbeliever'" The situation had changed dramati- 
cally by the eighth/fourteenth century, when the Sunni edifice became 
established in its most concretely permanent state. By that time some 
Hanafis had recast Muhammad b. al-Hasan al-ShaybanI as a proto- 
Sunni who had advocated the literal interpretation of the Qur'an and 
hadlth on issues of God's attributes.^ 



'' Abu al-Hasan al-Ash'an, Maqalat al-islamiyjin, ed. Helmut Ritter (Istanbul: Dar 
al-Funun, [1928]), 280-1. 

'' See, for example, al-Juwaynl, al- 'Aqida al-Nizamiyja ji al-arkdn al-isldmiyya, ed. 
Muhammad Zahid al-Kawtharl (Cairo: Maktabat al-Azhariyya li'1-Turath, 1412/1992), 
23, 32. 

' Abu Zur'a al-RazI, for example, is quoted as calling Abu Hanlfa, Muhammad b. 
Hasan al-ShaybanI and Abu Yusuf 'Jahml;' al-Khatib, Tdnkh Baghddd, 2:176. 

» Ibn al-Nadim, The Fihnst, 510-11; al-Khatib, fdnkh Baghddd, 2:425-5. 

' Hodgson, The Venture of Mam, 2:449; Ibn Ak al-'Izz al-Hanafi (d. 792/1390), Sharh 
al-'Aqida al-Tahdwiyya, 215. 



CONCLUSION 365 

This notion of Sunnism as a rhetorical touchstone within arm's reach 
of a variety of interpretive schools explains the tremendous, almost 
inconsistent, diversity within the later Sunni tradition. A phenomenon 
unimaginable in the fourth/tenth-century world of the ahl al-hadith and 
ahl al-ra'y is exemplified by MuUa 'All Qarl (d. 1014/1606), a loyal 
Hanafi who, in the space of one book, quotes Ibn Hanbal to condemn 
speculative theology and logic, embraces the Ash'ari figurative expla- 
nation of God's attributes and describes the Shafi'i/Ash'ari Sufi 'Abd 
al-Karim al-Q_ushayri as being on the path of the Salaf '" 

The development and function of the Sahihayn canon mirror the 
development of Sunni identity. What began as the limited interest of 
a network of Shafi'T scholars developed into a strong and shared iden- 
tification with these two hadlth collections among Shafi'T and HanbalT 
students of al-Hakim al-Naysaburl. Representatives from both these 
schools agreed on the Sahihayn as a common ground for identifying the 
Prophet's authentic legacy. The other schools of Sunni Islam gradually 
adopted this convention of al-Bukhari and Muslim as a measure of 
authenticity, authoritative reference and exemplum. Finally, even the 
Hanafis acceded to identifying with the Sahihayn as the common lan- 
guage for Sunni discussions of hadith. Although the Shafi'ls, Malikis, 
Hanballs and Hanafis had relied on their own bodies of hadiths in 
their elaboration of law and dogma, they all acknowledged the Sahihayn 
as rhetorically paramount in interactions between the schools. In the 
seventh/thirteenth and early eighth/fourteenth centuries, when the 
popular religious institutions of Sunnism such as Sufi brotherhoods were 
coalescing, the Scihihcyn too became vehicles for public ritual activity. 

By acknowledging the SdhWcyn as authoritative, the collection of legal 
and theological schools within Sunni Islam turned the two works into 
touchstones of communal identification. In order to understand how 
the forces of a developing sense of communalism created the canon, we 
must quickly review how the nature and needs of the Muslim scholarly 
community developed from al-Bukhari's and Muslim's lifetime to the 
mid-fifth/ eleventh century, when the Sahihayn canon found widespread 
use and acceptance. 



'" MuUa 'All Qarl, Shark al~fiqh al-akbar; 25—6, 28, 35, 63. For an expression of 
MuUa 'All's loyalty to the Hanafi legal school, see his Tashji' Jiiqahd' al-hanafiyya li-tashni' 
sufahd' al-shdji'iyya, Ms. 444, Yahya Tavfik Collection, Siileymaniye Library, Istanbul, 
fols. 82b-84b. 



366 CHAPTER TEN 

In the years after the deaths of the Shaykhayn, AbQ Zur'a and Abu 
Hatim al-Razi continued to ply their scholarly trade in their native 
Rayy. The two scholars were very conservative members of the trans- 
mission-based ahl al-hadith, drawing equally from the scholarship of 
Ibn Hanbal and al-Shafi'l. Although their study of legal texts like al- 
Muzanl's Mukhtasar or Ibn Hanbal's responsa certainly informed the two 
RazTs' legal and doctrinal opinions, their views were ultimately shaped 
by their own study and interpretation of hadiths back to the Prophet. 
Like the other major transmission-based scholars of their time, such as 
Abu Dawud, each scholar constituted his own school of hadlth criti- 
cism. When Muslim brought his freshly penned Sahih to Abu Zur'a, he 
looked through it with the eye of a scholar confidently following his own 
methodology of evaluating the authenticity of Prophetic reports. 

Two hundred years later, the scene of Sunni scholarship had trans- 
formed dramatically. Unlike the two Razis, scholars like the Shafi'i/ 
Ash'arl Abu Ishaq al-Shirazi were no longer willing to draw indiffer- 
ently from what had become the very distinct Hanbali and Shafi'i 
legal schools. Yet despite this solidification of boundaries, the Sunni 
universe had expanded beyond the excusive circle of self-sufficient, 
iiber-Sunni hadlth-based jurists to include figures like al-Juwayni, a 
practitioner of dialectical theology and a jurist loyal to a specific body 
of substantive law. Abu Zur'a and Abu Hatim al-Razi had personally 
vouched for the strength of their hadiths with the confidence their 
critical expertise inspired in their followers, but in the expanded Sunni 
world of the fifth /eleventh century a more institutionalized convention 
was required for discussing attributions to the Prophet. There existed a 
real need for a means to compel others to acknowledge a representa- 
tion of the Prophet's authoritative legacy. The Sahihayn provided this 
common measure of authenticity. Unlike the RazIs, al-Shirazi and 
al-Juwaynl were unable to critically vet their own corpora of hadiths; 
they needed to turn to authoritative references to provide commonly 
accepted reports. 

In the fifth/eleventh century, and later when the Hanafi school 
adopted the canon, the Sahihayn acted to both facilitate and define the 
expanded Sunni community. The two books provided a common source 
and reference through which different schools could address one another 
in debates and polemics. More importantly, however, the Sahihayn also 
functioned as a mantra of communalism. When the Shafi'i/Ash'aris 
Abu Ishaq al-Isfarayini and al-Juwayni, the Hanbali/iiber- Sunni Abu 
Nasr al-Wa'ili, and the Maliki Abu al-Walid al-Bajl proclaimed inde- 



CONCLUSION 367 

pendently that 'the community of Muhammad [al-umma)' had agreed 
on the Sahihayn as totally authentic vessels for the Prophet's authoritative 
legacy, they affirmed their own loyalty to that shared Sunni community 
More importantly, they acknowledged the membership of others who 
made that claim. When the Hanafi 'Abd al-'AzIz al-Bukharl attested 
that al-Bukhari's opinion on the authenticity of a hadith was absolutely 
definitive, he too took up this canonical mantra of Sunnism. When 
the Mamluks salaried scholars to read the Sahihayn for three months 
in the mosques of Cairo or placed al-Bukhari's collection at the van- 
guard of their army, the two books embodied Sunni ritual and political 
communalism. 

Although the pressures of communal identification create the canon, 
it is the canon that then defines the community As evident in al-Silafi's 
declaration that anyone who disagrees with the Five Book hadith canon 
places himself outside 'the Abode of Islam,' the canon could certainly 
delineate the boundaries of the Sunni pale. Although the permissibility of 
criticizing the Sahihayn constituted the norm for centuries, the perceived 
fragility of the Sunni community in early modern India led Shah Wall 
Allah to equate belittling al-Bukhari and Muslim with "not following the 
path of the believers." The ability of texts to determine and shape com- 
munity, however, is predicated on the compelling power of those books. 
Neither al-Silafi nor Shah Wall Allah could have made their statements 
before the canonization of the S^hihoyn at the dawn of the fifth/eleventh 
century. The relationship between canon and community is dialogic, but 
only after the community brings the canon into existence. 



III. Why Did the Canon Form at the Beginning of the 5th/ 11th Century? 

That the S^hihcyn canon formed and found its immediate application 
in the early fifth/ eleventh century is not accidental. The emergence 
of the canon as an institution was both a part and product of the 
coalescence of the new Sunni order in this period, one that was char- 
acterized by the institutionalization of education, modes of patronage 
and clearly delineated schools of thought. The frustrating ambiguity of 
the fourth/tenth century, with its fluctuating and languishing categories 
of the ahl al-hadith and ahl al-ra'y, and the regional laws school, faded 
as more concrete divisions solidified. The two strands of the transmis- 
sion-based school, the conservative iiber-Sunnis and the more moderate 
strain associated with the Shafi'l tradition, gelled into the guild-like 



368 CHAPTER TEN 

Hanball and Shafi'i schools. By approximately 425/1035 the Ash'ari 
school of theology had blossomed into a mature form. By 480/1090 
the Malikis, Hanafis, Shafi'ls and Hanballs had all composed defini- 
tive texts on legal theory, substantive law and hadlth and had staked 
their dogmatic positions in relation to one another The proliferation 
of madrasah, founded and funded by wealthy patrons often associated 
with the Seljuq state, furnished a new institutional setting for the study 
of the religious sciences. Unlike the merchant and landlord scholars of 
previous generations, the salaried teachers and stipended students in 
these madrasas could pursue scholarship in a professional setting. 

The institutionalization of Sunnism that spread rapidly from the 
fifth/eleventh century on occurred on a grand and massively important 
scale. As Marshall Hodgson recognized, it was in the period from 945 to 
1250 CE that Islamicate civilization grew from its adaptive adolescence 
into a viable institutional framework for a world-civilization." Richard 
BuUiet has seconded this emphasis on the theme of institutionalization 
in the fifth/eleventh-century emergence of Sunnism. He explains that 
this development was "actually the first stage in the dissemination of 
religious institutions and the standardization of Sunni religious norms 
that becomes the hallmark of later Islamic history"'^ In particular, 
BuUiet highlights the transition from the cultivation of hadiths with 
living isndds (BuUiet's 'orality')'' to the study of hadlth collections and 
the appearance of the madrasa system as the twin faces of the revolution 
that redefined Sunni Islam in the late fourth/tenth and fifth/ eleventh 
centuries. He links this institutionalization of education, in both the 
transition from living isndd% to books and the spread of the madrasa, 
with the formation of the Sunni hadlth canon, since madrasa% relied on 
these collections as part of their curricula.'* 



12 



Hodgson, The Venture of Islam, 2:3. 
BuUiet, Islam: The View from the Edge, 126-7. 
" I believe that the term 'living isnaiT more accurately describes the phenomenon 
that BuUiet addresses, namely a focus and reliance on direct chains of transmission 
back to the Prophet as opposed to collections of hadiths compiled by authors and then 
transmitted. A shift to employing books of hadiths did not obviate the oral nature of 
study. Even today, the study and transmission of these texts is an oral activity based 
on the communicative act of hearing the work read. 
'* BuUiet, Islam: The View from the Edge, 149. 



CONCLUSION 369 

Madrasa curriculum, however, cannot tell us why the Sahihayn achieved 
canonical status in this period. In cities like Q_azvin, hadith study gener- 
ally continued in large mosques, not madrasas. Furthermore, madrasa?, 
from Egypt to India utilized a large and varied selection of books 
for instruction. None of these, however, attained the ubiquitous and 
unparalleled status of the Sahihayn. Instead, we must look to the needs 
created by the Sunni scholarly community's act of self-delineation and its 
search for the tools required to facilitate internal coherence. Al-Bukhari's 
and Muslim's books had received concerted study in the long fourth 
century because they provided a network of influential Shafi'i scholars 
with the ideal vehicles for expressing the nature and quality of their 
command of the Prophet's legacy. Al-Hakim al-NaysaburT exploited 
this network's assiduous study of the Sahihayn to transform al-Bukharl 
and Muslim into widely recognized stamps of authenticity. This kandn, 
he claimed, met the authenticity requirements of both the Sunnis and 
the single greatest threat to their transmission-based worldview: the 
Mu'tazilite attempt to limit the role of Prophetic hadiths in elaborat- 
ing law and dogma. 

While the needs and contributions of the Sahihayn Network and 
al-Hakim in particular produced the canon, they cannot explain its 
wider proliferation. The canon flourished among al-Hakim's students 
and other major participants in the Sunni orthodoxy of the fifth/elev- 
enth century because the Sahihcyn fulfilled specific needs created by its 
solidification. The need for hadiths and hadith collections that could 
function as epistemologically certain loci of consensus, felt generally in 
the fourth/tenth century, became more pronounced when distinct legal 
schools that shared a common Sunni worldview required a common 
convention for their ceaseless debates over the proper interpretation of 
the Prophet's sunna. With the institution of the madrasa and the divi- 
sion of labor among Sunni scholars in the mid fifth/eleventh century, 
accepted references for hadith criticism also became necessary for non- 
hadlth specialists. The two books provided a common language and 
reference for discussing hadiths among the Maliki, Shafi'l and Hanball 
schools in the fifth/ eleventh century, with the Hanafi school adopting 
this convention only in the early eighth/fourteenth century. 

The adoption of the canon as a common convention for hadith 
study was certainly related to the shift from the living isndd to the 
transmission of books. It seems, however, that this shift occurred after 
the canonization of the S^hihciyn. In their biographical dictionaries. 



370 CHAPTER TEN 

al-Khalili (d. 446/1054) and al-Khatib al-Baghdadl (d. 463/1071), 
two scholars who readily employed the canon, still focused much 
more on living isndds, than books. Our sources for the second half 
of the fifth/ eleventh century, however, indicate that circa 465/1072 a 
marked shift occurred toward noting the hadlth books that scholars 
studied as opposed to their living isnads to the Prophet. In his history 
of Naysabur, 'Abd al-Ghafir al-Farisi (d. 529/1134-5) mentions only 
ten people studying the Sahihayn between 385/995 and 465/1072, but 
between 465/1072 and 545/1150 (some material was added after the 
author's death by al-Sarifini [d. 641/1243—44]) he mentions fifty-five 
(a 450% increase). Between 385/995 and 465/1072 he mentions only 
eight other hadith collections, such as the Sunans of al-Nasa'l and Abu 
Dawud, being studied. Between 465/1072 and 545/1 150 he mentions 
twenty (a 150% increase). In his Iraq-Khurasan-centric al-Muntazam, 
Ibn al-Jawzi mentions only nine instances of a scholar studying a hadlth 
book in the two hundred years between 285/898 and 485/1092. In 
the period of only eighty years between 485/1092 and 565/1170 he 
mentions seventeen (a 190% increase). Yet we know that despite these 
statistically dramatic changes, a strong attachment to the living isndd 
endured. Well into the 500s/ 1 100s, scholars like Ibn Funduq al-Bayhaqi 
(d. 565/1169-70) continued to define hadlth scholarship as the living 
transmission of individual hadiths from the Prophet as opposed to the 
study of hadlth collections. 

Although it is difficult to date precisely two such intangible events, it 
seems that the emergence of the Sahihayn canon in the early fifth/ elev- 
enth century preceded the first indications of a shift from living isndd% to 
the transmission of books by at least fifty years. We can see this clearly 
in the case of scholars who employed the canon while still depending 
wholly on their own living isndds, to the Prophet. Scholars like Abu Bakr 
al-Bayhaqi (d. 458/1066) and al-Khatib did not need hadlth books to 
provide the content of their hadlth works; these they filled with their 
own full-length living isndds. They did need collections like the Sahihayn, 
however, to guarantee the authenticity of these hadiths. The canon 
formed because scholars needed a stamp of approval for hadiths, and 
this could only come from consensus on a hadith collection. 



CONCLUSION 371 

IV Did the Canon Emerge from Ferment and Strife? 

Studies of canons and canonization have often identified periods of 
ideological ferment or strife as the seedbeds of scriptural canons.'-^ 
Just as a proclamation of orthodoxy arises as a response to perceived 
threats of interpretive plurality, so too a canon emerges as an attempt 
to dominate the textual landscape of a religious tradition. As a corol- 
lary, this emphasis on ideological ferment in canon studies has led to a 
focus on canons as "heavy weapons," tools for control and exclusion.'*" 
Western scholars have thus not fully appreciated the capacity of canons 
to create common convention and bridge rifts. Menzies alone argued 
that canons may well form in the reconstructive wake of conflict." 
Indeed, just as the Sahihayn provided a common language for Sunnism, 
the canon resulted from the institutional consolidation of an expanded 
orthodoxy in the wake of tumultuous plurality. 

The consistent intensification of the Sahihayn canonical culture after 
the careers of Abu Mas'ud al-Dimashqi and al-Khatib al-Baghdadi in 
the late fourth/tenth and mid-fifth/ eleventh centuries coincides with the 
consolidation of Sunnism. As Jonathan Berkey states, Sunnism of the 
fifth/ eleventh century was engaged in a process of minimizing "sources 
of contention."'" The dogged creed of communalism which, according 
to Hodgson, characterized Sunnism after this period perfectly describes 
the canonical culture's goal of suppressing opinions that threatened 
the institutional roles of al-Bukharl and Muslim. Sunni communalism 
demanded "loyalty to the community and its acknowledged symbols... 
even at the expense of all other values." Most assuredly, the canoni- 
cal culture required Sunnis to affirm the community's consensus on 
the Sahihayn at the expense of the established conventions of hadith 
criticism and the historical record of al-Bukharl's and Muslim's pre- 
canonical images.'^ 



Halbertal, 4—5; Hanaway, 3. 

Hanaway, 3; Kermode, "Institutional Control of Interpretation," 77. 

Menzies, 91. 

Berkey, The Formation of Islam, 189—90. 

Hodgson, TheVenture of Islam, 2:193. 



372 CHAPTER TEN 

V Was the Canon a Response to Shiism or the Product of the Seljuq State? 

Although the Sahihayn served as a unifying bond within the Sunni 
community, was this broad inclusivity the byproduct of an effort to 
exclude non-Sunnis? Many scholars have identified the emergence of 
institutional Sunnism in the fifth/ eleventh century as a defensive reac- 
tion to the tremendous power of Shiism in the fourth/tenth century. 
Did the ImamI Shiite Buyid dynasty's dominanation of the Abbasid 
caliphate in Iraq and Iran, and the meteoric rise of Fatimid power in 
Egypt, Syria and the Hijaz, catalyze the institutional consolidation of 
Sunnism? Was this reaction instigated and encouraged by the threatened 
Sunni Seljuq state, many of whose leading functionaries fell before the 
daggers of Isma'ill assassins? 

Some scholars have deemphasized the place of state sponsorship in 
the consolidation of Sunnism. One of the architects of the notion of 
the 'Sunni revival,' George Makdisi, viewed it as a victory of tradition- 
alism and credited it to the tremendous popular appeal of the Hanbali 
school in Baghdad, not to the Seljuq state. ^^ 

Others have understood the new Sunni order through a decidedly 
political lens. Hodgson associated it with Nizam al-Mulk's madrasa sys- 
tem, which epitomized the Seljuq-fostered framework that replaced the 
vanished Abbasid caliphal state with a new dispensation of uniformity. 
This state-sponsored madrasa system "carried on the task of maintain- 
ing essential unity in the community's heritage" as bequeathed by the 
Prophet and his Companions.^' 

The construct of a state-sponsored Sunni revival has been intimately 
bound to the Seljuqs' Shiite adversaries, both the ousted Buyids and the 
more immediately threatening Isma'ili Fatimids. Lapidus thus concluded 
that the fifth/eleventh-century institutionalization of a Sunni orthodoxy 
was a politically-led reaction to Shiite power. The Abbasid caliph al- 
Qadir, who promulgated the famously anti-Shiite Q_adiri creed in the 
twilight shadows of Buyid suzerainty, the Seljuqs and their successor 
dynasties of the Ayyubids and Mamluks all promoted an institution- 
alized Sunni orthodoxy as part of a drive to unite society around a 
state-embraced Sunni cause. This was exemplified by Nizam al-Mulk 



™ Makdisi, "Hanbalite Islam," 237-8. 

^' Hodgson, The Venture of Islam, 2:48, 192. 



CONCLUSION 373 

and Malikshah's efforts to mollify through patronage all the major 
non-Shiite factions in the various feuds on the Baghdad-Khurasan 
circuit: the Shafi'i/Ash'aris, Hanbalis and Hanafls.^^ BuUiet, however, 
rejects the equation of the Sunni revival with a reaction to Shiism. 
Instead, we should view it as an attempt to define Sunnism according 
to "centrally espoused dogma" (he thus admits that it is at least in 
some way the result of state policy). ^^ Jonathan Berkey follows BuUiet 
in downplaying the threat of Shiism or an anti-Shiite Seljuq policy as 
an engine for the crystallization of Sunnism. BuUiet and Berkey both 
point out that the Seljuqs often adopted a conciliatory attitude toward 
the powerful Imami Shiite interests in cities like Baghdad. For example, 
Nizam al-Mulk and his master Malikshah both married their daughters 
to Shiite nobles and appointed Shiite ministers.^* 

Neither BuUiet nor Berkey, however, pays sufficient attention to the 
fact that it was the Isma'ilis and not the relatively harmless Imami 
Shiites who alarmed the Seljuq state and Sunni scholars alike. Sunni 
firebrands such as the caliph al-Qadir certainly condemned Imami 
Shiites, but, as Abu al-Husayn Qazvlnl found himself insisting in his 
Ketab-e naqd, it was the Isma'llls whom the Sunnis truly feared. It was 
Isma'lll propaganda that proved so appealing to the intellectual elite 
in the major metropolises of the Seljuq realm, and Isma'ill assassins 
who represented the single greatest external danger to the stability of 
the Seljuq dynasty. This threat had earlier sparked an unlikely alliance 
between the Sunni caliph al-Q_adir, his Shiite Buyid overlords and the 
ImamT Shiite scholars of Baghdad. In 402/1011 they jointly promul- 
gated an anti-Isma'lli manifesto directed at the encroaching Fatimid 
state. ^-^ 

While the consolidation of Sunnism in the fifth/ eleventh century 
may well have been a response to the Fatimid threat and Isma'ili pro- 
paganda, we cannot identify any direct effect on the formation of the 
hadith canon. Shiism, whether ImamT or Isma'lll, never surfaces in 
the various discourses surrounding the authorization of the Sahihayn. 
The canon was, in fact, a boon to Imami Shiites like Qazvinl, who 
turned to al-Bukhari's and Muslim's compelling authority in attempts 



^^ Lapidus, ^ History of Islamic Societies; 164, 173-4. 

^' BuUiet, Islam: The Viewjrom the Edge, 126-7. 

^' BuUiet, Islam: The Viewjrom the Edge, 148; Berkey, Formation of Islam, 191. 

2' D. Sourdel, "al-Kadir," EP. 



374 CHAPTER TEN 

to trump Sunni opponents by using their own proof texts against them. 
Ultimately, the Sahihayn were more a unifying element within Sunnism 
than a tool for excluding the Shiite other 

To the extent that the Isma'ili threat and any resulting Seljuq patron- 
age of non-Shiite schools helped bring Sunnism to institutional maturity, 
the canon can be seen as part of a response to Shiism. This perspective 
holds true, however, only at the most global level of analysis. Those 
scholars who participated in the various discourses that produced the 
hadith canon did not exhibit any concern for a Shiite threat in their 
related writings or understand the Sahihayn as a tool for excluding non- 
Sunnis. To the contrary, the earliest recorded usages of the canon are 
directed at either Mu'tazilites or adherents of other Sunni schools with 
an emphasis on the inclusive consensus that those who wielded the 
canon claimed it enjoyed. Abu Nu'aym al-Isbahani, a member of the 
SaMhoyn Network who was very familiar with al-Hakim's work, thus did 
not refer to al-Bukhari and Muslim in his manual for debating Imami 
Shiites. Although Abu Nu'aym refers to hadlths he argues are agreed 
on by all Muslims, citing the Sahihayn would have had no proof value 
whatsoever for his Shiite opponents. 



VI. Was the Sahihayn Canon the Product of or Limited to a Specific Region? 

The Sahihayn canon germinated in the scholarly circles of Naysabur, 
Jurjan and Baghdad during the first half of the long fourth century. 
Its articulation and early usage took place in the writings and debates 
of scholars traveling between the great urban centers of the Nile-Oxus 
Islamicate heartlands. Beyond these early stages, however, the history of 
the Sahihayn canon does not diverge markedly from the course charted 
by Islamic history in general. Where Sunnism flourished, the canon 
followed. 

Roy Mottahedeh has pointed out the prominence of Khurasan! 
scholars in the articulation of the Sunni hadith tradition in the third/ 
ninth century^'' Richard BuUiet extends this geographical focus in both 
chronology and import, arguing that the institutions that characterized 



'^^' Roy Mottahedeh, "The Transmission of Learning. The Role of the Islamic North- 
east," Madrasa, eds. Nicole Grandin and Marc Gaborieau (Paris: Editions Arguments, 
1997), 68. 



CONCLUSION 375 

the Sunni revival in the great imperial center of Baghdad, such as the 
madrasa, were truly imports from the Iranian east.^' 

The hadlth canon, however, was not the product of eastern Iran 
alone. Certainly, figures central to the canonization of the two works 
such as al-Hakim al-Naysaburl resided mostly in Khurasan. The Sahihayn 
Network, however, that readied the two books for canonization, and 
the cadre of Shafi'l/Ash'arl and Hanbali scholars who first promoted 
the canon, were first and foremost participants in the highly mobile 
and cosmopolitan scholarly culture that dominated Islamic civilization 
from the third/ninth to the sixth/twelfth centuries. Khurasan was only 
one province in this wider world. Al-Daraqutnl never voyaged east of 
Baghdad, Abu Ishaq al-IsfarayTnl divided his career between the Abbasid 
capital and Khurasan, and both Abu Nasr al-Wa'ili and al-Juwaynl 
spent significant portions of their careers in the Hijaz. 

Furthermore, the expanded Sunni community to which the Sahihayn 
canon proved so useful in the mid-fifth/ eleventh century and beyond 
was just as present in North Africa, Baghdad, Egypt, or Isfahan as east- 
ern Iran. Scholars in any city on the great scholarly/mercantile circuit 
that ran from Mecca to Transoxiana or westward to Andalusia would 
have appreciated the need for a common measure of authenticity, an 
authoritative reference or a standard of excellence in hadith study. The 
Sahihayn canon was a product of these far-flung urban centers and dusty 
roads of the dominant Hijaz — Baghdad — Khurasan — Transoxiana 
circuit of the fourth/tenth and fifth/ eleventh centuries. 

Oddly, the tremendous geographical distance between Andalusia and 
the central Islamicate heartlands proved unimportant in the spread and 
usage of the canon. While the rugged mountains separating Jurjan from 
Naysabur restricted the movement of information on the Sahihayn in the 
first half of the fourth/tenth century, the vast expanses of desert, plain 
and ocean between Cordova and Baghdad were of little significance in 
the history of the canon. Not only did Andalusian scholars who had 
voyaged east, such as Qasim b. Asbagh of Cordova and Abu al-Walid 
al-Baji of Badajoz, participate visibly in the Scih^hcyn Network and early 
applications of the canon respectively, the Sahihayn attracted significant 
attention in Andalusia itself Sahih al-Bukhdri first arrived in Andalusia 
not long after it achieved fame in the East. Abu Muhammad 'Abdallah 
b. Ibrahim al-Aslll (d. 392/1002), a judge in Saragossa, received the 



BuUiet, Islam: The Viewjrom the Edge, 146. 



376 CHAPTER TEN 

book from Abu Zayd al-MarwazI in Mecca and brought it back to 
Andalusia.^" His teacher, Abu al-Hasan 'All b. Muhammad al-Qabisi 
(d. 403/1012), also brought the collection back to the North African 
city of Qayrawan.^^ Their student al-Muhallab b. Abl Sufra Ahmad 
al-Marlyyi (d. 435/1044), a judge in the Andalusian town of Almeria, 
wrote a commentary on Sahih al-Bukhdn that was in fact the first such 
work devoted to the book anywhere since al-Khattabl had written his 
A'ldm al-sunan fifty years earlier'"' 

Two generations later, al-Jayyani (d. 498/1105) participated in the 
study and development of the Sahihayn canon without ever leaving 
Andalusia.^' He collected six separate transmissions of al-Bukharfs 
Sahih through the author's senior student, al-Firabri, as well as another 
prominent transmission from Ibrahim b. Ma'qil al-Nasafl. Al-JayyanI 
had the two most famous transmissions of Sahih Muslim as well (those 
of al-Qalanisi and Ibn Sufyan).^^ In addition, he had copies of al- 
Hakim's Tdnkh Naysdbur and his Ma'nfat 'ulum al-hadith. Although he 
was writing only a few years after al-Khatib al-Baghdadi's death, al- 
Jayyanl also had a copy of the massive Tdnkh BaghdddP Some of the 
most influential studies of the Sahihayn, such as al-Jayyanl's study of 
al-Bukharl's teachers and al-Mazari's and al-Qadi 'lyad's commentaries 
on Sahih Muslim, came from the Maghrib. Although he was famously 
unaware of al-Tirmidhi's existence, Ibn Hazm rated the S^hihoyn as 
the two best collections of hadlth. After madrasas, were founded in the 
Maghrib, the Sahihayn became standard texts for hadith study among 
the majority Malikl school.'* 

To the extreme east of the classical Islamic world, the S^hihoyn canon 
was at the vanguard of hadith scholarship in South Asia from the 
seventh/thirteenth century on. The first Indian to leave any trace of 
studying the Sahihcyn was also the first renowned Indian hadith scholar 
in general. A native of Lahore, al-Hasan b. Muhammad al-SaghanI 



^^ M-\hxaiayAi, Jadhwat al-muqtabis, 240; al-Dhahabi, Siyar, 16:560. See also Maribel 
Fierro, "The Introduction of hadith in al-Andalus," Der Islam 66 (1989): 87. 

2« Al-Dhahabi, Siyar, 17:159. 

™ Al-Dhahabi, Siyar, 17:579. 

=" Al-Dhahabi, Tadhkirat al-hujfaz, 4:22. 

'^ Al-JayyanI, al-Tanbih 'aid al-awhdm al-wdqi'a fi al-musnad al-sahih li'l-Bukhdri, 22; 
idem, al-Tanbih 'aid al-awhdm al-wdqi'a fi Sahih al-imdm Muslim, 35—41. 

'' Al-JayyanI, al-Tanbih 'aid al-awhdm al-wdqi'a fi Sahih al-imdm Muslim, 30-34. 

" See Wadad al-Qadi, "al-Madrasa ft al-Maghrib ft daw' Kitab al-ml'ad li'l- 
Wansharl," in al-Fikr al-tarbawi al-isldmi (Beirut: Dar al-Maqasid al-Islamiyya, 1401/ 
1981), 147. 



CONCLUSION 377 

(d. 650/1252), penned a study of al-Bukharl's teachers, a commentary 
on his Sahih and a famous combined edition of the Sahihayn, the Mashdriq 
al-anwdrP Al-Saghani spent much of his time studying in the Hijaz and 
serving the Abbasid caliph al-Nasir, who sent him back to India from 
Baghdad as the Abbasid ambassador to the Delhi Sultanate. Otherwise, 
it was not untU the 700s/ 1 300s that any real study of the Sahihayn started 
in South Asia proper According to Muhammad Ishaq, the first men- 
tion of the two works comes in the work of Makhdum al-Mulk Sharaf 
al-Din sometime between 741/1340 and 786/1384?'^' 

This history of the Sahihayn in South Asia reflects the study of hadith 
in that region in general. Although there had been limited hadith 
scholarship in Lahore under the Ghaznavids in the late fifth/ eleventh 
and early sixth/twelfth centuries, it was the establishment of the Delhi 
Sultanate that marked the beginning of continuous Muslim scholarship 
in northern India. Even then, however, the study of hadith was limited 
to al-Baghawi's Masdbih al-sunna and al-Saghanl's Mashdriq al-anwdr (in 
effect, the Sahihayn), the two books that provided the narrow foundations 
of the hadith curriculum in the new Nasiriyya and Mu'izzI colleges 
in Delhi. ^' Hadith scholarship in northern India was thus buUt on al- 
Bukharl's and Muslim's canonical status as manifested in al-Baghawi's 
and al-Saghanfs digests of two works. 'Abd al-Awwal al-Husaynl al- 
Zaydpuri (d. 968/1560), who lived in Gujarat and Delhi, wrote the 
first Indian commentary on al-Bukhari's collection: the Fayd al-bdri fi 
shark Sahih al-Bukhdn}^ In the wake of 'Abd al-Haqq b. Sayf al-Dihlawi 
(d. 1052/1642), the Indian scholar who truly replicated the intense 
hadith scholarship of the Islamic heartlands in India, hadith study 
flourished in the subcontinent. From that point onward, almost every 
major Indian hadith scholar produced a commentary on al-Bukharl's or 
Muslim's Sohih- Many commentaries were written in Persian, with Siraj 
Ahmad al-Mujaddadi (d. 1815) even translating Sahih Muslim directly 
into Persian. ^^ In light of the prominent place of the Sahihayn in South 
Asian Islam, it is no surprise that the great Sufi scholar Nizam al-Din 
Awliya' (d. 725/1325) rebutted a hadith used against him in a debate 
by stating only the contents of the Sahihayn are assuredly authentic.*" 



Ishaq, India's Contribution to the Study of Hadith Literature, 230. 

Ishaq, India's Contribution to the Study of Hadith Literature, 11 . 

Ishaq, India's Contribution to the Study of Hadith Literature, 49. 

Ishaq, India's Contribution to the Study of Hadith Literature, 129. 

Ishaq, India's Contribution to the Study of Hadith Literature, 143. 

Amir Hasan Sijzl, Nizam ad-din Awliya: Morals for the Heart: Conversations of Shaykh 



378 CHAPTER TEN 

Conclusion 

The Muslim hadlth tradition and the manifold roles of hadlth in Islamic 
civilization can stretch the historian's analogical abilities to their limits. 
It is not difficult to imagine that reports from the Prophet Muhammad 
played a central role in defining Islamic doctrinal and legal thought. 
As different schools matured and competed, it was natural that the 
authenticity of hadiths became an issue of great communal import. Al- 
BukharT and Muslim remain enduring symbols of the system of hadlth 
criticism and authentication that Muslim scholars from Andalusia to 
Transoxiana developed on so daunting a scale and with such internal 
consistency that it deserves mention as a great accomplishment in intel- 
lectual history. Just as we admire the logical or ethical explications of 
Peripatetic philosophers regardless of the accuracy of their conclusions 
today, we need only shift our gaze slightly to examine in wonder the web 
of intersecting lines of transmission that weave downward and outward 
from the Prophetic singularity along the dome of time and space. 

Yet beyond the role of hadlth in law and doctrine, it seems almost 
incomprehensible how such a large number of people from all reaches 
of society could devote themselves so totally to collecting and sifting 
through reports from the Prophet. Histories like al-Khatlb's Tarikh 
Baghdad or al-Dhahabl's Tadhkirat al-huffaz are replete with normal indi- 
viduals who traveled for months simply to collect an additional version 
of a Prophetic report for which they already possessed one narration. 
Even more shocking is the obvious fact that most of these hadlth col- 
lectors had little concern for the actual authenticity of these reports. 

Perhaps, however, the question of the canonization of al-Bukhari 
and Muslim reminds us that such a distant and fantastic past is not 
actually far removed from us today. Even today, historical authenticity 
is not prized by all equally. Abu Zur'a al-RazI understood that in mak- 
ing authenticity paramount, one may sacrifice the tools necessary for 
communal cohesion. As al-Albani's conflict with the traditional schools 
of law demonstrates, there are real questions as to the extent to which 
the institutional needs of the community trump 'scholarly integrity' in 
the criticism of attributions to the Prophet. The Sahihayn canon was 
shaped by communal needs and priorities as they shifted over time. 
What does the Muslim community need today? 



Mzam ad-din Awliya recorded by Amir Hasan Sijzi, trans. Bruce B. Lawrence (New York: 
Paulist Press, 1992), 200. 



APPENDIX I 
REFERENCES FOR &4////i4r^" NETWORK CHART 



This appendix provides the references for the material presented in 
Chapter Four's Sahihayn Network Chart. It is organized by the regions 
shown in the chart, with chronological distribution within each 
region. 



Baghdad 

Ibn Rumayh Abu Sa'id Ahmad b. Muhammad al-Nasawi 

(d. 357/967-8): al-Khatib, Tdrlkh Baghdad, 5:210-11; al-Dhahabi, 
Tadhkirat al-huffdz, 3:96. 

Al-Daraqutni, 'All b. 'Umar (d. 385/995): al-Ghassani, Tanblh, 
39; Brown, "Criticism of the Proto-Hadith Canon." 
Al-Lalakat, HibataUah b. al-Hasan b. Mansur (d. 418/1027- 
28): al-Khatib, Tdrikh Baghdad, 14:71-2; al-Dhahabi, Tdnkh al-isldm, 
28:456-7; idem, Tadhkirat al-huffaz, 3:189. 

Al-Barqani, Abu Bakr Ahmad b. Muhammad b. Ahmad (d. 425/ 
1033-34): Al-Khatib, Tdrikh Baghdad, 5:137-40; Ibn al-Jawzi, al- 
Muntazam; 14:281-2, 333, 379, 15:242; Ibn al-Salah, Tabaqdt al-Juqahd' 
al-shdji'iyya, 1:363-5; al-Dhahabi, Siyar, 17:464-8; idem, Tadhkirat al- 
huffaz, 3:183. 

Al-Dimashqi, Abu Mas ud Ibrahim (d. 401/1010-11): al-Khatib, 
Tdrikh Baghdad, 6:170-1; al-Dhahabi, Tadhkirat al-hufffaz, 3:180. 
IQialaf b. Muhammad al-Wasiti (d. 400/1010): al-Khatib, Tdrikh 
Baghdad, 8:329-30; al-Dhahabi, Tadhkirat al-hufffaz, 3:179-80; al-Kattani, 
al-Risdla al-mustatrafa, 125. 

Al-IUiallal, Abu Muhammad al-Hasan b. Muhammad Abi 
Talib b. al-Hasan (d. 439/1047): al-Khatib, Tdrikh Baghdad, 
7:437-8; al-Dhahabi, Tadhkirat al-hufffaz, 3:205; idem, Tdrikh al-isldm, 
29:471-2. 



380 APPENDIX I 

Egypt and the Hijdz 

Ibn al-Sakan, Abu All Said b. Uthman al-Bazzaz (d. 353/964): 

al-Dhahabi, Tadhkirat al-huffaz, 3:100; idem, Tdnkh al-isldm, 26:88-9. 
Abu Dharr al-Harawi, Abdallah b. Ahmad (d. 430/1038): 

'Abd al-Ghafir al-FarisI, Tdnkh Naysdbur, 607; al-Dhahabi, Tadhkirat 
al-huffdz, 3:201-3, 244. 



Jurjan 

Muhammad b. Muhammad Abu Ahmad al-Jurjani (d. 373— 
74/983-85): al-Khatib, Tdrlkh Baghddd, 3:441; al-Dhahabi, Tdnkh 
al-isldm, 26:549. 

Ibn Adi, Abdallah Abu Ahmad (d. 365/975-6): al-Khalili, 
al-Irshdd, 291-2; al-Sahmi, Tdrikhjurjdn, 106; al-Dhahabi, Tadhkirat al- 
huffdz, 3:102-3; idem, Tdnkh al-isldm, 26:241. 

Al-Ismatli, Ahmad b. Ibrahim Abu Bakr (d. 371/981-2): 
al-Khalill, al-Irshdd, 291; al-Sahml, Tdrikh Jurjdn, 87; Ibn al-Jawzi, 
Muntazam, 14:281-2; Ibn al-Salah, Tabaqdt al-Juqahd' al-shdji'iyya, 417- 
418; al-Dhahabi, Tadhkirat al-hujfdz, 3:106-7; al-Subkl, Tabaqdt al- 
shdji'iyya, 3:8. 

Al-Ghitrifi, Abu Ahmad Muhammad b. Ahmad (d. 311/911-^): 
al-Khalill, al-Irshdd, 292; al-Sahmi, Tdrikhjurjdn, 488; al-Khatib, Tdrikh 
Baghddd, 5:43; al-Dhahabi, Tadhkirat al-hujfdz, 3:120-22; idem, Tdrikh 
al-isldm, 26:614-5. 



Naysabur 

Abu Bakr al-Fadl b. al- Abbas al-Sa'igh al-Razi (d. 270/883): 

Abu Zur'a al-Razi, Kitdb al-du'afi' wa ajwibatuhu 'aid as 'Hat al-Bardha'i, 
2:674; al-Khatib, Tdnkh Baghddd, 12:363; al-Dhahabi, Tadhkirat al-hujjdz, 
2:133-4; idem, Tdnkh al-isldm, 20:149-50. 

Ibn Raja', Abu Bakr Muhammad b. Muhammad al-Naysaburi 
al-Isfarayini (d. 286/899): Ibn al-Salah, Siydnat Sahih Muslim, 89; 
al-Dhahabi, Tadhkirat al-hujfdz, 2:186; idem, Tdnkh al-isldm, 21:288. 
Al-Bazzar, Abu al-Fadl Ahmad b. Salama al-Naysaburi (d. 286/ 
899): al-Khatib, Tdrikh Baghddd, 4:408; al-Dhahabi, Tdrikh al-isldm, 
21:59-60; idem, Tadhkirat al-hujjdz, 2:156. 



REFERENCES FOR X4//Iffi4rA'' NETWORK CHART 381 

Ibn al-Jarud, Abu Muhammad Abdallah b. All (d. 307/919- 

20): al-Dhahabi, Tadhkirat al-huffaz, 3:12-3. 

Al-Hiri, Abu Ja far Ahmad b. Hamdan (d. 311/923-4): al- 

Khatlb, Tdrikh Baghdad, 4:337-8; Ibn al-Salah, Siydnat Sahih Muslim, 88; 
al-Dhahabi, Tdnkh al-isldm, 23:402-3; idem, Tadhkirat al-hujfaz, 2:232. 
Abu Awana, Ya'qub b. Ishaq al-Isfarayini (d. 312/924-5-316): 
al-Dhahabi, Tdrikh al-isldm, 23:525-6. 

Al-Sarraj, Abu al- 'Abbas Muhammad b. Ishaq b. Ibrahim 
(d. 313/925): al-Khallll, al-Irshdd, 310-11; al-Khatib, Tdrikh Baghddd, 
1:264-7; al-Dhahabi, Tdrikh al-isldm, 23:462-4; idem, Tadhkirat al-huffdz, 
2:215. 

Ibn 'Ammar al-Shahid, Muhammad b. Ahmad b. Muhammad 
(d. 317/929-30): al-Dhahabi, Tadhkirat al-huffdz, 3:37; idem, Tdrikh 
al-isldm, 23:546-7. 

Al-Juwayni, Abu 'Imran Musa b. al- 'Abbas al-Naysaburi (d. 323/ 
934—5): al-Dhahabi, Tadhkirat al-hujfdz, 3:27; idem, Tdrikh al-isldm, 
24:139-40. 

Al-Baladhuri, Abu Muhammad Ahmad b. Muhammad b. 
Ibrahim al-Tusi (d. 329/940-1): al-Dhahabi, Tadhkirat al-hujfdz, 
3:72; idem, Tdrikh al-isldm, 25:169. 

Al-Qurtubi, Abu Muhammad Qasim b. Asbagh al-Maliki 
(d. 340/951): al-Dhahabi, Tadhkirat al-huffdz, 3:49; idem, Tdrikh al- 
isldm, 25:192-3; al-KattanI, al-Risdla al-mustatrafa, 20. 
Abu 'All al-Naysaburi (d. 349/960): Ibn Manda, Shurut, 71; al- 
Khatib, Tdrikh Baghddd, 8:70-2; al-Dhahabi, Tadhkirat al-huffdz, 3:80; 
Ibn Hajar, Hady al-sdri, 13. 

Al-Umawi al-Qazvini, Abu al-Walid Hassan b. Muhammad b. 
Ahmad (d. 344/955): al-'Abbadi, Kitdb Tabaqdt al-Fuqahd' as-Sdfi'iyja, 
74; Ibn al-Salah, Siydnat Sahih Muslim, 90; al-Dhahabi, Tadhkirat al-huffdz, 
3:75; idem, Tdrikh al-isldm, 25:417-8. 

Al-Tusi, Abu al-Nadr Muhammad b. Muhammad b. Yusuf 
(d. 344/955): al-'Abbadi, Kitdb Tabaqdt al-Fuqahd' as-Sdfi'iyja, 77; 
al-Dhahabi, Tadhkirat al-huffdz, 3:73; idem, Tdrikh al-isldm, 25:311-12; 
MuUa Khatir, Makdnat al-Sahihoyn, 176. 

Ibn al-Akhram, Abu 'Abdallah Muhammad b. Ya'qub al- 
Naysaburi (d. 344/955): al-Khallll, al-Irshdd, 315; al-Dhahabl, 
Tadhkirat al-huffdz, 3:55; idem, Tdrikh al-isldm, 25:312-3; cf. Ibn Manda, 
Shumt, 73. 

Al-Hiri, Abu Sa'id Ahmad b. Abu Bakr Muhammad (d. 353/ 
964): al-Khatib, Tdrikh Baghddd, 5:225-6; al-Dhahabi, Tadhkirat al-i 
3:89; idem, Tdrikh al-isldm, 26:84. 



382 APPENDIX I 

Abu al-Hasan al-Naysaburi, Muhammad b. al-Hasan (d. 355/ 
966): al-Dhahabi, Tadhkirat al-hujfaz, 3:68. 

Al-Shariki, Abu Hamid Ahmad b. Muhammad b. Sharik al- 
Harawi (d. 355/966): al-'Abbadi, Kitdb Tabaqdt al-Fuqahd' as-Sdfi'iyya, 
58; Ibn al-Salah, Siydnat Sahih Muslim, 89; al-Dhahabi, Tdnkh al-isldm, 
26:227-8. 

Al-Zaghuri, Abu 'All (d. 359/969-70): al-Khatib, Tdnkh Baghddd, 
13:102; Ibn al-Salah, Siydnat Sahih Muslim, 71. 

Al-Shammakhi, Abu 'Abdallah al-Husayn b. Ahmad (d. 372/ 
982): al-Khatib, Tdrikh Baghddd, 8:8-9; al-Dhahabi, Siyar, 16:360-1. 
Ibn Dhuhl, Abu 'Abdallah Muhammad b. al- 'Abbas al-HarawT 
(d. 378/988): al-Khatib, Tdrikh Baghddd, 3:335~7; al-Dhahabi, Tdrikh 
al-isldm, 26:634-5; idem, Tadhkirat al-hujjfdz, 3:141, 158. 
Al-Masarjisi, Abu 'All al-Husayn b. Muhammad (d. 365/976): 
'Izz al-Din Ibn al-Athir, al-Luhdh fi tahdhih al-ansdb, 2:147-8; al-Dhahabi, 
Tadhkirat al-huffdz, 3:110-11; idem, Tdnkh al-isldm, 26:337-8. 
Ibrahim b. Muhammad Abu Ishaq al-Muzakki (d. 362/ 
973): al-Khatib, Tdrikh Baghddd, 6:165-7; al-Dhahabi, Tdrikh al-isldm, 
26:289-90. 

Abu Ahmad Muhammad b. Muhammad b. Ahmad al-Hakim 
(d. 378/988): al-Hakim al-Naysaburi, Tdrikh MshdbUr, 187; al-Dhahabi, 
Tadhkirat al-huffdz, 3:123-4. 

Al-Jawzaqi, Abu Bakr Muhammad b. 'Abdallah b. Muhammad 
(d. 388/998): Ibn al-Salah, Siydnat Sahih Muslim, 89; al-Dhahabi, Tadh- 
kirat al-huffdz, 3:146; idem, Siyar, 16:493-4. 

Al-Armawi, Abu Ishaq Ibrahim b. Muhammad (d. 428/1036— 
7): al-'Abbadi, Kitdb Tabaqdt al-Fuqahd' as- Sdji'iyy a, 100; 'Abd al-Ghafir 
al-Farisi, 153; al-Dhahabi, Tdrikh al-isldm, 29:213. 
Ibn Manjawayh, Abu Bakr Ahmad b. 'All al-Isbahani (d. 428/ 
1036-7): al-Dhahabi, Tadhkirat al-huffdz, 3:191; idem, Tdrikh al-isldm, 
29:208-10. 



Isfahan 

Abu al-Shaykh Abu Muhammad 'Abdallah b. Muhammad b. 
Ja'far al-Isbahani (d. 369/979): al-Khatib, Tdrikh Baghddd, 10:117; 
Ibn al-Salah, Siydnat Scih^h Muslim, 61; al-Dhahabi, Tadhkirat al-huffdz, 
3:105-6; idem, Siyar, 16:276-80. 



REFERENCES FOR X4//Iffi4W NETWORK CHART 383 

Al-Shirazi, Abu Bakr Ahmad b. Abdan of Ahwaz (d. 388/998): 

al-Khalili, al-Irshad, 335; al-Dhahabi, Tankh al-isldm, 27:161. 

Ibn Manda, Muhammad b. Ishaq (d. 395/1004-5): al-Dhahabi, 

Tdnkh al-isldm, 27:320-4; idem, Tadhkirat al-huffdz, 3:158. 

Ibn M ardawayh, Abu Bakr Ahmad b. Musa al-Isbahani (d. 416/ 

1025-6): al-Dhahabi, Tadhkirat al-huffdz, 3:169. 

Abu Nu'aym al-Isbahani, Ahmad b. Abdallah (d. 430/1038): 

al-Dhahabi, Tdnkh al-isldm, 29:274-280; Ibn al-Najjar, Kitdb al-radd, 145; 

"Abu Nu'aym al-Isbahani," Dd'erat al-madref-e bozorg-e esldmt, 6:339. 

Al-Milanhi, Sulayman b. Ibrahim al-Isbahani (d. 486/1093): 

Ibn al-Jawzi, al-Munta^am, 17:6; al-Dhahabi, Tdnkh al-isldm, 33:17305; 

al-Albani, Fihris makhtutdt Ddr al-Kutub al-^dhiriyjia, 550. 



Transoxiana 

Abd al-Samad b. Muhammad Ibn Hayyawayh (d. 368/978-9): 

al-Khatib, Tdnkh Baghddd, 11:43; al-Dhahabi, Siyar, 16:290-1. 
Hamd b. Muhammad Abu Sulayman al-Khattabi (d. 388/998): 

Ibn al-Jawzi, al-Munta^am, 14:129; al-Subkl, Tabaqdt al-shdjiiyya al-kubrd, 

3:284-90; al-Dhahabi, Tdnkh al-isldm, 27:166-7; idem, Tadhkirat al-huffdz, 

3:149-150. 

Abu Nasr Ahmad al-Kalabadhi (d. 398/1008): al-Khatib, Tdrikh 

Baghddd, 5:201; al-Dhahabi, Tadhkirat al-huffdz, 3:154-5; idem, Tdrikh 

al-isldm, 27:355. 

Umar b. All Abu Muslim al-Laythi al-Bukhari (d. 466-8): 

al-Dhahabi, Tadhkirat al-huffdz, 4:24. 



APPENDIX II 

THE QUESTION OF THE ATTRIBUTION 
OF THE SAHIHATN 



Several scholars have argued that the texts of the Sahihayn did not sta- 
bilize until some time after the deaths of their authors. In light of such 
realities as "organic texts, pseudepigraphy and long-term redactional 
activity," Norman Calder claimed, "Apparently the product of the 
devoted and orderly activity of a single person, works like the Sahths 
of al-Bukharl and Muslim should probably be recognized as emerging 
into final form at least one generation later than the dates recorded 

for the deaths of the putative authors '" Based on his analysis of a 

partial fifth/eleventh-century manuscript of Sahih al-Bukhdri, Alphonse 
Mingana concluded that the text was still in a relatively fluid form at 
that point in time. Yet there is little available evidence suggesting that, 
beyond the normal permutations of manuscript transmission for texts 
as large and detailed as the Sahihayn, either al-Bukharfs or Muslim's 
books were altered substantially after their deaths. 

The Sahihayn are two massive works, and the vagaries of manuscript 
transmission introduced the possibility of frequent variation even for a 
text transmitted intact from its author. Several generations of editors, 
such as Abu Dharr al-Harawi (d. 430/1038), al-Saghani (d. 650/1252) 
and the Egyptian Hanball al-Yunini (d. 658/1260), thus played impor- 
tant roles in collating different transmissions of S^hih al-Bukhdri into 
vulgate editions.^ Such editorial review, however, was endemic to the 
pre-print world and does not reflect any instability specific to the 
Sahihayn. 

Mingana based his assertion that al-Bukhari's Sahih remained in 
fluid form through the early fifth/ eleventh century on his observation 



' Calder, Studies in Early Muslim Jurisprudence, 194. 

^ For discussions of these different editors and their contributions, see Mingana, 
An Important Manuscript of the Traditions of al-Bukhdri, 16—18; Rosemarie Quiring- 
Zoche, "How al-Buhan's Sahih was edited in the middle ages: 'All al-Ytlnlnl and his 
RumUz," Bulletin d'Etudes Orientales 50 (1998): 191-222; and Johann Fiick, "Beitrage 
zur Uberlieferungsgeschicte von Buharl's Traditionssammlung," ^eitschrift der Deutschen 
Morgenldndischen Gesellschcft 92 (1938): 60-82. 



THE C)UESTION OF THE ATTRIBUTION OF THE SAHlHAYN 385 

that two of the chapters of the manuscript that he examined were 
out of normal order and that each narration began with "al-Bukharl 
informed us...," a feature not found in the dominant recensions of the 
text.^ Yet Mingana's partial manuscript of the Sahih consisted of only 
three chapters. We have no evidence that the ordering of the remaining 
ninety-four chapters was irregular. 

Besides Mingana's unconvincing evidence, there are other indica- 
tions that al-Bukhari's Sahih varied slightly in content as it was trans- 
mitted from its author through his various students. We know from 
al-Kalabadhi that al-Bukharl was transmitting his Sahih during his 
own lifetime. Al-Kalabadhi informs us that al-Bukhari had been nar- 
rating his Sahih to students for at least eight years before his death.* 
This is corroborated by al-Bukharl's own student, al-Tirmidhi (d. 
279/892), who mentions al-Bukharl's S^hW in his, Jdmi':' As the author 
was almost certainly making adjustments to his work throughout his 
life, it should not surprise us that the different narrations of the S^hih 
from al-Bukharl's students varied from one another When compared 
with the enduring transmission of the Sahih from al-Bukhari's most 
famous student, al-FirabrI, his other student Hammad b. Shakir's 
(d. 290/902-3) recension of the text contained two hundred fewer nar- 
rations. Ibrahim b. Ma'qil al-Nasafi's (d. 295/907-8) was three hundred 
less.** But according to Ibn Hajar's count, the Sah^h contains a total of 
9,082 narrations of all sorts.' We should thus not consider a variation 
of three hundred narrations, roughly 3% of the S^hih, evidence of an 
incomplete or fluid text. 

The other major piece of evidence suggesting that al-Bukhari's Sahih 
was edited significantly after his death has been Abu Ishaq al-Mustamll's 
(d. 376/986—7) statement that, upon examining his teacher al-Firabrl's 
copy of