actually happened. In a way more interesting, however, is what they offer in terms of the attitudes and the memories different groups preserved and cherished about what had gone on.
The Middle East conquered by the Muslims in these early decades was a multicultural society, a world where different languages and religions coexisted and intermingled in the same geographical area. After the success of the conquests, the language of the new elite was Arabic. Even for government, however, the existing administrative languages - Greek in Syria and Egypt, Middle Persian (Pahlavi) in Iraq and Iran, Latin in Spain - continued to be used for the business of government. After a couple of generations, however, this began to change. Around the year 700, sixty or more years after the earliest conquests, the Umayyad caliph Abd al-Malik (685-705) decreed that Arabic and Arabic alone was to be used in the administration. The decree was surprisingly effective. From this time, anyone wanting a position in the expanding bureaucracy of the Islamic state, whether they were Arab or non-Arab by descent and upbringing, needed to be able to read and write in Arabic. The inscriptions on the new style, image-free coins and the roadside milestones were all in Arabic. There was no point for most people in learning Greek or Pahlavi because there were no career opportunities in them. It was around this time, in the early eighth century, that the Arabic traditions of the conquests began to be collected and written down.
The momentous events of the seventh and eighth centuries inspired an extensive Arabic-language literature which claimed to describe what had happened then. But the memories and narratives of the Muslim conquest were more than the records of ‘old forgotten far-off things and battles long ago’. They were the foundation myths of Muslim society in the areas that generated them. They were developed because they helped to explain how Islam had come to the land and to justify the defeat and displacement of the previous elites. These accounts did not deal with ethnogenesis, the birth of peoples, as Latin historians of the early medieval West did, but rather with the birth of the Islamic community. They preserved the names of the heroes who had led the armies of the conquest and were the founding fathers of the Islamic state in their area; the names of the companions of the Prophet, men who had met and heard Muhammad and brought with them a direct connection with his charisma; the names of the caliphs who had turned Islamic armies in their direction.
These narratives do provide information about the course of events, and just as interestingly they show how these events were remembered by later generations, how they saw the beginnings of the community in which they lived. Looked at as a form of social memory, the distortions and legends that can seem at first sight an obstacle to our understanding can be seen instead as reflecting the attitudes and values of this early Muslim society.
In the form in which they have come down to us, these accounts were edited in the ninth and early tenth centuries; that is, between 150 and 250 years after the events. The Arabic narratives are rarely simple accounts written by a single author and telling a straightforward account of events. They are actually multi-layered compositions that have gone through different stages of editing and elaboration for different purposes at different times. At the risk of oversimplifying a complex process, the narratives seem to have gone through three stages of development. The first was oral transmission of traditional stories of heroic deeds in battle. Such traditions were often preserved within tribes and kinship groups or among Muslims who had settled in particular areas. In part they may have preserved these memories as their predecessors had treasured accounts of the battles of the Arab tribes in the years before the coming of Islam. The ancient tradition