The Silent Qur’an and the Speaking Qur’an: Scriptural Sources of Islam Between History and Fervor

Value judgments aside, the ancient sage thus shares with modern researchers several fundamental ideas regarding Scriptures: namely, that they are bound to evolve as a result of ethnic, geographical, and linguistic factors as well as of historical events, the contingencies of reception, of being committed to written form and their transmission. In short, the very fact that they do have a history. Alterations to the prophetic message, its “falsification,” in Mānī’s terms, are an integral part of this history. And this too forms the theme of the fundamental text of al-Sayyārī that will be scrutinized in this chapter.² Of course, my intention here is not to trace, however briefly, the history of the redaction of the Qur’an but rather, to examine several determining factors in the varying representations that certain Muslim scholars, belonging to all sorts of political and religious tendencies, had of the Qur’anic revelation during the earliest centuries of the Hijra, and to do so through the complex of problems concerning falsification. It will thus be an attempt to set al-Sayyārī’s compilation within the perspective of a wider religious framework of the earliest period of Islam that was the scene of many violent conflicts as well as of numerous discussions and polemics over the Qur’anic text—controversies illustrative of a plethora of viewpoints which later “orthodoxy,” for obvious ideological reasons, sought to obscure. To that end, the presentation is divided into five sections: first, the questionings prompted by historical and critical research into the history of the Qur’an; then the vacillations and contradictions within the Islamic tradition itself and the disparities recorded by the primary sources on this very subject. Third, the attitude of Imami Shi’ites on the question of the Qur’an’s falsification will be examined. Finally, the two last parts will be devoted to al-Sayyārī and his work.³

Scrutiny of the complex of problems regarding the falsification of the revelation given to the Prophet Muḥammad is, to be sure, inseparable from the history and dating of the commitment of the Qur’an to written form. Clearly, the closer the definitive elaboration is to the period of the revelation, the less the risk of alteration. This is the principal reason why the most widely diffused “orthodox” tradition will assert that the decision to compile the Qur’an occurred right after the death of the Prophet in 11/632, in Abū Bakr’s time, and that the official version, utterly faithful to the revelations Muḥammad received, emerged during the caliphate of ʿUthmān, barely thirty years after the death of the messenger of Allāh. We shall return to this. It is fundamental, therefore, to ask when and under what conditions the Qur’an as it is known, and conventionally called the vulgate of ʿUthmān, was compiled. To what extent was it faithful to the “celestial messages” that the Prophet received? How did the first Muslims perceive these problems?

In this respect the manuscript tradition of the Qur’an is of no help to us, at least for the time being. No autograph manuscript exists from the hand of Muḥammad (it is now well known that he was not illiterate) and/or his scribes. The oldest complete versions of the Qur’an date from the third/ninth century. The oldest manuscripts from the pre-Abbasid period are extremely rare and their fragmentary nature makes dating them difficult and open to disputes among specialists. The few excerpts of the Qur’an discovered on papyrus and parchment have been dated by some experts to the end of the first and the beginning of the second century of the Hijra, but these hypotheses have been rejected by others. For more than a century no theory has yet managed to elicit unanimity among scholars.⁵

We are thus compelled to turn our attention to other sources of information, starting with studies in the historical philology of the Qur’anic text, if we wish to find some elements of an answer. In the most general way these may be found in the critical study of the history of writing in Islam. Already in the nineteenth century, in his renowned Life of Mahomet, Alois Sprenger chose a late date for the production of books as such. He made a sharp distinction in fact between “blocs-notes” or “aides-mémoires” which appeared quite early, and books, which he dates from the second–third/eighth–ninth centuries.⁶ This thesis was taken up and brought to completion by Ignaz Goldziher in his magisterial study on the formation and development of the Hadith;⁷ despite criticism by certain scholars, and often for good reason, this thesis would remain dominant for most scholars up until the beginning of the second half of the twentieth century.

Nevertheless, from the 1960s on, a decisive turning point occurred, and this thesis came to be radically contested. An initial reason was the publication of two monumental studies in support of the great antiquity of systematic writing among scholars in the Islamic world: first, Studies in Arabic Literary Papyri by Nabia Abbott and then the Geschichte des arabischen Schrifttums of Fuat Sezgin. Both maintained that a systematic commitment to written form, in poetry as it happens, existed among the Arabs from the pre-Islamic period and that a written corpus began to develop in a consistent fashion with the coming of Islam and particularly, during the Umayyad period, i.e., grosso modo between 40/660 and 132/750. The thesis of the antiquity of the book has been discussed expertly and at great length, especially by German scholars. In the 1990s Gregor Schoeler, a specialist in the transmission of texts in Islam, revisited and analyzed the earlier studies and refined the way in which the problem is examined to a considerable extent. By introducing a pair taken from Greek, syngrammata-hypomnema, Schoeler was able to establish, among other things, an apt distinction between the act of “writing,” which does not always imply a written publication, and the act of “publishing,” in the sense of “making public,” which for a long time occurred solely in oral fashion.⁸

The second reason for promoting the theory of “the late book” was the discovery and editing over several decades of a growing number of very old sources, some of which are monumental and date mainly from the second half of the twelfth/eighth century: al-Ridda wa’l-futūḥ, attributed to Sayf b. ʿUmar, al-Muṣannaf fī’l-aḥādīth wa’l-āthār of Ibn Abī Shayba, Ta’rīkh al-Madīna of Ibn Shabba, or even al-Muṣannaf of ʿAbd al-Razzāq al-Ṣanʿānī. The fact that systematic works of this order of magnitude do exist appears to indicate that the tradition of the book had already been established for a certain time and went back to at least the beginning of the second, if not to the end of the first, century of the Hijra.

Similar problems in defining periods are met equally by specialists in the text and the redaction history of the Qur’an. In confronting the problems and contradictions encountered in Islamic writings, and particularly the interval that separates the traditionally assigned date for the definitive commitment of the Qur’an to written form (during the caliphate of ʿUthmān: 23–35/644–656) and the earliest sources providing this information, two methodological viewpoints may be distinguished within Western research: the hypercritical and the critical, to use Schoeler’s terminology.⁹ Without claiming to be exhaustive, let us review some important instances and names embodying these two viewpoints.

The hypercritical method is held by scholars who completely, or almost completely, reject the Muslim textual tradition. To that end they invoke the protracted span of time of the transmission of written data, the abundant divergences and contradictions of the authors, the lack of an independent form of control of the transmission, the undeniable presence of historical errors and improbable accounts, the plethora of legends, etc. In this context, a decisive step was taken with the analyses of the Arabic and Syriac specialist Alphonse Mingana beginning with his study of the transmission of the Qur’an.¹⁰ In taking up and extensively developing the theses of Paul Casanova on the central role of the Umayyad caliph ʿAbd al-Malik b. Marwān (reg. 65–86/685–705) and his governor al-Ḥajjāj b. Yūsuf in the establishment of the final recension of the Qur’an,¹¹ Mingana first stresses the scarcely believable nature of the Islamic sources dealing with the history of the Qur’anic redaction, because of the gap of almost two centuries that, in his view, separates the age of the Prophet from the oldest sources containing the reports of the commitment of the Qur’anic text to writing—specifically the Ṭabaqāt of Ibn Saʿd (d. 229/844) and the Ṣaḥīḥ of al-Bukhārī (d. 256/870) (it is true that, as we have seen, other, much earlier sources have been edited since Mingana’s day; we will come back to this). He then painstakingly draws on a certain number of Syriac sources coming from Eastern Christian circles during the two first centuries of Islam to conclude that an official version of the Qur’an could not have existed before the end of the seventh century of the common era and that that version, called the Codex of ʿUthmān, must have appeared during the time of ʿAbd al-Malik.

The hypercritical approach counted other distinguished scholars among its ranks, culminating in the two resounding works of John Wansbrough: Quranic Studies and The Sectarian Milieu.¹² Like his predecessors, this author drastically disputes the historical character of the reports dealing with the Qur’anic recension as transmitted by Muslim tradition. He assumes that the Qur’an, created out of an array of texts issuing from traditions that were originally independent of each other, could only have taken its definitive form at the end of the second/eighth century, if not at the beginning of the third/ninth century. For a number of reasons, to be considered in due course, the very late dating of Wansbrough was decisively rejected later, and not only by those inimical to the hypercritical method but also by his own successors on the methodological level, such as Patricia Crone and Michael Cook, who, relying mainly on the inscriptions of the Dome of the Rock as well as on non-Islamic writings, appear to come to the same conclusions as Mingana on the dating of the final official version of the Qur’an, namely, to the time of the caliph ʿAbd al-Malik.¹³

Regarding the other approach, scholars using the critical method deem that a scientific examination makes it possible in an apposite way to distinguish within the heart of the Islamic textual tradition between reports and data that are more or less credible, on the one hand, and those, on the other hand, that cannot be accepted. Here, clearly, the core problem is to find adequate criteria of assessment to establish this distinction. Exact recognition of the political and religious tendencies of the authors or the currents within which this or that text has taken shape, for example, might provide a basic criterion for evaluation. This method has been adopted by most of the great nineteenth-century scholars as well as by their successors in the twentieth century. This is the case, first of all, of the authors of the illustrious work Geschichte des Qorāns, beginning with Theodor Nöldeke, who after 1860 adopted the traditional Muslim account of the history of the Qur’an, and followed by those who continued his work such as Gotthelf Bergsträsser and Otto Pretzl. Nevertheless, Friedrich Schwally, author of the revised edition of the work after 1909, in following the methods Goldziher applied to the study of the Hadith, rejected a considerable number of traditional accounts (such as the first compilation made by Abū Bakr or the Qurayshite dialect of the Qur’an) in the effort to retain only the thesis according to which a significant portion of the Qur’an assumed its final form during the lifetime of the Prophet himself, together with the thesis that the definitive compilation was made during the caliphate of ʿUthmān.¹⁴ The critical approach subsequently found some defenders of stature in Great Britain, in particular John Burton who provided it with one of its monuments, The Collection of the Qur’an, which appeared in the same year as Wansbrough’s Quranic Studies.¹⁵ Despite the fact that these two scholars base themselves on the methods of Goldziher and of Schacht to pose the fundamental problem of the trustworthiness of the Islamic tradition, with respect to what concerns the dating of the final redaction of the Qur’an, Burton comes to radically different conclusions from those of Wansbrough. Rejecting the reports regarding the codices first gathered together in Abū Bakr’s time and then in the time of ʿUthmān—“ideological reports” that he ascribes to later doctors—Burton makes a distinction between a very old “Qur’an document” and a later “Qur’an source” to draw the conclusion that the Qur’anic text as it is known appeared mainly during the lifetime of Muḥammad himself.¹⁶

Among the many advocates of the critical methodology, we may yet again cite Schoeler who in this domain presents himself explicitly as a follower of Nöldeke.¹⁷ With respect to the definitive recension of the Qur’an, Schoeler dismisses Wansbrough’s hypotheses, basing himself on the famous Qur’an of Ṣanʿā’ and the studies of the group under Gerd R. Puin who have ascertained, by Carbon 14 dating, that this manuscript may be dated to between 37/657 and 71/690, hence a short time after ʿUthmān.¹⁸ The problem is that in the absence of a scientific edition, it is still not known whether the many fragments of this manuscript constitute the entire text of the Qur’an or not. Moreover, it might be asked why, several decades after the major discovery of this Qur’an, Gerd Puin and his collaborators have published only a few brief articles on the subject.¹⁹

The method of the critical scholars can be summed up as follows: one must hold an ancient source or tradition to be authentic and its contents plausible just as long as no valid reasons have been found for rejection. Cook, a renowned representative of the hypercritical method, suggests the following procedure with regard to this approach: we shall probably have better success in coming close to historical truth if we reject everything that we have no specific reason for accepting.²⁰

Harald Motzki, a scholar professing the critical approach, illustrates very well the differences among Western scholars over the dating of the final version of the Qur’an by recounting the conclusions of four of the most decisive among them: taking them in chronological order, Schwally dates this version to the time of the caliph ʿUthmān, Mingana to the caliphate of ʿAbd al-Malik, Wansbrough to the beginning of the third/ninth century, and, at the same time, Burton dates it to the time of the Prophet Muḥammad.²¹

Aside from the problems posed by the Islamic textual tradition, several internal elements of the Qur’anic text have continually raised questions as well for Orientalists. For example, words and expressions that have perennially remained obscure not only for modern specialists but for medieval Muslim commentators themselves whose commentaries, as abundant as they are contradictory—at times in a single author—illustrate uncertainty if not downright ignorance. This is the case of the expression jizya ʿan yad (Qur’an 9:29), critical analysis of which extends from the pioneering study of F. Rosenthal ²² to the meticulous examination recently carried out by U. Rubin.²³ Meanwhile, the journal Arabica has served for years as a forum for learned discussions of this expression by C. Cahen, M. M. Bravmann, and M. J. Kister.²⁴ The word īlāf from Sūra 106 (Quraysh) has been discussed by H. Birkeland, M. Cook, P. Crone, and U. Rubin.²⁵ The term kalāla (Qur’an 4:12) has been studied at length by D. Powers in numerous publications.²⁶ So too with the term al-ṣamad from Sūra 112 (al-Ikhlāṣ), the problematic character of which F. Rosenthal, R. Paret, C. Schedl, U. Rubin, and A. A. Ambros, among others, have noted.²⁷ To this list might be added—without making it at all exhaustive—the word ḥanīf, or, indeed, the famous individual letters, “the Openers” (al-fawātiḥ), at the beginning of twenty-nine of the sūras,²⁸ or yet, as M. Cook stresses in his introduction to the Qur’an, such terms as abābīl, sijjīl, or al-qāriʿa.²⁹

With respect to the redaction of the Qur’an and its evolution, the basic question that may be asked is the following: for what reasons did Muslim scholars from a very early period, within a few decades after the Prophet’s death, not know, or no longer know, the meaning of these words, these expressions, these separate letters? M. Cook, in the aforementioned work, proposes the following: either the matter of the Qur’an was not made available as Scripture until several decades after the Prophet’s death or many of the terms found in the Qur’an were for one reason or another already obsolete during the Prophet’s’s lifetime.³⁰ It has to be acknowledged that the how and the why of this phenomenon are still far from being adequately explained. Nevertheless, this would demonstrate that there probably was an evolution in the redaction of the Qur’an, that it would have had a progressive elaboration in successive stages. We shall return to this.

Another breach involves the juridical domain. Already, in his great work The Origins of Muhammadan Jurisprudence, Joseph Schacht had stated that Islamic law, such as we know it from the second century of the Hijra, is surprisingly non-Qur’anic.³¹ Subsequently, the work of Burton on the rules of inheritance, the rights of widows, and even the stoning of adulterers,³² those already cited by David Powers on the right of inheritance,³³ Motzki’s work on the muḥṣanāt/muḥṣināt of Qur’an 4:24,³⁴ Gerald Hawting’s analysis on the rights of a repudiated wife during her “period of waiting” (ʿidda),³⁵ or Patricia Crone’s study of the Qur’anic word kitāb, in the overlooked sense of verse 24:33 where it has the meaning “marriage contract”³⁶—all this tends to show that less than a century after the Prophet’s time certain important aspects of the law had become not solely non-Qur’anic (as Schacht put it) but at times firmly anti-Qur’anic. In all the cases that have been studied the general impression is that Qur’anic expositions or those ascribed to Muḥammad were neglected and/or that their practical application was abandoned. Why? The reason may lie in the late and nonconsensual finalization of the Qur’an, but it could also be due to the heavy usage of ra’y, the jurist’s personal opinion and hence not deriving his reasoning directly from the Qur’an. The question remains open.

These questionings and breaches, which have still not been satisfactorily explained, underlie the hypercritical approach. Here it may be useful to recapitulate the theses of the most radical, and the most renowned, of its representatives: Wansbrough.³⁷ According to him, the Qur’an originated neither in Arabia nor even in Islam. The Arabs had not founded a new religious community of their own by the time that they set out from their native land to conquer other countries. It was outside of Arabia when, following the conquests, they discovered a “sectarian milieu” in the Middle East, and particularly in Iraq, and then began steadily to adopt this “milieu” and modify it for themselves, all the while rewriting its history by “arabicizing” its evolution. In this way the Qur’an emerged from a multitude of sources at the very heart of a process whereby popular preachers (qāṣṣ, pl. quṣṣāṣ) played a leading role. Indeed, the popular sermon would have served as the means both of the transmission and the explanation of pronouncements deemed prophetic, pronouncements that emerged from this very sectarian environment. The Qur’an, originating out of this composite material, detached itself from it only gradually. The establishment of its text took place so slowly that the date of the definitive version cannot be prior to around the year 800 CE (around the 180s of the Hijra) during the first Abbasid period.

Even though Wansbrough’s arguments are often forceful, and his theories as apposite as they are suggestive, his dating of the final version of the Qur’an no longer seems defensible, as we have already indicated. Apart from the reasons already given, on codicological, archaeological and epigraphical grounds, several discoveries render it equally indefensible: in addition to the Ṣanʿā manuscript, a fragment from Khirbet el-Mird citing Qur’an 3:102ff., described by A. Grohmann and analyzed by Kister, appears to prove that a stable text was already in existence toward the end of the Umayyad period.³⁸ The Nubian papyrus datable to 141/758, containing two Qur’anic verses, preceded by the formula “And God, may He be glorified and exalted, says in His book,” furnishes another indication of the same order.³⁹ There are also numismatic discoveries dating from the Umayyad period, studied by Crone and Hinds, that mention the name of Muḥammad as God’s messenger as well as Qur’anic citations.⁴⁰ Then there is the literary tradition along with textual evidence. Reports on the systematic collecting of the Qur’an, most especially during the caliphates of Abū Bakr and ʿUthmān, may be found in sources as early as al-Ridda wa’l-futūḥ attributed to Sayf b. ʿUmar (d. 184/800),⁴¹ al-Jāmiʿ of ʿAbd Allāh b. Wahb (d. 197/812), the Musnad of al-Ṭayālisī (d. 204/820), or the Kitāb faḍā’il al-Qur’ān of Abū ʿUbayd al-Qāsim b. Sallām (d. 224/838).⁴² In his recent aforementioned work Motzki convincingly shows almost all of these reports are attributable to the scholar and traditionist Ibn Shihāb al-Zuhrī (d. 124/742).⁴³ This brings us back, yet again, to the time of the first Marwānid Umayyads, as Mingana had maintained.⁴⁴ This is a very early date and still, even so, several decades after the time of the third caliph. A few dozen years that may be reckoned as several centuries so enormous were the consequences, between those two epochs, of the civil wars and the great blazing conquests which overwhelmed both the history and the mind-set of the earliest Muslims.

Let us look very quickly at the most recurrent account of Islamic tradition on the collection and the writing of the Qur’an.⁴⁵ According to the versions which were to become predominant, after the Prophet’s death there was no complete codex of the Qur’an that he duly authorized. Excerpts, some long, some short, were preserved by certain Companions, some of which were written on all sorts of material. An initial recension was determined by Abū Bakr, the first caliph, acting on the advice of the future second caliph ʿUmar and effected by Zayd b. Thābit, the Prophet’s scribe, though he was reticent at first. This gave rise to the compilation of one copy, a codex tradition often calls “the codex (that is) between the two covers.” After ʿUmar’s death, this copy remained in the family, since his daughter Ḥafṣa, one of the Prophet’s wives, had inherited it. Alongside this codex, other texts existed thanks to the initiative of highly placed individuals, the best known of whom are ʿAlī b. Abī Ṭālib, Ubayy b. Kaʿb, ʿAbd Allāh b. Masʿūd (Ibn Masʿūd), or Abū Mūsā al-Ashʿarī. On the advice of his renowned general Ḥudhayfa, ʿUthmān, the third caliph, decided to have an official edition of the text of the Qur’an established; this is what is called the “model edition” or the edition or vulgate of ʿUthmān. The task devolved upon Zayd b. Thābit once again, assisted this time by a commission made up of men from the Quraysh (oddly enough, the sources are not unanimous on either their number or their identity). This commission took as its working text the codex belonging to Ḥafṣa, ʿUmar’s daughter. The caliph then accorded an official and binding character to this edition. He dispatched copies of the reworked text into the different provincial capitals of the empire where they were to serve as reference copies. To conclude his work, he then gave the order to destroy the other editions. As we shall see subsequently, it took several centuries before this official edition came to be accepted by all Muslims as the textus receptus.

Let us now rapidly examine certain elements of the early reports which are questionable. First, on the philological plane, the most important terms in our inquiry appear to be ambiguous in the early period, beginning with the word qur’ān itself. The term can designate statements deriving from Muḥammad. In certain traditions qur’ān appears as a generic name to refer to everything that was heard from the Prophet (in this case we will write qur’an with a small q). According to a statement reported by Ibn Saʿd, Salima b. Jarmī said, “I have collected many qur’āns (qur’ānan kathīran) from Muḥammad.⁴⁶ Indeed, a clear distinction between hadith and Qur’an—the former indicating the Prophet’s statements and the latter the words of God—seems to be late.⁴⁷ Thus, in an epistle attributed to the Zaydī imam Zayd b. ʿAlī, we come upon two hadiths beginning with the phrase “Muḥammad said,” the respective contents of which may be found with slight linguistic differences in the Qur’anic verses 5:56 and 21:24.⁴⁸ Conversely, in verse 21:4, the readings waver between qāla and qul, and the commentators are anything but unanimous as to whether a statement of the Prophet or a divine injunction is at issue. Alfred-Louis de Prémare has studied another apt example of initial uncertainty between qur’ān and hadith in the famous sermons the Prophet gave shortly before his death that tradition has called the “farewell sermons” (khuṭbat al-wadāʿ). In fact, certain sentences in these sermons, and especially whatever pertains to women and to the sacred months, are to be found, with slight variations, in the Qur’an.⁴⁹ Finally, we have the puzzling example of the hadith qudsī, statements issuing from God, the transmission of which is ascribed to Muḥammad but which, nevertheless, are not to be found in the Qur’an.⁵⁰

De Prémare has emphasized as well the problematic nature of the root jamaʿa in the recurring expression jamʿ al-qur’ān.⁵¹ The root J-M-ʿ has the obvious meaning of “to collect, to assemble, to gather,” but many Muslim lexicographers tell us that it can also mean “to memorize, to retain, and to learn by heart,” though a specific verb with this last meaning exists, namely, ḥafiẓa. Hence, in quite a few accounts, we are told that the Qur’an was first “assembled in the hearts of men”—i.e., memorized—and then “collected”—i.e., gathered together—by one or another of Muḥammad’s Companions. To be sure, this may be merely a reflection of a dialectic between the oral and the written and yet it seems that the ambiguity affecting the verb jamaʿa is deliberate and that at its origin lies a concern to evade the most blatant contradictions that were current in the accounts of the editing of the Qur’an, or, indeed, there was an intent to conceal the conflicts that raged between different political and religious tendencies on this subject. Thus ʿAlī said that he collected a complete Qur’an in a codex immediately following the death of Muḥammad. But now Ibn Abī Dāwūd, who recounts this tradition in his Maṣāḥif , stipulates at once that here the expression jamaʿtu ‘l-qur’ān kullahu actually means “I have learned the Qur’an by heart in its entirety.”⁵² To complicate matters even further, the Qur’an itself states innā ʿalaynā jamʿahu wa-qur’ānahu (Qur’an 75:17; literally, “Upon us is its putting together / its memorization and its ‘qur’an’”). Who is speaking and what do the terms jamʿ and qur’ān mean here? Exegetes are far from unanimous on this.

Yet another puzzling point: many accounts report that when Abū Bakr and ʿUmar summoned Zayd b. Thābit to charge him with writing down the Qur’an, Zayd b. Thābit was at first furious and exclaimed, “Do you want to do what the Messenger of God himself did not do?”⁵³ Did Muḥammad then intend to preserve the Qur’an in the form of oral recitation (or indeed as a unique book), as was done with poetry?⁵⁴ In this scenario the studies of Fr. Edmund Beck regain all their relevance. Indeed, Beck was the first to have recognized the great similarity between the first “reciters of the Qur’an” (qāri’, pl. qurrā’) and the old rāwī, pl. ruwāt, those transmitters, especially in oral fashion, of the archaic Arabic poetry of the pre-Islamic period.⁵⁵ Now, for these ruwāt, the variants in the poetry, noticeable mainly during recitation, were not only no defect but rather were desirable since they allowed improvements in the poem. According to Beck, it is entirely possible that the first qurrā’, who were active up to at least the middle of the second/eighth century, may have considered the variants in different recitations, present for better or worse in the different versions of the Qur’an, as useful opportunities for improving the linguistic level of the text. This may account for the hadith attributed to the Prophet and especially appreciated by the first qurrā’: “In the muṣḥaf there are expressions in dialect (laḥn) but the Arabs will regularize them.”⁵⁶ These reader-reciters, experts in Arabic, are sometimes called in the sources “the masters of the Arabic language” (aṣḥāb al-ʿarabiyya); according to al-Ṭabarī, they reproached ʿUthmān fiercely for the official establishment of his vulgate: “There were several Qur’ans. You have abandoned them all except for one.”⁵⁷ It is interesting to observe that in al-Sayyārī’s text, the aṣḥāb al-ʿarabiyya are presented, among others, as artisans in altering the Qur’anic text.⁵⁸ An echo of the complaint addressed to the third caliph also occurs in a statement reported in several sources under differing forms and attributed to ʿAbd Allāh, the pious son of ʿUmar b. al-Khaṭṭāb and a figure admired by the qurrā’, when he said to ʿUthmān,”Let no one tell you that he has the entirety of the text of the Qur’an in his keeping. How can one know what the entirety of the text of the Qur’an is? Many things in the Qur’an have vanished forever (qad dhahaba minhu qur’ānun kathīrun; literally, “many qur’ans have vanished from it”).”⁵⁹ Other reports also refer to the suppression of parts of the Qur’an as well as to additions. First of all the suppressions: Abū ʿUbayd al-Qāsim b. Sallām (154–244/770–858) points out in his Faḍā’il al-Qur’ān a certain number of “censorings” applied to the Qur’anic text, some of which date from before the caliphate of ʿUthmān.⁶⁰ The two short suras “al-Ḥafd” and “al-Khalʿ,” which were part of the recension of Ubayy b. Kaʿb, along with other short texts ascribed to the codices of ʿAlī or of ʿUmar, were in the end not incorporated into the final version of the Qur’an.⁶¹ The same goes for certain verses: the verse on stoning (āyat al-rajm) or, indeed, the sentence: “If the son of Adam had two valleys of gold he would want a third / only earth can fill the belly of the son of Adam / God turns towards him who turns (towards Him).”⁶² Many other “missing texts” from the definitive version are mentioned, over several dozen pages, in the Geschichte des Qorāns, to cite only this work.⁶³ As for the additions, al-Sayyārī alludes to these in his work,⁶⁴ as does ʿAbd al-Masīḥ al-Kindī in his epistle as well as Ibn Abī Dāwūd al-Sijistānī in his K. al-Maṣāḥif.⁶⁵ Certain Khārijites, the ʿAjārida, considered the sura of Joseph to be apocryphal and added at a later date to the true revelations made by the Prophet.⁶⁶ Furthermore, Ibn Masʿūd deemed the first sura (al-Fātiḥa), together with the two last suras of the official vulgate, known as al-muʿawwidhatān, as prayers rather than Qur’anic revelations.⁶⁷

That the final codification of the Qur’anic text was linked to political and religious tensions is shown by an account reported by several sources, beginning with Bukhārī in his Ṣaḥīḥ or Ṭabarī in his Qur’an commentary:⁶⁸ Ḥudhayfa, the general who participated in the conquest of Armenia in the years 25–26/645–646, is alarmed by the disparities among his soldiers—most of them originally from Iraq—with respect to recitation of the Qur’an. He then entreats the caliph ʿUthmān to establish a unique and definitive written version of the Qur’an: “Unify this community before it diverges over its Book as the Jews and the Christians have diverged (over theirs).” Ṭabarī makes it clear that in this army a conflict had broken out over the Qur’anic text between the Iraqis and the Syrians (that is, between the partisans of ʿAlī and those of ʿUthmān, the Umayyads, or even between the clan of the Banū Hāshim and that of the Banū Umayya, the Umayyads?).

As de Prémare has aptly emphasized, this type of account appears to have more of a symbolic than an historical value.⁶⁹ Shortly after the Prophet’s death, the Arabs, skillful merchants, had turned into great conquerors, enjoying absolutely immense powers and riches within a few short years. A few of these extremely rich and powerful men seem to have concentrated within their own persons several elements that would prove decisive in the complex of problems arising from the definitive recension of the Qur’an. In synthesizing earlier studies while continuing to follow their methods—and particularly, those of Casanova, Mingana, and Crone and Cook—de Prémare specifically delineates the portraits of three personalities of the utmost significance:⁷⁰

1. First, ʿUbayd Allāh b. Ziyād, called Ibn Ziyād, the cruel and renowned Umayyad governor of Iraq (from 56 to 67/675 to 686). The adopted grandson of Abū Sufyān, hence a member of the powerful ruling family, he was one of those wealthy merchants who had become a mighty conqueror. After the conquest of Transoxania, he held the strategic post of governor of Iraq during the caliphates of Muʿāwiya and Yazīd I. He violently crushed anti-Umayyad revolts, especially those of the Alids, most conspicuously massacring the imam al-Ḥusayn, son of ʿAlī and grandson of the Prophet, along with almost all his family at Karbalā’. ʿUbayd Allāh b. Ziyād was, at the same time, an extremely literate man; he was one of the “experts in the Arabic language” (aṣḥāb al-ʿarabiyya). Abū al-Faraj al-Iṣfahānī, in his Aghānī,⁷¹ as well as Yāqūt, in his Udabā’,⁷² each devote several lines to him. Ibn Abī Dāwūd al-Sijistānī writes that ʿUbayd Allāh b. Ziyād had intervened during the establishment of the Qur’anic text. He obliged his Persian secretary Yazīd b. Hurmuz al-Fārisī to insert numerous additions: “ʿUbayd Allāh added (zāda) two thousand ḥarf (words, letters or expressions?) to the codex,” reported Yazīd.⁷³ To be sure, the secretary provides no details either on the identity of the codex in question or on the nature of the additions of which he speaks. Moreover, the ambiguity—without doubt, intentional—of the word ḥarf leaves the door open for all sorts of hypotheses.

2. The second person is also a governor of Umayyad Iraq, the no less celebrated al-Ḥajjāj b. Yūsuf, celebrated, sadly enough, for the horrendous tortures and massacres that he inflicted on the supporters of ʿAlī. He became governor during the reign of ʿAbd al-Malik b. Marwān (caliph from 65 to 86/685 to 705), the second and indubitably the most important caliph of the Umayyad Dynasty of the Marwānids. During this period the first ripples of doctrinal reflection began to appear following the conquests, all of which were deeply concerned with the problem of political and religious legitimacy. In this time of ceaseless civil wars, the issue of the scriptural texts took on major importance. Al-Ḥajjāj was also the conqueror of the rival caliph Ibn al-Zubayr who was based in Mecca. After more than ten years (62–73/681–692), the power of the caliph thus regained its unity. Al-Ḥajjāj was a man of letters as well and one of the great experts in the Arabic language. Ṭabarī in his History,⁷⁴ or Jāḥiẓ in his Bayān,⁷⁵ adduce some examples of his flamboyant prose. Moreover Ibn Abī Dāwūd al-Sijistānī, as well as Ibn Khallikān,⁷⁶ inform us that al-Ḥajjāj b. Yūsuf also effected massive interventions into the Qur’anic text. According to reports that are at times contradictory, he corrected divergent readings, arranged certain suras or verses, and improved the orthography by the introduction of the diacritical marks and vowels that had not previously existed. Based on a report by al-Samhūdī, the historian of the city of Medina, which has been masterfully analyzed by Mingana,⁷⁷ al-Ḥajjāj had established his own Qur’anic edition, a copy of which he sent to each of the provincial capitals of the Islamic empire in order to make it the official version at the expense of those earlier versions that he had in any case had destroyed in Iraq.⁷⁸ According to certain accounts, he was the first who took this initiative. But, according to other reports, he was replicating ʿUthmān’s action in having rival codices destroyed. According to still others, the other recensions remained in circulation, and it was under the Abbasids that the codex of al-Ḥajjāj was definitively set aside.⁷⁹ The name of this governor of Iraq is also to be found in the polemic of the Christian Arab apologist ʿAbd al-Masīḥ al-Kindī, written perhaps at the beginning of the third/ninth century.⁸⁰ In one of his epistles he (or whoever is writing under his name) states, “then, there was the intervention (i.e., in the Qur’anic text) by al-Ḥajjāj b. Yūsuf who left no version unseized. He had many verses dropped and he added others which in the view of some pertained to the men of the Banū Umayya and the Banū al-ʿAbbās, identified (in the text) by their own names.⁸¹ A version in conformity with the recension ordered by al-Ḥajjāj was made in six copies: one was sent to Egypt, another to Damascus, a third to Medina, a fourth to Mecca, a fifth to Kūfa, and the final copy to Basra. As for the other earlier editions, he plunged them into boiling oil and destroyed them, in so doing imitating ʿUthmān.”⁸²

3. The third personage is the caliph ʿAbd al-Malik himself. De Prémare’s historical investigation—like those before him of Hawting on Umayyad history or of Yehuda Nevo on inscriptions from the period—reveal that the period of ʿAbd al-Malik was a defining moment in the elaboration of Islamic doctrine.⁸³ With regard to the final establishment of the Qur’anic text, there is, to begin with, the caliph’s declaration reported by several sources: “I fear dying during the month of Ramadan; it was in this month that I was born; it was in this month that I was weaned and in this month I collected (jamaʿtu) the Qur’an.”⁸⁴ To be sure, here too we can reflect that the meaning of the root jamaʿa can be debated; yet what we already know about the interventions of the two governors of Iraq at this very period into the Qur’anic text appears to indicate that here the term does not mean “memorize, learn by heart” but rather “collect, put in order.”

Other noteworthy aspects of the reign of ʿAbd al-Malik stand persuasively in direct relation with our set of problems. The inscriptions on the Dome of the Rock, meticulously analyzed by Christel Kessler and Oleg Grabar, were executed by order of this caliph.⁸⁵ They appear to be the first actual dated illustration of a dogmatic definition of theological oneness in the Islamic context: “Say: God is Unique, He is the Impenetrable (?—the problematic term al-ṣamad is here at issue), He has not begotten nor has He been begotten and He has no associate” (qul huwa allāhu aḥad allāhu ṣ-ṣamad lam yalid wa-lam yūlad wa-lam yakun lahu kufu’an aḥad). The text is preceded on the inside of the Dome, on the outer southern face of the octagonal arcade, by the basmala (the formula: “In the Name of God the Compassionate the Merciful”) and the unitary formula “there is no God but Allāh who has no associate.” Apart from the final formulation, this of course constitutes Sura 112, al-Ikhlāṣ. And yet why this divergence from the Qur’anic text? Why is the formulation that immediately follows the basmala missing from the sura? There can be no question of carelessness in such a painstaking and protracted project. Is it because the Qur’anic text had not yet been definitively established?⁸⁶ Leaving aside its doctrinal aspect, the polemical character of this text is well known. Indeed, except for declarative texts, the inscriptions on the Dome of the Rock bear polemical texts of this type, directed quite specifically against Christian Trinitarian dogma and Christology.

Moreover, in accord with the studies by Amikam Elad and Yehuda Nevo, the first true glorifications of Muḥammad as the prophet of a religion independent of both Judaism and Christianity also date from the period of ʿAbd al-Malik.⁸⁷ In fact, several fundamental decisions in the early history of Islam are due to this caliph:

In this context the caliph and the other men in power who surrounded him could not be indifferent to that fundamental aspect of power which is the control of belief and the fact that the latter cannot be effective except through the control and hence the codification of the Scriptures, all the more so in a society in which dissident religious and political movements are quite numerous. A unique scriptural Book, independent of earlier Scriptures, specifically those of the Jews and Christians, codified in accord with state dogmas, is the surest guarantee of doctrinal, hence political, security. It is interesting to observe that the second scriptural source in Islam, that is, the Hadith, seems also to have begun to assume its systematic character from the time of ʿAbd al-Malik. Thanks to Goldziher, we know that a diffident writing down of Hadith began before this date,⁸⁹ but the studies of Michael Lecker and of Harald Motzki have demonstrated that the first person to have systematically written down Hadith was Ibn Shihāb al-Zuhrī by order of the caliph Hishām (reg. 105–127/724–743); and yet, he was already a court scholar in the time of the latter’s father, namely, ʿAbd al-Malik.⁹⁰ Thus, the initiative to establish an official Qur’anic codex, seemingly taken during the caliphate of ʿUthmān, seems to have been realized during the reign of ʿAbd al-Malik or perhaps a bit later.⁹¹ Meanwhile, together with the conquests and the intent of Muslim leaders to distinguish their religion clearly from both Judaism and Christianity, a continuing editorial activity—including composition, rewriting, stylistic refinement, elaboration, and correction, etc.—will have taken place.⁹² In addition, the point to be underlined, and that de Prémare did not emphasize is that the three persons I have presented (i.e., ʿUbaydallāh b. Ziyād, al-Ḥajjāj b. Yūsuf, and ʿAbd al-Malik b. Marwān) have two common traits: their belonging to the Umayyad camp and their violent hostility toward the Alids.

Islamic tradition, by an immense majority, insists on the very great antiquity of the establishment of the official version of the Qur’an (the initiative taken from the time of Abū Bakr and ʿUmar and finalized by ʿUthmān) since it is well aware that the later this establishment is located the greater the risk of taḥrīf (falsification). Nevertheless, despite all attempts to conceal the divergences on the part of “orthodox” authors, an examination of the hesitancies or contradictions that the sources bear along with them clearly shows that from the start a large protest movement took shape against the official version of the Qur’an. Indeed, in a society in which the different recensions might serve as the means of legitimation of diverse political and religious groups, the so-called ʿUthmānian version would have taken several centuries to be unanimously accepted by all Muslims. Without a doubt the most significant symbol is the double condemnation of Ibn Miqsam and Ibn Shannabūdh at the beginning of the fourth/tenth century for having practiced noncanonical “readings.” At the close of the same century, in 398/1007, a polemic broke out between the Sunnis and the Shi’ites of Baghdad over the licit or illicit character of the codex of Ibn Masʿūd. In the end a Sunni tribunal gave orders for its destruction.⁹³ According to Ibn Nadīm, a century earlier, there still existed copies of the recension by Ubayy b. Kaʿb circulating in the region of Baṣra.⁹⁴

Many thus believed on various grounds that the version of ʿUthmān was not faithful to the revelations accorded to the Prophet. In a well-documented article, Hossein Modarressi demonstrates that a certain number of persons, whom later Sunnism is averse to repudiating, harshly criticized the vulgate of ʿUthmān during the two or three first centuries of the Hijra.⁹⁵ Some of the first Muʿtazilites and the first ʿAjārida Khārijites did so as well. As we have seen, the latter conspicuously rejected the authenticity of Sura 12 (Yūsuf), largely because of the sensual resonance of the story of Joseph, and so considered it a later addition.⁹⁶ Nevertheless, the most direct, energetic, and numerous accusations of falsification of the Qur’an by the first three caliphs and their entourage, held up as enemies of ʿAlī, are to be found in Imami Shi’ite sources, particularly from the pre-Buwayhid period.⁹⁷ Briefly put, these held that ʿAlī, the only true initiate and the legitimate successor to Muḥammad, was in unique possession of the complete version of the revelation accorded to the Prophet. After the Prophet’s death and the seizure of power by ʿAlī’s enemies, this version, far longer than the official one, was set aside by the authorities mainly because it included explicit homage to the first Imam, his descendants and supporters as well as quite open attacks against his adversaries. Once rejected, ʿAlī concealed his recension, to be handed down secretly to the imams of his line (see the introduction and chapter 1). These beliefs, since at least the fourth/tenth century, form part of the main and constant accusations of Sunni or Muʿtazilite authors and heresiographers who spot in them one of the most flagrant elements of Shi’ite “heresy.” More specifically, the Imami or Twelver Shi’ites, with the drift of statements traceable back to their imams throughout the immense corpus of early Shi’ite hadith, have radically and explicitly cast doubt on the integrity of the vulgate of ʿUthmān and thus have accused non-Shi’ites of having falsified the Qur’an. This thesis, coherently set out in the great treatises with abundant historical reports and Imami doctrinal elements, is mostly recounted in pre-Buwayhid Imami sources. After this period, that is, the middle of the fourth/tenth century, for quite precise doctrinal and historical reasons, the majority of Twelver scholars, beginning with Ibn Bābawayh al-Ṣadūq (d. 381/991), started to go back on this thesis and to adopt the Sunni viewpoint.⁹⁸

Virtually all of the pre-Buwayhid Imami works that have come down to us, written as they often were by recognized religious authorities, evoke, at greater or lesser length, in direct or indirect manner, the falsified character of the version of the Qur’an known as the ʿUthmānian. Aside from al-Sayyārī’s work, let us cite those of al-Faḍl b. Shādhān al-Nīsābūrī (d. 260/874), Aḥmad b. Muḥammad al-Barqī (d. 274/887–88 or 280/893–94), al-Ṣaffār al-Qummī (d. 290/902–3), Saʿd ʿAbd Allāh al-Ashʿarī al-Qummī (d. ca. 300/913), Muḥammad b. Yaʿqūb al-Kulaynī (d. 328 or 329/939–40 or 940–941), Ibn Abī Zaynab al-Nuʿmānī (d. ca. 345/956 or 360/971), ʿAlī b. Aḥmad al-Kūfī (d. ca. 352/963 author of al-Istighātha fī bidaʿ al-thalātha) or the authors of the first Imami Qur’an commentaries such as Furāt b. Ibrāhīm al-Kūfī (d. 300/912), ʿAlī b. Ibrāhīm al-Qummī (d. ca. 307/919), Muḥammad b. Masʿūd al-ʿAyyāshī (third–fourth/ninth–tenth century).

The Buwayhid period marks the marginalization of the original “esoteric and nonrationalist” tradition, of which the sources we have just mentioned might illustrate the principal doctrines. With the close of the era of the historical imams, the surge of a certain rationalism in virtually all religious milieus, and especially the coming to power of Shi’ites at the very heart of the Sunni Abbasid caliphate, the Imami doctors, and particularly those in Baghdad, the capital, seem to have felt constrained to break at certain points with the original tradition in order to elaborate a new “rational theological and political” tendency. In seeking to come ever closer to the more and more firmly established Sunni “orthodoxy,” belief in the falsification of the official Qur’an, steadily sanctified and pronounced inseparable from the faith, could no longer be sustained.⁹⁹ The greatest Imami scholars of this period will declare that the vulgate of ʿUthmān is a faithful version of Revelation, and, at the same time, they will remain quite discreet about the religious authorities who preceded them and had maintained the contrary. Ibn Bābawayh, al-Shaykh al-Ṣadūq, seems to have been the first great Twelver author not simply to pass over silently or criticize by allusion the earlier traditions on this matter but, indeed, to have adopted a position identical to that of the Sunnis: “In our view (i.e., the Twelver Imamis), the Qur’an revealed by God to Muḥammad is identical to that which exists between the two covers [mā bayna al-daffatayn, i.e., the version of ʿUthmān declared official]…whoever claims that the revealed text was ampler than the established one is nothing but a liar.”¹⁰⁰ Thus our author says nothing about the quite plentiful traditions in which there is question of falsification, censorship, and alteration (taḥrīf, maḥw, tabdīl/taghyīr). His disciple and commentator, the Shaykh al-Mufīd (d. 413/1022) confines himself to speaking in certain of these writings of a change that has occurred in the order (ta’līf) of certain verses or suras or even of the elimination by certain Companions of the Qur’anic exegeses of ʿAlī, exegeses appearing on the margins of the first imam’s recension.¹⁰¹ Other thinkers, inimical to the thesis of falsification, will take up the same discourse again even while more or less developing it further, e.g., al-Sharīf al-Murtaḍā ʿAlam al-Hudā (d. 436/1044), in al-Masā’il al-Ṭarābulusiyyāt al-ūlā,¹⁰² or, indeed, al-Shaykh al-Ṭūsī (d. 460/1067), in al-Tibyān.¹⁰³

This position with respect to the Qur’an, in agreement with other Muslims, henceforth became the majority view within a Twelver Shi’ism dominated by that rationalist current later known as the Uṣūliyya. We limit ourselves to a few celebrated examples which, however, do not all belong to this rationalist tendency: al-Faḍl b. al-Ḥasan al-Ṭabrisī (d. 548/1153), Raḍī al-Dīn Ibn Ṭāwūs (d. 664/1266), al-ʿAllāma al-Ḥillī (d. 726/1325), al-ʿĀmilī al-Bayāḍī (d. 877/1472), Mullā Muḥsin al-Fayḍ al-Kāshānī (d. 1091/1680), al-Ḥurr al-ʿĀmilī (d. 1104/1692)…¹⁰⁴ According to these authors’ manifold arguments, the traditions maintaining that the vulgate of ʿUthmān underwent alteration, concocted in heterodox circles, are utterly untrustworthy and form no part of Shi’ite belief. Notwithstanding their lofty position in the transmission of doctrine, the compilers and scholars who reported these traditions without submitting them to a critical examination were lacking in both lucidity and vigilance.

Even so, there have always existed thinkers, more or less affiliated with the Akhbāriyya traditionalist current (the rivals of the rationalistic Uṣūliyya)—in a minority, to be sure, and yet enjoying a certain importance in the history of Shi’ite thought—who have maintained the thesis of taḥrīf precisely out of a respect for tradition. Among the most renowned after the Buwayhid period, mention may be made, in the sixth/twelfth century, of Aḥmad b. ʿAlī al-Ṭabrisī or Ibn Shahrāshūb. After the Safavid turning point during the eleventh/seventeenth century: Muḥammad Bāqir al-Lāhijī, Muḥammad Ṣāliḥ al-Māzandarānī, or Niʿmat Allāh al-Jazā’irī. In the twelfth/eighteenth and thirteenth/nineteenth centuries: ʿAbd Allāh al-Ḥusaynī al-Shubbar or Aḥmad al-Narāqī. Finally, three particularly important works need to be mentioned, namely, the Qur’an commentary Mir’āt al-anwār and the Ḍiyā’ al-ʿālamīn fī ‘l-imāma by Abū al-Ḥasan Sharīf al-ʿĀmilī al-Iṣfahānī (d. ca. 1140/1727) as well as the Faṣl al-khiṭāb of Mirzā al-Ḥusayn al-Nūrī al-Ṭabarsī/Ṭabrisī (d. 1320/1902).¹⁰⁵

At the same time, a close reading of the texts seems to indicate that the attitude of the Imami scholars toward the vulgate of ʿUthmān is at times extremely complex, due no doubt to a certain sense of embarrassment, and that the thesis of the existence or nonexistence of any falsification occasionally cuts across the rationalist/traditionalist divide. We have already chosen to list several great names associated with the traditionalist current, and specifically Ibn Ṭāwūs, al-Fayḍ al-Kāshānī, or al-Ḥurr al-ʿĀmilī among the opponents of taḥrīf even though in fact their positions are not always transparent. Conversely, several great names associated with the rationalist tendency also display an ambiguous attitude toward this unusually delicate issue: starting with the Buwayhid period, the Shaykh al-Mufīd (d. 413/1022);¹⁰⁶ in the Safavid period, the renowned al-Majlisī (d. 1110/1699 or 1111/1700);¹⁰⁷ or even, in the nineteenth and twentieth centuries, as R. Brunner has shown, such famous mujtahids as Aḥmad l-Narāqī, Shaykh Murtaḍā al-Anṣārī, Ākhund Muḥammad Kāẓim al-Khurāsānī, or, indeed, the Ayatallah Khomeini.¹⁰⁸

Of the works just cited, the Faṣl al-khiṭāb of Mirzā al-Nūrī al-Ṭabarsī/Ṭabrisī doubtless represents the most systematic attempt to uphold the thesis of the falsified character of the ʿUthmānian version of the Qur’an.¹⁰⁹ If al-Sayyārī’s work, which al-Nūrī draws on and quotes from extensively,¹¹⁰ may be considered the oldest monograph maintaining the Shi’ite notion of the falsification of the Qur’an, al-Nūrī’s work, by thus closing the circle, may be considered the last.

Even though al-Sayyārī was an important figure in the exegetical literature of early Shi’ism, biographical details about him are skimpy. His full name is Abū ʿAbd Allāh Aḥmad b. Muḥammad b. Sayyār. He is said to have served during the imamate of al-Ḥasan al-ʿAskarī as a secretary at the court of the Ṭāhirids (min kuttāb āl Ṭāhir),¹¹¹ whence the title al-kātib which is sometimes accorded him.¹¹² The Ṭāhirids ruled Khorasan on behalf of the Abbasids;¹¹³ al-Sayyārī must therefore have lived for a time in the capital of this region, the city of Nīsābūr/Nīshāpūr. Ibn al-Ghaḍā’irī (who lived in the first half of the fifth/eleventh century) is the sole author of early Imami works of rijāl who refers to our author, calling him “al-Qummī.”¹¹⁴ Al-Sayyārī’s connection with Qumm is discernible, in fact, through the names of certain of his masters and disciples. Hence he would have belonged to the Shi’ite scholarly circles that were quite active in that city.¹¹⁵

None of the Imami sources that have come down to us gives any exact information on the dates of al-Sayyārī’s birth or death. According to the Sunnite author Ibn Ḥajar al-ʿAsqalānī (d. 852/1449), he lived towards the end of the third/ninth century.¹¹⁶ Āghā Buzurg al-Ṭihrānī, citing Ibn Ḥajar’s report, appears to conclude that al-Sayyārī did not live into the fourth/tenth century though he does not without wholly exclude this possibility.¹¹⁷ Another contemporary author, the Iranian scholar Jaʿfar Subḥānī (Sobhani), declares the year 286/899 to have been al-Sayyārī’s death date.¹¹⁸ Since he gives no source, the accuracy of this date remains unproved. Still, it is true that several indications place the latter’s active life in the middle of the third/ninth century. One of the authorities he cites most frequently is, in fact, Muḥammad b. Khālid al-Barqī, who lived at the beginning of this century.¹¹⁹ His son Aḥmad b. Muḥammad al-Barqī transmits from al-Sayyārī in his Kitāb al-maḥāsin.¹²⁰ Further information on the period in which our author lived is furnished by Abū Jaʿfar al-Ṭūsī, who lists him among the disciples of the tenth and eleventh imams, ʿAlī al-Hādī (d. 254/868) and al-Ḥasan al-ʿAskarī (d. 260/873).¹²¹ From a letter (ruqʿa) attributed to the ninth imam Muḥammad al-Jawād (d. 220/835), it turns out that the latter knew al-Sayyārī personally; however, the Shi’ite scholar Muḥammad Taqī al-Tustarī (d. 1415/1995) maintains that the letter was written by al-Ḥasan al-ʿAskarī and that the text is incorrect in this respect.¹²² Al-Tustarī must be right as the text is reminiscent of the writings of the eleventh imam.¹²³ In this letter, written in response to a request, al-ʿAskarī states that al-Sayyārī does not hold the position that he claims to hold and forbids anyone from entrusting anything to him.¹²⁴ This cryptic message seems to indicate that al-Sayyārī had claimed to be the imam’s representative (wakīl), having obtained the right from him to collect money (for example, the khums, i.e., a religious tax of 20 percent levied on the incomes of the faithful).¹²⁵ According to the imam’s message, al-Sayyārī is not a wakīl and so no sum should be disbursed to him.

Our author was severely criticized by certain authors of prosopographical works (rijāl). He is accused mainly of two things. First, of having subscribed to extremist views.¹²⁶ In this regard, his contemporary Muḥammad b. ʿAlī Ibn Maḥbūb al-Ashʿarī in his Kitāb (nawādir) al-muṣannaf accuses him of believing in metempsychosis (tanāsukh).¹²⁷ Many scholars from the city of Qumm during the fourth/tenth century are said to have transmitted from al-Sayyārī only traditions from which any element of “extremism” had been removed. The second accusation is that the latter reported many traditions with incomplete chains of transmission and that in general he is an untrustworthy transmitter.¹²⁸ Nevertheless, despite these harshly critical judgments, a certain number of important authors from Qumm have transmitted from al-Sayyārī directly. Aside from Aḥmad b. Muḥammad al-Barqī, mention may be made of Muḥammad b. al-Ḥasan al-Ṣaffār al-Qummī (d. 290/902–903)¹²⁹ or of ʿAbd Allāh b. Jaʿfar al-Ḥimyarī (d. after 297/909–910).¹³⁰ Moreover, his name appears among the sources of two of the Four Books of the Imamis, namely, the Kāfī of al-Kulaynī and the Tahdhīb al-aḥkām of al-Ṭūsī.¹³¹

According to al-Ṭūsī, al-Sayyārī was the author of many works. He gives the titles of four: (1) Kitāb thawāb al-Qur’an, (2) Kitāb al-ṭibb, (3) Kitāb al-qirā’a, and (4) Kitāb al-nawādir.¹³² Al-Najāshī lists the same titles but with K. al-qirā’āt instead of K. al-qirā’a.¹³³ He mentions another work as well: Kitāb al-ghārāt. All these writings, he says, he has at hand.¹³⁴ Fragments of a work by al-Sayyārī are cited in the final section (entitled Mustaṭrafāt) of the Kitāb al-sarā’ir by Ibn Idrīs al-Ḥillī (d. 598/1202).¹³⁵ To judge from their contents, these fragments come from the K. al-nawādir. Except for these only the Kitāb al-qirā’āt (“The Book of Recitations” or “The Book of Variant Qur’an Readings”) seems to have come down to us.

The first author to cite that book is apparently Abū ʿAbd Allāh Muḥammad b. al-ʿAbbās Ibn Māhyār, known as Ibn al-Juḥām (d. after 328/939–940), the author of a voluminous Qur’an commentary entitled, among other variants, Ta’wīl mā nazala min al-Qur’ān al-karīm fī’l-nabī wa-ālihi (The Exegesis of What Was Revealed in the Noble Qur’an Concerning the Prophet and His Family).¹³⁶ The second of the two volumes of this work, containing exegetical traditions from the sura al-Isrā’ to the end of the Qur’an, was still available to the tenth-/sixteenth-century scholar Sharaf al-Dīn ʿAlī al-Ḥusaynī al-Astarābādī al-Najafī, who cites it copiously in his Ta’wīl al-āyāt al-ẓāhira fī faḍā’il al-ʿitra al-ṭāhira (Exegesis of the Manifest Verses on the Virtues of the Pure Family). These citations contain a respectable number of traditions drawn from al-Sayyārī’s book. Ibn al-Juḥām transmits from the latter through a single intermediary, a certain Aḥmad b. al-Qāsim al-Hamdānī (or Hamad(h)ānī).¹³⁷

During the four centuries after Ibn al-Juḥām’s day, the Kitāb al-qirā’āt seems not to have been cited. The only possible exception may be ʿAlī b. Mūsā Ibn Ṭāwūs (d. 664/1266), renowned for his huge library of Shi’ite texts. In two of his works he mentions the manuscript of an anonymous book entitled “The Book of Qur’anic Commentary Containing the Letter and the Spirit, the Abrogating and the Abrogated, the Clear and the Ambiguous in the Qur’an Along with the Rest of Its Words, Its Virtues and the Recompenses Linked with Them, Reported by Reliable Transmitters from Amongst the Veracious [in the plural, i.e., the holy imams; but the word may also be read in the dual, the Two Veracious, i.e., the fifth and the sixth imams] amongst the Descendants of the Messenger of God.” This manuscript, which belonged to Ibn Ṭāwūs, is described by him as “a magnificent old copy” or, again, as “an old bound Qur’an commentary.” He quotes a tradition taken from this manuscript regarding the interpretation of Qur’an 5:1 and he says that it contains four traditions regarding Qur’an 2:238 as well as two other traditions (probably on the same topic). Now, all of these hadiths may be found in al-Sayyārī’s work; moreover, the title given by Ibn Ṭāwūs corresponds to the contents of that work. It is tempting to suppose that the manuscript of Ibn Ṭāwūs contained a copy of al-Sayyārī’s text, but the information we are given does not allow for a sure identification.¹³⁸ The next author who cites our work is Ḥasan b. Sulaymān al-Ḥillī (fl. 802/1399–1400). His Mukhtaṣar (or Muntakhab) baṣā’ir al-darajāt¹³⁹ has three exegetical traditions (concerning verse 102:8) taken from al-Sayyārī.¹⁴⁰ Al-Ḥillī refers to the latter’s work under the title Kitāb al-tanzīl wa’l-taḥrīf. In classical Arabic literature it is common for the same work to be known under several titles,¹⁴¹ and the fact that in the sources that have come down to us al-Ḥillī is the first person to mention the book so titled is not inevitably an argument against the antiquity of the title. It is even possible to suppose that al-Ḥillī learned the title Kitāb al-tanzīl wa’l-taḥrīf from the second volume of Ibn Juḥām’s Ta’wīl mā nazala min al-Qur’an, which he owned, just as in the later case of al-Najafī, the author of the Ta’wīl al-āyāt al-ẓāhira.¹⁴² If this supposition proves to be true, it would mean that the book was in circulation shortly after al-Sayyārī’s death.

Direct citations from al-Sayyārī’s book are not to be found in the great compilations of exegetical traditions from the Safavid period. Nor is it mentioned directly in the monumental Biḥār al-anwār of Muḥammad Bāqir al-Majlisī from the same period,¹⁴³ and it does not appear in the list of his sources. Nevertheless, it is necessary to point out that after editing the Biḥār al-Majlisī, assisted by some of his students, kept on seeking out other Imami sources with the aim of assembling a supplement to his encyclopedia under the title Mustadrak al-biḥār.¹⁴⁴ Al-Majlisī died before he could edit this supplement, but the list of the sources he had discovered and intended to use was published by al-Nūrī al-Ṭabarsī/Ṭabrisī in his biography of al-Majlisī entitled al-Fayḍ al-qudsī fī tarjamat al-ʿallāma al-Majlisī. The tenth of the forty-eight works which constitute this list is identified as al-Tanzīl wa’l-taḥrīf li-Aḥmad b. Muḥammad al-Sayyārī wa-yuqāl lahu Kitāb al-qirā’āt ayḍan (“Revelation and Falsification [of the Qur’an]” by Aḥmad b. Muḥammad al-Sayyārī, Known as Well [under the title of] “The Book of [Variants] in [Qur’anic] Recitations.”¹⁴⁵

It is possible, therefore, to conclude that from the very beginning of the eighteenth century, at least one manuscript of al-Sayyārī’s work was discovered or rediscovered. Scattered citations drawn from it occur in certain eighteenth- and nineteenth-century works. For example, we might mention the Hāshiyat al-madārik of Muḥammad Bāqir al-Bihbahānī (d. 1206/1791–92 or 1208/1793–94),¹⁴⁶ the Ghanā’im al-ayyām fī masā’il al-ḥalāl wa’l-ḥarām of Abū al-Qāsim al-Jīlānī, known as al-Fāḍil al-Qummī (d. 1231/1815–16 or 1233/1817–18),¹⁴⁷ the Mustanad al-shīʿa of Aḥmad al-Narāqī (d. 1244/1828–29),¹⁴⁸ or, indeed, the Jawāhir al-kalām of Muḥammad Ḥasan al-Najafī (d. 1266/1850).¹⁴⁹ In the last three of these sources,¹⁵⁰ al-Sayyārī’s book is called Kitāb al-qirā’a or Kitāb al-qirā’āt, and the same traditions are reported from it.¹⁵¹

Thanks to al-Nūrī al-Ṭabarsī/Ṭabrisī in the second half of the nineteenth century, al-Sayyārī’s book came fully into its own. It attracted the attention of a large audience because of the fact that al-Nūrī quotes from it so extensively in two of his major works: the Mustadrak al-wasā’il (completed on 10 rabīʿ II 1319/27 July 1901), but prior to that, and most especially, in his Faṣl al-khiṭāb (completed on 12 shawwāl 1298/September 7, 1881).¹⁵²

The hadiths are divided into chapters in accord with the order of the suras. Sometimes—especially with respect to short suras—two or more of the suras are gathered together in the same chapter. More than once a chapter’s title does not actually reflect its contents.¹⁵⁴ Within the chapters the order of the verses does not invariably follow their order in the Qur’an. For example, the verses of Sura 3 (Āl ʿImrān) appear in the following order: 18, 7, 123, 110, 128, 33, 85, 140, 180, 183, 185, 110 (yet again), 81, 103, 92, 102, 97, etc. Moreover, verses belonging to a particular sura are occasionally cited in the chapter before or after that sura’s chapter.

The book has two opening chapters. The first (hadiths 1–17) is untitled and contains traditions in support of the thesis that the Qur’an was revealed in a single unique reading (qirā’a) and that the Qur’an as commonly known has been censored and thus contains omissions. The second (nos. 18–31, bāb mā jā’a fī bism Allāh al-raḥmān al-raḥīm) deals with the basmala (i.e., the formula “In the name of God the Compassionate the Merciful”) with particular emphasis on the duty to recite it aloud.

The characteristically Imami nature of the work becomes clear in a hundred or so exegetical traditions. In these ʿAlī plays a central part: he alone is the Commander of the Faithful (amīr al-mu’minīn; nos. 160, 440), superior to both Moses and Jesus (no. 270). He is the guide of the human race (al-hādī; nos. 233, 270) and he is the way (sabīl; no. 378) and the gateway (bāb) leading to God (nos. 92, 472) with whom he enjoys a position of privilege (no. 481). He is compassionate (no. 570) and he works miracles (no. 684). His jihād is identical to the Prophet’s jihād (nos. 225, 585). He is the sole legitimate successor to the Prophet (no. 162), and it is the believers’ duty to swear loyalty and love (walāya, ḥubb) to him (nos. 51, 88, 539). On Judgment Day, alongside the Prophet, he will hurl God’s enemies into the fire (no. 525). ʿAlī’s sons al-Ḥasan and al-Ḥusayn are descendants of Abraham (nos. 272, 278), and his foreknowledge of their tragic destinies made him ill (no. 279); just like their father before them, the two sons fell victim to the duplicity of the inhabitants of Kūfa (no. 450). Certain verses are interpreted metaphorically as alluding to Fāṭima, al-Ḥasan, al-Ḥusayn (no. 52). or ʿAlī (no. 672). The imams are those who possess authority (ulū ‘l-amr; no. 132); they warn mankind (no. 191), are their saviors (no. 118) and their ambassadors to God (no. 233). They have a perfect knowledge of the Qur’an (no. 451), and their knowledge constantly increases (no. 527). They carry out God’s order (qā’im bi-amr Allāh; no. 406) while awaiting the advent of the eschatological Qā’im (i.e., the Mahdi) (nos. 44, 98, 154, 406, 409, 558, 610, 681). Those who are faithful to ʿAlī (shīʿa) and to the other imams hold a privileged position; they are “those who have understanding” (ulū ‘l-albāb) (nos. 92, 472) and the best of mankind (no. 679); their faults will be forgiven (no. 473), and their places in paradise are assured (nos. 200, 468).

ʿAlī’s enemies are mentioned at several points, sometimes by name but more usually by contemptuous titles and pejorative appellations. Thus, Abū Bakr is “the first” (al-awwal; nos. 359, 383, 474, 522, 616, 617, 621, 660, 672, 684, 698); ʿUmar “the second” (al-thānī; nos. 197, 276, 286, 359, 383, 522, 598, 616, 617, 698); and ʿUthmān is “the third” (al-thālith; nos. 522, 600, 616, 617); elsewhere the two first caliphs are called “Such-a-one and Such-a-one” (fulān wa-fulān; no. 299), or, again, ʿUmar is called al-adlam (no. 380), Rumaʿ (no. 380), and Zufar (nos. 380, 382, 688, 725).¹⁵⁵ ʿĀ’isha is “the Pallid” (Ḥumayrā’; nos. 294, 408, 600). Aside from veiled references to the part she and Ḥafṣa played in the Prophet’s death (nos. 128, 590), her participation in the Battle of the Camel comes in for denunciation as well (nos. 294, 600). Obviously the Umayyads appear in a negative light (nos. 184, 299, 410, 511, 592, 598, 616, 677, 715, 716).

The text alludes to a certain number of Imami doctrines such as badā’ (“divine changeability,” nos. 351, 550), ʿiṣma (“infallibility,” nos. 214, 436, 508, 517), or rajʿa (“the return of the dead to life before the Resurrection,” nos. 113, 295, 303, 490), though the terms themselves are not employed.¹⁵⁶

Even so, a considerable number of exegetical traditions display nothing specifically Imami. For example, they deal with the lives of the prophets (qiṣaṣ al-anbiyā’), abrogated or abrogating verses, explanations of Qur’anic terms, the virtues (faḍā’il) of certain verses or suras.

As the title Kitāb al-qirā’āt indicates, different Qur’anic readings constitute the central subject of the work. In this respect the book differs from other exegetical writings of the pre-Buwayhid school in which the question of the qirā’āt occupies a lesser position. It can be said that approximately two-thirds of the traditions broach this problem. To the best of our knowledge, forty-nine of these readings are attested in no other source.¹⁵⁷

The variant readings can be divided structurally into several categories in accord with the following divergences with respect to the vulgate of ʿUthmān: (1) changes in vocalization, (2) the substitution of one or several words by one or several others, (3) a change in the order of the words, (4) the addition of one or several words, (5) a combination of elements in categories (1) to (4). As for the contents, the readings can be grouped into two main categories: (1) readings containing a specifically Imami message.¹⁵⁸ These qirā’āt, which may be termed “pro-Imami,” can be divided in turn into two types (though the distinction between the two is not always clear): (1a) explicitly pro-Imami readings. These are variants in which one of the following cases is mentioned: ʿAlī (cited by name, whether by one of his epithets such as walī, waṣī, etc.), the imams, the Prophet’s family, walāya, the enemies of the Prophet’s family (by name or by pejorative appellations), terms deriving from Imami law, doctrine, etc. (1b) Implicitly pro-Imami readings. These are variants, the Imami character of which can be recognized only by an initiate or revealed as a result of exegesis. (2) Neutral readings, that is, those that contain no Imami message, neither implicit nor explicit. Readings in category (1a) are almost never cited outside Imami literature (except in polemical contexts), whereas the variants in categories (2) and sometimes (1b) occur as well in non-Imami works.

In the present work the best way to demonstrate a reading that is genuinely Imami entails putting it within the context of a dialogue between the imam and his disciple. For example, the disciple recites (or refers to) a verse in a non-Imami reading; the imam corrects him then or approves it. In other instances, the imam recites a verse following an Imami reading, and when the disciple, who has never heard this reading, questions his master about it, the latter reaffirms its validity. The imam can recite a verse following the Imami reading and require his follower to do the same, or, again, he can forbid the follower from reciting it according to a non-Imami reading. Frequently the disciple reports that he has heard the imam recite a verse in accord with the Imami reading.

An unusual feature of the Kitāb al-qirā’āt is the relatively great frequency of references to the variant readings of Ibn Masʿūd. In Imami works of the pre-Buwayhid period such references are rather rare.¹⁵⁹ Al-Sayyārī, by contrast, retains eleven of them.¹⁶⁰ These are the only readings that do not derive from the imams, thus demonstrating the special status Ibn Masʿūd occupies for our author.

It is not certain that al-Sayyārī’s aim was to encourage his readers to recite the Qur’an in accord with the qirā’āt that he reports.¹⁶¹ Still, he must have regarded it as his special task to record readings deemed reliable by the Imami community.¹⁶²

The Kitāb al-qirā’āt betrays a remarkable affinity with the renowned Tafsīr (Qur’an commentary) of Abū al-Naḍr Muḥammad b. Masʿūd al-ʿAyyāshī.¹⁶³ Only the first part of his work, covering suras 1 to 18, seems to have come down to us. This surviving portion shares some eighty traditions with al-Sayyārī’s work; hence it is reasonable to suppose that the total number of traditions in common between the two books could have been twice as high. Al-ʿAyyāshī’s text as it stands is the kind of version in which the chains of transmission of the hadiths have been abridged and comprise only the name of the first transmitter—and occasionally the second as well—for each tradition.¹⁶⁴ When our two traditionists report the same hadith, and when al-Sayyārī furnishes a complete isnād, it is conceivable that that figured as well in the original version of al-ʿAyyāshī’s text. And so the text of the Kitāb al-qirā’āt may help in reconstructing certain of the latter’s chains of transmission.

Al-ʿAyyāshī lived at the end of the third/ninth and the beginning of the fourth/tenth centuries; hence he must have been younger than al-Sayyārī. He could have gained some portion of his information from the latter, though the reverse hypothesis is not impossible either. Finally it is conceivable too that both authors had a common source.

To recapitulate, each of the three titles under which al-Sayyārī’s work is known relates to a specific aspect of the book. The title al-Tanzīl wa’l-taḥrīf reflects the fact that many of its traditions are based on the Imami belief that the original text of the Qur’an was falsified. A certain number of works with similar titles are mentioned in the sources;¹⁶⁵ but, since none of them survives, nothing can be said of their possible resemblance to al-Sayyārī’s book. The title al-Qirā’āt demonstrates the central position that the question of variant readings occupies in the work. Many works bearing such a title are well known in both Sunni and Shi’ite literature. Lastly, the title al-Tafsīr—attested solely in certain late manuscripts—underlines the fact that the work properly belongs within the genre of Qur’anic exegesis.