Ẓafār

Royal capital of Ḥimyar (officially known as dhu-Raydān).

Ẓafār is located at 14°12’40”N, 44°24’10”, alt. 2740 m (map Robin & Brunner 1997: E8). Today, the site is still called Ẓafār after the village that settled there.

The first archaeologist to visit Ẓafār was the Austrian Eduard Glaser in early 1886; his unpublished notes are now available online (glaser.acdh.oeaw.ac.at).

The site was explored and excavated by a Heidelberg archaeological mission led by Paul Yule between 1998 and 2010 (Yule 2013). The inscriptions collected in the site’s small museum were the subject of a posthumous catalogue (Sima 2020). See also Schiettecatte’s synthesis (2011: 274-284, with bibliography) and the entry on Ẓafār in the second edition of the Encyclopedia of Islam (Müller 2000).

A. Epigraphic and manuscript sources

[By Ch. J. Robin]

Among the cities of Arabia, Ẓafār is one of those most often cited by epigraphic sources and manuscripts.

Epigraphic documentation from Ẓafār

The epigraphic documentation from Ẓafār itself is disappointing. Almost all the texts commemorate constructions and date from the period of the 4th-6th centuries (when the city was already full-grown). The almost complete absence of ancient texts and, in particular, texts from the temples, is probably explained by the fact that Ẓafār was occupied twice by the Abyssinians in the 3rd century and probably completely ruined.

For the 4th–6th centuries, when the city expanded considerably and became the capital of a large empire, the number of substantial inscriptions (which are mostly building commemorations) hardly exceeds twenty. To these must be added countless fragments of inscriptions that have been deliberately broken. It likely illustrates the ravages suffered by the city when it was taken by the Aksūmites (or Abyssinians) of the negus Kālēb Ella Aṣbəḥa in 525 or shortly thereafter. After this conquest, Ẓafār would still be inhabited as five Aksūmite inscriptions have been found (hypothetically post-525). It was abandoned around the middle of the 6th century.

Ẓafār’s epigraphic documentation includes no texts commemorating a polytheistic rite. This is highly unusual for a city that appears at a relatively early date. There was obviously no major polytheistic sanctuary in Ẓafār.

Epigraphic mentions of Ẓafār are found mainly in Qatabānite, Sabaean and Ḥimyarite inscriptions between 50 BCE and 400 CE.

Early epigraphic records (1st century BCE and CE)

The earliest mentions of Ẓafār, dating back to around the beginning of the Christian era, come from three Qatabānite towns: Maryamat^um, Haribat and Tamnaʿ.

In the city of Maryamat^um (today al-ʿĀdī in the Wādī Ḥarīb, Robin & Brunner 1997: G7, 140 km north-east of Ẓafār), a series of texts are sponsored by Maryamites established abroad who commemorate offerings in temples or the construction of defensive works within the city walls. These Maryamites, who were certainly merchants, resided in Tubnà (today Laḥj, ca. 30 km NNW of the port of ʿAdan), Shukaʿ^um (ca. 100 km N of ʿAdan) and Ẓafār.

The community of (traders of) Maryamat^um in Ẓafār financed the construction of a tower of the Maryamat^um rampart in the time of the Ḥimyarite prince Shammar Yuharʿish (ca. 25–1 BCE) (al-Ḥājj al-ʿĀdī 92), a second during the reign of the ḥimyarite king Karibʾīl Watār Yuhanʿim king of Sabaʾ and dhu-Raydān (ca. 40–70 BCE) (CSAI II, 6) and a third without date indication (al-Ḥājj al-ʿĀdī 90). Another tower is built by the community of (traders) from Raḥbatān (probably the town that is replaced by Ṣanʿāʾ in the 1st cent. CE) (al-Ḥājj al-ʿĀdī 91). Finally, a member of the ruling lineage of Maryamat^um makes ‘two offerings, one of silver whose weight is fifty shekels and one of bronze of ten shekels in the shrine Manʿ^um, (in respect of) taxes levied for their lord (the god) Ḥawkam in the two cities of Ẓafār and Maryamat^um’ (FB-Ḥawkam 3).

Regarding the Qatabānite town of Haribat (today Ḥinū al-Zurayr, Robin & Brunner 1997: G7-H7, 7 km south-east of Maryamat^um), its traders reside in Śawā^m (Robin & Brunner 1997 E10, ca. 100 km south of Ẓafār and 90 km east of the port of Mouza, RÉS 4329) and Ẓafār (Ry 497, MuB 554). They too make offerings or finance the construction of the rampart in their home city. Those of Ẓafār finance the construction of a tower during the reign of the Qatabānite king Warawʾīl Ghaylān Yhanʿim (ca. 25-40 CE, Ry 497) and make an offering in a temple (MuB 554).

In the necropolis of Tamnaʿ, capital of Qatabān, the base of a funerary statuette bears the inscription “Abode of Shamnar ibn Abḥaḍ in Riṣāf^um , the Ẓafārite” (Ja 358). The palaeography of the text makes it possible to date this monument approximately to the end of the 1st century BCE. This Shamnar ibn Abḥaḍ is probably a Ḥimyarite since his name and lineage are attested only in Ḥimyar: it is therefore very likely that the adjective “Ẓafārite” refers to the capital of Ḥimyar.

These inscriptions contain three important pieces of information. The first relates to chronology. All these texts date — approximately because all the dates are mainly based on palaeography— from the period between 50 BCE and 100 CE. The second indication concerns the orientation of this trade, which was clearly directed towards the ports, ʿAdan (170 km south-east of Ẓafār) and Makhawān/Mouza (150 km south-west of Ẓafār). The third is that Ẓafār was an important trading place at the turn of the Christian era.

Ẓafār and dhu-Raydān/Ḥimyar

The appearance of Ẓafār in epigraphic documentation can be related to two contemporary events.

The first is the break-up of the kingdom of Qatabān, whose western territories dissented and organised themselves into independent principalities: dhu-Raydān, Radmān, Maḍḥà^m, Khawlān, Sufār^um and Maʿāfir^um . The secession of dhu-Raydān could date back to 110 BCE (because this date is the inception of the Ḥimyarite era), but it is not recognised in inscriptions until around 50 BCE. It would seem that the other principalities were then more or less dependent on dhu-Raydān.

Dhu-Raydān (‘that of Raydān’ or ‘those of Raydān’) is the name of a princely lineage, deriving from Raydān, the dynastic palace of this lineage located in Ẓafār. The princes of dhu-Raydān have authority over a set of tribes-shaʿb called dhu-Raydān. This tribal set is also called Ḥimyar, although it is not clear what distinguishes Ḥimyar from dhu-Raydān. Very hypothetically, Ḥimyar could be the appellation of the tribes forming the initial core of the principality and dhu-Raydān the name of all the populations subject to the authority of the Banū dhu-Raydān.

A text from the 3rd century CE states that the tribe to which Ẓafār belongs is called Yaḥṣib (al-Miʿsāl 5/12); but it is not known whether this was already the case in the 1st century BCE.

The second event which can be related to the appearance of Ẓafār is the Roman expansionism in Arabia after the conquest and annexation of Egypt by the Roman ruler Augustus in 30 BCE. The Roman authorities wanted to take control of the aromatics trade in Arabia. They therefore needed to improve their geographical knowledge. To this end, a military expedition commanded by the prefect of Egypt Aelius Gallus was launched in Arabia in 25 BCE to seize the kingdom of Sabaʾ.

Around this time, the prince of Ḥimyar, Shammar, was at war with the three great kingdoms of southern Arabia (Sabaʾ, Qatabān and the Ḥaḍramawt — see CSAI, II, 14 = RÉS 4336). He was therefore a potential ally of the Romans. No source explicitly mentions this alliance between Ḥimyar and Rome, but it is plausible: firstly the princes of Ḥimyar benefited from the Roman expedition since they then seized the throne of Sabaʾ; secondly a slightly later Greek source (ca. mid 1st cent. CE) indicates that Ḥimyar was an ally of Rome (Peripl. M. Rubr. §23).

Roman expansionism in Arabia was also reflected in a rapid increase in the number of Roman ships travelling between Egypt and the Indian peninsula, calling at four South Arabian ports, Mouza (or Makhawān), Okêlis (in the straits), Eudaimôn Arabia (or ʿAdan) and Qanīʾ.

For this period, the names of three Ḥimyar princes are known: Shammar already named and two others who minted coins in the 1st century BCE, Mubahhal dhu-Raydān and Yuhabirr.

At an uncertain date in the two or three decades following the expedition of Aelius Gallus, the princes of dhu-Raydān toppled the king in Sabaʾ and ascended the throne. They assumed the title of ‘King of Sabaʾ and dhu-Raydān’. Their official residence is now in Marib, the capital of Sabaʾ, but dhu-Raydān retains the status of an autonomous principality with Ẓafār as its capital.

The first mentions of Ẓafār in manuscript sources

In external sources, the name Ẓafār appears in Pliny’s Natural History published around 77 CE, not in the description of Arabia itself, but, incidentally, in a revised version of the route between Egypt and India:

More in the interior (of the region producing incense), there is a city; the residence of the king is called Sapphar, and there is another city known by the name of Save.... None of these names of nations, ports, and cities are to be found in any of the former writers, which shows that the situation of the places has changed (HN VI, 104-105).

Similar information can be found in the Periplus of the Eritrean Sea:

(22) A three-day journey inland from Muza lies Savè, the city of the country called Mapharitis, that surrounds it. The governor (tyrannos), Cholaibos, has his residence there. (23) Nine days further inland is Saphar, the metropolis, residence of Charibael, legitimate king of the two tribes (ethnê), the Homerite and the one, lying next to it, called the Sabaean; he is a friend of the emperors, thanks to continuous embassies and gifts’ (adapted from Casson 1989).

It is likely that Pliny either summarised the Periplus — whose date is hypothetically slightly earlier than the Natural History — or used the same source as it.

In the following century, Ptolemy’s Geography describes Sapphara as a ‘metropolis’ (Geog. VI, 7, 41 and mentions it in a list of the main cities of ‘Arabia Eudaemon’, i.e., the Arabian Peninsula (Geog. VIII, 22, 16). Among the populations of the region, he cites the Sappharitai, namely the Ẓafārites (Geog. VI, 7, 25).

The city of Ẓafār in the 3rd century CE

We know of no details about the city of Ẓafār in the 1st and 2nd centuries of the Christian era. The first concrete information dates from 347 Ḥim./237–238 CE, with a fragmentary inscription commemorating the (re)construction of the rampart (ẒM 2263 + 2262 + 2264). The deities invoked by the unknown sponsor (maybe king Karibʾīl Ayfaʿ) are “their deities Walīl and Sumūyadaʿ, the mistress of the common dhāt[-... ...]” and various minor deities.

This (re)construction of the rampart came a few years after the Abyssinians (who then occupied the coastal regions of western Yemen) seized Ẓafār (Ja 631/16–36, ca. 230) and probably destroyed it.

The city of Ẓafār was again attacked and partially occupied in the early 260s as reported in the inscriptions al-Miʿsāl 3 (ll. 8–9) and al-Miʿsāl 5 (ll. 9–11). An inscription (Gr 27) commemorates the rebuilding of a palace that the Abyssinians had burnt down, probably during this occupation in the early 260s. Its sponsors also invoke ‘their deities Walīl and Su[mūyadaʿ ...]’.

These two deities Walīl and Sumūyadaʿ apparently occupy the first place in the Ḥimyarite pantheon. It is more difficult to say the same about the god Taʾlab (worshipped by the tribe-shaʿb of Samʿī north of Ṣanʿāʾ) whose name appears in a single inscription, ẒM 7, the palaeography of which seems to imply a date earlier than 3rd cent.

A Greek inscription, of which only an insignificant fragment survives, could date back to this period (Marek 2013).

Ẓafār in Late Antiquity (300-550)

The Ḥimyarites annexed Sabaʾ and conquered the Ḥaḍramawt around 275 and 300. They launched expeditions throughout the Arabian desert, a large part of which was annexed around 430-440. Ẓafār then became the capital of an immense empire. The city walls were enlarged on several occasions.

Around 380, the royal dynasty rejected polytheism and imposed an official religion inspired by Judaism. Although the state was not officially Jewish, it is likely that the kings were Jewish in a personal capacity. Between the end of the 4th century and the beginning of 6th, a fine series of inscriptions commemorate the construction of royal palaces, called Shawḥatān (RÉS 3383), Kilān^um (Garb Bayt al-Ashwal 2 ; see also ẒM 647) and Hargab (ẒM 1), and private palaces (ẒM 5+8+10; ẒM 2000; ẒM 579; Gar Bayt al-Ashwal 1; Gar nuove iscrizioni 3; Gar nuove iscrizioni 4; Robin-Viallard 1; Sirriyya 1). An inscription (Gar Bayt al-Ashwal 1) incidentally mentions a synagogue (mikrāb); the word is also found in a fragment (Bayt al-Ashwal 4 g). An inscription commissioned by king Marthadʾīlān Yunʿim, whose provenance is unknown (probably Ẓafār), commemorates the construction and fitting out of a synagogue (YM 1200). An inscribed block preserves a small part of a large inscription recounting the conquest and annexation of central Arabia (al-ʿIrāfa 1). A fine inscription (Gar BSE) and a few fragments (Garb Sab. Fragm. V, Ẓafār 10~002 and MQ-Minkath 1) include only the official list of reigning rulers. Finally, we have a few funerary texts (e.g., ẒM 772 A+B and RÉS 5094).

An ecclesiastical history relates the construction of a church in ca. 340s-350s. The king of Ḥimyar, to please an embassy sent to him by the Byzantine emperor Constantius II (r. 337-361), financed the construction of three churches, two in the ports and one in the capital Tapharon (Ẓafār), evidently intended for Romans travelling through the region (Philostor., Hist. Eccl. III, 4).

Ẓafār is the city in which the Christian priest Azqīr of Najrān, “who introduced a new religion” (Martyrdom of Azqīr, § 9), appears before King Sərābḥēl Yənkəf (Shuriḥbiʾīl Yakkuf, ca. 468-ca. 480) (§ 21: Ṣəfār). On the advice of the Jews in the king’s entourage, Azqīr is taken back to Najrān where he is beheaded.

Towards the beginning of the 6th cent., the kingdom of Ḥimyar became tributary to that of Aksūm in Africa. Its kings, who were henceforth chosen by the Negus, were Christians. An Aksūmite garrison of 300 men is installed in Ẓafār in a church (probably fortified) (Letter 1/ed. Guidi, 280 men; Letter 2/ed. Shahîd 300; Ry 507/4, 300).

After the death of the Christian king Maʿdīkarib Yaʿfur (c. June 522), the negus Kālēb Ella Aṣbəḥa aided a prince named Joseph (Yūsuf Asʾar Yathʾar) to gain the throne. His revolt, which took place in the autumn of 522, is reported by five inscriptions and several Syriac hagiographic texts, which agree on the sequence of events:

They assisted their lord the king Yūsuf Asʾar when he fought the Abyssinians at Ẓafār, killed them all and set fire to their church; then the king sent them with an army against the Abyssinians, Farasān and Mukhawān, and they killed them all, captured their daughters and the king at Ashʿarān’ (MAFSN-GDY01_57–1).

This church of the Aksūmites, which is burnt down according to the inscriptions, is changed into a synagogue according to the earliest Christian hagiographical account of the Najrān massacre (Letter 1/ed. Guidi); but later hagiographical sources correct this error and confirm the destruction by fire. It is interesting to observe that an erroneous datum recorded by ecclesiastical scholarly tradition may have been corrected by later editors (Robin 2011).

What happened next is known only from Christian hagiographic sources in Syriac and Greek, whose reliability is uneven. The negus Kālēb sends a large army that manages to land in Arabia and kills king Joseph. The Aksūmites immediately seize Ẓafār (Greek Martyrdom of Arethas, § 35: Taphar) and conquer the whole country as far as Najrān. Himyar receives a new Christian king, installed in Ẓafār, and is now officially acknowledged as a Christian kingdom.

A Christian hierarchy was created and churches were built, but the information transmitted by the sources is highly dubious. According to the Martyrdom of Arethas in Greek, King Kālēb transforms the dynastic palace of Ẓafār into a church:

Then the king dug into the palace with his own hands for seven days and built the most holy church that exists today’ (§ 38).

According to the Life of Gregentios, a late writing of even lower reliability, King Kālēb built three churches in Ẓafār (Tephar): the Great Church in the name of the Holy Trinity near the palace; the church in the name of the holy mother of God in Dana; the church of the holy apostles near the western gate (Berger 2006).

It is probable that the city of Ẓafār, on its conquest by Kālēb, was plundered and devastated. Almost all the sculpted and inscribed remains that have been collected there in recent years were reduced to fragments. However, it does not appear that the city was abandoned. It would even seem that it was still the seat of royal Ḥimyarite power for a time. Five inscriptions in vocalised Geʿez have turned up there:

• RIÉth 263 = ẒM 237
• RIÉth 264 = ẒM 308
• RIÉth 265
• RIÉth 266
• Ẓafār lz 08–077.

It is likely that these anodyne inscriptions date from the Aksūmite occupation after 525–530, although it is not completely impossible that one or other dates from an earlier period.

Ẓafār most probably lost its status as a royal residence at the time of Abraha (r. ca. 532–ca. 565), an Aksūmite army leader who came to power by a revolt and took the place of the Christian Ḥimyarite king appointed by Kālēb. Indeed, it would appear that Abraha settled in Ṣanʿāʾ. It was in this city that he had his Great Church, called al-Qalīs by the Arab-Muslim scholarly tradition, built around 560. But it is only in traditions reporting the struggles for power on the eve of Islam that the Ghumdān Palace of Ṣanʿāʾ is explicitly mentioned as the seat of power.

The capital of Ḥimyar is logically mentioned in Late Antique external sources, while Ḥimyar is the main regional power. In Greek, the name is transcribed Tapharon (Philostor., Hist. eccl. III, 4), Taphar (Martyrdom of Arethas in Greek, §§ 35 and 37), or Tephar (Life of Gregentios, ch. 44); in Latin, it is Tafra or Tarphara (Amm. Marc., Hist. XXIII, 6, 47, in the list of the principal cities of Arabia Felix); in Geez, it is Ṣəfār (Martyrdom of Azqīr, § 21); in Syriac, it is Ṭpr (Book of the Ḥimyarites, 7a twice, 7b, 8b, 19a) and Ṭypr (Letter 2/ed. Shahîd, III/11 f° IIA).

Interestingly, in Greek and Latin, the initial consonant of Ẓafār’s name is rendered as /s/ in ancient times, but as /t/ during Late Antiquity. The explanation may lie in the fact that knowledge about Arabia was no longer gathered directly on the spot, but passed on through Aramaic-speaking intermediaries.

The information that the Yemeni traditionist al-Hamdānī (d. after 970) passes on about Ẓafār is abundant, but of uncertain interpretation. It does not appear that al-Hamdānī personally visited the place; he refers to his master Abū Naṣr, the Ummayyad-era poet ʿAlqama ibn dhī Jadan and fictitious poetic fragments attributed to pre-Islamic Ḥimyarite kings. He lists the nine gates of the city (Iklīl 8, ed. Fāris: 26; trans. Fāris: 22–23) and cites several palaces (p. 23 and p. 20): dhū Yazan, Raydān, Shawḥaṭān, Kawkabān. One might think that the city is still standing and inhabited. However, al-Hamdānī does not mention any social group having its residence in Ẓafār. This is in clear contrast to the nearby market town of Mankath, residence of the Sukhṭiyyūn (Iklīl 8, ed. Fāris: 26; trans. Fāris: 22; Nashwān, Shams, ed. Aḥmad: 48, entry SH̲Ṭ) who are the descendants of King Joseph (Iklīl 2, ed. al-Akwaʿ: 59–64). It is therefore probable that Ẓafār was deserted in the 10th century and that the information transmitted by al-Hamdānī was of a scholarly nature.

The reliability of the information transmitted by al-Hamdānī is difficult to assess. The existence of the Shawḥaṭān palace is confirmed by an inscription (RÉS 3383). But it is quite difficult to recognize the nine gates (see below). More generally, none of the monuments mentioned in written, epigraphic and manuscript sources have yet been precisely located, except perhaps, very hypothetically, the dynastic palace of Raydān.

B. Archaeology

[By Paul A. Yule]

The two extinct volcanoes upon which ancient Ẓafār is built are known from mapping of the site and excavation which took place from 1998 to 2010, the latter largely on the SW slope of what the locals call the Ḥuṣn Raydān (Yule 2013). The 1000 x 1300 m site map reveals a heterogeneous population density. Ancient Maʾrib was larger, far more densely built and had a larger population.

Town defences

On the strength of the dense building debris, the settlement extended from the south and north of the city wall, the latter area was sparsely populated. The population lived both in and outside of the defences. The town defences consist of a double wall, of which 1% was proven by means of survey, ground penetrating radar and excavation (Fig. 1). The walls are most complete in the south. According to al-Hamdānī nine gates gave access to the town, the main one lying in the south, today which the east-west road transects. The city wall combined with a shearing off of rock faces to create defences. Remains of the defences in the south show them to be made at least partly of massive stone, repeatedly repaired.

Architecture

The main excavated structure lies to the SW of the Ḥuṣn Raydān in a slope known as al-Jaḥw (Figs 1-2). A stone building —perhaps a palace, or at some time a temple— existed over centuries, changing its form from time to time. The Ḥimyarite period ‘Stone Building’, built of limestone ashlar around a paved courtyard shows a high masonry standard. Its debris contained numerous reliefs which appear to be in secondary context and which rolled down the mountain into the structure. The excavation z300 and the Stone Building itself comprise the Stone Building site. This evidences five diachronic building states, beginning around the year -1/1 and continuing to shortly after 525 CE, on the strength of radiocarbon, stratigraphy, and the interpreted dating before and after the dated states 2 and 4 respectively (Yule 2013: 225).

Funerary architecture

Pre-Islamic tombs and graves carpet Ẓafār. Numerous tombs lie in and outside the defences (Fig. 3). Aside from large and elaborate chamber tombs (Figs 3.1, 3.3–3.8, 3.10), deeper and hypogea (Fig. 3.2) as well as smaller chamber tombs with a long entrance came to light (Fig. 3.9). Such a cemetery, zc01, 800 m west of the present-day village proved to be rifled. Larger and far more elaborate are the presumed royal tombs on the SW flank of the town, which reflect the highest-ranking members of the Ḥimyarite population. Lying on the west slope of al-ʿAṣabī, three large rock-cut tombs lie within 10 m of each other. They range from 6 to nearly 10 m in width (Figs 3.5-3.7) and lie outside the ancient defences.

Irrigation architecture

The ancient population needed an elaborate irrigation network to retain water over alternating periods of precipitation and dryness (Yule 2013: 31–32). 500 m SW of the main defences lies a dam (35 x 8 m) and a terrace complex necessary for the town water management (Fig. 4). At the west end a sluice and a dam (site z236), crafted from rough Ḥimyarite masonry, retain water for an agricultural catchment basin 1 km in length —the wettest place in the vicinity of the town. The agricultural base of the Ḥimyarite town was of prime importance.

Visual arts

At Ẓafār, the ground is littered with fist-sized, battered Ḥimyarite limestone reliefs, the most typical find, a vague echo of a once flourishing art industry. 991 registered fragments are on deposit in the site museum in its storage (Costa 1973; 1976; Yule 2009). Such far exceed the number of pottery fragments collected from the surface. In addition, excavation yielded nearly as many. The most striking relief assemblage is a 10 m relief wall (context z502) with its four sculpted registers in the eastern wall of the inner courtyard of the Stone Building (Yule 2013: 73, Fig. 3.14; 135, Fig. 7.1–7.2). It seems to postdate the year -1/1. The most striking find is the ‘crowned man’ fashioned from pale orange hard limestone in Byzantine style (Fig. 5) inserted into the SE corner of this wall. For a variety of reasons, a dating of ca. 525 has been ventured, but can be discussed.

Pottery

Local and import wares came to light as surface finds but more from excavation, the majority from Late Antiquity. Locally produced pottery reveals a plethora of different combinations of surface treatment, production methods, temper and shape. Most striking are storage amphorae in a tough, course Late Antique ribbed ʿAqaba ware presumably containing imported white wine (Fig. 6).

Plant remains

Hordeum vulgare (hulled barley) comprised more than 50 % of the eight different cereals (Rösch & Fischer 2013: 188). Less common were Avenna (oats), Triticum aestivum/durum/turgidum (free-threshing wheat), Triticum dicoccon (hulled wheat). Oil and fibre plants have a poorer survival rate.

Animal remains

6000 faunal remains derived mostly from the Stone Building, fewer from cemetery zc001 (Uerpmann & Uerpmann 2013). Cattle made up 90 % of the bones in the Stone Building, which may have served at this time as a slaughterhouse. Hybrid camels seem to be in evidence – a crossing of a Bactrian sire with a dromedary mare.

The Ḥimyarite population must have put heavy demands on the immediate mountain environment for fuel and for agriculture which were not sustainable.

Christian J. Robin, Paul A. Yule

References and suggested reading

Sources — South Arabian inscriptions

Refer to the DASI online database (http://dasi.cnr.it/). For texts not included in this database:

• al-Ḥājj al-ʿĀdī 90, 91: al-Ḥājj 2020.
• al-Ḥājj al-ʿĀdī 92: al-Ḥājj 2022.
• MAFSN-GDY01_57-1: unpublished, to appear in CRAI 2023.
• Al-Miʿsāl 3: unpublished.
• MQ-Minkath 1 = Mankaṯ 5: Sima 2020: 80; Robin 2014: 43, fig. 11.
• Sirriyya 1 = Sirrīyah 1: Sima 2020: 85, Abb 188 a, b.
• ẒM 7: Sima 2020.

Sources — Aksumite and Greek inscriptions

• RIÉth: Bernand et al. 1991; Drewes [& Schneider] 2019.
• Ẓafār lz 08-077: Müller 2012; Yule 2013, fig. 10.6, n° 5.
• ẒM 2021 : Marek 2013.

Sources — Manuscripts

• Ammien Marcellin, Hist.: Ammien Marcellin/Ed. Galletier et al. 1968–1999. Ammien Marcellin, Histoire. Ed., transl. and annot. by Ed. Galletier, G. Sabbah, J. Fontaine, M.-A. Marié. 6 vols. Paris: Les Belles Lettres.
• Life of Gregentios: Berger, A. 2006. Life and Works of Saint Gregentios, Archbishop of Taphar, Introduction, Critical Edition and Translation, edited by ..., with a contribution by Gianfranco Fiaccadori (Millenium Studien/Millenium Studies, 7). Berlin, New York: Walter de Gruyter.
• Al-Hamdānī/Ed. al-Akwaʿ, Iklīl 2: 1967. Kitāb al-Iklīl, al-juzʾ al-thānī, li-Lisān al-Yaman Abī Muḥammad al-Ḥasan b. Aḥmad b. Yaʿqūb al-Hamdānī, ed. By Muḥammad b. ʿAlī ʾl-Akwaʿ al-Ḥiwālī (al-Maktaba al-yamaniyya, 3). Cairo: al-Sunna al-muḥammadiyya.
• Al-Hamdānī/Ed. Fāris, Iklīl 8: 1940. Al-Iklīl (al-juzʾ al-thāmin), by al-Ḥasan ibn Aḥmad al-Hamdānī, edited with linguistic, geographic and historic notes by Nabīh Amīn Fāris (Princeton Oriental Texts, vol. VII). Princeton.
• Al-Hamdānī/Transl. Fāris, Iklīl 8: 1938. The Antiquities of South Arabia, Being a Translation from the Arabic with linguistic, geographic and historical notes of the Eighth Book of al-Hamdāni's al-Iklīl by Nabih Amin Faris (Princeton Oriental Texts, III). Princeton, 1938.
• Lettre 1/Ed. Guidi: I. Guidi, 1881. La lettera di Simeone vescovo di Bêth-Arśâm sopra i martiri omeriti. In Atti della Reale Accademia dei Lincei 278 (1880-1, Serie 3), Memorie della Classe di Scienze morale, storiche e filologiche 7: 471–515; repr. in Raccolta di scritti, 1, Oriente cristiano (Pubblicazioni dell’Istituto per l’Oriente). Rome: Istituto per l’Oriente, 1945: 1-60.
• Lettre 2/Ed. Shahîd: I. Shahīd, 1971. The Martyrs of Najrân. New Documents (Subsidia Hagiographica, 49). Brussels: Société des Bollandistes.
• Book of the Himyarites: A. Moberg, 1924. The Book of the Himyarites. Fragments of a hitherto unknown Syriac work, edited, with introduction and translation, by ... (Acta Reg. Societatis Humaniorum Litterarum Lundensis, VII). Lund: C.W.K. Gleerup.
• Martyrdom of Arethas in Greek: M. Detoraki & J. Beaucamp, 2007. Le martyre de Saint Aréthas et de ses compagnons (BHG 166), Marina Detoraki, ed., Joëlle Beaucamp, transl. Appendix on oriental versions by André Binggeli (Centre de Recherche d’histoire et civilisation de Byzance: Monographies, 27). Paris: Association des amis du Centre d’histoire et civilisation de Byzance.
• Martyrdom of Azqīr: A. Bausi, 2017. Il Gadla ʾAzqir. Adamantius 23: 341–380.
• Nashwān ibn Saʿīd al-Ḥimyarī/Ed. Ahmad, Shams: Die auf Südarabien bezüglichen Angaben Našwān’s im Šams al-ʿulūm, collected, alphabetically arranged and edited by ʿAẓīmuddīn Aḥmad (“E. J. W. Gibb Memorial” series, XXIV). Leiden, London: E.J. Brill, Luzac.
• Periplus of the Eritrean Sea: Casson, L. 1989. The Periplus Maris Erythraei. Princeton: Princeton University Press.
• Philostor., Hist. eccl. (ed.): Philostorgius, Kirchengeschichte, ed. by J. Bidez, 2^nd ed. by F. Winkelmann. Berlin, 1972 (Die Griechischen Christlichen Schriftsteller der ersten drei Jahrhunderte).
• Philostor., Hist. eccl. (transl.): Philostorgius, Church History, transl. with Introduction and Notes by Philip R. Amidon (Society of Biblical Literature, Writings from the Greco-Roman World, 23). Atlanta: Society of Biblical Literature, 2007.
• Pliny, H.N.: Natural history
• Ptol., Geog.: Stückelberger, A. & G. Grasshoff (eds) 2006. Klaudios Ptolemaios. Handbuch der Geographie. Basel: Schwabe Verlag.

Studies

• Bernand, É., A.J. Drewes & R. Schneider 1991. Recueil des inscriptions de l’Éthiopie des périodes pré-axoumite et axoumite. Tome I. Les documents. Tome II. Les planches. Paris: Académie des inscriptions et belles-lettres, De Boccard.
• Costa, P. 1973. Antiquities from Zafar (Yemen). AION 33: 185–206.
• Costa, P. 1976. Antiquities from Zafar (Yemen) - II. AION 36: 445–456.
• Drewes A.J. [& R. Schneider] 2019. Recueil des inscriptions de l'Éthiopie des périodes pré-axoumite et axoumite. Tome III. Traductions et commentaires, B. Les inscriptions sémitiques (Aethiopistische Forschungen 85). Wiesbaden: Harrassowitz [Title from a collection of the AIBL, used without permission].
• al-Ḥājj, M.A. 2020. al-Ahamiyya al-siyāsiyya wa-al-iqtiṣādiyya li-madīnat Ẓafār Ḥimyar fī al-qarn al-awwal al-mīlādī fī ḍawʾ naqshayn musnadiyayn jadīdayn. Majallat Ūrūk li-l-ʿulūm al-insāniyya 13: 711–731.
• al-Ḥājj, M.A. 2022. Naqsh qitbānī jadīd min ʿahd Shammar Yuharʿish dhī Raydān wa-muʿṭiyāt ukhrà ḥadītha ʿan taʾrīkh madīnatay Ẓafār wa-Maryama fī nihāyat al-qarn al-thālith al-mīlādī. Majallat al-Khalīj li-l-taʾrīkh wa-al-āthār 16: 111–133.
• Marek, C. 2013. Chapter 9. A Roman Period Inscriptions (sic) in South Arabia, in Yule (ed.) 2013: 163–165.
• Müller, W.W. 2000. Ẓafār, in Encyclopédie de l’Islam, 2^nd ed., vol. 11: 411–412.
• Müller, W.W. 2012. Äthiopische Inschriftenfragmente aus der himjarischen Hauptstadt Ẓafār. Aethiopica 15: 7–21. DOI: 10.15460/aethiopica.15.1.657.
• Raith, M.M., R. Hoffbauer, H. Euler, P.A. Yule & K. Damgaard 2013. The view from Ẓafār – An archaeometric study of the ʿAqaba pottery complex and its distribution in the 1st Millennium CE. ZOrA 6: 320–350.
• Robin, C.J. 2011. L’église des Aksūmites à Ẓafār (Yémen) a-t-elle été incendiée ? In Écriture de l'histoire et processus de canonisation dans le monde musulman des premiers siècles de l'Islam. Hommage à Alfred-Louis de Prémare (Revue des Mondes musulmans et de la Méditerranée, 129), Jul. 2011: 93–118. DOI: 10.4000/remmm.7101.
• Robin, C.J., 2014. Le roi ḥimyarite Thaʾrān Yuhanʿim (avant 325-v. 375). Stabilisation politique et réforme religieuse. Jerusalem Studies on Arabic and Islam 41: 1–95.
• Robin, C.J. & U. Brunner. 1997. Map of Ancient Yemen [1:1,000,000]. Munich: Staatliches Museum für Völkerkunde.
• Rösch M. & E. Fischer, 2013. Charred plant remains. In Yule (ed.) 2013: 187–193.
• Schiettecatte, J. 2011. D’Aden à Zafar. Villes de l’Arabie du Sud préislamique (Orient et Méditerranée, Archéologie, 6). Paris: De Boccard.
• Sima, A. 2020. Die sabäischen Inschriften aus Ẓafār. Aus dem nachgelassenen Manuskript herausgegeben von Norbert Nebes und Walter W. Müller (Jenaer Beiträge zum Vordereen Orient, 10). Wiesbaden: Harrassowitz.
• Uerpmann, H.-P. & M. Uerpmann, 2013. Animal remains from Ẓafār, in Yule (ed.) 2013: 194–219.
• Yule, P. 2009. Ẓafār photo archive, Heidelberg University Library, heidICON pool "SKVO Zafar", since 15.11.2009: http://www.ub.uni-heidelberg.de/Englisch/helios/digi/heidicon_zafar.html
• Yule, P. 2013. Late Antique Arabia Ẓafār, capital of Ḥimyar, rehabilitation of a ‘decadent’ society, excavations of the Ruprecht-Karls-Universität Heidelberg 1998–2010 in the highlands of the Yemen (Abhandlungen Deutsche Orient-Gesellschaft, 29). Wiesbaden: Harrassowitz.

_{Alternate spellings: Ẓafâr, Zafār, Zafâr, Zafar, Ẓfr}