Origins_of_Isma__039_ilism__A_Study_of_the_Historical_Background_of_the_Fatimid_Caliphate

8/7/2019 Origins_of_Isma__039_ilism__A_Study_of_the_Historical_Background_of_the_Fatimid_Caliphate

1/121

THE ORIGINS OFISMA'ILISMA study of the historical background of theFatirnid Caliphate

BYBERNARD LEWIS, B.A., PH.D. (London)

Lecturer In1lamu: History,School of Oriental and African Studies, University of London.

Et, une heure , jt' sues dcsccndu dans lemouncmcnl d'un boulevard de Bagdad oftdes compagnies ant chanie la joie dutrauail nouveau

ARTHUR RIMBAUD,Les Illumt naiions

CAMBRIDGEW. HEFFER & SONS LTD


2/121

FIRST PUBLISHED MARCH, J940

Printed and Hound ,n Great Britain at the Worh ojw. HhFFER " SONS LTD, CAMURID&E, ENGLANU


3/121

PrefaceIN every civilisation there are certain movements of socialand intellectual revolt, indicative of the reaction of thesuppressed and dissatisfied elements in that civilisation toprevailing conditions. The history of these movements,usually written exclusively by their opponents, is at onceof peculiar difficulty and peculiar value to the historian.In a predominantly religious society like that of mediaevalIslam, where State and Church were fused and treason andapostasy were synonymous terms, such movements neces-sarily assumed a theological colour, and appeared asheretical sects, revolutionary in doctrme and insurrectionaryin habits. In the early centuries of Islam we find a wholeseries of sects which, by their simultaneous challenge tothe Islamic faith and state, brought upon themselves theconcentrated wrath of rulers and theologians.Of these sects, most of which are known to us only fromthe writmgs of their enemies, by far the most significant isthat of the Isrna'Iliva, or Isma'Ilis (also called Batinis,Ta'Iimis, etc.). This sect rapidly acquired an Importancefar exceeding that of its competitors, and soon constituteda really serious threat to orthodox Islamic society. Forseveral centuries Ismfi 'Ili missionaries in all parts of theempire stirred up rebellion against the 'Abbasid Caliphs,and I sma'Ili philosophers elaborated a detailed system ofreligious thought, oftcnng a seductive alternative toorthodoxy. The importance of the sect is enhanced by itsconnection with the rise of the Fatirnid dynasty, the mostpowerful in mediaeval Egypt. This connection raisesmany problems, most of which have not yet been properlyelucidated.

Since the appearance of De Goeje's . 1 1 , [ emoire sur lesCarmathes in 1886, no detailed study has been publishedon the origins of the Isrna'ili movement, from the historicalviewpoint. A considerable amount of new material, ofvarious kinds, has become available since then, and althoughmany works still lie unknown in the Isrna'Ili libraries of

v


4/121

VI PREFACEYemen and India, enough have come to light to rendernecessary a restatement and reconsideration of the problemsinvolved.

I wish to record my indebtedness to Professor LouisMassignon, of the College de France, whose advice andassistance were of inestimable value to me throughout thecourse of my work. My thanks are also clue to ProfessorH. A. R. Gibb, now of Oxford, for much help and encourage-ment. I am grateful to the following for useful suggestionsand aid of various kinds :-Profp:;~or A. S. Tritton (who inaddition undertook the painful task of reading my proofs);Professor N. H. Baynes; Dr. P. Kraus; Dr. J Heyworth-Dunne; Dr. G. H. Sadirrhi , M. Abbas Eghbal; Mr. M. H.A'zami.

I should abo hke to place on record my gratitude to thelibranans of the School of Oriental and African Studies,London; the Ecole de Langue" Oricntale-, Vivantes, Paris;and the Egyptian Library, Cairo, for their unfailing courtesyand assistance. Finally, a word 1" due to my Isrna'Ilifriends in Masyaf, Salamiya and cls-wlu-n-. for their useful,if somewhat reticcn t _ co-operation.

This monograph IS a re-vised version of a thesis acceptedfor the degree of Ph.D. 1 1 1 the University of London. Myspecial thanks are due to the University of London Publica-tions Committee for a grant that rendered its publicationpossible.CAMBRIDGE,

January, 1940.


5/121

To My PARENTS


6/121

ContentsPREFACEINTRODUCTION -Introductory Remarks-Survey of Sources-Sunni, Shn, Isma'Ili=-Statemcnt of Problemsinvolved-Review of Various Solutions Offered-

Main Lines of a New InterpretationCHAPTER 1. ISMA'ILT OmGINS 23

(a) Intr(Jductory-The Beginnings of Shi'rsm->-ThE' Mahdi Idea-Early Shi'I Sects and Ten-dencies

(0) The Birth of Ismu'Ut.\m-Abii'l-KhaWi.b-Isma'Il=--Evcnts on the De-ath of Ja'farCHAPTEI< II. THE HIDDEN IMAMS A~DTHEIR HELPERS 44Spmtual Parcnthood=-Adoptive Imilmate and

Trusto ITJl:iIJl:Lte---Ong1l1s of the Fatl!nid Cali-plrate-c-Murmlin and 'Abdallah b. Maimiin al-Qaddll) --- UlIldin - The Hidden Imam- - TheDa' (ta-Syria and Mesopotanna-> Ycrnen

PAGEVI

CHAPI EI


7/121

THE ORIGINS OF ISMA'ILISM


8/121


9/121

2 THg ORIGINS OF ISMA'ILISMWith 'Alid legitimism as political programme (with

what important modification we shall see later), a syncretisthatch-patch of all faiths and philosophies, with a strongundercurrent of pure rationalism, as doctrine, and socialgrievances and organisation as an important part of itsactivities, the movement stood a good chance of canal isingthe whole social and religious discontent that was rife inthe mediaeval caliphate.

The characteristic Isma'ili combination of syncretismand esoterism render the investigation of the history of thesect a task of tremendous dtfficulty. This is not simplifiedby the prejudices and animosities of the hostile sourcesupon which we are forced to rely for most of this history.For the annals of the Isrna'Ili movement, at least in Itsearlier stages, must still be based largely on the records ofnon-Isma'Ilis, with the defects involved by way of ignoranceand bias.

The fault does not lie wholly with the Sunni andTwelver Shii writers who tell the story. The secretrvecharacter of the movement, its quasi-masonic organisation,and the haze of mvstery conceahng both doctrmcs andpersonalities from the uninitiated, did not facihtnte thetask of the historian, and have prevented until moderntimes a clear and accurate understanding of Isma'ili origins.

With the discovery of new documents, and the con-sequent revaluation of old ones, a critical study of earlyIsma'ili history has at last become possible, and scholarslike Massrgnon, Kraus, Ivanow, Hamclam and others havemade notable contributions. The line of study, however,has hitherto been mainly doctrinal and htcrary, and, savefor one or two spe-cial questions, the historical side hasremained substantially as De Sacy and De Goojc left it.The following pages constitute an attempt to clear theground for a general historical survey of the beginnings ofIsrna'Ilism.

SURVEY OF SOURCESIt has seemed best to start this study with a survey of

the primary sources at our disposal. Although many of


10/121

INTRODUCTION 3these sources have been studied before, it is necessary toattempt a revaluation of them in the light of the new sourcesavailable, several of which are here used for the first time.

Our first group consists of the Sunni historical writings.Here we can trace the successive stages in which the trueknowledge of Isma'il! doctrmes and the early, secret historyof the sect filters through to the Sunni world. It is in orderto illustrate the stages of this highly Important process thatI have considered these sources in some detail. The pene-tration of this knowledge can be classified in three stages-tho first, where the' historian knows of nothmg but thepublic activities of the scctarios-i-thc second, in which someinklings of what I" at stake have begun to penetrate, butno general conception IS a~ yet possible-vthc third, whendetaile-d, though not always accurate, knowledge of thesect, Its doctrines and it-, ongms, reach the Sunni world.The third ;.,tage IS marked by the famous Baghdad mani-festo, denouncing the Fatimid caliphs as impostors andschismatics.

SUNNI HISTORICAL SOl'RCESThe First Stage.

Tlu- earliest account we P0;'SC5;., is that of the greatSunni histonan Abii Ja'far Muhammad h. Jarir at-Tabari(d. 31I/922). Tabar! obviously represents the earlieststage of Sunnl knowledge rt'barumg the Batini movement.He has htt l acquamtancc WIth the doctrines of the sects,or with their inner differentiations. In spite of his notingthat the Carrnathian leaders m Syna claimed descent fromMuhammad b. Istna'il. he docs not think of connecting themwith the Fatimid pretender whose appearance he mentionsin N. Africa. HIS viewpoint on the whole may be taken asrepresenting that of tile average Baghdidl of his day,possessing little information beyond the actual reports ofevents and occasional "craps of doctrine. Tabari's account,which hegins in 27S/891, with the first Carrnathian missionin 'Iraq, and ends in 294/906 with the suppression of theCarmathian revolt in Syria, is purely factual, and, except


11/121

4 TH1~ ORIGINS OF ISMA'ILISMfor a brief notice of an alleged Carmathian book, makes noattempt to discuss the doctrines of the sect.Tabari's history was resumed and continued till 320/932by 'Arib b. Sa'd of Cordoba (d. 370/980). Under the years291, 293, and 294, 'Arib recapitulates the events of thoseyears as set forth in Tabari, anel continues the story withan account of the activities of the Carmathians in Bahrainunder Abu Sa'Id and Abu Tahir. Quite separately fromthis, 'Arib gives a detailed history of the Fatirnid da'ina inN. Africa, and the final victory of the Mahdi 'Ubaidallah,Itis to be noted that he too makes no connection betweenthe two movements.

Mas'Iidi (d. 344/956), in both his Tallbih and his Muruj,devotes a few pages to the Carrnathians, carrying theirhistory on to the death of Abii Tahir In332/944. Mas'Iidiobviously represents a later stage of Sunni knowledgeregarding the Carmathrans than does Tabari. He has readIbn Rizarn.! he is acquainted WIth Iifltini doctrine to thepoint of knowing about the esoteric ~y"tl'lI1of interpretationand the grades of initiation. Finally, he is aware of theconnections between the Carmathians and the Fatimidsin Yemen and N. Africa. 2

Unfortunately, Mas'Iidi does Iittk- more than mentionthese things, the pa",.,agp,., in the Tanbili being merely arecapitulation of his own mort- dctaili-d works on thesubject. These, alas, arc lost. Nor 1;' it possible to deter-mine exactly the sources of Ma-s'fidi's knowledge, though itwould seem, from his general tonc and from one or tworemarks, that much of it was obtained at first hand, fromconversations with the Carmathians themselve-s.

Although, therefore, in knowledge, :'1as'udi goes beyondthe other historians of the first stage, aud even beyondsome of the second, we must place him here' owing to thepaucity of detail that he gives us, and to the fact that hiswork does not go beyond the year 332/944.The only other important historian of this period isHarnza of Isfahan (Iourth zninth century). His account ofthe Carrnathians is limited to their military exploits, and

Iv. injra, p D. a T'anbih, 38'1 Tr. 495


12/121

INTRODUCTION 5makes no reference either to their beliefs or to their con-nections. It is interesting, however, for the picture it givesus of the disturbed state of Baghdad and the Empire duringthose troublous years, of the continuous discord and civilwar, the grave social and economic crisis, and the acuteconflict of classes between the Khli$$a (aristocracy) and the'}[mma (common people) of the capital. This picture doesmuch to explain the atmosphere of discord and discontentin which revolutionary Carmathianism emerged andflourished.The Second Stage.

For the period after J20, De Goeje was forced to rely onlater sources Iike the Kitdb al-'Uyiin and Ibn al-Alliir(d. 631/1234). Since then, however, a number of late'Abbasid chronicles have been published, notably those ofHilal as-Sabi (d. 447/1055) and Miskawaih (d. 421/T030),both of which contain much useful information.

During a recent visit to Cairo I had the good fortuneto acquire a manuscript of what appears to be a fragment ofthe lost history of Thabit b. Sinan the Sabian (d. 365/974).the grandson of the farnous physician Thabit b. Qurra.1

I shall reserve a more detailed discussion of the manu-script for an edition of the text which I hope to publishshortly. Suffice It to ~ay here that the manuscript is dated1057/1647, arid was made from a copy written in 577/nS1,which in turn was made from a copy of the author's ownmanuscript. There seems to me to hI' no good reason fordoubting the authenticity of the manuscript.

Thabit gives the history of the Carrnathians in Bahrain,Syria and Mesopotamia from their origins until the year ofhis death. His account of the war between the FatimidCaliph Mu'izz and the Carmathians is particularly detailed,and bears every sign of being a contemporary record.Thli.bit's account is followed closely by Miskawaih and Ibnal-Athir, of which it seems to have constituted the mainsource for the period.

Thabit represents a more advanced stage of knowledgeIOn Thiiblt and his history set' Chwolsson, 5,,,bIeY, I, 57Y


13/121

6 TH~ ORIGINS OF ISMA.'ILISMthan does Tabari, In several ways he shows himself to beaware of a connection between the Carmathians and theN. African Fatimids. Yet he does not say anything oftheir doctrine, nor is he aware of the earlier history of thesect, before its public eruption. The names of Abii'l-KhaHab, Maimiin al-Qaddah and 'Abdallah b. Maimiinal-Qaddah are not mentioned, and it does not occur to him toquestion the Fatimid legitimacy of 'Ubaidallah, to whom heinvariably refers as "al-Fdiimi al-l Alatei:"

Before passing to the third stage, a word is necessaryin justification of the inclusion of Thabit b. Sinan amongthe Sunni sources. As is well known, the author's familywere all Sabians, and he himself was of that faith. Likeother Sabian writers, however. he assimilates himself com-pletely to the Sunni point of view, and is not distinguishablefrom them, as are, for example, the Twc-lver Shi'is, in hischoice of sources and his method of approach. His historywas much appreciated and frequently utilised by Sunnischolars, and cannot be separated from the Sunni historicaltradition.The Thzrd Stage.

In the third stage, the knowledge of Isma'il! doctrmespossessed by Sunni writers has reached a far greater degreeof development. Fairly detailed information on Ismailidoctrines is available, and, notably, what we may term theproto-history of the sect--thc hi-tory of it;, early secretdevelopment, before its propaganda first burst upon publicnotice-is far better known.

The earliest known writer of this third stagp is Abii'Abdallah b. Rizarn (or Razzam), who hved probably in theearly part of the fourth/tenth century. Although he wouldseem to have been a theological rather than an historicalwriter, he may be included among the historical sources,owing to the fact that his account is preserved in workswhich belong rather to the historical group of wntings, andto the fact that he does mark the beginning of a definitetrend of histonography on the Carmathians. Among thehistorians who follow the lines laid down by him in their


14/121

INTRODUCTION 7accounts of the Carmathians, we may mention the Nizamal-Mulk, Ibn Shaddad, Abfi'l-Fida and RasJ.!id ad-Din.The original work of Ibn Rizarn is lost, but it was utilisedextensively by one Akhii Muhsin, an 'Alid who was roughlycontemporary with Mu'izz.' It is to Maqrizi that we owethe knowledge that Akhii Muhsin's account was copied fromthat of Ibn Rizam.! Maqrizi adds that he himself does notbelieve Ibn Rizam's account of Fatimid origins. In viewof this harsh judgment on Ibn Rizam and his followerAkhii Muhsin, it is curious that Maqrizi makes such ex-tensive and frequently unacknowledged use of it in hisvarious writings."The text of Akhii Muhsin is preserved for us in twoversions. The first of these l~ in Maqrizi, who gives thedoctrinal part in his Khitat' and the historical part in hisMuqafftP The other, and better, version is contained inthe Nihayat at Arab of Nuwairi (d. 732/133I), an encyclo-paedia of literature and history. Nuwairi's account is inthe part that has not yet been published, and is preservedin two manuscripts, one in Pans and one in Istanbul,"The version of Akhii Muhsin is also given briefly by Maqriziin his I!!2'ii? Ibn Rizam is quoted directly by the Fihrist.7He is mentioned in the list of writers on the Carmathiansof whom Mas'Iidi" speaks so slightingly. Casanova' regardsthis as "Ia condarnnation formelle" of Ibn Rizarn.

On the whole, however, one must judge the generalreliability of Ibn Rizam in a rather more favourable light.As has been already remarked, Maqrizl's censure of himloses greatly in effectiveness by the fact that Maqrizihimself makes extensive use of him. Casanova.P too, hasnoticed that despite Mas'udi's criticism of Ibn Rizam, hisown account differs little from that which he condemns.The discovery of actual Isma'ili documents has tended onthe whole to confirm Ibn Rizam's statements. Ivanow-t1Sacy, I, Tntro, 74, Itt .. ii: 11 Casanova, La Doctrine Secrete, p 9,note I.2 JII,'a!!, 12 Quatremt'r(', 1r7. Fagnan. 39 3E g Inthe Kl!ltat. I, 391. Tr. Casanova, La Doctrine Secrete. Trans. Quatremere and Fagnan See Bibhography. ' r86 Tanbih, 395 Tr 50r La Doctrine Secrete, 36.10 Ibui , 37. 11 Creed, 14.


15/121

8 T~'IE ORIGINS OF ISMA'ILISMnotes that the formula of oath given by Ibn Rizam corre-sponds closely with that still in use among the Isma'ills ofIndia, and that the doctrines he attributes to the Isma'Ilis"fit very well" with those of an Isma'Ili treatise by Nasirad-Din Tusi,l though, it should be observed, there is notrace there of the nine degrees of initiation mentioned byIbn Rizam, and also to be found in other writers,

Of the historical defects of Ibn Rizarn we shall speaklater.

With Ibn Rizam, the names of Mairniin al-Qaddah and hisson 'Abdallah appear in Sunni history books, the Fatimidmovement is traced back to its origins in their coterie, therelationship between Fatimids and Carmathians is estab-lished, and the pedigree of the Fatirmd caliphs called intodoubt, all for the first time,

In 40Z/IOII was published III Baghdad the famous anti-Fatimid manifesto, in which some of the assertions of IbnRizam were publicly proclaimed. 'Ubaidallah, the firstFatirnid caliph, is asserted to have been in reality oneSa'id, a descendant of Daisan, founder of the Daisani(Bardesanian?) sect, and himself a dualist and an infidel.'Abdallah b. Maimiin and his father arc not, however,mentioned.This manifesto, supplemented by the extra detailsgiven by Ibn Rizam, remained the basis of most anti-Fatimid writing.

Among the later Sunni sources only one need be noticedhere, as it is utilised for the first time in this study. In theYemenite chronicles published and translated by H. C. Kay,there is a brief account of the Carmathians in Yemen,extracted from the Kttcib as-Sulak of Baha ad-Din al-Janadi(d. 732/1331). AI-Janadi quotes as his sale authority oneAbu 'Abdallah Muhammad b. Malik b. Abi'l-Qaba'il, "ajurist of Yemen and a learned Sunni, He was one of thepersons who joined the Carmathian sect in the days of theSulaihi," and he acquired a thorough knowledge of its

'An isma,lllic Work by NaSIr. 534 The Isma'Ill da.'1 .All h. Muhammad (429-483/1037-1080), founderof the Isma'tlt Sulaihid dynasty 10 Yemen.


16/121

INTRODUCTION 9character. On becoming convinced of the depravity of theCarmathian doctrines, he abjured them, and he composeda celebrated treatise, in which he has described the principlesupon which they are founded, he demonstrates their wicked-ness, and warns his readers against their deceptions."!

Early in 1939 a work was published in Cairo which wouldappear to be the treatise in question. It is called KahfAsrar al-Ba#niya wa-Akhbar al-Qardmua, and the author'sname is given as Muhammad b. Malik b. Abi'l-Fada'il al-Hammadi al-Yamani, From the text it would seem thatthe author was a contemporary of the Fatimid CaliphMustansir (427-87/1035-94).

We have thus concluded our survey of the main trendsin Sunni historiography. There are, of course, many im-portant historians whose names have not been mentioned.These will be referred to where necessary."

Before proceeding, however, to a more detailed dis-cussion of the specific points at issue, it will be necessaryto extend our survey to include other groups of sources,namely, (r) the Sunni non-historical sources; (2) the TwclverShi'I sources; (3) the Isma'ill and near Isma'ili sources.

SlTNNI THEOLOGICAL SOURCESAs is to be expected, a movement as Widespread and as

subversive as Batinism occupies an important place inSunni theological literature. A considerable quantity ofpolemic writings was produced by orthodox theolog-ians,from which it is possible to gam, if not a knowledge ofBatini doctrines, at least an impression of the effect theyhad on Sunni observers.

Slightly less prejudiced are the works on the history ofreligions, a science that reached a high degree of develop-ment in mediaeval Islam.

For obvious reasons, the greater part of this literaturedeals with the theological and doctrinal aspects of Batinismrather than with the history of the movement, and thus fallsbeyond the scope of the present study. There are, however,

1Kay. 191 For a survey of sources on the rise of the Ffitrrruds, see Becker.Beltrage, I, 2--II


17/121

10 TIfE ORIGINS OF ISMA'ILISMscattered bits of historical information in these sourceswhich make an examination worth while,The main historical value of these works is that theydescribe not only the Ismaili movement in the strict senseof the word, but the whole welter of extremist Shi'I sectsfrom which Ismailisrn as an organised system evolved, witha wealth of detail which we cannot hope to find in thestrictly historical sources. The theologians, it is true,have many defects which vitiate their reliability. They arcviolently prejudiced against the sects they describe, andhave no interest in presenting a clear anti accurate pictureof their development. Often enough they arc ill-informed.and attribute to the sectaries doctrines which the latterthemselves would rejc'ct in horror. Mas'Iidi- speaks WIthscorn of the polemists, of their unwillingness to understandand their readiness to condemn. How then should oneexpect them to distinguish accurately between conflictinggroups of heretics? Then again, many of them are boundby a tradition attributed to the prophet, according to whichthe Islamic community would break up into seventy-threesects.> This number IS the Procrustean bvd into whichmost Islamic heresiologies have been fitted, by more or lessviolent means.

In spite of these fault" however, it is possible, by acareful comparison of the different versions, and=-mostimportant of all-by means of the Twelver Shii and Isma'Ilisources now available, to establish a critical account of thegrowth of the early extremist sects, and their culminationin the Isrna'Ili da'ioa. ThIS task will be attempted in ourfirst chapter. For the moment let us review briefly themost important of the Sunni theological sources, includingone not hitherto known."

The carhest extant classification of Islamic sects is theMuqdld: al-Tsldmiyin of the great schola-tic Abir'l-Hasanal-Ah'ari (d. 321/933), a work which attempts a detailed,and on the whole reasonable, survey of the subject. Ah'ariwas followed by Malati (d. 377/987), who is concerned with, Tanbih, 395, Tr 501 l Abu Da'ud, lS


18/121

INTRODUCTION IIrefuting rather than with explaining. but who neverthelesssupplies a few interesting details.More detailed than either of these is the theologianAbii Mansiir 'Abd al-Qahir b. Tahir al-Baghdadi (d. 429/1037). an A~'arite scholastic who. in his Farq bain al-Firaq,provides us with a detailed survey of the Shi'i sects. theirseparate histories and their doctrines. In his chapter onthe Batiniya, Baghdadi gives a lengthy historical account.apparently based, with some variations, on the Ibn Rizamversion.Brief accounts of Batini origins and doctrines are alsoto be found in the Baydn al-Adydn of Abu'l-Ma'ali (com-pleted in 485/1092) and in the first great Muslim history ofreligions, the Fasl [i'l-Milal iea'n-Nihal of Ibn Hazrn(d. 459/1064).Shahrastani (d. 548/II53). the successor of Ibn Hazm inreligious history, abo surveys the field. His attitude issurprisingly tolerant, and he has clearly made use of Isma'Ilisources. Shahrastfini's work, however, is concerned almostexclusively with doctrine, and supplies very little historicalinformation.In a passage recently edited and commented by Dr. P.Kraus.! Fakhr ad-Din Razl, the great Qur'anic commentator.makes an extremely interesting analysis of Shahrastani' swork. It is, he says, a piece of shameless plagiarism frombeginning to end, being merely a copy of a number of otherworks. The part with which we are concerned, namely,that dealing with the Islamic sects. IS, he says, based onthe Farq of Baghdadi, who, o..."ing to his extreme fanaticism,did not bother to reproduce faithfully the ideas of hisopponents. The only passage worthy of note is that dealingwith Hasan i Sabbah, for there Shahrastani used an authenticdocument. translating It into Arabic from the Persianoriginal.

As Dr. Kraus points out, Razi's strictures on Shahrastaniare in the main justified, though in a number of passages,and notably that dealing with the Kayyallya sect.Shahrastani does appear to have used other sources. His

1Les "Controverses," l05 fl. and 2U fl.


19/121

12 TI\E ORIGINS OF ISMA'ILISMgeneral attitude, too, is more liberal than that of Baghdadi,and his acquaintance with Isma'Ili doctrine far more detailed.Itmay be mentioned in passing that Shahrastani was himselfaccused of being an Isrna'Ili.!

One of the most important Sunni works dealing with theBatiniya is the "Refutation" of Ghazali (d. 50S/lIIl),analysed and partly published by Goldziher. This work,too, deals mainly with doctrine, and falls beyond ourpurview. The references to the general character of theBatini movement will be considered later. We may notein passing that Fakhr ad-Din Razi, in the passage quotedabove, accuses Ghaziili of having misunderstood the Ismaillthesis, and finds his refutation very faulty. There is extantan Ismaill reply to Ghazali's work by a Yemenite d{i'f of theseventh/thirteenth century."

Among the theologians we may include the Hanbalipolygraph Jamal ad-Din b. al-j awzl (d. 577/I20o), who intwo passages gives us a short account of the Batini move-ments, based in the main Oil Tabari, Ibn Rizam andGhazall.

To this list of theological sources must be added onewhich is still m manuscript. and is here utilised for thefirst time.

The author was the Qii


20/121

INTRODUCTION 1 3contains much new information. His work is quoted as anauthority by several later historians.'This, then, concludes our survey of the two main groupsof Sunni sources, illustrating the gradual growth of Sunniknowledge about the Batiniya. There are, of course, others.In travel-books, literary and philosophic writings, and otherworks, we shall find much valuable information, which willbe utilised in the following chapters. None of these, how-ever, warrants a separate consideration. It remains forus to review briefly the non-Sunni sources for our subject.TWEL VER ::;HI'I SOURCES

Apart from the above-mentioned, there is also availablea considerable mass of material relative to the early historyof the sects in the biographical and bibliographical com-pendia of the Twelver or moderate Shi'a, This literaturewas quite unknown to De Sacy, De Goeje and other earlyinvestigators, and indeed its importance has only recentlybeen pointed out by M. Massignon."

Naturally enough, the moderate Shi'a were in closercontact with the Isrna'Ilis than were the orthodox Sunnisand had a more detailed knowledge of their history andtenets, Abii'l-'Ala al-Ma'arris tells us that the TwelverShi'a still respected the memory of 'Abdallah b. Maimiinal-Qaddah, and quoted him as an authority for traditionsrelated by him before his "apostasy." This statement isamply borne out by a study of Shi'I biographical works,where 'Abdallah b. Maimiin and other notorious extremists(d!ultit) are cited as traditionists, and many valuabledetails given regarding them which cannot be found else-where.

The earliest and most interesting of these works is theMa'ri/at Akhbtir ar-Rijdl of Abii 'Arnr Muhammad b.'Umar b. 'Abd al-'Azlz al-Kasl1shi,which has come down tous in a later abridgement by Tus l . Kashw, who lived atsome time in the fourth/tenth century, was a pupil of therenowned Shi'I scholars Abii'l-Qasim Nasr b. Sabbah1Nujum, II, 443-6 Abu 8!5.ma,II, 201. SUYlltl-Ta'Yilill,3 Fagnan,Maqrlz! See further, Becker, Beitrage, 5 Brockelmann, I, 4II and 41B2 Esquasse and art Karmaitans m EI. 3 Ghuf.an, 156


21/121

THE ORIGINS OF ISM,A'ILISMal-Balkhi and Abu Nasr Muhammad b. Mas'Iid al-'Ayyasbias-Samarqandi, the former of whom was himself an ex-tremist. Kashshi's work, which is a collection of traditionsabout great men of the Shi'a sect, is a mine of inforrnation.!

Supplementing Kashshi we have a number of later works,all biographical and bibliographical compendia, amongwhich the most important are those of Naja.5hi (d. 4S0/IOS8),Tus i (d. 460/r067), Ibn Shahrashfib (d. S88/II9z) andAstarabadi (d. IOz8/I6IS).2

Apart from these compilations, there are two books byTwelver Shi'I authors which also shed some light on oursubject. These are two works on sects and religions, andare the only works on the subject by Shi'I authors that areextant. They reveal a closer knowledge of the subject thanis possessed by Sunnis, and are free from some, though notall, of the latter's prejudices.

The first of these ts the Ftraq a0l1-Shi'a, a work onShi'i sects, ascribed to Abu Muhammad al-Hasan b. Musaan-Nawbakhti (d. before 310/q22). This book is a surveyof Shi'I and related sects from the death of 'Ali until thedisappearance of the twelfth Imam of the Twelvers. Itiswritten in a remarkably dispassionate style, and containsmuch information on the early history of the Isma'Ilis.

The author is a rather lukewarm Shi'i, and appears touse Sunni sources as much as Shi'i ones. In some respectshe is no better mformed than his Sunni contemporaries, andhe docs not make any mention of 'Abdallah b. Maimiin.He does, however, provide us with a detarlcd picture of thepre-Isma'Ili sects clustering around Muhammad b. al-Hanafiya, Muhammad al-Baqir and Ja'far as-Sadiq, thusgiving us a historical framework into which we may fitthe biographical material in Kashshi and his followers.It should be mentioned in passing that Abbas Eghbal,in his brilliant monograph, Khdneddne Nawbakht,8 rejectsthe authorship of Nawbakhti, whose book of the same nameis, he says, lost. The true author was one Sa'd b. 'Abdallah

'On Kashshl, see Eghbal, Kl!_iinrdiine, 140, and Hadiyat, 226.2 For further details on these and other Twelver ~t'J authors, seeRieu, Unto Mus. Supp! Cat of Arabic MSS. P 422 fl. P. 140fl.


22/121

INTRODUCTION I Sal-Ash'ari (d. 299 or 30I/9II-9I3). M. Eghbal bases hisassertion on a number of passages in Kashshi and others,where texts identical with the Firaq are referred to A~'ari.

The other book is the Tabsirat al-'Awamm, a Persianwork, of uncertain authorship, but ascribed to SayyidMurtada b. Da'i Husni Razl (early seventh/thirteenthcentury).

ISMA'ILl SOURCESFor a long time the Ismailis were judged exclusively

on thc evidence of their bitterest opponents, their ownliterature, thanks to Sunni censorship and Isma'ili secrecy,having almost completely disappeared in the Near East.A few fragments collected among the Isma'Ilis of CentralSyria were published by Guyard, one or two others foundamong the Isma'Ilis of Central Asia were noticed in Russianperiodicals. In 1905 Griffini found a number of Isma'Iliworks in the Yemen, "orne of which he reviewed briefly inan article. During the following years, thanks to the effortsprincipally of W. Ivanow and H. F. Hamdani, the treasuresof the Ismiiili libraries of India were made known, a numberof Isma'Ili works 11 1 European libraries were Identified andothers acquired. By 110W a fair amount of this literature isavailable, in private and public collections, and a numberof texts have been published, translated and studied byvarious scholars.

By far the greater part of the new Isma'ili material is,however, religious and philosophical in character, and sur-prisingly little has been added to our knowledge of the earlyhistory of the sect. Moreover, no book earlier than thereign of the rust Hitimid caliph has yet becn found (withthe possible exception of the Umm al-Kitdb of the Isma'Illsof Central Asia), and the point of VIewthey represent is thatof the emasculated official Fatirnid da'wa rather than theearlier revolutionary period.

The principal Isma'Ili historical work is the 'Uyiinal-Akhbdr of the da't Idris of Yemen,1a seven-volumehistory of Isma'ilism from the marriage of 'Ali until the

1 Ivanow, Gu.de. 62


23/121

16 THE ORIGINS OF ISMA'ILISMauthor's own day (ninth/fifteenth century). This work,except for a few extracts published by Hamdani in DerIslam, is still in manuscript, and I have not been able toconsult a copy.

An earlier work is the Iftita/:t ad-Da'iua wa ibtidd ad-Datola, by the Q 1 i < ; l i Nu'man (chief Q M i of Mu'izz}.! Thiswork, a history of the Fatimid da'toa in Yemen and N. Africauntil the establishment of the caliphate, is of considerablevalue. It is still in manuscript, but I have a copy in mypersonal possession, and am preparing an edition of it.

Directly or indirectly based on these and other Isma'ilihistorical books are two modern works by Isrna'Ili writers,both in the Arabic language, but one written in India andthe other in Syria. The first of these is the Riyarj al-Jintinof Sharaf 'All Sidhpuri, printed in Bombay 1277/1860.This work is addressed to the general Islamic public, andthe author makes no reference to his being an Isma'Ili,The work, however, is strongly Isma'Ili in tendency, andthe historic chapters arc based on Isma'Ili sources. Thereare some interesting passages, though the exoteric characterof the work prevents any serious revelations from being made.

The other is al-Falak ad-Daunodr fi Samd al-A'immaal-Athdr, a general survey of Isma'ilism published in Aleppoin 1356/1933 by Shaikh 'Abdallah b. Murtada, an Ismailifunctionary in Khawabi, The author has apparently hadaccess to Bohra books, but the historical part is basedmainly on Sunni sources and is not of great value.

Apart from the professedly historical works, there arealso occasional histoncal notes and references in dogmaticand theological works. These, where they occur, arcprobably more reliable than the directly historical writings,because whereas the historical book" are considered to be?ahir, or exoteric, the secret books on Haqd'iq and "Aqa'id.are often esoteric, reserved for the chosen few, and thusmay contain information withheld from the masses. Astriking example of this is the short historical note in theGhayat al-Mawalid, a secret work on doctrine, in which it isadmitted that 'Ubaidallah al-Mahdi was not an 'Alid.

1Ibui , 40.


24/121

INTRODUCTION 17(See below.) This is, of course, strongly denied in thehistorical writings of the Qac. l i Nu'rnan (it is even possiblethat the Q M i , who was never initiated into the highestranks of the da'tna, was himself unaware of this), in the'Uyun al-Akhbar, and in modern Isrna'Ili writings.

Among the Isma'ili sources we must include two travellersboth strongly pro-Fatimid, and the latter at least a pro-foundly convinced Isma'Ili, In the travel-records of IbnHawqal (late fourth Itenth century) and Nasir i Khusraw(d. 481/r088), we have interesting first-hand descriptionsby unprejudiced observers of the Carmathian state inBahrain and its internal organisation.

A fair amount of information may be gathered also fromthe books of groups which, if not actually Ismaili, areclosely related to Isma'Ilism. The most notable of these isthe Rasd'tl lkhwan as-Safd, a store of information ondoctrinal and philosophic matters, containing quite a fewhints of histone value. More immediately useful are theDruze scriptures, representing as they do an early andrelatively pure Isma'ill current, and, owing to their specialcharacter as the scriptures of a secret revolutionary sect,less reserved than the official Fatirnid literature.Such, then, are the sources that tell the story of theBatini movement. It IS a story bristling with inner con-tradictions and inconsistencies. full of unanswered questions.The task before us is to attempt, by means of a criticalevaluation and confrontation of these sources, to give aclear and connected account of the events they describe.

Before proceeding to an enumeration of the problemsinvolved and an attempt at their solution, it is perhapsuseful to mention briefly the main difficulties that block ourpath. The first of these, already referred to, i" the ignoranceand sometimes deliberate bad faith of the anti-Batiniwriters who form one of our main sources of information.The second is the secret and esoteric character of the move-ment itself, concealing its doctrines and its personalities,not only from its opponents, but even from a large part ofits own adherents-those who are not initiated into theinner mysteries. Many of the Biitini works thus deliberately


25/121

18 THE ORIGINS OF ISMA'ILISMconceal and even falsify certain vital facts, if they areintended for general consumption. A third is the factthat owing to the conditions of difficulty and persecutionunder which the movement lived, a large number of its mostimportant leaders were constantly in hiding, and appearedin different places under a variety of aliases, which makestheir very identity uncertain. Finally, there is the curiousdoctrine of tafwi4, or spmtual adoption, according to whichwords like Abit and Ibn (father and son) may connote arelation of teacher and pupil rather than physical fatherand son. This, as will readily be seen, renders most Batinigenealogies suspect, and increases the difficulty of estab-lishing the true identity of various persons.Our problem is largely one of classification and identifi-cation. During the centuries of its actrvrty, the Batinimovement, owing to the wide dispersion and secret characterof its missions, was split into a large number of fairlydistinct groups, With different names and local traditions.Some of these form part of the central core of orgarusation->others are offshoots, dependent on the main centres. but,owing to the distance and the difficulty of communications,diverging in character under local influences Others, too,are sects, ongmally unconnected with the B~ltini movement,which have, however, thrown their lot in WIth it and acceptedits central authority. This authority, however, i~ oftenmerely nominal, and the groups go on very much in theirold way. Yet another group is composed of sects whichhave little or no actual connection WIth the Batiniya, andwhich have quite a separate origin and character. Thoughthere may occasionally have been a limited kind of co-operation, the connection of these with the Batiniya i-;largely a figment of anti-Batini imagination. In this groupwe can almost certainly plan' such sects as the Khurramiyaand other stnctly Iranian heresies.

This regional and doctrinal diversity has given us a hostof names, which are applied by most authorities indis-criminately to one sect or another, thus causing great con-fusion. Long lists of such names, with their regionalapplications, will be found in the Siyasat Name and Ibn


26/121

INTRODUCTION 19al-Jawzi. In order to avoid confusion. the word Batiniis here used to designate the whole group of movementsunder consideration. as being the widest and most generallyaccepted. Other terms used will be strictly defined.In the welter of sects and sub-sects revealed to us byour sources. it is possible to distinguish four chief groups.each of which forms a distinct unity. with a more or lesshomogeneous character and a separate tradition and history.to be found in most or all source". It is the relation betweenthese movements that constitutes t11(' crux of the problem.They are as follows:-

(I) The early Isma'Hi da'iea=-i.c, the movement which.in the second century A.H . centred on Isma'Il h. Ja'far. hisson Muhammad and the coterie Immediately surroundingthem. This is the group revealed to us by the Twelversu n sources.

(2) The da'tea which, beginning in the Yemen under thefamous Mansur, and continumg in N. Africa under Abu'Abdallah a~-Shi'i. culminated IIIthe establishment of theFfitirnid Caliphate.

(3) The Syro-Mesopotamian, movement sometimes calledCarrnathian. which, under the leadership of Zikrawaih b.Mihrawaih and his son", ravaged Syna from 289-294 A.H.(90I-906).(4) The Qaramita, or Carrnathians properly so-called,i.e. the movement IIIBahrain, under Abu Said, Abu Tahirand their successors.

These four group" appear ITl all our sources. Whereauthorities, both ancient and modern, differ is in the deter-mining of the relationships between them, and the allocationto one or another of prominent personalities such as 'Abdallahb. Maimiin. Upon the correct answer to these questionsdepends also the vexed problem of Fatimid legitimacy.In the following pages the four groups will be referred torespectively as Isma'Ili, Fatirnid, Syro-Mesopotamian, andCarmathian.Before stating our own answer to the question, it isperhaps useful to review briefly the answers given by others.The Arabic sources we have' already seen. For Tabari the


27/121

20 THE ORIGINS OF ISMA'ILISMFatirnids are what they claim to be, i.e. by implication,the successors of group 1. Groups 3 and 4 are one and thesame-quite unconnected with I and 2. Thabit suspectsa connection, but follows Tabari on the whole. Ibn Rizam,and all who follow him, identify the Fatirnids with theCarmathians, place 'Abdallah b. Maimiin among the latter,and make him the true ancestor of the Fatirnid Caliphs.They thus deny the Fatirnids or the Carmathians anyconnection with group I, or at best regard them as usurpersof Isma'Il's family claims. Most argument since has beenalong these lines, and those who accept the legitimacy ofthe Fatimids usually do so by denying them any con-nection with the Carmathians or WIth 'Abdallah b. Maimiin.

Among European scholars opinion has been divided."Earlier writers-De Sacy, Dozy, Hammer, Quatremere,Guyard, Blochet and De Goeje-all tended to base them-selves mainly on the Ibn Rizarn version in its broad lines.Isma'llis, Fatimids and Carrnathians were all differentnames for one single movement. Itbegan with the creationof a separate group supporting the claims of the ImamIsma'Il as against hIS brothers, but was completely re-organised and given its specific doctrinal and organisationalcharacter by the evil genius of 'Abdallah b. Maimiin al-Qaddal;, an anti-Islamic infidel who flourished some timeduring the third century A.H. Thus far there is unanimity.Sacy and Blochct are inchned to allow despite all this thatthe Fatirnid Caliphs were really 'Ahds-a claim which isrejected by Dozy, Quatremerl', Guyard and De Coeje.

The appearance of new source materials has caused acertain change of front among Orientalists, The firstserious divergence is to be found in Casanova. He, whilerecognising the Identity of the Isrna'Ili and the Fatirnidmovements, regards the Carmathians as a separate groupof earlier origin, which was later won over to Isma'Ilisrn.He places the life and activities of 'Abdallah b. Maimimal-Qaddah in the second century, A.H., and credits him withthe creation of the united Batini movement by weldingtogether the two great wings of the 'Alid movement, the

1For these general references see the Bibliography


28/121

INTRODUCTION 2IFatimiya, or followers of the descendants of Fatima, andthe Hanafiya, or followers of the line of Muhammad b.al-Hanafiya. The Carmathians stricto sensu belonged tothe latter group.

Ivanow, following the modern Isma'ili apologists, rejectsall contact between the Fatimid-Isma'ilis on the one handand the Carmathians on the other. The latter, he asserts,were as hateful to the Fatimids as to the orthodox, andthere was considerable doctrinal difference. The names of'Abdallah b. Maimim, Dindan, etc., are unknown inIsma'ili literature, thus showing thai they can have had noconnection with early Isma'Ilism. Ivanow has no hesitationin accepting the legitimacy of the Fatimid Caliphs.

For Massignon, Ismailis, Fatimids and Carmathians allform one and the same movement, which was founded in thesecond/eighth century by the Imam Isma'Il, his teacher andguide Abii'l-Khattab, Maimiin al-Oaddah and other con-temporanes. The Fatirrud Caliphs were descended tromMaimiin al-Qaddah. Massignon, however, stresses the im-portance of the Isma'Ili doctrine of spiritual adoption(v. infra), according to which the pupil, not the physicalson, is the true heir. The Fatimid Caliphs were thus'Alids in a spiritual sense, 'Abdallah b. Maimiin havingbeen adopted a" son and heir by Muhammad b. Isrna'Ilthe Imam.

For Guidi, the first impulse came from Abii'l-Khattab,a follower of Ja' far and Isma'Il, who was strongly influencedby the Khurr amdinan and other Iranian heresies, Itwashe who was the true founder of Ismailism. One of hisfollowers, 'Abdallah b. Maimfin, founded the' Carrnathiansect. The' Fatirnid Caliphs were descendants of 'Abdallah,and their da'taa a part of the Carmathian movement. Oncethe dynasty was established, however, the Carmath iandoctrines were moderated to some extent, and conflictensued with the purists 111 the East.

One other view which may briefly be mentioned here isthat of Prince Mamour. He admits the descent of theFatimid Caliphs from Maimiin al-Qaddah, but claims that thelatter himself was an 'Alid, none other than Muhammad b.


29/121

22 THE ORIGINS OF ISMA.'ILISMIsma'il. The Carmathians were a heretical offshoot of theIsma'ill da'wa. It has already been pointed out by Prof.H. A. R. Gibb! that by making this the last defence of Fatimidlegitimacy Mamour has put his case in a very weak position.Iremains then to sketch my own answer to the problemsinvolved. The following propositions, it will be seen, owemuch to the work of Massignon, Casanova, Ivanow,Hamdani and others, and also to a number of new sources.The following chapters of this study constitute an attemptto prove them,(I) The Isma'ili movement began in the personalcoterie of Ismail b. Ja'far. with the active co-operation ofIsrna'Il himself and his son Muhammad. Among the com-panions of Isma'Il and Muhammad, and the first organisersof the sect, were Abii'l-Khattab, Mairniln al-Qaddah and thelatter's son 'Abdallah.

( 2 ) The Fatirnid movement in Yemen and N. Africawas the direct continuation of that organised by Isma'il,Abii'l-Khattab and Maimiin al-Qaddah. 'Ubaidallah al-Mahdiwas the direct descendant and successor in leadership ofMaimiin. The Fatimid Caliphs were nevertheless genuine'Alids, the first of them being al-Qa'im, descendant of thehidden Imams for whom 'Ubaidallah and his Qaddal:lidprogenitors were working.

(3) It does not necessarily follow from this thatCarmathian = = Isma'ili. 'Abdallah b. Maimiin was oneof the founders of Ismailism. Itis his connection with theCarmathians which remains to be proved.(4) The Syro-Mesopotamian Carmathians were part ofthe Isrna'ili da'wa.(5) The Carmathian movement in Bahrain was separatein origin-possibly Hanafi, or dissident Isma'ili-but lateradhered as a group to the Fatimid Caliphs. while retainingits separate identity. The name Carmathian probablydates from after their conversion to Isrna'Ilism.(6) The later conflict between the Carmathians and theFatimids was due to a split between moderates and ex-tremists after the establishment of the Fatimid state.

IB S o.s. . VII, 984.


30/121

Chapter IISMA.'IU ORIGINS

(a) Int rodu ct ory .Before examining in detail the ongm'> of the Isma'Ili

movement, it is nece


31/121

2 4 THE ORIGINS OF ISMA'ILISMmll'll}(ili, the non-Arab converts to Islam, provided a fertilerecruiting ground for any revolutionary movement, and,once it had turned decisively towards them, the Shi'a sectsoon had a large mawla following in many parts of the empire.But the entry of large numbers of superficially IslamisedPersians, Aramaeans, Syrians and others necessarily workeda fundamental change in Shi'ism as a doctrine and as apurpose. Soon a welter of strange beliefs, brought overfrom Christian, Iranian and Old Babylonian heresies, foundtheir way into Shi'I theology. The movement came to bedominated by the mawali and other oppressed classes, andbecame the instrument of their social and religious revoltagainst the oppression of the orthodox state.Later, however, as the distinction between Arab andmawla came to correspond less and less with the economicdistinction between privileged and unprivileged, the revo-lutionary Shi'a ceased to represent the matodl; as such, andbecame the mouthpiece of all the oppressed classes, Theorthodox Persian Zoroastrians of the upper classes becameSunni and retained their privileges.' The impoverishedArabs of 'Iraq, Syria and Bahrain came under the influenceof extreme Shi'i ideas.

One of the characteristic marks of the change fromArab to mawlti Shi'ism is the appearanre of the Mahdi idea.From being a political candidate for power, the Shi'I Imambecame a mysterious figure of great religious importance, atfirst a Messiah and later an avatar of divinity. This intro-duction of eschatological ideas has been ascribed to varioussources. Darrncstetcr- attributes a Persian origin to thedoctrine. It was, he says, imported into Islam by themasses of imperfectly Islamised Persian converts, whobrought with them the Indo-Aryan idea of a chosen, God-begotten family, transmitting the Glory of God (FarriYa;:;dan) from generation to generation, and eventuallyproducing a Saomytint or Messiah. This conception wastransferred to the family of the prophet and the personalityof 'Ali. Snouck" prefers a Christian origin, based on the

lSad'glu, GR., P 613 Verspr gesch , I, 152 'Le Mahdi, p. 15 fl.


32/121

ISMA.'ILI ORIGINS 2 5idea of the return of Jesus and the end of the world. Guidi!follows Darmesteter, and goes still further in ascribing thegrowth of extreme ideas to a deliberate Iranian dualistpropaganda. Massignon- finally regards Mahdism as anautochthonous Muslim development, growing out of theQur'an, the Muslim tradition and Arab folklore, and stimu-lated by social conditions.Many of the Muslim historians ascribe the beginningsof revolutionary Shl'isrn to one 'Abdallah b. Saba, aYemenite Jewish contemporary of 'Ali, who preached thedivinity of 'Ali during the latter's lifetime and was burnedfor his pains. The beginnings of extreme Shi'isrn are thusascribed to him and through him to a Jewish origin. aModern research, however, has shown that this is ananticipation, a projection into the past by second-centurytraditionists of the conditions and ideas of their own day.Wellhauscn+ and Friedlander+ have shown by a critique ofsources that the conspiracy and mission ascribed to Ibn Sabaare a later invention. Caetani." too, in a well-argued chapter,has shown that a conspiracy with the ideology and organisa-tion ascnbcd to Ibn Saba was inconceivable in the tribal,patriarchal Arab world of 35 A.H., and obviously reflectsthe conditions of early 'Abbasid times, It required themurder of 'Ali, the dramatic martyrdom of Husain and hisfollowers at Karbala, and a considerable social changebefore Shi'ism of the revolutionary, messianic type couldappear.The word Mahdi may have been used in a purely honorificand political sense to describe 'Ali, Hasan and Husain,"but Its first indisputable use in a messianic context is in therevolt of Mukhtar in 66/685, who preached that Muhammadb. al-Hanafiya, a son of 'Ali by a Hanafl woman, wasMessiah. The revolt began in Kiifa-a city well fitted bycircumstances to be the starting point and centre of such a

1 R SO, 4'Lecture'> at Colh"ge de France, 193b- 37J Ka>h~1. 7()-71 Van Vloten was rruxlcd by his sources Into ascnbmg

an Important role to the Saba'Iya Stns, VI, 124-5 and 133 Rei Pol, 'JI Z.A., XXIII, 29(;, and XXIV, 1. Anna!!, VIII. 3l>fI7 Snouck, ISO Macdonald, EI, Art Mohd.


33/121

THE ORIGINS OF ISMA'ILISMmovement. A new and growing town, with a populationcomposed of men of innumerable creeds and races, allturbulent and discontented, hating the government, thereligion it represented, and the oppressing class whichmaintained it-such was Kiifa when Mukhtar began hisrevolt. It was an admirable breeding ground for syncretist,messianic movements of social revolt, and Mukhtar, byappealing to the ma1't i iT for support, began a movementof far-reaching sigmficance. As there was no 'AJiJ of theFatimid hue of suitable age available, Mukhtiir selected ashis Irnam-Mahdi Muhammad b. al-Hanafiya, whom hemade supreme representative on earth of the divine wisdom,divinely ordained succes-or of 'Ali and the Prophet. Dcsprtcthe suppression of ;\lukht~lr and the death of Muhammad,the movement spread rapidly, and many of its adherentspreached that Muhamrn,! was not rcally dead, but was IIIhiding, and would eventually return and "jill the worldwith justice and equity as It was full of injustice and op-pression."! Thus appear for the first tune th lIc;,;'Ianicdoctrines of Ghaiba (absence, concealment) and Raja(return), which are characteristic of almost all later Shi'isects.For the 70 odd years intcrvcmru; between the revolt ofMukhtar and the lwgiIllling,., of hm:-l'ih,.,m properly so-called,there are two main trend- among the revolutionary Slli'isects. These may conveniently be termed the Ffltimiyaand the Hanafiya (not to b,' confused with th late-r Iq,!,alschool of that nam). Thr- first of these were- the Iollow-r-,of the various Imam-, through the line of 'Ali by Fatima,i.e., Hasan, Husain and therr offspring-e-thc second werethe follower" of Muhammad b. al-Hanafiya and his descen-dants. These distincttons were never very firm, and groupsof followers seem to have flitted casrly from (JIll' -ido to theother. In the ear licr period the Ffitrmiya seem to haverepresented the conservative wing, and were alway" throwingoff small group" which graVitated towards the more extremeHanafiya. When the latter disappeared, the Fatimiya,after absorbing the Hanafi remnants, became themselves

IA mcsviaruc traditron In'qucnt 11 1 Shi'i works.


34/121

ISMA'ILI ORIGINS 27the more extreme section, and carried on the old rivalrywith the new moderate group of the Twelvers.After the accession of the 'Abbasids, the Hanaii grouplost its raison d' etre, and it was left to Isma'Il, the seventhFatimid Imam, and his associates, to unite and consolidatethe different, warring factions into one great movement,professing allegiance to one Imam and one doctrine.This survey, it should be noted, deals only with therevolutionary Shi'a, thc Ghuldt, and docs not cover thosewho, like the Zaidiya, continued, albeit in a much modifiedform, the traditions of early, political Shi'isrn. The latermoderate, or Twclver Shi'a are also excluded.

During the formative period mentioned, the sects aretoo numerous to be counted. Again and again variouspretender" of 'Ahd descent ri"e in revolt, and, on theirfailure, pas,> into mythology. Again and agam does sometoo zcalou-. mis-aonary forget hi" 'Ahd master and begin topreach on his own account. The story IS further com-plicated by a number of Iranian, Chrrstian and Jewishscct-, quite outside- Islam, which ncvor tlx-less "hare certaindoctrines WIth the Shi'i Chulat, and arc often confused withthem by Sunni historians. Particularly important amongthcs arc the Khurramdinan, the Mazdakis and the nco-Jewish 'Lsawiya.Let us begin With a brief ;,urV('y of the Hanafi line,which is of some interest as


35/121

THE ORIGINS OF ISMA'ILISMof this sect were Sa'id and Bayan,! the latter of whom latercreated a sect around himself.(2) Those! who believed that Muhammad b. al-Hanafiyawas hidden in Mount Radwa, in the Hijaz, and wouldreturn to establish justice. Among them was the poetas-Sayyid al-Himyari, who later turned to Ja'far as-Sadiq.

(3 ) Haw.irniya3-those who said that the Imamate hadpassed to Abu Haw.irn, the son of Muhammad b. al-Hanafiya.Some said that he was Mahdi, On his death in 98/710 hisfollowers split into four groups, viz.:

(a) He was succeeded by his brother 'Ali b. Muhammad,who in turn bequeathed hIS title to the 'Abbasidfamily.

( b ) He was succeeded by 'Abdallah b. Mu'awiya, adescendant of Abu Tahb, the father of 'Ali. Thisgroup were known as J:ffmfuiya, after 'Abdallah b.al-Hantjj, a hervnc of Madfl'm who at first founded asect of Ius own but later joined 'Abdallah b. Mu'awiya.They would "l'l'1l1 to be the "arne as the Harbiya andthe Janahiya 4

( c ) He was succeeded by Muhammad b 'Ali, of the houseof 'Abbas, to whom he bequeathed the Imamate inSyria. Thc-, are the Rawandiya, and art' intimatelyconnected with the r iso of the 'Abbasid dynasty."

(d ) He was Qfl'lm and Mahdi, and would "return."Bayan (v. su pra) joined these, and later claimed tobe the successor of Abu Ha~im. He was arrestedand crucified m 1H)!7.J7.8All these sects, according to our


36/121

ISMA'ILI ORIGINS 2 9functions to the Imam, and some of them, e.g. the Karbiya,going so far as to divinise him. To many lba/;/i, i.e. libertinetheories, are attributed-e.g. Hamza b. 'Umara is said tohave married his own daughter and to have taught that allforbidden things were permitted, the sole religious dutybeing knowledge of the Imam.' Ghaiba and Raj'a doctrinesplay an important part. Of interest is the idea that theImamate can be bequeathed. Thus, Abu Ha~im be-queathed his rights to the 'Abbasids, and Bayan claimed tohave been appointed successor by the Imam. This ideareceived a great development in the Isma'Ili system.

These, then, are the main sccts of the Hanafl line. Onthe accession to power of thc 'Abbasids, whose movementwas itself an offshoot of Hanafi Shi'i~m, the sect, after afew last nervous twitches, disappeared. Those who re-mained active were absorbed by the Fatimid da'iua.Before passing on to consider the Fatirnid Husainidline, we may note briefly the sects attached to the descen-dants of a third son of 'All, namely Al-Hasan. The linecomes into prominence with Muhammad b. 'Abdallah b.al-Hasan b. al-Hasan b. 'Ali (100-145/718-762), known asAn-Nafs az-Zakiya, the Pure-Soul. Muhammad An-Nafsaz-Zakiya, already at the age of 19, claimed to be Mahdi,His claims were advanced by Mughira b. Sa'id al-Tjli,founder of the :\lughiriya sect. These preached thatMuhammad An-Nat= az-Zakiya was Mahdi and Mu~ira aprophet. On the death of Muhammad they claimed thathe was in Ghaiba and would return. MU~lra himself waskilled in IIq/737 by Kh.lhd al-Oasri, whose freedman hewas. This sect seems to have had great influence in thedevelopment of Batini ideas, and is sornetimo- listed as asub-sect of the hlmttabiya (v. infra). Itshould be noted thatMuhammad an-Nafs az-Zakiya himself was not disposed torecognise Mu~ira.2

After the tragedy of Karbala, the Fatimid line was intemporary eclipse, owing to the absence of adult pretenders.J Naw, 26 Naw, 37. 5l. 54-5. 57 ~J.hr, IJ-l. Tr. ~o3 B.4;h. 02'1, Tr 4Y

Malatt, 122. Acl.!. I. 6 Ihn I:J Frted , 1. 5


37/121

30 THE ORIGINS OF IsMA'ILISMThe first serious movement is that of Zaid, son of 'Ali Zainal-'A.bidin, and his two sons Yahya and 'Isa. Zaid gavehis name to the Zaidiya wing of Shi'ism, which until to-daycarries on the tradition of Arab, political Shi'ism, differingvery little from the Sunnis as regards religion.

During the life of Muhammad al-Baqir, the fifth FatimidImam, the number of followers of the Fatimid line increasedgreatly, and at least one important sub-sect, the Mansiiriya,was given off. These were the followers of Abu Mansiiral-Tjli, who preached the Imamate of Muhammad al-Baqirand then of himself.' This group is believed to havepreached philosophic materialism, and Iibertinism (exceptby Nawbakhti), and to have advocated the use of terroristicmethods in dealing with opponents. In several importantaspects this movement anticipated Isma'Ili developments.It is listed by Ibn Hazrn as a sub-sect of the Khattabiya.Abu Mansur him-elf was killed in about I25/742 by orderof Yiisuf b. 'Umar aili-Thaqafi, governor of 'Iraq, and severerepressive measures taken agall1st the ~L'Ct.2On the death of Muhammad al-Bfiqir, some of hisfollowers seceded and supported the claims of Muhammadan-Nals az-Zakiya. The;-.('were de-nied by ja'Iar as-Sadiq,the sixth Fatimid Imam, who, WIth Iu-, followers, refused toaid the pretender." The rest supported the Imamate ofja'far. and continued to obey him. Promine-nt among thesupporters of .la'far is one Abu'l-Khatp.h, who founded asect of Ill;' own wluch in turn split into se-veral sub-sects.On the death of ja'Iar 11 1 l\Iadill


38/121

ISMA'ILJ ORIGINS 31through him the Twelver, or moderate Shi'a group.Their further sub-divisions fall beyond the scope ofour study.

(3) Isrna'Iliya-e-those who supported the claims ofIsma'Il, eldest son of la'far, and, after his death, ofhis son Muhammad b. Isma'Il,The first of these groups soon disappeared. The secondgrew into the Twelver, moderate Shi'a, The third, which

resumed within itself the main tradition of revolutionaryShi'ism, is the subject of our further studies.'(b) The Birth of I smd'iltsm,

In the second quarter of the second century of theMuslim era, conditions were npe for a reorganisation and re-orientation of the revolutionary movements. The accessionto power of the 'Abbasid dynasty ill 1}2/750 had marked-and caused=-scvcral ~ignificant changes. In the first place,as already noted, the success of the Ha~imiya faction, ofwhich the 'Abbfixid-, weft' the rcpresentat ivcs, and theconsequent officiahsation of the ~ect, finally ended the lineof Hanafi pntcndcrs, and left the fidd clear for the Fatimidline of which jatar wa-, the chid figure. The-n again. the'Abbasid n-volution marked a new stage in the social andeconomic history ( I f LIam. Tlu- assimilation of the non-Arab rullllg Cla""l'~ to the Arab, SUI1Ilistate, and the growing

1 In the .ihovc, I h av e not .h-alt w rtb uon-Jsl.inuc sects. Tw o of these,one of t liern prc-I sl.umc. howe-ver. ...hould 1)(' noted, 111 \-IC\" of the ~lg-mtic.urt ". .ly 1[1 w hu.h they ant u.ip.itcd and !'I'[haps mthu-no-d l sm.iIlldcv cloprru-n t-,lhc nrst of the-,e-. the -:\Lu'd,lkI~a, "ao;; all Ir.uu.m lu-revy w lnchpreached .l h.u ld of u-Iunou-, t'llllllllllIlbn Tho-,e w rrter s w ho at tribu tecornmurnst docinn('-.. .md Pt.!'1 til c- to the I..,nl,:i'ilis u~li.llly a t trrbu tethem to :\Lu:d.i1 ... Ul~ln ... \"'l'L' (~ 11 S,ldlI.4111. Lc; Jlozwt'Jld,'I.;s. etc) Thesecond I....lu- tu-o-jvw r-h ...e


39/121

32 THE ORIGINS OF ISMA'ILISMidentity of Arab and mawla subject classes, resulted in anew alignment of classes, on economic rather than, as in thefirst century, on racial lines. This process was accentuatedby the gradual transformation of the caliphate from anagrarian, military state to a commercial, cosmopolitanempire. This change was taking place throughout thesecond century, and by the third was wry far advanced.So tar-reaching a change m social conditions and alignmentsnaturally resulted in a reorganisation and expansion ofthose movements which expressed thl' revolt of the oppressedclasses and peoples. The crisis conditions of the third andfourth centuries produced a series of revolutionary out-breaks. The second century i~ what one might term aperiod of revolutionary incubation.A . b 1 t : i-Khat/lib.

The majonty of our sources seem to point to Abii'i-Khattab as having been the first to organise a movement ofspecifically Batin! type. He i~ known as Muhammad b.Abi Zamab, also as ~l!qla~ h. Abi'l-Khattab, and was amawlti of Asad.' He was a close a--sociat of the ImamsMuhammad al-Baqir and] a'Iar as-Sadiq, and was apparentlyone of their most intimate disciples until his rejection by thelatter. 2 The disavowal of Abu'I Khattab by ]a'far seemsto have caused great consternation among the Slli'a, andmany pages of Twelver Shi'I works are devoted to explana-tions of it.3The fullest and most reliable account of the activitiesof Abu'l-hnattab is that given by Nawbakhti.s Accordingto Nawbakhti, Abu'l-hllaHab began as a dti'i of Baqir andla'far, and as such made a number of extravagant claims-e.g. that he was a prophet, that he was herr to ]a'far, etc.He also taught a number of new, heretical doctrines. includ-ing Ibdha (libertinism) and taqiya" or secrecy, the right tolie and bear false witness in the interests of religion. Heand his followers taught also that the Heaven and Hell1Ka~l. p. 187. Ibn J : I . Fned, I, 6


40/121

ISMA'ILl ORIGINS 33mentioned in the Qur'an were merely persons, and had nosuper-terrestrial meaning. They introduced also the oldoriental "Light" theory of successive incarnations.' Thisis given in detail by Nawbakhti, and corresponds closelywith later Isma'Ili writings. A number of them, under theleadership of Abu'I-Khattiib himself, concentrated in Kiifa,and seventy of them were killed 10 the mosque there byorder of the governor, 'hii b. Musil. Abu'I-KhaWi.b wascaptured and crucified (138/755).2 Before that Ja'far hadalready publicly denied Abu'I-h.l1attiib, and dissociatedhimself from all his activities,

On the death of Abu'l-Khattiib, his followers transferredtheir allegiance to Muhammad b. Isma'Il, the grandson ofju'Iar. TIl(' Isma'iliya sect was actually Identical with theKhattfibiya.3

Such, then, are th main lmcs of Nawbakhti's account ofthe hnattabiya. A~ regard" doctrine it bcorroborated andamplified by most of the Sunni hcresiographers, Baghdadi!and Shahrastfini" give more details than Nawbakhti asto the sub-v-et- of the hllattiihiya movement after thedeath of Abu'l-Khattrlb. The? were (I) Mu'ammariya, (2 )Bazigljiya, (_:;)'Umairiya , (4) Muladdaliya. They differedon minor points of doct rme and on the personality of theleader.

Although most of lhe Sunni here-rographers do notmention any direct contact between the Khattabiya and theIsrna'Iliya. yl't the .iccount-, they give of KhaHabi doctrinesare such a" to corroborate Nawbakht i's statement of theidentitv of the two movements. Thus Baghdiidi8 andA2!'ari'7 attribute to the hllattflbiya the characteristicallyIsrna'Ili doctnne of "Silent and Speaking Imams" (~amitand Natiq).8 This j" supported by Maqrizl." Again, theIsrna'ili method of Ta'ioil, or allegoric interpretation of thescriptures is attributed to the Khattibis by Ibn I:Iazm,lO1AI,o InSlJahr, '3(), Tr 206'TillS I, the ddt" ~IH'n by K:J!iliI. 1


41/121

34 THE ORIGINS OF ISMA'ILISMShahrastani,l Maqrizi! and others, as well as by Nawbakhti.So, too, with the "Light" theory.Further information on KhaHabi doctrines is to be foundin the Twelver Shi'i works. Kashshi" relates a number oftraditions, telling of the claims of Abii'I-KhaHiib and thereasons for ja'far's denial of him. A few examples maybe noted.

la'far wrote to Abii'I-KhatFib, saying: "It has reachedme that you believe that adultery is a man, wine IS a man,prayer is a man, fasting is a man and sin is a man."'Anbasa b. Mus'ab- said: "ja'Iar asked me, 'What didyou hear from Abii'l-Khattab ? " I answered: 'I heard himsay that you (]a'far) had plaC(d your hand on hie; breastand said to hirn : 'Remem bcr and do not forget! You knowthat which is hidden,' and that you told him he was 'thehidden thing of our knowledge, tilt' place of our secret,responsible for our hying and our dead.''' All this J a'Iardenied ernphaticallv.sOther traditions describe how Abu.'l-KhatFlo preachedthat la'far was divine and pO""t'~sed occult powt'r~, how hetaught reincarnation ( taI1l1sukh) and ddled the expresscommands of .la'far.

Historical, as dr-tinct from cloctnnal evidence, of theconnection be-tween tlu- l\"hatt~lbl~'a and t lu- hlll~l'mya isprovided by a number of sourc--c. Ibn l{iZ~1I117,ays thatthe Mairniiniya, or follower" of Mal l11Un al-QadddJ,l, weredisciples of Abii'l-hl1att:ib. Ibn al-Alliir8 mentions Ahu'l-Khattilb as the f irvt of the se-ct (miwal man [a'ala dlllihka),anel Maimiin al-Oaddah as his follower. Nuwairi, on theauthority of Ibn 5'hadd:ld,' al-.o de-cribes :\lalIl1llI1 a~ afollower of Abu'l-hnaWi.b, and ascribe- to the Khattflbi~the doctrine, of Tuu.il, Bdiin , etc. The moveme-nt spreadby Maimirn and lus son 'Abdallah was substantially that of

"l6,Tr 206 2 Khllat, II, 352 'IR7-I


42/121

ISMA'ILI ORIGINS 3 5Abii'l-Khattab. Rashid ad-Dint also mentions Abii'l-KhaWib as founder, and considers Maimiin and 'Abdallahas his disciples. Finally we may note a reference in aZaidi work, the Kztlib Haqa'iq at-Ma'rifa, of the Imamal-Mutawakkil, Zaidi Rassi Imam of Yemen, 532-566/II37-1170. H(' speaks of the "Isma'iliya, ieho are theMubarakiya and the KhaWihiya."2

At this point we may well ask ourselve-s what referencesto Abii'l-h'haqfib and the J{hattabiya arc to be found inIsma'ili writing". He-re we arc at a rli-advantage. Almostall the bma'ili documents at our disposal belong to thelater, officiahvd period of the da'sca, when the apologistswere intercst.! 1 1 1 (It'nying all connection with such dIS-reputable pl'r;,ol1" _i'


43/121

THE ORIGINS OF ISMA'ILISMhim on a par with Salman. The statement is clear andunequivocal:- "The Isma'Ili religion is that which wasfounded by the descendants (disciples?) of Abii'I-KhaHab,who gave their lives in love of the descendants of Ja'faras-Sadiq (and) Isma'Il.!

Similar statements and doctrines are to be found in theNusairi writings studied by Massignon. These, too, makeAbii'l-KhaHab a founder of the sect and Maimiin al-Qaddah adisciple of his. They attribute to him most of the character-istically Isrna'ili doctrines."

The denial of Abii'l-Khattab by Ja'far, and the violentdeath of the former, seem to have caused much perturbation.Kashshi and the other Twelver Shi'I compilers relate numeroustraditions explaining and justifying this denial. Thefollowing are particularly noteworthy.

'Abdallah ar-Rijani said: "I spoke to Ja'far of Abii'l-Khattab and his death. While doing so I was touched withcompassion and wcpt. Ja'far asked: 'Do you mourn him?'I answered: 'No. But I have heard it "aid that after 'Alihad killed the men of the river." his companions began toweep. He asked them: "Do you mourn them?" They said:"No, but we remember the friendship that we had and thedisaster that has befallen them. Therefore we weep."And he said: "It is not wrong." "4

'Isa Shalqan5 said that he had asked Abii'l-Hasan, whenthe latter was still a boy, why his father Ja'far had orderedthem first to be friendly to Abii'l-Khattab and then torenounce him. The answer, later approved by ja'far, wasas follows: "There are some whom God has created forprophecy, and they cannot be anything but prophets.There are some created for belief, and they cannot beanything but believers. Some God has entrusted with belief.If He wish He may perfect and if He wish He may withdrawit. Abu'l-KhaWi.b was of those to whom God lent belief,

1Text ell, Notes 4zHMassignon, Salman and Art Nusairis in EI. Alii an-Nahr, Probably for Ahl an-Nahrawiin, Ie., the ID.!awanJ. Kashshl, 189 'lsa b. Abl Mansur Shalqan was one of the most trusted followers ofJa'far (K~I, HI)


44/121

ISMA'ILI ORIGINS 37and when he falsified my father's words God withdrewit. "1Q1i.sim as-Sairafi said: "I heard ] a'far say: 'There arepeople who think that I am their Imam. By God, I amnot their Imam, may God curse them. Whenever I havehidden anything they have revealed it. May God revealtheir hidden things. I say thus, and they interpret it thus.I am the Imam of those who obey me.' "2

All the Twelver Shi'i and most of the Sunni sources agreein relating the repudiation of Abii'l-Khattab by ]a'far,3and it would seem that serious doctrinal differences didexist between the two. The subsequent deposition ofIsma'il by ]a'far, the close association between the descen-dants of Isrna'Il and the disciples of Abii'l-Khattab, and thefundamental doctrinal split between the followers of Isma'Iland those of Musa al-Kazirn, the seventh Twelver Imam,would all seem to point in the same direction, and so placeIsma'll himself among the rebels against ja'Iar's authority.'Ism(i'U.

This bnngs us to another, and difficult question. Whatwas the part played by Isma'Il himself in all these activities?Unfortunately, our knowledge about Isrna'Il is extremelylimited. In the Isma'Ili sources, Isma'il IS the Imam, ofsemi-divine status, absolute Lord (az arbdb i i!ldq), and isthus placed on so high a pedestal that any information givenregarding him is of little historical value. Moreover, thereis a considerable amount of contradiction between theIsma'Ili sources themselves, presumably corresponding tothe varying degrees of secrecy of the books in question.Thus, in the Kalarn i Pir" we read: " ... after him (Ja'far)Musa al-Kazim, on the same conditions as Imam Hasan after(Imam) 'Ali,-he had not the privilege of transmitting theI Ka~l. 191 The distmctron between M ustatrda' and "UlIslaqarr

plays an Important part III later hma'lli doctrrnes See Kaliim-i-Pir , 782 Kash,hi. IQ4 Tlus outburst of indignation has a nng of truth.3 Among those who do not IS Ibn Hazrn Fried. I, 6 9 - Among modern scholars who place _\hu!-!.iliattab among the first

founders of the Isrna'Illya, we may mention Massignon (Bib! and EL Art.Karrnatrans and Nusams), Margohouth (EI Art !.iliattiiblya). andFriedlander (Ibn Hazm, II. 106)

Kaliim s Pir, 75 P. 75


45/121

THE ORIGINS OF ISMA'ILISMN4$$, or heritage. After him Imam Isma'il, who wasabsolute Lord. . . . Since Isma'Il knew that the Imamatewas to remain with his descendants, he agreed to the na$~of Miisa al-Kazim, so that they were not opposed to eachother." This story obviously contradicts the numerousother works according to which (I) Isma'll died beforeja'Iar, (2 ) Miisa al-Kazim was not recognised by the Isma'llisas Imam at all.!

The Sunni and Twelver sources regard Isma'il as areprobate, and an unworthy son of his father. The latterin particular seem to adopt an attitude of "the less said thebetter." The following two points constrtutc almost allthe information that can be gathered from them. Uponthese points all arc agreed. They arc (I) Isma'Il diedbefore his father. (2) Isma'Il was deprived of his rightof succession to the Imamate by ja'far, because of his evilhabits.Most authorities confine them-elves to the simplestatement of these facts, juvamj! adds that Isma'Il diedin 145. A Shi'I work" places his death in 138. H.a5.hidad-Din' and juvaini- both ~ay that ja'Iar ordered the bodyof Isma'il to be publicly exposed and his death attested bynumerous witnesscs-s-prcsumably to prevent the spread oflegends. These measures were unsuccessful. Both Ra5.hidad-Din" and juvaini? attribute to the' Isma'ilis the statementthat Isma'Il did not die, that Ill' was alive year" after hisfather's death, and that he performed several miracles.This is confirmed by the pro-Fatirrud Dustier al-Afuna}jimin,8according to which Isrnail was the first hidden Imam. Hisconcealment began in 145, but Ius death did not occur tillseven years later.The general statement is that Isma'il was deposed byJa'far because of his excessive inclination to drink." Thereare, however, in our sources two statements which would

1 Falak , 125 Taqsim (Druze MS), fol 1I7v Fyzee, List, p 8 146 'U'mdat as-Tiiiib, p. 223. Quoted III lvanow, Ismailitica Levi. 521 14() 521 ' 146 De GOCjC, 203 Blochet, 57-8t Levi. 519. juv , 145. Tabart, III. 154 and 2509


46/121

ISM.A'ILI ORIGINS 39seem to imply that there were more serious reasons for thechange.The first of these, the importance of which was firstpointed out by Massignon.! occurs in KasbshP

.. 'Anbasa said: 'I was with Ja'far h. Muhammad at thegate of the Caliph Abu ] a'Iar (= = Mansiir) in Hira whenBassarn" and Isma'il b. ]a'far b. Muhammad were broughtand taken before the Caliph. Bassam was brought outdead. Then Isma'il was brought out. Ja'far raised hishead towards him and said, "You have caused it to be done,o evil-doer. I consign you to Hellfire." , "From this it is clear that Bassarn and Isma'Il wereassociated in some seditious enterprise of which ]a'far dis-approved strongly. This will explain the second passage,in which ]uvami attributes to ]a'far a phrase which, itmust be admitted, is Vt'fy much out of proportion if Isma'Il'ssole offence' was his lack of sobriety. "]a'far said: 'Isma'Ilis not my son, but a devil who has assumed hIS form.' '"Massignon has gone as far as to suggest that the kunya"Abu Isma'Il" which Kashshl- give,; to Abu'l-KhaHab refers

to Isma'Il b. ]a'far, and that Abu'l-Khattab was the adoptive,"spiritual" father of Isma'il."

In his account of Mufaddal b. 'Umar al-]u'fi, Ka~~i7relates a number of traditions which would also seem toimply a close association between Ismiii] and the Khattabis,and express the resentment of Ja'far against those who wereleading his son into heresy and danger. "]a'far said toMufaddal b. 'Urnar: '0 unbeliever. 0 idolater, what is therebetween you and my son?' meaning Isma'tl.':" And later,"What do you want from my son? You want to kill him... "9

1Plan de K i i r " , 350-351. c 15


47/121

THE ORIGINS OF ISM.A.'ILISMIn view of all this there would seem to be a strong

prima facie case for the hypothesis that Isma'il was closelyassociated with the extremist. revolutionary circles thatfounded the sect that bears his name. and that his depositionby Ja'far was due to this association. This impression isconfirmed by the close relation which. as we shall see.existed between Muhammad, the son and heir of Isma'Il,and the extremist disciples of Abu'l-Khatt1i.b, Maimiinal-Qaddah and his son 'Abdallah.

We can now turn to consider the course of events on thedeath of la'far. For it is at this stage that the great splitoccurs between the two wings of the Shi'I movement. Mostof our sources confine themselves to the simple statementthat one group followed Musa al-Kazirn, whereas anothersupported the claims of Ismail and his son Muhammad.For practical purposes we may ignore the minor groups thatfollowed other sons of ja'Iar, as these wen' of no great im-portance, and failed to perpetuate the-mselves.Some of our sources, however, provide us with a moredetailed account, and of these we must make a closerexamination.Nawbakhti- divides those who accepted the claims ofIsma'Il on the death of la'far into two groups. (I) Thosewho denied the death of Ismail during ja'Iar's lifetime.saying that he had been hidden by ja'far. They preachedthat Isma'll was Qti'im. that he was hidden (Ghatba) andwould return (Raj'a). These were the l smd'Iliya properlyso called. the pure Ismailiya iAl-Lsnuiiliya al-Khtili$a).

(2) Those who accepted the Imamate of Muhammad b.Isma'il. According to them Ismail succeeded to theImamate during hi" Iathcr's Iifetime, On Isrna'Il's deaththe Imamate passed to his son Muhammad, The Imamate,they said. cannot fall from brother to brother after the caseof Hasan and Husain. This group was called M'ubdrakiya,after a leader of theirs called Mubarak, a mawlti of Isma'Il,These were joined by the Khattabi". They then split intoseveral sub-groups. One of them, the Carrnathians, beganby preaching the doctrines of the Mubarakis, but then

1P. 57 ft.


48/121

ISMA.'ILI ORIGINS 41diverged. They were named after a Sawadi, Nabataeanleader called Qarmatuya. They said that the spirit passedfrom Ja'far to Abu'l-Khattab and from him to Muhammad b.Isrna'Il and his descendants. Later they developed asystem of their own, according to which Muhammad b.Isma'il was Qti'im, Mahdi, immortal, and the seal of theprophets. Nawbakhti gives some details regarding theirdoctrine, and mentions that in his own day they numberedsome 100,000 souls, and were especially strong in Yemenand in the region of Kiifa.Majlisi,' quoting an early authority, mentions threegroups.(I) Isma'il was al-Qa'im. al-il1unla;;ar. His "death"was a ruse.

(2 ) Isrna'Il died during his father's lifetime, but hadalready appointed his son Muhammad, who succeeded himas Imam. These arc the Carmathians, or Mubarakiya,"Al-Qaramita iea-hum al-Miibarakiya." The name "Car-mathian" derives from a Sawadi called Qarmatiiya, and"Mubarakiya" from Mubarak, a mawld of Ismail. TheCarmathians are the successors of the Mubaraklya, "JVa'[-Qardmita aklzhi/ al-Mubdrahiy ua'l-M'ubdrakiva salafuhum:"

(3) Muhummad was appointed by Ja'far himself.Those thre-e form the I


49/121

42 THE ORIGINS OF ISMA'ILISMSunni sources we can add a little to the above narrative,which is compiled exclusively from Twelver and Zaidis u n sources.

According to the Siydsa: Name,l Mubarak was a l;Iijaziand a servant of Muhammad b. Isma'Il, He was expert incalligraphy of the type called "Muqarmat," thanks to whichhe was given the name Qarmatuya, by which he was known.Mubarak was seduced by 'Abdallah b. Maimiin al-Qaddah,and with him founded and propagated the sect which wasknown as Mubaraki or Carmathian, after the two names ofMubarak.

In view of the earlier and more reliable evidence to thecontrary, this identification of Mubarak and Qarrnatuyamust, I think, be rejected.

Ash'arp Ba~dadi,3 Shahrastani,' and Maqrizi- all men-tion Mubarak, and in the main corroborate our ShI'I sources.

In view of his importance, there is surprisingly littleabout Mubarak in the Isma'Ili sources, The only referenceto him that I have been able to find occurs in the DusUtral-Munajjlmln, which mentions him as a client ofMuhammadb. Isma'il, who recogmscd him as Imam." Itis possible thathe was known in Isma'Ili circles by another name.

We may now attempt a brief recaprtulation of theorigins of the Isrna'Iliya, as revealed by the documents wehave studied.

During the lifetime of ja'far, Abu'l-Khattab and Isma'il,probably in collaboration, elaborated a system of doctrinewhich served as the basis of the Isma'Ili religion of laterdays. They also worked for the creation of a revolutionaryS U i ' ! sect which should gather all the minor Shi'i groupstogether, around the Imamate of Isma'Il and his descendants.On the death of Abu'I-Khattab, Isma'Il and Ja'far, theirorganisation split into several sub-sects, with conflictingideas and leaders. These were rallied around the personof Muhammad b. Isma'Il, who succeeded, with the aid ofvarious supporters, notably Mubarak and 'Abdallah b.

1P 183, Tr. 269 27. 16 and 128, Tr 24 and 193. De Goeje, 203 Blochet, 58

47 Kh,tal, II, 351


50/121

ISMA'ILI ORIGINS 43Maimiin al-Qaddah, in welding into one movement most ofthe followers of Isma'Il, including the greater part of theKhatt1ibiya. whose doctrine was taken over with somemodifications.' Around Muhammad b. Isma'il was createdthe historical Isma'Ili movement."1Questions of doctrinal evolution and change fall beyond the scope ofthe present work A discussion of them will be found InMassignon, Art,Karmatians I Maimiin, 'Abdallah b Marmun, Muhammad b Isma'Il and the

Carmathian sect will all be discussed in later c

Documents

Origins_of_Isma__039_ilism__A_Study_of_the_Historical_Background_of_the_Fatimid_Caliphate