The Faith Instinct

by Wade, Nicholas

  Traditionalists believe that the collection of Islamic documents, although full of internal inconsistencies, contains essential historical truths, which can be extracted by diligent study. A small number of revisionists, also known as rejectionists, view the documents as “salvation history,” or nonhistorical literature designed to wrap a theological message in historical trappings; they see non-Islamic sources and archaeology as more reliable evidence of the period. “It is not generally appreciated,” writes the Islamic historian Patricia Crone, “how much of our information on the rise of Islam, including that on Meccan trade, is derived from exegesis of the Qur’an, nor is it generally admitted that such information is of dubious historical value.”204 According to another historian of the Islamic world, Jonathan Berkey, “The reader should at least understand that the usual accounts of the origins of Islam are based on sources of dubious historical value.”205

  Here, for instance, is a recent account, based on the traditionalist method, of the battle of Yarmuk of A.D. 636. Islamic sources hold the battle to have been a turning point in the Arab conquest of the Byzantine empire’s holdings in the Near East. “The battle of Yarmuk is, along with the battle of Qadisiya in Iraq, one of the major conflicts that has come to symbolize the Muslim victories in the Fertile Crescent,” writes the historian Hugh Kennedy. “As with Qadisiya, the Arab accounts are extensive and confused and it is difficult to be clear about exactly what happened. There is no contemporary or reliable account from the Byzantine point of view.”206

  The revisionist view of the event is this: there is confusion among the Arab accounts of this allegedly crucial battle, and no Byzantine account, for a simple reason—there was no battle of Yarmuk, nor indeed an Arab conquest.

  The revisionist view of early Islamic history is described in more detail below because, if true, it furnishes a case study of how a religion can be adapted with great success to a state’s purposes. The conclusions of this minority view may not yet be widely accepted, but its methodology of giving serious weight to archaeology, and to non-Islamic texts, seems a reasonable approach.

  According to traditional Islamic history, Muhammad converted his followers to Islam before his death in 632. The first caliph Abu Bakr, ruling from Mecca in the Hijaz, the western region of the Arabian peninsula, directed Muslim armies northward to conquer the Near Eastern provinces of the Byzantine empire in the name of Islam. Many have noticed the parallel between this account and that of the Hebrew Bible; Muhammad, like Moses, died in the desert without seeing the Promised Land, and Abu Bakr, like Joshua, was the trusted general who implemented the prophet’s design.

  The revisionist view is different. Arabs didn’t invade Syria and Palestine because they were already there, say Yehuda Nevo, an archaeologist at the Negev Archaeological Project, and his colleague Judith Koren of the University of Haifa. And the Arabs could not have been Muslims at that time; the word Islam does not appear in history until inscriptions on the Dome of the Rock in Jerusalem, built in 692, and its meaning even there is disputed.207

  The policy of the Byzantine empire was to settle Arab tribes in border areas of Syria and Palestine, both to prevent them from raiding and to rely on them for defense. Eventually the Byzantines decided to withdraw altogether from their Near Eastern provinces, establishing their line of defense just south of Antioch in northern Syria. In 632 they even stopped payments to their Arab allies. Syria and Palestine were on their own.

  The Arabs who had been settled in those provinces found that they had only to push and one territory after another fell into their hands. There was a struggle for dominance between rival Arab tribes until one ruler, Mu‘awiyah, defeated all the others. Coins suggest he was the first Arab ruler, reigning from Damascus from 661 to 680. No caliph’s name before Mu’awiyah is mentioned in non-Islamic manuscripts, meaning that there is no independent evidence of their existence. And if there was a great invasion “it would seem that, at the time, nobody noticed,” Nevo and Koren assert.

  A similar conclusion has been reached by Peter Pentz, an archaeologist at the National Museum of Denmark, who describes the Arab takeover as “the invisible conquest” because of the lack of historical or archaeological evidence to support the Muslim accounts of invasion.208

  The proposal that the Arab takeover came about through an uprising in place, not an invasion from Mecca, would ease several geographical problems in the traditional account. Archaeological and literary evidence shows that some early mosques in Iraq and Egypt were oriented to an unknown sanctuary in northwest Arabia, not toward Mecca.209 The Hijaz was very sparsely populated during the period. Though the traditional account portrays Mecca as a thriving trade center, the historian Patricia Crone has shown it did not lie on the trade routes from Yemen and is not mentioned by the classical geographers.

  The Qur’an itself has several details that point to its composition in a setting farther north. Muhammad’s opponents are said to have grown grain, olives, grapes and dates, but Mecca is unsuitable for any kind of agriculture. The pagans are invited to reflect on the destroyed cities of Lot’s people, given that “you pass by them in the morning and in the evening,” suggesting a location near the Dead Sea.210

  A broader point is that the Qur’an assumes its readers are familiar with the Pentateuch and the Psalms. It contains many polemical passages, which give the impression of having been developed within a rich environment of Christian-Jewish theological discourse. The Hijazi desert seems less likely as the locale for this development than somewhere in the more populous regions of Palestine, Syria or Iraq. “Islam is obviously part of the Semitic monotheistic tradition and must have arisen within its matrix, and it is not futile to attempt to define rather more precisely how that happened,” writes G. R. Hawting, a historian at the London School of Oriental and African studies.211

  The canons of both the Old and New Testaments took shape over several centuries before being frozen in their present forms. It would be no surprise if the same were to be true of the Qur‘an. But the traditional account allows little leeway for such a process. In traditional belief, the Qur’an was dictated by an angel to Muhammad, preserved on palm leaves, flat stones and in people’s memories, and assembled from these disparate sources within a few decades of Muhammad’s death in 632; whereupon the Caliph ‘Uthman, who had directed that a standard text be prepared, “gave orders to burn every leaf or codex which differed from it.”212 In the revisionist view, the process took much longer, and at least in part in the setting of sectarian religious communities somewhere in Palestine, Syria or Iraq.

  The founder of the revisionist school was John Wansbrough, a historian at the London School of Oriental and African Studies whose principal works were published in the late 1970s. After a textual analysis of the Qur‘an, Islamic interpretations of the Qur’an, and the Sirah, he concluded that all belong to the genre of “salvation history,” meaning that the writing is a literary description of religious events, not a historical account. Without independent corroboration, a historian simply could not assess how much of the Islamic corpus corresponded to fact. “With neither artifact nor archive, the student of Islamic origins could quite easily become victim of a literary and linguistic conspiracy,” Wansbrough wrote.213

  Wansbrough concluded that the Qur‘an was assembled over a long period of time and probably did not take final form until around 800, more than 150 years after Muhammad’s death. The interpretations of the Qur’an, called Tafsir, and other Islamic documents were shaped with a specific purpose, in his view: “Tafsir traditions, like traditions in every other field, reflect a single impulse: to demonstrate the Hijazi origins of Islam.”214 The insistence that Islam originated in the Hijaz, in Wansbrough’s view, suggested there had been an internal debate, now lost, as to whether or not that was the case.

  Why should the Qur‘an and Muhammad’s life have been located in the Hijaz if in fact the origin of both was elsewhere? The revisionists’ proposal is that the shapers of the Qur’anic canon, who
were perhaps scholars working under the caliphs, believed the prophet of the new religion would fittingly have had an Arabian identity and needed to have lived in a place that clearly distanced Islam from both Judaism and Christianity.

  Wansbrough’s work is not widely known, in part because of the obscurity with which he wrote. One scholar has accused him of “relentless opacity,” a charge against which even his admirers may be reluctant to defend him.215 His ideas are known mostly through the writings of his students and others. Following up on his work, the historians Patricia Crone and Michael Cook reviewed all the non-Islamic sources they could find and tried on that evidence to reconstruct early Islamic history.

  Accounts from the 640s in Syriac, the version of Aramaic spoken in Syria, refer to the Arabs as Mahgraye, which can be rendered as Hagarenes—descendants of Hagar—in English, and the equivalent word Magaritai appears in Greek sources. Crone and Cook suggested the roots of Islam lay in Hagarism, a Jewish-Arab movement to repossess Jerusalem from the Byzantines, whom both peoples had reason to resent. But after Jerusalem was captured, the Arabs broke with their Jewish allies and entertained an alliance with Christianity instead. The rapprochement was temporary but is vividly captured in the surprisingly pro-Christian inscriptions on the Dome of the Rock in Jerusalem. The rulers of the new Arab empire then decided they needed to develop a faith of their own, independent of both Judaism and Christianity.

  Nevo and Koren have developed a different perspective on Islamic origins. From study of coins and inscriptions on rocks and buildings, they have defined a period of Arab monotheism during which writings often included the tawhid, the assertion that Allah is the one god, which was intended to contradict the Christian belief in the Trinity. Strangely, the earliest tawhid inscriptions make no reference to Muhammad. The first known historical reference to Muhammad may occur on an Arab-Sassanian coin minted in Damascus in 690/691, depending on how the coin’s legend—muhammad rasul allah—is translated. “Muhammad is the messenger of Allah” is the obvious translation but another, to be considered later, is “The messenger of God is to be praised.”

  Even if the coin does refer to Muhammad, its date is 70 years after the Islamic era began, according to the traditional account, with the migration or hegira of Muhammad and his followers to Medina in 622. “Before 71 A.H. [After the Hegira] he is not mentioned; after 72 A.H. he is an obligatory part of every official proclamation,” Nevo and Koren write, using the Islamic dating system.216

  Mu‘awiyah, the first Arab ruler recorded by non-Islamic sources, and his eventual successor, Abd al-Malik, had created a large state, most of whose subjects were Monophysite or Nestorian Christians. Now they needed a unifying religion. All early documents about the formation of the Qur’an have disappeared. One very interesting text that reflects the emergence of Islam survives. It is made not of paper but of stone, and was completed probably in 692, many decades before the earliest known copy of the Qur‘an. Constructed by ‘Abd al-Malik on the site of the Jewish temple in Jerusalem, the Dome of the Rock, like other early Islamic buildings, does not face toward Mecca. It has eight equal sides, which point if anywhere to itself as the center of worship.

  The octagon’s outer inscription contains statements that are mostly in the Qur‘an and a few that are not, such as “There is no God but Allah alone, he has no associate (lā ilāha ilā llāh wahdah, lā shark lahu).” These phrases suggest that the text of the Qur’an had not yet been finalized. The inner inscription addresses the interminable Christian disputes as to whether Jesus’ nature was human, divine or some admixture of the two. Abd al-Malik takes what Nevo and Koren characterize as the Judeo-Christian position, that Jesus was a true prophet but merely human. His inscription expresses, in another non-Qur‘anic statement, the highest respect for Jesus: “Allah, incline unto your messenger and servant Jesus son of Mary and let peace be upon him the day he was born and the day he dies and the day he shall be raised alive.” The inscription continues, in mostly Qur’anic language, “The following is the truth about Jesus son of Mary, about whom you dispute: why should Allah acquire a son?”

  The Dome of the Rock inscription had several purposes, Nevo and Koren write. “It called for an end to dissension, and for the population to unite into one community under their caliph, now firmly in control after several years of civil war. As the reason and justification—and framework—for this communal consensus, it presented an official religion: a form of Judaeo-Christianity, with particular emphases. To this end it took issue with, and rejected, the tenets of Trinitarian Christianity. And finally, it set within this framework an element which became the focal point of that religion—the Arab prophet.”217

  Nevo and Koren raise the possibility that the Arab prophet did not in fact exist, based on their speculation that the word muhammad—which is used only 4 times in the Qur’an, compared with 79 mentions of Abraham, 136 of Moses and 24 of Jesus—could have meant “the chosen one” and was not in this context a proper name. Their inference is that the corpus of Islamic literature consists of layer upon layer of stories each of which builds further detail about the life and sayings of the chosen one, developing a personality, biography and whole salvation history from a single word.

  There is indeed an accretionary process evident in Islamic writings, whereby a later writer is somehow able to supply the interesting historical details which an earlier writer had neglected to include. “If one storyteller should happen to mention a raid, the next storyteller would know the date of this raid, while the third would know everything that an audience might wish to hear about it,” writes Crone. Thus the Islamic historian Al-Waqidi, born in 748, relates far more copious information about Muhammad’s life than does the earlier historian Ibn Ishaq, born in 704. “No wonder that scholars are fond of al-Waqidi,” Crone writes: “Where else does one find such wonderfully precise information about everything one wishes to know? But given that this information was all unknown to Ibn Ishaq, its value is doubtful in the extreme. And if spurious information accumulated at this rate in the two generations between Ibn Ishaq and al-Waqidi, it is hard to avoid the conclusion that even more must have accumulated in the three generations between the Prophet and Ibn Ishaq.”218

  Crone, however, rejects Nevo and Koren’s suggestion that the entire life of Muhammad has sprouted from a mere title, saying Muhammad’s existence is well attested by several early texts, including an Armenian chronicle written in 660 and ascribed to a Bishop Sebeos. “Most importantly,” she adds, “we can be reasonably sure that the Qur’an is a collection of utterances that he made in the belief that they had been revealed to him by God.”219 Nevo and Koren, on the other hand, say the reference to Muhammad in Sebeos’s history is probably “a later explanation added by a copyist who saw that Sebeos did not know what he was talking about.”220

  Estelle Whelan, a critic of Wansbrough and the rejectionist school, argues that the Dome of the Rock’s inscriptions that appear to be non-Qur’anic are abbreviations of the canonical text, made to fit the limited architectural space. Nevo’s failure to find any mention of Muhammad in early Arabic inscriptions in the Negev simply reflects the fact that Islam was developing far away in the Hijaz, Whelan argues.221

  The Arab prophet may have lived, as Hawting suggests, somewhere within the matrix of Judeo-Christian monotheism, but if so the locale of his ministry, at some stage during the development of the Qur’an, was transferred from Palestine or Syria or Iraq to the purely Arab background of the Hijaz.

  An Alternative Hypothesis About Islam

  The Islamic era began in 622, a date held to mark Muhammad’s flight from Mecca to Medina. But why is this era dated to a mere shift of residence rather than to the prophet’s date of birth, for instance? The year 622 was indeed of the greatest significance in the Arab world, but for a reason that has been allowed to recede from historians’ sight: it was the date on which Arab independence began.

  For a century beforehand the Byzantine empire had been locked in a generally l
osing struggle with the Sassanid rulers of Iran. The Arab populations of Syria, Mesopotamia and Iran were caught in the middle of these rival power centers, both of which set up and manipulated vassal Arab buffer states. The Byzantines controlled the Ghassanids in Syria while the Sassanids were allied with the Lakhmid dynasty based in southern Iraq. In terms of the complicated Christian politics at the time, the Ghassanids were Monophysite Christians (Christ has only one nature-divine) whereas their Byzantine patrons were Chalcedonian (Christ has two natures, one divine, one human). The Lakhmids, on the other hand, were Nestorian Christians (Christ has not two natures in one person but two persons, one human, one divine).

  What happened in 622 was that the Byzantine emperor Heraclius, a superb commander, decisively defeated the Sassanid army under its great general Shahrvaraz, who seven years earlier had captured and sacked Jerusalem. The defeat was so devastating that the Sassanid empire collapsed a few years later. But the Byzantine empire also had systemic weaknesses and Heraclius, rather than taking possession of the captured buffer states, simply withdrew from them. For the first time, the Arab peoples in the region were on their own. Their first ruler, who was naturally a Christian like his people, certainly recognized the importance of 622: he measured his reign from this foundational date, and a Greek inscription in his name at Gadara (in present-day Jordan) records the year in terms of kata Arabas—Greek for “according to the Arabians.”