some mention is usually made of possible causes, the underlying explanation that is given remains the same: the Central Eurasians were hungry, poor, and cold, but they were also aggressive, energetic, and naturally ready for war. They took advantage of the opportunity to plunder the weak agrarian peoples to the south and had unexpected success, such that they were able to establish their own states in the region. This characterization of the history of the period is misleading at best. It is based more or less completely on the mistaken belief that the Central Eurasians (excluding refugees from war, of course) were indeed hungry, poor, and so on in their homelands and imputes to them motives for which there is no evidence. (See the discussion in the epilogue.) Being human, the Central Eurasians undoubtedly did attack their enemies—as much as their neighbors attacked them. But the simple fact is that we do not know why the Great Wandering of Peoples took place. Nevertheless, enough is known about the events themselves in the peripheral cultures, so the reason seems perhaps to be potentially discoverable.
62. The history of Arabian internal and foreign trade, Muḥammad, and the early Islamic expansion is an extremely contentious field. See the differing treatments of Shaban (1970, 1971, 1976), Crone (1987), and Peters (1994). The present treatment largely follows Shaban and, in part, Crone, particularly her conclusion on the primary driving force behind the expansion out of Arabia. She argues, via a process of elimination, that the putative natural belligerency of the Arabs (q.v. endnote 63) cannot explain the unique history of the foundation of Islam and the subsequent conquests, leaving as the only explanation the “foreign penetration” of Arabia (Crone 1987: 245–250): “Muḥammad’s Arabia had thus been subjected to foreign rule on a scale unparalleled even in modern times” (Crone 1987: 246). On the early conquests, see also Donner (1981).
63. Crone (1987: 243–245) and others claim that the Arabs were greedy, rapacious conquerors: “Tribal states must conquer to survive, and the predatory tribesmen who make up their members are in general more inclined to fight than to abstain” (Crone 1987: 243). So too, in her view, were the Muslims: “Muḥammad had to conquer, his followers liked to conquer, and his deity told him to conquer: do we need more?” (Crone 1987: 244). These statements do not seem to have scientific justification. Their remarkable similarity to the received ideas about Central Eurasians discussed in the epilogue is not accidental; see the discussion of similar views on the Islamic comitatus in Beckwith (1984a).
64. It is often stated that there are so few books in Middle Persian or in any Persian literary language before New Persian because the Arabs destroyed the “great library of Ctesiphon.” In fact, so few books in early Persian have survived because the Persians simply wrote few books, at least in Persian, before they adopted Islam and got the habit of writing from the Arabs. When the Arab Empire began dissolving in the early ninth century, a highly Arabicized literary language, New Persian, developed. The Persians thenceforth wrote copiously, like the Arabs. The story seems to have arisen to explain the paucity of books in Middle Persian by contrast with the great number in Arabic and, eventually, New Persian. This myth belongs on the dustheap of history along with the one that claims the Arabs destroyed the great library of Alexandria, which actually had disappeared centuries before the Arab conquest.
65. The Western and Southern subregions of Central Asia were conquered by Persian empires several times in recorded history, but those regions were never ruled directly by them for long. The local peoples were not Persians by culture or language. In fact, they spoke entirely different languages (Bactrian, Sogdian, and so on). Although those languages are related to Persian in the