of recording the triumphs and tragedies of pre-Islamic warfare certainly coloured the way in which the battles of the first Islamic conquests were remembered. Like their ancestors in the jhiliya (time of ignorance) before the coming of Islam, they composed and preserved poems and songs to celebrate heroic deeds. As well as these ancient, traditional themes, the Muslims could also remember their victories as clear evidence that God was on their side, the deaths of their enemies and the vast quantities of booty they amassed all being evidence of divine favour: no one could question the essential rightness of what they had done. They also preserved, elaborated and even made up accounts to serve new purposes, to justify claims to stipends or rights to enjoy the proceeds of taxation. Men who could prove that their ancestors had participated in the early conquests felt entitled to salaries from public funds; the inhabitants of cities might hope for lighter taxation because they had surrendered peacefully to the Muslim armies. In short, the stories of the conquests were preserved, not because of interest in producing a clear historical narrative, but because it was felt to be useful. Correspondingly, material that was not useful, the exact chronology of events, for example, was consigned to oblivion.
The next stage was the collection and writing down of this oral material. It is not easy to say exactly at what stage this occurred because Arabic, like English, uses expressions like ‘He says (in his book)’ so verbs of speech may actually refer to writing, but the process was certainly begun during the eighth century. These collections seem to have been made for antiquarian reasons, to preserve the record of the early years of Muslim rule in Iraq or Egypt when the memories were fading and there was a risk that much of this important story would fall into oblivion. The practical considerations that had led to the preservation of these traditions in the first place had, by now, become largely irrelevant, but of course the collections assembled by these editors necessarily reflected the purposes of the earlier narrators.
The ninth and early tenth centuries saw a vast explosion of writing and book production. The introduction of paper to replace parchment (dried animal skins) as the main writing material 1 meant that writing became both quicker and cheaper. Historical writing increased as part of this, reflecting a growing demand for historical information, both in the circles around the courts of the caliphs and among the wider literate society of Baghdad and the rest of Iraq. In Baghdad, where there was a real book trade, it became possible to make a living writing for a wider public, not just for a rich patron. Knowledge became professionalized, in the sense that men could make a career out of it.
Knowing your history, being an authority, could lead to an appointment at court. The historian Balādhurī, whose Book of Conquests is one of the main sources on which we rely, seems to have made a living as a nadīm or ‘boon-companion’ at the Abbasid court. Every boon-companion was expected to bring some knowledge, expertise or talent to the party: some were poets, some authorities on quaint or unusual Arab vocabulary or the characteristics of different geographical areas. Surely Balādhurī owed his position to the fact that he knew so much about the conquests and other areas of early Islamic history, for he was a great authority too on the genealogies of the ancient Arab tribes. This was all despite the fact that he does not seem to have come from an important family and was not himself a descendant of the participants. The greatest of these compilers was Tabarī (d. 923). He was a Persian who came from a landowning family in the area along the south shores of the Caspian Sea. He spent most of his adult life in Baghdad and became a great authority on two of the most important areas of Muslim learning, the