Читать книгу The Death of a Prophet - Stephen J. Shoemaker - Страница 10

CHAPTER 2

The End of Muhammad’s Life in Early Islamic Memory

The Witness of the Sīra Tradition

Any effort to reconstruct the life of Muhammad and the origins of the religious movement that he founded must confront the difficult problem that there are only a handful of Islamic sources from the early period that convey any information regarding his life—or death, for that matter. Particularly troubling is the complete absence of any accounts from the first Islamic century. While the traditions of the Qurʾān rather probably belong to the first Islamic century, they convey virtually no information concerning the life of Muhammad and the circumstances of his prophetic mission.¹ Admittedly, many of Muhammad’s later biographers claim to relate traditions on the authority of earlier sources, identifying their alleged informants in the chains of transmission, or isnāds, that generally accompany individual traditions about the prophet. Nevertheless, in the Islamic tradition such claims of authenticity through appeal to ancient experts are notoriously unreliable. Isnāds and the ḥadīth (that is, prophetic traditions) that they claim to validate were subject to forgery on a massive scale in early and medieval Islam, as discussed in more detail below, and among the most highly suspect and artificial elements in this system of legitimation are the transmitters named at the earliest stages, that is, the first-century “Companions of the Prophet” and their “Successors.”² Moreover, while some later sources ascribe written biographies of Islam’s prophet to certain renowned authorities from the later first century AH, many other reports offer contradictory testimony, and the balance of the evidence would appear to favor the latter. The issue of writing itself was the subject of considerable controversy in earliest Islam, and even though some more optimistic scholars have accepted at face value such testimonies of early written biographies, there is general consensus against the written transmission of traditions prior to the second Islamic century.³ Despite some hints that early traditionists may have kept written notes for their own personal use, the transmission of knowledge remained almost exclusively oral for more than one hundred years after Muhammad’s death.⁴ ʿUrwa ibn al-Zubayr (d. 712), a renowned early authority on Muhammad’s biography, is among those most frequently alleged to have written a narrative of Muhammad’s life, but most scholars remain deeply skeptical of such reports.⁵

Nevertheless, a small group of researchers has recently attempted to locate certain biographical traditions credibly within the first Islamic century, focusing especially on traditions ascribed to ʿUrwa.⁶ Avoiding the question of whether ʿUrwa actually wrote a biography of Muhammad, these scholars seek to identify ʿUrwa as the author of a corpus of oral tradition that is often assigned to his authority by much later sources. Yet despite a well-developed methodology and some very thorough analyses, their arguments are not persuasive. Indeed, the general failure of this approach to identify a significant corpus of early material presents one of the most troubling problems for efforts to reconstruct the history of primitive Islam on the basis of traditional Islamic sources.⁷ The late formation of the earliest accounts of Islamic origins thus raises significant questions concerning their reliability as historical sources, particularly when they are studied in isolation from other non-Islamic witnesses. Excepting only the decidedly “ahistorical” witness of the Qurʾān, there are essentially no Islamic accounts describing the formation of Islam that can be convincingly dated prior to the turn of the second Islamic century, a circumstance greatly limiting historical-critical investigation of the beginnings of Islam.⁸

The manifold shortcomings of the early Islamic historical tradition, particularly with respect to the period of origins, invite the strong possibility that the beginnings of Islam differed significantly from their representation in the earliest biographies of Muhammad. Not only were the narratives first composed at only an arresting distance from the events that they describe, but modern scholarship on the traditional biographies of Muhammad has repeatedly found them to be unreliable sources. These writings present a highly idealized image of Muhammad and the early community suited to the beliefs and practices of Islam at the beginning of its second century and conformed to a number of literary and theological tendencies. Most importantly, however, the chronology of these narratives has long been recognized as one of the most artificial and unreliable aspects of Muhammad’s canonical biographies, allowing for the real possibility that the sources considered in the previous chapter may indeed preserve an earlier tradition regarding the final years of Muhammad’s life. The traditions of Muhammad’s death contained in the oldest biographies are rather minimal, and in their earliest state they seem to have lacked any specific geographic or chronological context: these elements would appear to have been added only with the composition of the first written biographies around the middle of the eighth century. Consequently, these relatively recent documents cannot exclude the possibility that Muslims of an earlier age may indeed have remembered their prophet as leading his followers as they left Arabia and first entered into the land that had been promised to Abraham and his descendants. To the contrary, their failings as historical sources almost require that we look elsewhere to supplement our knowledge about the beginnings of Islam.

The Earliest Islamic Sources for the Life of Muhammad

The single most important early biography of Muhammad remains the Maghāzī, or Campaigns, of the Prophet by Ibn Isḥāq (d. 767), an account of Islamic origins compiled around the middle of eighth century, approximately 120 years after Muhammad’s death.⁹ Unfortunately, however, Ibn Isḥāq’s biography does not itself survive: it is known only through later recensions, the most important of which are the Sīra, or Life, of the Prophet by Ibn Hishām (d. 833), composed at the beginning of the ninth century, and al-Ṭabarī’s History from the turn of the tenth century. The mediated nature of Ibn Isḥāq’s traditions must constantly be born in mind, particularly inasmuch as Ibn Hishām does not always reproduce Ibn Isḥāq’s biography faithfully but has “abridged and vigorously edited” his source.¹⁰ Nevertheless, through comparison of Ibn Hishām’s transmission with that of al-Ṭabarī and others, it is frequently possible to recover significant amounts of Ibn Isḥāq’s lost biography of Muhammad: when the sources coincide, it is highly likely that the material in question derives from Ibn Isḥāq’s vanished Life. Of the various other early Islamic scholars who were reportedly engaged in the production and transmission of the sīra and maghāzī traditions (the two terms being largely interchangeable in this period), we generally know little more than their names. It would appear that only a handful of these early authorities actually produced written accounts, and with the exception of Ibn Isḥāq’s biography, as mediated primarily by Ibn Hishām’s later redaction, these early documents are witnessed by only a couple of fragments. A papyrus, for instance, has been discovered that relates traditions ascribed to Wahb b. Munabbih (d. 728), and while it remains uncertain whether these traditions actually derive from Wahb, there is no question that this document witnesses to early traditions, inasmuch as the artifact itself is contemporary with Ibn Hishām.¹¹ Unfortunately for present purposes, however, this fragment relates no information concerning the end of Muhammad’s life.¹²

Working backward from the later recensions of Ibn Isḥāq’s biography, one finds Ibn Shihāb al-Zuhrī (d. 742) frequently identified as one of Ibn Isḥāq’s primary sources. Al-Zuhrī was a renowned Medinan authority on the life of Muhammad from the generation immediately prior to Ibn Isḥāq, and on the whole it seems likely that many of the traditions related by Ibn Isḥāq ultimately derive from al-Zuhrī’s teaching, at least in terms of their general content. While it certainly is not at all impossible that later transmissions of Ibn Isḥāq’s sīra have occasionally inserted al-Zuhrī’s name on the basis of his reputation as a great scholar,¹³ the probability that much of Ibn Isḥāq’s information depends on al-Zuhrī seems rather high. In some instances, traditions from al-Zuhrī are further ascribed to ʿUrwa ibn al-Zubayr, and while it is not inconceivable that certain reports about Muhammad took their origin from ʿUrwa’s teaching, this possibility has not been successfully demonstrated and remains highly speculative. It is doubtful that al-Zuhrī himself wrote either a history of early Islam or a biography of its prophet,¹⁴ but several of his students in addition to Ibn Isḥāq composed biographies of Muhammad on the basis of traditions related from al-Zuhrī, the most important of these disciples being Mūsā b. ʿUqba (d. 758) and Maʿmar b. Rāshid (d. 770). Often by correlating traditions independently ascribed to al-Zuhrī in these and other sources with similar reports from Ibn Isḥāq’s Maghāzī, it is possible to establish a measure of probability that al-Zuhrī may in fact have taught some of these traditions to his students.

Unfortunately, however, like Ibn Isḥāq’s lost biography, neither Mūsā’s or Maʿmar’s Maghāzī survives, and we must rely primarily on the evidence of later writers for indirect knowledge of their contents, including especially al-Wāqidī (d. 823) and his disciple Ibn Saʿd (d. 845), as well as al-Ṭabarī (d. 923) and al-Balādhurī (d. 892). The only exception is perhaps a brief fragment purporting to transmit extracts from Mūsā’s Maghāzī, which relates nineteen short and disconnected traditions concerning the life of Muhammad. Nevertheless, the authenticity of this document has been disputed, and given the paucity of its contents, the bulk of Mūsā’s early biography must otherwise be derived indirectly from much later sources.¹⁵ Despite the lack of a similar artifact, the prospects of recovering traditions from Maʿmar’s Maghāzī are in fact much better than for Mūsā’s lost work. Al-Wāqidī’s Maghāzī from the close of the second Islamic century forms a particularly important witness to Maʿmar’s biography, which seems to have served as one of its primary sources. Although al-Wāqidī’s collection is somewhat marred by his occasionally irregular use of isnāds, as well as by the very strong possibility that he has made extensive—and often unacknowledged—use of Ibn Isḥāq’s Maghāzī, al-Wāqidī’s Maghāzī transmits considerable material from Maʿmar, at a chronological distance roughly equivalent to Ibn Hishām’s separation from Ibn Isḥāq.¹⁶ Unlike many earlier maghāzīs, however, al-Wāqidī’s work is true to its title, taking focus on the campaigns of Muhammad during the period from his flight to Medina until his death, an event mentioned only briefly in passing.¹⁷

Al-Wāqidī is reported to have written several other works on Muhammad’s life, including a collection on the Death of the Prophet (Kitāb wafāt al-nabī), but none of these writings is extant.¹⁸ Presumably, many of the traditions from these lost works survive in the biography of Muhammad prepared by al-Wāqidī’s student Ibn Saʿd. In the modern edition of the latter’s Ṭabaqāt, the first two volumes comprise an extensive collection of traditions regarding the life of Muhammad, which seems to have been prepared by Ibn Saʿd himself (as opposed to his students). Although Ibn Saʿd has drawn from a number of authorities in compiling this biography, a large number of its traditions are given on al-Wāqidī’s (and Maʿmar’s) authority, many of which were likely taken from al-Wāqidī’s now lost sīra works. Ibn Saʿd’s collection is thus of particular importance since, as Horovitz notes, he is “the earliest author, after Ibn Isḥāq, from which a complete biography of the Prophet has come down to us.”¹⁹ In contrast to Ibn Isḥāq, however, Ibn Saʿd devoted considerable attention to the end of Muhammad’s life, allotting roughly the last quarter of his biography to traditions concerning his death and burial. Here al-Wāqidī again figures prominently, and while his work on the Death of the Prophet was almost certainly a major source, Ibn Saʿd “has very greatly amplified” al-Wāqidī’s earlier collection.²⁰ Thus, beyond its significance as a likely witness to al-Wāqidī’s lost work on Muhammad’s death, this final section of Ibn Saʿd’s biography of Islam’s prophet preserves the most extensive early Islamic collection of traditions about the end of Muhammad’s life, first assembled nearly two centuries after the events themselves.

Despite the enormous value of both al-Wāqidī’s and Ibn Saʿd’s biographical works, the more recently published Muṣannaf of ʿAbd al-Razzāq al-Sanʿānī (d. 827), a collection of ḥadīth addressing a number of topics, presents a much more promising source for recovering some semblance of Maʿmar’s lost Maghāzī. While this text includes a wealth of biographical traditions ascribed to Maʿmar, its attribution to ʿAbd al-Razzāq remains somewhat controversial, and there are significant unresolved issues regarding its authenticity. Most of this Muṣannaf is known only as transmitted by a somewhat later writer, Isḥāq al-Dabarī (d. 898), who in many respects can be seen as its potential author and furthermore seems to have been much too young to receive its contents directly from ʿAbd al-Razzāq, as alleged.²¹ Nevertheless, Harald Motzki has recently argued that the published edition of ʿAbd al-Razzāq’s Muṣannaf is in some sense “authentic” and can be relied upon as a source of traditions deriving from ʿAbd al-Razzāq in one form or another. Motzki willingly concedes that both Wansbrough and Calder are correct in noting considerable problems concerning the authorship of much early Islamic literature, acknowledging that “if the work [the Muṣannaf] is considered as a book with a definitely fixed text composed by ʿAbd al-Razzāq, the question must be answered in the negative.” But Motzki argues that it was ʿAbd al-Razzāq who “spread the traditions” now compiled in the Muṣannaf, and in this more limited sense his authorship can be largely accepted.²² Thus with some care it may be possible to recover early traditions from this collection, including perhaps many that were originally derived from Maʿmar’s lost biography of Muhammad.

In its modern edition, the fifth volume of this Muṣannaf includes a sizable collection of traditions about Muhammad’s maghāzī, here used in its broader sense to encompass the full span of Muhammad’s life.²³ The overwhelming majority of these biographical traditions are ascribed to Maʿmar, suggesting that this section of the Muṣannaf may very well preserve a selection of traditions drawn from Maʿmar’s Maghāzī. Although ʿAbd al-Razzāq is said to have studied with Maʿmar himself, in light of their considerable difference in age, one wonders if perhaps ʿAbd al-Razzāq has instead relied on a written version of Maʿmar’s biography or some other intermediate source. ʿAbd al-Razzāq is reported to have died roughly fifty-seven years after Maʿmar, and if he studied with Maʿmar for seven or eight years as alleged by the later tradition, Maʿmar must have lived to at least eighty years old, instructing ʿAbd al-Razzāq just prior to his death while in his late seventies: although clearly not impossible, this detail certainly invites some question regarding the nature of ʿAbd al-Razzāq’s relationship with Maʿmar.²⁴ Nevertheless, the prospect that the maghāzī section of the Muṣannaf transmits at least some material from Maʿmar’s now lost work seems likely, particularly in those instances where other early sources can confirm this attribution.²⁵ Moreover, in contrast to al-Wāqidī’s Maghāzī, ʿAbd al-Razzāq’s Muṣannaf includes a short section of traditions related to Muhammad’s death, a collection that adds an important and roughly contemporary supplement to Ibn Isḥāq’s early assemblage of death and funeral traditions.²⁶ In making use of this source, however, it will be essential to bear in mind Motzki’s caution that Ibn Hishām’s Sīra must not be used “as if it were Ibn Isḥāq’s original text,” a warning that applies all the more so to ʿAbd al-Razzāq’s unequaled witness to traditions from Maʿmar’s early biography.²⁷

Two additional early muṣannaf collections, the Muṣannaf of Ibn Abī Shayba (d. 849) and the Ṣaḥīḥ of al-Bukhārī (d. 870), include sections on the maghāzī, and while al-Bukhārī adheres to the more narrow meaning of this term, focusing largely on Muhammad’s campaigns, both authors relate traditions about the end of his life. Like many other sources, these ḥadīth collections often convey traditions not otherwise attested in early Islamic literature, yet given the relatively late origins of all the surviving biographical compilations, it is difficult to assess the historical significance of such isolated traditions. The primary worth of these writings for recovering the earliest Islamic traditions about Muhammad’s life and death tends to be found mainly in their convergences, when they collectively attest to a tradition that may be traced to an early authority, such as Ibn Isḥāq, Maʿmar, or perhaps even al-Zuhrī. Occasionally, however, some less well-attested reports may also be judged as early on the basis of their content, their “matn,” particularly when they run counter to the prevailing doctrinal and literary tendencies of the earliest sources: such dissimilarities suggest an early tradition that has been preserved against these ideological interests on account of its relative antiquity. By carefully sifting these earliest sources according to such principles, it is possible to identify a rather basic account of the end of Muhammad’s life as it was remembered by Medinan scholars of the mid-eighth century.

As will be seen, the resulting sketch of Muhammad’s death and burial is disappointingly meager, and despite the frequent illusion of detailed specificity, the earliest accounts disclose remarkably little about the historical circumstances of Muhammad’s departure from this life. Although repeated attention to concrete details can give these reports a feel of authenticity, their focus on minutia often comes at the expense of broader historical context. As much is in fact typical of the early Islamic historical tradition, whose fragmented, atomistic nature is one of its most characteristic features. By consequence, individual traditions, despite their occasionally remarkable attention to detail, are commonly transmitted without any connection to a broader historical narrative.²⁸ This quality often leaves the sequence of events uncertain, and accordingly, as will be discussed further below, the chronology of the early Islamic historical tradition is widely recognized as one of its most artificial features. Likewise, the narrative detail that occasionally seems to bring these biographical vignettes to life is a common literary device, named by Roland Barthes “the reality effect.”²⁹ With specific regard to the early traditions of Muhammad’s death, Leor Halevi observes that their attention to seemingly trivial details serves “to give the religious narrative a sense of verisimilitude, a certain tangibility that only such casual details could provide.”³⁰ As interesting as such details are for what they reveal about the conceptual world of Muslim believers at the beginning of Islam’s second century, one must take care not to be seduced by these nuances into accepting the veracity of these reports. Rather than validating the accounts in which they occur, they are instead very likely a sign of their literary construction.

Later Biographical Sources: Isnāds, Forgery, and Isnād Criticism

There are, of course, in addition to these early collections, innumerable traditions about the life of Muhammad that survive in only later sources, a great many of which concern his death. One need only consider, for example, the sizable collection of death and burial traditions gathered by Ibn Saʿd in his Ṭabaqāt, the vast majority of which find no parallels in other early Islamic sources.³¹ More recent works, such as Ibn Kathīr’s Sīra, are even more extensive in their knowledge of Muhammad’s life and death: somewhat paradoxically, it would seem that as the distance from Muhammad’s lifetime increased, so too did the Islamic tradition’s knowledge of what he had said and done.³² Each of the biographical traditions in these collections of course bears an isnād vouching for its authenticity, and these chains of transmitters generally conclude with an early authority, such as al-Zuhrī or ʿUrwa, or even ʿĀʾisha or some other Companion of the Prophet, who is identified as the ultimate source of the report in question. In light of the attribution of these reports to such early authorities, one may perhaps wonder why they are not equally valued as witnesses to the life of Muhammad and the history of Islamic origins. Should not these traditions be taken for what they purport to be, namely, reports from the earliest authorities on the beginnings of Islam, including many who were themselves participants in these very events? While there is certainly no reason to exclude the possibility that some early traditions may survive in these later collections, and unquestionably some do, the endemic forgery of ḥadīth and isnāds in medieval Islam means that neither these traditions nor their alleged transmissions can be taken at face value.

Consequently, the countless traditions ascribed to al-Zuhrī and ʿUrwa (among others) by later sources most likely do not reflect actual transmissions so much as the reputation of these two scholars as the earliest and most important authorities on Muhammad and the rise of Islam. Traditions conveying what the community believed to be true about earliest Islam would have been attracted magnetically to their names by sheer virtue of their fame. One need not imagine some sort of conspiracy or even a willful falsification, as some have wrongly maintained, to explain such developments: members of the Islamic community would rather “naturally” have assumed that traditions about the Prophet held to be true must have originated with one of these two sagacious men. As Harris Birkeland comparably observes with regard to Ibn ʿAbbās, whose reputation as a great authority on tafsīr inspired later transmitters to attribute a “great ocean” of exegetical traditions to his authorship, “so it is even today, for instance in traditionalistic, rural communities in Norway. Every accepted religious opinion is attributed to Christ, Paul, or Luther.” He continues to note, perhaps even more tellingly, that “it would provoke great indignation if anybody should happen to express the opinion that Luther ever believed in pre-destination. Every believing peasant would deny that statement most decidedly.”³³ Surely this is not the result of some widespread conspiracy to deceive. Accordingly, one would in fact expect to find that the chains of transmission in the sīra literature regularly ascribe much of their material to ʿUrwa and al-Zuhrī, and consistent attribution of traditions to these early authorities does not necessarily indicate the authenticity of these attributions. It is instead altogether likely that established patterns of authoritative transmission had become fixed according to traditional forms rather early on, and these patterns provided paradigms for the isnāds that were attached to later traditions. Insofar as the Islamic community believed such later traditions to be true, there was not so much a need to invent phony isnāds to justify their authenticity; rather, the “truth” of the traditions themselves would make their attribution to authoritative scholars such as ʿUrwa and al-Zuhrī mostly a foregone conclusion.³⁴

In the face of such concerns, the methods of isnād criticism, especially as developed by Joseph Schacht, G. H. A. Juynboll, and, most recently, Harald Motzki, can often be somewhat helpful for assessing the probability of attributions to such early authorities. Through an extensive correlation of the different isnāds assigned to a particular tradition in later sources, one can occasionally identify a plausible date for the tradition, as well as the individual who was most likely responsible for initially placing it into circulation. The Islamic tradition itself of course has long-established methods of isnād criticism designed to assess the authenticity of the numerous ḥadīth ascribed to Muhammad, the vast majority of which have been regarded as spurious even in the Islamic faith. Yet modern scholarship on Islamic origins generally approaches these chains of transmitters with a great deal more skepticism than the Islamic tradition, and consequently it has developed its own methods for evaluating both the isnāds themselves and the various traditions, or matns, to which they are attached. There is certainly warrant for such suspicion, since forgery of ḥadīth and their isnāds was pandemic in early Islam: the ninth-century Islamic scholar of ḥadīth al-Bukhārī is said to have examined 600,000 traditions attributed to the Prophet by their isnāds, and of these he rejected over 593,000 as later forgeries.³⁵ Matters are even worse in regard to the sīra traditions, which medieval Islamic scholars regarded as having even less historical reliability than the rest of the ḥadīth.³⁶ With good cause, modern scholarship on Islamic origins has merely intensified the Islamic tradition’s own internal skepticism of prophetic traditions in its efforts to reconstruct the beginnings of Islam.

Ignác Goldziher and Schacht after him were among the first Western scholars to draw attention to the artificial and historically problematic nature of very many isnāds that the Islamic tradition viewed as credible, casting considerable doubt on the authenticity of the traditions that these isnāds claimed to validate.³⁷ Schacht, however, developed a method of analysis that allowed for the extraction of historically valuable information from these partially fabricated lists of transmitters. This approach, generally known as common-source analysis, compares all the various isnāds assigned to a particular tradition in different sources in order to identify the earliest transmitter on whom all the highly varied chains of transmission converge, the so-called common link.³⁸ As Schacht rather reasonably concludes, this figure is most likely the person who first placed a particular tradition into circulation, since numerous isnāds all unanimously identify him as a source. Otherwise, it is difficult to explain how these highly variegated chains of transmission could converge on this single individual as their earliest common source. The alternative, that somehow all of these different isnāds have by chance invented the same early transmitter, is comparatively unlikely. Thus some degree of confidence may be placed in identifying the common link with the earliest history of a particular tradition, although as will be seen in a moment, even this seemingly fail-safe method is not without significant problems and uncertainties.

An inherent skepticism pertains to the list of transmitters preceding the common link, however. By definition these figures do not vary in any (or almost any) of the isnāds transmitting a particular tradition, which could on the surface seem to speak for their authenticity. Nevertheless, there are considerable reasons for doubting the historical accuracy of these earliest transmitters, and it seems rather likely that these oldest links in these chains were invented early in the process of transmission in order to give these traditions sanction by linking them with Muhammad and other revered figures from the earliest history of Islam. Particularly important is Schacht’s famous observation that isnāds tend to grow backwards. Schacht has argued rather compellingly that the earliest links of many isnāds, particularly those identifying the Prophet, the Companions, and the Successors as sources, are in fact the most likely to be falsified. Moreover, he concludes that the closer the original source of the tradition is to the Prophet himself, the more likely that the isnād and the tradition itself are counterfeit and late, making traditions ascribed directly to Muhammad both the latest and most likely to be forged.³⁹ Recent studies by several scholars who are otherwise sympathetic to Schacht’s methods have cast significant doubt on his second principle, and the notion that isnāds ascribed to earlier authorities are categorically more likely to be both recent and inauthentic has come into question.⁴⁰ Yet while many of these studies have shown that such traditions are not necessarily more recent than others, they nonetheless generally confirm that their ascriptions to early authorities are overwhelmingly false, verifying the most important aspect of Schacht’s hypothesis. Suspicion of these earliest transmitters is further warranted by the fact that prior to the second Islamic century isnāds usually were not used in the transmission of early Islamic traditions, including the sīra traditions in particular.⁴¹ At this late stage, chains of transmission suddenly had to be constructed, as is evident in Ibn Isḥāq’s use of only a very basic and nascent form of isnāds.⁴² When there was uncertainty regarding a tradition’s origin, which surely was often the case after over a century of anonymous transmission, traditions were ascribed to great figures from the past, and from this chronological distance it seems rather likely that the nearer the isnād approaches to Muhammad, the less likely it is to reflect an actual pattern of historical transmission.

Despite the apparent promise of Schacht’s approach, however, significant unresolved issues remain concerning its reliability, and several recent studies have raised important concerns about the accuracy of common-link analysis for dating early Islamic traditions. The most dramatic challenge to the method has come from an article by Michael Cook, which demonstrates that in certain instances where one can actually test the reliability of common-link analysis through alternate means of dating, the method fails to date material accurately.⁴³ Cook’s study examines several early Islamic eschatological traditions, all of whose dates can be determined from their content, using a rather standard method for dating apocalyptic material. These traditions all purport to predict the future, and up to a certain point they exhibit astonishing accuracy, which is undoubtedly due to the fact that they were written after the events that they correctly predict. Then, suddenly, the author’s prognostic powers fail, and his predictions of the future no longer correspond with the historical record. The point at which this transition occurs reliably indicates the time of the tradition’s composition: here is where its author has truly begun to speculate regarding the future. This moment of the prophetic spirit’s departure can thus be compared with the date of the tradition as determined by common-link analysis of the isnāds, and for each of the three traditions that Cook considers, the common link fails completely as a means of dating. How could such a seemingly well-reasoned method perform so poorly?⁴⁴

The most common explanation for the common link’s failure to provide consistently accurate and reliable dating of early Islamic traditions involves the so-called spread of isnāds during the process of transmission.⁴⁵ As Schacht first recognized, it is altogether likely that these authoritative chains of transmission were altered by the complications of transmission over an extended period of time as well as by the editorial interests of an evolving Islamic tradition. The result is that many isnāds are contaminated and do not preserve an accurate record of historical transmission, particularly in the earliest stages of this process. According to Schacht, the “spread of isnāds” involves “the creation of additional authorities or transmitters for the same doctrine or tradition.” This phenomenon is particularly evident in material ascribed to Successors of the Prophet, and it can often create the illusion that the common link, and thus the tradition itself, circulated earlier than it actually did.⁴⁶

Nevertheless, Motzki and others advocating the reliability of this method have largely rejected out of hand such concerns about any significant spread of isnāds, inasmuch as their approach demands accurate records of transmission. As these scholars seek to mine ever deeper within these transmission histories in hopes of securing traditions even closer to the beginnings of Islam, a much more optimistic view concerning the reliability of these textual genealogies is required, particularly in regard to the early transmitters. While occasionally this approach has convincingly dated certain traditions to the beginnings of the second Islamic century, Motzki often argues aggressively for an even earlier dating, to the first Islamic century. Yet in doing so he generally must engage in special pleading on behalf of early tradents,⁴⁷ and as several critics have noted, these efforts to push certain traditions into the seventh century are methodologically problematic and not very convincing.⁴⁸ Motzki seeks to further enhance these claims of authenticity by raising the stakes and forcing a decision between either accuracy and genuineness or outright forgery and vast conspiracy. If the reliability of these pedigrees is to be doubted, then one must suppose the existence of a widespread and deliberate conspiracy of forgery within the early Islamic community on a scale that is historically improbable.⁴⁹ The rhetorical effect of this position is effectively to shift the burden of proof, requiring any skeptics to account for what is reckoned to be the only alternative to “authenticity,” a grand conspiracy of forgery.

These are not, however, the only two possibilities, as many less sanguine scholars have remarked, and generally one would not want to insist on such a severe bifurcation in analyzing the formative period of a religious tradition.⁵⁰ G. R. Hawting, for instance, has critiqued this falsely posed either/or well in his review of Motzki’s book, the full extent of which is worth quoting:

It seems unlikely that this stark contrast is an adequate view of what is a religious tradition, produced during a relatively long period of social and political disruption when the institutions for safeguarding the transmission were only beginning to be formed, subject to the vicissitudes of a still mainly oral culture, and committed to writing in the form in which we have it at the beginning of the third century of Islam at the earliest. Motzki seems to have little time for the effects of the continuous reworking of the tradition, the introduction of glosses and improvements, the abbreviation and expansion of material, the linking together of reports which originated independently, the adaptation of traditions which originate in one context with a particular purpose so that they may be used in another, let alone simple errors of scribes and narrators. One cannot rule out real forgery but what that might be in a society which revered authority and tradition above independence and innovation is not obvious. Students of the historical tradition (taʾrīkh) have been able to demonstrate the way in which the tradition could be manipulated to give significantly different messages even after it had been recorded in writing (cf., for example, the way in which al-Ṭabarī was used by the later compilers like Ibn al-Athīr). This sort of creative reinterpretation must have been much more possible in the stages before the appearance of written texts.⁵¹

One can in fact identify a variety of interests and tendencies within the early Islamic tradition, as well as certain features of the process of transmission itself, that may have effected the manipulation of isnāds. Michael Cook presents perhaps the most detailed explanation of this phenomenon, and he describes numerous mechanisms by which isnāds likely spread, none of which, it is important to note, involves a grand (or even modest) conspiracy of forgery. On the contrary, Cook identifies several very ordinary events from the process of transmission that likely have introduced the spread of isnāds, and all of these are “thoroughly in accordance with the character and values of the system [of transmission].”⁵² Patricia Crone additionally explains how the rivalries between various centers of early Islamic scholarship (that is, Medina, Mecca, Kūfa, Baṣra, Syria) likely brought about the spread of isnāds in many instances.⁵³ Likewise, Norman Calder’s study of the early Islamic legal tradition identifies still more factors that likely influenced the process of transmission and caused the spread of isnāds. Calder focuses particularly on doctrinal differences as a vector for such changes, and he presents a compelling example from the ḥadīth that clearly evidences the spread of isnāds occasioned by inter-Islamic dogmatic disputes.⁵⁴ Any one or a combination of these factors could easily have inspired adjustments to these records of transmission, introducing distortions that would lead to the identification of false common links and, by consequence, inaccurate datings.⁵⁵ Thus, while the use of common-link analysis to date material may be accepted somewhat provisionally, one must always bear in mind the failures of this method when it has been tested and the potentially deviating effect of the spread of isnāds.⁵⁶ In order to guard against such inaccuracies, this approach can be applied effectively only to traditions bearing an extremely dense pattern of transmission from multiple, intermediate common links, a threshold that few traditions prove capable of meeting. Moreover, while this approach has shown some success in locating a number of traditions at the beginnings of the second Islamic century, for many of the reasons noted above, it has not proven very effective for identifying traditions from the first century with much credibility.⁵⁷

Despite these problems, a small group of scholars has recently applied this method to a selection of sīra traditions, not in an effort to recover early traditions from much later sources, where it may perhaps prove effective, but instead with the intent of securing elements of Muhammad’s biography to figures from the first Islamic century. In particular, they have aimed at exhuming a core of tradition that can be assigned to ʿUrwa ibn al-Zubayr’s authorship, thus fixing the outline of Muhammad’s career to this scholar from the end of the seventh century and the beginning of the eighth. In this way they would attempt to establish the historical accuracy and authenticity of at least some of the basic events from the traditional narrative of Islamic origins. At issue is the general reliability of the early sīra traditions for knowledge of Muhammad’s life and the beginnings of Islam: the historical veracity of these accounts stands very much in question. As already noted, the narrative traditions of Muhammad’s life were rather late in forming, and even the earliest sources, such as they are, can be known only indirectly through more recent transmissions. Accordingly, one must reckon with the fact that during the century that elapsed between the end of Muhammad’s life and the first recoverable narratives of Islamic origins, the Islamic faith almost certainly underwent significant changes in its beliefs and practices. As the chapters to follow will argue, Islam’s transformation during this first century seems in fact to have been considerable, involving the shift from an imminent eschatological belief focused on Jerusalem to become the religion of a global empire with a sacred geography centered on the Hijāz. Such developments were bound to have an effect on Islam’s self-image, including particularly how it recalled its formative period and even perhaps how it remembered the ending of its founder’s life. Indeed, as is widely conceded, the image of Muhammad presented in these early biographies reflects not so much a historical figure from the early seventh century as an idealized portrait of Islam’s founding prophet designed to suit the needs and concerns of eighth- and ninth-century Islam.⁵⁸ Taking the second edition of the Encyclopaedia of Islam as a reflection of the opinio communis, the sīra traditions are here judged as being essentially worthless for reconstructing a historically credible biography of Muhammad or for the history of early Islam more generally.⁵⁹

With this assessment, Muhammad runs the risk of vanishing from history, and taking with him any reliable knowledge concerning the origins of Islam. Against this general consensus, Gregor Schoeler and Andreas Görke, and to a lesser extent Motzki, have applied the methods of isnād criticism to several individual sīra traditions, in the hopes of preventing such an epistemological collapse.⁶⁰ If their analysis is correct, then the “basic framework” of Muhammad’s biography, presumably including at least some of its chronology, may be ascribed to ʿUrwa and perhaps some other early figures. Since ʿUrwa was a nephew of Muhammad’s favorite wife, ʿĀʾisha, as Görke and Schoeler frequently remind their readers, one can safely assume, they would argue, that his account is largely accurate. Although Görke and Schoeler do not include traditions concerning the end of Muhammad’s life among their alleged corpus of ʿUrwan material, their proposal, if correct, would be of some significance for estimating the reliability of these biographical sources. In such a case it would certainly be more difficult, although by no means impossible, to raise significant doubts concerning the accuracy of the traditional Islamic memory of Muhammad’s death. Nevertheless, the approach fails to deliver what its proponents have promised, largely because the biographical traditions generally lack the dense networks required to identify meaningful nodes of transmission, leaving them rather unsuited for this method of analysis. Consequently, Görke and Schoeler’s claims that ʿUrwa may be identified as the author of a significant corpus of sīra traditions are not especially persuasive. Ultimately their investigations do little to advance our knowledge of the sīra traditions beyond what may already be determined from Ibn Isḥāq’s Maghāzī and other early sources.

For example, Motzki applies this isnād-critical approach to a tradition in which Muhammad orders the assassination of a Jewish opponent, Ibn Abī l-Ḥuqayq, and while he convincingly assigns the tale to al-Zuhrī, his efforts to identify an earlier source are not persuasive.⁶¹ To do so, he must conflate two traditions that in fact appear to be quite distinct and ignore the deeply problematic nature of one of his tradents, Abū Isḥāq.⁶² Schoeler makes a similar analysis of the traditions of the beginnings of Muhammad’s revelations (the iqraʾ episode) and the rumors that ʿĀʾisha had committed adultery (ḥadīth al-ifk),⁶³ while Görke has investigated the reports of Muhammad’s treaty at al-Ḥudaybiya.⁶⁴ Görke and Schoeler have also published together a very brief article on an extensive tradition complex purportedly associated with the events of Muhammad’s hijra.⁶⁵ In each instance they attempt to identify these traditions with ʿUrwa, whose biography of Muhammad they aim to reconstruct using the methods of common-link analysis.⁶⁶ While al-Zuhrī and occasionally other authorities of his generation can be persuasively linked with these traditions, the reach back to ʿUrwa is generally not convincing. Their arguments often require a great deal of optimism regarding the accuracy of certain isnāds and an occasional willingness to accept hypothetically reconstructed lines of transmission. In the case of the complex of traditions linked with the hijra, for instance, a large body of material transmitted by only a single source is identified as genuine, while isnāds belonging to only specific parts of the alleged tradition complex are occasionally represented as authenticating the entire block of material.⁶⁷

Görke and Schoeler are most successful in arguing that the traditions of Muhammad’s experience of visions and voices at the onset of his revelations and a basic narrative of his flight to Medina in the face of opposition had begun to circulate at the end of the seventh century. Likewise, the story of ʿĀʾisha’s suspected adultery and her subsequent acquittal is persuasively dated to this period through the study of its isnāds. Yet one should recognize just how meager these results are, particularly given the amount of effort involved. Even if all the methodological questions regarding such an isnād-critical approach to the sīra traditions are placed to the side, the resultant biography of Muhammad is disappointingly minimal. Motzki himself ultimately expresses some doubt whether “the outcome will justify the time and energy needed for such an enterprise,” and he forecasts that “the historical biography which will be the outcome of all these source-critical efforts will be only a very small one.”⁶⁸

Perhaps even more important is the failure so far of this arduous method to reveal anything particularly new about the historical Muhammad that could not already be determined using simpler approaches. For instance, there can be little doubt that the early Muslims believed that Muhammad had been the recipient of divine revelation, and its representation as a vision of light and auditions merely reflects a well-established biblical pattern.⁶⁹ Moreover, dating according to the hijra is attested by early documentary sources, signaling the importance of a tradition of Muhammad’s “flight” for the early Muslims.⁷⁰ The accusations against ʿĀʾisha are also credibly early, inasmuch as they reflect negatively on a figure who later came to be revered as the “mother of the faithful,” and one would thus imagine that the story had begun to circulate before ʿĀʾisha had attained this status in Sunni piety.⁷¹ Even if one were to accept the more problematic arguments presented on behalf of the traditions of al-Ḥudaybiya and Ibn Abī l-Ḥuqayq’s murder, ultimately very little is added to our portrait of Muhammad. It is certainly believable that Muhammad may have concluded an unfavorable treaty regarding fugitives or ordered the assassination of an opponent, as reported in these accounts. But these traditions reveal almost nothing about the nature of Muhammad’s religious movement and its early history. In these areas the sīra traditions remain not only unproven but suspect, leaving modern scholars with the difficult choice of either taking these biographies more or less at face value or looking elsewhere for more reliable evidence of primitive Islam. Such are the circumstances one must face in evaluating the early Islamic traditions of the end of Muhammad’s life.⁷²

Ibn Isḥāq’s Account of Muhammad’s Death and Burial

Taking Ibn Isḥāq’s early biography as a basis, we gain a clear sense of how the Muslims of the mid-eighth century imagined the death and burial of their founding prophet and what they thought was important to “remember” about these events.⁷³ The story begins just as a band of soldiers under Usāma b. Zayd’s leadership is dispatched to attack Syria, more specifically the region of Transjordan and the coastal plain of Palestine, in a report given without attribution according to Ibn Hishām’s transmission.⁷⁴ Then, suddenly Muhammad became ill after returning home from Medina’s graveyard, where he had offered prayers for the dead, an act that foreshadows the prayers offered over his own grave after his burial.⁷⁵ According to one account, Muhammad was posed with a choice at the cemetery, presumably by God, who offered him either “the keys of the treasuries of this world and long life here followed by Paradise,”⁷⁶ or the chance to meet the Lord in Paradise at once. As Ibn Isḥāq later explains, Muhammad often said that “God never takes a prophet to Himself without giving him the choice.” Deciding for the latter option, Muhammad returned home to ʿĀʾisha, and then while making rounds among his wives, he suddenly fell ill in the house of Maymūna. Muhammad asked his wives for their permission to return to ʿĀʾisha’s house and be cared for there by her, and when they agreed he was taken to ʿĀʾisha and spent his final days with her. At Muhammad’s request, she placed him in a tub, and together with al-Faḍl b. al-ʿAbbās and ʿAlī, she poured “seven skins of water from different wells” over him until he cried “Enough, enough!” As ʿAlī left Muhammad’s house, al-ʿAbbās warned him that Muhammad would soon die, and ʿAlī would find himself “a slave.” Al-ʿAbbās suggested that they should go to Muhammad and ask him either to declare them as his successors or, if he had chosen someone else, “to enjoin the people to treat us well.” Thereafter Muhammad went and “sat in the pulpit,” revealing the choice that he was offered as well as his decision. Abū Bakr expressed alarm at the news, but Muhammad reassured him and underscored their unique bond of friendship, directing that all the doors to the mosque should be closed except for the one from Abū Bakr’s house. According to Ibn Hishām (but not al-Ṭabarī), Muhammad also took the occasion to encourage the people to join in Usāma’s expedition to Palestine, ordering that it be dispatched immediately.

When Muhammad returned home, his illness intensified, and he lost consciousness. His wives agreed to administer a medicine that had been brought from Ethiopia, and once Muhammad awoke, he was irritated and demanded to know who had forced the medicine upon him. They explained that they were afraid that he would develop pleurisy without the medicine. Muhammad protested that God would never afflict him with such a shameful disorder, and as punishment, he forced each of his wives to take the medicine themselves. Ibn Hishām then relates several stories in which Muhammad declares his preference that Abū Bakr should lead the community in prayers in his stead, some of which insist quite deliberately that Abū Bakr, rather than ʿUmar, was to fill this role. Although al-Ṭabarī also reports two similar traditions, he fails to do so on Ibn Isḥāq’s authority, raising the question of whether these endorsements of Abū Bakr appeared in Ibn Isḥāq biography.⁷⁷ Nevertheless, the appearance of one of these traditions in al-Balādhurī’s Ansāb al-ashrāf on Ibn Isḥāq’s authority perhaps confirms its place in his Maghāzī.⁷⁸ A pair of related traditions further note that Muhammad peeked into the mosque while Abū Bakr was leading the prayers and was seen by the people one last time: according to one tradition Muhammad sat beside Abū Bakr as he led the prayers, concluding with an admonition to adhere strictly to the Qurʾān and to it alone, laying nothing to his charge. Abū Bakr and Muhammad returned to their houses, and Muhammad laid his head on ʿĀʾisha’s bosom. When someone from Abū Bakr’s family brought a toothpick (siwāk), ʿĀʾisha offered it to Muhammad and “chewed it for him to soften it and gave it to him. He rubbed his teeth with it more energetically than [she] had ever seen him rub before.” Then, after this final act of oral hygiene, Muhammad cried out, “Nay, the most Exalted Companion is of Paradise,” signaling his resolve to depart from this world, and he expired in ʿĀʾisha’s arms. A rather peculiar story then follows, in which ʿUmar refuses to believe that Muhammad has died, insisting that, like Moses, he had ascended to God only temporarily and would soon return. Although more will be said about this intriguing episode especially in the following chapter, Abū Bakr arrives from his house and silences ʿUmar by citing a Qurʾānic verse predicting Muhammad’s death. Astonishingly, however, Ibn Isḥāq reports that no one had ever heard that verse before Abū Bakr recited it at that very moment.

Muhammad’s burial is then deferred by the ensuing struggle over who was to succeed him as the community’s new leader. In a gathering at the hall (saqīfa) of the Banū Saʿīda, the prominent men of the community jockeyed with one another to determine Muhammad’s successor, ultimately choosing Abū Bakr, who served as the first caliph. In the transition then to Muhammad’s burial, ʿUmar twice offers apologies for his frenzied denials of Muhammad’s death, one given immediately after the saqīfa meetings and a second ascribed to ʿUmar during the time of his own caliphate. In both accounts ʿUmar explains his behavior as a result of his firm belief that Muhammad would remain alive and leading his people until the arrival of the eschatological Hour. ʿAlī, it would appear, remained behind while Abū Bakr and the others contended over the caliphate, attending to Muhammad’s body and preparing it for the grave. Assisted by al-ʿAbbās and his sons al-Faḍl and Qutham, as well as Usāma b. Zayd and Shuqrān, one of Muhammad’s freedmen, ʿAlī washed Muhammad’s body. When they could not decide whether or not to remove Muhammad’s clothing before washing his corpse, divine intervention made clear that he should remain clothed. Following the washing, Muhammad’s body was wrapped in three garments, and two gravediggers, a Meccan emigrant and a Medinan, prepared his grave in the characteristic Medinan style, with a distinctive niche. A dispute arose over Muhammad’s place of burial that was resolved by Abū Bakr, who recalled Muhammad as having said, “No prophet dies but he is buried where he died.” Thus the grave was dug immediately beneath his bed, in ʿĀʾisha’s house, and the people came and began to pray over Muhammad. ʿAlī, the sons of al-ʿAbbās, al-Faḍl and Qutham, descended into the grave, as did Shuqrān and al-Mughīra b. Shuʿba, who purposefully dropped his ring in the grave as a ruse to allow him to descend and embrace Muhammad’s body one final time. The scene draws to a close with recollections of Muhammad’s censure against those “who choose the graves of their prophets as mosques” and his final injunction to eliminate all religions other than Islam from the Arabian Peninsula.

While the broader context of these events is somewhat obscured by the episodic and disconnected nature of the individual ḥadīth, it is clear that they collectively relate Muhammad’s death in an urban setting, which is easily recognizable as the Medina of Muslim tradition. Moreover, Ibn Isḥāq’s presentation of these events within the sequence of his collection locates them before the full-scale assault on Palestine had begun, although Usāma’s expedition to Palestine just before Muhammad’s illness and death presents an intriguing anomaly to be addressed later in this chapter. Muhammad’s death seems to follow closely on his “farewell pilgrimage” to Mecca, which Ibn Isḥāq appears to locate in the year 10 AH.⁷⁹ Yet nothing in the death and burial traditions themselves specifies such timing, and the reports alone offer no clear indication of when Muhammad died, either in relative or absolute terms: this information must be derived from Ibn Isḥāq’s arrangement.⁸⁰ His Maghāzī is the first witness to this chronology, and while there is no basis for concluding that the sequence is entirely Ibn Isḥāq’s invention, the reports that he has gathered fail to present any evidence of its existence prior to his collection. The brief fragment purporting to relate selected traditions from Mūsā ibn ʿUqba’s Maghāzī affords no confirmation of his ordering of events, inasmuch as these extracts contain nothing relevant to the end of Muhammad’s life.⁸¹ ʿAbd al-Razzāq’s Muṣannaf presents a relative chronology similar to Ibn Isḥāq’s, but there is likewise no indication that his sequence reflects any earlier source. As is the case with so much of our information concerning the origins of Islam, it is not possible to date this chronology of Muhammad’s death before the beginning of the second Islamic century. Perhaps Ibn Isḥāq inherited this schema from al-Zuhrī, but there is no evidence to indicate this. In any case, the received chronology of Muhammad’s death in the Islamic historical tradition cannot be shown to have existed prior to the middle of the eighth century, over a century after the events in question took place.

Muhammad’s Death According to al-Zuhrī

If this chronology is first witnessed only by Ibn Isḥāq, there are a number of death and burial traditions that can, with some measure of credibility, perhaps be linked with al-Zuhrī’s teaching.⁸² For instance, several other early sources link al-Zuhrī with the report of the sudden onset of Muhammad’s illness while visiting his wives, in the house of Maymūna, after which his wives gave permission for him to be nursed in ʿĀʾisha’s house, where al-Faḍl b. ʿAbbās and ʿAlī assisted her as they poured water from seven wells over him. An account of these events almost identical to Ibn Isḥāq’s is ascribed to al-Zuhrī through different channels in ʿAbd al-Razzāq’s Muṣannaf, Ibn Saʿd’s Ṭabaqāt, and al-Bukhārī’s Ṣaḥīḥ.⁸³ The convergence of these transmissions on al-Zuhrī suggests a likelihood that the tradition originated with his teaching. Likewise Muhammad’s preaching in the mosque during his illness, in which he praises Abū Bakr as his closest friend and orders all the doors of the mosque closed except for Abū Bakr’s, is also ascribed to al-Zuhrī by ʿAbd al-Razzāq and Ibn Saʿd.⁸⁴ Muhammad’s statement that “God never takes a prophet without offering him a choice” is imputed to al-Zuhrī by al-Bukhārī and Ibn Saʿd, as well as by a collection of traditions from al-Zuhrī surviving on a papyrus of the early ninth century.⁸⁵ The basic elements of the Ethiopian medicine story are placed under al-Zuhrī’s authority by ʿAbd al-Razzāq and Ibn Saʿd,⁸⁶ and both identify al-Zuhrī as having circulated Muhammad’s command to establish Islam as the only faith in the Arabian Peninsula.⁸⁷ Al-Bukhārī, ʿAbd al-Razzāq, and Ibn Saʿd all attribute Muhammad’s denunciation of those who make the graves of their prophets into places of worship to al-Zuhrī,⁸⁸ and all three impute to him the traditions concerning Muhammad’s appointment of Abū Bakr (rather than ʿUmar) as the community’s new prayer leader.⁸⁹ The tradition of Muhammad peering into the mosque while Abū Bakr led the prayers on the day of his death was also known to ʿAbd al-Razzāq, al-Bukhārī, Ibn Saʿd, and al-Balādhurī from al-Zuhrī.⁹⁰ The conversation between ʿAlī and al-ʿAbbās concerning their status after Muhammad’s death is also widely attested on al-Zuhrī’s authority, appearing in ʿAbd al-Razzāq, Ibn Saʿd, and al-Bukhārī, as well as in al-Ṭabarī’s History, where he cites the ḥadīth both from Ibn Isḥāq and according to a second, independent line of transmission from al-Zuhrī.⁹¹ Finally, al-Zuhrī is credited with teaching the story of Muhammad and the toothpick by Ibn Saʿd, although ʿAbd al-Razzāq records only Muhammad’s final words from this scene, “with the most Exalted Companion!,” an exclamation that Ibn Saʿd and the ninth-century al-Zuhrī papyrus instead link with the tradition of Muhammad’s choice.⁹²

The same sources also agree in assigning to al-Zuhrī the story of ʿUmar’s refusal to accept Muhammad’s death and his correction by Abū Bakr, who persuaded ʿUmar and the others that Muhammad had indeed died through the recitation of a Qurʾānic verse that no one had ever heard before. Versions almost identical to Ibn Isḥāq’s account appear in ʿAbd al-Razzāq, Ibn Saʿd, and al-Bukhārī.⁹³ A shorter version, which relates only ʿUmar’s protests, absent any rebuttal from either Abū Bakr or the Qurʾān, is transmitted from al-Zuhrī through different channels by Ibn Abī Shayba and Ibn Saʿd, as well as by ʿAbd al-Razzāq.⁹⁴ Likewise, ʿUmar’s initial apology for his actions in the wake of the saqīfa meetings is also ascribed to al-Zuhrī by ʿAbd al-Razzāq and Ibn Saʿd, while al-Balādhurī credits al-Zuhrī with ʿUmar’s second explanation, allegedly given while he was caliph.⁹⁵ Presumably, some version of this story, at the very least in its shortened form, and ʿUmar’s subsequent apologies belonged to al-Zuhrī’s teaching. The nature of the story itself suggests a particularly early origin: it seems improbable that Muslims of the early second century or later would have invented such a strange tale, involving ʿUmar’s violent denials of Muhammad’s death. ʿUmar’s mistaken rant casts this “rightly guided” caliph in a rather unfavorable light, and it does not fit with the tendencies of the early Islamic historical tradition. Accordingly, this tradition’s preservation is a likely token of its early formation: its transmission by al-Zuhrī despite its awkwardness is most likely a consequence of the story’s well-established status in the community’s historical memory already by his time. As much is equally if not more true of Abū Bakr’s Qurʾānic correction, particularly in light of the crowd’s alleged ignorance concerning the recited passage. It is hard to imagine the invention of a tradition that so pointedly raises the question of the Qurʾān’s integrity during the mid-eighth century.

There is, however, another account of ʿUmar’s denial ascribed to Ibn ʿAbbās that appears to be even older than the al-Zuhrī version, a report that, although absent from Ibn Isḥāq’s Maghāzī, is witnessed by ʿAbd al-Razzāq, Ibn Saʿd, and al-Balādhurī.⁹⁶ According to this tradition, it was al-ʿAbbās, rather than Abū Bakr, who opposed ʿUmar’s ravings, countering them not with a Qurʾānic proof-text but instead with the observation that Muhammad’s body had begun to stink. As Wilferd Madelung has argued, the chronology of Muhammad’s burial in relation to the saqīfa meeting in this account favors the antiquity of the al-ʿAbbās version.⁹⁷ Moreover, failure to make recourse to the Qurʾān also strongly suggests its priority: it is difficult to account for the subsequent invention of a tradition that so inelegantly argues for Muhammad’s mortality on the basis of his pungent corpse if Abū Bakr’s Qurʾānic riposte was already in circulation. Al-ʿAbbās’s complaints of Muhammad’s stench are in fact seemingly belied by a widely circulated tradition from Ibn Isḥāq’s collection that underscores the exceptional nature of Muhammad’s body in death as well as life: such a body presumably would not stink so offensively immediately after dying.⁹⁸ Indeed, the sweet fragrance of Muhammad’s incorruptible body after death is a frequent theme of the ḥadīth that seems to have developed over the course of the eighth century through influence from the Christian hagiographical tradition.⁹⁹ Moreover, the earliest Christian accounts of Muhammad’s death and burial also note that his corpse began to stink when his followers did not bury him soon after his death: accordingly it would appear that these narratives show an awareness of the early Islamic tradition regarding ʿUmar and al-ʿAbbās, as well as the reported delay in burying Muhammad’s festering body.¹⁰⁰ Consequently, a tradition indicating the initial denial of Muhammad’s death by at least some within the earliest Islamic community not only can be traced back to al-Zuhrī’s teaching, but there is evidence of an even older version of the story that ultimately seems to have required the invention of a Qurʾānic rebuttal to silence the protests of ʿUmar—and presumably others as well. These reports of a controversy surrounding the reality of Muhammad’s death reflect perhaps the earliest extant Islamic traditions about the end of Muhammad’s life, and as we will see in the following chapter, they appear to be intimately linked with the imminent eschatological expectations of Muhammad and his earliest followers.

ʿAbd al-Razzāq and Ibn Saʿd also ascribe to al-Zuhrī a tradition that Muhammad sought to write something down just before his death, a report that Ibn Isḥāq has possibly suppressed.¹⁰¹ As Muhammad’s illness grew worse, he asked for something to write on, in order to leave behind a document that would prevent his followers from going astray. ʿUmar opposed the request, suggesting that Muhammad’s illness was clouding his judgment and that the existence of the Qurʾān obviated the need for any additional document to guide the community. Others, however, began to argue that Muhammad should be given something to write with. When the ensuing noise and confusion eventually began to disturb Muhammad, he dismissed the throng and ultimately failed to produce a document. While it is certainly possible that al-Zuhrī taught something of this nature, the absence of any ascription to al-Zuhrī independent of Maʿmar suggests that possibly the latter is its author. Nevertheless, in light of the controversies surrounding the issue of writing in earliest Islam, as noted above, as well as the politically volatile nature of the tradition with regard to issues of succession to Muhammad, it is certainly conceivable that Ibn Isḥāq may have chosen to omit the story from his collection.

As for the washing of Muhammad’s corpse and his burial, Ibn Isḥāq’s account of these events largely departs from al-Zuhrī’s authority, ascribing its dozen or so reports mainly to other traditionists. Other early collections, such as al-Wāqidī’s Maghāzī, Mālik’s Muwaṭṭaʾ, al-Bukhārī’s Ṣaḥīḥ, Muslim’s Ṣaḥīḥ, and Abū Dāʾūd’s Sunan, show little interest in the details of Muhammad’s burial, and al-Ṭabarī’s History merely reproduces Ibn Isḥāq’s rather meager assemblage of funeral traditions, which appears to be the earliest such compilation. The reticence of these early sources on this topic suggests that perhaps Muhammad’s burial did not arouse the interests of his earliest biographers until a relatively later date, and Ibn Isḥāq’s shift away from al-Zuhrī at this point seems to signal that the latter did not concern himself particularly with this subject. One wonders if perhaps this early silence is somehow related to the tradition from the East Syrian Baḥīrā legend that Muhammad’s followers knew nothing about his grave.¹⁰² Nevertheless, the eventual proliferation of traditions about Muhammad’s funeral can be witnessed especially in Ibn Saʿd’s Ṭabaqāt, as well as to a lesser extent by ʿAbd al-Razzāq, Ibn Abī Shayba, and al-Balādhurī.¹⁰³ Although Ibn Saʿd ascribes a significant number of funeral traditions to al-Zuhrī, the absence of any parallel transmissions from either earlier or contemporary collections makes it extremely difficult to judge the accuracy of these attributions. It may well be that as traditions about Muhammad’s burial began to develop, they were attracted to al-Zuhrī’s name and assigned to him largely on the basis of his reputation as an authority on Muhammad’s biography.

Of the burial traditions gathered by Ibn Isḥāq, only a single report is given on al-Zuhrī’s authority, a notice that after Muhammad’s corpse had been washed, it was wrapped in three garments, “two of Ṣuḥār make, and a striped mantle wrapped one over the other.”¹⁰⁴ Ibn Saʿd, ʿAbd al-Razzāq, and al-Balādhurī report a similar tradition on al-Zuhrī’s authority, indicating that Muhammad was buried in three pieces of cloth, two white and one striped.¹⁰⁵ The early ninth-century papyrus also ascribes to al-Zuhrī a tradition that Muhammad was buried in a striped woolen garment, adding further credence to the possibility that Ibn Isḥāq inherited such information from him.¹⁰⁶ Nevertheless, the early Islamic traditions about Muhammad’s burial clothes vary widely, and as Halevi observes, these reports reveal a great deal more about the culture of burial in early Islam than they do about actual events from the early seventh century.¹⁰⁷ Ibn Saʿd, ʿAbd al-Razzāq, and al-Balādhurī also ascribe ʿAlī’s exclamation while washing Muhammad’s corpse (“you were excellent in life and in death”) to al-Zuhrī,¹⁰⁸ while Ibn Abī Shayba joins these three collectors in ascribing to al-Zuhrī the traditions that ʿAlī, al-ʿAbbās, al-Faḍl, and Ṣāliḥ (that is, Shuqrān) participated in Muhammad’s burial.¹⁰⁹ ʿAbd al-Razzāq and Ibn Abī Shayba further relate here the installation of a brick monument to mark the location of Muhammad’s grave. Nevertheless, in both instances the report is given on Maʿmar’s authority, and thus the attribution to al-Zuhrī is somewhat questionable.¹¹⁰ Muhammad’s statement that “no prophet dies but he is buried where he died,” reported by Abū Bakr, and the ensuing decision to bury him beneath his bed are ascribed by ʿAbd al-Razzāq, Ibn Abī Shayba, and al-Balādhurī not to al-Zuhrī but to one of Ibn Isḥāq’s contemporaries, Ibn Jurayj (d. 767).¹¹¹ ʿAbd al-Razzāq likewise joins Ibn Isḥāq in assigning the tradition of Muhammad’s burial in the middle of the night between Tuesday and Wednesday to another contemporary traditionist, ʿAbdallāh b. Abī Bakr (d. 753).¹¹²

One has the sense then that with perhaps the exception of Muhammad’s burial in three garments, Ibn Isḥāq (and possibly Maʿmar as well) is collecting these funeral traditions for the first time, and on the whole, his account of Muhammad’s burial consists of what appear to be “idealized memories of Muḥammad” aimed at normalizing Islamic funeral practices and distinguishing them from the practices of their non-Islamic neighbors.¹¹³ Moreover, both these burial traditions and the traditions of Muhammad’s illness and death are heavily overlaid by the political and sectarian struggles of early Islam that ensued immediately after Muhammad’s death. As both Madelung and Halevi observe, the cast of characters and their various roles in Muhammad’s death and burial are designed to bolster the claims of one party or the other in the contest for authority within the earliest community.¹¹⁴ Muhammad’s sickbed provided, as Juynboll remarks, a frequent topos for the expression of these and other interests.¹¹⁵ Thus, many of the details from these accounts should be viewed as governed by such ideological concerns, rather than reflecting actual historical events.

Yet these observations aside, Ibn Isḥāq transmits a mosaic of traditions from al-Zuhrī that seem to envision Muhammad’s death within an urban context, where his wives live in separate dwellings and the faithful gather regularly in a central mosque for prayers. In contrast to the implied witness of the non-Islamic sources, Muhammad does not appear to have been out on campaign when he suddenly became ill and died; rather his death is situated within a thoroughly domestic setting, where Muhammad is surrounded by the constant care and attention of his friends and family. While this backdrop certainly bears a credible resemblance to the Medina of Islamic tradition, the city itself is never named in the death and burial traditions ascribed to al-Zuhrī. Is it then possible that the later tradition has supplied this location and its urban ambiance as the setting for Muhammad’s departure from this world, transferring these events from an original context somewhere outside the Ḥijāz? Could it be that the early Muslims had re-remembered the circumstances of Muhammad’s passing so dramatically just under a century after the event itself? There are in fact reasons to suspect a possible relocation of Muhammad’s death to Medina, but their consideration must be deferred until a later chapter. The remainder of this chapter will instead examine the historical reliability of the sīra tradition more broadly, focusing especially in the issue of its chronology. As it turns out, the chronology of Muhammad’s life is one of the most artificial and unreliable features of these early biographies, a point that is widely conceded by modern scholarship and even, to a certain extent, by the Islamic tradition itself. Moreover, modern scholarship has identified a variety of literary tendencies that have markedly shaped Muhammad’s traditional biography, and some of these may have influenced the early Islamic memory of his death, determining certain aspects of both its timing and location. Finally, we will consider certain passages from Ibn Isḥāq’s Maghāzī that seem to link Muhammad with a military campaign in Palestine during the final years of his life. These textual anomalies may possibly reveal traces of an older Islamic tradition concerning the end of Muhammad’s life that would comport with the witness of the non-Islamic sources.

Sīra Chronology and Its Reliability