The evidence about pre-Islamic Arabian populations emanates from two perspectives: (1) the writings of peoples from outside Arabia who, across the 1,500 years from the Assyrians in the ninth century BCE to Islam's rise in the seventh century CE, recorded many stories about Arabians, and (2) voices from within the Arabian Peninsula itself, preserved in inscriptions from as early as the eighth century BCE. Both bodies of sources contain numerous and intriguing references to an array of ancient peoples whose names resemble ‘Arab’, and it may seem logical enough that Arab history can be written by synthesising the material, but, curiously, it has not transpired this way. Several generations’ worth of modern analysis produced a number of different narratives, and the field remains divided between surprisingly divergent opinions. Different approaches to the evidence enabled some to argue that Arabs existed across Arabia since time immemorial as its original (or at least very early) inhabitants, while another imagines Arabs as a distinct militarised religious community that formed around the ninth century BCE. Another posits that Arabs only emerged as a group circa the first century BCE in south-central Arabia and only consolidated a sense of political unity in the fourth century CE; whereas a further body of scholars imagines the first Arabs as a conglomeration of north-west Arabian Bedouin who formed a loose sense of a community around a shared oral/poetic culture between the fourth and sixth centuries CE. Yet another theory argues that Arab ethnogenesis occurred not in Arabia, but on the Byzantine–Syrian frontier in the fifth and sixth centuries CE, whereas other recent observations radically hint that ‘Arab’ communities did not emerge until after the dawn of Islam. The theories are rather mutually exclusive, but the origins of an Arab community must lie somewhere, and hence the array of options are in need of better resolution. The task inspires this chapter.

As Christian Robin and Michael Macdonald each note, we cannot recreate Arab history by simply searching for instances of the word ‘Arab’ in ancient records, as this produces an unmanageable array of references, many of which do not point to what can be understood as a community of Arabs.