The Return of the Black Box

The revisionist idea of the Hijaz as a barbarian no-man's-land that couldn't produce Islam has made a late comeback

In a review of Patricia Crone’s Roman, Provincial and Islamic Law: The Origins of the Islamic Patronate (1987), Wael Hallaq once noted a theme running through the eminent historian’s controversial early work, from Hagarism (1977), to Slaves on Horses (1980), to Meccan Trade and the Rise of Islam (1987) — the notion of a stark civilizational and identitarian divide between the settled people of the Levant and Mesopotamia, on the one hand, and Bedouin communities of the Arabian Peninsula, on the other. In these studies the Fertile Crescent belongs typologically to the world of ancient Greece and Rome, while the southern regions are assumed to be a barbarian no-man’s-land. So when the culturally deprived inhabitants of the desert conquered the northern areas, they could only imitate and adapt the high culture of their subjects, including the system of clientage (walāʾ) and manumission that Crone argues came from Roman provincial law. But as Hallaq demonstrated, Crone was ignoring the “constant state of flux between North and South” in the centuries before the Islamic era, preferring instead a vision of discrete geographic zones and ethnic groups that scholars of the Roman-Byzantine Near East had already begun to overturn.

Crone’s wider point was a historical comparison of nomadic or semi-nomadic societies who were able to use their domination of the Eurasian desert and steppe zones to launch major assaults on settled centres of civilization. She was thinking of the Gothic invasions that brought down Rome and the later Mongol invasions, but while in terms of the broad lines of human history she was onto something — further developed in her masterful study Pre-Industrial Societies: Anatomy of the Pre-Modern World (1989) — when you get down to the details the model becomes misleading. Hallaq was able to point to a raft of studies that have established how the Roman and Byzantine Levant in particular was full of Arabic-speaking communities, both settled and migratory, retaining their own customs and language, yet living symbiotically with the Syriac (Aramaic) speakers of most settled towns and villages. The polities in question, which were often linguistically hybrid in nature, included the Palmyrene state of Queen Zenobia, as well as the Nabataean territories (where papyrus discoveries from 1993 show that Arabic-speakers were still resident in Petra in the late 500s) and the Ghassanid tribal confederation based at al-Jabiya in the Hawran region overlooking the Golan in southwest Syria. In the immediate pre-Islamic period Christianity was widespread among these Levantine communities, mainly through two churches, the Greek-language Chalcedonian rite of Constantinople and the (West) Syriac Jacobite rite of the Levant, both of which competed with the (East) Syriac Mesopotamian Church of the East which was under Sasanian patronage.

Islam as a Late Antiquity Religion

By the time Crone passed in 2015 Islamic historians had managed to catch up with these ideas spreading in other historical fields; the notion of Islam as a Late Antiquity phenomenon had taken hold. In simple terms, what this means is that Islam should be understood as emerging from the intellectual world of Late Antiquity, or to put a more colloquial spin on it, as in conversation with those religious, social, legal and political systems. But the crucial question is where contact between these ideas and the Hijazi Islamic movement took place. While the earlier Patricia Crone and other revisionist scholars would say this meeting occurred in the Levant, during the era of the conquests when they suspected much of the Quran was composed, now the trend is to see the Levant as having come to the peninsula. Indeed, it has become popular to argue that the culture of the Fertile Crescent — especially its Christian texts and languages — had permeated the peninsula by the time of the conquests, an idea to which many Islamic historians were for long oddly resistant. What this would mean is that the various Christian churches were competing for souls inside the peninsula from the centres they had established in Yemen and the Gulf littoral.

Crone was among the revisionists who changed their thinking on a range of issues related to the fundamental question of the Islamic tradition’s reliability due to the fact that its version of history was written down over a century after the events in question and thus reflects the changing worldview of a new empire and culture. Among the issues she rethought is the notion of Arabia as an inscrutable black box. Indeed, she famously wrote in a series of open access web articles published from 2007 to 2009: “Arabia seems to have been a much more developed place than most Islamicists (myself included) had ever suspected — not just in the north and south, but also in the middle.” There are still some outliers — independent researcher Dan Gibson thinks Muhammad lived in Petra not Mecca — but among serious Islamic scholars the most prolific has been Stephen Shoemaker, and I want to take a look at some of his writing here. I should state at the outset that I’m a fan of Shoemaker’s work. It’s hard not to be given the meticulous research he’s done that challenges newly formed and forming conventions. But there is an underlying theme there that needs interrogation.

The New Radicals

A professor at the University of Oregon, Shoemaker began as a historian of Near Eastern Christianity with early research such as The Ancient Traditions of the Virgin Mary’s Dormition and Assumption (2002) that looked at Christian apocryphal literature related to Mary. One of his first publications in the Islamic field concerned an obscure Byzantine church that once existed between Bethlehem and Jerusalem, the Church of the Kathisma of the Theotokos, or Church of the Seat of the God-Bearer (i.e. Mary). Uncovered during excavations in 1997, the church contains liturgical and iconographic traditions regarding Christ’s Nativity that appear to explain Quran 19:22-27, specifically the story of Mary resting under a date palm and eating from it during the flight to Egypt. Shoemaker’s key argument is that Muhammad could not have encountered these traditions during his lifetime in the Hijaz (the Red Sea coast and mountain region down to ʿAsir), so these verses must have been composed during the conquests (Jerusalem fell in 637/8).

After publishing his article in 2003, Shoemaker went on to author a series of studies looking at non-Muslim sources of the early Islamic era which give the impression that the conquests began as a jihadist apocalyptic movement promising its diverse followers an Abrahamic divine right to rule Palestine. His latest book, Creating the Qur’an: A Historical-Critical Study (2022), is a tour de force of insistent revisionism. Radiocarbon dating of a series of early Quranic fragments since 2009 has helped create a new scholarly consensus that the Quranic text we have today was finalized during the reign of ʿUthman (644-656), just as the Islamic tradition said it was. But Shoemaker comes up with many valid criticisms of radiocarbon dating to argue that the text circulated orally until it was finally gathered and edited as part of an imperial codification project under ʿAbd al-Malik (685-705), who was content to promulgate it as ʿUthman’s work since ʿUthman was his kinsman amongst the early rulers.

Indeed, it is under ʿAbd al-Malik that the conquest society becomes recognizably Islamic. Coins, buildings, milestones and documents from his era show that it was at this time, following his reassertion of Umayyad family rule after a civil war, that the state publicly promulgated terminologies such as “caliph”, “Islam” and “Muslim” for the first time, whereas previously “emigrants” (muhājirūn) and “believers” (mu’minūn) were the apparent preferred self-designations. The language of administration shifted from Greek to Arabic. The Dome of the Rock — which ʿAbd al-Malik completed in Jerusalem in 691/2 — promotes a public definition of what this Muslim faith and identity involved by means of its inner and outer wall inscriptions containing Quranic statements of faith (verses 37:35, 47:19, 48:29, 112:1-4) and declarations that Christ was a messenger of God, not His son. Rebels had challenged the Umayyad right to rule through a nativist discourse that impugned Umayyad piety, claiming to be the true heirs of Muhammad through acts such as issuing their own coins which for the first time carried Muhammad’s name and seizing the Meccan pilgrimage site as their stronghold. During nineteen years in power, the earlier Umayyad ruler Muʿawiya had drawn close to the majority Christians after moving the seat of government from Medina to Damascus — holding his coronation in Jerusalem and then praying the same day at Golgotha and Gethsemene (if the 7th century “The Maronite Chronicle” is to be believed) and drawing Christians into his bureaucracy and army. By contrast, ʿAbd al-Malik drew boundaries around the religious offerings, stamping his regime with an unambiguously Islamic imprint.

Hijazi Hinterland

However, one of the key problems with Shoemaker’s argument is his insistence that the Quran’s Late Antique character cannot have resulted from origins in a Hijazi setting. “It does not seem possible for the region to have been steeped in the rich cultural heritage of the late ancient Roman and Sasanian worlds,” he states. Reasons for this in his telling include Hijazi illiteracy and lack of education, only a few references to Medina in pre-Islamic sources and none to Mecca, lack of evidence that either town was on the ancient world’s trade routes, the Quran’s sole and ambiguous reference to Mecca as “Bakka”, lack of evidence that Quranic Arabic is a product of the Hijaz, and frequent references to non-Hijazi features such as grain, grapes, olives, date palms (!), pomegranates, seafaring, and fishing, as well as sheep, goats, cows, oxen, camels, mules, donkeys, and horses. “How is any of this compatible with Mecca and Yathrib [Medina] in the Hijaz?” he asks. Since Mecca and Medina both sit inland, “some one to two-hundred kilometres from the Red Sea respectively, in the middle of a vast, barren desert,” he claims these references to the civilized world could not possibly have resonated with a Hijazi audience. In short, Mecca — if it existed for Muhammad — was a backwater. There was no knowledge to be had in Mecca, at least of a non-Arabian variety, and its residents were incapable of traveling to acquire any.

These arguments typify a kind of abstract approach to the culture and society in question seems to me divorced from reality. In his own book A Prophet Has Appeared: The Rise of Islam Through Christian and Jewish Eyes (2022), Shoemaker notes the confirmation in early non-Muslim reports of Muhammad as a trader in the Levant who became a learned man through his sojourns there. What makes it unlikely, then, that he or others from the deepest Hijaz would acquire knowledge about Christianity, its theology and its disputes, or the wars of the Byzantines and Persians? And if Muhammad could travel to the Levant, why couldn’t he spend time by the sea at al-Shuʿayba, south of modern Jeddah, which functions fairly typically as the coastal centre to the hinterland pilgrimage sites around Mecca? The port is mentioned by early Abbasid historians al-Azraqi and al-Fakihi as “Mecca’s coastline”, though Crone ignores this in her Meccan Trade. What about the green mountains of Taif, further inland from Mecca, and ʿAsir and Jizan to the immediate south? Those are not exactly the desiccated zones that Shoemaker describes in Mecca, but somehow he thinks residents of the Meccan dustbowl would never in their lives have thought of taking the weather elsewhere. But if the Hijazi Arabs knew nothing of the sea, why were they heading to Cyprus in naval raids organized by Muʿawiya from 649 when he was governor of Syria? Arabia seems to be for Shoemaker another world, something close to the Barbarabia that Hallaq excoriated in Crone’s early work.

Leafy Taif sits inland from Mecca and Jeddah

The idea that the various tribal, settled and pastoral communities of the peninsula wouldn’t have interacted with each other is ludicrous to anyone who knows them today. Firstly, anthropologists specializing in the Arabian Peninsula are well aware of the historical phenomenon of tribes moving from south to the north. A pre-Islamic Hijazi poet such as Hassan bin Thabit — only some of whose material was produced later under false attribution — spent time at the Levantine court of the Ghassanids in al-Jabiya. So even if Shoemaker can claim that, “although Christianity had literally encircled the central Hijaz by Muhammad’s lifetime, there is no indication whatsoever of a Christian community in either Mecca or Yathrib [Medina],” the point is that there didn’t have to be for Muhammad’s religious community to develop their own take on Abrahamic monotheism. Shoemaker also over-relies on the dictum that lack of evidence is evidence of absence. Hatoon al-Fassi touches on the lifeless semblance of history that this can lead to in her study of the Nabataean caravan trade that ran the length of the peninsula, Women in Pre-Islamic Arabia: Nabataea (2007). Even if inscriptions haven’t provided the proof that empirical maximalists demand to establish the centrality of the southern Hijaz for the early Islamic movement, experts in Arabian epigraphy have noted that monotheistic epithets (such as al-ilāh, raḥmān) begin to appear the length of the breadth of the peninsula throughout the 500s, attesting to a pre-Islamic shift in religiosity that is hard to disassociate from the apparent spread of the rival churches throughout the region.

Give It Back

In short, the 1980s are calling and want their revisionist history back. Shoemaker doesn’t go quite as far as those claiming that “Mecca is Petra” — a theory that would require rethinking the geographic origins of everyone in the Islamic tradition described genealogically as Qurayshi. But he leans heavily into some arguments that most scholars have been content to set aside. Shoemaker has certainly thrown down the gauntlet to the recent radiocarbon dating fetish in Quranic Studies, providing an alternative theory of how the text could have been finalized during ʿAbd al-Malik’s caliphate. But he is also speaking to the question of whether the Hijaz or the Levant is the intellectual background for a fair chunk of the Quranic text, arguably the most pressing issue in the field at present, and on that score the idea that the Hijaz could not produce people mobile enough to access knowledge from near or far seems incredibly naïve, and those for whom this makes little sense need to push back.