‘Rahman’ before Muhammad: A pre-history of the First Peace (Sulh) in Islam

Abstract

This article examines how the divine epithet ‘Rahman’ was invoked in public inscriptions and oaths in pre-Islamic Yemen. Between the first and the sixth centuries ce, with the spread of Christianity across the Roman empire and Abyssinia, and the subsequent rise of Jewish and Christian kingdoms in southern Arabia, the use of ‘Rahman’ was gradually biblicized. By tracing this history, this article opens a window into the use of this theonym in the Quran and the controversy surrounding its use in the first formal treaty in Islam, the Peace Treaty (Sulh) of Hudaybiyya.

Introduction

According to numerous classical Islamic traditions, the Meccans forced the Prophet Muhammad to omit the oath ‘in the name of Allāh, al-Raḥmān al-Raḥīm’—a defining feature of Islam—from the Peace Treaty (Sulh) of Hudaybiyya.¹ As noted in the framework article for this special issue, the Prophet was attempting to negotiate terms so that he and his followers would be allowed to enter Mecca for their pilgrimage. At the moment when Muhammad dictated the oath, Suhayl b. ‘Amr, the Qurayshi emissary, said ‘I do not recognize this one. Write rather, “In thy Name O God”.’² This use of ‘/allāhumma/’ can be explained by the fact that it refers not to a specific deity, but rather is a generic term for the highest god, which perhaps even carried within it a ‘pantheistic’ notion of divinity. In any case, Muhammad agreed to this compromise over his own followers’ strenuous objections. The matter did not end there. When the treaty was about to be signed, Suhayl refused to accept Muhammad's epithet ‘Messenger of Allāh’ and insisted that Muhammad must sign with his familial name, effectively giving up his status as a (biblical) prophet.³ Essentially, Suhayl forced Muhammad to sign a treaty to be able to perform pilgrimage to the Ka‘ba as an ordinary Meccan clansman defined by kinship, not as a biblical prophet with a privileged connection to a deity. Although the Qurayshi messenger proclaims that he ‘does not recognize’ any Rahman, I argue that he most definitely did, but as an identifiable god that was not Meccan nor accepted in the Ka'ba. For Suhayl, al-Rahman was clearly marked as a monotheistic god that mandated conversion and rejected the system of worshipping several deities (shirk) established around the Ka‘ba in Mecca. When the Prophet Muhammad encountered issues with using al-Rahman in his oath at Hudaybiyya, it was because al-Rahman represented something very specific to the Meccan Suhayl b. ‘Amr who was to share in the oath with Muhammad. To Suhayl, al-Rahman was not compatible with a polytheistic worldview but rather signalled an exclusionary epithet. However, the name not only signified a god different from those venerated at that time in the Ka'ba, it was also affiliated with a Christian warrior king from the South who had threatened Mecca with invasion a generation earlier, intending to destroy the Ka'ba. This incursion from the South had threatened not only the political power of the Quraysh, but also the status of their sacred sanctuary.

The event at Hudaybiyya represents a culmination in the development and gradual biblicization of the term ‘Rahman’. For the purposes of this article, I take biblicization to mean the emergence of what Jan Assmann calls the ‘Mosaic distinction’. This distinction is produced by the regulative idea that all other religions and gods are false, that the divine name cannot be translated across pantheons, and the only path to the divine lies through ‘faith’ in the truth of biblical doctrine.⁴ There is no ‘faith’ without biblicization. Furthermore, in theoretical terms, biblicization can be understood as the emergence of transcendentalization, that is, a shift away from the belief that the divine is immanent and palpable in the key elements of cosmos and nature.⁵ From a transcendentalist perspective, immanentist cultures make the mistake of interacting with the divine via concrete and ritual means, such as placating and entreating divinities with sacrificial gifts, instead of keeping ‘faith’ and awaiting salvation in the afterlife. While a detailed discussion of this theoretical framework is outside the scope of this article, it is used here to underline the observation that while the Quranic use of al-Rahman fully reflects a transcendentalized deity, earlier pre-Quranic occurrences of the theonym reflect an emerging religiosity that is on the cusp of the Mosaic distinction, but not yet fully there. In other words, al-Rahman in the Judaicizing texts from South Arabia can, at times, be read as distinctively immanentist and monolatric, even though by the time Islam arrives, it would undeniably take on a transcendentalist and monotheistic form.

In what follows, I outline the origins of the term ‘Rahman’, its use in monolatric (worshipping only one god while recognizing the existence of others), Jewish, and Christian South Arabian texts, and how it eventually came to signify a monotheistic transcendentalized god. All of this transpired in a political and religious context that was, by comparison, very different to the one that produced the Mughal idea and institution of ‘Total Peace’ (Sulh-i Kull). In the case of Hudaybiyya and al-Rahman, the biblical monotheistic paradigm was clearly predominant. By the seventh century, imperial Christianity had brought an end to temple paganism across the Mediterranean and related regions, including East Africa and South and North Arabia. In the words of Aziz al-Azmeh, Arabia Deserta was the last ‘reservation’ of paganism left in the region.⁶ The pagan Arabs were fully aware of the fact that the religious world outside the desert was a hostile one and that making ‘peace’ on terms of biblical monotheism meant ‘converting’ to it.

Based on extant material and inscriptional evidence, I demonstrate how the kingdom of Himyar (in modern day Yemen and southwestern Saudi Arabia) made use of this theonym to describe their god as Rahman-an during the fourth to sixth centuries ce.⁷ While Rahman-an was used as an epithet of the main god in Himyar in the fourth century as forms of Judaism spread through the region, it only came to mean the ‘one true’ god and be exclusionary of other gods in the sixth century, a generation before the rise of Islam when South Arabia came under the sway of a militant Abyssinian Christianity. This late incarnation of Rahman-an in Christianized form threatened the status and sanctuary of the Ka'ba, which was most likely why Suhayl b. ‘Amr rejected the use of Rahman so vehemently in the oath-making of Hudaybiyya.

It is salient to point out that, based on an approximate chronological dating of the Quranic suras, theonyms in the Islamic scripture seem to have evolved in three phases. In the earliest phase, the Quran uses rabb, shifting to al-Rahman, and finally culminating in an almost exclusive use of Allah in the later suras.⁸ Rabb simply meant ‘Lord’ and was used for immanent betylic divinities. Its use in the earliest parts of the Quran also corresponds to a monolatric and immanentist usage. By contrast, al-Rahman was clearly associated with Moses in the Quran and the rejection of image-worship, which appears in later Meccan verses. Eventually, however, Allah became the universal theonym, subsuming both Rabb and al-Rahman, in the service of an Abrahamic and fully biblicized monotheism that took shape in Medina.⁹

Al-Rahman also appears briefly in pre-Islamic poetry, where at times it is used interchangeably with Allah.¹⁰ However, this usage is infrequent and might simply reflect a later Islamicization of these orally transmitted sources. Moreover, the pre-Islamic poetic uses of al-Rahman does not use the theonym in a transcendetalized manner, that is, declaring all other gods as false. Even in several South Arabian inscriptions, discussed below, Rahman is equated to the generic name for god (īlān/īlahān) but without implying the Mosaic distinction. Overall, inscriptional evidence is more significant for this study than poetry for two key reasons: inscriptions are more reliably datable than pre-Islamic poetry, and inscriptions were not Islamicized later. Finally, and most importantly, the South Arabian inscriptions demonstrate the use of Rahman in oaths and invocations, which presages its use in the oath-making at Hudaybiyya.

The nature and origin of Rahman before the Quran

We can only guess at an origin for the South Arabian use of the epithet Rahman to indicate a god. As in Quranic Arabic, ‘Rahman’ occurs in the South Arabian¹¹ texts with a definite article, rendering the translation ‘The Merciful’. A number of surrounding cultures and languages used a variation of this word to describe their deities and, as such, its use is not novel. The Quranic al-Rahman can be possibly linguistically traced to the South Arabian Rahman-an, though it cannot be definitively proven that South Arabia was the only or direct source for this theonym since many of the surrounding cultures used the same root to name and praise divinities. For instance, the use of the root r-ḥ-m used for a divinity can be found in Palmyra and in both the Babylonian and the Jerusalem Talmud, along with the Targums. Similarly, it can be found in Syriac and even in the Hebrew Bible, albeit in the different form raḥūm.¹² While we might assume that the name Rahman is borrowed from the surrounding Judeo-Aramaic cultural milieu, it was not, at least initially, used to designate a transcendental and uncompromising god, which is why there is still some sense in exploring the biblicization of the term. If we are engaging in a tracing of the root itself, it is worth noting that it is also found in Akkadian in the form rêmum and likewise used in contexts of religion.¹³ A good indication for the direct transfer of the Sabaic Rahman-an into Arabic as al-Rahman is that Rahman occurs with the definite article alif-lam in the Arabic, echoing the use of the definite article in Sabaic.¹⁴ Thus, both names have a definite meaning and can be translated as ‘The Merciful’.

In the rest of this article, I will present inscriptional evidence surrounding the use of Rahman-an and introduce the historical and religious context necessary to interpret it. In summary, the theonym Rahman-an only appears in Late Sabaic texts dated from the fourth to sixth centuries ce. Because Himyar began to embrace Judaism around the fourth century, this is also known as the monotheistic period. The process may indeed have begun much earlier when Jews increasingly began emigrating to the area after the fall of the Jewish temple in 70 ce. But it is only in the fourth century and later that the area of Himyar openly began to embrace Judaism, as evidenced in the increasing number of inscriptions that mention the God of Israel and/or Rahman-an. Whereas earlier South Arabian cultures seem to have a pantheon of gods with specific deities affiliated with certain tribes and their centres, Rahman-an was not one such god, and his worship in Jewish Himyar is somewhat exclusionary of other gods. Instead of the names of other gods, the texts that name Rahman-an pair it with descriptions of anonymous gods such as ‘Lord of Heaven’ and ‘Lord of Life and Death’. Temple paganism had gradually declined in South Arabia towards the fourth century, coming to an end by the sixth century.¹⁵ Yet, as we shall see below, it is an overreach to perceive the period as fully monotheistic. Instead, we might think of it as monolatric or, at best, henotheist, at least until the arrival of Christianity towards the end of the period. The monolatry of Jewish Himyar under the reign of Joseph dhu Nuwas (517 ce to circa 525 ce) is different from the monotheism embraced under Abyssinian Christian rule (circa 525 ce to maybe 570 ce) in that we do see oaths in inscriptions favouring a certain deity, but not a full cessation of the use of names that might indicate other deities. In these early texts, Rahman-an becomes aligned to the biblical deity but, at the same time, retains a limited degree of translatability and can be interchanged with or invoked at the same time as other epithets. Rahman has not been fully transcendentalized as he would be in classical Islam.

The early ‘monolatrous’ era of Rahman-an

It is important to note that out of the approximately 58 extant instances of the theonym Rahman-an in Late Sabaic inscriptions, none of them is in a text or context that we could clearly label ‘pagan’ or polytheist. The inscriptions do not indicate a variety of gods as is the case in earlier Sabaic texts. Those earlier inscriptions indicate small pantheons centred around major gods which were different, depending on which kingdom or main city they were affiliated with.¹⁶ These pantheons and their temples made way for a monolatric religion around the fourth century ce with the character of Rahman-an at its centre.¹⁷ This change seems to have happened quite abruptly and was likely not to have been the result of a gradual change.¹⁸ In prior centuries, the previously independent kingdoms of Saba’ (280 ce), Qataban (200 ce), and Hadramawt (300 ce) were superseded by Himyar, which at the time spanned the entirety of the southwestern part of the Arabian Peninsula. By the time Judaism made its way to the area, Himyar was already dominant in terms of language, culture, and religion.

There are a number of inscriptions in which Rahman-an is not the only deity mentioned, which indicates that there was some acceptance of the existence of other gods or that Rahman-an had several expressions that could be invoked publicly. Some of these cases include texts that mention Rahman-an along with the god of the Jews. These texts might have been written by an immigrant Jewish population wishing to honour a local god as well as their own god. At a minimum, these inscriptions—in which the use of Rahman-an occurs in oaths and praises—demonstrate that biblical religion and oaths were beginning to be accepted by the ruler and the ruling class, though not exclusively. Some inscriptions mention Rahman-an alone and, as such, do not have an obvious Jewish connection, but they do seem to perceive Rahman-an as the primary god, which has led some earlier scholarship to assume that there was a native monotheism with Rahman-an at its centre called Rahmanism.¹⁹ However, the current consensus is that Rahman(-ism) was not a monotheist religion that arose independently in the Arabian Peninsula; rather, it was a monolatric or henotheistic religious current affiliated with the ruling class of Himyarites and with a rejection of polytheism from the public sphere (that is, in the public inscriptions that we are able to read today).²⁰ In the following section, one group of inscriptions will be discussed, namely those that mention only Rahman-an in terms of a divine figure. (Inscriptions that mention Rahman-an in various Jewish or Judaizing contexts are discussed in the section ‘Rahman-an in Jewish or Judaizing texts’.)

Rahman-an alone as a local divinity?

Quite a few of the inscriptions that mention only Rahman-an and do not have clearly identifiable Jewish traits survive in incomplete form, with the stone broken and parts missing. Others are simply short and thus only contain a brief thanks to Rahman-an. It is worth considering that the incomplete texts could have contained language that would clearly mark them as Judaizing, but if that is the case, we have lost those clues. One example is the broken and short inscription ATM 425,²¹ in which Rahman-an is mentioned twice in five lines that lack the first and last parts of each line, and which may have been broken into pieces to be repurposed in later buildings. The inscription asks Rahman-an to hear, presumably, a question regarding the protection of wives and daughters, and Rahman-an is invoked again later to (perhaps) bring prosperity.

Another example of an inscription that only mentions Rahman-an but from which we cannot tell much more is the prayer text CIH 539, which talks about forgiveness of sins and protection from illness. It also contains the expression in line four w-mrḍym l-s¹m rḥmnn ḏ-Klʿn, which could be translated as ‘… and in the name of Rahman-an of Klʻn …’. It has been argued that ḏ-Klʿn refers to a place, which, if correct, indicates that Rahman-an could be perceived as the deity of a certain place.²² However, the reading is difficult and it would be the only case of Rahman-an being connected to a geographical location.

Rahman as a Jewish or Judaizing divinity

As is perhaps clear by now, the inscriptions that we might categorize as monolatric, as opposed to Jewish, are very difficult to separate from one another. It is also difficult to call them, in good conscience, ‘Jewish’. This is partly because the specifics and ritual of late antique South Arabian Judaism and Jewishness are unclear. Furthermore, many of the texts that we do have can only be categorized as ‘Jewish-like’ or ‘Judaizing’ due to a few signs in the texts. It has been suggested that there are two kinds of Judaism present here, namely ‘Judaism and a “bare” form of monotheism, the religion practiced by the rulers in their inscriptions.’²³ I suggest that there is another element at work: the Judaism which came in the form of the diaspora, that is, the Judaism of immigrants escaping difficult conditions in Roman-ruled Palestine.²⁴ Southern Arabia was already a known trading post, and there is evidence of contact between the area and Palestine.²⁵ These immigrant groups produced texts and inscriptions mentioning their rituals and places of worship and using Hebrew or Aramaic loanwords in their Sabaic texts. At times, bilingual texts²⁶ were produced or Hebrew vignettes were added to Sabaic texts containing religious language.²⁷ In short, the process at work was not one of conversion but of gradual transcendentalization or, more precisely, Judaicization. However, this Judaicization is not quite a fully transcendentalist religion, but rather it is on the verge of it. As Jewish groups moved to South Arabia, other communities in the region—and even the ruling class—adapted this Judaism, but in a more open-ended and less orthodox manner, that is, we find Jewish-like expressions in royal and building texts but not the same kind of specific references to rituals which occur in clearly Jewish texts.

In the next section, I will summarize the texts that seem most Jewish and clearly monotheistic in nature. We can divide the texts in roughly three groups: (1) texts that mention Rahman-an in a context that appears like an immigrant or conversion Judaism, (2) texts that include Rahman-an but seem like a local Judaicizing genre, and (3) the texts from the time of the openly Jewish king Joseph dhu Nuwas. The texts from the rule of dhu Nuwas are certainly part of the second group as well, but they are different nonetheless as they are the largest, longest, and the last of the Judaicizing inscriptions. They are thus treated as their own category.

The inscriptions in the first group contain verbiage and expressions that seem clearly Jewish, even keeping in mind that the Judaism of Late Antiquity and of this region was itself a changing tradition. There are a significant number of Jewish texts in Late Sabaic, and only some of these actually use the epithet Rahman-an. We can infer that the people behind those inscriptions were Jewish because they mention synagogues and ritual baths.²⁸ In some cases, they use expressions like ‘Lord of Heaven’, occasionally equated with Rahman-an, but it is important to note that Rahman-an is used very rarely. This indicates that at this stage, Rahman-an was only beginning to be equated with biblical monotheism. Here, I will focus only on the texts that are both overtly Jewish and contain the epithet Rahman-an, since it is not a question of if there were Jews in Himyar, but rather, to what degree they used the ‘indigenous’ epithet Rahman-an and whether Rahman-an was considered to be the same as or different from the Jewish god.

It must be stated upfront that, owing to a lack of historical context, it is difficult to ascertain without any doubt whether an inscription is related to immigrant or conversion Jews. Here I discuss two such inscriptions: one whose reading is contested and one that seems to be written by a local South Arabian who converted to Judaism. RES 4109 is a very short text in which a man begins his inscription thus: ‘May Rahman-an hear (the prayer) of Hmdm Ks¹dyn’. Ks¹dyn is noteworthy since it has been argued that the name might mean that the man invoking Rahman-an is a Chaldean, that is, a Jew.²⁹ It has been argued, equally convincingly, however, that it could simply be a local place name.³⁰

The second inscription, Ibrahim al-Hudayd 1, is slightly longer and gives more context. Dating to around 470 ce, it is a construction text invoking protection for a newly built house and the family meant to live within it.

w-b-rdʾ w-ḫyl mrʾ-hmw ʾln bʿl s¹myn w-ʾrḍn w-b-rdʾ s²ʿb-hmw ys³rʾl w-b-rdʾ mrʾ-hmw s²rḥ(b)ʾl mlk s¹bʾ w-ḏ-rydn w-ḥḍrmwt w-l-(ḫ)mr-hmw b-hw rḥmnn ḥywm ks³ḥ[m]

With the help and the power of their Lord God (they built the house), master of Heaven and Earth, and by the help of their tribe, Israel, and by the help of their lord, Sarahb'il, king of Saba’, dhu Raydan and Hadramawt, and may Rahman-an give it (i.e. the house) and unblemished life.

Two things are notable in this text. One is that we clearly have a text commissioned by a family who considers themselves part of the tribe of Israel while also honouring King Sarahb'il and living in his kingdom. Moreover, based on the Himyarite-sounding names of these individuals, they are likely to have been not itinerant Jews but converts to Judaism.³¹ There is a clear translation of divine names taking place in this inscription in which two divinities are being invoked, one which belongs to the ‘tribe of Israel’, namely the Lord God, and one which belongs to the ‘King of Saba’, namely Rahman-an. Although they are clearly perceived as different deities, their invocations are not mutually exclusive. Unlike a fully biblicized oath, these local Jews had no problems invoking more than one deity. Once the Mosaic distinction becomes pervasive with the spread of biblical religion, translatability ceases since ‘false gods cannot be translated’.³² Nevertheless, even though the authors of Ibrahim al-Hudayd 1 clearly perceived themselves as Jewish, they had no issue communicating with other gods and might have perceived their own god as analogous to the local god, Rahman-an.

Secondly, it is worth noting that Rahman-an is invoked to protect the physical building of the house after the Lord of Heaven and Earth was invoked, having helped the family complete the house, so that they can prosper in it. This pattern re-emerges later in the texts from the time of Joseph dhu Nuwas. A transcendent deity is invoked to protect the lives of inhabitants or soldiers, whereas Rahman-an is invoked to protect the physical structure of a house or an inscription.

Another good example of a text that contains Rahman-an and expressions that are Judaicizing is CIH 543, which is an inscription invoking protection for a newly built house and the family members living within it. The first lines read: ‘Bless, and may (it, that is, the house) be blessed, the name of Raḥmānān, who is in heaven, and Israel and their god, the lord of the Jews, who assisted their servant S²hrm’.

Arguably this text invokes two different deities, namely Rahman-an and ‘The Lord of the Jews’.³³ Gajda even suggests that the text is an example of a sort of religious syncretism in which Rahman-an is the main god of Himyar, but he is not exclusively a God of the Jews. Robin specifically notes the expression ‘their God’ ʾlh-hmw which indicates a difference between Rahman-an and the Lord of the Jews. Perhaps the house-owner, S²hrm, made sure to call for protection from both deities in order to be fully covered. As we shall see in the texts from the rule of Joseph dhu Nuwas, there is a sense of Rahman-an having had more of an immanent character and having been a protector of physical structures like a house or a large inscription.

Similarly, there exists another short construction text, Ry 403, which thanks the ‘Lord Rahman-an’ because the house was built and completed. In the final line it ends with: ‘… and may Rahman-an give life to them and their house. Amen’. This little ‘amen’ is the only Judaicizing character to it, and it occurs in just seven texts in total. Some of these are construction texts as well (for instance, ZM5+8+10), but one is from the same area and time period as the texts from the rule of Joseph dhu Nuwas (Ry 513). The expression ‘Amen’ does not show up in texts earlier than Late Sabaic, and so we might assume that it occurs through the influence of Jewish groups in Himyar.

Finally, in the building inscription, Gar antichità 9d, we have Rahman-an with the expression ‘Lord of Heaven’, which occurs in a number of monolatric texts from this period. Sometimes Rahman-an is equated to the Lord of Heaven, but at other times the two expressions seem to denote different aspects of a god. Sometimes we find the expression ‘Lord of Heaven and Earth’ and in the Yehuda Inscription (Gar Bayt al-Aswal 1), this Lord, who is also the lord of life and death, is described as creating the earth. Later in the Gar antichità 9d inscription, the house owner thanks the ‘grace (zkt) of Rahman-an’. Again, it is in the smallest words that we find Judaizing elements as the word zkt is novel to Sabaic and seems like a loan from Hebrew or Aramaic.

Inscriptions from the reign of the Jewish king Joseph dhu Nuwas

Although the effect of Joseph dhu Nuwas's rule was momentous in the history of the southern part of the Arabian Peninsula, we have few clues as to his background or how he came to power in Ḥimyar. Similarly, we know little of how he met his end after the successful Christian Abyssinian conquest (presumably he had been killed by the end of the war).³⁴ Notably, no Jewish sources mention the rise and fall of the Jewish Ḥimyarite kingdom, even though they do mention Jews in southern Arabia.³⁵ There are, however, later Christian and Muslim sources that recount dhu Nuwas's reign. Unsurprisingly, the Christian sources paint Joseph dhu Nuwas as an evil villain and painstakingly outline how he massacred the Christian population mainly centred in the northern city of Najran.³⁶ Similarly, Ethiopic Christian sources perpetuate an enduring memory of dhu Nuwas as the mortal enemy of the saintly king Kaleb who led the war against dhu Nuwas.³⁷

Conversely, the Muslim memory of these events is less antagonistic towards dhu Nuwas. For instance, Ibn Ishaq (704–767 ce) recounts in his famous The Life of Muhammad (sirat al-nabi, circa 750), which became the definitive biographical account of the Prophet, that dhu Nuwas came to power by tricking and overthrowing an unjust usurper, a cruel sodomite.³⁸ Further, al-Tabari (839–923 ce) notes that dhu Nuwas battled Abyssinian missionaries who were converting people to Christianity in Najran. According to al-Tabari, dhu Nuwas gave the inhabitants an ultimatum: revert to Judaism or be killed.³⁹ The citizens of Najran chose the latter and thus, one could argue, were in part responsible for their own demise. Accordingly, this massacre incurred the wrath of King Kaleb, who launched an attack on Himyar and prevailed after a hard-fought campaign.

During the Abyssinian campaign, one of dhu Nuwas's generals wrote inscriptions to commemorate the events, all of which invoke Rahman-an. This group of texts comprises three large and long descriptions written over a couple of months during the war in 524 or 525 ce. They are still located in situ around an area called Hima in southwestern Saudi Arabia. Evidently, the general who commissioned them was stationed there with some of the army while Joseph dhu Nuwas was off with another part of the army, trying to defend the coastline and prevent the Abyssinians from landing on the shores of Himyar. A few smaller inscriptions are scattered around this area and are thought to be from the same period because they either mention the general or other names found in the larger inscriptions.

In these texts, the meaning of Rahman-an still seems ambiguous, and one could argue that there is still no clear indication of whether the texts use Rahman-an as just another name for the Jewish god or whether Rahman-an is perceived as a local, more immanent entity working at the same time as the transcendent Lord of the Jews. For instance, in the formulaic invocations below (Ja 1028, Ry 507, and Ry 507), it seems as if the writer is pleading with the Lord of the Heaven and Earth for the protection of the army but then turns to Rahman-an to ensure that the inscription itself is protected—that is, as if the local deity in charge of the place of the inscription was Rahman-an.

Ja 1028

l-ybrkn ʾln ḏ-l-hw s¹myn w-ʾrḍn mlkn Yws¹f ʾs¹ʾr Yṯʾr mlk kl ʾs²ʿbn w-l-ybrkn ʾqwln

May the God, to whom belong the heavens and the earth, bless the king Joseph ʾAs¹ʾar Yaṯʾar, king of all the tribes, and may he bless the generals…

w-s¹t mʾtm w-k-b-ḫfrt s¹myn w-ʾrḍn w-ʾʾḏn ʾs¹dn ḏn ms¹ndn bn kl ḫs¹s¹m w-mḫdʿm w-rḥmnn ʿlyn b n kl mḫdʿm ḏn ms¹[ndn] wtf w-s¹ṭr w-qdm ʿly s¹m rḥmnn wtf tmmm ḏ-hḍyt rb-hd b-mḥmd

For the protection of the heavens and the earth and of the strength of the men was this inscription against those who would harm and degrade. May Raḥmānān, the Highest, protect it against all those who would degrade. This inscription was placed, written, executed in the name of Raḥmānān. Tmm of Ḥḍyt placed. The Lord of Jews. By the Highly Praised.

The following two inscriptions are from the same vicinity and time as the abovementioned inscription (Ja 1028). One of these—the inscription Ry 507—equally invokes the God of Heaven and Earth, and while it does not actually mention Rahman-an, it does describe how the inscription, like a talisman, was put up in order to invoke protection from the creator of Heaven and Earth.

Ry 507

l-ybrkn ʾlhn ḏ-l-hw s¹[myn w-ʾrḍn mlkn Ys¹f ʾs¹ʾr Yṯʾr mlk kl] ʾs²ʿbn [w-l-ybrk]n ʾql-hmw w-mrʾs¹-hmw [w-]ʿrbn-hmw

May God to whom belong heavens and earth bless king Joseph ʾAs¹ʾar Yaṯʾar, king of all the tribes, and might He bless their generals and their leaders and their nomads…

w-s¹ṯ mʾtm w-b-ḫfrt [s¹]myn w-ʾrḍn w-ʾḏn ʾs¹dn ḏn ms³ndn bn kl ḫs¹s³m s¹lm ʿly mlkt Ḥmyrm

For the protection of the creator of heavens and earth, and the hosts, is this inscription, from any damage. Peace to the king of Ḥimyar.

The final inscription invokes the God to whom belongs the heavens and the earth and then it invokes Rahman-an to protect the inscription from harm. However, it also implores Rahman-an to spread mercy over the world and calls him ‘the Lord’.

Ry 508

w-s¹ṯ mʾtm w-ʾʾlhn ḏ-l-hw s¹myn w-ʾrḍn l-yṣrn mlkn Ys¹f b-ʿly kl ʾs²nʾ-hw w-b- ḫfr Rḥmnn (ḏ)n ms¹ndn bn kl ḫs¹s¹{s¹}m w-mḫdʿm w-trḥm ʿly kl ʿlm Rḥmnn rḥm-k mrʾ ʾt

May the God, to whom belong the heavens and the earth, give to the king Joseph the victory over all his enemies. For the protection of Rḥmnn this inscription against those who would harm and degrade. Spread all over the world, Rḥmnn, your mercy. You are the Lord.

It is worth noting that Ry 508 invokes God by using the plural ʾʾlhn, instead of the singular ʾlhn found in Ry 507. A reasonable explanation for this could that it is a calque echoing the plural in the Hebrew 'elohīm.⁴⁰

The smaller inscriptions scattered near these three inscriptions also include Rahman-an. First, the small epigraph Ry 515 mentions the same general who commissioned the large inscriptions and is thus contemporary with them. Of note is the fifth line which reads: rb-hwd b-rḥmnn, ‘Lord of the Jews, with/by Raḥmānān’, as it echoes the phrasing in some of the larger inscriptions. The Lord of the Jews is here coupled with Rahman-an, but whether we should interpret it as the Lord being invoked with Rahman-an or through him is unclear.

Ry 513 must also be mentioned as it is written by the person who wrote some of the large inscriptions, and it uses expressions similar to theirs.

tmm mqtw lḥyʿt yrḫm ḏ-gdnm w-trḥm ʿly ʾbny mlkm ḏ-gdnm rḥmnn w-ʾmn

Tmm, officer of Lḥyʿt Yrḫm ḏ-Gdnm. May He be merciful regarding the sons of Mlkm ḏ-Gdnm. Raḥmānān. Amen.⁴¹

Rahman-an, the epithet, is used in this invocation of mercy, and the final word of the inscription is the Judaicizing expression ‘amen’. Rahman-an is invoked when it comes to the protection of physical structures like houses and monumental inscriptions but also when one asks for the protection of people. This is reminiscent of the invocation of Rahman-an in Ibrahim al-Hudayd 1, the Jewish/convert building inscription, discussed in an earlier section, where the Master of Heaven and Earth is thanked for making it possible for the family to build the house, while Rahman-an is invoked to protect the physical structure of the building from any blemish.

Rahman-an under Christian rule

The very last years of the Himyarite kingdom saw the ascendance of Christian kings to the throne. Most important of these was Abraha, who became king around 532–535 ce⁴² and ruled until at least 558 ce⁴³ and possibly later. A few years before Abraha took the throne, the kingdom of Himyar had been embroiled in a war with Abyssinia and, upon losing, the previously Jewish kingdom became a Christian vassal state. First, a man called Sumyafa‛ Ashwa‛ was placed on the throne as a vassal king, but he was soon replaced—potentially during a coup—by Abraha. Abraha's origins are obscure, but it is likely that he was a soldier who was part of the earlier war with the Himyarites. Procopius writes that Abraha was a former slave of a Byzantian trader,⁴⁴ but it seems more likely that he was a soldier already in Himyar, as he would have needed the backing of his military comrades in order to successfully oust a king appointed by Abyssinia. Abraha ruled Himyar for an extended period of time, during which he conducted several incursions into central Arabia, fortified his kingdom, and built an important sanctuary in Sana'a. After Abraha's rule, his sons seem to have taken over, but the kingdom quickly came under Sassanian rule and remained so until the coming of Islam. Effectively, Abraha was the last great independent king of Himyar.

It is important to note that the Abyssinians continued dhu Nuwas's practice of using Rahman-an as the name of the biblical deity. This is the case even before Abraha. A few texts survive from the reign of his predecessor, Sumyafa‛ Ashwa‛, and in them, we find Rahman-an used for the Christian god, the Father. The most complete text is Ist 7608 bis, which, although broken in some areas, allows us to read a recounting of how Sumyafa‛ Ashwa‛ came to power as a negus, an Abyssinian vassal king, meant to keep order in Himyar. The inscription is rife with Christian vocabulary and Ethiopic loan words, and in the final lines we read the expression: s¹m Rḥmnn w-bn-hw krs³ts³ ġlbn (‘In (the) name of Rahman-an and his son, Christ the victorious’). For the first time, Rahman-an is unequivocally a biblicized and transcendental God, that is, there is no notion of an immanent Rahman-an deity alongside a transcendent god of Israel. Here, Rahman-an simply is God, the Father of Christ. The era of the translatability of divine names had apparently come to an end.

The practice of using Rahman-an as the epithet for the biblical god is continued in the texts of Abraha, as Rahman-an is used as the name for the Christian god of the Trinity. We have many more texts from the rule of Abraha than from Sumyafa‛ Ashwa‛ and some of them, such as the wonderfully well-preserved two-and-a-half metre tall stela CIH 541, are long narratives describing Abraha's military excursions and power moves in some detail. CIH 541 commemorates the restoration of the great dam at Marib by Abraha in circa 543 ce. The stela has 136 lines, but mentions Rahman-an only twice, the first at the very beginning of the text when Abraha takes the word and writes:

b-ḫyl w-[r]dʾ w-rḥmt rḥmnn w-ms¹ḥ-hw w-rḥ [q]ds¹ s¹ṭrw ḏn ms³ndn ʾn ʾbrh

By the power, the aid, and the mercy of Rahman-an, and his Messiah, and the Holy Spirit, I, Abraha, wrote this inscription.

Rahman-an appears again in line 93 when Abraha describes a plague which subsides ‘thanks to Rahman-an’. Ry 506, another important text from Abraha's reign dating to 552 ce, is also important for our purposes for two reasons. First, the continued use of Rahman-an proves the point that the theonym was synonymous with the Christian god and, second, its description of military campaigns stretching far into the centre of the Arabian Peninsula suggests that Abraha, or the memory of him, was what Meccans associated with the epithet al-Rahman.

Now let us turn to Rahman-an which, as we expect, occurs in the initial line and again in the final line of an inscription commemorating military excursions against Arab tribes. The first line reads b-ḫyl Rḥmnn w-ms¹ḥ-hw mlkn ʾbrh z-b-Ymn …s¹ṭrw ḏn s¹ṭrn (‘By the power of Rahman-an and his Messiah, the king Abraha, who is in Yemen…wrote this inscription’). In this particular text, we do not get the full trinity, as the holy spirit is not invoked.

The final line of the inscriptions reads w-qflw bn Ḥl[b]n [b-]ḫyl Rḥmnn (‘…and he returned from Hlbn by the power of Rahman-an’), highlighting that Abraha returns victoriously by the power of God. In these inscriptions it is clear that Rahman-an is simply the word used for the Christian god and he has lost any distinction he might have had as a local god occurring either alone or alongside the Jewish god. This Christian/biblicized Rahman-an is the one with whom the Prophet Muhammad's contemporaries were familiar. One could speculate that had Rahman-an retained his veneer of ambiguity—that is, the possibility of translatability—perhaps the Meccans would have been less resistant to the use of his epithet in oath-making. However, the antagonistic feelings the epithet Rahman invokes were closely related to how he was portrayed during Abraha's reign. This brings us to another significant insight from Ry 506.

The second important point related to the inscription Ry 506 depends on a close reading of this text. In an article from 2012, Christian Robin has re-read and reinterpreted this text,⁴⁵ which was previously only known from inaccurate readings from the first expeditions to the area of Murayghan in southwestern Saudi Arabia, where the inscription was found in situ on the mountainside. The earlier misreading of the text suggested that it was written after Abraha's military failure in central Arabia, which led to the conclusion that, since the text was dated 552 ce, this was probably Abraha's last attempt to rule central Arabia. In his new and correct reading of the actual inscription, however, Robin suggests that the text does in fact portray a successful excursion to the tribes in central Arabia and the appointment of an Abraha-friendly governor in the area. The text also describes this excursion as being the fourth one, indicating that Abraha was a frequent visitor to the area. More importantly, Robin argues that this was not the end of Abraha's rule over these areas but rather the beginning, and that his influence certainly extended well beyond 552 ce—perhaps even to sometime after the last dated inscription from his reign, which was 558 ce, that is, only a dozen or so years before the birth of Muhammad.

Abraha is well-known in the Islamic tradition because his memory was preserved by the early Quranic exegetes as the military antagonist of the Meccans in the short 105th sura, the Sura of the Elephant. In this chapter of the Quran, an unnamed group, simply called ‘lords (or companions) of the elephant’, tries to attack Mecca, apparently around the time of the birth of the Prophet Muhammad, but the attack is unsuccessful. The attacker and his elephants are destroyed miraculously—crushed like ‘eaten straw’—by birds throwing pellets of clay.

Have you not considered, [O Muhammad], how your Lord dealt with the companions of the elephant?

Did He not make their plan into misguidance?

And He sent against them birds in flocks,

Striking them with stones of hard clay,

And He made them like eaten straw.⁴⁶

Traditionally, Muslim commentators have described the leader of the elephant army in the sura as Abraha. In historical terms, the latter may indeed have been antagonizing areas close to Mecca⁴⁷ and Medina.⁴⁸ Moreover, with Robin's revised chronology, Abraha's failed attempt at taking Mecca would be some time after 558 ce, which puts it very close to the assumed date of birth of the Prophet Muhammad in 570 ce.⁴⁹

Of special note is the fact that Abraha had launched the construction of the cathedral of Sana'a. According to the historian al-Tabari, this Christian site was meant to compete with the other main religious pilgrimage centre of the area, the Ka'ba in Mecca.⁵⁰ This cathedral was known in the Arabic tradition as al-Qalis (from the Greek ἐκκλησία, meaning ecclesia or church), and its foundations can still be seen today in Sana'a.⁵¹ By all accounts, it was a magnificent building of stone and marble, sumptuously decorated with Byzantine-style mosaics, perhaps even resembling the Church of the Nativity in Bethlehem. According to Arab historians, Abraha also constructed sleeping quarters for pilgrims, and eventually the Qalis did become a real competitor to the Ka'ba and was visited by members of Arabian tribes—much to the chagrin of the Meccans.⁵²

Robin argues—convincingly—that Abraha was so tied up in internal political turmoil that he could not have begun construction on this large edifice until rather late in his reign. One would assume that inscriptions from his time would mention the prestigious and politically important cathedral or that, if they do not, building had not yet begun.⁵³ However, it is likely that the cathedral was finished by the time Abraha embarked on his ill-fated march against Mecca, when he wanted to destroy the Ka'ba and move the sphere of religious influence further south.⁵⁴ According to later Muslim tradition, the Meccan tribe of Quraysh, in response to the prolonged threat of Abraha, and especially after the events described in the Sura of the Elephant, organized themselves in the cultic alliance known as Hums. It was an alliance based on shared ritual duties and taboos among the various tribes during the annual Hajj pilgrimage to the Ka'ba.⁵⁵ If this organization of power was indeed a direct response to Abraha's final campaign, it follows that the memory of Christian invasion under the banner of Rahman-an was strong in Hudaybiyya's time. Indeed, in the Muslim tradition, Muhammad's grandfather Abd al-Muttalib—leader of the Quraysh in 570—reportedly prayed in front of the door of Ka'ba, grasping its door handle and asking the lord of the shrine for protection against the cross-bearing invaders.

Oh Lord! As a man protects his house; protect your sanctuary!

So that not their cross⁵⁶ and their power vanquish your glory.⁵⁷

The creation of the Hums further increased the importance of the Ka'ba for Arabia Deserta in the years around Muhammad's birth, especially enhancing its political significance. At Hudaybiyya, Abraha's legacy of crusading in the name of Rahman-an is likely to have served as a problematic memory when Muhammad tried to sign a pact with the pagans at Mecca.⁵⁸ Not only was Muhammad's use of the name clearly monotheistic and transcendentalized, it also evoked memories of antagonism and warfare in the name of the biblical god against the holiest pilgrimage site in Central Arabia. By this time, Rahman-an, or al-Rahman, had transformed into the name of the biblical deity that did not allow the use of its name in pagan oaths of peace.

Conclusion

In the three centuries preceding Islam, the theonym Rahman appeared in southern Arabia where it gradually evolved into the name of the transcendentalized and uncompromising biblical god—first the God of Israel and then the Father in the Christian Trinity. From the earliest attestations, Rahman-an designated a main god, but it is unclear whether he was an independent local god or an alternate designation of the emerging Jewish god in ancient Yemen, that is, Himyar. The texts that appear to be clearly Jewish (and not Judaicizing) hardly use the epithet, indicating that it was the native Himyarites who equated Rahman-an with the Jewish god, whereas immigrant Jews conceived of him as a local god. The Jewish and Judaicizing inscriptions culminate in the three major public inscriptions from the reign of Joseph dhu Nuwas who was the last ‘Jewish’ king of the area. Dhu Nuwas attacked local Christian groups and eventually lost the ensuing retaliatory war against Christian Abyssinia around the year 525 ce. After this event, Himyar became Christian and so did Rahman-an. Any ambiguity regarding his monotheistic nature ceased, and he became the one exclusive God of the Bible.

It is this version of Rahman-an who is invoked in large monumental Christian inscriptions wherein Abraha praises his god's glory in Himyar. I argue that this is the version which elicited such a strong reaction in the mind of the Meccan emissary who was sent to strike the peace with Muhammad at Hudaybiyya.⁵⁹ For Abraha was the military leader who, in the year of Muhammad's birth, had attempted to destroy the Ka'ba’ and to funnel pilgrim traffic and revenue to his own sanctuary of (Christian) Rahman-an.

As the framework article in this special issue explores, the biblically monotheist model of making treaties with pagans normally only allowed for a conversion. Biblical monotheism considers its god to be untranslatable and rejects earlier models, which allowed for oath-making across pagan religions. Although monolatrous Judaism was introduced into Himyar abruptly, in the case of oath-making, the Himyarites were far from exclusionary in their invocations. In fact, authors of inscriptions unproblematically used language that indicated various deities with various roles before the far more established monotheist Abyssinian Christianity took over the ruling class and its religious vocabulary. Although the untranslatability of the biblical god is a hallmark of biblical monotheism, centuries passed before this feature fully took hold, at least in Himyar. The culmination of this gradual development was the Peace at Hudaybiyya. This event constitutes the beginning of the genealogy of the sulh-i kull of the Mughal empire which had to accommodate oath-making with non-monotheist partners. Additionally, it marks the end of the genealogy of South Arabian Rahman-an, which culminated with the biblical kingship of Abraha and the subsequent adoption of Rahman by another biblical king: the Prophet Muhammad.

Acknowledgements

I am grateful to A. Azfar Moin for countless thoughtful comments and feedback; to Blake Pye for the copy-edits; and to the second anonymous reviewer of Modern Asian Studies for their highly useful feedback which undoubtedly increased the quality of this article.

Competing interests

None.

¹ Michael Fishbein (trans.), The Victory of Islam: Muhammad at Medina A.D. 626–630/A.H. 5–8, 40 vols, Vol. 8, The History of Al-Tabari (Albany, NY: SUNY Press, 1997).

² /bismaka allāhumma/.

³ Al-Ṭabari, Tafsih, 1546, in the English translation by Fishbein, The Victory of Islam, Vol. 8, pp. 85–86. Al-Azmeh notes that Suḥayl was ‘clearly perceiving it could avert complications brought about by mention of individual divinities’. Aziz Al-Azmeh, The Emergence of Islam in Late Antiquity: Allāh and His People (New York: Cambridge University Press, 2014), p. 230.

⁴ Jan Assmann, ‘The Mosaic Distinction: Israel, Egypt, and the Invention of Paganism’, Representations 56, Special Issue: The New Erudition (Autumn 1996), p. 50.

⁵ See Strathern for an outline of the attributes of transcendentalist religion and how they might present themselves. Alan Strathern, Unearthly Powers: Religious and Political Change in World History (Cambridge: Cambridge University Press, 2019), Chapter 1.

⁶ Al-Azmeh, Emergence, p. 40.

⁷ The theonyms īlān/īlahān were used as well throughout the inscriptions of the fourth to sixth centuries. They both translate as ‘the God’.

⁸ Gerhard Böwering, ‘Chronology and the Qur’ān’, in Encyclopaedia of the Qur’ān (Leiden: Brill, 2001), p. 329.

⁹ The initial reluctance to use the theonym Allah might have been due to its polytheistic origins. Ibid.

¹⁰ Nicolai Sinai, Rain-Giver, Bone-Breaker, Score-Settler: Allāh in Pre-Islamic Poetry, American Oriental Series 15 (New Haven, CT: American Oriental Society, 2019), p. 69.

¹¹ These texts are written in the Semitic language of Sabaic which is a subdivision of Old South Arabian (OSA). Sabaic itself is often divided into Early, Middle, and Late Sabaic. Late Sabaic corresponds roughly to the time period of the fourth to sixth centuries ce. Although the name could indicate otherwise, OSA is not a direct precursor to Arabic. The language was written in the Old South Arabic script which may have been the precursor of Ethiopic script. See R. Hasselbach, ‘Old South Arabian’, in Languages from the World of the Bible, (ed.) Holger Gzella (Berlin: De Gruyter, 2011), pp. 160–193. For the classification of the Semitic languages, see John Huehnergard and Aaron D. Rubin, ‘Phyla and Waves: Models of Classification’, in The Semitic Languages: An International Handbook on Their Structure, Their History and Their Investigation, (ed.) Stefan Weninger (Berlin: De Gruyter, 2011).

¹² J. Payne Smith, A Compendious Syriac Dictionary, Ancient Language Resources (Eugene, OR: Wipf and Stock Publishers, 1999), p. 537, and Francis F. Brown, S. R. Driver, Charles A. Briggs and Wilhelm Gesenius, The Brown-Driver-Briggs Hebrew and English lexicon: based on the lexicon of William Gesenius, as translated by Edward Robinson, and edited with constant reference to the thesaurus of Gesenius as completed by E. Rödiger, and with authorized use of the German editions of Gesenius' Handwörterbuch über das Alte Testament (Peabody, MA: Hendrickson Publishers, 1996).

¹³ Chicago Assyrian Dictionary, Vol. R, p. 575ff. The suggestion by Beeston that the root r-ḥ-m is later in Sabaic and modelled on the Palmyrene and replaces the ‘genuine Sab. root… rḫm’ is worth considering, but perhaps not fully convincing, since r-ḫ-m appears to occur only in personal names. Unless there is compelling evidence to the contrary, we should consider the roots separate. See Beeston, ‘Foreign Loanwords in Sabaic’, p. 42. It is remarkable that the use of the root r-ḥ-m, which is then fully functional, does not occur until well into the monotheistic period, but the personal name Yrḫm using the root r-ḫ-m coexists well into the monotheistic period and in texts in which Raḥmānān occurs (for instance, in Ja 1028, one of the texts describing Joseph dhu Nuwas's military success).

¹⁴ The suffix /-ān/.

¹⁵ To summarize the religious changes in the area: the pagan temple religions of the pre-Himyarite South Arabian kingdoms (until the fourth century ce), the Jewish monolatry of the Himyarite kingdom (fourth century to 525 ce), the monotheistic Christianity of the Abyssinian-inspired kingdoms (525 ce to circa 570 ce), the interim period of Sassanian-controlled Himyar, and, finally, the Islamic period which began in 630 ce with the arrival of Ali in the area.

¹⁶ For instance, the kingdom of Saba centred on the main god Almaqah, but inscriptions honour other gods such as ʻAthar, Haubas, dhat-Himyam, and dhat-Badan as well. For more information about South Arabian pantheons before the kingdom of Himyar, the following is a good place to start: Christian J. Robin, ‘Before Himyar: Epigraphic Evidence’, in Arabs and Empires before Islam, (ed.) Greg Fisher (Oxford: Oxford University Press, 2015), pp. 90–126.

¹⁷ At least, this is what the stone inscriptions from this time indicate. There is now some evidence that remnants of polytheism survived among the people of Himyar who did not belong to the elites. This population did not produce stone inscriptions; rather, they wrote in a hard-to-decipher minuscule on reeds. These reed letters have only been the focus of research in the past 20–30 years. The texts contain everyday interactions, as opposed to the monumental stone inscriptions. Iwona Gajda, ‘Remarks on Monotheism in Ancient South Arabia’, in Islam and Its Past, Jahiliyya, Late Antiquity, and the Qur'an, Oxford Studies in the Abrahamic Religions, (eds) Carol Bakhos and Michael Cook (Oxford: Oxford University Press, 2017).

¹⁸ Christian J. Robin, ‘Ḥimyar, Aksūm, and Arabia Deserta in Late Antiquity. The Epigraphic Evidence’, in Arabs and Empires, (ed.) Fisher, p. 129.

¹⁹ A. F. L. Beeston, ‘The Religions of pre-Islamic Yemen’, in L'Arabie du sud histoire et civilisation. Vol. 1: Le peuple Yemenite et ses racines, (ed.) J. Chelhod (Paris: Maisonneuve and Larose, 1984), pp. 267–269.

²⁰ Robin, ‘Ḥimyar, Aksūm, and Arabia Deserta in Late Antiquity’, p. 129.

²¹ ATM 425 and all of the following inscriptions quoted in this article are available in the online Digital Archive for the Study of pre-Islamic Inscriptions: http://dasi.cnr.it/, [accessed 18 August 2021].

²² Iwona Gajda, ‘Ḥimyar gagné par le monothéisme (IVe-VIe siècle de l’ère chrétienne). Ambitions et ruine d'un royaume de l'Arabie méridionale antique’, PhD thesis, Aix-en-Provence, 1997, p. 157.

²³ Robin, ‘Ḥimyar, Aksūm, and Arabia Deserta in Late Antiquity’, p. 129.

²⁴ Gordon D. Newby, A History of the Jews in Arabia (Columbia, SC: The University of South Carolina Press, 1988).

²⁵ See, for instance, the Hebrew/Aramaic/Sabaic epitaph found near Jerusalem: G. Wilhelm Nebe and Alexander Sima, ‘Die Aramäisch/Hebräisch-Sabäische Grabinschrift Der Lea’, Arabian Archaeology and Epigraphy 15, no. 1 (May 2004), pp. 76–83.

²⁶ Rainer Degen, ‘Noch einmal: Die hebräische Inschrift DJE 23 auf dem Jemen’, in Neue Ephemeris für Semitische Epigraphik, 2 vols (Wiesbaden: Otto Harassowitz, 1974), Vol. 2, pp. 166–168.

²⁷ The Yehuda inscription (Gar Bayt al-Ashwal 1) is an example of this. In the fourth to sixth centuries ce, Hebrew had long ceased to exist as a spoken language yet survived in literature and religious expressions. For a more thorough description of the history of Ancient Hebrew, see Angel Sáenz-Badillos, A History of the Hebrew Language (Cambridge: Cambridge University Press, 1993).

²⁸ The texts corroborate archaeological evidence which indicate the same. See Yosef Tobi, ‘The Jews of Yemen in Light of the Excavation of the Jewish Synagogue in Qanî’, Proceedings of the Seminar for Arabian Studies 43 (2013), pp. 349–356; Avraam Lundin, ‘The Jewish Communities in Yemen during the 4th–6th Centuries (According to Epigraphic Material)’, in Judaeo-Yemenite Studies Proceedings of the Second International Congress, (eds) Ephraim Isaac and Yosef Tobi (Princeton, NJ: Institute of Semitic Studies, 1999), pp. 17–25.

²⁹ François Bron, ‘Palmyréniens et Chaldéens en Arabie du Sud’, Studi Epigrafici e Linguistici 3 (1986), pp. 95–98.

³⁰ Hani Hayajneh, ‘Erneute Behandlung einiger altsüdarabischen Inschriften’, Arabian Archaeology and Epigraphy 13, no. 2 (2002), pp. 193–222.

³¹ Iwona Gajda, ‘Une nouvelle inscription juive de Ẓafār’, Scripta Yemenitica 2004, pp. 197–202. Arguably, the family could have taken Himyarite names in their new homeland, but that hardly makes sense, unless they were also shedding their religion. Further, there are plenty of examples of inscriptions of people with very Jewish-sounding names.

³² Assmann, ‘The Mosaic Distinction’, p. 50.

³³ Gajda, ‘Remarks on Monotheism in Ancient South Arabia’, pp. 253–254; Christian Robin, ‘Quel judaïsme en Arabie?’, in Le judaïsme de l'Arabie antique, Judaïsme ancien et origins du christianisme 3, (ed.) Christian Robin (Turnhout: Brepols, 2015), pp. 297–329.

³⁴ We do have an inscription mentioning the Ḥimyarite king being killed by the Ethiopians, CIH 621 line 8: ‘…when the Abyssinians sent their expedition task force to the land of Ḥimyar, when they killed the king of Ḥimyar and his princes (qayls), Ḥimyarites and Raḥbatites’. Translation from Robin (ed.), Le judaïsme de l'Arabie antique. However, we do not know where and when this would have taken place although the fragmentary Ethiopic inscription RIE 195 might give some clues. This inscription was erected in the city of Marib and recounts the Ethiopian king killing the ‘deposed king of Himyar’. Right after, it describes the burning of the ‘palace of Saba’. See the following for a detailed treatment of the fragmentary text: Jacquelinne Pirenne and Gigar Tesfaye, ‘Les deux inscriptions de Negus Kaleb en arabie du sud’, Journal of Ethiopian Studies 15 (August 1982), pp. 105–122.

³⁵ For instance, Rabbi Akiva reportedly travelled in the peninsula, and wrote of the Jewish communities there; see Newby, Jews in Arabia, p. 39. Although it seems certain that Akiva went to the Arabian Peninsula, whether he went there on a political errand to strengthen the Jewish communities is less certain and it does not confirm the existence of Jews in southern Arabia before Philostorgius's fourth century ce account of the Christian embassy sent to convert the Ḥimyarite king, but who found Jews at the court.

³⁶ The Martyrdom of Arethas recounts these events and exists in several versions. An older Greek version from the sixth century ce exists: see Irfan Shahid, The Martyrs of Najran: New Documents, Subsidia Hagiographica 49 (Bruxelles: Société des Bollandistes, 1971), p. 206, as well as a later Ethiopic/Arabic version which might date from ninth century ce. Similarly, the Syriac Book of the Himyarites recounts the events of dhu Nuwas's reign and casts him in a less than favourable light: Axel Moberg, The Book of the Himyarites: Fragments of a Hitherto Unknown Syriac Work (Lund: C. W. K. Gleerup, 1924).

³⁷ Venerated as St Elesbaan, Kaleb is celebrated in the martyrologies of the Oriental, the Roman Catholic, and the Eastern Orthodox churches (in addition to the Ethiopic Church, of course). According to these traditions, he resigned from the throne after fighting Joseph dhu Nuwas and spent the rest of his life as an ascetic monk.

³⁸ A. Guillaume, The Life of Muhammed: A Translation of Ibn Ishaq's Sirat Rasul Allah (Oxford: Oxford University Press, 1967), pp. 13–14.

³⁹ T. Nöldeke, Geschichte der Perser und Araber zur Zeit der Sasaniden aus der arabischen Chronik des Tabari: Übersetzt und mit ausführlichen Erläuterungen und ergänzungen Versehn (Leiden: E. J. Brill, 1879).

⁴⁰ Iwona Gajda, Le royaume de Himyar à l’époque monothéiste: L'histoire de l'Arabie du Sud ancienne de la fin du IVe siècle de l’ère chrétienne jusqu’à l'avènement de l'islam (Paris: Mémoires de l'Académie des Inscriptions et Belles-Lettres, 2009), p. 239.

⁴¹ I would argue that the entity which the inscription addresses is perhaps not an implied ‘God’, but Raḥmānān, although the word order is a little curious.

⁴² Although Abraha is not mentioned in an inscription until 541 ce (CIH 541), most scholars assume, based on sources like Procopius and the Martyrium of Saint Arethas, that he rose to power not long after 531 ce.

⁴³ The latest dated inscription that is clearly from Abraha's reign is Ja 547+546+544+545, dated to November in the year 668 in the Himyarite calendar, which is 558 ce.

⁴⁴ G. W. Bowersock, The Crucible of Islam (Cambridge, MA: Harvard University Press, 2017), pp. 19–20.

⁴⁵ Christian J. Robin, ‘Abraha et la reconquête de l'Arabie déserte: un réexamen de l'inscription Ryckmans 506 = Murayghan 1’, Jerusalem Studies in Arabic and Islam 39 (2012), pp. 2–94.

⁴⁶ English translation from Sahih International.

⁴⁷ Turaban (modern-day Turaba), mentioned in Ry 506, is about 300 kilometres from Mecca. Gajda, Himyar, p. 142.

⁴⁸ The text Murayghan 3, believed to be from a little before Ry 506, describes Abraha as ‘taking possession’ of a number of areas in Central Arabia, one of them being Yathrib, the earlier name for Medinah. See Robin, ‘Abraha et la reconquête’ for more.

⁴⁹ There is some debate as to when Muhammad was actually born, since all that we know from the tradition was that he began receiving visions at the age of 40. Furthermore, some have argued that the age of 40 is a literary topos and not necessarily the exact age of Muhammad. Lawrence I. Conrad, ‘Abraha and Muḥammad: Some Observations Apropos of Chronology and Literary “Topoi” in the Early Arabic Historical Tradition’, Bulletin of the School of Oriental and African Studies 50, no. 2 (1987), pp. 225–240.

⁵⁰ According to al-Tabari, Abraha was provoked by a person from Mecca defecating at his marvellous new cathedral. For annotations, see Nöldeke, Geschichte der Perser und Araber zur Zeit der Sasaniden aus der arabischen Chronik des Tabari, pp. 201–203.

⁵¹ Gajda, Himyar, p. 124.

⁵² Ibid., pp. 123–124.

⁵³ Christian J. Robin, ‘La Grande Église d'Abraha à Ṣanʿāʾ. Quelques remarques sur son emplacement ses dimensions et sa date’, in Interrelations between the Peoples of the Near East and Byzantium in Pre-Islamic Times, Semitica Antiqva 3, (ed.) Vasilios Christides (Cordoba: Oriens Academic, 2015), p. 121.

⁵⁴ Robin suggests that the inscription CIH 325, which mentions the construction of a building with ‘alabaster’ and ‘decorative stone’, actually describes the construction of the cathedral. CIH 325 is dated to 559–60 ce thus pushing Abraha's final military acts to after 560 ce. Robin, ‘La Grande Église d'Abraha à Ṣanʿāʾ’, pp. 120–121.

⁵⁵ Walter Dostal, ‘Mecca before the Time of the Prophet—Attempt of an Anthropological Interpretation’, Der Islam 68, no. 2 (1991), pp. 193–231.

⁵⁶ Wüstenfeld suggests that the ‘cross’ refers to the crosses of the Abyssinian standards. Regardless, it is the Arabic word ṣalīb which is used. To this day, the word simply means ‘cross’.

⁵⁷ al-Azraqi, Die Chroniken der Stadt Mekka. Gesammelt und auf Kosten der Deutschen Morgenländischen Gesellschafft herausgegeben von Ferdinand Wüstenfeld (Leipzig: F. A. Brockhaus, 1859), Vol 1, p. 96 in Arabic and Vol 4, p. 47 in German.

⁵⁸ This memory of the competing sanctuary in Sana'a may have lingered for some time; although the Qalis was not destroyed until circa 753 ce, parts of it were taken to Mecca and incorporated into the Ka'ba, perhaps as a semi-conscious act meant to underline the final victory of the Ka'ba as the most important shrine of the Arabian Peninsula. See Gajda, Himyar, pp. 125–126.

⁵⁹ Some, like al-Azmeh, argue that it was rather a desire to not be too specific in the oath, and to avoid using a particular deity. Granted, Allahumma is a broader term, but, if the Meccans were familiar with al-Rahman, it seems odd that they would deny its use simply on the basis that it was the name of a ‘foreign’ deity. Al-Azmeh, Emergence, pp. 229–231.