The social organisation of pre-Islamic Arabia was closely bound up with considerations of religion, and it is in this area that problems of methodology and source criticism are most acute. Issues such as borrowing from more advanced civilisations, the starting-points and relative antiquity of religious forms, the roles of animism and totemism, and differences between sedentary and nomadic peoples have been and remain highly controversial. In many cases important arguments involve value judgements about nomads and, similarly, supposed distinctions between ‘high’ and ‘low’ forms of religious expression. There is also the problem that the Arabic sources, where the vast bulk of our source material is to be found, can hardly be said to offer an objective view of pre-Islamic religion. The folly of idolworship and the credulity of its adherents are routinely stressed in stereotyped ways. One tale describes how a tribe fashioned an idol out of dried curd mixed with dates and clarified butter (h. ays) and worshipped it for a time, but eventually devoured it during a famine, leading to a poet’s wry comment: The tribe of Hanifa ate their lord When dearth and hunger swept the land, Fearing naught for consequences From their lord’s avenging hand.33 Inspired by Koranic criticisms,34 Arabic sources also present bedouins as indifferent to matters of faith.35 Arabian polytheism took several forms,36 one of which was stoneworship. Greek and Syriac sources presented this as adoration of lifeless rocks, but such objects were not deities in themselves, but their dwellingplaces or the focus of the rituals of the cult. Offerings were made at the site, and ritual observances included circumambulation of the stone. The best-known example is of course the Ka‘ba in Mecca, but we are told that other places had such cultic foci.37 These foci were often surrounded by a sacred territory, usually calledh. aram in the north andh. awa in the south. These were precincts associated with the sanctity of worship and sacrifice; violence and killing, including hunting, were forbidden there. Holy men were in charge of these precincts, and their descendants enjoyed special religious esteem.38 Also prominent was religious observance revolving around idols – again, with the idol probably representing the deity being worshipped. The names of many idols are known from ancient poetry and from later prose works drawing on this verse. Important new details pertaining to Yathrib (Medina) may be indicative of a more general pattern: there, clans each had an idol in a room belonging to the whole clan, where the idol was venerated and sacrifices made to it. People also had wooden idols in their homes, making similar observances at that level. To offend the idol was an offence against the honour of the head of the house and a matter for retaliation, and there is some evidence that these idols were intended to be figures of ancestors. There was thus a hierarchy of idols, corresponding to the social status of their owners.39 There is good evidence of star-worship and astral divinities as well. The widely venerated al-Lat (a sky goddess) and al-‘Uzza (possibly the morning star) may have been representations of Venus, and Byzantine polemics against Islam claim that the Islamic slogan All¯ahu akbar (‘God is great’) has as its origin a cry of devotion in astral religion.40 The worship of astral divinities has also been connected with the veneration of idols. The attitude of the ancient Arabs towards their gods was entirely empirical and pragmatic. Although they did consider problems of human existence and the meaning of life,41 they did not look to their deities for the answers. They regarded their gods as the ultimate sources of worldly phenomena beyond human control, such as disease, rain, fertility, and personal and communal adversities of various kinds; they worshipped the gods in expectation of their assistance, but they did not revere them or consider that they owed unwavering commitment to them.42 Monotheistic religion was also known in Arabia from an early date. The influx of Jews into Arabia is difficult to trace, but probably had much to do with the failure of the Jewish revolt and the destruction of the Temple in ad 70, and the gradual spread of Christianity over the next three centuries. In south Arabia, Judaism enjoyed considerable success in the fifth and early sixth centuries, and to the north there were various important Jewish communities, notably at Yathrib. Judaism seems to have had deep and powerful roots there, judging from reports that in pre-Islamic times the Jews there had three times as many fortified compounds (qus. ¯ur) as all the other non-Jewish clans combined,43 and that in the latter half of the sixth century the Jewish clans of Qurayza and al-Nadir collected taxes from the other tribes.44 The question of Jewish influences in Arabia and on Islam has become highly sensitive in modern scholarship, but there can be no doubt that such influences were profoundly important; the Koran itself contains many tales and accounts of Jewish origin, as also do early Islamic religious lore and scholarship.45 The Christianisation of the Roman empire in the fourth century opened the way for the large-scale spread of the faith along and beyond the empire’s frontiers, including Arabia.46 Along the Syrian desert fringe from the Red Sea to the Euphrates, it spread to the Arab tribes via monasteries and wandering missionaries, primarily monophysite. In some cases, as with the Banu Taghlib and Ghassanids, entire tribes converted; some tribal settlements such as al-Jabiya and Jasim, south of Damascus, also became ecclesiastical centres. These tribes were familiar with at least basic observances, yet remained completely within Arab tribal culture as well.47 Along the Iraqi frontier the spread of Christianity was somewhat slower, perhaps because a network of Nestorian monasteries in the area took longer to appear than had been the case among the monophysites.48 Still, the Lakhmid base of al-Hira was the seat of a bishopric by 410.49 Further south, there were major Christian communities at such centres as Najran and Sanaa, and small ecclesiastical outposts along the Arabian coast of the Persian Gulf. Specifically monophysite or Nestorian forms of Christianity were practised in such centres, but elsewhere the Arab tribesman’s main contact with the faith was through individual monks and hermits, and there confessional boundaries may have been less sharply drawn.50 Two other beliefs – which were influenced by Judaism and Christianity, yet remained distinct from both – revolved around a ‘high god’ and around h. an¯ıf¯ıya. Little can be said about belief in a ‘high god’ in ancient Arabia, apart from the fact that, as elsewhere in theMiddle East,51 some held that a god called Allah had a certain dignity and status above the other deities of the Arabian pantheon and was extolled as a god to whom one could turn in case of particular need.52 Onh. anfya there is more information.53 The Koran makes it the religion of Abraham and associates it, on the one hand, with belief in a single God and, on the other, with rejection of idolatry and repudiation of worship of the sun, moon and stars. In particular, and most importantly, h. an¯ıf¯ıya reflects not the pragmatic attitude towards religion described above – in which the god(s) were worshipped in expectation of help with worldly needs beyond an individual’s control – but rather a submissive devotion to and faith in God for his own sake. Nevertheless h. anfya is distinct from Judaism and Christianity: in several passages of the Koran, its adherent (ah. anf ) is equated with aMuslim, and in one variant to the Koranic text, h. anfya replaces Islam as the ‘true religion’.54 Other sources suggest that there were h. anfs in various parts of Arabia, that the movement was one of individuals rather than religious communities, and that Mecca was important to its adherents. Other details are less reliable, and there is no evidence to link h. anfya with south Arabian inscriptions attesting to worship of a god called al-Rahman, ‘the Merciful’, one of the Islamic names for God. But the fact that the tradition on theh. anfs makes some of them doubters or enemies of Muhammad suggests that it should not be dismissed entirely as later prophetic annunciation or the tidying up of a pagan past. Of interest in this respect is the testimony of Sozomen, who died before 448; writing from the vantage-point of Gaza in southern Palestine, he offered the following comments on Arab religion: It seems that the Saracenswere descended fromIshmael, son ofAbraham, and hence were originally called Ishmaelites. Their mother Hagar was a slave, so in order to hide the shame of their origin they took the name of Saracens, pretending to be descended from Sarah, the wife of Abraham. As such is their descent, they practise circumcision like the Jews, abstain from eating pork, and adhere to numerous other Jewish observances and practices. In so far as they in any sense diverge from the observances of that people, this arises from the passage of time and their contacts with other neighbouring peoples . . . It seems likely that with the passage of time their ancient customs fell into disuse as they gradually took to observing the customs of other peoples. Eventually, when some of their tribe came into contact with the Jews, they learned from them the facts of their true origin and returned to observance of Hebrew custom and law. In fact, even at the present time there are some of them who live their lives in accordance with the Jewish law.55 The connection with Judaism may reflect an inclination to associate false belief with the machinations of Jews.56 As to theAbrahamic religion attested in the text, while the connection is circumstantial and Sozomen wrote long before the testimony of the Koran, the Islamic scripture may refer to continuing monotheistic trends in Arabia that it wishes to distance from earlier monotheistic faiths now viewed as rivals.