Pastoralist peoples exploding out of the their lands and conquering farming peoples is a recurring feature of human history.
The Indo-Europeans did so with great historical consequences, expanding into Europe, the Iranian plateau and North India. They were the first of a recurring waves of steppe conquerors: Huns, Turks, and Mongols most famously. The Maghreb also generated at least two waves of Berber conquest across the Maghreb and into Spain.
The Arabian Peninsula produced the most startling, and also profoundly historically momentous, such wave of pastoralist conquest, the conquests of 632–750. But it did this only once. Only once was the Arabian peninsula the source of a major wave of conquest.
And the traditional story of how this happened, which is the Muslim traditional narrative, makes remarkably little sense. A story whose surviving textual sources start two centuries after the stated death of Muhammad in 632.
Mecca is an enormously implausible starting point. There is a dramatic paucity (pdf) of historical references to Mecca. (For the problems of the historicity of Mecca, see here, here, here, here and here.) It is a small settlement with a single well and no agricultural hinterland, that was not on any major trade route and well away from any imperial frontier.
Yathrib (Medina) is a bit better, but not much. It is larger, has an agricultural hinterland and is on a trade route. But is still too far away from the relevant frontiers and has no history of major political organisation. The Hejaz generally, particularly the section that Medina and Mecca are in, makes little sense as a breakout centre as there is simply not enough there. The most recent waves of conquest in the Arabian Peninsula, those of the al-Saud, go towards the Hejaz, not away from it.
Building a new history out of contemporary sources
This lecture by Peter von Sivers on the interactions between Christian theological controversies and struggles with what was happening with the Arabs in Northern Arabia, creates a hugely more plausible context. The action moves to the frontiers with the Roman (“Byzantine”) Empire and Sassanian Empire. Both Empires had had Arab buffer-client states (Ghassanids and Lakhimids) that had been either much reduced or effectively eliminated by their Imperial sponsors by the start of the last and greatest of the Roman-Sassanid Wars, which lasted from 602 to 628. A decades-long struggle that exhausted both empires and left the Sassanian Empire mired in civil war and instability but the Arabs largely unaffected.
The unification of these former buffer states — areas used to significant political organisation and familiar with the practices, strengths and weaknesses of the exhausted Imperiums — into a single Arab kingdom provides a far more plausible basis for the Arab breakout.
Tom Holland’s discussion of the broader similarities between the processes on the borders that saw the collapse of the Western Roman Empire in the C5th with the Arab conquest of over half the Eastern Empire in the C7th fits nicely into this picture.
As does Dan Gibson’s argument that Petra is the origin city, not Mecca. As indicated by the fact that the Qibla (direction of prayer wall) of all the original mosques point to Petra, not Mecca. (Of course, a self-published scholar does not have the same cachet.)
Nevertheless, Petra (a major trade and religious centre) is in the right place, has the right history and fits the descriptions of the city of the Prophet’s birth.
If we add in Prof. Fred Donner’s lecture on trying to contextualise (i.e. assemble a history based on contemporary evidence) early Islam, we also get a picture compatible with what von Sivers and Holland are arguing (and, for that matter, Dan Gibson). Islam becomes a religion assembled out of the needs of imperial control to justify, first the Arabs as a ruling people, and then the Abbasids ruling as a Muslim dynasty.
A process started by the Umayyad Caliph Abd al-Malik ibn Marwan (r.644–705) who builds the original Dome of the Rock in 691–2, which contains the earliest Quran verses and the first explicit reference to Muhammad.
The written-down two centuries later, hundreds of miles away, traditional story of the origins of Islam and the Islamic conquests makes remarkably little historical sense. But we seem to be groping towards a picture, based on contemporary evidence, that makes a lot more sense.
It still leaves the Arab breakout and its consequences as a most extraordinary eruption into history. But not a nonsensical one.