Forgotten Arabia: Himyarite Yemen and Early Islam

By ANDREW HAMMOND NEWSLETTER

Pre-Islamic Yemen has attracted relatively little attention from scholars of early Islamic history, including those who regard themselves as radically revisionist. The temptation to overlook south Arabia is understandable in that the familiar Late Antiquity regions of the Levant, Mesopotamia and Egypt were the main target of the Hijazi conquests outside Arabia. Those regions provide a form of contemporary written record of the early empire’s fortunes (controversies over its evaluation aside) and they are where the formation of classical Islam of the legal schools, theological debates and philosophy takes place. Fortunately, they are also regions that have always been relatively easy to access for historians. Yet the epigraphic evidence regarding the far south is considerable, including thousands of graffiti and rock drawings in mountains north of Najran in southern Saudi Arabia, as well as within modern Yemen. More and more of this material is being published, offering the chance to refine understandings of south Arabian history during the sixth century CE and examine further south Arabia’s role in the early rise of the Islamic movement.

Sixth century Arabia was politically volatile as the site of intense imperial rivalry. Greek Byzantium and Sassanid Persia competed territorially, and to that end patronized different Christian sects – the Byzantines had their Chalcedonian rite and the Persians, though patronizing Zoroastrianism, had the Church of the East (sometimes called Nestorian) under their control. Within Byzantine territories another breakaway church had established itself, the Jacobite rite. Constantinople and Ctesiphon also had their allies among the Arabic-speakers of the Levant (the Ghassanids based at al-Jabiya) and Mesopotamia (the Lakhmids based in al-Hira) who were expected to keep a rein on the pastoralist tribes further out in the desert regions. The empires used the churches and their proselytization efforts as bridgeheads of political influence throughout the Arabian Peninsula. For example, the various Church of the East sites established the length of the Gulf littoral functioned as outposts of Sassanian power and outposts of the Jacobite church did the same for Byzantium. Both are thought to have established a presence in Najran on the edge of Yemen. But they were also of course agents of intense theological conflict over the relationship of Christ to God.

United Yemen Extends Northward

In Yemen around 300 CE the Ḥimyarite state based in Ẓafār (near modern Yarim in Ibb governorate) managed to unify the country, bringing under its control Saba (Marib), Hadramawt and the south, imposing one calendar, one script (Sabaic) and one religion. That religion was Judaism, which replaced the previous polytheistic local cults in the public record of inscriptions. But it was what French scholar Christian Julien Robin calls a “bare monotheism” in that most statesmen who erected monuments did not use the language of communal belonging to the notion of Israel and its tribes. Indeed, the late Semitic philologist A. F. L. Beeston called this Ḥimyarite religion “Raḥmānism”, in reference to the south Arabian word for the monotheistic God; Robin prefers “Judaeo-Monotheism”. Importantly, this choice amounted to a rejection of the Christianities on offer, each of which brought implications of outside suzerainty, whether Byzantine, Persian – or Aksumite, the north Ethiopian state that adopted Christianity around the same time that Ḥimyar united Yemen.

Ḥimyarite Yemen’s links to central Arabian regions existed from an early stage. It extended into areas of the three regions outside immediate Byzantine and Persian control – the Hijaz in the west, Najd in the centre and the Ahsa oasis in the east – and the Kinda tribal confederation appears to have acted as a local proxy. However, things get interesting in the 500s when Aksum, located just across the Mandab Strait at the south of the Red Sea, managed to install a Christian vassal on the throne around 519. His successor Yūsuf Asʾar Yathʾar (known as Masrūq in Syriac and Dhū Nuwās in Arabic) tried to return the state to Judaism, famously killing a large number of Najran Christians in 523 – an event commemorated in various Christian church calendars and understood by the Islamic tradition to be the event in which believers are burned in a trench (ukhdūd) in the Quran (85:4-8). An Aksumite army under King Kālēb then intervened to depose Yūsuf and restore Christianity, but by 535 a man called Abraha, the local leader of the Aksumite army, had seized control himself, continuing the Ḥimyarite state but as a satrapy of Aksum until he died sometime around 565-570.

Inscriptions indicate that Abraha managed to win Aksumite, Byzantine and Sassanian recognition of his rule and reestablishment of Ḥimyarite power in central Arabia. One that lies 230 km north of Najran and dated to 552/3 shows that he had extended control over various areas including Yathrib (Medina). These campaigns may provide the historical context for another Quranic story detailed in the Islamic tradition in explanation of Sura 105’s reference to an incident in which God sent birds to pelt unspecified men-on-elephants with stones. Histories such as that of al-Tabari claim Abraha’s campaign targeted the Meccan shrine and in the year of the Prophet’s birth, which could be correct, or may amount to a salvation history reworking of the known earlier event. The latest dated Ḥimyarite text, from 559/560, appears to refer to the Sanaa church known in Arabic as al-Qalīs (or al-Qulays), which the Islamic tradition claims prompted Abraha’s desire to subjugate Mecca as a rival attraction for pilgrims.

From here to the Islamic era the fate of Ḥimyarite Yemen is a blur. Yemen shifted to Persian tutelage in the 580s when it began to be administered directly from Ctesiphon, while the Islamic tradition talks of south Arabian tribes dispersing northwards, which would be one way to explain their Arabization. Robert Hoyland and Fred Donner note that the leading military figures during the Arab-Islamic conquests appear to have been from Yemen and the oases towns of west and central Arabia, and a large portion of the fighters in early battles such as Yarmouk in 636 were from Yemeni tribes.

The Fate of Ḥimyarite religion

The key question is what happened to Ḥimyarite religion in these decades before the Islamic dispensation. Robin makes an important observation from the inscriptions – that the language of the Trinitarian invocation shifts notably between the time of Kālēb’s Christian restoration and the period of Abraha, from “in the name of Raḥmān, his son Christ the Victor and the Holy Spirit” to “in the name of Raḥmān, his Messiah, and the Spirit of Holiness”. There are some variations but they also avoid mention of the Son and tend to use Messiah (using the Syriac mešiḥā, a cognate of the Arabic masīḥ). Robin notes that this change conforms to the Judaeo-Monotheistic Ḥimyarite tradition of no public declaration of God having progeny, which would be more familiar and acceptable to its ruling elites during the Christian era. Yet it is also in harmony with the Quranic dogma regarding Jesus, whose status as a son of God is rejected in favour of being His messenger (4:157) and who is called al-masīḥ on nine occasions (al-masīḥ ibn maryam, 5:17, 5:72, 5:75. 9:31; al-masīḥ ʿīsā ibn maryam, 3:45. 4:157, 4:171; al-masīḥ, 4:172, 9:30).

In two articles published in 2016, Patricia Crone argued that most of the list of non-standard beliefs about Christianity provided in the Quran – that he was a prophet, that he was not the son of God, that he was not crucified, that Mary was of Levite descent – came from within what she called Jewish Christianity, that is, Jews who upheld devotion to Jesus while retaining adherence to Jewish identity and law. She concluded that Jews upholding such beliefs “must in fact have been present in the localities in which the Messenger was active”. But Crone only discussed these contacts in terms of Medina, where the Islamic tradition locates the Jewish tribes that entered into a covenant with Muhammad after he moved there with his followers. Yemen, where Jewish and Christian theology most certainly coexisted and jousted, and whose rulers consistently spread their authority over tribal Arabia from the 300s to 500s, is not mentioned at all. The Yemeni experience could also speak to the Quranic concept of the ḥanīf, the epithet of an undefined basic monotheism that is applied to Abraham eight times and Muhammad twice (10:105, 30:30). Abraham is explicitly described as “neither Jewish nor Christian but ḥanīf” (3:67) and “not among the polytheists (mushrikūn)” (3:95).

Further, epigraphy scholar Ahmad Al-Jallad has argued that the Quranic basmala (bismillāh al-raḥmān al-raḥīm) has its origins in South Arabian monotheism. Analysing a graffito from Al-Ḍāliʿ in southwest Yemen published in 2018, Al-Jallad rereads it to suggest it dates to the late 500s or early 600s, written in a form of the usual Sabaic script though possibly in the Ḥimyarite vernacular or a form of Old Arabic, but again reflecting a monotheism that is neither Judaic nor Christian. Rather than “in the name of Allāh, the raḥmān , the raḥīm [merciful], lord of the heavens,” he suggests it says “in the name of Allāh, (who is) the Raḥmān/Have mercy upon us, O lord of the heavens” and that the Quranic terminology was originally intended with the same meaning, before raḥmān was reinterpreted as an adjective. The intent behind such a terminological strategy would have been to unite the northern and southern Arabian names for God in one invocation, equating Allāh, the monotheistic god of Arabic-speaking northwest Arabia, and Raḥmān of south Arabia as merely different names for the same single divinity. This political acted expressed in theology came amid a period of intense imperial conflict as the Byzantines and Sassanians launched into war from 602 to 628. What degree of cultural affinity the tribes of Yemen felt for those living to the immediate north isn’t clear, but within a few decades many of them were participants in the Islamic conquests and presumably speaking Arabic. Al-Jallad notes that the Quran twice mentions the fact that there are some among its intended audience who refer to God as al-raḥmān (17:110, 25:60). Indeed, the final section of Sura 17, addressing either Muhammad or a congregation of believers, is explicit that God can be addressed as Allāh or al-Raḥmān in prayer. In later understandings the latter term can only be read as an epithet of God.

We also have the fact that Ethiopic is one of the source languages for foreign words in the Quran. The extent to which foreign words had already been nativized and or were innovations of the Hijazi movement is difficult to know. Most of the terms in question are Syriac in origin, but Ethiopic also features, one of the most notable being ḥawārī for the disciples of Christ, while words from south Arabian languages such as wathn (idol) are also present. Earlier studies by modern scholars suggested such words came to Muhammad through events such as the refuge that some of the early Meccan community sought in Aksum around 616, as recounted in Tabari’s History. But as Arthur Jeffery noted in The Foreign Vocabulary of the Qur’an (1938), there is no reason why such words would not have been extant in the southern Hijaz before the era of the Prophet, as the extensive Yemeni involvement in Arabia throughout the 500s would suggest. The tradition itself is littered with references to Yemen/Aksum, even if they are, as Hoyland says, the “haziest of recollections”, from the first muezzin Bilal al-Habashi, whose mother was said to be an Ethiopian captured during Abraha’s expedition, to the Yemeni companion and Jewish convert Abū Isḥāq Kaʿb ibn Māniʿ al-Ḥimyarī, also known as Kaʿb al-Aḥbār

Levantine Obsessions

This has been no more than a tour d’horizon of issues indicating the relevance of Yemen to the early Islamic period and the Quranic thinking on monotheism. But it is enough, I hope, to suggest the extent of misplaced energy that some revisionist scholars have given to north Arabia and attempts to tie the Hijazi movement to Palestine, and Petra in particular – from Stephen Shoemaker, to Dan Gibson, to an upcoming book by Peter Von Sivers. Their concern has been to locate the Prophet’s upbringing in the north or to establish the Levant as the intellectual environment which produced much if not not most of the Quran. Which seems to be an approach based on a pre-existing assumption that civilizational strength lay in the knowable lands with the accessible literature, while everything to the south had little to offer but varying degrees of cultural aridity.