world “almost a holy writ, whose reliability was accepted almost without asking questions.”34 The result is a deep and widening gap between Western and Islamic interpretation of both the sīra traditions and the ḥadīth more generally: the methodological skepticism that guides much modern scholarship on these topics is often rejected out of hand by traditional Islamic scholarship and occasionally seen as an attack on Islam itself.
This divide between modern secular and traditional Islamic scholarship over the historical reliability of the sīra and ḥadīth presents an important context for understanding both the nature and intensions of this investigation into early Islamic history. This study and the methods that it employs are in no way aimed at casting doubt on the religious truth of the Islamic tradition. Instead, this book explores a particular aspect of formative Islam from a point of view outside the Islamic tradition, with explicit commitments to the principles of modern, secular historical criticism and the hermeneutics of suspicion rather than fidelity to the traditions and values of the Islamic faith. When approached from this secular perspective, with its specific concerns and commitments, the formative period of Islam will rather obviously look quite different than it does from within the umma. It is important to recognize, however, that both perspectives on Islamic origins are certainly valid, and within their own contexts and communities they are rightly understood to disclose truth. One approach interprets early Islam from the outside, confessing the skepticism of the secular academy, while the other presents a sacred history of formative Islam, a narrative that both shapes and is shaped by the Islamic faith and its community.
Neither perspective then can claim to represent an unbiased account of Islamic origins that somehow is obvious to any objective observer: both understandings are fully intelligible only within the particular interpretive communities that produce them. Moreover, one perspective does not necessarily invalidate the other, and the conclusions of traditional Islamic and modern secular scholarship can both rightly claim to be valid within their own cultural and intellectual contexts. In fact, it is quite possible for an individual to approach a particular issue simultaneously from both a secular point of view and a confessional one, as numerous Western scholars have demonstrated.35 What must be conceded on all sides, however, is that truth depends on the context of an interpretive community, be it religious or secular, and there is no objective truth that will appear as such to every individual and in every cultural context. This approach then does not negate the truth of Islamic accounts of formative Islam: they are in fact true for those whose worldview has been and continues to be shaped most fundamentally by the Islamic tradition. Likewise, those outside the Islamic faith community will not necessarily find Islam’s representation of its own early history to be true in the same way that Muslims do.36 In similar fashion, however, secular knowledge must also recognize the situatedness of its own truth claims: it may only claim to be objective perhaps in the somewhat limited sense that it approaches Islam, for instance, from the outside and thus as an object of study.37
Finally, if some readers may perhaps think it entirely implausible that the Islamic tradition has incorrectly preserved something as significant as the time and place of its founder’s death, a quick glance at formative Christianity is instructive. Undoubtedly many scholars of early Islam will want to persist in maintaining the accuracy of the traditional Islamic accounts of Muhammad’s death and burial, regarding the deviant reports considered in this study as simply misinformed errors coming from those outside of the Islamic community. Yet it is not at all clear why the traditional Islamic narratives of Muhammad’s death should warrant such implicit confidence, particularly in the face of this alternative early tradition. The simple fact that the Islamic accounts were produced by insiders in no way guarantees the accuracy of their information, any more so than one would presume that the Christian gospels accurately record the life and death of Jesus on the basis of their production by insiders. Indeed, to the contrary, it is for this very reason that New Testament scholars are generally suspicious of the gospel accounts, seeking to test them whenever possible by quality evidence drawn from external sources. This sharp contrast with the study of early Islam is seen quite clearly in F. E. Peters’s recent comparative study, Jesus and Muhammad, where the discussion of Jesus begins with evidence from the “pagan” and Jewish sources, while evidence from non-Islamic sources for the beginnings of Islam is rather strangely ignored.38
The earliest extant gospels were written between forty and seventy years after the death of Jesus, based in part on earlier literary sources that had begun to form perhaps some twenty years after his death, a considerably smaller interval than the time elapsed between Muhammad’s death and his earliest biographies. Yet despite the fact that Jesus’ biography took written form more quickly than did Muhammad’s, the gospels have significant disagreements in chronology, including perhaps most famously the differences between the synoptic and Johannine gospels regarding the length of Jesus’ ministry. Likewise, the date of Jesus’ death, for instance, can only be known approximately: 28–33 CE.39 Yet perhaps more comparable with the tradition of Muhammad’s death in Medina are actually the accounts of Jesus’ birth. These reveal that only half a century after Jesus’ death, the early Christians had created a historically improbable tradition of his birth in Bethlehem to serve the needs of Christian salvation history.40 Still more apt is the comparison of Islam’s apostle with early Christian traditions about its apostles. Take, for example, the apostle Peter, whose death and burial are located in Rome by multiple, independent reports written just over a century after the fact: there is even an early tomb identified as the site of this burial. Yet there is considerable debate as to whether Peter was ever even in Rome, and the most recent analysis argues rather persuasively that in fact he was not.41 Likewise, traditions from the second century identify Ephesus as the apostle John’s final resting place, some of which are allegedly based on oral transmission spanning only two generations. Yet the strong consensus of New Testament scholarship rejects the accuracy of these reports.42 If then early Christian traditions concerning Jesus and the apostles could be subject to such manipulation over the course of just a century or even less, how much more so might one expect to find similar developments in the early Islamic biographies of Muhammad, whose contents are widely regarded as highly stylized and untrustworthy.
Such adjustments to a religious tradition’s memory of its early history are in fact not at all unusual and need not be judged as either deceptive or the product of some insidious conspiracy (as some scholars of early Islam have wanted to insist). To the contrary, it is quite common to find that a religious community has revised certain important aspects of its formative history to comport with its most cherished theological principles, as the Christian Nativity traditions bear witness. Often such revisions serve to extend and intensify the interpretive power and cohesion of a religion’s core narrative by incorporating various important religious symbols and practices into the story of its origins. The early Christian gospel writers, like Muhammad’s early biographers presumably, simply were not interested in writing an objective description of past events in the fashion that modern history values. Their narratives urgently seek to communicate the truth about Jesus Christ and the meaning of his life, death, and resurrection: to expect a dispassionate inventory of events would be both anachronistic and absurd. Moreover, the pious fictions of early Christian literature would be wrongly condemned as frauds or deceptions: to the contrary, they undoubtedly were efforts to proclaim the truth, as seen by the authors and their communities, with perfect clarity.43 One would only expect that similar impulses and developments are to be found in the nascent Islamic tradition, and as I will argue, the early Islamic traditions of the end of Muhammad’s life (much like the Christian Nativity traditions) appear to have adapted the arc of his biography to fit the needs of early Islamic identity and salvation history nearly a century after his death. Consequently, our knowledge of exactly when Muhammad died is not nearly as certain as much previous scholarship has assumed, and it seems we must accordingly adjust our historical estimate for the end of his life to sometime more approximately within the period 632–35 CE.
The first chapter of this study examines the various sources from the seventh and eighth centuries that attest to Muhammad’s survival and leadership at the time of the initial assault on the Roman Near East, circa 634–65. Although later sources, particularly from the Christian tradition, continue to repeat this tradition, this chapter focuses on witnesses from the