7. The Judaism of the Ancient Kingdom of Ḥimyar in Arabia: A Discreet Conversion

Christian Julien Robin (CNRS, Membre de l’Institut)

1.0. Introduction

Yemenite Judaism can be described as ‘rabbinic’ from the moment sufficient sources are available in the later Middle Ages.¹ It had probably been so for many centuries. One notes, for example, the epistolary links between Yemen’s Jewish communities and Moses Maimonides (d. 1204 CE), who sent them his celebrated Epistle to Yemen.

By contrast, the Judaism of Ḥimyar, the kingdom gradually extending its domination to the whole of ancient Yemen and even, between 350 and 570 CE, over a large proportion of the deserts of Arabia, seems to be different. That is what I shall attempt to demonstrate in this paper. I suggest a reappraisal of the entire file on Ḥimyarite Judaism in order to answer as fully as possible the two main questions: is it possible to claim that Ḥimyar converted to Judaism, and, if so, which type of Judaism was adopted by the Ḥimyarites?

Knowledge of the history of the kingdom of Ḥimyar (whose capital was located at Ẓafār, 125 km south of Ṣanʿāʾ) is relatively recent. Information is derived mainly from the inscriptions discovered following the opening of both Yemeni states to archaeological research at the beginning of the 1970s. A comparison between Hermann von Wissmann’s seminal 1964 article and Iwona Gajda’s 2009 book illustrates this complete change of perspective, which has resumed at a fast pace in recent years despite the war in Yemen.²

In the political field, it appears that Ḥimyar was the leading power in Arabia between approximately 350 and 570 CE, imposing its rule on the entire Peninsula (or at least a large part of it), except during the crisis years of 523–552 CE. In the religious field, the inscriptions illustrate in increasingly clear manner that Judaism was dominant in the kingdom of Ḥimyar from the fourth century CE until around 500–530 CE; they then show that Christianity became predominant, remaining the official religion for some forty years (530–570 CE). These discoveries do not agree with the data from the Arab-Muslim tradition, which emphasizes pre-Islamic Arabia’s isolation, polytheism, anarchy, and intellectual and material poverty.

Dealing with Ḥimyarite Judaism is no easy matter because religious identities are still fluid and difficult to distinguish in the fourth and fifth centuries CE. Furthermore, documentation is scarce and consists essentially of monumental inscriptions that only make the vaguest of allusions to religion. The archaeological remains cannot compensate for the laconic aspect of epigraphic material. One could even say that they are of no assistance at all, since no assuredly Jewish monument has been identified to this day. As for manuscripts, their utility is marginal.

My approach will necessarily be empirical. I will not attempt to answer the many questions that can be asked, but only to outline what is known today. As I already published all the available data on the nature of Ḥimyar’s Judaism in 2015,³ I will recall only the most significant facts here. I will then complete the discussion by examining to what extent the kingdom of Ḥimyar can be described as ‘Jewish’.

2.0. Sources on Religious Practices in the Kingdom of Ḥimyar

Shortly before the end of the fourth century, between 380 and 384 CE, a religious change of considerable importance took place in the kingdom of Ḥimyar. In January 384, the ruling kings, who had just built two palaces, commemorated these events in two inscriptions. The invocation formula concluding these two texts is, in itself, a break with the past: it no longer mentions the support of ancestral deities, as was previously the case, but of a new God: “With the support of the Lord, the Lord of the Sky.”

At first glance, the formula may seem banal and of no great consequence. Several polytheistic deities have a similar name. In South Arabia the great god of Najrān is called ‘The one of the Heavens’ (dhu-Samāwī or dhu ʾl-Samāwī).⁴ In Eastern Arabia a goddess is called ‘She who is in the Heavens’ (dhāt bi-[ʾl]-Samāwī),⁵ and in Syria an important god is ‘Master of the Heavens’ (Baʿal-Shamîn, with various orthographical variants of this name in different languages). By looking a little closer, one finds that the break with previous religious practices was a radical one, particularly evident in the evolution of terminology. One is assuredly dealing here with the establishment of a new religion.

Before highlighting this break with previous periods, it is quite useful to recall the nature of the available sources for Arabia’s religious history during the 250 years preceding the formation of Islam.⁶ These sources belong to three heterogeneous categories: Ḥimyarite inscriptions, external manuscript sources (mainly in the Greek and Syriac languages), and the ‘Arab-Muslim Tradition’ collected during the eighth and ninth centuries CE (second and third centuries of the Hijra).

2.1. Ḥimyarite Inscriptions

Ḥimyarite inscriptions do not inform us beyond 559–560 CE, the date of the most recent text. For the period between 380 and 560 CE, a total of some 150 texts are available, often fragmentary. Some three-fifths of these have a more or less precise chronology, with a date or reference to a known person or event. If one focuses on religious changes, relevant texts are only a few dozen in number. Most often these commemorate building activities.

One can infer the religious orientation of the inscriptions both through their invocations of celestial powers at the end (and, once, at the beginning) of texts and through their petitions. The formulation, which is always concise and stereotyped, and the onomastics are also illuminating.

2.2. External Sources

External sources are of real assistance only in the case of one episode of Arabian history: the long period of political and religious disorder that shook the kingdom of Ḥimyar in the first decades of the sixth century and led to its demise (c. 500–570 CE). Around 500 CE, the kingdom of Ḥimyar, where Jews enjoyed a dominant position, was placed under the tutelage of the Ethiopian kingdom of Aksūm. From then on, it was the (Christian) Negus who designated the ruler. When the Ḥimyarite Christian king died in 522 CE, the Negus nominated a successor. This prince, called Joseph (Masrūq in Syriac and Zurʿa dhū Nuwās in Arabic) soon rebelled. He massacred the Aksūmite garrison sent to Ẓafār by the Negus and then began to spread terror in the regions favourable to the Aksūmite party. He enjoyed the support of the Jewish party, but also of some Christians (apparently those of the Church of the East, called ‘Nestorian’).

Joseph’s vengeful policy provoked the dissidence of Miaphysite (or ‘Monophysite’) Christians in Najrān, who had refused to provide troops. Joseph repressed their rebellion through cunning and deceit and eventually exterminated them, no doubt reckoning that they were a threat on account of the close links they had established with Syria’s Byzantine provinces. Syria and Egypt’s ecclesiastical authorities seized the opportunity to make these victims martyrs of the faith and demanded a rapid response. With their assistance, Aksūm’s Negus gathered ships to carry his army across the Red Sea. Upon their arrival (sometime after Pentecost Day, 525 CE), Joseph was killed. Ḥimyar’s conquest, completed around 530 CE, brought the Negus as far as Najrān. It was followed by the systematic massacre of Jews. The country then became officially Christian. Churches were built and an ecclesiastical hierarchy was established. The conflict, which (at least in the beginning) seems to have been political in nature, is presented in ecclesiastical sources as a war of religion. This account is often quoted uncritically in historical works, especially since historical reports of the Arab-Muslim Tradition have adopted it.

The only documents contemporary with the events—some ten inscriptions written in June and July of 523 CE by the general and officers of the army sent by Joseph to repress the Najrān revolt—make no clear mention of religion. They do not explicitly claim to be Jewish; they do not quote the Bible; they do not boast that the army was invested with a sacred mission by religious authorities. To detect the Judaism of their authors, one can rely only on a small number of terms and turns of phrases meaningful only to specialists.⁷ Focusing largely on military operations, these documents are mainly aimed at terrorizing insurgents. It is clear that their purpose is political and not religious.

External sources mentioning Late Antique Arabia include above all the historical chronicles in Greek (particularly those of Procopius, Malalas, and Theophanes), and Syriac (like those of the Zuqnin monastery and of Michael the Syrian). One of the Greek chronicles, written by the Egyptian John of Nikiû, is known only in a Geʿez (classical Ethiopian) translation. Another, in Syriac, whose author remains unknown, has reached us only in its Arabic version (the Seert Chronicle). The summary of a Byzantine diplomatic report written by ambassador Nonnosus is also available. Emperor Justinian (527–565 CE) sent Nonnosus to Arabia and Ethiopia at an unknown date, probably in the early 540s. This summary appears in the Bibliotheca of Patriarch Photius (who died in 891 or 897 CE).⁸

The Ḥimyarite crisis is also known via Greek and Syriac texts produced by churches to celebrate the martyrs of South Arabia and to establish their cults: these are stories in the form of letters (the Guidi Letter, attributed to Simeon of Beth Arsham,⁹ and the Shahîd Letter in Syriac¹⁰), homilies, hymns, and hagiography (the Book of the Ḥimyarites in Syriac¹¹ and the Martyrdom of Arethas in Greek¹²). Two documents refer to events prior to the crisis of 523 CE: a hagiographical text in Geʿez, probably translated from Arabic, celebrating a priest of Najrān who was persecuted by the king of Ḥimyar Shuriḥbiʾīl Yakkuf (c. 468-480) (the Martyrdom of Azqīr),¹³ and the consolation letter written by Jacob of Serugh (who died in 521 CE) in honour of the Ḥimyarite martyrs.¹⁴

Apart from this Ḥimyarite crisis, the only significant event known to us is the dispatch of an embassy by the Byzantine Emperor Constantius II (337–361 CE) to convert the king of Ḥimyar. The account of this embassy can be found in Philostorgius’s fragments of the Ecclesiastical History transmitted by Photius: Philostorgius, an Arian ecclesiastical historian, was interested in this embassy because one of its leaders, Theophilus the Indian, was himself an Arian Christian.

As a general rule, external sources dealing with Late Antiquity do not focus on South Arabia at all. At most, Byzantine chroniclers make a passing note of desert Arabs when they launch forays into the Empire’s eastern provinces (which make up the Diocese of the Orient) or when the Empire asks them to join an alliance against Sāsānid Persia.

Since Eastern Arabia was conquered by Ḥimyar on two occasions—in 474 CE and 552 CE—one can incidentally mention that the proceedings of the Nestorian Church’s synods, known under the name Synodicon Orientale, and the correspondence of the heads of this church in the Syriac language, include precious information on the bishoprics of the Arab-Persian Gulf until the year 677 CE (i.e., some fifty years after the Islamic conquest).¹⁵

In sum, Greek and Syriac sources emphasize that Jews already exerted influence on the kingdom of Ḥimyar around the mid-fourth century CE and then enjoyed a dominant position until approximately the early sixth century CE, at the time of king Joseph.¹⁶

2.3. The Arab-Muslim Tradition

In order to reconstruct the history of pre-Islamic Arabia, other data is available from the ‘Arab-Muslim Tradition’, a convenient appellation for the set of texts recorded or written during Islam’s first centuries. These are not really internal sources; rather, they are diverse traditions collected and assembled in the schools of the Islamic Empire located mainly outside Arabia more than two centuries after the events. This tradition is particularly precious for the tribal geography and the study of place names. It has also preserved multiple individual testimonies of the events as experienced by Muḥammad’s companions or their immediate ancestors. This collective memory, however, is flimsy with regard to questions of general import, such as chronology, the pre-Islamic religions, or even the beginning of Arabic writing.

As concerns the Judaism of Ḥimyar, the Tradition retained that a king, Abū Karib Asʿad the Perfect (al-Kāmil), had introduced this religion into Yemen, and that another, Yūsuf Zurʿa dhū Nuwās, had become a Jew and had forced the Christians of Najrān to choose between conversion to Judaism or death. It incidentally signals that various other characters were also Jewish. Finally, four scholars of the Tradition give lists of the regions in which Jews could be encountered. These are: Ibn Qutayba (d. 889 CE),¹⁷ al-Yaʿqūbī (d. 897 CE),¹⁸ Ibn Ḥazm (d. 1064 CE),¹⁹ and Ibn ʿAbd al-Barr (d. 1071 CE).²⁰ Unsurprisingly, it appears that Judaism was solidly rooted in northwestern Arabia (the north of the Ḥijāz) and in the southwest of the Peninsula (in Yemen). More precisely, there were apparently Jews in Ḥimyar (or in Yemen), Kinda, banū ʾl-Ḥārith b. Kaʿb, Kināna, Ghassān, Judhām, al-Aws, al-Khazraj, and Khaybar. Sometimes one of these scholars considers that such-and-such a tribe included Jews in large numbers, while another gives a lower estimate, and a third says nothing on the matter. One should moreover note that the Jewish tribes of Yathrib (today al-Madīna)—al-Naḍīr, Qurayẓa, and Qaynuqāʿ—are not mentioned. This is undoubtedly due to the fact that these tribes were not included in the Great Genealogy of the Arabs, written in the second and third centuries after the Hijra.²¹

It bears emphasising that the sources just listed were first produced in a Christian environment and then in a Muslim one. None is of Jewish origin.

3.0. The Institution of an Official Religion as Revealed by Inscriptions

For a precise perception of the nature of the new religion established by Ḥimyar’s rulers—I shall come back later to the points proving we are effectively dealing with a new religion—only inscriptions are available, and these are not very many.

3.1. Four Categories of Monotheistic Inscriptions

The corpus on which we rely comprises all the texts later than the official establishment of the new religion and earlier than the final conquest of Ḥimyar by Christian Aksūmites. These are therefore the texts of the period 380–530 CE, whose number is roughly 140.

These inscriptions can be classified into four sets, corresponding to the institutional position of their authors: (1) official inscriptions, whose author is the king; (2) inscriptions whose author is a high official in the king’s service; (3) inscriptions whose author is a prince, ruling a territorial principality; and, finally, (4) inscriptions whose author is a private individual. It seems necessary to distinguish these diverse categories to appreciate as precisely as possible these documents’ meaning and exact scope. Only royal inscriptions define the official orientation used as a model in the entire country. The others provide complementary glimpses that are all the more precious since their composition was not subjected to the same constraints.

3.1.1. Royal Inscriptions

For the period 380–530 CE, sixteen royal inscriptions are available,²² a number that can be reduced to twelve if one discards four fragments that are too small to contribute any substantial information.²³ Four texts out of these twelve are particularly significant because they are long and complete, though they make no reference to religion at all. They share two remarkable traits. First of all, they do not originate from Yemen, but from the deserts of Arabia.²⁴ Moreover, they commemorate victorious military campaigns in these deserts. Two others celebrate the building of a place of worship without an invocation to God, either securely in one inscription (Ja 856 = Fa 60) or hypothetically in the other (YM 1200, which is fragmentary). A last text merely lists the ruler and his co-regents with their official title (Garb BSE). Royal texts that contain a religious invocation are five in number:

Garb Bayt al-Ashwal 2 (Ẓafār, capital of Ḥimyar), January 384 CE, dhu-diʾwān 493 ḥim. (Fig. 1): a commemoration of the construction of a palace in the capital by king Malkīkarib Yuha ʾmin and his co-regents,²⁵ these being his sons Abīkarib Asʿad and Dhara ʾʾamar Ayman: 

...b-mqm mrʾ-hmw mrʾ s1m(4)yn

With the support of their lord, the Lord of the Sk(4)y

RES 3383 (Ẓafār), January 384 CE, dhu-diʾwān 493 ḥim.: a commemoration, with the same date, of the construction of a second palace in the capital by these same rulers, king Malkīkarib Yuha ʾmin and his co-regents, his sons Abīkarib Asʿad and Dhara ʾʾamar Ayman:

...b-mqm m(4)rʾ-hmw mrʾ (s1my)[n]

With the support of ⁽⁴⁾ their lord, the Lord of the Sky

YM 327 = Ja 520 (Ḍahr, 10 km northwest of Ṣanʿāʾ): a commemoration at an uncertain date of a building several stories high by king Abīkarib Asʿad, then in co-regency with his brother Dhara ʾʾamar Ayman and his sons Ḥaśśān Yuʾmin, Maʿdīkarib Yunʿim, and Ḥugr Ayfaʿ:

[…](5)(n) l-ḏt ẖmr-hmw rḥ[mnn ...]

[…]⁽⁵⁾ so that Raḥ[mānān] may grant them […]

CIH 540 (Ma ʾrib , 120 km east of Ṣanʿāʾ), January 456 CE, dhu-diʾw 565 ḥim. (Fig. 2): the commemoration of an important restoration of the Marib Dam²⁶ by king Shuriḥbiʾīl Yaʿfur:

...b-nṣr w-rdʾ ʾlhn b(82)ʿl s1myn w-ʾrḍn

With the aid and help of God (Ilāhān), ow(82)ner of the Sky and the Earth

ẒM 1 = Garb Shuriḥbiʾīl Yaʿfur (Ẓafār), December 462 CE, dhu-ālān 572 ḥim. (Fig. 3): a commemoration of the construction of a palace in the capital by the same king, Shuriḥbiʾīl Yaʿfur:

...b-nṣr w-rdʾ w-mqm mrʾ-hmw rḥmnn bʿl (13) s1myn (w-ʾ)rḍ(n)

With the help, aid, and support of their lord Raḥmānān, owner (13) of the Sky and the Earth²⁷

It is remarkable that these five texts contain no dogmatic formulation indicating a precise religious affiliation. From this viewpoint, they are quite different from royal inscriptions later than 530 CE, which begin with an invocation to the Holy Trinity.²⁸

3.1.2. Inscriptions by High Officials in the King’s Service

Several texts of the period 380–530 CE are more explicit regarding their authors’ beliefs. Of these, the most important are the inscriptions written by high officials in the service of the king.

Garb Bayt al-Ashwal 1 (Zafār), undated, whose author does not invoke the ruling king (Abīkarib Asʿad), but only a co-regent, Dhara ʾʾamar Ayman (around 380–420 CE), which makes one think that he is in the service of the latter. The author, Yehuda Yakkuf, is a Jew, as proven by a small graffito in Hebrew incised in the central monogram. As the language bears the imprint of Aramaic,²⁹ he might be of foreign origin (Fig. 4 and 5):

...b-rdʾ w-b-zkt mrʾ-hw ḏ-brʾ nfs1-hw mrʾ ḥyn w-mwtn mrʾ s1(3)myn w-ʾrḍn ḏ-brʾ klm w-b-ṣlt s2ʿb-hw ys3rʾl

With the assistance and the grace of his Lord who created him, the Lord of life and of death, the Lord of the Sk(3)y and the Earth, who created everything, with the prayer of his commune Israel

Ry 508 (Ḥimà, 100 km northeast of Najrān), June 523 CE, dhu-qiyāẓān 633 ḥim. (Fig. 6): a proclamation by the army general whom the Jewish king Joseph (mentioned in the text) has sent to crush the Najrān revolt. The text, which recalls the miltary events of the previous year, implicitly incites the insurgents to submit:

...w-ʾʾlhn ḏ-l-hw s1myn w-ʾrḍn l-yṣrn mlkn ys1f b-ʿly kl ʾs2nʾ-hw w-b-(11)ẖfr rḥmnn *ḏ*n ms1ndn bn kl ẖs1s1{s1}m w-mẖdʿm w-trḥm ʿly kl ʿlm rḥmnn rḥm-k mrʾ ʾt

May God (Aʾlāhān = Elôhîm), to whom the Sky and the Earth belong, grant king Joseph (Yūsuf) victory over all of his enemies. With ⁽¹¹⁾ the protection of Raḥmānān, that this inscription [may be protected] against any author of damage and degradation. Extend over the entire universe, Raḥmānān, your mercifulness. Lord you are indeed

Ry 515 (Ḥimà), undated, but assuredly contemporary with Ry 508, because it is carved to the left of the latter and is written by officers of its author (Fig. 7):

...rb-hwd b-rḥmnn

Lord of the Jews, with Raḥmānān

Ja 1028 (Ḥimà), July 523 CE, dhu-madhra ʾān 633 ḥim.: a new proclamation by the author of Ry 508, but written a month later (Fig. 8 and 9):

⁽¹⁾ l-ybrkn ʾln ḏ-l-hw s1myn w-ʾrḍn mlkn yws1f ʾs1ʾ (vac.) r yṯʾr mlk kl ʾs2ʿbn

May God (Īlān), to whom the Sky and the Earth belong, bless the king Joseph (Yūsuf) Asʾar Yathʾar, king of all the communes

...w-l-ybrkn rḥmnn bny-hw (line 9)

May Raḥmānān bless his sons

...w-k-b-ẖfrt s1myn w-ʾrḍn w-ʾʾḏn ʾs1dn ḏn ms1ndn bn kl ẖs1s1m w-mẖdʿm w-rḥmnn ʿlyn b(12)n kl mẖd(ʿ)m bn m(ṣ).. wtf w-s1ṭr w-qdm ʿly s1m rḥmnn wtf tmmm ḏ-ḥḍyt rb-hd b-mḥmd

With the protection of the Sky and the Earth and the capacities of men, may this inscription [be protected] against any author of damage or degradation, and Raḥmānān Most-High, ag⁽¹²⁾ainst any author of degradation [... …] The narration of Tamīm dhu-Ḥaḍyat was composed, written, and carried out in the name of Raḥmānān, Lord of the Jews, with the Praised One

Ry 507 (Ḥimà), the same date as Ja 1028, July 523 CE, dhu-madhra ʾān 633 ḥim.: another proclamation by the author of Ry 508 and Ja 1028:

(1)l-ybr(kn ʾl)hn( ḏ-)l-h(w s1)[myn w-ʾrḍn mlkn ys1f ʾs1ʾr Yṯʾr mlk kl ]ʾs2ʿb(n)

May God (Ilāhān), to whom the S[ky and the Earth] belong, [bless the king Joseph (Yūsuf) Asʾar Yathʾar, king of all the] communes

...w-b-ẖfrt (11) [mrʾ s1]myn w-ʾrḍn

With the protection of [the Lord of the S]ky and the Earth

3.1.3. Inscriptions by Princes at the Head of Territorial Principalities

Inscriptions written by princes ruling a principality also yield useful information on the topic. Two examples will suffice here:

Ry 534 + Rayda 1 (Rayda, 55 km north of Ṣanʿāʾ), August 433 CE, dhu-khirāfān 543 ḥim. (Fig. 10): text commemorating the construction of a mikrāb by a Hamdānid, prince of the Ḥāshid^um and Bakīl^um (dhu-Raydat fraction) communes, under the reign of Abīkarib Asʿad with his four sons as co-regents:

...(brʾ)w w-hs2qr mkrbn brk l-ʾl (2) mrʾ s1myn w-ʾrḍn l-wfy ʾmrʾ-hmw … (3) … w-l-ẖmr-hm ʾln mrʾ s1myn w-ʾrḍn (4) ṣbs1 s1m-hw w-wfy ʾfs1-hmw w-nẓr-hmw w-s2w[f-h]mw b-ḍrm w-s1lm

(The author) has built and completed the synagogue Barīk for God (Īl),(2) Lord of the Sky and the Earth, for the salvation of their lords … ⁽³⁾ … so that God (Īlān), Lord of the Sky and the Earth, may grant them ⁽⁴⁾ the fear of his name and the salvation of their selves, their companions and of their subj[ects,] in times of war and peace

Ry 520 (according to the text, from Ḍulaʿ a few kilometres northwest of Ṣanʿāʾ), January 465 CE, dhu-diʾwān 574 ḥim.: commemorating the construction of a mikrāb by a Kibsiyide prince of the Tanʿim^um commune, 25 km east of Ṣanʿāʾ, probably at the time of king Shuriḥbiʾīl Yaʿfur (who is not mentioned):

...hqs2b(4)w mkrbn yʿq b-hgr-hmw ḍlʿm l-mrʾ-hm(5)w rḥmnn bʿl s1myn l-ẖmr-hw w-ʾḥs2kt-(6)hw w-wld-hw rḥmnn ḥyy ḥyw ṣdqm w-(7)mwt mwt ṣdqm w-l-ẖmr-hw rḥmnn wld(8)m ṣlḥm s1bʾm l-s1m-rḥmnn

(The author) has built from ne⁽⁴⁾w the synagogue Yaʿūq in their city of Ḍulaʿ^um for his lor⁽⁵⁾d Raḥmānān, owner of the Sky, so that Raḥmānān may grant him, as well as to his wi⁽⁶⁾fe and to his sons, to live a just life and to ⁽⁷⁾ die a worthy death, and so that Raḥmānān may grant them virtuous ⁽⁸⁾ children, in the service for the name of Raḥmānān

3.1.4. Inscriptions by Private Individuals

The file also contains a few texts whose authors are private individuals or officials who do not mention their responsibilities or their duties.

ẒM 5 + 8 + 10 (Ẓafār), February 432 CE, dhu-ḥillatān [5]42 ḥim. (Fig. 11): a commemoration of the construction of two palaces under the reign of Abīkarib Asʿad (who is not mentioned):

...b-zkt rḥ[mnn w-b-rdʾ w-...] (5) ʾmlkn ʾbʿl byt[n] rydn w-mrʾ s1my(n)[... ] (6) ḥyw b-ʿml-hmw ʾks3ḥ ṭwʿ ʾfs1-h(m)[w ... ... mrʾ] (7)s1myn bn kl bʾs1tm w-l-yẖmrn-hmw mw[t …] (8) w-ʾmn

With the grace of Raḥ[mānān and the help and ... ] ⁽⁵⁾ of kings, owners of the palace Raydān, and the Lord of the Sky [ ... ... ...] ⁽⁶⁾ a life with their works, exemplary(?) of the submission of their selves [... ... the Lord] (7) of the Sky against all evil, and that he may grant a deat[h ... ... ...] ⁽⁸⁾ and āmēn

ẒM 2000 (Ẓafār), April 470 CE, dhu-thābatān 580 ḥim. (Fig. 12): a commemoration of the construction of a palace under the reign of king Shuriḥbiʾīl (Yakkuf):

...w-b (6) rdʾ w-ẖyl mrʾ-hmw ʾln (7) bʿl s1myn w-ʾrḍn w-b-rdʾ (8) (s2)ʿb-hmw ys3rʾl w-b-rdʾ mrʾ-hmw s2rḥ(b)(9)ʾl mlk sbʾ w-ḏ-rydn w-ḥḍrmwt w-l-(ẖ)(10)mr-hmw b-hw rḥmnn ḥywm ks3ḥ[m]

With ⁽⁶⁾ the assistance and the power of their lord God (Īlān) (7) owner of the Sky and the Earth, with the assistance ⁽⁸⁾ of their commune Israel and with the assistance of their lord Shuriḥbi⁽⁹⁾ʾīl king of Saba ʾ, dhu-Raydān and of Ḥaḍramawt. May ⁽¹⁰⁾ Raḥmānān give them here (in this house) an exemplary life

CIH 543 = ẒM 772 A + B (Ẓafār), undated; the purpose of this text is unknown:

[b]rk w-tbrk s1m rḥmnn ḏ-b-s1myn w-ys3rʾl w-(2)’lh-hmw rb-yhd ḏ-hrdʾ ʿbd-hmw…

[May it bl]ess and be blessed, the name of Raḥmānān, who is in the Sky, Israel and ⁽²⁾ their God, the Lord of the Jews, who has helped their servant…

Garb Framm. 7, of unknown provenance and date: a fragment of an inscription commemorating the final stage of a construction under the reign of Abīkarib Asʿad, ruling in co-regency with his brother Dhara ʾʾamar Ayman and his son Ḥaśśān Yuha ʾmin:

...b-(r)[dʾ mrʾ-hw mrʾ s1myn w-b] (2) [rdʾ s2ʿb-](h)w Ys3rʾl

With the he[lp of his lord, the Lord of the Sky, with] ⁽²⁾ [the help of his commu]ne Israel

3.2. A Radical Reform

The religious reform that took place around the year 380 CE reveals a radical aspect. From this date, all royal inscriptions became monotheistic. What is even more remarkable is that polytheistic inscriptions disappeared almost immediately.³⁰ Only two such texts are known from the two decades following the reform. However, they are not from the capital, where the power structure controlled public expression, but from the countryside.³¹

Even if the corpus of documents is not very substantial, the break with the past is radical in terms of both lexicon and phraseology. The most prominent change is the manner of designating God and places of worship, as we shall see later.³² One also notes the radical change in the lexicon relating to the human self. Traditionally, inscriptions mentioned various components, mainly the ‘capacities’ (ʾʾḏn) and the ‘means’ (mqymt), as in Ir 12 / 9 (Ma ʾrib, text going back to the reign of Shaʿr^um Awtar, early third century CE):

...w-l-ẖmr-hw ʾlmqh bry ʾʾḏnm w-mqymtm

And may (the god) Almaqah grant them capacities and means to the fullest

This vocabulary also appears in a single monotheistic inscription, CIH 152 + 151 (Najr, near ʿAmrān, 45 km northwest of Ṣanʿāʾ):

[...].t ʾ(ḥṣ)n w-bn-hw s2rḥʾl bnw mrṯdm w-qyḥn br(ʾ)[w w-] (2) [.....] mkrbn l-wfy-hmw w-ẖmr-hmw ʾln bry ʾʾḏnm w-mqymtm [...]

[...].. Aḥsan and his son Shuriḥbiʾīl banū Murāthid^um and Qayḥān have bu[ilt ... ... ⁽²⁾ ... ...] the synagogue so that God (Īlān) may save them and grant them capacities and means to the fullest [...]

The inscription is undated and relates to the new religion, since it commemorates the construction of a mikrāb and addresses a prayer to the One God, called Īlān here. It still makes use of the vocabulary of the traditional religion, particularly the substantive nouns ʾʾḏn and mqymt and the verb ẖmr. Later on, only the verb ẖmr (‘to grant’) is still employed. One might suppose that the inscription CIH 152 + 151 goes back to a transitional period between the old and new practices, perhaps around the mid-fourth century CE.

In addition to the change in terminology, one should also note the appearance of some twenty terms and proper nouns borrowed from Aramaic and Hebrew.³³

While the inscriptions employ new religious terminology after the religious reform, one nevertheless notices a certain continuity in their structure. Traditionally, inscriptions first mention their authors; they then recall, in the third person, the deeds they accomplished; lastly, they invoke the celestial and terrestrial powers who favoured or supported the operations mentioned. The inscriptions of the period 380–500 CE preserve the same structure. It is only after 500 CE that one observes a radical transformation, illustrated by the invocation to God occasionally placed at the beginning of the text. During the period 500–530 CE, one finds it in a dated Jewish inscription (Ja 1028, Ḥimà, July 523 CE, dhu-madhra ʾān 633 ḥim.):

l-ybrkn ʾln ḏ-l-hw s1myn w-ʾrḍn mlkn yws1f ʾs1ʾ (vac.) r yṯʾr mlk kl ʾs2ʿbn w-l-ybrkn ʾqwln/(2) lḥyʿt yrẖm w-s1myfʿ ʾs2wʿ w-s2rḥʾl yqbl w-s2rḥbʾl ʾs1ʿ (vac.) d bny s2rḥbʾl ykml ʾlht yzʾn w-gdnm

May God (Īlān), to whom the Sky and the Earth belong, bless the king Joseph (Yūsuf) Asʾar Yathʾar, king of all the communes, and may He bless the princes ⁽²⁾ Laḥayʿat Yarkham, Sumūyafaʿ Ashwaʿ, Sharaḥʾīl Yaqbul and Shuriḥbiʾīl Asʿad, sons of Shuriḥbiʾīl Yakmul, (of the lineage) of Yazʾan and Gadan^um

The same change can be noticed in a dated inscription where no explicit sign of religious orientation is apparent (Garb Antichità 9 d, Ẓafār, March 509 CE, dhu-maʿūn 619 ḥim.):

[b-nṣr w-](b-)ḥmd rḥmnn bʿl s1myn w-b (2) [rdʾ ](mr)ʾ-hmw mlkn mrṯdʾln ynwf

[With the help and] the praise of Raḥmānān, owner of the Sky, and with ⁽²⁾ [the aid] of their lord king Marthadʾīlān Yanūf

Lastly, one notes this change in two undated inscriptions, one of them Jewish (CIH 543 = ZM 772 A + B, already quoted),³⁴ and the other devoid of any explicit religious orientation (RES 4109 = M. 60.1277 = Ja 117 = Ghul-YU 35, of unknown provenance):

l-ys1mʿn rḥmnn (2) ḥmdm ks1dyn

May Raḥmānān answer the prayers of ⁽²⁾ Ḥamīd^um the Kasdite

Changing the location of the invocation to God in the text becomes systematic in Christian inscriptions, all of which are later than 530 CE. This change no doubt emphasizes that God is now conceived of as the main player in earthly matters and that nothing can be accomplished against His will.³⁵

If the religious break with the past around 380 CE is both radical and systematic, it is also the final stage of an evolution observable over several decades. Only half of the inscriptions from the fourth century prior to 380 CE continue to celebrate or invoke ancient deities, which was previously the norm for all inscriptions. The others have already adopted the One God or abstain from making any reference to religion Those postdating 380 CE invoke no divinity other than the One God, with the possible exception of a single text whose precise date is uncertain.³⁶

Most temples were already deserted during the third and fourth centuries CE.³⁷ More precisely, one ceases to find in these places of worship inscriptions commemorating offerings, which implies that the wealthiest worshippers no longer entered them. The only temple that still received offerings after the mid-fourth century CE was Marib’s Great Temple, dedicated to the great Sabaean god Almaqah. In this temple, excavators have uncovered some eight-hundred inscriptions for the period between the first and fourth centuries CE. The last dated inscription comes from 379–380 CE.³⁸ It is likely that the authorities closed the temple immediately after this date, since official policy from then on was clearly unfavourable to polytheism. But it cannot be excluded that the closure was a little later and that the temple had been visited discreetly by worshippers for some time. One can moreover notice that the entrance hall was refurbished around this period, as attested by the inscribed stelae reused in the paving.³⁹ This redevelopment is probably related to a new use of the monument.

Of course, inscriptions, whose conception and carving were costly and whose authors belonged to the elite class, do not reflect exactly the religious practices of the entire society. One may even suspect that they do not even reflect these elites’ real religious practices, but only those the authorities encouraged. It is indeed quite difficult to believe that the entire group of princely lineages unanimously and simultaneously rejected polytheism in order to convert to a new religion. Inscriptions teach us above all that in public space, from 380 CE, only the new religion could be mentioned.

The date of the break can be pinpointed with a certain measure of precision. It occurred for certain before January 384 CE and probably a little before. Since the last polytheistic inscription in Marib’s great polytheist temple bears the date of 379–380 CE,⁴⁰ I shall retain the interval 380–384 CE. It is not impossible, however, that the official establishment of the new religion took place a little earlier, if indeed one supposes that it did not immediately entail the abandonment and closure of polytheistic temples.⁴¹

An external source—and an imprecise one, at that—nevertheless agrees quite well with the data from the inscriptions. The already-mentioned Ecclesiastical History of Philostorgius recalls that the Byzantine Emperor Constantius II (337–361 CE) sent an embassy to Ḥimyar’s king to invite him to convert to the Christian faith.⁴² One can therefore surmise that Constantius II had been informed that Ḥimyar was favourable to such as invitation. The embassy’s date is not known for certain, but it can probably be dated to the early 340s CE. One of the embassy’s leaders, the Arian Christian Theophilus the Indian, recalls that the embassy did not achieve its aims because of the Jews in the king’s entourage, but that the king (whose name is not given) agreed to build with his own funds three churches in the capital and in two of the country’s ports (implicitly for the Romans residing there).⁴³

3.3. Problems the Change of Religion Solved

The adoption of a new religion is not a trivial or insignificant act. This was the antique equivalent of a modern revolution. The fourth century CE was a period where radical change of religion became a surprising trend in the manner of the nineteenth century liberal revolutions. Armenia paved the way, followed by Caucasian Iberia (Georgia), the Roman Empire, Ethiopia, the Arabs (of the Syrian desert and the Sinai), and then Ḥimyar.

The reasons why the king of Ḥimyar established a new religion are a matter of guesswork. The authorities’ main ambition was to reinforce the cohesion of the empire and ensure the regime’s stability. Prior to Ḥimyar’s conquests, religious diversity was great. Each kingdom had its own great god and its own pantheon (that is to say, a small number of deities that were the focus of official worship practiced collectively). The great god had his great temple in the capital and an additional temple in each of the kingdom’s major regions, with the exception of those where a local god could be worshipped in place of the great god, this being a more or less formally declared assimilation.

In Saba ʾ, the great god was Almaqah, who had his great temple in Marib; in Qatabān, it was ʿAmm, with his great temple in Tamnaʿ; and in Ḥaḍramawt, it was Sayīn, whose great temple was in Shabwat. In these kingdoms founded in remote antiquity (before 700 BCE), the distribution of rites could be completely superimposed on the political map. In other words, in any kingdom, only the subjects of this kingdom would participate in official rites; reciprocally, belonging to a kingdom (particularly following an annexation) implied participating in the rites in honour of the kingdom’s great god.

In the kingdom of Ḥimyar, founded in the first century BCE, matters were different. Political unity did not (apparently) entail the establishment of official collective rites. Each of the kingdom’s regions preserved its traditional rites, with the god ʿAthtar in the north and the god ʿAmm in the southeast.

Ḥimyarite expansionism, which had resulted in the annexation of Qatabān, Saba ʾ, and Ḥaḍramawt (between 175 and 300 CE), did not immediately affect religion. Pilgrimages to Almaqah and Sayīn continued to be held as normal for a certain time. Religious diversity nevertheless did not go without posing some practical issues. As a result of the redistribution of territories, princedoms often united communes worshipping different deities. The Ḥimyarite ruler was obviously fearful of ancient cults being used by political competitors to organize hostile forces.

Despite not having been very interventionist in religious matters, the Ḥimyarite ruling class decided to change policy radically around 380 CE. This was perhaps because new problems had then arisen. Three of these can be recognized.

First of all, the rejection of ancient religious practices seems to have been a general phenomenon, at least in the princely lineages of the mountains. Reform could therefore be a response to the demand for a more personal and spiritual religion.

Secondly, the king of Ḥimyar was firmly requested by both Sāsānid Persia and Byzantium to choose his camp at a moment when these two powers were fighting over control of the Peninsula. As early as the 340s CE, as already mentioned, Byzantium had sent an embassy with sumptuous gifts to convince the Ḥimyarite ruler to accept baptism; moreover, the Christian mission was beginning to gain followers in the Arab-Persian Gulf. Ḥimyar finally refused to join Byzantium’s alliance because its hereditary enemy, the Ethiopian kingdom of Aksūm—a traditional ally of the Romans—was already well on its way to conversion to Christianity. In such a context, the choice of a new religion could be a way of resisting Byzantine pressure precisely at a moment when the Byzantine throne was weakened.⁴⁴

One should also take financial aspects into account. In ancient Arabian society, authorities benefitted from three available sources of revenue. Of these, the most important consisted of taxing a certain proportion of harvests and the natural growth of herds. Temples were responsible for this form of taxation, which went back to very ancient times, even as ancient as the very development of agriculture, perhaps as early as the third millennium BCE. Inscriptions distinguish two types of taxes, called ʿs2r⁴⁵ and frʿ, whose nature and amount are unknown.⁴⁶

In South Arabian temples, archaeologists have discovered a large number of inscriptions commemorating offerings. It would appear that a large fraction of these offerings were not spontaneous gifts thanking the deity for a favour or the accomplishment of a promise, but an ostentatious means of paying taxes. Indeed, one should note that offerings were habitually placed on a stone base on which the donor had carved an inscription; for the donor, this inscription, theoretically commemorating the rite, was an occasion to flaunt his status.

Temples possessed not only an immense treasury, consisting of innumerable accumulated offerings, but also property (no doubt in the form of landed estates, livestock, and financial means). It is therefore likely that they played an important part in economic life. Many monetary emissions show a divine symbol. These symbols appear particularly on the coinage of Saba ʾ (where all minted coins carry the symbol of the great Sabaean god Almaqah) and of Ḥaḍramawt (where many series bear the name of Sayīn). We are not yet, however, in a position to assess how the part played by the temple in coinage was reconciled with that of the king.

The second source of revenue consisted of custom duties on trade, mainly taxes on markets and passports, to which one can add the benefits of services (accommodation, food, water, storage, security). Apparently, this source of income, which only became substantial in the first millenium BCE, was a prerogative of political power. Trade was a matter for the king only, as he controlled markets and the circulation of goods. A few inscriptions in temples, however, indicate that the offering being commemorated was financed with the benefits of trade. It is not known in this case whether the authors of inscriptions paid a tax to the deity or whether they were showing their gratitude for returning safe and sound from a perilous journey after making comfortable profits.⁴⁷

The third source of income was the seizure of war booty. This booty was habitually destined for political rulers, but sometimes also for the temple. Thus, a handful of inscriptions, all dating from a brief period of the early third century CE, commemorate offerings made in the great temple of the god Almaqah in Marib with the booty taken from Shabwat and Qaryat^um. The meaning of this exception is unknown. Did the king at the time dedicate his share of the booty to the god to thank him for an exceptional favour?

This brief reminder shows that taxes deposited in the temples played an important part in economic life. Most temples ceased receiving offerings commemorated by inscriptions—no doubt those that had the greatest value—sometime during the third or fourth century CE. In tandem with the crisis of polytheism, they also lost part of their financial resources and could not play the same important role in the economy.

As for the landowners of estates and herds who rejected ancestral religious practices, they were, by the same act, freeing themselves of taxes they owed the temple. State intervention was therefore necessary to reorganize public finances. Nothing is known, unfortunately, of this reorganization. One can only notice that no South Arabian emission of coins postdates the religious reform.

In summary, this religious reform had several aims. The first was to re-establish the old correspondence between political groups and the distribution of religious rites. The second was to resist Byzantine pressure. The third consisted in replacing the temple as the beneficiary of taxation. One can undoubtedly add a last goal: the conversion to a new religion, which transformed the past into a tabula rasa and obliterated past times, enabled the monarchy and principalities to seize treasures accumulated in polytheist sanctuaries.

4.0. The New Religion’s Main Traits

The most noteworthy novelties brought by the new religion were threefold: the appearance of a single God with multiple appellations, clearly distinguishable from the innumerable deities of the past; the institution of a new place of worship; and, finally, the appearance of a new social entity called ‘Israel’.

4.1. One God

A single God replaced the old polytheistic deities of South Arabia: Almaqah, ʿAthtar, Ta ʾlab, Wadd, Sayīn, dhāt-Ḥimyam, dhāt-Ẓahrān, al-ʿUzzà, Manāt, al-Lāh, al-Lāt and many others. This single God was designated in multiple ways. The earliest attestations called him ‘Owner of the Sky’ (bʿl s1myn), ‘Lord of the Sky’ (mrʾ s1myn), ‘God’ (īlān, ʾln), or ‘God, Lord of the Sky’ (ʾln mrʾ s1myn). This new God was fundamentally a celestial power. However, very quickly, it was specified that this God of the Sky also ruled the Earth: He was “the Lord of the Sky and the Earth, who has created all things” (mrʾ s1myn w-ʾrḍn ḏ-brʾ klm).

All these denominations are interchangeable because they are evenly distributed in the various inscription categories I have determined.⁴⁸ The name Īlān includes the root ʾl, which means ‘god’, and the suffix definite article -ān. It deserves a few words of explanation. In the Near East of the second millennium BCE, a supreme god named Ēl or Īl was worshipped; from his name the appellation īl ‘god’ was derived (if indeed the derivation did not occur in the opposite way).

In South Arabia, this Near Eastern heritage took two forms. In Sabaʾ, a god Īl was worshipped in very ancient times, from around the eighth to sixth centuries BCE. Nevertheless, to designate a divine being, a derivative ʾlh (vocalized probably as ilāh) was used. It is found, for instance, in the very common syntagm “dhu-Samāwī god of Amīr^um” (ḏ-s1mwy ʾlh ʾmrm). This appellative ʾlh preserves the same spelling when a suffix is added. See, for example, “his god dhu-Samāw⁽⁴⁾ī owner of Baqar^um” (ʾlh-hw ḏ-s1mw⁽⁴⁾y bʿl bqrm)⁴⁹ or “his god Qaynān owner of Awtan” (ʾl⁽⁴⁾h-hw qynn bʿl ⁽⁵⁾ ʾwtn).⁵⁰ With the definite article, ʾlhn (ilāhān) means ‘the god’ in a polytheist context. See, for instance, “the sanctuary of the god dhu-Samāwī, god of Amīr^um” (mḥrm ʾlhn ⁽³⁾ [ḏ-s1mw]y ʾlh ʾmrm).⁵¹ ʾlhn is also attested as one of the names of the monotheist god already mentioned in CIH 540 as “God (Ilāhān), ow⁽⁸²⁾ner of the Sky and the Earth” (ʾlhn b⁽⁸²⁾ʿl s1myn w-ʾrḍn). The noun ʾlh is assuredly a derivative of ʾl with a consonant added to fit the triliteral mould, as indicated by the unusual form of its plural: ʾlʾlt, which was formed by the doubling of the root ʾl.

In Qatabān, where the god Īl is not attested, one notices a substantive noun ʾl meaning ‘god’, often designating the tutelary god (called s2ym in Sabaic):

…s1qnyw l-ʾl-s1m w-mrʾ-(3)s1m ḥwkm nbṭ w-ʾlh-s1ww ʾlhy bytn (4) s2bʿn

[the authors] have offered to their god and to their ⁽³⁾ lord Ḥawkam Nabaṭ and to his deities, the deities of the temple Shabʿān⁵²

The noun ʾl can also be used for the god of a region: “with (the god) ʿAmm, with (the god) Ḥawkam and with Ḥbr god of Shukaʿ^um” (b-ʿm w-b-ḥwkm w-b-ḥbr ʾl s2kʿm).⁵³ Finally, it can refer to any god whom it is not necessary to name if the context is clear: “[the authors] carried out the restoration of the basin belonging to the treasury of the god at Bana ʾ” (…s1ḥdṯ ṣʾrtn bn mbʿl ʾln b-bnʾ).⁵⁴ The plural of ʾl, attested only in the construct state, is ʾlhw or ʾlhy.

In polytheistic Ḥimyarite inscriptions, written in a Sabaic showing certain peculiarities, the usual term for ‘god’ is the substantive noun ʾl, without /h/, as in Qatabānic. See, for example, “(the author) has offered to his god and his lord Rgbn mistress of Ḥaẓīrān…” (hqny ʾl-h⁽⁴⁾w w-mrʾ-hw rgbn bʿlt ḥ⁽⁵⁾ẓrn).⁵⁵

The One God of the Ḥimyarites, sometimes called Īlān ‘the God’ in the earliest inscriptions, soon received a new name derived from Aramaic, Raḥmānān ‘the Merciful’. Its oldest attestation dates from approximately 420 CE. Between 420 and 450 CE, Raḥmānān became increasingly frequent, but would freely alternate with six other names. Among these, the most significant was ʾʾlhn, for which only one attestation is known (Ry 508). One can analyse ʾʾlhn as a noun of the ʾfʿl scheme, which expresses a plural. God is therefore designated here by a plural of ʾlh, which is not the usual plural (in general, ʾlʾlt, and twice ʾhlht).⁵⁶ The term ʾʾlhn (perhaps to be vocalized as Aʾlāhān) is therefore particularly interesting, since it is an innovation that apparently closely copies Hebrew ʾĕlōhīm.

The name Raḥmānān, which one can find in Qurʾānic Arabic under the form al-Raḥmān, refers to the quality of mercy.⁵⁷ This quality, which in Judaism is initially less commonly associated with the idea of God,⁵⁸ became common in Late Antiquity.⁵⁹ As a name for God, it is frequent in the Babylonian Talmud, but less so in the Jerusalem Talmud. It is attested in the Targum; one can also find it in Christian Palestinian Aramaic and in Syriac.⁶⁰ The fact that one of the names for God in the Qurʾān refers to the idea of mercy (or, rather, of beneficence⁶¹) appears to be significant. Muḥammad began his mission with apocalyptic overtones by announcing the End of Time and the Last Judgment. In such a context, the qualities of God are rather anger and intractable justice. The adoption of al-Raḥmān as a name of God (or as one of His names) no doubt reflects a shift that can be associated with the foundation in 622 CE of the theocratic principality of al-Madīna. From then on, the End of Time is not as close as previously believed, because God has shown himself to be compassionate. Muḥammad now prepares for the long term and worries more about the functioning of his community.

The name Raḥmānān is sometimes rendered more explicit by a qualifier. In a clearly Jewish text dating to July 523, he is described as “Most-High” (rḥmnn ʿlyn in Ja 1028 / 11). Elsewhere, it is the adjective ‘merciful’ that one can find in a text whose religious orientation is unclear (rḥmnn mtrḥmn, in Fa 74 / 3, Ma ʾrib, July 504). Finally, in a text with Jewish undertones, but dating to the Christian period, one finds “Raḥmānān the King” (rḥmnn mlkn, Ja 547 + 546 + 544 + 545 = Sadd Ma ʾrib 6, November 558 CE). Only once the reference to Raḥmānān is made explicit by a second term, bhṯ (Robin-Viallard 1 = Ja 3205, Ẓafār, May 519 CE, dhu-mabkarān 629 ḥim.). Unfortunately, the meaning of the latter is uncertain:

...w-l-ys1mʿn-h(5)mw rḥmnn w-kl bhṯ-hw w-ʾẖw(6)t-hmw

May ⁽⁵⁾ Raḥmānān with all His powers (?) listen to them, and to their bro⁽⁵⁾thers

It is quite remarkable that the names of the one God evolved in comparable ways in both the kingdoms of Ḥimyar and Aksūm. In the inscriptions written by king ʿEzānā following his official conversion to Christianity towards the beginning of the 360s CE,⁶² one notes the use of neutral names appealing to many different religious orientations. In particular, one finds the reference to God as a celestial power: “the Lord of the Sky who, in the Sky and on the Earth, is victorious for me” (ʾɘgzīʾa samāy [za-ba] samāy wa-mədr mawāʾī līta); then shortened as “the Lord of the Sky” (ʾɘgzīʾa samāy); “the Lord of the Universe” (ʾɘgzīʾa kwelū); “the Lord of the Earth” (ʾɘgzīʾa bəḥēr) (RIÉth 189 in vocalized Geʿez and RIÉth 190 in the South Arabian script). By contrast, in the sixth century CE, the Trinitarian faith appears to have become strongly rooted when one looks at RIÉth 191 (king Kālēb, around 500 CE); RIÉth 195 ([king Kālēb], around 530 CE); and RIÉth 192 (king Waʿzeb, in the years 540 or 550 CE). It is sufficient to quote here the beginning of the first inscription:

God is power and strength, God is powerful ⁽²⁾ in battle.⁶³ With the power of God and the grace of Jesus Christ, ⁽³⁾ son of God, the Victor in whom I believe, He who gave me a kingdom ⁽⁴⁾ of power with which I subjected my enemies and trampled the heads of those who hated me, he who watched ⁽⁵⁾ over me since my childhood and placed me on the throne of my forefathers, who has saved me. I have sought protection ⁽⁶⁾ from Him, Christ, so I succeed in all my endeavours and live in the One who pleases ⁽⁷⁾ my soul. With the help of the Trinity, that of the Father, the Son, and the Holy Spirit (RIÉth 191 / 1–7).

4.2. A New Place of Worship Called the mikrāb

The new religion had its own place of worship, an expression I shall return to shortly. In polytheistic inscriptions, places of worship were described by a whole series of terms, the most common being maḥram (mḥrm) ‘sanctuary’ and bayt (byt) ‘temple’. After 380 CE and until approximately 500 CE, the place of worship was systematically called mikrāb (mkrb). After 500 CE, two new terms appeared: bīʿat (bʿt) and qalīs (qls1), both meaning ‘church’, the first a loan from Syriac, bîʿotô ‘dome’ (from the word for ‘egg’), and the second from the Greek ekklêsia.

The term ‘place of worship’ must be understood as a generic name for all consecrated monuments and spaces where individual or collective religious rituals (oracular consultations, offerings, sacrifices, prayers, atonement) were performed at determined moments or at any time of the year. Many places of worship had other functions, especially for studying, teaching, or hosting travelers; some played the part of a banking institution for the faithful or the local economy. These secondary functions are difficult to pinpoint. In the case of the mikrāb, they are never explicitly mentioned in sources. They cannot even be confirmed by archaeological observation, because no mikrāb has yet been identified. The hypothesis suggesting that a building in Qanīʾ is a synagogue rests on meager evidence that does not appear to be decisive.⁶⁴

The vocalization of mkrb is certainly mikrāb. This can be deduced from attestations of the word in Yemen’s dialects (as noted by two nineteenth-century travelers) and in Geʿez. According to Eduard Glaser,⁶⁵ in eastern Yemen (Mashriq), the noun mikrāb was used (but also mawkab and muqāma) to designate a polytheistic temple. As for Ḥayyīm Ḥabshūsh, he noted that in Haram (in the Jawf), mikrab was the term used to describe the portico of an ancient temple.⁶⁶ Though the two travelers indeed recorded the same word, they differ on the length of the vowel /a/. The most likely vocalization is that given by Glaser, who had a robust philological background; moreover, Glaser took notes in the field, while Ḥabshūsh wrote from memory more than twenty years after his journey. The noun mikrāb is also attested in Geʿez under the form mək^wrāb, which designates a synagogue or the Temple of Jerusalem.⁶⁷

The meaning of the root KRB, to which the noun mkrb and other South Arabian words are related—in particular, the title of mkrb (traditionally vocalized as mukarrib) borne by rulers enjoying a dominant position in South Arabia—has been a matter of discussion for quite some time. That KRB expresses the notion of blessing⁶⁸ is a reasonable assumption, both in monotheistic texts and in earlier polytheistic written sources. Clearer attestations can be found in the greetings at the beginning of correspondence, some of which have survived as copies on wooden sticks. See as polytheistic examples YM 11738 = X TYA 15 / 1-2 or YM 11733 = X TYA 9 / 2:

...w-s2ymn (2) l-krbn-k

May the divine Chief (i.e., the god Aranyadaʿ of Nashshān) ⁽²⁾ bless you

or

...w-s2ymn l-krbn-kmw

May the divine Chief bless you

For a monotheist example, see X.SBS 141 = Mon.script.sab 6 / 3:

... w-rḥmnn ḏ-b-s1myn l-ykrbn (4) tḥrg-kmw b-nʿmtm w-wfym

May Raḥmānān, who is in the Sky, bless ⁽⁴⁾ your Lordship with good fortune and well-being⁶⁹

The noun mkrb can therefore mean ‘place of blessing’.

The root KRB of Sabaic is apparently related to the Hebrew and Arabic root BRK, which also expresses the notion of ‘blessing’. This is one of the most secure instances of a metathesis in a Semitic root. Sabaic is the only language where the two roots are attested at the same time, both the local root KRB and the root borrowed from the Jewish-Aramaic BRK in the times of monotheism.

Attestations of the noun mikrāb number ten. The mikrāb is on six occasions built by well-known figures, the king or the prince.⁷⁰ A text details that a mikrāb called Yaʿūq included an assembly room (ms3wd) and porticoes (ʾs1qf).⁷¹ A second document, which is unfortunately fragmentary, suggests that another mikrāb included a kneset, apparently another type of assembly room.⁷²

Of the five mikrāb whose names have come down to us, three of them bear a name borrowed from Hebrew or Judaeo-Aramaic. They are (once) Ṣwryʾl,⁷³ from Hebrew ṣūrīʾēl, ‘God is my rock’, the name of a person in Num. 3.35; and (twice) Brk (or Bryk), from Aramaic barīk, ‘blessed’.⁷⁴ The mikrāb are the only South Arabian buildings for which names of foreign origin are attested.

One of the mikrāb is located in a cemetery meant exclusively for Jews. The inscription of Ḥaṣī (220 km southeast of Ṣanʿāʾ, MAFRAY-Ḥaṣī 1, Fig. 13) mentions the transformation of four plots to create a cemetery only for Jews. It details that a fourth plot was added to the three plots and the well already conceded to the mikrāb Ṣūrīʾel. The mikrāb, which is entrusted to a custodian (ḥazzān), drawing its subsistence from the revenues of a well, owns landed estates.

The name mikrāb is not merely the transposition of one of the Greek terms used to name a synagogue, the proseuchê, literally ‘prayer’, or sunagogê, literally ‘meeting’. The mikrāb would therefore be an original institution and not just a copy of an institution of the Mediterranean Jewish Diaspora.

4.3. A New Social Entity Called ‘Israel’

Together with the new religion, a new social entity called ‘Israel’ appeared for the first time in South Arabia. The authors of three inscriptions mention “their commune Israel.”⁷⁵ One is Ḥimyarite and one is apparently of foreign origin. In the third (fragmentary) text, the author’s name is lost. In these inscriptions, the invocation of Israel seems to replace the old invocations of the commune of origin. Thus, one can hypothesize that the Jews—Jews of Judaean origin as well as converts (or proselytes) and perhaps ‘sympathizers’—were reunited in a new social entity called ‘the commune Israel’.

It is probable that this commune Israel was conceived as a way of unifying tribal society and replacing the old communes. However, as Jérémie Schiettecatte has pointed out to me, it is only attested in the capital’s cosmopolitan environment. In the provinces, local power was always held by princes, who never failed to mention the communes over which these princes exerted authority (communes which, indeed, appear to have still been in existence).

The new entity, whose name suggests it was based on religion, was not a simple copy of the ancient communes. It had a quasi-supernatural dimension since, in the blessing formula introducing a text, it appears between two names for God (CIH 543 = ẒM 772 A + B):

[May it bl]ess and be blessed, the name of Raḥmānān, who is in the Sky, Israel and ⁽²⁾ their god, the Lord of the Jews, who has helped their servant…

The name Israel is quite significant. It undoubtedly betrays the hope of a restoration of the historical Israel. One also notices that Israel is a name that can only come from Jews of Judaean origin, since this is how Judaean Jews designate themselves. Logically, in these invocations, the commune Israel is invoked before the king himself.

4.4. A New Monotheistic Religion Shared by All?

Having examined the main aspects of the new religion, how do we know we are speaking of a single religious creed and not of several?

At first glance, the variety of the names given to God suggests diversity rather than unity. It quickly appears, however, that these names are interchangeable, since two or more are often mentioned together.⁷⁶ One can thus find in the same text:

Raḥmānān and ‘Lord of the Sky’: ẒM 5 + 8 + 10; Ry 520; CIH 537 + RES 4919 = Louvre 121; Garb Antichità 9, d

Raḥmānān and ‘Lord of the Jews’: Ry 515; Ja 1028; CIH 543 = ẒM 772 A + B

Raḥmānān and ‘God (Īlān) master of the Sky and the Earth’: ẒM 2000

Raḥmānān and ‘God (Īlān) to whom the Sky and the Earth belong’: Ja 1028

Raḥmānān and ‘God (Aʾlāhān) who owns the Sky and the Earth’: Ry 508

The unity of this corpus is moreover founded on the fact that it presents notable differences not only with respect to the inscriptions that precede it, but also with respect to those that follow, i.e., Christian inscriptions of the period 530–560 CE. These Christian inscriptions can be distinguished by a new way of designating God, a new name for places of worship, and a new place in the inscription for invocations.

One still notices that the faithful of various tendencies visit the mikrāb. This building was intended for observant Jews, since one was located in the Jewish cemetery of Ḥaṣī. It is probable, however, that the mikrāb was also open to others, this conclusion deriving from the fact that kings and princes intended to build them everywhere.

Unfortunately, there is no doctrinal term that allows one to isolate a group of inscriptions and contrast it with another, apart from the fact that some royal inscriptions are more laconic than others, an observation to which I shall return. It is true that the corpus is too restricted to make this point imperative.

On these grounds, there is no reason to surmise that the inscriptions of the period 380–530 CE do not form a homogeneous group. In all likelihood, they refer to a single religion.

5.0. A Variety of Judaism

If one asks about the nature of this religion, there is no doubt that it is a form of Judaism. Among lexical, onomastic, and doctrinal indexes allowing one to place the new religion within the religious panorama of the Near East (polytheistic, Jewish, Christian, Manichaean, Gnostic, or Zoroastrian), many emphasize proximity with Judaism only; some point towards both Judaism and Christianity; but none suggest a link with Christianity only or with another type of religious worship.

5.1. Proofs of Judaism

The most decisive proofs of the proximity to Judaism are the four attestations of the name Israel (ys3rʾl) and the three attestations of the syntagm ‘Lord of the Jews’, a matter on which I wish to return. One can add to these the discovery of two texts in Hebrew: the already-mentioned Hebrew graffito in the monogram of Yehuda’s inscription and the list of priestly families in charge of the divine service in the Temple of Jerusalem (mishmarōt) (Fig. 14 ).⁷⁷

The ritual exclamations amen (ʾmn) and shalom (s1lwm) provide another argument in favour of Judaism. Amen (ʾmn) and salām (s1lm), however, can also be found in Christian inscriptions. It is therefore only the spelling s1lwm with the mater lectionis /w/ that securely points to Judaism.⁷⁸

Most of the lexical borrowings from Aramaic could originate from either Jewish-Aramaic or Syriac and Christian Palestinian Aramaic. Two loanwords, expressing the notions of ‘prayer’ (ṣlt) and ‘favour, (divine) grace’ (zkt), are particularly interesting because they are also found in the Qurʾān some two hundred years later with the meanings ‘prayer’ (in Arabic, ṣalāt) and ‘legal alms’ (in Arabic, zakāt), names of two of the five pillars of Islam.⁷⁹ This does not mean these Aramaic terms were borrowed by Ḥimyar and, from there, passed into Arabic.⁸⁰ Patterns of transmission were no doubt diverse. It is remarkable nevertheless that some Qurʾānic loan-words were already rooted in Yemen well before Islam.

The Ḥimyarite anthroponymy has three names that come from the Hebrew Bible. Among them, one, Yehuda (yhwdʾ, ywdh), is always Jewish,⁸¹ but two others, Joseph (Yūsuf, ys1wf or ys1f) and Isaac (Yiṣḥaq and Isḥāq, yṣḥq and ʾsḥq), can also be Christian. The spelling of Isaac varies by language: in Sabaic, it is yṣḥq, exactly like ancient Hebrew; but in pre-Islamic Arabic, like in Aramaic, it is ʾsḥq.⁸² The most conservative spelling, yṣḥq, is probably evidence of an affiliation with Judaism.

On this matter, one notices that the conservation of the initial /y/ (replaced by a vocalic glottal stop in Aramaic and Arabic) can also be seen in the spelling of the name Israel as ys3rʾl.

Mention should lastly be made of epigraphic texts proving people traveled between Ḥimyar and Palestine, and some Ḥimyarites expressed a strong bond with the Land of Israel. First of all, a passing reference should be made to the grave owned by the Ḥimyarites in a collective tomb at Bet Sheʿarim in the Galilee.⁸³ Another example is a funerary stele written in Aramaic, probably originating from a necropolis close to the Dead Sea, whose author is Yoseh son of Awfà, who

passed away in the city of Ṭafar (= Ẓafār) ⁽³⁾ in the Land of the Ḥimyarites, left ⁽⁴⁾ for the Land of Israel and was buried on the day ⁽⁵⁾ of the eve of the Sabbath, on the 29th ⁽⁶⁾ day of the month of tammûz, the first ⁽⁷⁾ year of the week [of years], equivalent ⁽⁸⁾ to the year [400] of the Temple’s destruction’ (Naveh-Epitaph of Yoseh = Naveh-Ṣuʿar 24).

Ḥimyar’s conversion to Judaism was not a simple parenthesis in time before its very brief conversion to Christianity and then to Islam. It left a durable mark on Yemen. A first proof of this is the importance and influence of Yemen’s Jewish community until modern times.⁸⁴ A second indication (obviously indirect) is provided by the works of the greatest of Yemeni scholars, al-Ḥasan al-Hamdānī, who lived in the tenth century CE: as opposed to what all of Arab literary production says, he expresses an astonishing religious neutrality when speaking of Yemen and of Arabia, as if he wanted to emphasize that in Yemeni history, Muḥammad and Islam were but one episode following many others.

5.2. A Non-Rabbinic Form of Judaism

If indeed inscriptions reveal that Ḥimyar converted to Judaism, it is relevant to ask what type of Judaism Yemenis were following. For quite some time, the prevailing opinion was that the various orientations of the Second Temple period (Sadducees, Pharisees, Essenes, Zealots), well-known thanks to Flavius Josephus, did not survive the destruction of the Temple in 70 CE. In recent decades, however, a hypothesis stressing that some older currents survived has become dominant; as a consequence, ‘rabbinization’ would not be an immediate consequence of the Second Temple’s destruction but a long process that concluded only at the very end of Late Antiquity or even in Islam’s early years.

If indeed the existence of several currents of Judaism after 70 CE is generally accepted, opinions differ strongly as to their number, their definition, and their names (rabbinic, scriptural, priestly, Hellenistic, synagogal, etc…). This is not surprising since they diverged on a whole series of central questions relating to Judaism’s history, beginning with the date and composition of the Torah and the origins of the synagogue.

Since I am not a specialist on these matters, I will not give a definite opinion on post-70 Judaism but shall restrict my scope to writing an inventory of characteristics Ḥimyar’s Judaism shared with such-and-such a current.

On at least one point of doctrine (the issue of resurrection after death), Ḥimyar’s Judaism seems to differ from that of rabbis. Five inscriptions conclude with petitions concerning the end of their authors’ lives. And yet none of them mention resurrection.

In one text, certain nobles, who are otherwise unknown and who are commemorating the construction of their palace in Ḥimyar’s capital, conclude their inscription with the following invocation (Garb Nuove icrizioni 4, Bayt al-Ashwal [Ẓafār]):

...b-(7)rdʾ rḥmnn bʿl s1myn l-ẖmr-(8)hmw qdmm w-ʿḏ(r)m ks3ḥ(m ʾ)mn

With the help of Raḥmānān, owner of the Sky, so that He may grant ⁽⁸⁾ a pure beginning and a pure end, amen

The authors ask God to guard their lives on Earth, particularly their end, but they ask for nothing in the afterlife, which leads to the thought that they do not believe in an existence after death. The same conclusion can be drawn from two other documents cited above. The first of these commemorates the construction of a mikrāb by a princely family of the region of Ṣanʿāʾ. The prince provides detailed reasons for his patronage (Ry 520, from the vicinity of Ṣanʿāʾ):

...l-ẖmr-hw w-ʾḥs1kt-(6)hw w-wld-hw rḥmnn ḥyy ḥyw ṣdqm w-(7)mwt mwt ṣdqm w-l-ẖmr-hw rḥmnn wld(8)m ṣlḥm s1bʾm l-s1m-rḥmnn

In order that Raḥmānān may grant him, as well as to his wi⁽⁶⁾fe and his children, to live a just life and to ⁽⁷⁾ die a just death, and that Raḥmānān may grant him virtuous childre⁽⁸⁾n in the service for the name of Raḥmānān⁸⁵

The second document’s author was a Jew called Yehuda Yakkuf, already mentioned, who appears to not have been from Ḥimyar. He commemorates the construction of a palace in the capital. In his invocations, Yehuda seeks to give details on the main traits of his God (Garb Bayt al-Ashwal 1 [Ẓafār]):

...b-rdʾ w-b-zkt mrʾ-hw ḏ-brʾ nfs1-hw mrʾ ḥyn w-mwtn mrʾ s1(3)myn w-ʾrḍn ḏ-brʾ klm

With the assistance and grace of his Lord who has created him, the Lord of life and death, the Lord of the S⁽³⁾ky and the Earth, who has created all⁸⁶

Once more, the afterlife is not mentioned. This is, no doubt, an argument from silence, but it cannot be dismissed since, in principle, the afterlife is a constant preoccupation of those who believe in it.

A third document is more ambiguous. It is a bilingual grave stele, of unknown provenance, written in Aramaic and Sabaic. The fact that the Jewish-Aramaic text is written first (before the one in Sabaic carved underneath) suggests that the stele comes from a Jewish necropolis of the Near East and not from Yemen.⁸⁷ The document is ambiguous, because the first text explicitly mentions resurrection, while the second one does not (Naveh-Epitaph of Leah):

The Aramaic text reads:

...nšmt-h l-ḥyy ʿwlm (3) w-tnwḥ w-tʿmwd l-gwrl ḥyym lqṣ (4) h-ymyn ʾmn w-ʾmn šlwm

May her soul (rest) for eternal life, ⁽³⁾ and it will rest and become [ready] for resurrection at the en⁽⁴⁾d of days. Amen and amen, shalom

The Sabaic text reads:

...l-nḥn-hw rḥmnn (7) ʾmn s1lwm

May Raḥmānān grant her rest. Amen, shalom

Among the various scenarios that one could contrive to explain this difference in formulation, the most likely is that the stonecutter was content to copy the standard formulae on hand or those provided by Leah’s family. This could mean that Ḥimyarite Jews did not believe in an afterlife (or were not in the habit of mentioning it in their grave inscriptions), while the Jews of the Levant did believe in it. We cannot dismiss that one of the two formulae was written or chosen by Leah’s family, but if one accepts such a hypothesis, nothing allows favouring one version over the other.

One must set aside the Aramaic grave stele in the name of Yoseh son of Awfà, which has already been mentioned (Naveh-Epitaph of Yoseh = Naveh-Ṣuʿar 24):

...ttnyḥ nfšh d-ywsh br (2) ʾwfy d-gz b-ṭfr mdynth (3) b-ʾrʿhwn d-ḥmyrʾy w-nfq (4) l-ʾrʿh d-yśrʾl

May the soul of Yoseh son ⁽²⁾ of Awfà, who passed away in the city of Ṭafar ⁽³⁾ in the Land of the Ḥimyarites and left ⁽⁴⁾ for the Land of Israel, rest in peace⁸⁸

The deceased passed away in the Land of the Ḥimyarites, yet nothing certifies that he is himself a Ḥimyarite. At most one notes that he bears an Arab patronym. Noteworthy, however, is the fact that no allusion is made to resurrection.

The fifth inscription, Ja 547 + 546 + 544 + 545 = Sadd Ma ʾrib 6 (Ma ʾrib, November 558 CE, dhu-muhlatān 668), mentioned above, also poses problems of interpretation. Dating from the reign of the Christian king Abraha, it can be considered Christian; in fact, a small cross is carved at the end of lines 10, 13, and 14. One suspects that the authors introduced themselves as Christians without really belonging to the faith. The crosses are very discreet and placed in such manner that they can be thought of as letters. Moreover, the invocations to God make no reference to the Holy Trinity (“In the name of Raḥmānān, Lord of the Sky and the Earth” [w-ʿl-s1m rḥmnn mrʾ s1my(n) w-ʾrḍ⁽⁸⁾n] and “In the name of Raḥmānān, the King” [ʿl-s1m rḥmnn mlkn], line 10). Finally, the authors come from a commune very strongly marked by Judaism. The text ends with the petition:

...l-ẖmr-hmw ḥywm ks3ḥm (14) w-mrḍytm l-rḥmnn (cross)

May [Raḥmānān] grant them a life of dignity ⁽¹⁴⁾ and the satisfaction of Raḥmānān

Once more, life after death is omitted. If the authors are Jews rather than Christians, this silence is not surprising. If the authors are true Christians, however, this could mean that the afterlife is not a topic that one mentions in inscriptions, whatever one’s religious orientation.⁸⁹

In short, all the texts available seem to show that the afterlife was not a matter of concern for Ḥimyarite Jews, who probably did not believe in the resurrection of the dead. According to the Mishnah, those who denied resurrection belong to the three groups excluded from the world to come: “[Here are] those who have no part in the world to come: the one who says there is no resurrection of the dead, [the one who says] that the Torah does not come from heaven, and the Epicurean” (m. Sanh. 10.1).⁹⁰ According to the rabbis, the most severe punishment in the world to come will be meted to:

Those belonging to sects (minim), apostates (meshummadim), traitors (mesorot), Epicureans, those who have denied [the divine origin of] the Torah, who have gone astray from the community’s ways, who have doubted the resurrection of the dead, who have sinned and have made the community (ha-rabbim) sin like Jeroboam, Ahab, and those who established a reign of terror over the land of the living and have extended their hand over the House [i.e., the Temple] (t. Sanh. 13.5).

This is therefore a first clue that Ḥimyar’s Judaism was not rabbinic. On this matter, it should be recalled that one of the main reasons Muḥammad, the founder of Islam, reproached his opponents was their disbelief in Judgment Day and in the resurrection. One supposes that these opponents were followers of the old religion of Makka; the example of the Jews of Ḥimyar, however, shows that his opponents were plausibly followers of other religious currents. After Arabia’s conversion to Islam, the change was immediate: in the oldest Islamic inscriptions in Arabic, the author frequently “demands paradise”.

A second point of doctrine that would distinguish Ḥimyar’s Judaism from that of the rabbis is the issue of ‘binitarianism’. This is more problematic, because it mainly rests on a single inscription of somewhat enigmatic meaning (CIH 543 = ẒM 772 A + B, Ẓafār):

[b]rk w-tbrk s1m rḥmnn ḏ-b-s1myn w-ys3rʾl w-(2)ʾlh-hmw rb-yhd ḏ-hrdʾ ʿbd-hmw s2hrm w-(3)ʾm-hw bdm w-ḥs2kt-hw s2ms1m w-ʾl(4)wd-hmy ḍmm w-ʾbs2ʿr w-mṣr(5)m...

[May it bl]ess and be blessed, the name of Raḥmānān, who is in the Sky, Israel and ⁽²⁾ their God,⁹¹ the Lord of the Jews, who has helped their servant Shahr^um,⁽³⁾ his mother Bdm, his wife Shams^um, their chil⁽⁴⁾dren [from them both] Ḍmm, ʾbs2ʿr and Mṣr⁽⁵⁾m...⁹²

The blessing in the introduction associates God (“Raḥmānān, who is in the Sky”) with Israel and the Lord of the Jews (two divine entities and Israel). It is legitimate to ask whether one finds here an instance of deviance denounced by the rabbis, the one that states there are “two powers in heaven”.⁹³

This blessing is, therefore, a call to question the relationship between Raḥmānān and the “Lord of the Jews”, who is found in two other invocations:

rb-hd b-mḥmd

Lord of the Jews, with the Praised One (Ja 1028 / 12, Ḥimà, Fig. 7 )⁹⁴

rb-hwd b-rḥmnn

Lord of the Jews, with Raḥmānān (Ry 515, Ḥimà)⁹⁵

One should first of all notice that the authors of these three texts, who use the title ‘Lord of the Jews’ (Rabb-Yahūd, written rb-yhd, rb-hd, and rb-hwd),⁹⁶ are proven or plausible Ḥimyarites, successively invoking the deity under two different names, as if dealing with two gods: the ‘Lord of the Jews’ and Raḥmānān or the ‘Lord of the Jews’ and Mḥmd. It is quite unlikely that a title like ‘Lord of the Jews’ would be used by Jews of Judaean ancestry, since they prefer the self-designation ‘Israel’ to Yahūd. The term ‘Jew’ is above all used by Gentiles; when Jews use it, it is in exchanges with people outside the community.

Incidentally, the term Mḥmd given to God is intriguing. It perhaps echoes a text invoking “Raḥmānān and Ḥmd-Rḥb” since, in the second name (unfortunately, also enigmatic), one finds the same root ḤMD.⁹⁷ The spelling of the deity’s name Mḥmd seems identical to that of Islam’s prophet. One cannot be sure this identity is significant because the vocalization of these two names may differ (for example, Maḥmūd and Muḥammad). We know that some reformers were nicknamed after the deity they claimed to worship; this could also have been the case with Muḥammad (whom the Qurʾān also calls Aḥmad).⁹⁸

A second observation is that the name ‘Lord of the Jews’ probably refers to the Jewish Adonai, reflected from the outside. The ‘Lord of the Jews’ would therefore be YHWH, the God of the Hebrew Bible, the God who dictated the Law to Moses.

If Raḥmānān is different from the ‘Lord of the Jews’, the first could be the God of those not considered fully Jewish, i.e., the ‘candidates’ who aspire to become Jews and the ‘sympathizers’.⁹⁹ Or, more doubtfully, the first could be the God of converts—or proselytes—as opposed to the God of Jews of Judaean origin.

To identify which current of ancient Judaism was practiced in Ḥimyar, we can once more draw attention to the fact that some traits are shared by various kinds of Judaism of the Mediterranean world, while others are not. Ḥimyar’s Judaism, like other forms of Judaism in the Mediterranean world, uses the local language and script but not Hebrew, which is strictly confined to symbolic texts.¹⁰⁰ By contrast, Ḥimyar lacks the menorah and other symbols found in the synagogues of Galilee and elsewhere in the Mediterranean world.¹⁰¹

Another singular trait of Ḥimyar’s Judaism is the famous list of mishmarot (or ‘guards’) of Bayt Ḥāḍir, mentioned above.¹⁰² It enumerates the twenty-four families of the priesthood in charge of the divine service in the Temple of Jerusalem following the Babylonian Exile, and it associates these family names with residences in Galilee. The fact that it originates from social backgrounds vouching for the Temple’s restoration is not doubtful; just as secure is the fact that its function was to legitimate the priestly pretentions of lineages then settled in Galilee. Yemen is the only country outside of Palestine where such a list was carved in stone. This is not banal, since the making of such a beautiful inscription was very expensive.

We can only hypothesize as to why such a document was copied and carved in Yemen. It may have had symbolic meaning, like the public statement of an indefectible attachment to the Temple, or the claim that only priests are legitimate to manage the community. It could have also been propaganda benefitting families of the priesthood who were effectively present in Yemen. The list of the Bayt Ḥāḍir mishmarot, which is not explicitly dated, certainly goes back to a time when the power stakes were high; it is therefore very likely that it is from the period 380–530 CE.

Finally, Ḥimyar’s Jews transcribe proper nouns according to Biblical Hebrew (and not according to later texts, notably in Aramaic). The impression is that one is dealing with a conservative form of Judaism, attached not only to the Temple but also to a literal interpretation of the Holy Scriptures. Since Ḥimyarite Jews, like the Sadducees (the priestly party at the end of the Second Temple period), apparently rejected belief in the resurrection, one has good grounds to characterize Ḥimyar’s Judaism as ‘priestly’, all the more so since nothing recalls rabbinic Judaism.

The case of Yathrib—the future al-Madīna—in the seventh century is entirely different. Haggai Mazuz has recently demonstrated in quite convincing fashion that the Judaism of the Yathrib Jews had much in common with that of the rabbis.¹⁰³ One could therefore surmise the existence of different orientations in South Arabia and the Peninsula’s northwest. Due to the difference in dates, however, this is not the most likely hypothesis.

It is plausible that in the fifth century CE the Judaism of the Ḥijāz was similar to that of Ḥimyar. First of all, Ḥimyar ruled the entire Peninsula. Moreover, it was the only Jewish state, a feature that makes it difficult to believe that Ḥimyarite Judaism was not the reference point and the model for the smaller Jewish communities in the region.

By the seventh century (c. 620 CE), Jewish power in the kingdom of Ḥimyar had long since vanished. The reference points for Judaism were now located in Mesopotamia and Galilee. The radiance of these centres was even at its zenith, since the Sāsānid Persians, having expelled the Byzantines from the Near East in 614 CE, were supported by the Jews as they consolidated their domination of the Levant. It was therefore logical that the small Jewish community of Yathrib was inspired by the teachings of the Sages of Mesopotamia and Galilee, among whom the rabbis already enjoyed a dominant position.

A favourable prejudice towards the priests nevertheless remained. It is the Arab-Muslim Tradition that suggests this, since it systematically ascribes a priestly ascendancy to the Jews occupying eminent positions, as no doubt the Jews of Yathrib themselves did.¹⁰⁴

6.0. The Extent of Ḥimyar’s Conversion to Judaism

If indeed Judaism was the primary religion of the kingdom of Ḥimyar for a duration of 150 years, can one say that this kingdom was Jewish or, more precisely, that it converted to Judaism, since the majority of its population was not of Judaean origin?

Clues indicating Ḥimyar’s adherence to Judaism abound. The leaders of the main princely families—or, at least, some of them—wrote inscriptions that included specific signs of conversion. Jews or rabbis belonged to the Ḥimyarite king’s entourage who received an embassy led by Theophilus the Indian in a year close to 344 CE and were later counselling the ruler who sentenced the priest Azqīr of Najrān to death (c. 470 CE).

Proofs of royal adherence to Judaism, however, are tenuous and fragile. If one puts aside the case of king Joseph, who rebelled against the Aksūmites in 522 CE, the only text explicitly indicating that the Ḥimyarite king was Jewish is the Ethiopian synaxarion, which summarizes the martyrdom of Azqīr, but this is not an original source, only a late abridgment.¹⁰⁵ One can also mention that the Arab-Muslim Tradition attributes to king Abīkarib the introduction of Judaism to Yemen.¹⁰⁶ Finally, the inscription (YM 1200) of king Maʿdīkarib Yunʿim (c. 480–485 CE) commemorates the construction of a mikrāb while using the term kneset (kns1t), which seems more Jewish than Christian in the fifth century CE.¹⁰⁷

The political and religious authorities’ gradual trend towards radicalism could also be interpreted as increasingly visible adherence to Judaism. Christian sources contain several allusions to a policy of anti-Christian repression that developed from 470 CE onwards. This policy resulted in, first of all, the trial against the priest Azqīr of Najrān, a man accused of “introducing a new religion [into] the country.” There are then various allusions to persecutions against the Christians of Najrān prior to those of 523 CE.¹⁰⁸ Finally, onomastics is a clue since only the last Jewish king bears a biblical name.

In fact, the element causing the most difficulties is the absence of royal inscriptions explicitly referring to Judaism.

6.1. Stages and Purposes of Conversion

Conversion to a new religion is not an isolated event but the result of a long process, generally extending over several generations. In some measure, one can even say it is a process with no end.

The last centuries of Late Antiquity provide several comparable examples of a foreign religion adopted by marginal groups, which then gained followers in the ruling classes and finally became the established religion. One can distinguish four stages:

1. The hindered diffusion of the new religion among marginal groups.
2. The acceptance by authorities of the new religion as licit.
3. The adherence of the ruling classes to the new religion, which becomes the dominant religion of reference.
4. The elevation of the new religion to official status, more or less exclusively. It is only when the third stage is reached that one can speak of conversion.

To better illuminate certain stages in Ḥimyar’s conversion to Judaism, I will compare them with those of Christianization in the Roman Empire, close to the conversion of Arabia in both time and space. One must take into account a huge difference in the nature of available sources, since in the case of Ḥimyar we have at hand only a very specific source, epigraphy.

The first stage is the opposition to the spread of the new religion among marginal groups. In the Roman Empire, this was the time of great persecutions, during which Roman authorities fought with determination against the spread of Christianity, all the more so since its followers frequently resorted to provocation.¹⁰⁹ It is moreover frequent that authorities see the introduction of a new religion (a potential source of social disorder) in a negative light.

The second stage corresponds to the acceptance of the new religion by the authorities, who now recognize it as legitimate. As a result, many people close to the circles of power adhere to the new religion. The religion ceases to be perceived as divisive and becomes one of the components of the religious landscape. Such an evolution in the Roman Empire occurred via Galerius’ Edict of Serdica (311 CE), later endorsed by Constantine and Licinius in June 313 CE as the Edict of Milan. This put an end to all anti-Christian measures still implemented in the Empire’s territory. The Empire was not yet Christian. It was not more so under the reign of Constantine, although he favoured the Christian faith and requested baptism, an event that took place on the eve of his death in the year 337 CE.

The third stage is reached when the new religion becomes the official state religion. In the Roman Empire, this occurred when Constantius II, the son of Constantine (337–361 CE), ascended to power. From then on, one can say that the Empire had become Christian, and therefore it had ‘converted’. In 341 CE, Constantius II, who was the first ruler brought up in the Christian faith, forbade sacrifices. In 346 CE, he ordered the closure of pagan temples, whose property was then handed over to the imperial treasury. At the same time, the emperors pampered the Christian clergy. One must not be misled by repressive decisions that were rarely implemented and were mainly political posturing to gain the support of religious authorities. Even though Constantius II was careful to reinforce the Church’s unity by firmly intervening in controversies on the nature of the Holy Trinity, the break with the past was not yet consummated: Constantius II was still pontifex maximus and fulfilled his duties as a leader of traditional cults when he travelled to Rome.

When the ‘conversion’ is taken for granted, it must be consolidated and made irreversible by making the new religion compulsory and exclusive. This is the last and fourth stage, whose ultimate goal is the population’s unanimous adherence to the new religion.

To convert the stubborn, the use of force and, as a last resort, massacre or expulsion, is quite common. Even in the case of Islam, which historically has rather acted as a protector of minorities,¹¹⁰ one today notices extreme movements promoting the total eradication of all other religions. What is also observable is that a religion never durably keeps a hegemonic position; in the most monolithic of societies, seeds of dissent swiftly sprout. Total conversion is therefore a goal that one tries to achieve but that is never completely reached.

In the Roman Empire, Christianity became the compulsory religion through the Edict of Thessalonica, enacted on 28 February 380 CE by Gratian (359–383 CE) and Theodosius (379–395 CE). At this point, Gratian resigned from his pagan office of pontifex maximus. As a result of this edict, later Christian emperors no longer favoured non-Christian beliefs and avowedly reduced religious diversity within the Roman Empire.

If one looks at the inscriptions only and not at the entire documentation, it is only during the fourth stage that a change in religion appears in a significant manner. Pierre-Louis Gatier has demonstrated this clearly with the example of the land around Antioch, one of the major centres of christianization in the East.¹¹¹ Much time was necessary for the new cult, together with its network of specialized constructions and its hierarchy, to organize in cities and then spread to the countryside.

Coinage also reflects new trends after a period of delay. Some coins from the reign of Constantius II include Christian symbols, but one must wait for the advent of Theodosius II (408–450 CE) to see coinage become truly ‘christianized’.

For individuals the adoption of a new religion is also a complex matter. It implies the dismissal and rejection—or at least the abandonment—of the previous religion, the religion of one’s parents and ancestors and many other people to whom one was attached through affection or solidarity. Changing one’s religion constitutes a break with the past, a break that could easily be considered a form of ingratitude or even treason.

This break is most often personal, involving close relatives or other kin. It can also be a spontaneous collective process, following the initiative of a prince, a chief, or a magnate.

The change of religion was certainly a response to the appearance of new moral and spiritual ideals. The idealization of justice led to the expectation of divine judgment, either individual, immediately following death, or collective, at the End of Time. If there were to be a judgment, a punishment or a reward would obviously be necessary: a paradise and a hell were thus needed, and why not even a purgatory for more complicated cases? To implement this judgment, the body of each person had to be resurrected, which raised the question of the state of the body after being resurrected: either as a glorious, eternally young and vigorous body that could be imagined with clothing or as a body completely identical to that of the deceased immediately prior to death.

The question of resurrection, judgment, and retribution is one of the greatest issues of Late Antiquity. Judging by the condemnations of the rabbis, it was a matter of debate. According to the Qurʾān and the Yathrib Document, this was the major controversy between the conservatives from Makka and the reformer Muḥammad.¹¹²

Change of religion has not only a spiritual dimension but also a political one.¹¹³ In short, those in charge of the matter are faced with two options. The first is to reform the religious practices of old, to make them better in order to answer new aspirations. The second option is to abandon these religious practices for an entirely new set of beliefs. This second option was the one frequently chosen for centuries. It had three advantages. First, by choosing a religion originating from outside, the reformer did not leave any space for accusations of partiality. Second, in a kingdom with diversity—and even more so in an empire with multiple traditions—the choice of a new religion could be a unifying factor. Finally, the abandonment of old sanctuaries allowed leaders to seize treasuries that had accumulated there. This factor was probably the most decisive one.

6.2. Ḥimyar’s ‘Discreet Conversion’

The first stage for Ḥimyar, the one of initial diffusion, remains almost completely elusive. At most, what is known is a modest inscription carved on a reused pillar, apparently earlier than the third century CE. This inscription might be Jewish.¹¹⁴

In the second stage, the first adherents of the new religion, which can only be Judaism, belonged to the ruling classes. One may suppose they were converted by Judaeans or by the descendants of Judaeans, people who had taken refuge in Arabia after the disastrous revolts of 70 and 135 CE and who would have quickly ascended to leading positions in the oases of the northern Ḥijāz.¹¹⁵

These first followers are known via five ‘monotheistic’ inscriptions carved before 380 CE:

Schiettecatte-Nāʿiṭ 9, around 320, under the reign of Yāsir^um Yuhanʿim II:

...w-l-ẖmr-hmw mrʾ s1(m)[yn …]

May it be granted to them by the Lord of the S[ky …]

The authors of this fragmentary text, which invokes the king, are probably the banū Hamdān, princes of Ḥāshid^um.

Ag 3 = Gorge du Haut-Buraʿ 3, c. 325–350 CE:

...w-ʾln bʿl s1myn l-yrdʾn-hmw

As for God (Īlān), owner of the Sky, may He assist them

The author is a client of the banū Haṣbāḥ, princes of Maḍḥà^m, and belongs to the Ḥimyarite lesser nobility.

Ag 2 = Gorge du Haut-Buraʿ 2, c. 355 (± 11 CE):

...b-rdʾ ʾln bʿl s1myn

With the assistance of God (Īlān), owner of the Sky

This text, written by the son of the author of the previous text (Ag 3), is also by someone belonging to the Ḥimyarite lesser nobility.

YM 1950 (vicinity of Ṣanʿāʾ), August 363 or 373 CE, under the reign of Tha ʾrān Yuhanʿim with one or several of his sons (Fig. 15 ):

[...w-mr](ʾ)-hmw bʿl s1myn l-s1(mʿ) ʿnt w-[...]

[ … As for their lor]d, the Owner of the Sky, may he answer the plea and [...]

...w-l-ys1mʿn bʿl-(s1)[myn ...]

May he answer, the Owner of the S[ky …]

The authors of this fragmentary text are the princes of Ḥumlān (i.e., the banū Bataʿ, whose name has disappeared in the gap).

CIH 152 + 151 (Najr, near ʿAmrān), of unknown date:

...w-ẖmr-hmw ʾln bry ʾʾḏnm w-mqymtm

May God (Īlān) grant them fullness of capacities and means¹¹⁶

The authors of this text originate from the banū Murāthid^um, princes of Bakīl^um dhu-ʿAmurān, but they do not mention this title here.

When one examines the entire group of inscriptions of this same period (320–380 CE), it becomes apparent that religious practices of old were undergoing a crisis. Simultaneously, visits to polytheistic temples inexorably declined.¹¹⁷

It was during this second stage that Constantius II sent an embassy to the king of the Ḥimyarites.¹¹⁸ The Ecclesiastical History by Philostorgius, which describes this embassy, tells us that Ḥimyarites are polytheists, but “quite a large number of Jews are living among them” (3.4). Because of the Jews, the embassy was unsuccessful in convincing the king to accept baptism. The passage is unfortunately incomplete: “Upon reaching the Sabaeans, Theophilus tried to persuade the ruler of their people to worship Christ and to dissociate themselves from pagan error. But the schemes typical of Jews [… ]” (3.4).

The third stage begins with Ḥimyar’s official conversion to Judaism. Between 380 and 384 CE, royal inscriptions reveal the rulers adhering to a new religion whose nature is not made explicit. In other inscriptions, however, all clues as to the nature of the monotheism practiced by the Ḥimyarites point towards Judaism. As there is no doubt whatsoever that only one religion had the right to express itself from this date, one can safely conclude that the kingdom of Ḥimyar officially converted to Judaism.

No royal inscription, however, explicitly proclaims this. No Jewish authority is ever invoked. The Bible is never quoted. One is therefore dealing with a conversion whose sole apparent effect in royal propaganda is the rejection of polytheism. This is why one can describe it as ‘discreet’. The fourth stage is not yet reflected in the documentation available.

Comparing Ḥimyarite inscriptions with those of the Roman Empire can perhaps fill in the gaps. In the epigraphy of the region of Antioch (the cradle of Christianity), it was only quite late, towards the end of the third stage and at the inception of fourth, that explicit references to Christianity appeared, as we have seen.¹¹⁹ Due to social inertia, time was necessary for religious innovations to be reflected by epigraphy (just as with coin emissions). If this observation also holds true for Ḥimyar, this would mean that Judaism was more profoundly rooted in Arabia than at a first glance, considering the small number of significant epigraphic texts.

6.3. Religious Minimalism in Ḥimyarite Royal Inscriptions

Monotheistic inscriptions of the period between 380 and 530 CE can be separated into two sets. The first one is made up of texts whose author is the ruler and that, as a result, can be considered official documents; none explicitly states that the ruler adheres to Judaism. The second set, all other texts, presents variable religious formulae. Some are as laconic as those of the royal inscriptions, while others clearly show the mark of Judaism; between the two groups, one finds the entire range of intermediary documents. The most disturbing trait of these inscriptions is the minimalism of official inscriptions.

The variability in religious formulae has been explained by the existence of several possible levels of adherence to Judaism. In theory, there must have been three main ones: the ‘sympathizers’, who shared with Jews some fundamental beliefs and some rules of social behaviour, but did not imagine themselves becoming Jews; ‘candidates’, who aspired to become members of the community; and, finally, the faithful who were Jews by birth or by conversion (the latter also called ‘proselytes’).

In the Roman world, sources confirm the existence of several levels. At Aphrodisias, the capital of the province of Caria in Asia Minor, three categories are mentioned in the lists recorded on a stele bearing two inscriptions, dating from the fourth or fifth century CE: sixty-eight Jews, three proselytes, and fifty-four God-fearers (theosebeis).¹²⁰ Latin literature (e.g., Juvenal) mentions metuentes;¹²¹ Greek writers, including Josephus and the author of Luke-Acts, refer to sebomenoi ton Theon¹²² and theosebeis.¹²³ These different terms, based on verbs meaning ‘to fear’, can be applied to people who ‘fear (God)’ and thus reject polytheism. It is difficult to say whether the God-fearers belong to the category of ‘sympathizers’ or ‘candidates’.

In Yemen, there were certainly observant Jews who respected the Law of Moses and were scrupulous about ritual purity, as shown by the existence of the cemetery reserved for them at Ḥaṣī. One can suppose that these Jews were in part Ḥimyarite converts (or proselytes) and in part foreigners settled in the kingdom of Ḥimyar, even if it is not always easy to distinguish these categories from one another.

There were also people who were inclined towards Judaism. We are not in a position to know whether they were about to convert to Judaism or whether they formed a stable group unwilling to go beyond simple respect for the ‘natural’ morality of the Noahide laws and certain rules of life, since strict observance of the Law (particularly the weekly day of rest and food prohibitions) were hardly compatible with traditional social life. These hypothetical sympathizers had a unique role model in Abraham, the first convert, well before the revelation of the Torah.

Just as in the Roman world, it is possible that these Ḥimyarite sympathizers or candidates may have been called ‘God-fearers’. The notion of ‘fear of God’ is indeed found in an inscription (Ry 534+ Rayda 1), with ṣbs1, a loan from Greek sebas, ‘reverential fear’, as indicated by the meeting of the consonants ṣ and s1 in the same root, which is perfectly irregular in Semitic phonetics:

...w-l-ẖmr-hm ʾln mrʾ s1myn w-ʾrḍn (4) ṣbs1 s1m-hw

And so that God (Īlān), Lord of the Sky and the Earth, may grant them ⁽⁴⁾ fear of his name¹²⁴

Since this inscription uses a Greek term, it surely reflects a notion Mediterranean in origin. It is not unthinkable that a second inscription (Ry 520, in the vicinity of Ṣanʿāʾ) refers to the fear of God in the expression:

...wld(8)m ṣlḥm s1bʾm l-s1m-rḥmnn

virtuous children,⁽⁸⁾ in the service of the name of Raḥmānān¹²⁵

The difficulty lies here in the meaning of the word s1bʾ. This word has been at first rendered as ‘fighter, militant’, because ʾs1bʾ was usually translated as ‘warriors’, but it is surely established today that ʾs1bʾ is the plural of s1bʾy and refers to the ‘Sabaeans’. The meaning of the verb s1bʾ ‘to go on an expedition’ could point to the idea of ‘being on a mission, in the service of’. Another interpretation is possible, however; one could see in s1bʾ another Sabaic transcription of the Greek sebas. No doubt the transcription of the Greek sigma by the Sabaic letter s1 was an irregular occurrence, yet it is attested: ‘Kaisar, Caesar’ is rendered by Qys1r in MB 2004 I-123, while the Arabic regularly transcribes Qyṣr with an emphatic letter.

It is noteworthy that the notions of ‘fear (of God)’ (taqwà) and of ‘God-fearers’ (muttaqūn) are found not only in the Qurʾān,¹²⁶ but also in the Yathrib Document,¹²⁷ which I will speak of later.

The two degrees of adherence to Judaism could have given birth to two series of religious rites, some open to all (as part of the official religion), and the others meant solely for Jews, as I have previously suggested.¹²⁸

I would now like to explore another explanation for the minimalism of official inscriptions. These were not attempting to give an exact and faithful picture of the religious situation. They were political propaganda in the service of the ruling power. They are therefore to be interpreted in political terms.

One can easily admit that the main preoccupation of the Ḥimyarite ruler was his throne’s stability. He therefore needed to be backed by a large base of supporters likely to provide him with troops and other forms of assistance. However, the establishment of another religion, aiming to unite the populace and reduce potential dissidence among followers of other beliefs, was initially a source of division.

The ruler undoubtedly had the active support of the Jewish party and its sympathizers. This party, during the new religion’s establishment, was probably a minority, even among the ruling classes. To counter opposition, it was therefore necessary for conversions to increase or, at least, for groups as large as possible to pledge allegiance, even if they did not adopt the new religion.

Late Antiquity provides many examples of religion being used as a tool in political life and international relations. It was therefore logical that religion be used for the formation of political alliances. Since the main fissure placed the backers of ancient rites against supporters of monotheism, one can suppose that the kings of Ḥimyar sought to create a federation of all monotheistic religious currents who would submit to them.

Such a hypothesis is not as gratuitous as it first appears. It is confirmed for at least one reign, that of the Jewish king Joseph, when he seized power and repressed the revolt of the pro-Byzantine Najrān Christians. As this was a period when tensions were exacerbated, the support provided to Joseph by Christians is particularly significant.

Syriac hagiographers celebrating Najrān’s victims mention first of all two Christians, the first Ḥimyarite and the second from al-Ḥīra, who assisted king Joseph and acted as emissaries during the violent takeover:

He <and his followers> fought the Ethiopians (kwšyʾ) who were in Ẓafār (ṭyfr), in the church that the Ethiopians had built there. When he saw that he was no match for their army in war, he sent them a Levite priest from Tiberias (ṭybʾryʾ), together with a man of Nagrān (nygrn) whose name was ʿAbd Allāh (ʿbdʾlh), son of Mālik (mlk), who was considered a Christian in name, and another man whose name was Kônb (kwnb), son of Mawhûbâ (mwhwbʾ), from Ḥirtā of Nuʿmān (ḥrtʾ d-nʿmn), who was also a Christian. He sent with them pledges written to the Ethiopians [saying]: “No harm will befall you if you come forth to me of your own volition and if you surrender to me the city of Ẓafār,” and he promised with oaths that he would send them alive to the king of the Ethiopians. When they received the written oaths, they came out to meet him—three hundred men with the Ethiopian archpresbyter, whose name was Abābawt. This Jew welcomed them, he treated them kindly and distributed them among his chieftains, telling them: “May each of you kill the Ethiopian who is in your home.” On this same night, all were slain. At daybreak, all their corpses were discovered thrown upon one another. He immediately sent men to the city of Ẓafār, who burnt the church where the Ethiopians had gathered, two hundred men. Thus [the number of] all the Ethiopians killed, at the beginning or thereafter, reached five hundred clerics and laymen.¹²⁹

Mālik son of ʿAbd Allāh from Najrān (and probably also Kônb [Kalb?] son of Mawhûbâ from al-Ḥīra) is described as ‘Christian in name’ (b-šmʾ krysṭynʾ).¹³⁰ This expression means that the inhabitants of Najrān, for whom the hagiographer is the spokesperson, do not consider him a real Christian. This same expression is used to speak of the Nestorian Christians of the Gulf, who in the seventh century CE rejected the authority of the catholicos: krsṭynʾ d-šmʾ.¹³¹

A third Christian called Gaḥsanâ saved the life of Joseph during a previous Ethiopian invasion of the land of Ḥimyar. This deed is mentioned by Maḥyâ (mʾḥyʾ), the “indiscreet and perverted” servant of Ḥārit (the revolt’s chief) in a speech to king Joseph:

“But as for you [Joseph], all the Ḥimyarites know the shame Gaḥ[sanâ] the merchant of Ḥirtâ of Nuʿmān inflicted upon you, since he saved you from death [at the hands] of the Ethiopians [at the time of battle].” This very same Gaḥsanâ was present in the land of the Ḥimyarites at the moment when Ethiopians had gone out and had persecuted the Ḥimyarites. They had surrounded him [and wanted to kill him]. But this Gaḥsanâ stood up [and] swore by the Holy Gospel that he [Joseph] was a Christian. It was in this way that this Jew escaped death. Now, after having ascended the throne and persecuted the Christians, he [Joseph] sent part of the loot from Christians to the same Gaḥsanâ in Ḥirtâ of Nuʿmān, with a letter and a blessing. This why all the Christians hated this Gaḥsanâ, and it was because of him that the blessed one reviled the king, as has been written above’.¹³²

Incidentally, this text indicates that Joseph was not killed during a massacre of Jews because someone guaranteed he was Christian. One can easily suppose that Joseph himself, when he was interrogated and threatened with execution, pretended he was Christian. This observation raises the question of whether Joseph, before his coup, was not officially Christian. Indeed, one must remember that according to the Greek Martyrdom of Arethas, it is the Negus himself who placed him on the throne.¹³³

Syriac hagiographers do not explicitly say that Gaḥsanâ was a Ḥimyarite. This nevertheless seems plausible. In any case, hagiographies twice mention that some Christians supported Joseph.

The same sources suggest that Joseph kept good relations with Nestorian Christian authorities. In the Greek Martyrdom of Arethas, the king is supposed to have declared to Arethas and his companions: “Would you therefore be superior to the Romans called Nestorians, who are in our land and teach us this...”¹³⁴ The Christians of Najrān belonged to two very distinct and occasionally antagonistic communities. There was, first of all, a community maintaining close links with anti-Chalcedonian Byzantines of North Syria,¹³⁵ who are called today ‘Miaphysites’ (or Monophysites). There was also a community attached to the Church of the East (or Nestorian Church) of Sāsānid Persia, whose tutors were in al-Ḥīra in the lower valley of the Euphrates.¹³⁶ It is not to be doubted that it was the Nestorians who backed the Jewish party and the Miaphysites who opposed to it.

According to the Greek and Syriac sources relating the wars between Byzantium and Sāsānid Persia in the sixth century CE, many Arabs participated in the conflict, either in the Byzantine camp or that of the Persians. Sources call them ‘Arabs of the Romans’ and ‘Arabs of the Persians’.¹³⁷ One could likewise state there were ‘Christians of the Romans’ and ‘Christians of the Persians’.

If one accepts that in the Jewish kingdom of Ḥimyar there was effectively a coalition of Jews and Nestorian Christians, the minimalism of Ḥimyarite royal inscriptions can possibly be explained by this coalition’s existence: when publicly communicating, the ruler took into account the political and religious leanings of his allies. It is not known what this political-religious coalition, uniting those who believed in one God, was called. This interpretation of the minimalism of royal inscriptions is all the more likely since similar or comparable practices are noted in Abraha’s Christian kingdom and in the first Muslim State, each time during the years following a new religion’s establishment, as we shall see.¹³⁸

The minimalism of Ḥimyarite royal inscriptions is therefore not an argument to be used to deny Ḥimyar’s conversion to Judaism. It signals only that the ruler was never capable of publicly stating his adherence to Judaism, no doubt because his power rested on a coalition of groups who were not all Jewish. I suggest describing Ḥimyar’s conversion to Judaism as ‘discreet’ because it was never explicitly translated into royal propaganda.

The religious policy of the kings of Ḥimyar, understood here as a conversion to Judaism, was previously interpreted in a different manner. In 1984, A. F. L. Beeston, from a corpus of texts notably more restricted in quantity, supposed that the Ḥimyarite rulers adhered to a peculiar form of monotheism independent of both Christianity and Judaism.¹³⁹ To name this belief, Beeston reemployed the term ‘Raḥmānism’, coined by D.S. Margoliouth for the monotheism of the (so-called, according to him) Jews of Yathrib.

A. F. L. Beeston admitted that a few inscriptions were indeed Jewish, but this was not the case for every inscription that included monotheistic formulae. This did not prevent him from finding in ‘Raḥmānism’ elements of “Jewish inspiration”¹⁴⁰ confirmed by Arab traditions relating to the conversion of the Ḥimyarites in the days of king Abū Karib.¹⁴¹ This ‘Raḥmānist’ hypothesis had the advantage of providing a plausible origin for the ḥanīf of the Arab-Muslim Tradition, postulating that pre-Islamic Arabs could have chosen monotheism without adhering to one of the great established religions. The few Islamic scholars who paid attention to Beeston hypothesis (like Andrew Rippin) were unconvinced.¹⁴²

7.0. A Few Similar Examples

In order to better convince scholars of the plausibility of two of the hypotheses formulated in this paper (the existence of a political coalition around the Jewish ruling elite, based on the belief in one God; a certain form of tribal restructuring on a religious basis, outlined by the creation of the commune Israel), I will show that these have parallels in both Arabia and Ethiopia around the same period.

7.1. Minimalist Official or Public Expression

The minimalism of official (or royal) Ḥimyarite inscriptions undoubtedly reflects the beliefs that other members of the coalition backing the ruling elite shared with them. In other words, in its political propaganda, the Jewish ruling class did not impose its own particular convictions, but only those that created a consensus within the coalition.

7.1.1. Religious Invocations in Abraha’s Inscriptions

In Abraha’s kingdom, just like in Ḥimyar’s Jewish state, the formulation of religious invocations diverges from what scholars would at first expect.¹⁴³

First of all, it is necessary to summarize the historical context. Following the defeat and death of king Joseph (525–530 CE), the Negus placed on Ḥimyar’s throne a Ḥimyarite Christian. The only inscription of this Ḥimyarite proclaims the perfect Trinitarian orthodoxy of the new regime (around 530 CE) wih the following introduction (Ist 7608 bis + Wellcome A 103664):

[b-s1]m w-s2r[ḥ rḥmnn w-bn-hw krs3ts3 ḡlbn w-mn]fs1 qds1

[In the na]me and with the safe[guarding of Raḥmānān, of His son Christ the Victor, and of the Ho]ly Spirit

And again in the conclusion:

[…]b-s1m rḥmnn w-bn-hw krs3ts3 ḡlbn [w-mnfs1 qds1]

[...] In the name of Raḥmānān, of His son Christ the Victor, [and of the Holy Spirit]

Shortly after, Abraha, general of the Aksūmite occupation troops, seized power by force, perhaps in 532 CE or in the following years. For fifteen years his power was threatened by two punitive expeditions of the Negus of Aksūm and by internal dissent. Only in 547–548 CE did his rule stabilize. Between 548 and 560 CE, he had seven inscriptions made, three containing an opening invocation to God. These three invocations are:

— CIH 541:

b-ẖyl w-[r]dʾ w-rḥ(2)mt rḥmnn w-ms1(3)ḥ-hw w-rḥ [q]ds1

With the power, assistance, and merci⁽²⁾fulness of Raḥmānān, of his ⁽³⁾ Messiah, and of the Spirit of Holiness

— DAI GDN 2002/20 = Sadd Ma ʾrib 4:

b-ẖyl w-n(ṣr) (2) w-rdʾ rḥmnn (3) mrʾ s1myn (4) w-ms1ḥ-h(w)

With the power, the support, ⁽²⁾ and help of Raḥmānān,⁽³⁾ Lord of the Sky,⁽⁴⁾ and of His Messiah

— Ry 506 = Murayghān 1:

b-ẖyl rḥmnn w-ms1ḥ-hw

With the power of Raḥmānān and of His Messiah

These inscriptions can be distinguished from the first one by a significant change: the word ‘son’, designating the second person of the Holy Trinity, is replaced by that of ‘Messiah’. This alteration means that the second person of the Trinity is not of divine essence but a human being who received divine anointment. Moreover, one shall note the absence of any reference to the Holy Spirit in two of these three texts of Abraha.

Under the reign of Abraha, Ḥimyar, now an unquestionably Christian state, was certainly leaning towards Miaphysitism (or Monophysitism) and maybe even towards Julianism, its most extreme version, both of which firmly defended the divine nature of the second person of the Holy Trinity.

To explain why dogmatic formulae in Abraha’s inscriptions diverge from Miaphysitism, the most plausible explanation is that this inflection is a result of internal policy. One can reckon that a significant part of the population, despite the massacres, remained attached to Judaism and did not accept that God had a son or was constituted of several beings. Abraha displayed a minimalist Christology and sought the support of not only all Christian currents (particularly Najrān Nestorians and those of the Gulf under his control), but also some Jews and perhaps even other sects. His propaganda shows that he was concerned with obtaining or preserving the adherence of Jews who were ready to take a step in his direction by accepting Jesus as Messiah, even though there is no proven instance of Jews paying allegiance to him or in his service.

The minimalism of Abraha’s dogmatic formulations was a response to the same necessities as those of Jewish royal inscriptions. It reflects the fragility of a regime in the third stage of conversion.

7.1.2. Religious Invocations in the Inscriptions of ʿĒzānā at Aksūm

The kingdom of Aksūm, not located in the Arabian Peninsula but claiming sovereignty over South Arabia (as shown by the titles of its rulers, which includes ‘king of Ḥimyar’), is an interesting case of discrepancy between two official doctrines on display, while the king’s true religious beliefs are not known precisely.

King ʿĒzānā, under whose reign Aksūm converted to Christianity (apparently around the early 360s CE), expressed his new beliefs in a very allusive manner in his inscriptions in the local script and language:

[By] the power of the Lord of the Sky (ʾəgzīʾa samāy), who in the Sky and on the Earth is victorious for me, ʿĒ⁽²⁾[zā]nā son of ʾƎlē ʿAmīda the man of Ḥalən, king of Aksūm, Ḥəmē⁽³⁾r, Raydān, Saba ʾ, Salḥēn, Ṣəyāmō, Bəgā,⁽⁴⁾ [of] Kāsū, king of kings, son of ʾƎle ʿAmīda, who is not vanquished by the enemy.⁽⁵⁾ [By the pow]er of the Lord of the Sky, who has granted me [kingship], the Lord of the Universe in whom I [believe],⁽⁶⁾ [I] the king who is not vanquished by the enemy, may no enemy place himself in front of me and may no enemy ⁽⁷⁾ follow me. By the power of the Lord of the Universe, I waged war on the Noba… (RIÉth 189, in vocalized Geʿez).¹⁴⁴

Further in the same text, God is also called “the Lord of the Earth” (ʾǝgzīʾa bəḥēr), which later became the name for the One God. Nothing in this text reveals ʿĒzānā’s true religious beliefs.

ʿĒzānā, however, in a contemporary Greek inscription,¹⁴⁵ announces his adherence to perfect Nicene orthodoxy:

In the faith in God and the power of the Father, the Son and the Holy Spirit, to the one who preserved for me the kingdom through faith in His Son Jesus Christ, to the one who came to my assistance and still does, I, Azana, king of the Axômites, of the Homêrites, of Reeidan, of Sabaeans, of S[il]êel, Kasô, of the Bedja and of Tiamô, bisi Alêne, son of Elle-Amida, and servant of Christ, I thank the Lord my God, and I cannot fully say his graces, for my mouth and spirit cannot [express] all the graces He did for me: He has given me strength and power; He has granted me a great name by His Son in whom I believe; and He has made me guide of all my kingdom because of my faith in Christ, by His will and by the power of Christ. It is He who has guided me, I believe in Him, and He made Himself my guide. I came out to fight the Nôba (RIÉth 271, in Greek).

Greek, a language inaccessible to the local population, was nonetheless understood by passing foreign travelers. In this language, the king was presenting himself as an exemplary Christian, watchful of the injunctions of Byzantium’s political and ecclesiastical authorities. In Geʿez, however, it was local politics that took precedence: the king chose formulations that non-Christians could adhere to.

The minimalism of public expression noticed in the earliest Christian inscriptions in local tongues truly seems to be of identical nature to that of Ḥimyarite Jewish royal inscriptions. Once more, inscriptions reflect the regime’s fragility, a characteristic typical of the third stage of conversion.

7.1.3. The Minimalism of the Oldest Muslim Inscriptions

A last parallel is also quite enlightening. As indicated several times above, Ḥimyar’s religious history is known only through inscriptions. It is therefore interesting to examine what a study of Islam’s formation through the exclusive prism of inscriptions would produce for scholars.¹⁴⁶

Islamic inscriptions of the two first centuries of the Hijra, whose number has spectacularly increased during the last decades, reveal several unexpected traits. The most significant is that the name Muḥammad does not appear during the first sixty-six years of the Hijra, and that there is no mention of either an apostle (rasūl) or a prophet (nabī).¹⁴⁷ During this early phase, the very repetitive formulae implore God’s forgiveness and clemency and ask for paradise. Qurʾānic formulations or quotes, which would securely characterize these texts as Muslim, only gradually appear.¹⁴⁸

For this period, only two inscriptions of a semi-official character are available to us. Both commemorate the construction of dams in the ruler’s name, and both date to the reign of Muʿāwiya b. Abī Sufyān (661–680 CE). The one found in al-Ṭāʾif, which dates from 58 AH (677–678 CE), soberly indicates that works were carried out “with the permission of God” and asks God to “grant pardon to the servant of God Muʿāwiya, Pr⁽⁵⁾ince of the Believers, strengthen and assist him”.¹⁴⁹ The second text, from al-Madīna, is undated but a little more explicit.¹⁵⁰ In it, one finds at the beginning “In the name of God, al-Raḥmān, the merciful”, and, a little further on “O God, bless it for him, Lord ⁽⁵⁾, master of the Skies and of the Earth”.

A third text cannot be quoted here on account of its Christian environment (as attested by the cross on the top left), its provenance (Ḥammām Gāder or Jādir, in the Yarmūk valley, at the foot of the Jawlān), and finally its language (Greek); it nonetheless dates to 5 December 662 CE, under the rule of ʿAbdalla Maauia Amêra ⁽²⁾ al-Moumenêna’.¹⁵¹

In addition to the already published texts of the first generations of Muslims, it is possible to add about a hundred others showing the same traits, all from North West of Saudi Arabia and the Najrān valley. Now, we know that at Najrān the Christian and Jewish communities (the former from the Church of the East) were both still political forces in the ninth century CE. The inscriptions therefore date to a period when Najrān was enjoying genuine religious pluralism. This raises the question of whether their authors are all Muslims.

The earliest Islamic inscriptions consist of a small core of texts (all after 70 AH and therefore quite late), including explicit adherence to a well-identified and exclusive religion and many more documents that could have been written by adherents of many different religions. One has the impression that public religious expression was as neutral as possible to avoid upsetting a union of all religious currents sharing the belief in one God and Judgment Day.

Until the accession to the throne of ʿAbd al-Malik b. Marwān (685–705 CE), no inscription of a caliph is known, while private texts are plentiful. This strange absence can perhaps be explained in the same way. The ruler did not order any inscriptions because the political situation was unsettled and official phraseology still uncertain. As soon as the regime stabilized, however, change took place immediately, illustrated, for instance, by the inscription on the Dome of the Rock in Jerusalem, from 72 AH (691–692 CE), which solemnly proclaims an official doctrine in breach of all other monotheisms.¹⁵² This is, in fact, the transition from the third stage to the fourth, with peculiarities characteristic of Islam.

These interpretations recall the hypothesis of Fred Donner who, using a completely different approach based on a critical examination of the Qurʾān and the Yathrib Document (or the Constitution of Medina), postulates that Muḥammad founded at the very time of his arrival in Yathrib a ‘Community of Believers’ (muʾmin), a federation of  the disciples he taught (the muslim) and the adherents of other religions who shared the belief in one God and the ideal of a virtuous life.¹⁵³ Fred Donner quotes Q 5.65–66:

[65] Had the People of the Book (ahl al-kitāb) believed and been pious, we would have erased for them their evil deeds and would have made them enter the Garden of Delight [on Judgment Day]. [66] Had they abided by the Torah, the Gospel, and what has been given to them from above from their Lord, they would have eaten what is above them and beneath their feet. Among them there is a provident/moderate community walking a straight path (ummatun muqtaṣidatun). [But for] many of them, what evil they do!

He concludes: “This passage implies strongly that those individuals among the ahl al-Kitāb who embrace right belief and right action will be welcomed among the Believers”.¹⁵⁴

Donner accepts that the interpretation of texts with a theological purpose perhaps does not permit drawing conclusions on social and communal organization.¹⁵⁵ Our intention is not to reopen this complex case, but only to show how another approach can lead to a similar result, which evidently strengthens its plausibility.

7.2. The Tribal Coalition founded by Muḥammad

I have interpreted the minimalism of official (or royal) Ḥimyarite inscriptions as the formulation of beliefs shared by a political coalition uniting the Jews (adepts and sympathizers) and other groups adhering to various monotheistic beliefs. The existence of such a coalition around Ḥimyarite Jews rests on only a few tenuous clues. By contrast, we are in possession of the founding text by which Muḥammad created a coalition of this type upon his arrival at the oasis of Yathrib in 622 CE.¹⁵⁶

The oasis at that time comprised five main tribes (or, more accurately, clans), all of relatively modest size, and many secondary groups. Among the principal ones, three were Jewish (Qurayẓa, al-Naḍīr, Qaynuqaʿ). They were settled on the best land in the centre of the oasis and could be considered local aristocracy. The other two, al-Aws and al-Khazraj, were in principle allies of the Jewish clans, but wished to be emancipated from them and to redistribute the wealth, leading to the invitation of Muḥammad.

Upon his arrival in Yathrib, Muḥammad no doubt had the support of al-Aws and al-Khazraj, but this backing, which was not even unanimous, was evidently insufficient to control the oasis and organize its defence in case of attack from the people of Makka. He therefore decided to conclude an alliance with other groups residing in Yathrib. The founding text of this alliance, which calls itself a ṣaḥīfa, ‘document’, was fully transmitted to posterity via several channels. Two versions are available (with variants that are of little significance) and have been attentively and thoroughly studied, notably by Michael Lecker.¹⁵⁷ Almost all scholars consider the Ṣaḥīfat Yathrib (the ‘Yathrib Document’) to be authentic, despite apparent modifications.

The document includes two sections, which Michael Lecker calls “the treaty with the muʾmin” and “the treaty with the Jews”. The relevant groups are mentioned in the first clause: “This is an agreement written upon the initiative of Muḥammad the prophet between the muʾmin and the muslim originating from Quraysh and from Yathrib and those who follow them, are linked to them, and fight with them.”¹⁵⁸ The entire set of parties, called “the people of this treaty” (clause 45),¹⁵⁹ are a tribal coalition (a group linked by rules of solidarity) called umma, as detailed by clause 2: “They form a single umma, to the exclusion of others.”¹⁶⁰

The treaty with the muʾmin explicitly mentions that the alliance’s ideological basis is “to believe in God and in the end of the world” or “in the day of resurrection.”¹⁶¹ It is therefore not Muḥammad’s teachings that are the point of reference but only two fundamental (or, rather, minimalist) principles. What follows in the text provides some clarifications: participants originating from Quraysh are also called muhājir (clause 3); the muʾmin are twice described as muttaqūn, ‘God-fearers’ (clauses 14 and 22). Finally, the adherence to the umma implies allegiance to Muḥammad (mentioned in clauses 1, 26, 52, and 63) and the renunciation of previous tribal solidarities.

The treaty with the Jews (Yahūd) explicitly mentions seven groups (clauses 28–34). One of the clauses indicates that each party keeps its own rules (strictly linked to religion): “The Jews have their law and the muslim theirs” (clause 28).¹⁶²

The concrete meaning of all these terms (muʾmin, muslim, muhājir, and umma) has been the focus of several studies whose conclusions very much differ. I shall limit myself to a few remarks.

Muḥammad founded a new tribal coalition whose perimeter went beyond that of the followers of his teachings. This coalition, based on adherence to a few fundamental religious principles, is designated by the term of umma. The meaning of umma is contentious. The Qurʾān gives this noun a mainly religious dimension, but it can also be found in a profane context, with the meaning of ‘tribe’.¹⁶³ It is not yet a proper noun but would become one.

The core of the new coalition appears to be made up of muslim, whom the muʾmin and the Jews have joined. The first are apparently disciples following Muḥammad’s teachings, and the muʾmin are those who abide only by a few general principles. If this is true, it is not surprising that the latter are described as ‘God-fearers’, like the sympathizers and candidates of Judaism in the Mediterranean world.¹⁶⁴

Concerning the Jews, one can suppose that these, just like the muʾmin, believe in the end of the world and in the day of resurrection, while those of Yemen two centuries earlier did not believe in these.

What one sees in the Yathrib Document is therefore an example of a tribal coalition uniting the followers of a new religious orientation and their allies, similar to what one notices in the kingdom of Ḥimyar in the time of Joseph and Abraha.

Tribal restructuring taking place in al-Madīna, with the creation of the umma, also recalls the ‘commune Israel’ of Ḥimyar’s Jewish kings.¹⁶⁵ In both cases, the adherents to a new religious orientation break loose of their old tribe to enter into a new structure.

7.3. Tribal Restructuring on a Religious Basis: The Example of al-Ḥīra

I have surmised on several occasions that tribal coalitions and restructuring were based on adherence to such-and-such a religion or to common beliefs. To illustrate this process, in addition to the example of Yathrib, we have that of al-Ḥīra, a city on the lower reaches of the Euphrates, where a vassal of the Sāsānid kings resided in the sixth century CE. Al-Ḥīra’s population consisted of three tribal groups:

1. Tanūkh, the original tribe, whose kings had dominated the middle Euphrates valley since the end of the third century CE.
2. Al-Aḥlāf, a conglomerate of tribes like those found in various cities, particularly in Najrān.
3. Al-ʿIbād, a second conglomerate uniting Christians of different tribal origins.

Gustav Rothstein collected all meaningful sources on the topic.¹⁶⁶ It is unnecessary to mention them again in this paper. Even though the origin of ʿIbād is not a matter of general consensus, scholars admit that ʿIbād is a new tribal formation. What one sees in this city after the arrival of groups rallying around the king is a process of tribal reorganization, with two tribes uniting foreign groups, the Christians (or some Christians) on one side and various other people on the other.

The patterns of such a trend, which is not exceptional at all, are not usually explained. In the specific case of Ṣanʿāʾ, however, they were examined by the Yemeni Muslim scholar al-Ḥasan al-Hamdānī (d. 945 CE), who explained which tribe newcomers were related to:

Ṣanʿāʾ is divided between the banū Shihāb and the Abnāʾ [the descendants of the Persians who settled in Yemen between 575 and 630 CE]. The man who originates from Nizār [Arabs of the north] is attached to the Abnāʾ; but the people of the land, and the man who originates from Qaḥṭān, are attached to the banū Shihāb.¹⁶⁷

Al-Hamdānī explains that the tribal structuring is fixed; new groups are attached to pre-existing tribes. The tribes themselves, however, are of relatively recent formation: the Abnāʾ appeared following the Sāsānid occupation in the late sixth and early seventh centuries CE, as al-Hamdānī seems to know; as for the banū Shihāb, one knows nothing of them before the tenth century.

Another enlightening example is the city of Ṣaʿda in the tenth century CE, which is also mentioned in al-Hamdānī. Its population is made up of two groups, Ukayl and Yarsum.¹⁶⁸ The banū Ukayl are the chiefs (sayyid) of the main sub-fraction of the northern Khawlān, the large tribal confederation of northern Yemen, of which Ṣaʿda is the centre;¹⁶⁹ we are therefore dealing with the local population.

Yarsum is a very different case: it is the commune of Sabaean princes who conquered Khawlān and annexed it to Saba ʾ in the second century CE. One could therefore suppose that these princes (the banū Sukhaym) settled in Ṣaʿda, a garrison composed of men of their commune.¹⁷⁰ Yarsum is therefore at the origin of a non-native group from the vicinity of Ṣanʿāʾ. Al-Hamdānī provides us with the detail of its composition in the tenth century CE:

Yarsum, a group made up of thirteen houses who have taken the name of Yarsum from (tarassamat ʿalà) Yarsum b. Kathīr and from Yarsum the first (Yarsum al-ūlà). ʿAbd al-Malik b. Yaghnam gives details: “There are three houses at the origin of Yarsum. These are al-ʿUmayrāt [uncertain vocalization], from the offspring of dhū Sukhaym, and two other houses of the old Yarsum of Ḥimyar. Within Yarsum, there is a house of the Āl Dhuwād [uncertain vocalization] who belong to al-Abnāʾ, a house of Hamdān, [more precisely of] Ḥāshid, a house of al-Khawlī, a house of the banū Hilāl, a house of Kināna, a house of the banū Ḥanīfa, a house of the people of Najrān, a house of Madhḥij, a house of Quḥāfa belonging to Khathʿam, a house of ʿUwayr.”¹⁷¹

One can observe the same process as in Ṣanʿāʾ: the population is divided into two groups, the locals and those who come from elsewhere. Moreover, it is the second group that attracts newcomers: people from Yemen (Yarsum, Hamdān, and the Abnāʾ), Najrān (Najrān and Madhḥij), and both Western (Kināna) and Central (Ḥanīfa) Arabia.

The examples of Ṣanʿāʾ and Ṣaʿda illustrate the way in which tribal affiliations undergo a process of reformation. The case of al-Ḥīra shows that, as for Ḥimyar, tribal restructuration can be based on religious affiliation.

7.4. Concerning the ḥanīf

In this paper, I have not dealt with the ḥanīf, who have at times been considered the heirs of Ḥimyarite monotheism inspired by Judaism. The reason for this resides in the fact that these ḥanīf are probably historical ghosts and that the link with Ḥimyarite Judaism rests on an obsolete interpretation of the earliest known Jewish Ḥimyarite inscriptions.

In Muḥammad’s Arabia, there were supposedly believers called ḥanīf with no specific religious affiliation. Texts of the Arab-Muslim Tradition mention a number of them, particularly in Makka and al-Madīna. They were living at the same time as Muḥammad or slightly earlier. These ḥanīf allegedly adhered to a form of monotheism identified with the religion of Abraham.

Many of today’s scholars, however, doubt that the ḥanīfiyya ever existed. Instead, they are seen as the result of a late reconstruction based on scattered data and the enigmatic Qurʾānic term ḥanīf. In the Qurʾān, seven out of twelve occurrences of ḥanīf refer to Abraham; the others describe the exemplary behaviour that Muḥammad and true believers must adopt. Tradition may have invented the hanīfiyya to give more consistency to Abraham’s religion and to respond to Muslim fears that their ancestors were damned. On the other hand, Uri Rubin has noted that in the Tradition, the hanīf are often Muḥammad’s opponents, which is incompatible with the hypothesis postulating that they were a late invention.

In any case, in 1984, A. F. L. Beeston connected the religion of the Ḥimyarites, who wrote monotheist inscriptions without indicating adherence to a precise creed, to that of the ḥanīfs. Recognizing the same reservations towards foreign beliefs, Beeston surmised that the ḥanīfiyya in the days of Muḥammad was a relic of a religious current that developed in the kingdom of Ḥimyar, 250 years earlier. To designate this religious current, he employed the term ‘Raḥmānist’.¹⁷²

Against Beeston, Andrew Rippin emphasized that no tangible evidence supported his hypothesis.¹⁷³ One can add that this thesis implies that, after 380 CE, there were two different religious currents in the kingdom of Ḥimyar: Judaism and Raḥmānism. Even now it is difficult, not to say impossible, to distinguish between those two currents. Nothing permits identification of the ḥanīfiyya with one of the religions attested in Arabia prior to Islam.

8.0. Conclusion

At the end of this inquiry, it appears that all power structures behave in the same way after radical religious reform. For one or two generations they advance while remaining undercover, revealing only minimal signs of reform so as not to antagonize potential opponents. This can be seen in the Roman Empire, in Ḥimyar during the Jewish and Christian (Abraha) periods, in Aksūm, and in the Islamic Empire. Official inscriptions do not refer to the new religion but only to a few general principles.

The most apparent of these principles are the uniqueness of God, a God who rules the Sky and the Earth, a God who is the author of Creation, and, finally, a God who metes out reward and punishment at the End of Time. They distinguish between those who have rejected pagan religious practices and those who have preserved them, even under a reformed manner close to monotheism. In Arabia, it is easier to recognize that the same God is worshipped because this God bears the same name whatever the religious beliefs adopted: al-Raḥmān or Raḥmānān.

The powers in place advance surreptitiously because they are a minority and are faced with forms of opposition. They therefore need to gain allies and obtain the support of new groups. In the context of religious reform, it is logical that alliances translate into religious terms.

These alliances were (or could be) formalized with a genuine written contract. This can be seen at al-Madīna, where the contract creating a tribal confederation that included muslim and other groups has survived until the present day.

These tribal coalitions were not meant to last. They were typical of a transitory period and were intended to facilitate the strengthening of the newly-founded power structure. Once stability was ensured, they were no longer useful. Nevertheless, in Ḥimyar’s Jewish kingdom, this stabilization process did not occur. One may suppose that the regime failed to produce a confederation of new supporters sufficient for it to display its true nature. Hence Ḥimyar’s conversion to Judaism, which seems to be proven, was not reflected in royal propaganda and remained ‘discreet’. For lack of a stabilizing process, the regime collapsed quite rapidly: around 500 CE, Ḥimyar became a tributary of the Christian kingdom of Aksūm. It is hard to doubt that internal divisions provoked this humiliating outcome.

As a final note, I shall return to the initial question of the rabbinization of Ḥimyarite Judaism. Yemenite Judaism was rabbinic from the early days of Islam, but pre-Islamic sources suggest that in the fourth and fifth centuries CE this was not the case. In Late Antiquity, the situation would rather have been similar to that observed in the Mediterranean world but perhaps with a stronger attachment to the priesthood. The rabbinization of Yemenite Jews thus took place at a date later than 520 CE. This date is difficult to pinpoint precisely, but it could be close to the time of Islam’s formation.

9.0. Addendum

As this contribution was being finalized, Mrs Sarah Rijziger, an independent scholar carrying out epigraphic investigations in Yemen, sent me a photograph of an inscription she discovered in Naʿḍ, 35 km southeast of Ṣanʿā.ʾ¹⁷⁴ The text is particularly interesting, since it mentions for the first time the expression ‘God of Israel’:

1 [… … …]mr w-Yws3f w-ʿb(d)ʾln w-Y(ḥ) =

2 [… … …] w-hqs2bn w-hs2q(r)n byt-h =

3 [mw … … …]. bʿ(l) S1myn ʾlh Ys3r =

4 [ʾl … … …]..d mlkn w-b-rd(ʾ) mrʾ-(h) =

5 [mw … … … s1](b)ʿy w-ẖms1 mʾtm (flower) (or: … ʾr](b)ʿy w-ẖms1 mʾtm)

1 [… … …]mr, Joseph, ʿAbdʾīlān and Yḥ =

2 [… have …, …], built anew and completed [their] palace

3 [… … …].. owner of the Sky, God of Isra =

4 [el … … …]…. of the king, and with the aid of [their] lord

5 [… … … seven]ty and five hundred (or: … for]ty and five hundred)

(SR-Naʿḍ 9)

The authors of this text, which dates from 54[.] or 57[.], i.e., 430–440 or 460–470 CE, are all Jewish: they bear, respectively, a name that is certainly biblical (Joseph) and another that is also perhaps related to ancient Israel (yḥ[…], Yoḥannan) and, finally, another which is monotheistic and theophoric (ʿAbdʾīlān). It is plausible that these people, who built a palace in a small town in the countryside, are princes or local lords.

The name that these people give to God is incomplete: “[…] owner of the Sky.” One can reconstruct it in three different ways, depending on which known texts one uses for extrapolation:

ʾln bʿl s1myn ‘God (Īlān), owner of the Sky’;

rḥmnnn bʿl s1myn ‘Raḥmānān, owner of the Sky’;

or perhaps

ʾlhn bʿl s1myn ‘God (Ilāhān), owner of the Sky’ (which is attested only in the more elaborate formula ʾlhn bʿl s1myn w-ʾrḍn ‘God, owner of the Sky and the Earth)’.

The expression ‘God of Israel’ is an apposition to the name of God, providing an element of clarification. It is not impossible that ‘Israel’ here refers to the commune Israel discussed above. One cannot exclude, however, the possibility that one is dealing here with something totally different, an identification of the God of the text’s authors with the God of historical Israel.

Bibliography

Aḥmad, ʿAẓīmuddīn, ed., Die auf Südarabien bezüglichen Angaben Našwān’s im Šams al-ʿulūm (Leiden: Brill, 1916).

al-Akwaʿ al-Ḥiwālī: see Ḥiwālī.

Alexander, Scott C., ‘Fear’, in Encyclopaedia of the Qurʾān, ed. by Jane McAuliffe, 5 vols. (Leiden: Brill, 2001), II, 194–98.

ʿAṭā, Sālim Muḥammad, and Muḥammad ʿAlī Muʿawwaḍ, eds., al-Istidhkār al-Jāmiʿ li-Madhāhib Fuqahāʾ al-Amṣār, by Ibn ʿAbd al-Barr (Beirut: Dār al-kutub al-ʿilmiyya, 2000).

Bausi, Alessandro, ‘Il Gadla ’Azqir’, Adamantius 23 (2017): 341–80.

Beaucamp, Joëlle, Françoise Briquel-Chatonnet, and Christian Julien Robin, ‘La persécution des chrétiens de Nagrān et la chronologie ḥimyarite’, Aram 11 (2000): 15–83, https://dx.doi.org/10.2143/aram.11.1.504451.

Beaucamp, Joëlle, Françoise Briquel-Chatonnet, and Christian Julien Robin, eds., Juifs et chrétiens en Arabie aux Ve et VIe siècles: Regards croisés sur les sources (Paris: Association des amis du Centre d’histoire et civilisation de Byzance, 2010).

Beeston, Alfred F. L., ‘Himyarite Monotheism’, in Studies in the History of Arabia II: Pre-Islamic Arabia, ed. by Abdelgadir Abdalla, Sami al-Sakkar, and Richard Mortel (Riyadh: King Saud University Press, 1984), 149–54.

———, ‘The Religions of Pre-Islamic Yemen’, in L’Arabie du Sud, histoire et civilisation I: Le peuple yéménite et ses racines, ed. by Joseph Chelhod (Paris: Maisonneuve et Larose, 1984), 259–69.

Bernand, Étienne, Abraham J. Drewes, and Roger Schneider, Recueil des inscriptions de l’Éthiopie des périodes pré-axoumite et axoumite, 3 vols. (Paris: de Boccard, 1991–2000).

Bowersock, Glen W., Martyrdom and Rome (Cambridge: Cambridge University Press, 1995), http://dx.doi.org/10.1017/cbo9780511518546.

Caskel, Werner, Ǧamharat an-nasab: Das genealogische Werk des Hišām ibn Muḥammad al-Kalbī, 2 vols. (Leiden: Brill, 1966).

Chabot, J. B., ed., Synodicon orientale, ou, Recueil de synodes nestoriens (Imprimerie nationale, 1902).

Chaniotis, Angelos, ‘The Jews of Aphrodisias: New Evidence and Old Problems’, Scripta Classica Israelica 21 (2002): 209–42.

Conti Rossini, Carlo, ‘Un documento sul cristianesimo nello Iemen ai tempi del re Šarāḥbīl Yakkuf’, Rendiconti della Reale Accademia dei Lincei 14 (1910): 747–50.

Crone, Patricia, ‘Pagan Arabs as God-Fearers’, in The Qurʾānic Pagans and Related Matters: Collected Studies in Three Volumes, Volume 1, ed. by Hanna Siurua (Leiden: Brill, 2016), 315–39, https://dx.doi.org/10.1163/9789004319288.

Darles, Christian, Christian Julien Robin, Jérémie Schiettecatte, and Ghassan El Masri, ‘Contribution à une meilleure compréhension de l’histoire de la digue de Ma’rib au Yémen’, in Regards croisés d’Orient et d’Occident: Les barrages dans l’Antiquité tardive, ed. by François Baratte, Christian Julien Robin, and Elsa Rocca (Paris: de Boccard, 2014), 9–70.

Degen, Rainer, and Walter W. Müller, ‘Ein hebräisch-sabäische Bilinguis aus Bait al-Ašwāl’, in Neue Ephemeris für semitische Epigraphik, ed. by Rainer Degen, Walter W. Müller, and Wolfgang Röllig, 3 vols. (Wiesbaden, 1972–1978), II, 117–23.

Detoraki, Marina, ed., and Joëlle Beaucamp, trans., Le martyre de saint Aréthas et de ses compagnons (BHG 166) (Paris: Association des amis du Centre d’histoire et civilisation de Byzance, 2007).

Donner, Fred McGraw, ‘From Believers to Muslims: Confessional Self-Identity in the Early Islamic Community’, Al-Abhath 50–51 (2002): 9–53.

———, Muhammad and the Believers: At the Origins of Islam (Cambridge, MA: The Belknap Press of Harvard University Press, 2010), http://dx.doi.org/10.2307/j.ctvjf9w3h.

Drewes, Abraham J., ‘The Meaning of Sabaean MKRB, Facts and Fiction’, Semitica 51 (2003): 93–125.

Drewes, Abraham J., and Jacques Ryckmans, Les inscriptions sudarabes sur bois dans la collection de l’Oosters Instituut conservée dans la bibliothèque universitaire de Leiden, ed. by Peter Stein and Harry Stroomer (Wiesbaden: Harrassowitz, 2016).

Eraqi Klorman, Bat-Zion, The Jews of Yemen in the Nineteenth Century: A Portrait of a Messianic Community (Leiden: Brill, 1993).

Gajda, Iwona, 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: de Boccard, 2009).

Gatier, Pierre-Louis, ‘La christianisation de la Syrie: L’exemple de l’Antiochène’, Topoi 12 (2013): 61–96.

Gimaret, Daniel, Les noms divins en Islam: Exégèse lexicographique et théologique (Paris: Éditions du Cerf, 1988).

Glaser, Eduard, Mittheilungen über einige aus meiner Sammlung stammende sabäische Inschriften, nebst einer Erklärung in Sachen der D. H. Müllerschen Ausgabe der Geographie al Hamdânî’s (Prague, 1886).

Goitein, Shelomoh D., ed., Travels in Yemen: An Account of Joseph Halévy’s Journey to Najran in the Year 1870, Written in Sanʿani Arabic by His Guide Hayyim Habshush (Jerusalem: Hebrew University Press, 1941).

Gorea, Maria, ‘Les classes sacerdotales (mišmarôt) de l’inscription juive de Bayt Ḥāḍir (Yémen)’, in Le Judaïsme de l’Arabie antique: Actes du Colloque de Jérusalem (février 2006), ed. by Christian Julien Robin (Turnhout: Brepols, 2015), 297–329, http://dx.doi.org/10.1484/m.jaoc-eb.4.00035.

Greenfield, Jonas C., ‘From ʾlh Rḥmn to Al-Raḥmān: The Source of a Divine Epithet’, in Judaism and Islam: Boundaries, Communication, and Interaction: Essays in Honor of William M. Brinner, ed. by Benjamin H. Hary, John L. Hayes, and Fred Astren (Leiden: Brill, 2000), 381–93.

Grohmann, Adolf, Arabic Inscriptions: Expédition Philby-Ryckmans-Lippens en Arabie (Leuven: University of Leuven, 1962).

Guidi, Ignazio, ed., ‘La lettera di Simeone vescovo di Bêth-Arśâm sopra i martiri omeriti’, Atti della Reale Accademia dei Lincei 278 (1881): 471–515.

Hadas-Lebel, Mireille, Rome, la Judée et les Juifs (Paris: Picard, 2009).

al-Hamdānī : see Ḥiwālī, Löfgren, Müller.

Hārūn, ʿAbd al-Salām Muḥammad, ed., Jamharat Ansāb al-ʿArab by Ibn Ḥazm (Cairo: Dār al-Maʿārif, 1977).

Henry, René, ed., Bibliothèque, tome I: Codices 1–83 by Photius (Paris: Les Belles Lettres, 2003).

Hillers, Delbert R., and Eleonora Cussini, eds., Palmyrene Aramaic Texts (Baltimore, MD: Johns Hopkins University Press, 1995).

Hirschfeld, Yizhar, and Giora Solar, ‘The Roman Thermae at Ḥammat Gader: Preliminary Report of Three Seasons of Excavations’, Israel Exploration Journal 31 (1981): 197–219.

Ḥiwālī, Muḥammad b. ʿAlī al-Akwaʿ al-, ed., Kitāb al-Iklīl, al-Juzʾ al-Awwal by al-Hamdānī (Cairo: al-Sunna al-muḥammadiyya, 1963).

Hoftijzer, Jacob, and Karel Jongeling, Dictionary of the North-West Semitic Inscriptions, 2 vols. (Leiden: Brill, 1995), http://dx.doi.org/10.1163/9789004294240.

Horovitz, Josef, Jewish Proper Names and Derivatives in the Koran (Hildesheim: Georg Olms, 1964).

Houtsma, Martijn Theodoor, ed., The History (Taʾrīkh) by Ibn Wāḍiḥ al-Yaʿqūbī, 2 vols. (1883; reprint Leiden: Brill, 2018).

Hoyland, Robert, Seeing Islam as Others Saw It: A Survey and Evaluation of Christian, Jewish, and Zoroastrian Writings on Early Islam (Princeton, NJ: Darwin Press, 1997), http://dx.doi.org/10.31826/9781463239602.

———, ‘Insider and Outsider Sources: Historiographical Reflections on Late Antique Arabia’, in Inside and Out: Interactions between Rome and the Peoples on the Arabian and Egyptian Frontiers in Late Antiquity, ed. by H. F. Dijkstra and Greg Fisher (Leuven: Peeters, 2014), 265–78.

Imbert, Frédéric, ‘L’Islam des pierres: L’expression de la foi dans les graffiti arabes des premiers siècles’, Revue des mondes musulmans et de la Méditerranée 129 (2011): 57–78, http://dx.doi.org/10.4000/remmm.7067.

Jastrow, Marcus, A Dictionary of the Targumim, the Talmud Babli and Yerushalmi, and the Midrashic Literature (New York: Judaica Press, 1996), http://dx.doi.org/10.31826/9781463232344.

Klein-Franke, A., ‘Tombstones Bearing Hebrew Inscriptions in Aden’, Arabian Archaeology and Epigraphy 16 (2005): 161–82, http://dx.doi.org/10.1111/j.1600-0471.2005.00256.x.

Kozah, Mario, ‘Ishoʿyahb of Adiabene’s Letters to the Qataris’, in An Anthology of Syriac Writers from Qatar in the Seventh Century, ed. by Mario Kozah, Abu-Husayn Abdulrahim, Saif Shaheen Al-Murikhi, and Haya Al Thani (Piscataway, NJ: Gorgias Press, 2015).

Lecker, Michael, The “Constitution of Medina”: Muḥammad’s First Legal Document (Princeton, NJ: Darwin Press, 2004).

Leslau, Wolf, Comparative Dictionary of Geʻez (Classical Ethiopic): Geʻez-English, English-Geʻez, with an Index of the Semitic Roots (Wiesbaden: Harrassowitz, 1987).

Leszynsky, Rudolf, Die Juden in Arabien zur Zeit Mohammeds (Berlin: Mayer & Müller, 1910).

Löfgren, Oscar, ed., Al-Iklīl, Erstes Buch, by al-Hamdānī, in der Rezension von Muḥammed bin Nas̆wān bin Sa’īd al-Ḥimyarī (Uppsala: Almqvist & Wiksells, 1954).

Maraqten, Mohammed, Altsüdarabische Texte auf Holzstäbchen: Epigraphische und kulturhistorische Untersuchungen (Beirut: Orient-Institut, 2014).

Mazuz, Haggai, The Religious and Spiritual Life of the Jews of Medina (Leiden: Brill, 2014), http://dx.doi.org/10.1163/9789004266094.

Moberg, Axel, ed., The Book of the Ḥimyarites: Fragments of a Hitherto Unknown Syriac Work (Lund: C. W. K. Gleerup, 1924), http://dx.doi.org/10.31826/9781463223366.

Müller, David Heinrich, ed., Al-Hamdâni’s Geographie der arabischen Halbinsel, nach den Handschriften von Berlin, Constantinopel, London, Paris und Strassburg zum ersten Male herausgegeben (Leiden: Brill, 1968).

Muṣṭafà, Kamāl, ed., al-Ḥūr al-ʿayn, li-l-Amīr ʿallāmat al-Yaman Abū Saʿīd Nashwān al-Ḥimyarī (Cairo: Maṭbaʿat al-Saʿāda, 1942).

Nallino, Carlo Alfonso, ‘Ebrei e cristiani nell’Arabia preislamica’, in Raccolta di scritti editi e inediti, ed. by Maria Nallino, 6 vols. (Rome: Istituto per l’Oriente, 1939-1948), III, 87–156.

Nashwān al-Ḥimyarī: see Aḥmad, Muṣṭafà.

Noman, Khaldon, ‘A Study of South Arabian Inscriptions from the Region of Dhamār (Yemen)’ (PhD diss., Università di Pisa, 2012).

Prioletta, Alessia, ‘Le pilier de Tan‘im: La plus ancienne inscription juive du Yémen? Une approche paléographique’, in Le Judaïsme de l’Arabie antique: Actes du Colloque de Jérusalem (Février 2006), ed. by Christian Julien Robin (Turnhout: Brepols, 2015), 331–58, http://dx.doi.org/10.1484/m.jaoc-eb.4.00036.

Prioletta, Alessia, and Mounir Arbach, ‘Ḥimyar en Arabie déserte au Ve siècle de l’ère chrétienne: Une nouvelle inscription historique du site de Maʾsal, Arabie saoudite’, Comptes rendus des séances de l’Académie des Inscriptions et Belles-Lettres 160 (2016): 917–54.

Rāshid, Saʿd b. ʿAbd al-ʿAzíz al-, Dirāsāt fī ʾl-Āthār al-Islāmiyya al-Mubakkira bi-ʾl-Madīna al-Munawwara (Riyadh: Muʾassasat al-Ḥuzaymī, 2000).

Reynolds, Joyce Maire, and Robert F. Tannenbaum, Jews and God-Fearers at Aphrodisias: Greek Inscriptions with Commentary: Texts from the Excavations at Aphrodisias Conducted by Kenan T. Erim (Cambridge Philological Society, 1987).

Rippin, Andrew, ‘Rḥmnn and the Ḥanīfs’, in Islamic Studies Presented to Charles J. Adams, ed. by Wael B. Hallaq and Donald P. Little (Leiden: Brill, 1991), 153–68.

Robin, Christian Julien, ‘Arabie méridionale: L’État et les aromates’, in Profumi d’Arabia: Atti del Convegno, ed. by Alessandra Avanzini (Rome: L’Erma’ di Bretschneider, 1997), 37–56.

———, ‘La mosquée al-ʿAbbās et l’histoire du Yémen’, in De l’or du sultan à la lumière d’Allah: La mosquée al-ʿAbbās à Asnāf (Yémen), ed. by Solange Ory (Damascus: Institut français d’études arabes de Damas, 1999), 15–40.

———, ‘Le judaïsme de Ḥimyar’, Arabia 1 (2003): 97–172.

———, ‘Ḥimyar et Israël’, Comptes rendus des séances de l’Académie des Inscriptions et Belles-Lettres 148 (2004): 831–908, http://dx.doi.org/10.3406/crai.2004.22750.

———, ‘La réforme de l’écriture arabe à l’époque du califat médinois’, Mélanges de l’Université Saint-Joseph 59 (2006): 319–64.

———, ‘Joseph, dernier roi de Ḥimyar (de 522 à 525, ou une des années suivantes)’, Jerusalem Studies in Arabic and Islam 34 (2008): 1–124.

———, ‘Les Arabes des “Romains”, des Perses et de Ḥimyar (IIIe–VIe s.è. chr.)’, Semitica et Classica 1 (2008): 167–202, http://dx.doi.org/10.1484/j.sec.1.100252.

———, ‘Nagrān vers l’époque du massacre: Notes sur l’histoire politique, économique et institutionnelle et sur l’introduction du christianisme (avec un réexamen du Martyre d’Azqīr)’, in Juifs et chrétiens en Arabie aux Ve et VIe siècles: Regards croisés sur les sources, ed. by Joëlle Beaucamp, Françoise Briquel-Chatonnet, and Christian Julien Robin (Paris: Association des amis du Centre d’histoire et civilisation de Byzance, 2010), 39–106.

———, ‘Les signes de la prophétie en Arabie à l’époque de Muḥammad (fin du VIe et début du VIIe siècle de l’ère chrétienne)’, in La raison des signes: Présages, rites, destin dans les sociétés de la Méditerranée ancienne, ed. by Stella Georgoudi, Renée Koch Piettre, and Francis Schmidt (Leiden: Brill, 2012), 433–76, http://dx.doi.org/10.1163/9789004210912_017.

———, ‘Sabaʾ et la Khawlān du Nord (Khawlān Gudādān): L’organisation et la gestion des conquêtes par les royaumes d’Arabie méridionale’, in Arabian and Islamic Studies: A Collection of Papers in Honour of Mikhail Borisovich Piotrovskij on the Occasion of his 70th Birthday, ed. by Alexander V. Sedov (Moscow: Gosudarstvennyj Muzej Vostoka, 2014), 156–203.

———, ‘Le roi ḥimyarite Thaʾrān Yuhanʿim (avant 325–c. 375): Stabilisation politique et réforme religieuse’, Jerusalem Studies in Arabic and Islam 41 (2014): 1–96.

———, ‘Quel Judaïsme en Arabie?’, in Le Judaïsme de l’Arabie antique: Actes du Colloque de Jérusalem (Février 2006), ed. by Christian Julien Robin (Turnhout: Brepols, 2015), 15–294, http://dx.doi.org/10.1484/m.jaoc-eb.5.108285.

———, ‘Ḥimyar, Aksūm, and Arabia Deserta in Late Antiquity: The Epigraphic Evidence’, in Arabs and Empires before Islam, ed. by Greg Fisher (Oxford: Oxford University Press, 2015), 127–71, http://doi.org/10.1093/acprof:oso/9780199654529.003.0004.

———, ‘L’arrivée du christianisme en Éthiopie: La ‘conversion’ de l’Éthiopie’, in Saints fondateurs du christianisme éthiopien: Frumentius, Garimā, Takla-Hāymānot et Ēwosṭātēwos, trans. by Gérard Colin (Paris: Les Belles Lettres, 2017), xxii–xliii.

———, Christian Robin, ‘al-ʾIlāh et Allāh : le nom de Dieu chez les Arabes chrétiens de Najrān au 6^e siècle de l’ère chrétienne’, Hawliyāt (Faculté des Lettres et des Sciences, Université de Balamand, Liban) (titre arabe Ḥawliyyāt), 19, 2020, Special issue, 57-109.

———, ‘Langues, écritures, tribus et religions à Najrān à la veille de l’Islam’, forthcoming.

Robin, Christian Julien, ʿAlī I. al-Ghabbān, and Saʿīd F. al-Saʿīd, ‘Inscriptions antiques récemment découvertes à Najrān (Arabie séoudite méridionale): Nouveaux jalons pour l’histoire de l’oasis et celle de l’écriture et de la langue et du calendrier arabes’, Comptes rendus des Séances de l’Académie des Inscriptions et Belles-Lettres 158 (2014): 1033–128.

Robin, Christian Julien, and Sarah Rijziger, ‘“The Owner of the Sky, God of Israel” in a New Jewish Ḥimyaritic Inscription Dating from the Fifth Century CE’, Der Islam 95 (2018): 271–90, http://dx.doi.org/10.1515/islam-2018-0050.

Rothstein, Gustav, Die Dynastie der Laẖmiden in al-Ḥîra: Ein Versuch zur arabisch-persischen Geschichte zur Zeit der Sasaniden (Hildesheim: Georg Olms, 1968).

Schröter, Robert, ‘Trostschreiben Jacob’s von Sarug an die himjaritischen Christen’, Zeitschrift der Deutschen Morgenländischen Gesellschaft 31 (1877): 360–405.

Segal, Alan F., Two Powers in Heaven: Early Rabbinic Reports about Christianity and Gnosticism (Leiden: Brill, 1977), https://dx.doi.org/10.1093/0199246122.003.0004.

Shahīd, Irfan, The Martyrs of Najran: New Documents (Brussels: Société des Bollandistes, 1971).

Stein, Peter, Die altsüdarabischen Minuskelinschriften auf Holzstäbchen aus der Bayerischen Staatsbibliothek in München, 2 vols. (Tübingen: Wasmuth, 2010).

Toral-Niehoff, Isabel, ‘The ʿIbād of Al-Ḥīra: An Arab Christian Community in Late Antique Iraq’, in The Qurʾān in Context: Historical and Literary Investigations into the Quranic Milieu, ed. by Angelika Neuwirth, Nicolai Sinai, and Michael Marx (Leiden: Brill, 2009), 323–48, http://dx.doi.org/10.1163/ej.9789004176881.i-864.4.

———, al-Ḥīra: Eine arabische Kulturmetropole im spätantiken Kontext (Leiden: Brill, 2013), http://dx.doi.org/10.1163/9789004260245.

ʿUkāsha, Tharwat, ed., al-Maʿārif by Ibn Qutayba (Cairo: Wizārat al-Thaqāfa wa-ʾl-Irshād al-qawmī, al-Idāra al-ʿāmma li-l-thaqāfa, 1960).

Walker, John, A Catalogue of the Arab-Sassanian Coins: Umaiyad Governors in the East, Arab-Ephthalites, ʿAbbāsid Governors in Ṭabaristān and Bukhāra (London: The British Museum, 1941).

Wissmann, Hermann von, ‘Himyar: Ancient History’, Le Muséon 77 (1964): 429–99.

Abbreviations

Ag (Fahmī) al-Aghbarī & Khālid al-Ḥājj

BSS Bayerische Staatsbibliothek

CIH Corpus Inscriptionum Semiticarum: Inscriptiones Himyariticas et Sabaeas Continens

DAI GDN Deutsches archäologisches Institut, Grosser Damm Nord

DJE Deutsche Jemen-Expedition

Fa (Ahmed) Fakhry

FB François Bron

Garb (Giovanni) Garbini

Gl (Eduard) Glaser

Ir (Muṭahhar) al-Iryānī

Ist Istanbul

Ja (Albert) Jamme

L Leiden

MAFRAY Mission archéologique française en République arabe du Yémen

MAFSN Mission archéologique franco-saʿūdienne de Najrān

MAFY Mission archéologique française au Yémen

MIbb Museum of Ibb

MS Mohammed al-Salami

RES Répertoire d’épigraphie sémitique

RIÉth Recueil des inscriptions de l’Éthiopie

Ry (Gonzague) Ryckmans

SR Sarah Rijziger

YM Yemen Museum

ẒM Ẓafār Museum

Inscriptions Quoted

1. South Arabia

Inscriptions with a paragraph mark (§) are found in Robin, ‘Quel judaïsme en Arabie?’.

Those with a star (*) are accessible on the DASI website http://dasi.humnet.unipi.it/.

*Ag 2 = Gorge of the Upper-Buraʿ 2.

*Ag 3 = Gorge of the Upper-Buraʿ 3.

*al-ʿĀdī 21.

*al-ʿIrāfa 1.

*CIH 152 + 151 (§); 534; 537 + RES 4919 = Louvre 121; 540; 541; 543 = ẒM 772 A + B (§); 560; 620.

*DAI GDN 2002 / 20.

DJE 23 (Hebrew): See most recently Maria Gorea, ‘Les classes sacerdotales (mišmarôt) de l’inscription juive de Bayt Ḥāḍir (Yémen)’, in Le judaïsme de l’Arabie antique: Actes du colloque de Jérusalem (février 2006), ed. by Christian Julien Robin (Turnhout: Brepols, 2015), 297–329.

FB-Ḥawkam 3.

*Garb Antichità 9 d; Bayt al-Ashwal 1 (§); Bayt al-Ashwal 2 (§); Framm. 3; Framm. 7 (§); BSE ( = Garb Minkath 1); Nuove iscrizioni 4; Shuriḥbiʾīl ( = ẒM 1 (§)).

*Gl 1194.

*Haram 8; 53.

Ḥimà-Sud PalAr 2: Christian Julien Robin, ʿAlī I. al-Ghabbān, and Saʿīd F. al-Saʿīd, ‘Inscriptions antiques récemment découvertes à Najrān (Arabie séoudite méridionale): Nouveaux jalons pour l’histoire de l’oasis et celle de l’écriture et de la langue et du calendrier arabes’, Comptes rendus des séances de l’Académie des Inscriptions et Belles-Lettres 158 (2014): 1033–128 (1092–93). Ḥimà-Sud PalAr 8: Ibid., 1099–102.

*Ibrāhīm-al-Hudayd 1 ( = ẒM 2000 (§).

*Ir 12.

*Ist 7608 bis + Wellcome A 103664; for a possible reconstruction of the text, integrating the various fragments, see Robin, ‘Joseph, dernier roi de Ḥimyar’, 96–100, and Robin, ‘Ḥimyar, Aksūm and Arabia Deserta in Late Antiquity’, 163–64 (translation only).

*Ja 516.

*Ja 547 + 546 + 544 + 545; see most recently Ja 547 + 546 + 544 + 545 = Sadd Maʾrib 6 in Christian Darles, Christian Julien Robin, and Jérémie Schiettecatte, with a contibution by Ghassan el Masri, ‘Contribution à une meilleure compréhension de l’histoire de la digue de Maʾrib au Yémen’, in Regards croisés d’Orient et d’Occident: Les barrages dans l’Antiquité tardive, ed. by François Baratte, Christian Julien Robin, and Elsa Rocca (Paris: de Boccard, 2014), 9–70.

Ja 856 = Fa 60 (§); 1028 (§); 2484.

Khaldūn-ʿIlbij 1: unpublished text, see Khaldon Noman (Khaldūn Hazzāʿ ʿAbduh Nuʿmān), ‘A Study of South Arabian Inscriptions from the Region of Dhamār (Yemen)’ (PhD diss., Università di Pisa, 2012).

L001: Drewes and Ryckmans, Les inscriptions sudarabes, by issue.

Maʾsal 3: Alessia Prioletta and Mounir Arbach, ‘Ḥimyar en Arabie déserte au V^e siècle de l’ère chrétienne: Une nouvelle inscription historique du site de Maʾsal, Arabie saoudite’, Comptes rendus des séances de l’Académie des Inscriptions et Belles-Lettres 160 (2016): 917–54.

*MAFRAY-Ḥaṣī 1 (§).

*MAFY-Banū Zubayr 2.

MB 2004 I-123: unpublished text (American excavations in Maʾrib).

MB 2004 I-147: unpublished text (American excavations in Maʾrib).

*MIbb 7.

*Murayghān 1 = Ry 506.

*Murayghān 3.

*MS-Tanʿim al-Qarya 9.

Naveh-Epitaph of Leʾah (§): Robin, ‘Quel judaïsme en Arabie?’, 191–92.

Naveh-Epitaph of Yoseh = Naveh-Ṣuʿar 24 (Aramaic) (§): Robin, ‘Quel judaïsme en Arabie?’, 192–93.

*RES 3383 (§); 4105; 4109 = M. 60.1277 = Ja 117 = Ghul-YU 35 ; RES 4176.

Robin-Viallard 1 = Ja 3205.

*Ry 506 = Murayghān 1; Ry 507; Ry 508 (§); 509; 510; 515 (§); 520 (§); 534 + Rayda 1 (§).

Schiettecatte-Nāʿiṭ 9: Robin, ‘Le roi ḥimyarite’, 62–63 and fig. 20 (93).

*Sharʿabī al-Sawā 1.

SR-Naʿḍ 9: Addendum, above.

X.SBS 141 = Mon.script.sab 6: Stein, Die altsüdarabischen Minuskelinschriften, by issue.

YM 327 = Ja 520: Robin, ‘Quel judaïsme en Arabie?’, 270, fig. 3.

*YM 1200 (§); 1950; 11733 = X TYA 9; 11738 = X TYA 15; YM 14556= CSAI 1, 114.

*ẒM 1 (§) (see Garb Shuriḥbiʾīl); 5 + 8 + 10; 2000 (§) (see *Ibrāhīm-al-Hudayd 1 (§)).

2. Ethiopia

RIÉth 189, 190, 191, 192, 195 (South Arabian and Geʿez); 271 (Greek): Étienne Bernand, Abraham J. Drewes, and Roger Schneider, Recueil des inscriptions de l’Éthiopie des périodes pré-axoumite et axoumite, Tome I: Les documents; Tome II: Les planches (Paris: de Boccard, 1991); idem, Tome III: Traductions et commentaires A: Les inscriptions grecques (Paris: de Boccard, 2000).

Illustrations

Fig. 1: One of the two earliest royal inscriptions invoking the One God; it comes from Ẓafār, Ḥimyar’s capital (Garb Bayt al-Ashwal 2, January 384 CE). Photograph by Christian Julien Robin. © All rights reserved.

Fig. 2: Two fragments of the inscription of King Shuriḥbiʾīl Yaʿfur commemorating an important reconstruction of the Ma ʾrib Dam (Ma ʾrib, CIH 540, January 456 CE). Photograph by Christian Julien Robin. © All rights reserved.

Fig. 3: Inscription commemorating the building of a royal palace in the capital (Ẓafār, ẒM 1 = Garb Shuriḥbiʾīl Yaʿfur, December 462 CE). Photograph by Christian Julien Robin. © All rights reserved.

Fig. 4: Inscription commemorating the building by a Jew of a palace in the capital (Garb Bayt al-Ashwal 1, between 380 and 420 CE). Photograph by Christian Julien Robin. © All rights reserved.

Fig. 5: Garb Bayt al-Ashwal 1: Hebrew graffito in the central monogram. Photograph by Christian Julien Robin. © All rights reserved.

Fig. 6: Inscription commemorating the blockade of Najrān in June–July 523 CE by the Ḥimyarite army sent by King Joseph (Ḥimà, al-Kawkab, Ry 508, June 523 CE). Photograph by MAFSN. © All rights reserved.

Fig. 7: Inscription carved by the two chiefs of the Ḥimyarite army sent by King Joseph (Ḥimà, al-Kawkab, Ry 515, June 523 CE). Photograph by MAFSN. © All rights reserved.

Fig. 8: Another inscription commemorating the blockade of de Najrān in June–July 523 CE by the Ḥimyarite army of King Joseph (Ḥimà, the wells, Ja 1028, July 523 CE). Photograph by MAFSN. © All rights reserved.

Fig. 9: Detail of Ja 1028, July 523 CE: The last line is to be read rb-hd b-mḥmd, ‘Lord of the Jews with the Praised One’. Photograph by MAFSN. © All rights reserved.

Fig. 10: (above and bottom) Princely inscription commemorating the building of a mikrāb: Ry 534 + Rayda 1, reprinted back to front to facilitate reading (Rayda, 55 km north of Ṣanʿāʾ, August 433 CE). Photograph by MAFY. © All rights reserved.

Fig. 11: Inscription commemorating the building of two palaces in the capital (Ẓafār, ẒM 5 + 8 + 10, February 432 CE). Photograph by Christian Julien Robin. © All rights reserved.

Fig. 12: Another inscription commemorating the building of a palace in the capital; in line 8, its authors claim to belong to the commune Israel (Ẓafār, ẒM 2000, April 470 CE). Photograph by Ibrāhīm al-Hudayd. © All rights reserved.

Fig. 13: Princely inscription creating a cemetery intended for Jews (Ḥaṣī, some 220 km southeast of Ṣanʿāʾ, MAFRAY-Ḥaṣī 1). Drawing by Maria Gorea. © All rights reserved.

Fig. 14: List of Mishmarot of Bayt Ḥāḍir, 15 km east of Ṣanʿāʾ. Drawing by Maria Gorea. © All rights reserved.

Fig. 15: Princely inscription commemorating the construction of a mikrāb in the vicinity of Ṣanʿāʾ; dated August [36]3 or [37]3 CE, it is from before the ruler’s conversion to Judaism (YM 1950). Photograph by Iwona Gajda. © All rights reserved.

Fig. 16: Inscription commemorating the construction of a palace at Naʿḍ, 35 km southeast of Ṣanʿāʾ (SR-Naʿḍ 9, the date has been mutilated, but it is either 430–440 or 460–470 CE). Photograph by Sarah Rijziger. © All rights reserved.

Maps

Map 1: South Arabia. Drawing by Daniel Stoekl, Hélène David-Cuny, and Astrid Emery (2020).

Map 2: The Arabian Peninsula. Drawing by Daniel Stoekl, Hélène David-Cuny, and Astrid Emery (2020).