Читать книгу The Death of a Prophet - Stephen J. Shoemaker - Страница 9

CHAPTER 1

“A Prophet Has Appeared, Coming with the Saracens”

Muhammad’s Leadership during the Conquest of Palestine According to Seventh- and Eighth-Century Sources

At least eleven sources from the seventh and eighth centuries indicate in varied fashion that Muhammad was still alive at the time of the Palestinian conquest, leading his followers into the Holy Land some two to three years after he is supposed to have died in Medina according to traditional Islamic accounts. As will be seen, not all of these witnesses attest to Muhammad’s leadership with the same detail: some are quite specific in describing his involvement in the campaign itself, while others merely note his continued leadership of the “Saracens” at this time. When taken collectively, however, their witness to a tradition that Muhammad was alive at the time of the Near Eastern conquests and continuing to lead his followers seems unmistakable. The unanimity of these sources, as well as the failure of any source to contradict this tradition prior to the emergence of the first Islamic biographies of Muhammad beginning in the mid-eighth century, speaks highly in their favor. In fact, no source outside the Islamic tradition “accurately” reports Muhammad’s death in Medina before the invasion of Palestine until the early ninth-century Chronicle of Theophanes, a text that shows evidence of direct influence from the early Islamic historical tradition on this point as well as others.

It would appear that this tradition of Muhammad’s continued vitality and leadership during the campaign in Palestine circulated widely in the seventh- and eighth-century Near East. Although the majority of the relevant sources are of Christian origin, collectively they reflect the religious diversity of the early medieval Near East, including witnesses from each of the major Christian communities as well as a Jewish, a Samaritan, and even an Islamic witness to this discordant tradition. This confessional diversity is particularly significant, insofar as it demonstrates the relative independence of these accounts and the diffusion of this information across both geographic distance and sectarian boundaries. Indeed, the multiple independent attestation of this tradition in a variety of different sources demands that we take seriously the possibility that these eleven sources bear witness to a very early tradition about Muhammad. Presumably, it was a tradition coming from the early Muslims themselves, since it seems highly improbable that all of these sources would have so consistently stumbled into the exact same error concerning the end of Muhammad’s life. If this deviant report arose simply through misunderstanding, one would accordingly expect that at least some sources would have managed to understand these events “correctly.” At the very least, this evidence seems to indicate that a tradition of Muhammad’s death at Medina before the invasion of Palestine had not yet become clearly established prior to the beginnings of the second Islamic century.

It should again be made clear from the outset, however, that the existence of this tradition invites much more than an opportunity simply to extend the longevity of Muhammad by a mere two or three years, and the discrepancy of the source materials on this point instead calls for some sort of explanation. Why are there very different memories concerning Muhammad’s relation to the expansion of his religious movement outside of Arabia and his followers’ invasion of Roman territory in Syro-Palestine? Admittedly, one cannot entirely exclude the possibility that the difference is simply the result of a collective misunderstanding, but as this chapter will argue, the nature of the sources in question renders this solution improbable. The fact that no source, Islamic or non-Islamic, from the first Islamic century locates Muhammad’s death before the Near Eastern invasions indicates that it is not simply a matter of having guessed incorrectly. Possibly the esteem expressed for Muhammad by members of this new religious movement may have led each of these non-Islamic writers to the false assumption that he remained in charge for a few years longer than had actually been the case. Such a scenario is certainly not inconceivable, but it would imply that a profound and prolonged ignorance regarding the basic “facts” about Islam’s founding prophet remained pervasive in the various non-Islamic religious communities of the seventh and early eighth centuries. Indeed, if the earliest Muslims had clearly recalled from the start that Muhammad died two years before their invasion of Syria and Palestine, it is hard to imagine that not a single one of the early non-Islamic sources (not to mention the letter of ʿUmar) would manage to get this right. Alternatively, as this study proposes, dramatic changes in the faith of the early Muslims may have given rise to these divergent traditions and could potentially explain the eventual displacement of one tradition by the other. Indeed, as will be seen in chapters to come, there appears to have been some effort initially to deny the reality of Muhammad’s death within the earliest community. Likewise, there is considerable evidence to suggest that primitive Islam transformed rapidly from a non-confessional monotheist faith with an extremely short eschatological timeline into an imperial religion grounded in a distinctively Arabian and Arab identity. Such changes, as we will see, provide a credible context for the apparent shift in early memories about the end of Muhammad’s life.

Doctrina Iacobi nuper Baptizati (July 634 CE)

The earliest extant text to mention Muhammad is the Greek account of a dialogue that purportedly took place in July 634 in Roman North Africa, in the context of the empire’s forced conversion of North African Jews in 632. The text, entitled Doctrina Iacobi nuper Baptizati, was most likely written very soon after the events that it describes, as seems to be required by its concern to address the specific issue of the forced baptism of 632, as well as by references to contemporary political events that suggest a time just after the first Arab attacks on the Roman Empire.¹ The text identifies its author as Joseph, one of the participants in the dialogue, but its central character is Jacob, a Jewish merchant from Palestine who had recently been coerced into baptism while on an ill-timed business trip to Africa. As the text begins, Jacob addresses the other Jews who have been forcibly baptized and explains that he has come to see the truth of Christianity through a miraculous vision and careful study of the scriptures. After extensive instruction and dialogue with his audience, he successfully persuades these newly baptized Jews to commit with their hearts to the faith that they have received through compulsion. Several days later, and approximately midway through the text, a new character appears: Justus, the unbaptized cousin of one of Jacob’s pupils, who has recently arrived from Palestine. Justus is upset that his cousin and so many other Jews have accepted their Christian baptism, and he is persuaded to debate the issue with Jacob before the group. Unsurprisingly, given that this is a Christian text, the story ends with Justus’s conversion. Yet despite this rather clichéd conclusion, the text is a rich source for understanding the history of the eastern Mediterranean world during the crucial period just after the Persian occupation and at the beginnings of the Islamic conquest.

Among other things, this remarkable text is one of our most important resources for understanding relations between the Jewish and Christian communities in the Byzantine provinces, since, unlike so many other early Byzantine writings on Jews and Judaism, the Doctrina Iacobi is regarded as a particularly reliable and accurate source. Anti-Jewish polemics were especially popular during the early Byzantine period, and for the most part this literary tradition is replete with stereotypes and rhetoric, bearing a complicated and very tenuous link with the historical realities of the day. Although these texts usually give the appearance of being directed at converting the Jews, this cannot have been the actual cause for their production, since they frequently misrepresent or misunderstand Judaism so badly that they would have little hope of effectively reaching this audience. These texts are instead best understood as insider literature, intended to reassure the Christian faithful of the truth of their faith by demonstrating (in Christian terms) the superiority of Christianity to Judaism, which was Christianity’s main religious rival in the pre-Islamic Near East.²

Nevertheless, the Doctrina Iacobi defies most of the literary conventions—and conventional interpretations—of the adversus Iudaios genre: it is, as David Olster explains, “the exception that proves the rule.”³ The Doctrina Iacobi is distinguished from its kin most especially by the accuracy with which it portrays Judaism and Jewish life in the late ancient Mediterranean. Whereas most anti-Jewish literature from this period presents a highly stereotyped construct that is rhetorically designed to demonstrate the superiority of Christianity, the Doctrina Iacobi presents a highly detailed and realistic depiction of late ancient Judaism. It is in fact so accurate and nuanced that Olster concludes not only that the Doctrina Iacobi was most likely written with a Jewish audience in mind, but also that its author was almost certainly a converted Jew; otherwise, it is difficult to conceive how the text could have such depth of insight into seventh-century Jewish life.⁴ Moreover, the Doctrina Iacobi’s author displays considerable knowledge of Palestinian geography, as well as of the contemporary situation in North Africa, lending credibility to the text’s genesis among a group of Palestinian Jews who found themselves in Roman Africa at this inopportune time.⁵ In addition, the text details the business dealings of both Jacob and Justus, and even the circumstances of its own production, creating a high level of verisimilitude.⁶ Even if the latter elements are merely in place to enhance “the reality effect” of the story, the author’s descriptions of contemporary social and political life are astonishingly accurate when compared with other sources.⁷ The Doctrina Iacobi stands out within its genre for its careful and accurate representation of such historical details and, more remarkably, for the thorough and thoughtful contextualization of its dialogue within this broader historical setting.⁸

An important part of this backdrop is the appearance of a new prophet in Palestine, who, although he is unnamed, is unquestionably to be identified with Muhammad. The passage in question follows Justus’s conversion, and, like the rest of the dialogue, it is remarkable for its attention to certain details:

Justus answered and said, “Indeed you speak the truth, and this is the great salvation: to believe in Christ. For I confess to you, master Jacob, the complete truth. My brother Abraham wrote to me that a false prophet has appeared. Abraham writes, ‘When [Sergius]⁹ the candidatus was killed by the Saracens, I was in Caesarea, and I went by ship to Sykamina. And they were saying, “The candidatus has been killed,” and we Jews were overjoyed. And they were saying, “A prophet has appeared, coming with the Saracens [ὁ προφήτης ἀνεφάνη ἐρχόμενος μετὰ τῶν Σαρακηνῶν], and he is preaching the arrival of the anointed one who is to come, the Messiah.” And when I arrived in Sykamina, I visited an old man who was learned in the scriptures, and I said to him, “What can you tell me about the prophet who has appeared with the Saracens?” And he said to me, groaning loudly, “He is false, for prophets do not come with a sword and a war-chariot. Truly the things set in motion today are deeds of anarchy, and I fear that somehow the first Christ that came, whom the Christians worship, was the one sent by God, and instead of him we will receive the Antichrist.¹⁰ Truly, Isaiah said that we Jews will have a deceived and hardened heart until the entire earth is destroyed. But go, master Abraham, and find out about this prophet who has appeared.” And when I, Abraham, investigated thoroughly, I heard from those who had met him [Καὶ περιεργασάμενος ἐγω Ἀβραάμης ἤκουσα ἀπὸ τῶν συντυχόντων αὐτῷ] that one will find no truth in the so-called prophet, only the shedding of human blood. In fact, he says that he has the keys of paradise, which is impossible.’ These things my brother Abraham has written from the East.”¹¹

What can one make of this passage, which mixes vivid historical detail with obvious polemic? Is its indication that Muhammad was still alive and leading the invading Arabs as they entered Palestine of any historical significance or has the author (or one of his sources) simply made a mistake? To a certain extent, this judgment will depend on whether other independent witnesses also credibly describe Muhammad as alive at the time of the invasion of Palestine, and as this chapter will demonstrate, a number of such sources exist. In its own right, however, the Doctrina Iacobi is a historical source of particularly high quality that was written very close to the events that it describes. Since the Doctrina Iacobi has repeatedly shown itself to be a reliable source with regard to various other matters, perhaps one should initially give its near contemporary report of Muhammad’s involvement in the conquest of Palestine at least the benefit of the doubt.

For example, comparison with other historical texts confirms the accuracy of the Doctrina Iacobi’s reference to a candidatus Sergius of Caesarea who was killed by the Arabs. Two other sources report the death of Sergius the candidatus in combat with the Arabs: the Syriac Common Source, a now lost chronicle from the mid-eighth century discussed below, and a Syriac chronicle from the year 640.¹² In the Doctrina Iacobi we seem to have an almost contemporary witness to Sergius’s defeat by the Arab army as described in these later sources.¹³ While this by no means ensures that the passage is accurate in all of its other details, the verification of this point by independent sources is a testimony in favor of its general reliability as a historical source. Likewise, the Doctrina Iacobi’s report that Muhammad claimed to possess the “keys of paradise” seems to reflect a very early Islamic tradition that was later abandoned. Not only do other Byzantine sources repeat this tradition, but certain Islamic sources preserve it as well, although the latter attempt to soften the audacity of Muhammad’s claim by reducing it to a metaphor.¹⁴ Perhaps even more important, however, is the high level of conformity between the Doctrina Iacobi and other witnesses to the social, political, and religious events of the early 630s noted already above. As Olster’s persuasive analysis of this text demonstrates, the Doctrina Iacobi’s accurate representation of its historical circumstances is precisely what makes it so remarkably different from other anti-Jewish writings of the same period. Thus, while one certainly cannot assume that this source is reliable in every detail, we nevertheless may take some confidence in the fact that the Doctrina Iacobi has been shown to be generally trustworthy through comparison with other sources from the period. The fact that it was probably written so close to the time it describes only adds to its credibility.

Particularly significant in this report is the Doctrina Iacobi’s notice that this prophet who arrived in Palestine with a Saracen army was “preaching the arrival of the anointed one who is to come, the Messiah.” As Crone and Cook observe, this earliest witness to Muhammad’s religious message from outside of the Islamic tradition portrays him as preaching Jewish messianism. Although Cook and Crone initially characterize this idea as “hardly a familiar one,” thanks in large part to their own work, it has become much less unfamiliar.¹⁵ Most importantly, the seventh-century Jewish apocalypse preserved in the Secrets of Rabbi Shimʿōn (discussed below) confirms that there were in fact Jews who understood Muhammad and his message as the fulfillment of Jewish messianic expectations. Theophanes’ Chronicle echoes this information at a greater distance, and the report in Sebeos’s Armenian History of Arab and Jewish unity during the assault on Palestine, discussed in the final chapter, may also point indirectly to such beliefs.¹⁶ Moreover, the Qurʾān itself would appear to substantiate these reports: as discussed below in Chapter 3, the Qurʾān’s unmistakable eschatological urgency reveals that Muhammad and his early followers believed themselves to have been living in the final moments of history, just before the impending judgment and destruction that would soon arrive with the Hour. In Jewish ears, this forecast of the eschaton’s proximate arrival would inevitably awaken expectation of the messiah’s advent, which was expected to precede the Final Judgment. As will be seen in the final chapter, substantial evidence signals the presence of a significant Jewish element among Muhammad’s earliest followers, and undoubtedly these Jewish “Believers” would have understood his eschatological preaching through the lens of their own traditions. Thus, while Fred Donner is certainly correct to note that the early Islamic sources do not reveal any clear belief in a coming messianic figure, as both he and Suliman Bashear rightly conclude, the Jewish members of the early community of the Believers undoubtedly would have interpreted Muhammad’s eschatological message according to their own messianic expectations.¹⁷

Hoyland’s criteria ask that we push beyond these conclusions, however, and scrutinize the source’s source, as it were. In this regard the situation is less than ideal, but it is much better than it might be. In the best possible case, we would have the statement of an eyewitness (or better still, eyewitnesses). In the Doctrina Iacobi, we find instead what essentially amounts to third-hand testimony, although the account is allegedly based on reports from eyewitnesses. Jacob, the author, heard this report of the Arab invasion of Palestine from Abraham’s letter, which Abraham’s brother Justus read aloud in his presence. Abraham, who was living in Palestine, identifies the source of his information in interviews that he had personally conducted with “those who had met him [that is, Muhammad].” Despite these intervening steps, we may take some measure of confidence in Jacob’s report: according to this genealogy, it derives from the testimony of multiple eyewitnesses and was then quickly committed to writing before reaching Jacob. Moreover, the report’s close proximity to the actual events themselves stands further in its favor: mere months seem to have transpired since the invasion. On the whole, these circumstances present a much more credible line of transmission than the pedigrees that accompany the earliest Islamic traditions about Muhammad and the conquest. As will be seen in the following chapter, their chains of transmission (isnāds) are notoriously unreliable and often highly artificial, purporting to document transmission over multiple generations. By comparison, the transmission of Jacob’s report is both immediate and relatively uncomplicated.

Admittedly, there are elements of polemic in this passage, including especially the diatribe against Muhammad as a false prophet. But by and large the details are descriptive and often can be confirmed by other sources, as seen in the case of Sergius the candidatus and the report that Muhammad claimed to hold the keys to paradise: although the latter is potentially polemical, as noted above, later Byzantine and Islamic sources corroborate this characterization. Even the allegation that Muhammad was preaching the advent of the messiah seems to be more or less accurate, reflecting a Jewish understanding of his eschatological message that is evident in other early sources. In similar fashion, the Doctrina Iacobi’s indication that Muhammad was still alive and coming with the Arabs during the Palestinian campaigns of 634 seems to be a descriptive, non-polemical observation that is confirmed by a number of other sources. It is, moreover, information that could have been known to Abraham’s informants, “who had met him,” as he reports, and potentially to others as well who had experienced the Arab invasion of Roman Palestine.

More importantly, there is no obvious apologetic or polemical reason for the Doctrina Iacobi’s author (or his sources) to have invented Muhammad’s leadership during the campaign in order to serve a broader ideological purpose.¹⁸ Hoyland suggests, somewhat half-heartedly it seems, that the widespread Christian reports of Muhammad’s participation in the conquest of Palestine may stem from an effort “to emphasize his un-prophetlike behavior.”¹⁹ This would certainly fit with the Doctrina Iacobi’s polemic against Muhammad as a false prophet, since, as the “old man” says, “prophets do not come with a sword and a war-chariot.” Nevertheless, as Hoyland himself concedes within the very same sentence, “the essence of [this representation] is already encountered in the very foundation document of the Muslim community, the so-called Constitution of Medina, which unites believers under the ‘protection of God’ to fight on his behalf.”²⁰ Moreover, as Hoyland notes elsewhere, the Qurʾān itself attests that “coming with sword and chariot” was an integral part of Muhammad’s message: “That religion and conquest went hand in hand in Muḥammad’s preaching is clear from many passages in the Qur’an which command: ‘Fight those who do not believe in God and the Last Day … until they pay tribute’ (ix.29) and the like.”²¹ It is thus highly unlikely that the Doctrina Iacobi, along with the various other non-Islamic sources that will be examined, has falsely represented Muhammad as alive at the time of the Islamic invasion of Palestine in order to discredit him by portraying him as a prophet who preached a message of conquest. The Islamic sources themselves preserve this image of Muhammad rather well, and there would have been little need for these authors to invent data in order to emphasize a point that otherwise emerges quite clearly from both the Qurʾān and the early Islamic tradition. On the whole then, the Doctrina Iacobi generally fares well in regard to Hoyland’s criteria and should accordingly be taken seriously in its report of a tradition that as late as 634 Muhammad came to Palestine “with the Saracens.”²²

Of course, one cannot completely exclude the possibility that the sources behind Doctrina Iacobi may have simply misunderstood Muhammad’s relation to the invasion of Palestine. Perhaps Muslim confessions of Muhammad as a religious prophet whose teachings they followed were mistakenly understood as indications that he was a still-living military and political leader of the Muslims. Since Muhammad was a bellicose prophet who had preached jihād, it is possible that the conquered peoples of Palestine and the Near East merely assumed that he was leading the jihād that subdued their territory and brought it under the dominion of his religious movement. Nonetheless, as will be seen in the remainder of this chapter, the wide range of sources conveying this tradition strongly suggests that such a misunderstanding is unlikely to be the origin of this difference between the Islamic and non-Islamic sources. If such confusion were the cause of Muhammad’s representation as still living at the beginning of the Palestinian campaign, then one must assume that a large number of independent sources have somehow separately made the same mistake. While this certainly is not impossible, it becomes increasingly improbable with each source, and the broad geographic spread of this tradition across the various religious communities of the early Islamic world instead suggests more probably a primitive tradition that underlies these reports. Likewise, the fact that no source “correctly” locates Muhammad’s death before the Palestinian invasion or otherwise clearly separates him from these events before the emergence of his official Islamic biography in the middle of the eighth century is a strong indication that this association of Muhammad with the conquest of Palestine reflects an early tradition that circulated widely among the different religious groups of the Mediterranean world in the seventh and eighth centuries. There are, as will be seen in chapters to follow, other more likely explanations for the discrepancy between these early sources and the later Islamic tradition on this issue. Consequently, even if Muhammad did not in fact survive to personally lead the invasion of Palestine, as the Doctrina Iacobi reports, the convergence of so many sources on this point seems to reveal what is likely an early tradition, presumably coming from within Islam itself, that Muhammad led his followers into the Abrahamic land of promise. There they seem to have anticipated that he would guide them to meet the eschaton’s impending arrival, signaled here by Jewish expectations of the messiah’s appearance.

The Apocalypse of Rabbi Shimʿōn b. Yoḥai (635–45?)

As Crone and Cook are quick to note in Hagarism, certain medieval Jewish apocalyptic traditions ascribed to Rabbi Shimʿōn b. Yoḥai form an important compliment to the Doctrina Iacobi’s witness, particularly in providing further evidence of a messianic understanding of the Islamic conquests among many contemporary Jews.²³ Nevertheless, Crone and Cook fail to note the parallel indication by these Jewish visionary texts that Muhammad led his followers in the invasion of Palestine, an oversight owing itself most likely to their dependence on Bernard Lewis’s translation of a key passage in 1950.²⁴ While Lewis’s translation is certainly not incorrect, it is problematic inasmuch as it obscures certain grammatical ambiguities that are essential for the present question of Muhammad’s relation to the invasion of Palestine. As will be seen, the full complement of witnesses to these Rabbi Shimʿōn b. Yoḥai traditions indicates that this early Jewish vision of the Islamic conquests identified Muhammad as the leader of the Ishmaelite army that was believed to be the agent of Israel’s divine deliverance from Roman oppression in Palestine.

Several closely related apocalyptic texts describe Rabbi Shimʿōn’s visions of the Islamic conquests, each giving a slightly different version of events that seems to depend on an earlier common source. The earliest of these works, and also the most important, is The Secrets of Rabbi Shimʿōn b. Yoḥai, an apocalypse written sometime around the middle of the eighth century whose visions cover the period between the Islamic conquests and the ʿAbbāsid revolution. As The Secrets begins, Rabbi Shimʿōn reflects on the “Kenite” of Numbers 24:21, which is revealed to him as a prediction concerning the Ishmaelites and their coming dominion over the land of Israel.²⁵ When he cries aloud with frustration, asking if the Jews had not yet suffered enough oppression at the hands of Edom (that is, Rome), the angel Metatron comes to him and reassures him that God will use the Ishmaelites to free the Jews from Byzantine oppression. “Do not be afraid, mortal, for the Holy One, blessed be He, is bringing about the kingdom of Ishmael only for the purpose of delivering you from that wicked one (that is, Edom [Rome]). In accordance with His will He shall raise up over them a prophet. And he will conquer the land for them [], and they shall come and restore it with grandeur. Great enmity will exist between them and the children of Esau.”²⁶ The revelation continues as Metatron responds to Rabbi Shimʿōn’s questions by equating Israel’s liberation through this Ishmaelite prophet to the messianic deliverance foretold by Isaiah’s vision of the two riders (Isa. 21:6–7).²⁷ This identification of Muhammad as the fulfillment of Jewish messianic hopes is remarkable, and it offers important corroboration of the Doctrina Iacobi’s report that the Saracen prophet was “preaching the arrival of the anointed one who is to come, the Messiah.” Predictions concerning the various Umayyad rulers then follow, including a prophecy that Muhammad’s successor, apparently the caliph ʿUmar, would restore worship to the Temple Mount.²⁸ The apocalypse then concludes with the ʿAbbāsid revolution, which is identified as the beginnings of an eschatological confrontation between Israel and Byzantium that will result in a two-thousand year messianic reign, followed by the Final Judgment.²⁹

In view of this rather positive assessment of Muhammad’s prophetic mission and the early years of Islamic rule, numerous scholars have observed that The Secrets of Rabbi Shimʿōn almost certainly depends on a much earlier source for its description of these events.³⁰ It is hard to imagine that a Jewish author of the mid-eighth century would have written so glowingly of the advent of Islam, painting Muhammad and his followers in such messianic hues over a century later. Moreover, as Crone and Cook rightly observe, “the messiah belongs at the end of an apocalypse and not in the middle” as one finds in The Secrets, an anomaly that also seems to indicate the inclusion of older material.³¹ On the whole, the character of this section of the apocalypse strongly suggests that The Secrets here has incorporated some very lightly edited traditions from an older Jewish apocalypse that was roughly contemporary with the events of the conquests themselves, possibly written in the first decade after the Arab invasions. Moreover, this lost apocalypse appears to relate the perspective of a Jewish group either within the early Islamic movement or closely allied with it. We have long known from the Islamic tradition itself that in the early stages Jewish groups were welcomed into Muhammad’s new religious community while maintaining their Jewish identity. Yet according to Muhammad’s early biographers, this was a brief experiment limited to certain Jewish tribes of Medina that was quickly abandoned after it failed. There is increasing evidence, however, that for the first several decades Muhammad’s followers comprised an inter-confessional, eschatological religious movement focused on Jerusalem and the Holy Land that welcomed Jews and other monotheists within the community, as will be seen further in the final chapter. The older apocalypse of Rabbi Shimʿōn echoed in this more recent text almost certainly derives from this milieu: otherwise, it is difficult to understand its proclamation of the invading Arabs as divinely appointed “messianic” deliverers who would restore worship to the Temple Mount.

It is of special note that this seventh-century apocalypse of Rabbi Shimʿōn appears to have described this Ishmaelite “messiah,” unmistakably here Muhammad, as leading this conquest of the Holy Land and liberating it from the Romans. Yet Lewis translates the crucial passage, cited above, as follows: “He raises up over them a Prophet according to His will and will conquer the land for them and they will come and restore it in greatness.”³² Lewis’s translation determines God, “the Holy One,” as the actor who will conquer the land for the Ishmaelites. The Hebrew, however, is in fact ambiguous on this point. The verb in question is an imperfect third-person singular (יכבוש), and thus its subject is potentially either God or the prophet that God will raise up. Lewis has determined to understand God as the one who will conquer the land, and while this certainly is a possibility, it seems more likely that the prophet is in fact intended: God will raise up the prophet, but it is the prophet who will lead the conquest of the land. No doubt Lewis was inspired to translate the passage as he did by the Islamic historical tradition, which relates Muhammad’s death in Medina prior to the invasion of Palestine. Writing in 1950, Lewis was presumably unaware of this counter-tradition that Muhammad led his followers in the initial assault on Palestine; consequently, he not unreasonably assumed that The Secrets and its source envisioned God, rather than Muhammad, as subduing the land, since according to the received Islamic tradition Muhammad was already dead by this time. Nevertheless, in view of the new sources that have now come to light, one must seriously consider the possibility that, as the text seems to suggest, it is the prophet who conquers the land. Inasmuch as God raised up this Ishmaelite prophet for the purpose of delivering the Jews from the Romans, it seems implicit that the prophet was to achieve this divine mission by leading the conquest of the land himself.

Indeed, the reading that it is the prophet, rather than God, who conquers the land seems highly preferable here, as is confirmed by the other witnesses to this seventh-century Jewish apocalypse, all of which preserve a memory of the Ishmaelite prophet, rather than God, as the one who conquers the land. For example, a fragment preserving the opening section of The Secrets of Rabbi Shimʿōn survives among the Cairo Geniza texts, and according to this version, “He raises over them a crazy prophet, possessed by a spirit, and he conquers the land for them [] and they come and seize dominion in greatness and there will be great enmity between them and the sons of Esau.”³³ Here Lewis translates the passage so that the prophet (“he” instead of “He”) is identified as conquering Palestine, which seems to be indicated by the context: surely God would not conquer the land for this crazy, possessed prophet and his followers. The same reading is also confirmed by another manuscript in Munich, which preserves a version very similar to that of the Geniza fragment.³⁴ Thus these other manuscript witnesses to The Secrets clearly relate this prophecy as describing the conquest of the land by an Ishmaelite prophet, whom the circumstances clearly identify as Muhammad.

Other closely related sources convey a similar understanding of Muhammad’s role in the conquest of Palestine, namely, the Ten Kings Midrash and The Prayer of Rabbi Shimʿōn b. Yoḥai, both of which seem to have drawn independently on the now lost seventh-century apocalypse of Rabbi Shimʿōn.³⁵ Judging from the historical figures identified in the Ten Kings Midrash, it is roughly contemporary with The Secrets of Rabbi Shimʿōn, placing its composition sometime not long after the reign of al-Walīd II (d. 744).³⁶ And while the Ten Kings Midrash appears to have also made direct use of this older apocalypse of Rabbi Shimʿōn b. Yoḥai, there is no mention in the Ten Kings Midrash of God raising up an Ishmaelite prophet, and the original Ishmaelite “messianism” of this source has been slightly rearranged. Nevertheless, the relevant section concludes with the following prediction concerning Muhammad. “At the beginning of his dominion, when he goes forth, he will seek to do harm to Israel, but great men of Israel will join with him and give him a wife from among them, and there will be peace between him and Israel. He will conquer all the kingdom and come to Jerusalem and bow down there and make war with the Edomites and they will flee before him and he will seize the kingship by force and then he will die.”³⁷ The indication that Muhammad led the conquest of Palestine and would die only afterward is unambiguously clear here, confirming what we have seen in the Doctrina Iacobi and The Secrets of Rabbi Shimʿōn, but the notice that he would actually come to Jerusalem to “bow down there” is otherwise unprecedented to my knowledge. Nevertheless, this feature would appear to comport with the exalted status of Jerusalem in earliest Islam, as is further discussed in the final chapter.

Likewise, The Prayer of Rabbi Shimʿōn b. Yoḥai, a more recent text dating from the time of the First Crusade, also describes Muhammad as leading the invasion of Palestine. Although The Prayer of Rabbi Shimʿōn has transformed the relevant prophecy so that it relates to the events of the Crusades, its words clearly echo both the Cairo and Munich versions of The Secrets, noting that “a crazy man possessed by a spirit arises and speaks lies about the Holy One, blessed be He, and he conquers the land, and there is enmity between them and the sons of Esau.”³⁸ While the Ishmaelite prophet is here portrayed in strongly negative terms, this leader, originally Muhammad one must assume, is said to conquer the land. Thus, despite the change of historical context, The Prayer of Rabbi Shimʿōn has reused this older tradition of the Ishmaelite prophet’s conquest of Palestine, applying it unchanged to the new circumstances presented by the Crusades.

The persistence of this particular theme, Muhammad’s conquest of the land, across all of these sources, despite their heavy revisions to this prophecy, rather strongly suggests that this was an original feature of the earlier seventh-century apocalypse on which they have all drawn. Although each has altered the originally positive, messianic assessment of Muhammad and his religious movement that was present in their now lost source, their convergence in reporting Muhammad’s leadership of the Arab invasion of Palestine seems to confirm that this feature was a primitive element of this near contemporary apocalyptic vision of the Islamic conquests. Thus, this complex of texts bears witness to a tradition of Muhammad’s continued vitality and leadership during the invasion of Palestine within the context of Jewish messianic expectations, seemingly recorded, like the Doctrina Iacobi, close to the time of the Arab conquests themselves. It is an impressive convergence on this point, which appears to reflect a very early memory from the Palestinian Jewish community of Muhammad’s leadership during the invasion of Palestine.

It is somewhat difficult to assess the quality of this witness according to Hoyland’s criteria, particularly in light of its apocalyptic genre. On the one hand, the source of its information is identified as the angel Metatron, and it presents the Ishmaelite prophet’s invasion of Palestine within a totalizing narrative of Israel’s deliverance at the hands of this prophet and his followers. On the other hand, it would appear that this notice, despite its obvious literary conventions, originated within a context that was either very close to or perhaps even inside the primitive Islamic community itself. The early Jewish apocalyptic vision of the Islamic conquest that has been collectively adopted by these later texts clearly seems to have anticipated Jewish redemption through the invading Ishmaelites and their prophet. The seventh-century Jewish group that produced the original apocalypse of Rabbi Shimʿōn at a time close to the events of the Arab conquest themselves appears to have placed its faith in the “Islamic” prophet and the early caliphs as deliverers raised up by God. In its acceptance of Muhammad’s divine guidance, the apocalypse thus seems to reflect a viewpoint that in some sense is that of an insider.

While this perspective is perhaps difficult to comprehend in light of the confessional boundaries that have long since separated Islam and Judaism, recent research into Islamic origins has revealed that such divisions were likely not as important during Islam’s first decades. The apocalypse of Rabbi Shimʿōn is itself important evidence of the early Islamic community’s openness to other monotheist confessions, and it would seem that it preserves the visionary hopes of a Jewish group that joined cause with the invading Arabs and the message of their prophet, whom they saw as their liberator. There are in fact strong indications that Islam’s sacred geography originally focused not on Mecca and the Ḥijāz, but instead on Jerusalem and Palestine, which Muhammad’s earliest followers seem to have regarded as the promised land of their inheritance, a holy land rightfully belonging to Abraham’s descendants, Jews and Arabs alike. The Islamic invasion of the Holy Land thus seems to have been conceived at least in part as the liberation of the Abrahamic patrimony from Roman rule and oppression, an undertaking that would have aligned the Arab cause with Jewish apocalyptic hopes.

Consequently, one would imagine that this apocalypse of Rabbi Shimʿōn reflects the perspective of a Jewish group that was sympathetic to, if not even allied with, the invading Arabs and their prophet. Its prediction that this prophet would lead the conquest of the Holy Land thus seems to reflect the perspective of contemporary eyewitnesses who themselves had some experience of the invasion and early Islamic rule over Jerusalem. Whatever the precise nature of the community behind this text may have been, it clearly describes the invading Arabs and their prophet in positive terms, as divinely appointed agents of deliverance. The sharp dissonance of this favorable assessment of Muhammad and his devotees with later Jewish attitudes toward Islam speaks not only to the antiquity of the source itself; this quality also diminishes the possibility that Muhammad’s participation in the conquests was contrived to serve some polemical purpose. To the contrary, it is extremely difficult to envision a later Jewish redactor inventing the idea of Muhammad’s divinely appointed liberation of the Holy Land. It is instead much easier to understand such sentiments as reflecting the impressions of contemporary Jews whose apocalyptic expectations aligned them, at least for a time, with the invading Muslims and their prophet.

The Khuzistan Chronicle (ca. 660 CE)

Further evidence of Muhammad’s leadership during the conquest of Palestine occurs in a brief, anonymous Syriac chronicle that was probably composed in the Khuzistan region of southwestern Iran, where most of its events take place. The Khuzistan Chronicle, as this text is often called,³⁹ is generally dated to around 660 on the basis of its contents, including most notably the fact that it makes no clear reference to any event after 652.⁴⁰ The chronicle’s account of the Islamic conquests is somewhat unusual in that it describes the events of the conquests twice and in two very different contexts. The chronicle first gives a rather general notice of the conquests according to its chronological sequence, and then near its conclusion, the author returns to a discussion of the Islamic invasions outside of the broader chronology and in more detail, focusing especially on Islamic military activity in Khuzistan. This doublet reflects a rather peculiar feature of the chronicle’s general organization. Most of the chronicle adheres to a strict chronological order in relating events, marking time according to the succession of both the Persian emperors and the leaders of the East Syrian (that is, “Nestorian”) church. But following accounts of the reigns of Emperor Yazdgerd III (632–52) and Patriarch Maremmeh (646–49), the work suddenly alters its structure. The chronicle’s final entries include, in order, “an account of the miraculous conversion of some Turks by Elias of Merv (d. after 659), a list of towns founded by Seleucus, Semiramis and Ninus son of Belus, a portrayal of the Arab conquests (630s-40s), and a short survey of Arabian geography.”⁴¹ This sudden departure from chronological sequence has led many interpreters to suggest that these final sections are the work of another author, who has appended this material, including the second description of the Islamic conquests, to an earlier chronicle that originally concluded with Patriarch Maremmeh’s death.⁴² Other factors, however, suggest that both sections are in fact the work of the same author, and as Hoyland has proposed, “it may be, then, that the disjuncture is not an indication of a change in author, but of a change of focus and/or source.”⁴³ It seems plausible that upon reaching the end of his historical narrative, the chronicler turned in his conclusion to focus on topics of special significance for mid-seventh-century Khuzistan and East Arabia. This shift of focus presumably reflects the author’s interest in his own milieu, and consequently, there is a very real possibility that the information in this section is based on eyewitness reports or even the author’s own personal knowledge.

In this final section, the chronicle’s second account of the Islamic conquest narrates the Arab invasion of northern Khuzistan, focusing especially on the capture of the cities Shush and Shushtar. The account is so rich in detail that it almost certainly derives from eyewitness reports,⁴⁴ but inasmuch as it describes the conquest of Mesopotamia instead of Palestine and does not identify Muhammad as leading the invasion, this vivid account is unfortunately irrelevant to the matter at hand. The chronicle’s initial notice of the Islamic conquests, however, which appears according to chronological sequence, is more valuable in this regard. Here the chronicler describes the initial Arab assaults against both Persia and the Byzantines in Syro-Palestine, reporting these events as follows:

And Yazdgerd, who was from the royal lineage, was crowned king in the city of Estakhr, and under him the Persian Empire came to an end. And he went forth and came to Māḥōzē and appointed one named Rustam as the leader of the army. Then God raised up against them the sons of Ishmael like sand on the seashore. And their leader was Muhammad [], and neither city walls nor gates, neither armor nor shields stood before them. And they took control of the entire land of the Persians. Yazdgerd sent countless troops against them, but the Arabs destroyed them all and even killed Rustam. Yazdgerd shut himself within the walls of Māḥōzē and in the end made his escape through flight. He went to the lands of the Huzaye and the Mrwnaye,⁴⁵ and there he ended his life. And the Arabs took control of Māḥōzē and all the land. They also went to the land of the Byzantines, plundering and laying waste to the entire region of Syria. Heraclius, the Byzantine king, sent armies against them, but the Arabs killed more than one-hundred thousand of them.⁴⁶

The structure of this passage seems to indicate Muhammad’s leadership of the Arabs during their initial attacks against the Persians and the Romans. After first naming the Persian “king,” Yazdgerd, and then identifying the leader of the Persian army in Rustam, the chronicler describes the Arab invasion of Persia, designating Muhammad as the Arab leader in this specific context. Muhammad’s positioning alongside of these other leaders in the conflict, including Heraclius, strongly suggests Muhammad’s participation in the initial phase of the Near Eastern conquest.

Unfortunately, we know very little regarding the sources from which the author obtained his information that Muhammad was leading the Muslims at the time of the conquests. The chronicler identifies his sources rather generally as the ecclesiastical and secular histories from the period between the death of Hormizd son of Khosro and the end of the Persian Empire.⁴⁷ But since the Khuzistan Chronicle was composed so soon after the events of the Islamic conquests, it is certainly possible that its author relied on reports of eyewitnesses rather than written sources for knowledge of the Islamic invasion, particularly since he appears to have relied on eyewitness testimony for his description of Khuzistan’s conquest. In any case, the author himself was most likely not an eyewitness to the Palestinian campaign or to Muhammad’s role therein: how he came by the information that Muhammad was leading this powerful army is not known. But in the Khuzistan Chronicle’s favor are the facts that its author seems to have taken a special interest in recording the events of the Islamic conquest, and that he had access to eyewitness testimonies (or perhaps even personal experience?) for at least some of his information. As for the character of the observation, it is not polemical, nor does the notice of Muhammad’s leadership during the conquests serve any sort of grand narrative within the chronicle; here the chronicle makes descriptive observations about events that took place less than thirty years prior. Thus the Khuzistan Chronicle forms an additional witness to a tradition of Muhammad’s continued leadership of the Muslims as they began their conquest of Rome and Persia, in a source written outside the boundaries of Rome, at the heart of the recently fallen Persian Empire.

Jacob of Edessa, Chronological Charts (691/92 CE)

Jacob of Edessa was a prolific author of the later seventh century, of whom it has been said that his importance in Syriac Christian culture is equivalent to that of Jerome in Western Christendom.⁴⁸ Jacob’s contributions to the medieval West Syrian (that is, “miaphysite”) church are extensive. In his day he was particularly renowned, or perhaps more accurately, notorious, for his work in canon law: in addition to producing a number of important works on the subject, he famously burned a copy of the ecclesiastical regulations while bishop of Edessa to protest the laxity of their observance in the church, after which he (perhaps wisely) withdrew to a monastery. Jacob was also instrumental in standardizing aspects of Syriac grammar, and the West Syrian tradition of indicating vocalization was his invention. Like Jerome, he labored to produce a more accurate version of the biblical text, and he wrote numerous biblical commentaries in addition to various theological and philosophical works. In his youth Jacob had gone to Alexandria to undertake advanced study of Greek, which enabled him to translate, among other things, the works of Severus of Antioch from Greek into Syriac and the Categories of Aristotle. He also authored a number of liturgical texts, and his extensive correspondence with people across Syria also survives.⁴⁹ But our primary concern in the present context is Jacob’s Chronicle, or his Chronological Charts as the text is perhaps more accurately named: these present a somewhat complicated, but nevertheless important, witness to the tradition of Muhammad’s leadership during the invasion of Palestine.

Jacob’s Chronological Charts were prepared with the intent of covering the interval from the end of Eusebius’s Church History up until the end of the seventh century by presenting “in brief the events of the time and the years of empires … placed facing each other so that it might be for those coming to it [to see] who were at a certain time the kings, generals, scholars, writers.”⁵⁰ Unfortunately, much of this chronicle is lost: only a series of extracts has survived, preserved in a single manuscript from the tenth or eleventh century. Among the missing sections is Jacob’s record of events from 631 until 692, the year that he composed the chronicle. Ordinarily, this would considerably limit the text’s value for assessing the date of Muhammad’s death, since it breaks off just before the traditional date of his death in 632. Nevertheless, in this case we are the beneficiaries of a rare and fortunate error in Jacob’s chronology. According to Jacob’s charts, in 620/21 “the first king of the Arabs, Muhammad, began to reign for seven years.”⁵¹ Seven years later, the chart records in 627/28 the beginning of Abū Bakr’s reign as the second king of the Arabs, which lasted for two years and seven months.⁵² This of course places Muhammad’s death in 627/28, four to five years before the traditional date. Jacob’s lapse in chronology here is surprising, given the fact that Jacob’s chronicle is otherwise highly regarded for its accuracy.⁵³ Nonetheless, a list of caliphs compiled between 705 and 715 gives the same dates for Muhammad’s reign, perhaps having followed Jacob’s charts, as does the Hispanic Chronicle of 754, discussed below.⁵⁴

At first glance, one would hardly think that Jacob’s report could strengthen an argument for extending the time of Muhammad’s death beyond its traditional date, since according to Jacob, it occurred even earlier. No doubt this is why Cook, Crone, and Hoyland do not include this witness among the Christian sources identifying Muhammad as still alive during the campaign in Palestine. While Jacob may have had erroneous knowledge of the length of Muhammad’s reign, he nevertheless is in complete harmony with these other sources in recording the onset of the Islamic conquest of Palestine while Muhammad was still alive and leader of the Muslims. Beginning with Muhammad’s reign, Jacob’s chronological charts are ordered into four columns that count the years of the various Roman, Persian, and Islamic leaders, alongside a count of the years since Jacob’s charts began in the twentieth year of Constantine’s reign. On both sides of these charts are comments noting important historical events that coincide with the regnal years tabulated in the charts. Beside the year 625/26, on the left side of the chart, Jacob records that “the Arabs began to make raids in the land of Palestine.”⁵⁵ The sentence begins next to 625/26 and ends on the following line, beside the year 626/27, but it is clear that this comment identifies the beginning of the Islamic campaign in Palestine with the fifth year of Muhammad’s reign, two years before his death.

Conceivably, one could interpret this notice as possibly referring to the first minor skirmish between a small Muslim force and an army of Arab tribes allied with the Byzantines at Muʾta in 629, if one were determined to bring this report more in line with the traditional chronology. But this solution does not seem very likely, at least according to the traditional understanding of this early confrontation between Muhammad’s followers and the “Romans.”⁵⁶ Although Muʾta was technically in Palestina tertia, about twenty-five miles south of the province of Arabia in what is today Jordan, it was certainly very much on the margins of the Byzantine Empire. There is no evidence of any Roman troops at Muʾta during the sixth and seventh centuries, and the forces that actually would have engaged the Muslims were from a variety of Arab confederates. The battle itself was rather minor: it consisted of a single engagement in which the Muslims were soundly defeated. A small skirmish on the fringes of the empire between Christian and Muslim Arabs at a minor outpost hardly seems worthy of Jacob’s notice of Islamic raids in Palestine. Perhaps more importantly, the conflict at Muʾta is unattested in the Syriac historical tradition, or in any Syriac text at all to my knowledge. The sole reference to the battle of Muʾta outside of the Islamic historical tradition is the Greek Chronicle of Theophanes, written in the early ninth century, and Theophanes almost certainly relied on Islamic sources for his account of this battle.⁵⁷

By contrast, the initial phase of the Islamic conquest of Palestine sounds very much like the “raids” that Jacob envisions. In 633–34, the Islamic army moved into southern Palestine, in the province of Palestina prima, and made a number of smaller engagements, mostly with local garrison forces in the countryside. But initially there were no major confrontations with the Byzantine army, and the towns and cities remained under Byzantine control.⁵⁸ These circumstances more credibly reflect the events that Jacob describes as “raids” in Palestine, two years, according to his count, before Abū Bakr succeeded Muhammad as the leader of the Muslims after the latter’s death. Therefore, even though Jacob is mistaken in the length he assigns to Muhammad’s reign, his chronicle provides yet another witness to the tradition that Muhammad was still alive as the conquest of Palestine began. We do not know the source of Jacob’s information in this instance, although we can assume that an individual in Jacob’s position would have had access to a number of different sources, both written and oral. His general reliability in sifting through these sources makes Jacob’s error concerning the precise date and number of years that Muhammad ruled quite surprising. There are moreover no signs of any apologetic agenda or totalizing explanation in the terse outlines of these charts, which rather dryly signal the beginnings of the Palestinian conquest before Muhammad’s death.

The History of the Patriarchs of Alexandria: The Life of Patriarch Benjamin (before 717 CE)

The History of the Patriarchs of Alexandria is a rather complex text that was first compiled in late antiquity, but over the centuries it has continually been augmented, revised, and updated as new patriarchs have sat on the throne of St. Mark: the most recent update was added in 1942.⁵⁹ Most of the material covering the first millennium was originally composed in Coptic, and in the tenth century this was all translated into Arabic, which has been the language of composition ever since. Through a careful analysis of editorial notes scattered throughout the earliest extant versions of this text, David Johnson has been able to identify various redactional layers from the first thousand years of its history.⁶⁰ The earliest portion of the text derives from a Coptic History of the Church that today is known only in fragments. This first segment covers the period from the founding of the Egyptian church up to the reign of Dioscorus (first century–451).⁶¹ This section is followed by a second redactional unit that was composed by a certain George the Archdeacon, who narrates the interval between Patriarch Cyril (d. 444) and the reign of the caliph Sulaymān (715–17).⁶²

As a part of his contribution to the History of the Patriarchs of Alexandria, George the Archdeacon includes a life of Patriarch Benjamin (626–65), during whose lengthy reign the Muslims conquered Egypt along with the rest of the Byzantine Near East. George begins his account of the Islamic conquests with a dream of Heraclius, which warned him, “Truly, a circumcised nation will come upon you, and they will defeat you, and they will take possession of the land.”⁶³ Mistakenly thinking that the dream warned against the Jews, Heraclius ordered all the Jews and Samaritans in the Roman Empire to be baptized. The narrative then explains his mistake with the following account of the rise of Islam:

And after a few days, there arose a man among the Arabs, from the southern regions, from Mecca and its vicinity, named Muhammad. And he restored the worshippers of idols to knowledge of the one God, so that they said that Muhammad is his messenger. And his nation was circumcised in the flesh, not in the law, and they prayed toward the south, orienting themselves toward a place they call the Ka‘ba. And he took possession of [وملك] Damascus and Syria, and he crossed the Jordan and damned it up.⁶⁴ And the Lord abandoned the army of the Romans before him, because of their corrupt faith and the excommunication that was brought against them and because of the Council of Chalcedon by the ancient fathers.⁶⁵

This passage identifies Muhammad as leading the conquest of “Damascus and Syria,” crossing over the river Jordan with his followers and into Palestine, where the Roman armies fell before him. We do not know the source of the information, since the various biographies that comprise the History of the Patriarchs generally draw on earlier, individual vitae while adding some supplementary material.⁶⁶ In view of this fact, it is quite likely that this report of Muhammad’s involvement in the conquests antedates George the Archdeacon’s addition to the History of the Patriarchs of Alexandria: George probably has taken this information from an earlier vita of Benjamin. Muhammad’s capture of Damascus and Syria and his crossing into Palestine are reported here in a direct, matter-of-fact manner that is in no way polemical. While there are references to broader historical narratives, particularly the Council of Chalcedon (which is unsurprisingly condemned), there is no trace of any anti-Islamic or apologetic agenda in this account of the rise of Islam.

The Spanish Eastern Source (ca. 741 CE)

During the earliest years of Islamic rule in Spain, two Latin chronicles, the Byzantine-Arab Chronicle of 741 and the Hispanic Chronicle of 754, were written almost simultaneously. Surprisingly, these are the only surviving Latin historical works composed during the many centuries of Islamic dominion in southern Spain. Although there are considerable differences between the two chronicles, some of which we will note, both have drawn on a common source for most of their information regarding the history of Islam.⁶⁷ Inasmuch as the information that concerns us derives from this shared source, we will consider these two related chronicles together in order to ascertain the witness of their earlier source regarding Muhammad’s role in the conquest of Palestine. The precise nature of this source, however, remains something of a mystery.

The Spanish Eastern Source, as we will name this shared document, is perhaps most surprising for its rather favorable treatment of Muhammad and the early Islamic caliphs. This comes through most clearly in the Byzantine-Arab Chronicle of 741, which, although it shows signs of having abbreviated the Spanish Eastern Source, does not add any sort of polemic to its source’s consistently positive descriptions of the Islamic leaders. This is in contrast to the Hispanic Chronicle of 754, which “often adds a pejorative remark or omits the notice altogether if it is too positive, as with that on Muhammad.”⁶⁸ The Spanish Eastern Source’s positive representation of Islam led one early interpreter to suppose that its author must have been a Spanish Christian who had converted to Islam, but for numerous reasons, this hypothesis seems unlikely.⁶⁹ Roger Collins suggests instead that the author was a Christian writing in Spain or North Africa, and that the rather favorable treatment of the Islamic leaders was a necessary condition of writing under Islamic rule. Since the Spanish Eastern Source generally avoids religious topics and limits its discussions of Islam strictly to political matters, it is conceivable that a Christian could have written it. The positive representation of Islam may simply reflect the need to appease the Islamic authorities.⁷⁰

While it is difficult to exclude completely the possibility that the Spanish Eastern Source was composed in the Islamic West, its production in the eastern Mediterranean, and Syria in particular, seems far more likely for a variety of reasons. Theodor Nöldeke was the first to propose this, arguing in an “Epimetrum” to Theodor Mommsen’s edition that this Spanish Eastern Source was most likely written in Greek by a Syrian Christian close to the center of Umayyad power.⁷¹ More recently, this position has been argued by Hoyland, who explains that the Spanish Eastern Source “must have been composed in Syria, since the Umayyad caliphs are each described in a relatively positive vein, all reference to ʿAlī is omitted, Muʿāwiya II is presented as a legitimate and uncontested ruler, and the rebel Yazīd ibn al-Muhallab is labelled ‘a font of wickedness.’”⁷² Moreover, the Spanish Eastern Source shares a number of parallels with the Byzantine chronicle tradition, and if we suppose its composition in Spain, it is difficult to explain the circulation of so many Byzantine sources in Spain (or North Africa for that matter) at this time. By contrast, it is much easier to imagine that a single Eastern historical source had reached eighth-century Spain, most likely written in Greek, as this was the most common language of cultural exchange between East and West at the time.⁷³ Hoyland additionally identifies a number of common features shared by this Spanish Eastern Source and the Syriac Common Source, a now lost chronicle written around 750 by Theophilus of Edessa, whose contents are known from the extant chronicles of Theophanes, Agapius, Michael the Syrian, and the Syriac Chronicle of 1234, all of which depend on the Syriac Common Source (see the discussion below). Hoyland suggests the possibility that perhaps these Spanish chroniclers made use of the same Greek translation of the Syriac Common Source that Theophanes must have used when composing his Greek chronicle at the beginning of the ninth century.⁷⁴ While he makes this proposal somewhat tentatively, such apparent connections further indicate an eastern Mediterranean origin for the Spanish Eastern Source. Although much admittedly remains uncertain, Nöldeke’s original suggestion of a Greek source written by a Syrian Christian still remains the most likely solution.

Of the two Spanish chronicles, the Byzantine-Arab Chronicle is generally regarded as the earlier, believed to have been written in 741. More accurately, however, this is not the date of the Byzantine-Arab Chronicle itself but is instead the date of the final entry from its eastern source. This would indicate that the Spanish Eastern Source, rather than the Byzantine-Arab Chronicle, was most likely produced in 741, while the Byzantine-Arab Chronicle was likely composed sometime later on the basis of this earlier source. For a western European chronicle of its time, the Byzantine-Arab Chronicle is rather peculiar in its overwhelming focus on events in the eastern Mediterranean, while devoting very little attention to either Spanish affairs or western Europe. According to Hoyland, only 9 percent of its contents concern Spanish affairs: there are six brief entries on the later Visigothic kings near the beginning (all taken from Isidore of Seville’s History of the Goths), a brief mention of the conquest of Spain later on, and, near the chronicle’s end, a description of the battle of Toulouse in 721.⁷⁵ Roughly one-third (29 percent) of the chronicle is devoted to Byzantine affairs, consisting of slightly more substantial notices regarding the Byzantine emperors from Phocas (610) to Leo III (717), although the reign of Heraclius alone commands approximately two-thirds of the total Byzantine material.⁷⁶ The majority of the chronicle, almost two-thirds of its total content (62 percent), focuses on Islamic history, with extended, favorable accounts of each ruler from Muhammad to Yazīd II (720–24).

Regarding Muhammad and the rise of Islam, the Byzantine-Arab Chronicle is remarkably favorable and free from polemic. As is the case in both chronicles, the account focuses largely on political matters, leaving religious affairs entirely to the side. Muhammad, however, is very clearly identified as the political leader of the Muslims at the time of the Islamic conquests of the Roman Near East. “When a most numerous multitude of Saracens had gathered together, they invaded the provinces of Syria, Arabia, and Mesopotamia, while one named Muhammad held the position of leadership over them [Syriae, Arabiae et Mesopotamiae prouincias inuaserunt supra ipsos principatum tenente Mahmet nomine]. Born of a most noble tribe of that people, he was a very prudent man and a foreseer of very many future events.”⁷⁷ After a brief description of the conquest of Syro-Palestine,⁷⁸ the chronicle notes Muhammad’s death and succession by Abū Bakr, who continued the conquests. “When Muhammad, the previously mentioned leader of the Saracens, had finished 10 years of rule, he reached the end of his life. [He is] the one whom they hold in such high regard and reverence until this day that they declare him to be the apostle and prophet in all their rituals and writings. In his place Abū Bakr of the Saracens (from which his predecessor also arose) was chosen by them. He organized a massive campaign against the Persians, which devastated cities and towns, and he captured very many of their fortifications.”⁷⁹ The entire passage is extraordinarily positive for a Christian chronicle written under Islamic occupation. It is rather peculiar, however, in its apparent division of the Islamic conquest of the Near East into two successive stages: the first stage was begun by Muhammad in the “provinces of Syria, Arabia, and Mesopotamia,” which context makes clear are Roman provinces, while the second stage commenced after Muhammad’s death, when Abū Bakr led a massive campaign of conquest against the Persian Empire.⁸⁰

This two-fold structure can perhaps be explained as the author’s attempt to harmonize two different accounts of the Islamic conquest of the Near East, one an older tradition ascribing leadership to Muhammad, witnessed in the Christian historical tradition, and the other an ostensibly emerging Islamic tradition that identified the beginning of the Near Eastern conquests with Abū Bakr’s reign. Roughly contemporary with the composition of the Spanish Eastern Source is the earliest Islamic biography of Muhammad, Ibn Isḥāq’s Sīra of the Prophet, compiled sometime not long before the author’s death in 767. According to Ibn Isḥāq’s seminal account, Muhammad died in 632 in Medina and was not involved in the conquest of Syro-Palestine, as discussed further in the following chapter. During the mid-eighth century then, an Islamic biography of Muhammad had begun to form in the eastern Islamic lands, where the Spanish Eastern Source was most likely composed, and, as Lawrence Conrad has demonstrated, some Christian historical writers appear to have had access to these nascent Islamic traditions and occasionally made use of them.⁸¹ The events of the Near Eastern conquests, however, were “only beginning to receive systematic historical attention” in the mid-eighth century, according to Conrad, and the Islamic historical tradition at this time could at best be characterized as “an emerging discipline.” Nevertheless, it would appear that the earliest Islamic traditions of Muhammad’s life and the Near Eastern conquests had possibly begun to circulate at this time, even though they may not have been written down yet, and some of these reports seem to have affected Christian historical writing of the period.⁸²

Thus it seems possible that the author of the Spanish Eastern Source may have been aware of emerging Islamic traditions reporting Abū Bakr’s leadership at the beginning of the conquests, and his two-stage account of the Islamic conquests could accordingly be understood as an effort to synthesize two divergent traditions that were circulating in his milieu.⁸³ An early tradition of Muhammad’s leadership during the Palestinian campaign has perhaps come into contact here with the traditional Islamic account of Abū Bakr’s leadership during the conquest of the Near East after Muhammad’s death. The author of the Spanish Eastern Source has possibly preserved both traditions and harmonized them by locating Muhammad’s leadership of a campaign against the Roman Near East slightly earlier in time, before the traditional date of his death, and then having Abū Bakr organize and execute the campaign against the Persian Empire only after Muhammad’s death. This solution results in a somewhat inaccurate chronology, in seeming to make the Islamic conquest of the Roman Near East commence somewhat earlier than it actually did, instead of extending the date of Muhammad’s death beyond its traditional date.⁸⁴ In any case, the Spanish Eastern Source clearly preserves the tradition of Muhammad’s leadership at the beginning of the Islamic conquest of Syro-Palestine.

The same division of the conquest into two stages is also preserved in the Hispanic Chronicle of 754, although this chronicle adopts a considerably different attitude toward Islam, and consequently, its preservation of the Spanish Eastern Source differs in some significant details. In comparison with the Byzantine-Arab Chronicle, the Hispanic Chronicle is rather polemical, occasionally adding derogatory comments and, more frequently, omitting material from the Spanish Eastern Source that portrays Islam too favorably, as can be seen especially by comparing the citations that follow with those above. This chronicle is also considerably longer than the Byzantine-Arab Chronicle, since it includes extensive material on the Visigoths and focuses much more squarely on the Iberian Peninsula, while drawing on the Spanish Eastern Source to set events in Spain within a more global context. Moreover, the Hispanic Chronicle continues its record of eastern events until approximately 750, prompting the suggestion that perhaps the Spanish Eastern Source originally continued to this point, and the Byzantine-Arab Chronicle has for some reason truncated its source in 741.⁸⁵ Yet it is not at all clear why the author of the Byzantine-Arab Chronicle would have done this, and so it is just as likely that the Hispanic Chronicle has somehow supplemented the Spanish Eastern Source with additional information from another source.

In its basic outline, the Hispanic Chronicle’s account of the rise of Islam and the Islamic conquests of the Near East largely repeats that of the Byzantine-Arab Chronicle, but it clearly has edited the Spanish Eastern Source to reflect much more negatively on Islam.

The Saracens rebelled in 618, the seventh year of the emperor Heraclius, and appropriated for themselves Syria, Arabia, and Mesopotamia, more through trickery than through the power of their leader Muhammad [Siriam, Arabiam et Mesopotamiam furtim magis quam uirtute Mammet eorum ducatore rebellia adortante sibi], and they devastated the neighboring provinces, proceeding not so much by means of open attacks as by secret incursions. Thus by means of cunning and fraud rather than power, they incited all of the frontier cities of the empire and finally rebelled openly, shaking the yoke from their necks. In 618, the seventh year of Heraclius, the warriors invaded the kingdom, which they forcefully appropriated with many and various consequences.⁸⁶

Like the Byzantine-Arab Chronicle, the Hispanic Chronicle follows with a brief description of the Islamic conquest of Palestine, after which it notes the death of Muhammad and his replacement by Abū Bakr: “When Muhammad had completed his tenth year, Abū Bakr, from his own tribe, succeeded to the throne, and he too launched major attacks against the power of the Romans and the Persians.”⁸⁷ Excepting the marked difference in tone, this report is remarkably similar to the Byzantine-Arab Chronicle’s description of the same events. Muhammad is clearly identified as the leader of the Muslims at the time of the initial campaign in Palestine, and there seems to be a two-fold structure to the campaign, beginning in Syro-Palestine and then expanding into Persia during the reign of Abū Bakr.

The most significant difference between these two accounts of the Islamic conquests is the Hispanic Chronicle’s indication that Abū Bakr led attacks against both the Romans and the Persians, in contrast to the Byzantine-Arab Chronicle, which describes Abū Bakr’s massive new campaign against the Persians only. Yet this is not a particularly serious discrepancy, and in actuality it does not contradict the two-stage presentation of the Arab conquests found in the Byzantine-Arab Chronicle and, almost certainly, in the Spanish Eastern Source as well. Presumably, the author of the Hispanic Chronicle reflects here the fact that Islamic military operations against the Roman Empire did not cease with the conquest of Syro-Palestine. After taking control of Syria and Palestine, the Arabs continued to make advances against the Byzantines, proceeding to conquer Egypt, North Africa, and eastern Anatolia, and laying siege to Constantinople itself in 674.⁸⁸ Nonetheless, the Hispanic Chronicle clearly presents the Persian campaign as something begun only in the reign of Abū Bakr, and thus as a second stage in the Islamic conquest of the Near East. More importantly, the Hispanic Chronicle’s adherence to a two-phase description of the conquest is indicated again in its summation of the earliest Islamic conquests, which follows immediately after the notice concerning Abū Bakr. “After the tenth year of Muhammad’s rule had expired in 628, in the seventeenth year of the emperor Heraclius, they chose the aforementioned Abū Bakr, of Muhammad’s own tribe, in his place, and the Arabs fought with sword against Persia, which had been abandoned by the Roman empire. Abū Bakr ruled for almost three years, powerfully waging war.”⁸⁹ Here the Hispanic Chronicle mentions Persia specifically in connection with Abū Bakr, essentially identifying the beginning of the campaign against Persia with the commencement of his reign. Furthermore, this passage bears a striking similarity to the Byzantine-Arab Chronicle’s description of Abū Bakr’s reign, making it rather probable that this two-stage account of the Islamic conquests was present in the Spanish Eastern Source. Most likely then, this lost Greek chronicle described the initial Islamic assault on Palestine and Syria as occurring under Muhammad’s leadership, while presenting the assault of Persia as a second stage in the conquests that commenced under Abū Bakr. Thus, the Hispanic Chronicle’s attribution of attacks against the Romans to Abū Bakr is best understood as an addition by its author, who no doubt was aware that conflict between the Byzantines and Muslims continued into the first caliph’s reign and beyond.

In conclusion then, the Spanish Eastern Source was most likely a Greek chronicle written in Syria, sometime very close to 741. It is clear that this chronicle described the Islamic conquest of Syro-Palestine under Muhammad’s leadership and then represented the conquest of Persia as a second stage of the conquests that commenced under Abū Bakr. This two-fold conception of the Islamic conquests possibly reflects an effort to reconcile an earlier tradition of Muhammad’s leadership during the assault on Palestine with an emerging Islamic tradition that separated Muhammad from the Near Eastern conquests and identified their beginning with the reign of Abū Bakr. We do not know the source of the Spanish Eastern Source’s information regarding the Islamic conquests and Muhammad’s role therein, but given its later date and its apparent connections with the Eastern chronicle tradition, its report may derive from earlier literary sources.

There is no indication that this account of the rise of Islam has been doctored to suit any grand narrative, and perhaps most remarkably there is no trace of any apology or polemic in the Spanish Eastern Source, at least insofar as it is represented by the Byzantine-Arab Chronicle. There is hardly any reason to suspect that the redactor of the latter document was responsible for this favorable depiction of Islam, particularly in view of the comparative data afforded by the Hispanic Chronicle. In fact, so positive is the Spanish Eastern Source’s view of Islam that it is tempting to suspect that somehow there are Islamic sources lying just behind it. Perhaps some now lost early Islamic (Umayyad?) historical traditions also preserved a primitive tradition of Muhammad’s leadership during the Palestinian campaign, such as we find attested in the non-Islamic sources. The Letter of ʿUmar discussed below certainly suggests this possibility. Furthermore, as noted above, the remarkably positive representation of Islam and its early leaders in the Spanish Eastern Source probably reflects an expectation of scrutiny by Islamic readers. In light of this, it seems rather unlikely that its author would either deliberately misrepresent Muhammad as the leader of the Palestinian conquests or would include information widely regarded as false by the Islamic authorities. This source in fact seems to be very close to the center of Umayyad power, and its use by these two early medieval Spanish chroniclers demonstrates not only that the tradition of Muhammad’s leadership during the campaign in Palestine remained current in Christian historical writing over a century after the events themselves but also that this tradition had spread even to the West in early Islamic Spain.

The Syriac Common Source: The Chronicle of Theophilus of Edessa (ca. 750 CE)

The Syriac Common Source is a now lost medieval chronicle that we have already mentioned briefly in discussions of the Doctrina Iacobi and the Spanish Eastern Source. The first traces of this vanished chronicle began to emerge in the later nineteenth century, when it was discovered that the Greek chronicle of Theophanes (written 814 CE) and the Syriac chronicle of Michael the Syrian (written 1195 CE) had used a common source in compiling their notices for the seventh century and much of the eighth, the so-called Eastern Source, or Syriac Common Source, as we have determined to call it. In Michael’s case, it was further known that he had used this lost source at second hand, as it had been mediated to him through yet another lost chronicle, the Chronicle of Dionysius of Tellmahre (d. 845), which Michael implies was the only substantial source available to him for the seventh and eighth centuries.⁹⁰ The subsequent publication of the Christian Arabic chronicle of Agapius (written ca. 940) and the anonymous Syriac Chronicle of 1234 have added further clarity to the picture. Agapius depends almost entirely on the lost Syriac Common Source for his description of events during the years 630–754, providing now a third independent witness to this missing source.⁹¹ The Chronicle of 1234, in contrast, presents a second source that has drawn its seventh- and eighth-century material almost exclusively from Dionysius of Tellmahre’s lost chronicle, preserving its contents in what many think is a less heavily edited version than is found in Michael’s chronicle. Since Dionysius’s chronicle is believed to have best preserved the Syriac Common Source, this anonymous thirteenth-century chronicle is an invaluable resource for reconstructing the contents of this now lost text.⁹²

All of this makes determining the contents of the Syriac Common Source a rather complex and at the same time fairly straightforward endeavor. Since it is generally assumed that the Chronicle of 1234 has most faithfully preserved the Syriac Common Source, via Dionysius of Tellmahre’s vanished chronicle, one begins by looking at this chronicle, but at each point, one must also compare the data from Theophanes, Agapius, and Michael. Only after evaluating the various testimonies from all of these sources both with one another and with the tendencies of each individual chronicle can one come to a judgment as to what the Syriac Common Source most likely reported. When several sources converge very closely, we can be quite certain that this material has been faithfully preserved from the Syriac Common Source. By this means, an outline of this lost chronicle can be restored, as evidenced in Hoyland’s very helpful summary of its contents.⁹³ Moreover, we now know the author of this important history of the seventh and eighth centuries to have been Theophilus of Edessa, an eighth-century Maronite scholar who served as court astrologer to the ʿAbbāsid caliph al-Mahdi.⁹⁴ Theophilus is said to have written several works on astrology, and his knowledge of Greek was such that he translated the Iliad and perhaps the Odyssey into Syriac, but all of these works are now lost, except for a few surviving fragments and excerpts. Most importantly for the present purposes, however, Theophilus also composed a chronicle, which, as Conrad has convincingly demonstrated, is almost certainly to be identified with the lost Syriac Common Source.⁹⁵

Unfortunately, Theophilus of Edessa’s account of Muhammad’s life and the rise of Islam is somewhat difficult to determine, since the various witnesses to his Chronicle themselves preserve different descriptions of these events. Hoyland nicely summarizes the situation as follows: “Theophanes almost totally ignores Theophilus for his notice on Muhammad, drawing instead, indirectly, on Jewish and Muslim sources. Agapius abridges Theophilus, as he himself acknowledges, and supplements him with material from the Muslim tradition. That leaves Dionysius, who seems to me to best preserve Theophilus’ entry.”⁹⁶ Luckily, Dionysius’s account of the rise of Islam is well preserved in both Michael the Syrian’s Chronicle and the Chronicle of 1234: the two are either identical or very close in wording at this point. Michael’s text does contain a few passages not found in the Chronicle of 1234, many of which are polemical in nature, but these are more likely to have been added by Michael than deleted by the latter.⁹⁷ Thus we may with some confidence regard the following passage from the Chronicle of 1234 as representing something very close what once stood in Dionysius’s Chronicle, and in turn as reflecting more or less what Dionysius likely found in Theophilus’s now lost mid-eighth-century Chronicle.

Therefore this Muhammad, while in the measure and stature of youth, began to go up and come down from his city Yathrib to Palestine for the business of buying and selling. And while he was engaged in this region, he encountered the belief in one God, and it was pleasing to his eyes. And when he went back down to the people of his tribe, he set this belief before them, and when he persuaded a few, they followed him. And at the same time he would also extol for them the excellence of the land of Palestine, saying that “Because of belief in the one God, such a good and fertile land has been given to them.” And he would add, “If you will listen to me, God will also give you a fine land flowing with milk and honey.” And when he wanted to prove his word, he led a band of those who were obedient to him, and he began to go up and plunder the land of Palestine, taking captives and pillaging. And he returned, laden [with booty] and unharmed, and he did not fall short of his promise to them.

Since the love of possessions drives such behavior to become a habit, they began continually going out and coming back for plunder. And when those who were not yet following him saw those who had submitted to him becoming wealthy with an abundance of riches, they were drawn to his service without compulsion. And when, after these [raids], the men following him became numerous and were a great force, he no longer [went forth but] allowed⁹⁸ them to raid while he sat in honor in Yathrib, his city. And once they had been sent out, it was not enough for them to remain only in Palestine, but they were going much further afield, killing openly, taking captives, laying waste, and pillaging. And even this was not enough for them, but they forced them to pay tribute and enslaved them. Thus they gradually grew strong and spread abroad, and they grew so powerful that they subjugated almost all the land of the Romans and the kingdom of the Persians under their authority.⁹⁹

The indication that the initial Islamic attacks on Palestine began during Muhammad’s lifetime and under his leadership is quite clear here, and comparison with Michael’s Chronicle confirms that Dionysius must have written something very similar in his early ninth-century Chronicle. Since Dionysius is believed to best preserve Theophilus’s lost chronicle, it is further likely that this account bears a strong resemblance to Theophilus’s description of the rise of Islam. Nevertheless, Theophanes and Agapius are not able to confirm the presence of this report in Theophilus’s Chronicle, since they have both utilized other sources in their descriptions of the rise of Islam.¹⁰⁰ Fortunately, another source is available to verify that Theophilus’s Chronicle almost certainly contained a passage similar to the one above and, more importantly, that it described Muhammad’s leadership during the initial phase of the conquest of Palestine. The East Syrian Chronicle of Siirt, written in Arabic during the tenth century, also depends on Theophilus’s lost Chronicle for its knowledge of many early seventh-century events, including the rise of Islam in particular.¹⁰¹ Although many details found in Dionysius’s account do not appear in the Chronicle of Siirt, the latter similarly indicates Muhammad’s leadership during the initial assault on the Roman Near East in a report that almost certainly depends on Theophilus’s earlier Chronicle. The Chronicle of Siirt begins its account of the rise of Islam by introducing Muhammad’s appearance among the Arabs and briefly describing his religious teachings. Then it continues to relate the events of the conquests: “And Muhammad ibn ʿAbdullah was a strong and powerful leader. In the eighteenth year of Heraclius [627/28], Emperor of the Greeks, the year in which Ardasir the son of Siroe the son of Khosro Parvez reigned [629/30], the Arabs began their conquests, and Islam became powerful. And after that Muhammad no longer went forth in battle, and he began to send out his companions.”¹⁰² This passage is obviously much more terse than the account in Dionysius’s Chronicle, but it is sufficient to confirm that the Chronicle of Theophilus, which was their common source, described Muhammad as initiating Near Eastern conquests and then withdrawing, entrusting the command during further expansions to others among his followers.

It would seem that Theophilus has perhaps here also combined two separate traditions about Muhammad’s relation to the Near Eastern conquest: one reporting his direct involvement, as indicated in the first section, and a second that remembered Muhammad as remaining behind, sending forth his followers instead to assault the Roman and Persian empires. Quite possibly, this structure reflects an effort to merge the divergent accounts of the Christian historical tradition with the early biographies of Muhammad that were just beginning to emerge at this time. As Conrad has demonstrated, Theophilus appears to have had access to the nascent Islamic historical tradition in some form, and one would imagine that this was the source of his second tradition separating a still-living Muhammad from later events of the conquests.¹⁰³ Thus, in a schema that offers an intriguing parallel to the Spanish Eastern Source, which also seems to have had knowledge of the early Islamic historical tradition, Theophilus has possibly harmonized these disparate memories according to a two-stage narrative of the Islamic conquests that begins with Muhammad’s leadership of the initial attacks on Palestine and then is followed by his withdrawal to Medina and a more extensive conquest of the Near East after his death under the leadership of Abū Bakr. Like the Spanish Eastern Source, Theophilus achieves this structure by advancing the onset of the Islamic conquests several years in order to place the initial Islamic attacks on Palestine within the traditional lifespan of Muhammad, that is, before 632, a date that Theophilus may also have learned from his Islamic sources.¹⁰⁴

Theophilus’s Chronicle is certainly not free from polemic in its description of the rise of Islam, at least if the Chronicle of 1234 at all represents his account accurately. The earliest followers of Muhammad are depicted as being interested only in plunder, and their successful conquest of the Near East is ultimately accredited to their excessive greed. Moreover, Muhammad’s early travels to Palestine as a merchant are clearly linked with a greater narrative having an apologetic agenda. These trips introduced him to the monotheistic beliefs of the Jews and Christians living there, and the chronicle identifies these as the source of his religious inspiration. The clear implication seems to be that Islam represents nothing more than a rehashing of the Judeo-Christian monotheistic traditions that Muhammad picked up during his visits to Palestine. Nonetheless, Muhammad’s leadership during the conquest of Palestine plays no discernable role in this polemical narrative of Islamic origins: only Muhammad’s travels to Palestine as a merchant are enlisted to mark Islam as derivative of Judaism and Christianity. Moreover, in contrast to his followers, Muhammad does not act out of greed but instead because of his devotion to the monotheistic traditions that he encountered in Palestine. Thus, his leadership of the initial attacks on Palestine is not ascribed to the covetous motives of his followers but instead to a prophetic call to lead them to the land of divine promise. In any case, Theophilus’s identification of Muhammad as alive and leading the initial assaults on Palestine is clear, and the fact that he preserves this tradition perhaps in the face of new information issuing from the nascent Islamic historical tradition is a testament to how deeply engrained the tradition of Muhammad’s leadership during the Palestinian campaign remained in Christian historiography approximately one century after the events.

The Short Syriac Chronicle of 775 (ca. 775 CE)

Among several short Syriac chronicles from the eighth century is an anonymous chronicle sometimes known by the title that it bears in the unique manuscript preserving it: “An Account of the Generations, Races, and Years from Adam until the Present Day.” This chronicle runs very quickly through the main events and figures of the Bible, following these with a list of Roman emperors and the length of their reigns. When it reaches the seventh century, the chronicle interrupts the reign of Heraclius with a brief mention of the Islamic conquests; then it continues to give a list of the early Islamic rulers and the number of years that each reigned, up until the accession of the caliph al-Mahdi in 775, which is the likely date of the chronicle’s completion. The chronicle’s transition from Roman and Muslim authorities, which hinges on the Islamic conquests, is related as follows:

Maurice, 27 years and 6 months; Phocas, 8 years; Heraclius, 24 years. In the year 930 of Alexander, Heraclius and the Romans entered Constantinople. And Muhammad and the Arabs went forth from the south and entered the land and subdued it []. The years of the Hagarenes and the time when they entered Syria and took control, from the year 933 of Alexander. Each one of them by name as follows. Muhammad, 10 years; Abū Bakr, 1 year; ʿUmar, 12 years; ʿUthman, 12 years; no king, 5 years; Muʿāwiya, 20 years; Yazīd, his son, 3 years; no king, 9 months; Marwan, 9 months; ʿAbd al-Malik, 21 years; Walīd, his son, 9 years; Sulaymān, 2 years and 7 months; ʿUmar, 2 years and 7 months; Yazīd, 4 years, 10 months, and 10 days.¹⁰⁵

Unfortunately, the chronicler’s knowledge of early seventh-century chronology was rather poor. As Palmer writes, “This text is full of oddities. Of the Byzantine emperors only Phocas reigned for a period approximately equivalent to that shown here. Of the Arab caliphs Abū Bakr is curtailed and ʿUmar I is prolonged.”¹⁰⁶ Perhaps the most peculiar item of all, however, is the implication that the Islamic conquest of Palestine took place in the year 618/19. While some of the Christian historical sources place the Islamic conquests before 632, none of them locates it this early: the date precedes even the hijra by three years. Strangely enough, however, a Syriac inscription from a north Syrian church dated to 780 bears the same information: “In the year 930 the Arabs came to the land.”¹⁰⁷ Although Palmer and Hoyland both speculate as to possible explanations for this date, it remains a mystery. Nonetheless, for the present purposes the text is clear: in spite of its rather idiosyncratic dating, this short chronicle identifies Muhammad as leading the Islamic invasion of the Roman Near East. While the source of this information is completely unknown, it is conveyed without polemic and in the absence of any sort of apologetic agenda or totalizing explanation.

The Zuqnin Chronicle (ca. 775 CE)

Roughly contemporary with the preceding text is an anonymous chronicle written at the monastery of Zuqnin near Amida (modern Diyarbakır) sometime around 775. Unfortunately, these two chronicles have more in common than just their date of composition: the Zuqnin Chronicle’s chronology is also very weak during the period of the Islamic conquests. In fact, its author warns his readers that he was unable to find reliable sources for most of the seventh and eighth centuries: “From that point (574 CE) up to the present year (775 CE) … I have not found [a history] concerning events which is composed on such solid foundations as the former ones [that is, Eusebius, Socrates, John of Ephesus].”¹⁰⁸ In view of the author’s own awareness of the rather poor sources at his disposal, one can hardly fault him for his mistakes in chronology.¹⁰⁹ In describing the rise of Islam, the Zuqnin chronicler, in spite of his expectedly weak chronology, nevertheless maintains the tradition of Muhammad’s leadership during the invasion of Palestine:

In 621 the Arabs conquered the land of Palestine all the way to the Euphrates River, and the Romans fled and crossed over to the east of the Euphrates, and the Arabs ruled over them in it [that is, Palestine]. Their first king was a man from among them whose name was Muhammad. They also called this man a prophet, because he turned them away from cults of every sort and taught them that there is one God, the maker of creation. And he established laws for them, because they were especially devoted to the worship of demons, the veneration of idols, and especially the veneration of trees. And because he had shown them the one God, and they had defeated the Romans in battle under his leadership [], and he had established laws for them according to their desire, they called him a prophet and a messenger of God.¹¹⁰

The entry for this year concludes with some brief polemical remarks accusing the Arabs of being “an especially greedy and carnal people,” who follow only such laws as suit their desires.¹¹¹

Excepting these final remarks, the Zuqnin Chronicle’s account of the rise of Islam is relatively free from polemic: it does not serve any obvious apologetic agenda and is not linked with any sort of totalizing explanation. With regard to Abū Bakr, the chronicle notes only his death and the length of his rule (five years), without any indication of his involvement in the conquest of the Near East. The conquests are not mentioned again until the second year of ʿUmar’s reign, when “the Roman Emperor Heraclius went down to Edessa, and the battle of Gabitha took place, and the Persians were defeated and they left Mesopotamia.” Following this is a notice that four years later the Arabs crossed into northern Mesopotamia and defeated the Romans there.¹¹² The chronology is in fact rather chaotic here as elsewhere in the chronicle, but its identification of Muhammad as alive and leading the Muslims during the conquest of Palestine is unmistakable.

A Report from the Continuatio of Abū l-Fatḥ’s Samaritan Chronicle (seventh century?)

Among the sources signaled by Crone and Cook as witnessing to Muhammad’s leadership of the assault on Palestine is the Samaritan Chronicle compiled by Abū l-Fatḥ al-Sāmirī al-Danafī at only the rather late date of 1355.¹¹³ Yet despite the comparative youth of this collection, it is widely acknowledged that Abū l-Fatḥ’s chronicle assembles much earlier material from a variety of older sources, several of which Abū l-Fatḥ identifies at the beginning of his composition.¹¹⁴ Commissioned by the Samaritan high priest, the chronicle of Abū l-Fatḥ spans the period from Adam through the appearance of Muhammad, and it is generally regarded as one of the most important sources for the history of the Samaritan people. Although it was compiled only relatively recently, this chronicle is broadly recognized as preserving a great deal of much older material.¹¹⁵ Abū l-Fatḥ’s original composition concluded with Muhammad’s appearance, drawing to a close with a Samaritan version of the Baḥīrā story, an Islamic legend according to which a Christian monk named Baḥīrā met the young Muhammad and identified him as a prophet on the basis of a distinctive birthmark on his back. In Abū l-Fatḥ’s version, three astrologers, a Jew, a Christian, and a Samaritan, discerned Muhammad’s appearance from the stars, and traveling together to his hometown, they each spoke with the young man, but it was (of course) the Samaritan who identified the sign on his back.¹¹⁶ Immediately thereafter, Abū l-Fatḥ’s chronicle appends a list of Samaritan high priests up until the appearance of Muhammad, concluding with the date at which the chronicle was completed.¹¹⁷ Nevertheless, several of the most important manuscripts continue beyond this point, extending the narrative either to the reign of Hārūn al-Rashīd (786–809) or, in one manuscript, until the time of the caliph al-Rāḍī (934–40). There is a clear consensus that Abū l-Fatḥ’s chronicle came to a close with Muhammad’s discovery by the three astrologers,¹¹⁸ and thus the account of the Islamic conquests often preserved in this Samaritan chronicle was not originally part of Abū l-Fatḥ’s late medieval compilation. Rather, these reports belong to another anonymous Samaritan chronicle, known as the Continuatio, that has been appended to Abū l-Fatḥ’s composition to extend its scope into the early Islamic period.

This Continuatio has recently been translated and subject to careful historical analysis by Milka Levy-Rubin, who determines that despite its distinction from Abū l-Fatḥ’s original compilation, the Continuatio is in fact a particularly important source for the history of Palestine in the early Muslim period.¹¹⁹ Levy-Rubin translates the most complete version of the Continuatio, known from only a single manuscript, which ends with the rule of al-Rāḍī, and her arguments for the value of this unique witness are convincing. The manuscript is reproduced following the translation, in lieu of an edition. Nevertheless, in the section covering the period between the Islamic conquests and the reign of Hārūn al-Rashīd, for which additional witnesses exist, Levy-Rubin has made comparative use of the other relevant manuscripts, as reflected in her extensive critical annotations. Even though almost nothing is known regarding the provenance or date of this nameless chronicle, Levy-Rubin’s careful analysis has demonstrated the exceptional value of its witness to the history of early medieval Palestine.

The Continuatio opens with the events of the Islamic conquest of Palestine, and it names Muhammad as a key participant in the assault. Immediately after the “Baḥīrā” legend from Abū l-Fatḥ’s chronicle, the Continuatio describes the Arab invasion and its consequences for the Samaritans in some detail.

After this the Ishmaelites, Muḥammad and all his army, went forth to wage war against the Byzantines; they conquered the land and defeated the Byzantines and killed them as they fled before them. The imām¹²⁰ in those days was ʿAqbūn ben Elʿazar, who lived in Bayt Ṣāma. When the Muslims attacked and the Byzantines fled, all of the Samaritans who lived along the coast fled with the Byzantines from the advancing Muslims, [thinking] that they would return. When the Samaritans began to leave with the Byzantines for Byzantium (Rūmīya), they came to the raʾīs ʿAqbūn ben Elʿazar, to Bayt Ṣāma, because he lived there, and said to him, “You are a trustworthy man, so we will deposit our possessions with you until we return,” thinking that they would be returning soon…. The people who deposited [their wealth] were the people of Caesarea, Arsūf, Maioumas, Jaffa, Lydda, Ascalon, Gaza, and all of the interior villages and those along the coast. And after this they left for Byzantium and remained there and have not returned to this day. The Muslims rose and entered the land of Canaan, and took control of it; they seized all the cities and inhabited them, and ruled over all the places until there was no place left which they had not taken over but Caesarea, which rebelled and did not submit to them because it was called the mother of cities and took precedence over them. [The Muslims] set up camp against it and besieged it for six years before they conquered it…. After they captured it, every place else stood in awe of them.¹²¹

As Levy-Rubin observes, this account has much to recommend it, and even at considerable historical distance from the events in question its verisimilitude is impressive.¹²² Excepting only the indication that Muhammad participated in the assault, which Levy-Rubin regards as an error adopted from the Syriac chronicle tradition,¹²³ the details of this narrative comport well with the current understanding of how the conquest of Palestine unfolded. The Continuatio reports that while the Samaritans living on the coast felt threatened by the invaders and fled with the Byzantines, the inland areas were not as disrupted by the incursion: in fact, the region was sufficiently tranquil that the coastal Samaritans decided to entrust their belongings to the high priest living there. This description agrees with the apparent concentration of the Arab forces on the Byzantine cities along the coast, and the decision by many inhabitants to abandon their cities rather than offer resistance is consistent with the increasing recognition that the conquest of Palestine was largely a nondestructive affair.¹²⁴ Both literary evidence and the archaeological record suggest a picture of the Arab takeover as a mostly peaceful transition: numerous recent excavations have revealed “no sign of any traumatic break or crisis in the seventh century” that would indicate a pitched struggle for control of the region.¹²⁵ Moreover, the Continuatio’s indication that Caesarea in particular offered fierce resistance to the invaders is also confirmed by other sources, which describe the city’s capture only after a long and arduous siege, as reflected in the text.¹²⁶

More importantly, as Levy-Rubin notes, the author of this account “seems to have been familiar with the layout of the Byzantine city [that is, Caesarea], and was well informed about the story of its conquest.”¹²⁷ Such knowledge of the city’s plan as it existed during the Byzantine period is an impressive indication that this account was likely written by someone very close to the events described, perhaps with firsthand knowledge of what he relates.¹²⁸ This determination comports with the broader character of the Continuatio, whose reports generally exhibit “close proximity, both in time and place, to the events described in the text,” often seeming to relate accounts provided by firsthand witnesses.¹²⁹ Although it is not known when or by whom this chronicle was first stitched together, its individual reports, as Edward Vilmar was the first to observe, appear to be contemporary with the events that they describe.¹³⁰ Comparison with the Islamic historical tradition reveals the Continuatio to be a reliable source in general, but with regard to events and activities in early Islamic Palestine, this Samaritan chronicle offers a unique source of particularly “detailed and trustworthy information.”¹³¹

In view of the Continuatio’s overall quality as a historical source, and the general credibility of its description of the conquest of Palestine more specifically, one should perhaps reconsider Levy-Rubin’s somewhat hasty dismissal of its report concerning Muhammad’s involvement in the initial invasion. Levy-Rubin rejects this notice simply out of hand, on the basis that it contradicts the Islamic historical tradition, which consistently reports Muhammad’s death prior to the assault on Palestine. Inasmuch as the Samaritans used a dialect of Aramaic as their primary language during the early Middle Ages, she proposes that this “mistake” owes itself to Samaritan knowledge of the Syriac historical tradition. Yet she does not elsewhere show evidence of influence from the Syriac tradition, nor does the Continuatio manifest any significant dependence on Christian historiography. Quite to the contrary, Levy-Rubin frequently appeals to the Continuatio’s independence and the uniqueness of its witness as evidence of its exceptional importance. To be sure, the Continuatio knows the same tradition regarding Muhammad’s participation in the invasion of Palestine that is reflected in the Christian sources and the Shimʿōn b. Yoḥai complex. Nevertheless, there is no evidence to suggest that the Continuatio’s knowledge of this tradition is contingent on any of these other texts. Instead, the Continuatio seems to be an independent witness to this early tradition, which appears to have circulated among the different religious communities of early Islamic Palestine and the Near East more generally. Such an assessment fits well with the detailed and local character of the Continuatio’s report, and the apparent credibility of this account of the Palestinian conquest on other points invites some confidence in its notice of Muhammad’s involvement. If this remark were merely the isolated witness of an anonymous Samaritan chronicle, it would rightly be disregarded. But when placed in the context of these other sources, it seems that the Continuatio confirms their collective witness, and together with the apocalypse of Rabbi Shimʿōn b. Yoḥai, it offers important evidence that this tradition was not simply a collective delusion of Christian historiography.

The Continuatio’s account of the Arab conquest is surprisingly free from polemic, and it does not interpret either the Muslim invasion or Muhammad’s participation in it according to some apologetic interest or a totalizing narrative. On the whole, the Continuatio is quite favorable to the Arabs, and as Levy-Rubin observes, it exhibits a “positive evaluation concerning both conditions in Palestine during the Umayyad period and the positive attitude of these rulers towards the local population.”¹³² The Arab expulsion of the Byzantines is described with approval, and the terms of Islamic governance are met with neutral acceptance. Of Muhammad, the Continuatio says, rather astonishingly, that “the prophet of Islam did not cause anyone distress throughout his life. He would present his belief before the people, accepting anyone who came to him, [yet] not compelling one who did not.” His immediate successors, the chronicle continues, ruled “according to what he had enjoined upon them; they did no more or less, and did not harm anyone.”¹³³ It is a portrait of Islam’s emergence within Palestine that comports rather well, as Levy-Rubin notes, with what can otherwise be known about this period.¹³⁴ Indeed, it is difficult to imagine that such a favorable account would have been composed much beyond the first several decades of Islamic rule, after which social and economic pressure on the dhimmis (that is, non-Muslim peoples) was increased. Consequently, when all the relevant factors are taken into consideration, this view of the early Islamic conquests from Samaria has much to recommend it, and its notice of Muhammad’s involvement during the invasion warrants its inclusion alongside these other early witnesses to this tradition.

An Early Islamic Witness: ʿUmar’s Letter to Leo (Eighth Century)

Important confirmation of this tradition of Muhammad’s leadership during the invasion of the Near East emerges from a recently rediscovered early Islamic text, the alleged letter from the caliph ʿUmar II (717–20) to the Byzantine emperor Leo III (717–41). This letter was already known, albeit somewhat indirectly, from a précis of ʿUmar’s correspondence composed by the Armenian chronicler Łewond in his eighth-century History.¹³⁵ Other historical sources, including the chronicles of Theophanes and Agapius make reference to this epistle, which ʿUmar purportedly sent in hopes of converting the emperor, but the original text was long presumed lost.¹³⁶ Leo’s “reply,” however, has been known since the beginning of the sixteenth century, when a brief Latin translation made from “Chaldean” (presumably Arabic) was first published.¹³⁷ The full extent of Leo’s letter subsequently came to light only in Łewond’s History, where it follows his summary of ʿUmar’s letter. This Armenian translation of Leo’s letter is rather lengthy, and alone it amounts to more than one-fourth of Łewond’s chronicle.¹³⁸ Its size not only revealed the Latin translation to be a mere summary of Leo’s letter but also invited suspicions that the original version of ʿUmar’s letter was likely of similar extent.

Fortunately, the complete text of ʿUmar’s letter has recently come to light, having been pieced together from two partial manuscripts in different languages by Jean-Marie Gaudeul.¹³⁹ The second half of ʿUmar’s letter was the first to be discovered, but since this fragment lacks the opening epistolary framework, the nature of this early Islamic text was not immediately recognized. In the mid-1960s, Dominique Sourdel found among a collection of materials from Damascus at the Turkish and Islamic Arts Museum in Istanbul ten stray parchment folios containing an Arabic text that appeared to be quite old. Sourdel published the text as an “Anonymous Muslim Pamphlet” against the Christians, and on the basis of the manuscript itself and the contents of the text, he convincingly argued for its composition sometime before the end of the ninth century.¹⁴⁰ Not long thereafter, Denise Cardaillac published a manuscript of Muslim anti-Christian polemics from the National Library of Madrid that includes the beginnings of a letter ascribed to ʿUmar, written to “Lyon, king of the Christian infidels.”¹⁴¹ Like the other polemics of this collection, the letter survives in Aljamiado, that is, a Romance dialect written using the Arabic script. Cardaillac compared this letter with Arthur Jeffrey’s translation of ʿUmar’s letter in Łewond’s History, and, believing that Łewond’s version was in fact the original, she concluded that the Aljaimado text had been more recently composed by Moriscos, using ʿUmar’s letter as a basis and expanding it considerably.¹⁴² Clearly, however, Łewond gives merely a “summary” (), as he himself says, of ʿUmar’s letter, and thus his account cannot form a reliable basis for such judgments.¹⁴³

Gaudeul first came to suspect that Sourdel’s “Anonymous Pamphlet” should be identified with ʿUmar’s letter after comparing Leo’s letter in Łewond very broadly with early Islamic polemical writings against the Christians from the ninth and tenth centuries. Gaudeul noted that many of the same themes and even similar expressions were found in both Leo’s letter and the Anonymous Pamphlet, leading him to conclude that these two texts were in dialogue with one another and, by consequence, that the Anonymous Pamphlet was indeed the second half of ʿUmar’s lost letter.¹⁴⁴ This hunch was confirmed unmistakably when Gaudeul began to compare ʿUmar’s Aljaimado letter with Leo’s letter in Łewond. At first Gaudeul began to notice connections between the Aljaimado text and Leo’s letter that were similar in nature to the former’s parallels with the Anonymous Pamphlet. Then, in the final pages of the Aljaimado letter, Gaudeul found that its contents suddenly began to overlap with the first few pages of the Anonymous Pamphlet and that their contents were nearly identical.¹⁴⁵ This discovery revealed that the Aljaimado text was in fact no Morisco forgery but instead a very faithful translation of this early Arabic text, validating Gaudeul’s identification of the Anonymous Pamphlet with ʿUmar’s lost letter. Thanks to Gaudeul’s meticulous research, ʿUmar’s letter has now been recovered from these two manuscripts, thus restoring the other side of this interreligious debate from the early medieval Near East.

On the basis of this newly recovered text, Hoyland has introduced some important refinements to the dating of this early Islam polemic. Although Gaudeul largely follows Sourdel’s initial dating of the Anonymous Pamphlet in assigning ʿUmar’s letter to the late ninth century,¹⁴⁶ Hoyland’s more thorough analysis of the Leo-ʿUmar tradition complex convincingly identifies the eighth century as the likely milieu for this epistolary contest.¹⁴⁷ First, Hoyland answers Stephen Gerö’s proposal that Łewond’s letter of Leo is a medieval Armenian forgery added to the text by a later reviser. According to Gerö, Łewond’s History as we now have it is the work of an eleventh- or twelfth-century redactor, who heavily revised a now lost chronicle that was actually written by Łewond in the late eighth century. Among his amendments was the introduction of this epistolary exchange, inspired by brief mention of such correspondence in Thomas Artsruni’s early tenth-century Armenian chronicle.¹⁴⁸ Łewond’s editor, however, wanted to incorporate a more detailed account of the Leo-ʿUmar correspondence than he found in his source.¹⁴⁹ Consequently, Gerö postulates that the redactor took an existing Armenian anti-Islamic polemical tract and reshaped it to create the illusion of an exchange of letters. The scheme involved forging a letter from ʿUmar that corresponded with the main points of the anti-Islamic treatise and then “lard[ing] the Christian tract with allusions to the ʿUmar letter.”¹⁵⁰

Gerö’s theories regarding Łewond’s History and the Leo-ʿUmar correspondence in particular have not found much acceptance. Experts on the Armenian historical tradition continue to regard Łewond’s chronicle as an authentic work of the late eighth century, and its genuine witness to an early tradition of a polemical exchange between Christians and Muslims in the guise of letters authored by Leo and ʿUmar seems widely conceded.¹⁵¹ Nevertheless, Hoyland responds to each of Gerö’s arguments point by point and convincingly demonstrates both that Łewond’s History as we now have it is a work of the late eighth century and that his version of Leo’s letter is not an adaptation of an Armenian work, but in fact translates an older Greek text that was part of an early tradition of epistolary polemic between Muslims and Christians.¹⁵² Gaudeul’s study, published subsequent to Gerö’s work, is particularly decisive in this regard. Gaudeul’s recovery of ʿUmar’s letter leaves Gerö’s scenario rather improbable, and the close rhetorical connections between this Muslim text and Łewond’s account of the correspondence suggests that they reflect an actual polemical exchange between Christians and Muslims in the early medieval Near East.¹⁵³

Although Gaudeul (and Sourdel) would locate this exchange as late as the end of the ninth century, the date of Łewond’s chronicle, the late eighth century, would seem to indicate that it had reached a fairly mature state more than a century earlier. Many of the main themes from this confrontation are in fact, as Hoyland notes, paralleled in other sources of the late eighth century and the early ninth.¹⁵⁴ Moreover, both letters have the appearance of responding to an earlier tradition of correspondence, which leads Hoyland to propose that over the course of the eighth century a series of Leo-ʿUmar / ʿUmar-Leo letters were composed, and “what has come down to us is a compilation from or rehashing of such works.”¹⁵⁵ Perhaps most importantly, however, the Aljaimado text of ʿUmar’s letter begins with an isnād, that is, a chain of the text’s early transmitters. Although such efforts to authenticate Islamic traditions by providing an intellectual pedigree were frequently forged and are thus generally viewed with a high measure of suspicion, Gaudeul and Hoyland are both correct to note that in this instance the letter’s isnād seems worthy of some historical consideration.¹⁵⁶ The isnād identifies a series of three scholars who are known to have been active in Ḥimṣ (Homs in western Syria), and the fact that the isnād does not attempt to link the letter with ʿUmar himself seems to speak for its authenticity. The earliest of these transmitters died in 798, a date that would be consistent with the origins of these epistolary polemics in the eighth century. On the whole then, as Hoyland rightly concludes, the evidence strongly favors the emergence of a literary tradition of polemical correspondence between Leo and ʿUmar, and more specifically the composition of ʿUmar’s letter, sometime before the end of the eighth century. This would make ʿUmar’s letter one of the oldest Islamic documents to have survived, making it a precious witness to the beginnings of Islam.

The relevant passage of ʿUmar’s letter for the present question comes at the very end of the text, in the early Arabic fragment published by Sourdel. As the letter draws to a close, it undertakes an extended defense of Muhammad’s prophethood. Invoking passages from the Qurʾān as evidence, “ʿUmar” contends that Muhammad was not taught by the Christian monks of the Baḥīrā legend but instead received his teaching directly from God. The nature of Muhammad’s message is also defended. Muhammad brought the truth of monotheism to a people that “had never before received any prophet or any scripture, a nation of ignorant people … worshipping idols.”¹⁵⁷ Muhammad’s success in the face of his countrymen’s immorality and infidelity is adduced as proof of the divine origin of his message. The conversion of these barbarous and faithless men to prayer, fasting, piety, and faithfulness verifies the authenticity of Muhammad’s prophetic call and his preaching: “Indeed, only prophets, God’s messengers, and the best of His servants can lead men in this way towards good, prescribing it, exhorting to it, while forbidding sins and transgressions.”¹⁵⁸

The letter then shifts to the Islamic conquests, which comprise its final theme. At God’s command Muhammad taught his followers to fight against those who “give partners to God, refuse to recognize Him and worship another god until they come to honour the only God, the only Lord, adopt the one religion”; those who fail to do so are to pay the jizya, by which God will teach them to realize their infidelity (citing Qurʾān 9.29). As a consequence of this instruction, the letter explains that Muhammad led his followers forth out of Arabia against the Byzantine and Persian empires. “In this way, with him in whom we trust, and in whom we believe, we went off [فخرجنا معه تصديقا به وإيقانا به], bare foot, naked, without equipment, strength, weapon, or provisions, to fight against the largest empires, the most evidently powerful nations whose rule over other peoples was the most ruthless, that is to say: Persia and Byzantium.”¹⁵⁹ Thus, this early Islamic text seems to confirm the witness of the non-Islamic sources that Muhammad was still leading his followers as they went forth and invaded the Byzantine Empire. Since ʿUmar’s letter is a Muslim text, Hoyland’s questions are largely irrelevant: although it is a polemical text, there is no reason to think that the literary confrontation with Christianity has somehow determined Muhammad’s involvement in the invasions. While the key passage unfortunately does not identify Muhammad specifically by name, using instead the third-person singular suffix pronoun, the immediate context leaves little doubt that he is the one with whom they went forth to fight, and both Sourdel and Gaudeul agree in translating the passage thus.¹⁶⁰

Consequently, we have in ʿUmar’s letter to Leo an early Islamic text roughly contemporary with (or at least within a few decades of) Ibn Isḥāq’s biography that appears to preserve a memory of Muhammad’s leadership at the beginning of the Near Eastern conquests. This strongly invites the possibility that ʿUmar’s letter bears witness to the same early tradition signaled by the non-Islamic sources. Quite possibly, the tradition of Muhammad’s leadership during the invasion of Palestine was still remembered by the Muslims of western Syria at the end of the eighth century, even as the Medinan traditions of Muhammad’s pre-conquest death at Medina received official sanction at the court in Baghdad, in the form of Ibn Isḥāq’s imperially commissioned biography.¹⁶¹ Perhaps the author of ʿUmar’s letter did not yet know the new contours of Muhammad’s biography as they were being formed in Medina and authorized at the ʿAbbāsid capital. Or it may be that ʿUmar’s letter adheres to this tradition because it is in dialogue with the Christians, who seem to have known this early tradition rather well. In confronting these religious rivals, it would not be helpful to introduce revisionist history: such dramatic changes to the narrative of Islamic origins would likely not persuade Christians of the truth of Islam. Moreover, western Syria is precisely the location where one might expect to find such a traditional holdout: as will be seen in subsequent chapters, the tradition of Muhammad’s leadership during the assault on Palestine seems to reflect the sacred geography of the earliest Muslims and Umayyads in particular. On the whole, the letter of ʿUmar to Leo offers important and early confirmation from the Islamic tradition that the combined witness of the non-Islamic sources is not simply the result of an unlikely collective mistake. Instead, this anti-Christian polemical treatise, seemingly one of the earliest Islamic texts to have survived, vouches for the antiquity and authenticity of the tradition witnessed by these Christian, Jewish, and Samaritan sources.

Conclusion

These eleven witnesses from the seventh and eighth centuries all indicate in various ways that Muhammad was alive and leading the Muslims when the Islamic conquest of Palestine began.¹⁶² While occasionally the years assigned to various events are not correct, this is not at all uncommon in medieval chronicles: such errors occur frequently in the chronicle tradition, and similar mistakes in chronology characterize the Islamic historical tradition as well.¹⁶³ Nonetheless, even when their chronologies are confused and inaccurate, medieval historical sources such as these are often reliable for their relative sequencing of events, and the consistency displayed with regard to Muhammad’s involvement in the Near Eastern conquests is impressive to say the least. Some of these documents merely indicate Muhammad’s death sometime after the Near Eastern conquests had begun, while others are more descriptive in noting Muhammad’s actual leadership during the invasions. But when all of the sources are considered together, as they are here, their collective witness to Muhammad’s continued leadership of the early Islamic community during the assault on the Roman Near East is unmistakable. While many scholars have rejected or ignored the evidence of these sources, no one has disputed that they do in fact report this.

For the most part, these reports are free from polemic and apologetic interests, and even when these qualities are evident elsewhere in a given text, they do not affect the notice of Muhammad’s vitality and leadership of the military campaign in Palestine. None of these texts connects its report of Muhammad’s leadership during the Near Eastern conquests with any sort of “totalizing explanation” of Islam or an apologetic agenda. Although a few of the authors display marked ideological tendencies elsewhere in their writing, in no instance are these themes linked with their observations that the conquest of Palestine or the Roman Near East began during Muhammad’s lifetime. In every case, the notice of Muhammad’s survival and leadership during the Near Eastern campaigns is mentioned almost in passing, so unobtrusively that its dissonance with the received tradition could easily be overlooked, as indeed it generally has been. The neutral, matter-of-fact manner with which the various sources convey this information suggests that this was the chronology that the authors had collectively received (or perhaps in some cases experienced?) rather than something that they were trying to impose onto their narratives. There is then little cause to suspect that any or all of these writers have invented a report locating the Islamic conquest of the Near East within Muhammad’s lifespan to suit some broader ideological agenda: no evidence would suggest this, nor is there any obvious reason for them to have fabricated such information. Likewise, the possibility of a collective error by all eleven sources seems highly improbable, particularly in the case of the Letter of ʿUmar. While such an interpretation of course cannot be entirely excluded, it does not offer a very compelling explanation for the persistent and seemingly independent manifestations of this tradition linking Muhammad with the invasion of the Roman Near East.¹⁶⁴

Several of these documents are of particularly high quality, including the first two and the final two especially. The Doctrina Iacobi, written within months of the invasion of Palestine it would seem, bears near contemporary witness to Muhammad’s presence among the invading “Saracens.” Although the text itself was composed in North Africa, its report concerning recent events in Palestine is said to rely on a document sent by a Jewish resident of Palestine, Abraham, who allegedly obtained his information about the Arabs and their prophet from eyewitnesses. In light of Abraham’s notice that Muhammad was preaching the imminent arrival of the messiah, one wonders if some of his informants were among those Jews who saw Muhammad’s religious movement as the fulfillment of their eschatological hopes. Such contemporary Jewish faith in Muhammad as a divinely appointed deliverer and herald of the messiah is clearly witnessed in the apocalyptic traditions ascribed to Rabbi Shimʿōn b. Yoḥai. Although these early traditions survive only in slightly more recent texts, their identification of Muhammad as one who conquers the land at God’s will is so anomalous with later Jewish attitudes toward Muhammad and Islam that, as numerous scholars have noted, this apocalyptic vision must have been composed very close to the events of the conquest itself. Similarly, the Continuatio of the Samaritan Chronicle, which very clearly relates Muhammad’s leadership during the invasion of Palestine, seems to preserve a particularly early account of the Islamic conquests, despite its survival only in a relatively late collection. Like the apocalypse of Rabbi Shimʿōn, the Continuatio’s strikingly positive attitude toward Muhammad and the invading Muslims, as well as its seemingly detailed and accurate knowledge of the conquests, suggests that its source for events of the mid-seventh century must have been composed in close proximity to the events themselves, perhaps on the basis of eyewitness accounts.

Finally, ʿUmar’s Letter to Leo provides important confirmation of this early tradition from the Islamic side. Although this epistolary polemic is one of the later texts that we have considered, it is for an Islamic source particularly early. By way of comparison, the oldest extant narrative of Islamic origins, Ibn Isḥāq’s biography of Muhammad, was composed only in the middle of the eighth century, and it is known only in two later recensions by ninth- and tenth-century authors. Moreover, as Hoyland notes, ʿUmar’s letter shows signs of having compiled earlier “exchanges,” and thus perhaps its tradition of Muhammad’s leadership during the invasion of the Near East is even older than this version of the letter itself, similarly predating Ibn Isḥāq’s biography. The apparent composition of ʿUmar’s letter in western Syria is especially important, inasmuch as the Medinan traditions of Muhammad’s pre-conquest death at Medina may have first spread into the Near East only at a later date largely through influence of Ibn Isḥāq’s biography, a work composed by this Medinan scholar in Baghdad at the caliph’s request. As a whole the canonical accounts of Islamic origins were composed under ʿAbbāsid rule almost entirely on the basis of Medinan and Iraqi authorities, and accordingly Syrian and (pro-)Umayyad traditions are very scarce in these eighth- and ninth-century collections.¹⁶⁵ Yet by contrast, ʿUmar’s letter originates in the same geographic region as most of the sources considered in this chapter, that is, Syro-Palestine, the center of Umayyad rule. Quite possibly, this early Islamic apology preserves a common early memory of Muhammad’s role in the invasion of Syro-Palestine from this region and this era, shared by Muslim, Christian, Jew, and Samaritan alike. Of utmost importance is the independence of these four reports from one another, which makes their convergence regarding Muhammad’s leadership at the beginning of the Near Eastern conquests quite impressive to say the least. While it is of course possible that someone might have misunderstood Muhammad’s significance for the invading Muslims, it is extremely unlikely that these four documents and their sources would all have made the same mistake independently, particularly in the case of ʿUmar’s letter. Thus, in view of their high quality, these sources alone are compelling enough to warrant serious reconsideration of the traditional Islamic memory of Muhammad’s death.

The remaining seven reports all come from the Christian historical tradition, whose accounts no doubt depend on earlier oral and written traditions about the Islamic invasions. Nonetheless, several of these texts bear witness even more clearly to Muhammad’s leadership at the onset of the Near Eastern conquests. This agreement suggests that we are not misreading the earlier sources, or, at the least, we are interpreting their reports in the same way as the next generation of Near Eastern Christians and their historians. Like the previous four documents, these sources also represent the diverse religious communities of the early medieval Near East. Although one document, the short Syriac chronicle written in 775, was produced in an unknown context, the others were composed by authors from the Coptic, Maronite, East Syrian, and West Syrian communities, while one set of traditions survives in the Christian chronicles of early Islamic Spain. And most importantly, each of these witnesses appears to transmit this information independently.

We would add here briefly a later indication from the Islamic biographical tradition identifying Syria as the land of Muhammad’s rule. In a report assigned to Kaʿb al-Aḥbār, a legendary bearer of Jewish lore in the early Islamic tradition, Ibn Saʿd identifies Mecca as the place of Muhammad’s birth, Medina as the place of his migration, and Syria as the land of his rule ().¹⁶⁶ Although it certainly is possible that this tradition merely reflects the eventual dominion of Muhammad’s followers in Syria shortly after his death, this notice that Muhammad ruled over Syria is rather intriguing in light of the information above. The statement, which Kaʿb claims to know from “the Torah,” identifies Syria as the area in which Muhammad established his political authority, seemingly in the same fashion that Mecca should be recognized as the place where he was born and Medina as the place to which he fled. Such parallels would appear to suggest that rule over Syria was one of the hallmarks, indeed the climax, of Muhammad’s career: while other tendencies may have inspired this formulation, one certainly should not exclude the possibility that this report bears witness at greater distance to an earlier tradition associating Muhammad with the conquest of Syro-Palestine.

Yet while each of these early sources indicates in various ways the same chronology of Muhammad’s survival into the period of the Near Eastern conquests, it should be noted that none of them actually relates any specific information concerning the manner and circumstances of his death. There are, however, a few Christian reports from the eighth or ninth century that in fact purport to describe the events of Muhammad’s death. As one might expect, these accounts are highly polemical, offering narratives of Muhammad’s demise that have been deeply colored by the Christian imagination. One of these, the Latin Istoria de Mahomet, is a brief biography of Muhammad that seems to have come into circulation in Spain sometime prior to the middle of the ninth century, when Eulogius of Cordova incorporated it into his Liber apologeticus martyrum.¹⁶⁷ Interestingly enough, like the sources considered above, the Istoria de Mahomet also seems to present the Islamic conquest of the Roman Near East within Muhammad’s lifetime. This concise Christian “Life of Muhammad” begins with Muhammad “as an avaricious usurer,” whose frequent business travels brought him into contact with Christian communities.¹⁶⁸ After drinking deeply of what he learned from the Christians, Muhammad was approached by “the spirit of error … in the form of a vulture,” who persuaded Muhammad that he was the angel Gabriel and directed Muhammad to present himself to his people as a prophet. Muhammad then began to preach, convincing many to abandon idolatry and ordering them “to take up arms on his behalf, and … to cut down their adversaries with the sword.” Then we learn that “first they killed the brother of the emperor who held dominion over the land and in recognition of the triumph of victory, they established the Syrian city of Damascus as the capital of the kingdom.” Immediately thereafter, the Istoria continues to describe how Muhammad fabricated the Qurʾān, followed by notice of his somewhat irregular marriage to the wife of Zayd, which took place after he had already “subjected her to his lust.”¹⁶⁹ Then with the commission of such a heinous sin, “the death of his soul and body approached simultaneously,” and the Istoria concludes with an account of Muhammad’s death.

The flow of the narrative certainly seems to suggest that the conquest of Syria took place during Muhammad’s lifetime. It appears in this biography of Muhammad amid other major themes from his life, such as his career as a merchant, his doctrine of religious conquest, his composition of the Qurʾān, and his “irregular” marital life. If one did not know any better (from reading the accounts of the traditional Islamic sources), one would presumably understand the assault on Syria as also falling within Muhammad’s lifespan. Standing squarely at the center of this polemical vita, the conquest of Syria seems very much to belong among the accomplishments of Muhammad’s prophetic career. It is thus tempting to suppose that we meet here yet another witness to the early tradition of Muhammad’s survival during the invasion of Syro-Palestine, albeit at a slightly greater chronological distance. Although the source is admittedly a hostile one, there is no obvious polemical motive for placing the Near Eastern campaign within—as opposed to immediately after—Muhammad’s lifetime. That Damascus is here in focus, rather than Jerusalem, merely reflects its status as the first capital of the Islamic empire, and its capture in 634–35 fits with the time frame envisioned by the sources considered above. Likewise, Heraclius’s brother Theodore did in fact lead, unsuccessfully, the defense of Syria, even though there is no evidence that he died in battle against the Arabs, as suggested by the Istoria.¹⁷⁰ Nevertheless, it must be admitted that the text does not explicitly associate Muhammad directly with the assault on the Roman Near East, and Damascus did not become the Islamic capital until 661. While these events do rather strangely intrude at the center of Muhammad’s life story here, one cannot exclude the possibility that the author has “cut to the chase” by introducing what his audience would otherwise have known to be the final outcome of Muhammad’s militant message. Yet by the same token, the Istoria does not otherwise clearly separate Muhammad from the conquests, and in light of the early tradition placing them within his lifetime, it seems very possible that we have here another relatively early witness to this rival tradition.¹⁷¹

As the Istoria continues to relate Muhammad’s death, it explains that when he sensed that death had come upon him (immediately after his “sin” with Zayd’s wife), he predicted that he would be resurrected three days after his death by the angel Gabriel. Following his death, Muhammad’s followers maintained a vigil, guarding his body and awaiting its resurrection. When three days later this did not transpire, Muhammad’s body began to stink, and his followers convinced themselves that their presence was preventing the angel’s appearance. So they left the body alone, “and immediately instead of angels, dogs followed the stench and devoured his flank”; his disappointed followers then buried what was left of the body.¹⁷² The Syriac versions of the Baḥīrā legend, a medieval Christian counter-narrative of Islamic origins, share a similar story, according to which Muhammad declared himself the Paraclete. By consequence, it seems, his followers expected that three days after his death “he would go up to heaven, to Christ, who sent him.”¹⁷³ When he died, they brought his body to a large house and sealed it inside. Three days later, they returned only to find that they could not even enter the house on account of the stench of Muhammad’s rotting corpse. Barbara Roggema, the text’s most recent editor, dates this particular tradition tentatively to the eighth or ninth century, largely on the basis of its similarities to the Istoria de Mahomet, while Krisztina Szilágyi suggests a similar dating on the basis of the Baḥīrā legend’s literary history.¹⁷⁴ It certainly seems possible, as Roggema suggests, that a Christian polemical tradition ascribing failed predictions of a bodily resurrection to Muhammad arose quite early, and that this episode from the Baḥīrā legend thus bears witness to an early anecdote about the end of Muhammad’s life. As much would certainly seem to be suggested by an early Islamic tradition, discussed in the following chapters, that when ʿUmar initially refused to allow Muhammad’s burial after his death, seemingly in hopes of his resurrection, al-ʿAbbās intervened to insist on his burial, noting that Muhammad’s corpse had begun to stink.

Unfortunately, however, the Syriac Baḥīrā legend affords no indication of the timing of Muhammad’s death in relation to either the Near Eastern conquests or any other major events from the history of early Islam. Nevertheless, the most striking feature of this alternative account of Muhammad’s demise is its indication, in the East Syrian recension at least, that Muhammad’s followers do not know anything about his grave, including, one would presume its location.¹⁷⁵ This feature would seem to suggest a particularly early date for this tradition, sometime before the tradition of Muhammad’s death and burial in Medina had become well established. More to the point, particularly for present purposes, is that this brief polemical account seems to recall a time when Muhammad’s followers were perhaps uncertain as to the location of his grave. It is difficult to imagine a Christian polemicist fabricating such Islamic ignorance concerning the site of Muhammad’s death, particularly if the tradition of his death in Medina had been well established from early on. It is certainly not obvious, for instance, how this would serve the tendencies of this polemic: there is no reason why the location of the house where Muhammad failed to resurrect would need to remain a mystery, and indeed, the absence of a known grave could seem to validate an Islamic claim to his resurrection. Admittedly, this source is problematic on a number of fronts, and its polemical character raises substantial questions regarding its reliability. Nonetheless, its suggestion that there was a time when Muhammad’s followers did not know the location of his grave is more than a little intriguing, and it certainly adds lateral support to the notion that Muhammad’s life may have ended in rather different circumstances than his traditional biographies remember it.

In summary then, from 634 onward, the various religious communities of the Near East repeatedly report a memory of Muhammad’s continued leadership of the Islamic community at the beginning of the Islamic conquests of the Near East. The consistency of this tradition and its persistence across confessional boundaries and over considerable distances are themselves quite persuasive. Moreover, there is no obvious reason for these authors to have fabricated this information, and the nature of the sources that transmit this information suggests that on this particular matter they are as reliable as one could reasonably expect of any historical source. To my knowledge, the earliest non-Islamic text to indicate that Muhammad died before the onset of the Near Eastern conquests is in fact Łewond’s Armenian chronicle from the end of the eighth century, although Łewond’s chronology of the conquest is itself highly erratic. Łewond locates the conquest of Syria and Palestine after Muhammad’s death, although a little too far thereafter: according to Łewond, the Muslims did not invade Palestine until after the death of Heraclius, that is, 641.¹⁷⁶ This would place the invasion of Palestine well into ʿUmar’s reign, which cannot be right.

Perhaps a more successful effort to “correct” the Christian historical tradition so that it would agree with the emergent Islamic historical tradition can be seen in the Greek chronicle of Theophanes, written at the beginning of the ninth century.¹⁷⁷ Although Theophanes is clear in signaling Muhammad’s decease before the onset of the Palestinian campaign, Theophanes, or perhaps more correctly one of his sources, has made use of Islamic traditions for knowledge of the chronology of Muhammad’s life, as Conrad has shown.¹⁷⁸ Thus, this Christian witness to the traditional Islamic chronology does not in fact offer independent attestation of Muhammad’s death prior to the conquest but almost certainly reflects the author’s direct knowledge of the emergent Islamic historical tradition and its memory of Muhammad’s death in Medina in 632. Nevertheless, despite these “corrections,” Theophanes additionally relates that Muhammad’s life ended with his “slaughter” or “wounding” (σφαγή): could this anomaly perhaps suggest some vestige of an earlier tradition that Muhammad died in battle, possibly leading his followers in the conquest of the Holy Land?¹⁷⁹ To be sure, such a proposal is highly speculative, but the further indication in this passage that Muhammad’s “slaughter” took place against a backdrop of Jewish messianic expectations would seem to comport with many of the early reports from the sources discussed above, as well other related traditions to be considered in Chapter 4. In any case, despite the eventual establishment of the canonical Islamic narratives of origins, the tradition of Muhammad’s leadership during the conquest of Palestine died a hard death, and it continued to figure prominently in the Syriac historical tradition, where it appears in both Michael’s Chronicle of the later twelfth century and the anonymous Chronicle of 1234, as we have already seen. Likewise, Thomas Artsruni’s Armenian History from the turn of the tenth century also places the conquest of Palestine within Muhammad’s lifetime.¹⁸⁰ Perhaps the tradition continued even later.

On the whole then, when considered purely on its own merits, the tradition that Muhammad survived to lead the invasion of Palestine would appear to be both early and trustworthy. The only problem, however, is that the Islamic historical tradition invariably reports Muhammad’s death at Medina in 632, almost two full years before the Islamic armies first invaded Palestine and the rest of the Near East. Since these Islamic sources were essentially the only accounts of Islam’s earliest history consulted or even available prior to the last century, the traditional Islamic account of the end of Muhammad’s life has dominated Western historiography for centuries.¹⁸¹ Now, however, thanks to the considerable efforts of both Western and Near Eastern scholars over the past century and a half, the literary heritage of other religious communities from the medieval Near East is becoming better known, and their writings have disclosed new perspectives on the rise of Islam. While much that these sources report is of use only for understanding internal responses to Christian defeat and the transition to Muslim rule, some of the information preserved by these texts also has value for understanding the earliest history of Islam itself, and the tradition of Muhammad’s leadership at the beginning of the conquest of Palestine quite possibly stands among the latter. The high quality of the evidence demands that we take this witness seriously. But what are we to make of these two conflicting reports? To pursue this question further we must first and foremost consider both the nature and reliability of the sources responsible for transmitting the Islamic tradition of Muhammad’s death in Medina as we have just done for the non-Islamic sources, a task to which we now turn in the following chapter.