through a social structure that amounted to a caste system of four social estates: priests (a class that included judges), landed gentry (the warrior class), cultivators, and artisans (a class that included merchants). As long as the four estates were kept discrete, the Sasanians did not ordinarily persecute members of other religious communities, because the communities of Jews, Buddhists, Manicheans, and Christians simply had no place in the stratified hierarchy. The Romans and Iranians, then, cherished two imperial visions that were never really going to get along with one another.
The roots of the animosity between Rome and Iran stretched back to Rome’s encounter with the Parthians (248 BC–AD 224), the predecessors of the Sasanians, as the Romans expanded into the eastern Mediterranean. The Sasanians (ca. AD 224–ca. AD 650), originally a landholder family from the highlands of southwest Iran, inherited this animosity and pursued it with zest. At their greatest extent, Sasanian rule extended from the Oxus River in the northeast to the Euphrates in the Fertile Crescent, giving them effective control of the silk trade with the Mediterranean.
The sixth century witnessed an increase in hostilities. In AD 531, Khusro I Anusheruwan (known as Kisrā Anushīrwān in Arabic sources) (r. AD 531–79) wrested control of Sasanian Iran from his father. Between AD 540 and 562, during the reign of Emperor Justinian (r. AD 527–65), Khusro’s Sasanians invaded Syria, formally a province of the Roman Empire. A peace treaty was negotiated in AD 562. In the last years of his rule, between AD 570 and 578, Khusro I conquered the kingdom of Ḥimyar in the Yemen and expelled its Axumite overlords. But by the end of the century, the Sasanian emperor Khusro II Aparviz (r. AD 591–628) had to depend upon troops provided by the Roman emperor Maurice in order to recapture his throne, yet when Maurice was assassinated in AD 602, Khusro II was quick to invade Roman Syria.
The Romans were not idle during the sixth century. Their involvement in Arabia was largely through the manipulation of proxies, including the Axumite kingdom, the Hujrids of Kindah, or the Jafnids of Syria. In part, this was a natural consequence of established Roman policy in the region implemented through the province of Roman Arabia, but it was more immediately a consequence of geography: the Syrian desert, devoid of food and water, was not the place for an army to cross, whether Roman or Sasanian. With the terrain so inimical to conventional warfare, both sides resorted to the cultivation of alliances and diplomacy.
The Romans and the Iranians developed links with two powerful Arabic-speaking clans at either end of the northernmost points of Arabia: the Jafnids of Ghassan in the west and the Nasrids of Lakhm in al-Ḥīrah near the Euphrates in the east. The Romans made the Jafnids into imperial foederati, confederates, bestowing a kingship upon them and recognizing them as phylarchs (tribal leaders). The Jafnids were charged with restraining the Arabic-speaking tribes and preventing them from interfering with trade routes and the collection of tax tribute. They supplied the Roman army with troops and waged war against the Nasrids of Lakhm, who acted on behalf of the Sasanians. Nasrid influence stretched along the eastern Arabian littoral and even into Oman. Their influence has been detected in Yathrib (the settlement that under Islam was to become Medina), to the extent that in the sixth century the Nasrids may have appointed a governor there. Roman and Iranian interest in the Arabian peninsula did not stop with the recruitment of elite warrior-rulers from the north to do their dirty work for them. Their activities extended as far as Yemen and Ethiopia, or the kingdoms of Ḥimyar and Axum.
AXUM AND ḤIMYAR
We do not know much about the kingdom of Axum.9 Its ruler was known as the negus and his territory included modern Eritrea and the northern part of Ethiopia. It may even have stretched farther west into Sudan. Christianity took hold in Axum in the fourth century. From the fourth to sixth centuries, Axum grew astonishingly rich in African products such as gold, ivory, rhinoceros horn, and tortoiseshell.
The fate of Axum is closely tied up with the history of South Arabia. In about 110 BC the South Arabian tribe of Ḥimyar formed itself into a kingdom and brought South Arabia under its control. By the third century AD, under the rule of Shammar Yuharʿish, Ḥimyar had conquered the southern Arabian region of Ḥaḍramawt and expelled the Axumites from the Yemeni coast. In the following centuries, the Ḥimyarites sought to extend their influence over the tribes of the interior, venturing deep into the Yamāmah and maybe even as far as Ḥajr (modern-day Riyadh) (see Maps). During the fourth century they converted to Judaism, and in the fifth century they exerted their dominion over Maʿadd, the main tribal confederation of the northern Arabs of Najd, by installing the Hujrids as their proxies under a chieftain of the powerful tribe of Kindah.
With Roman help, Kaleb Ella Asbeha, negus of Axum (r. ca. AD 520–40), invaded Ḥimyar, placing a Christian king on the throne. This led to a reprisal from the Jewish royal family, and the new Himyarite ruler, Yūsuf, slaughtered the Axumite garrison and in AD 523 executed several hundred Christians, who became known as the Martyrs of Najrān. This event led to an Axumite invasion in AD 525, the death of Yūsuf, the eventual replacement of the Himyarite kingdom with an Axumite protectorate, and enforced conversion to Christianity.
One of the Axumites who had remained in Ḥimyar after the return of the negus Kaleb to Ethiopia was a man named Abraha (in Arabic sources: Abraḥah), who assumed control of the protectorate. In AD 547, he received ambassadors from Rome, Iran, and Ethiopia, and from the Nasrids and Jafnids. In ca. AD 550, he constructed the Christian cathedral of Sanaa, and five years later mounted a major expedition into central Arabia, but that expedition resulted in his defeat and retreat.
Perhaps the most notable construction project undertaken by the kingdom of Ḥimyar was the Maʾrib Dam, which was 650 meters wide and 15 meters high. Maʾrib (presumably the church and not the dam) was where, in AD 552, Abraha chose to receive the delegations of ambassadors, but sometime between AD 575 and 580, during the childhood of Prophet Muḥammad, the dam is reported to have burst and not been repaired. The collapse of the dam signaled the end of the kingdom of Ḥimyar and may have led to a massive influx of mercenaries and professional soldiery maintained by the kingdom into central and northern Arabia.10
During the second half of the sixth century, the frontiers between Rome and Iran were destabilized, and the interior of Arabia was thrown into turmoil. On the eve of the advent of Islam—and toward the end of ʿAntarah’s life—the Jafnids were overthrown by Rome in AD 573, and the Nasrids by the Sasanians in AD 602. Ḥimyar had been unable to repair the dam that it so crucially depended on. And in AD 604, the Sasanian army was defeated by an army of Arabian tribesmen at the Battle of Dhū Qār.
ARABS IN ARABIA
The term “Arab” is apparently an old one. Its earliest appearance is thought to occur in Assyrian texts from the seventh century BC, though this has been disputed. But there is no indication in this or any of its subsequent occurrences that it is an ethnonym, i.e., the name of an ethnic group. In fact, it is likely that for many centuries inhabitants of Arabia were not widely or even automatically known as Arabs. Other names, such as the Greek names homēritai (i.e., Ḥimyarites) and sarakēnoi (Saracens), predominate—presumably they are designations of specific groups of inhabitants of regions of Arabia.
Two basic solutions to the identity of the Arabs has been proposed, one maximalist, the other minimalist. The maximalist solution is to take “Arab” as an atemporal designation of transhumant tribespeople, that is, nomadic pastoralists who herded camels, or other domestic animals such as goats and sheep, in designated winter and summer camping grounds. Yet the probable percentage of the population of the Arabian peninsula that may at any one time have been nomadic was small, even allowing for returns to nomadism after a period of sedentarization. As Donner notes, “Most Arabians … are, and have been, settled people.”11 If the “Arabs” were transhumant tribespeople, they would not have been particularly numerous. And if the “Arabs” were nomads, we should not presume that they would necessarily be Bedouins, i.e., people of the desert. Bedouins could, counterintuitively, be settled for much of the year, and could also share some of the features of the nomad’s lifestyle, such as camel pastoralism.12
The word “Bedouin” represents the hinge point at which the maximalist solution becomes minimalist: its historical frame of reference is specifically the fifth and sixth centuries AD; its geographical frame is North