To whom do the dead belong?
Preliminary observations on the cemetery of Tsomar, eastern Tigray (Inscriptiones Arabicae Aethiopiae 2)
p. 647-672
Résumés
Islamic networks played a critical role in the regional integration of the Ethiopian highlands to the Red Sea basin, after the eighth-century collapse of the kingdom of Axum. However, the history of Ethiopia’s earliest Islamic settlements is still poorly known. Knowledge has increased over the last two decades, with the identification of several Islamic sites with the help of the local Muslim communities’ memory. In 2018‑2019, a cluster of several ancient Muslim cemeteries was surveyed by a French-Ethiopian team in the area of Arra (eastern Tigray). The preliminary results of the survey of Tsomar, a Muslim cemetery in use in the late 13th century, are presented here. Since no excavations have yet taken place, the study is based on surface observation, drone zenithal views and the analysis of nine Arabic inscriptions found on the Tsomar site.
Les réseaux islamiques ont joué un rôle déterminant dans l’intégration des hauts plateaux éthiopiens au bassin de la mer Rouge après l’effondrement du royaume d’Axum au viiie siècle. L’histoire des premiers établissements islamiques d’Éthiopie n’en reste pas moins mal connue. Depuis deux décennies, plusieurs sites islamiques jusque-là ignorés ont pu être identifiés en Éthiopie en s’appuyant sur la mémoire des communautés musulmanes locales. C’est ainsi qu’en 2018‑2019 un ensemble de plusieurs cimetières musulmans anciens a été étudié par une équipe franco-éthiopienne dans la région d’Arra (Tigray oriental). Les résultats préliminaires de l’étude du cimetière de Tsomar, actif à la fin du xiiie siècle, sont présentés dans cet article. En l’absence de fouilles, l’étude s’appuie sur des observations de surface, des vues zénithales prises par drone et sur l’analyse de neuf inscriptions arabes retrouvées sur le site de Tsomar.
Entrées d’index
Mots-clés : Islam, Éthiopie, Tigray, cimetière, marqueur de tombe, épigraphie arabe, stèle funéraire, Coran, onomastique arabe
Keywords : Islam, Ethiopia, Tigray, cemetery, tomb marker, Arabic epigraphy, funerary stele, Qur’an, Arabic onomastics
Texte intégral
Introduction: in search of early Islam in the Ethiopian highlands
1The quest for a better understanding of the history of early Islam in the Ethiopian highlands has made significant progress in the past two decades.1 Several Islamic sites have been surveyed, and some excavated, from the Tchertcher mountains in eastern Ethiopia and the eastern cliff of the central highlands, up to eastern Tigray in the northern part of the country.2 Some of them have been identified by chance during searches for earlier stages of occupation. For example, the presence of Axumite remains drew attention to sites marked by centuries of continuous settlement.3 More recently, however, evidence of previously unknown Islamic settlement has been brought to light through the collective memory of local Muslim communities.4 This is not meant to imply that present-day Muslims are perforce the direct heirs of their fellow-believers who lived in the same area centuries ago, nor that their “memory” of their predecessors is an undistorted version of the latter’s actual history. Although one should always be wary of discourses of precedence and seniority based on local memory, especially in a country that has experienced major demographic disruptions in the past five centuries, the appropriation of the past can provide valuable clues in the search for forgotten histories.
2This approach has proven particularly fruitful in eastern Tigray, where the memory of early Islam has long been overshadowed by the history of the Christian kingdom of Ethiopia, to which the area was of critical importance for centuries.5 Archaeological surveys were conducted by a French-Ethiopian team in both areas of Kwiḥa and Arra (eastern Tigray) in 2018‑2019, with the support of the Authority for Research and Conservation of the Cultural Heritage (ARCCH, Addis Ababa), the Tigray Culture and Tourism Bureau (TCTB, Mekelle) and the ERC project HornEast.6 They resulted in promising findings, some of which are introduced in the present paper.7
3The village of Arra (eastern Tigray, district of Hintalo Wejjerat) is located at the eastern end of the plain of Adigudom, at an altitude of 2,200 m on the volcanic ridge overlooking the rolling hills that slope down toward the Danakil depression (fig. 1). The village is known for its church, dedicated to Mikael, which was built by king Yoḥannǝs IV (r. 1872‑1889) to thank its inhabitants for their help during his previous exile.8 However, Arra is currently inhabited by both Muslims and Christians. The formers claim the spiritual legacy of Faqīḥ Muḥammad, a leading figure who played a major role in the reintroduction of Islam to Tigray, according to oral tradition collected by the team, and to whom the building of several mosques in the village is credited. Faqīḥ Muḥammad most likely lived in the beginning of the 19th century and was buried in Arra. His funerary mosque spurred the formation of a large Muslim cemetery. The critical importance of Arra in the history of Islam in the area, its value until recently as a place for Muslim pilgrimage (ziyāra), as well as its location at the centre of a network of Muslim communities associated to the descent of Faqīḥ Muḥammad and scattered throughout the Tigray, came to the attention of the authors in 2018 through a personal communication from Dr Fesseha Abraham of Addis Ababa University. In addition to the figure of Faqīḥ Muḥammad, Dr Fesseha Abraham emphasised the large extent of Muslim cemeteries in Arra and its vicinity, and the previous recording of Arabic funerary inscriptions on the tombs.
Fig. 1 – Map of south-eastern Tigray (Ethiopia) showing the location of medieval Muslim cemeteries (S. Dorso).

4Local informants met by the team in December 2018 and October 2019 (Shaykh ‘Abd al‑Raḥmān ‘Abd al‑Qādir, imām of Adigudom, and Shaykh Muḥammad ‘Alī Kahsay, his deputy in Arra) indeed remembered that funerary stelae with Arabic inscriptions were collected decades ago by Muḥammad Saraj, an Ethiopian Muslim scholar foreign to the village, and deposited with the “old mosque” of Arra. An interview with Muḥammad Muḥammad Saraj, imām of Godjabele (southern Tigray), conducted by the team in October 2019, confirmed the latter information. According to him, his father, the muftī Muḥammad Saraj, graduated from al‑Azhar University (Egypt) around 1350 H/1931 AD and being related by marriage to Faqīḥ Muḥammad’s family, came to Arra and there collected ancient Muslim funerary stelae. Searches undertaken by the team in, on top of, and around the “old mosque” of Arra proved unsuccessful. However, Shaykh Muḥammad ‘Alī Kahsay, head of Arra’s Muslim community, has been keeping a Muslim funerary stele in his custody, which was dated to 584 H/1189 AD, along with a fragment from a second stele, most likely of the same period, that may come from Muḥammad Saraj’s collection, and showed them to the team in December 2018. These were the first clues to the very existence of an ancient Muslim community in Arra or its vicinity, prior to Faqīḥ Muḥammad’s reintroduction of Islam in the area in the beginning of the 19th century.
5In the absence of any excavations in Arra, no evidence of a medieval stage of village occupation has been brought to light thus far. Even the 2019 discovery of an Arabic stele dated to 551 H/1156 AD on top of a tumulus in Qelqel Rway, an area located in the northern part of the village, cannot be considered decisive evidence, since it might have been displaced and reused to adorn the structure. A survey of Ḥabera Mountain, located about 3 km north-east of Arra as the crow flies and 200 m lower, on the other side of the May Zelefo valley, and of its surroundings, revealed a substantially different picture. Conducted by the team in two stages, in December 2018 and October 2019, under the guidance of Shaykh ‘Abd al‑Raḥmān ‘Abd al‑Qādir, Shaykh Muḥammad ‘Alī Kahsay and several other members of Adigudom Muslim community, it brought to light a group of five cemeteries (fig. 2). Some of them are currently used as stations for a Muslim pilgrimage circuit leading to the cemetery of Ḥabera, on top of Ḥabera Mountain. Three of these cemeteries (Meyda Zelegat and Ḥabera, visited in 2018 and 2019, and Tsomar, visited in 2019) still contain funerary stelae with Arabic inscriptions (tab. 1), evidencing the presence of Muslim people in the area in the course of the 13th‑14th centuries AD. With a total of 44 epigraphic stelae (including fragments) found during survey work in 2018 and 2019, the village of Arra and the Ḥabera Moutain area cemeteries have provided a collection of Arabic inscriptions equivalent to that of Bilet, a medieval Muslim cemetery located 30 km north of Arra, close to the modern city of Kwiḥa. Kwiḥa has been known since the beginning of the 19th century for its Arabic funerary stelae, however they were found out of context, and original location of the medieval cemetery was only identified by the team in March 2018.9 Together, both collections represent over 70% of the corpus of medieval Arabic inscriptions from inland Ethiopia (excluding the Dahlak islands).10
Fig. 2 – Arra and Ḥabera Mountain cemeteries location map (S. Dorso).

Tab. 1 – Arabic Muslim funerary stelae from medieval Tigray.
Location | Sub-location | Time of discovery | Number | Total |
Bilet (Kwiḥa) | found between 1937 and 2000 | 19 | ||
found in 2018 | 23 | |||
42 | ||||
Arra | Private collection | showed in 2018 | 2 | |
Qalqel Rway | found in 2019 | 1 | ||
Ḥabera Mountain | Tsomar | found in 2019 | 9 | |
Meyda Zelegat | found in 2018-2019 | 8 | ||
Ḥabera | found in 2018-2019 | 24 | ||
44 |
To whom do the dead belong?
6The number and proximity of the cemeteries located at the foot, sides and top of Ḥabera Mountain, raises a major question: to whom do the dead belong? Nearby Arra, likely inhabited in the 13th‑14th centuries, is about 3 km away as the crow flies – but any actual journey would have been much longer, over uneven ground. What would have driven Arra’s inhabitants to bury their dead in places located at several hours’ walk from home? A dozen of graveyards have been located in the immediate surroundings of the village, some of them possibly dating back to the Middle Ages. The most obvious answer would be to suppose that Ḥabera Mountain’s cemeteries were connected to another or several other settlements. However, no traces of ancient dwellings have been found in the area thus far, with the sole exception of a few ruined houses in Ona Addi, a narrow plateau overlooking the cemetery of Grät Weizero, about 2 km as the crow flies north-east of the foot of Ḥabera Mountain. Without excavation, these cannot be dated. It also cannot be excluded that Ḥabera Mountain’s cemeteries were used and visited by non-permanent settlers, such as nomadic groups trading between the Danakil depression and the highlands.
7Be that as it may, it is also not certain that the five Ḥabera Mountain cemeteries were all, and exclusively, used by ancient Muslim communities. Legal norms and common practice observed in the medieval period in both Christian and Muslim lands usually urged a strict separation of graves according to religious allegiance, with a few exceptions, generally in the case of communities retaining an existing cemetery after conversion to Islam.11 Was this the case at Ḥabera Mountain? In the absence of studies of Christian medieval cemeteries in Tigray, it is not possible to establish a comparative typology of Christian and Muslim tomb markers and thus to investigate the question.12 At this stage, it is worth noting that the modern Christian cemetery of Arra, located near Mikael church, presents the same type of tomb marker – that is, rectangular platforms – as the 11th‑12th century Muslim tombs excavated in Bilet cemetery in 2018.13 The mere fact that some of the five Ḥabera Mountain cemeteries are nowadays stages in the same seasonal pilgrimage circuit, followed by Muslim women, is not sufficient evidence to assume that they were all Muslim burial grounds in the 13th‑14th centuries.
8Bearing these uncertainties in mind, this paper aims to be an introduction to the broader study of Ḥabera Mountain, using as a case study Tsomar, one of the three cemeteries that preserve epigraphic stelae. Located at the northern base of the mountain, the Tsomar site was surveyed in October 2019 under the guidance of Shaykh Nūr Ḥusayn Muḥammad, member of Adigudom Muslim community. The following preliminary remarks are based on surface observations and drone photography only.14 A return to the field and test excavations may drive the authors to revise their provisional conclusions in the near future. Despite these restrictions, however, the data already collected in Tsomar cemetery are pioneering enough to deserve a first presentation. This includes the edition and translation of the unpublished Arabic inscriptions of nine stelae or fragments of stele found during the survey of Tsomar cemetery.
Tsomar cemetery: an overview
9The Tsomar site is located at the bottom of the valley (fig. 3), at an altitude of about 1,900 m, north of Ḥabera ridge (site coordinates: 13.27226/39.59100). It lies on a natural communication route along which the cemeteries of Meyda Zelegat, and possibly Grät Weizero, are also to be found. According to Shaykh Nūr Ḥusayn Muḥammad, the name “Tsomar” refers, in Tǝgrǝñña, to a local flower. The site looks mostly out toward the north-west, i.e. toward the hills and the depression of the Ha‑Imele River, where another cemetery surveyed in 2019, Maetso Wedi Gomo, is located.
Fig. 3 – General view of Tosmar cemetery from the south (Bilet Mission).

10In the course of the last decade, Tsomar cemetery has been bisected by a modern motor track oriented west-north-east (fig. 4). The northern part of the site lies against a natural depression created by a seasonal brook. Water and rain have eroded the upper ground and stones down to the bedrock, possibly damaging part of the cemetery. The southern part of the site stands 2 to 4 m above the track, on the northern slope of a very shallow mound partly covered by high grass and bushes (fig. 5). The cemetery currently extends over 0.15 ha. It is fairly likely, however, that it used to be larger before its destruction by the track and the seasonal rains. Its western part might also have been damaged by a crop field located to the west.
Fig. 4 – Tsomar cemetery bisected by the modern motor track (N. Baker).
Fig. 5 – View of the southern part of Tsomar cemetery (Bilet Mission).
11Two complete epigraphic stelae have been found standing in place in the southern part of the cemetery, along with six fragments of stele resting on the ground, all within a 15 m radius. In addition, two fragments belonging to the same stele have been found resting on the ground in the northern part of the cemetery, 2 m apart from each other: the break is unlikely a fresh one (fig. 6). All the stelae found in Tsomar were made of sandstone, unlike in Ḥabera and Meyda Zelegat cemeteries, which also included basalt blocks. Both basalt and sandstone are present in the area where they naturally outcrop. By decision of the ARCCH and TCTB experts, to prevent looting or damage, all the stelae from Tsomar have been moved to the mosque of Adigudom, in the custody of the local Muslim community (fig. 7).
Fig. 6 – Spatial dispersion of epigraphic stelae in Tsomar cemetery (S. Dorso).

Fig. 7 – Epigraphic stelae from Ḥabera, Meyda Zelegat and Tsomar cemeteries deposited in the mosque of Adigudom in October 2019 (Bilet Mission).
The tomb surface markers
12Several types of tomb surface markers have been observed in Tigray Muslim cemeteries in Bilet, Arra and Ḥabera Mountain: rectangular platform, circular marker, box-like marker (square chamber structure), nave-shaped marker and tumulus.15 Most of the surface markers in Tsomar cemetery belong to the low circular type, consisting of a single ring of stones with an average diameter of 2 m (fig. 8). Most commonly, these stones are orientated longitudinally toward the centre of the circle (subtype 1). However, it is possible to identify subtypes with surface markers built on two or three rows of stones (subtype 2). Sometimes, the surface marker is totally covered by a low pile of stones (subtype 3) and on a few occasions, a box-like structure is visible in the middle of the stone circle (subtype 4). Unlike in other neighbouring sites, it seems that in Tsomar circular markers were not associated with a central standing stele, whether inscribed or anepigraphic.
Fig. 8 – Typology of low circular tomb markers in Tsomar cemetery (S. Dorso).

13Circular surface markers are clearly present in the northern part of the cemetery, suggesting its continuation toward the south and most probably the destruction of the central part of the cemetery by the modern motor track (fig. 9). It was not possible to determine the total number of markers in this area, although there should be at least a dozen.
Fig. 9 – Tsomar cemetery: preliminary plan (S. Dorso).

14In the southern part, the overall aspect of the cemetery differs from the northern part. In addition to the common circular markers, it also includes two standing epigraphic stelae, the inscribed face of which is oriented west-south-west (fig. 10). It is noteworthy that both stelae were found on the highest point of the southern area, where vegetation is denser. Several large flagstones were also identified on the surface, at the bottom of both stelae, oriented east-west. If these flagstones marked the actual location of the graves, this would imply that the stelae were placed at the eastern end of the surface marker, with the inscription facing the interior of the latter. However, these flagstones were not perfectly aligned, and an examination of the larger drone view suggests that both stelae may have been part of nave-shaped enclosures located on the opposite side and partly eroded along the slope (fig. 11). This second hypothesis would place the stelae at the western end of the surface marker, with the inscription facing the exterior of the latter. By way of comparison, the only stele (dated to 431/1040) excavated in Bilet in connection with its surface marker was placed at its eastern end, and faced the exterior.16
Fig. 10 – The two standing stelae and surrounding flagstones in the southern part of Tsomar cemetery (Bilet Mission).
Fig. 11 – The two standing stelae, in red, and surrounding tomb surface markers: preliminary plan (S. Dorso).

15The position of the stelae and of both assumed markers nevertheless indicate that the two graves followed a rough east-west orientation, in line with the Muslim burial practice consisting in burying the dead facing the qibla (the direction of the Ka‘ba in Mekka),17 that is, facing the north when in Tsomar. Any small deviation between both graves may have been caused by the topography. The surface markers associated to the two standing stelae found in situ in the southern area therefore seem to differ from the more common circular markers observed in the rest of the cemetery. They may belong to a different phase of occupation, during which Muslims were buried on top of a cemetery which may or may not have been a Muslim burial ground. The varied architecture of the markers might also be unrelated to the religious belonging of the deceased, instead reflecting differing social status or a chronological evolution in tomb surface markers. In addition, it is noteworthy that the southern part of the cemetery preserved far more epigraphic stelae and stele fragments (8) than the northern area (2 fragments belonging to the same stele): this might point to a sectorisation of Muslim graves adorned with epigraphic stelae within the cemetery. That area also seems to have had a higher concentration of other nave-shaped markers.
16Without further cleaning or excavation, it is too early at this stage, on the sole basis of surface observation, to formulate any remarks on the internal organisation of the graves or the size and orientation of the pits. However, the deciphering and study of the Arabic inscriptions preserved on Tsomar’s epigraphic stelae already provide important information about the history of the cemetery.
The Arabic funerary inscriptions of Tsomar cemetery: edition and translation
17The October 2019 survey of Tsomar cemetery resulted in the discovery of 2 complete epigraphic stelae, 2 fragments belonging to the same incomplete stele, and 6 other fragments that cannot be related to each other. Based on the preserved sections of the epitaphs, the latter belong to at least 4 different stelae. The corpus of Arabic inscriptions of Tsomar cemetery thus consists in 9 different items, coming from at least 7 different stelae (fig. 12‑13). They are featured below in chronological order for the 3 dated inscriptions, then in decreasing length for the 4 others, since the variations in writing style on the stelae is too limited to provide evidence of relative dating. Order in which the stelae were discovered is specified by the letters TS (for Tsomar) followed by a number: this is the system by which they are currently preserved in the mosque of Adigudom. Qur’anic quotations in the text and translation are marked by superscript stars * at the beginning and end of each quotation. The English translations of the Qur’an are taken from M.A.S. Abdel Haleem’s edition.18
Fig. 12 – Stelae 1, 2, 3 from Tsomar cemetery (S. Dorso/Bilet Mission).

Fig. 13 – Stelae 4, 5, 6, 7, 8, 9 from Tsomar cemetery (S. Dorso/Bilet Mission).

Stele 1
18Location: found standing in place in Tsomar cemetery, preserved in the mosque of Adigudom under the number TS02.
19Material: sandstone.
20State of preservation: complete, 9 lines readable.
21Forms of writing: transition style, Naskhi script incised, no diacritical dots.
22Decoration: none.
23Dimensions: visible height: 24‑29 cm; width: 35‑42 cm; thickness: 4.5‑7 cm.
24Previous publication: unpublished.
25Photography: fig. 12: 1.
١. بسم الله الرحمن الرحيم
٢. *كل نفس ذائقة الموت وانما توفون اجوركم
٣. يوم القيامة فمن زحزح عن النار وادخل (sic)
٤. فقد فاز وما الحياة (sic) الا متاع الغرور* توفيت
٥. زهرة بنت عبرة الى رحمة الله تعالى توفيت يوم
٦. الخميس في شهر شوال يوم الخامس والعشرون
٧. من شهور سنة ستمائة وثمانية وثمانين من هجرة
٨. النبوة غفر الله لها ولجميع المسلمين امين
٩. على النبي افضل السلام
1. In the name of God, the Lord of Mercy, the Giver of Mercy
2. *Every soul will taste death and you will be paid in full only
3. on the Day of the Resurrection. Whoever is kept away from the Fire and admitted to (the Garden)
4. will have triumphed. The (present) world is only an illusory pleasure* (Qur’an, 3, 185).
5. Zahra bt. ‘Ubra was called back to the mercy of Almighty God. She died on
6. Thursday the 25th of the month of Shawwāl
7. of the year 688 of the hijra
8. of the Prophet (11 November 1289). May God forgive her and all the Muslims, Amen!
9. Upon the Prophet the best salutation!
26Remarks:
27l. 1. The basmallah only covers two thirds of the space available.
28l. 2‑4. The Qur’anic quotation (Qur’an, 3, 185) has been twice misstated: two words are missing. The word الجنة (“the garden”) is missing between وادخل (“admitted to”) and فقد فاز (“will have triumphed”). The word الدنيا (“present”) is missing between الحياة (“the world”) and الا متاع الغرور (“an illusory pleasure”).
29l. 5. The name of the deceased’s father reads عںره. The reading عُبرة (‘Ubra) we suggest here is tentative. In his Kitāb al‑Ikmāl, Ibn Mākūlā (d. 1082?) provides the following names formed on the same ductus: عُبرة (‘Ubra), عَترة (‘Atra), عِترة (‘Itra), عُترة (‘Utra), عَنَزَة (‘Anaza), or غِيَرة (Ghiyara).19
30l. 6‑7. The wording of the date is grammatically incorrect: the century’s number is given before that of the year and of the decade. Shawwāl 25th 688 did not fall on a Thursday but on a Friday.
Stele 2
31Location: found out of context resting on the ground of Tsomar cemetery, preserved in the mosque of Adigudom under the number TS10.
32Material: sandstone.
33State of preservation: lower left fragment, 3 lines partly readable.
34Forms of writing: transition style, Naskhi script incised, no diacritical dots.
35Decoration: none.
36Dimensions: height: 22 cm; width: 32 cm; thickness: 4‑7 cm.
37Previous publication: unpublished.
38Photography: fig. 12: 2.
١. [...] [شهـ]ـر المحر[م] [...]
٢. [...][سـ]ـنة احد وتسـ[ـعين]
٣. [...] [محمـ]ـد النبي واله افضل السلم (sic)
1. […] of the month of al‑Muḥarram […]
2. […] in the year [?]91
3. […] (upon) Muḥammad the Prophet and his family the best salutation
39Remarks:
40l. 1‑2. The reading of the date is incomplete: century is missing. Despite some discrepancies in the size and design of the letters, the stele’s script appears to be closer to the two complete stelae found in its immediate vicinity in Tsomar (respectively dated to Shawwāl 688 and Muḥarram 691), than to the two complete stelae preserved in Arra (respectively dated to Jumādā II 551 and Dhū l‑ḥijja 584). This, as well as the closeness of the years, strongly supports the reading of the date as Muḥarram 691 (24 December 1291-22 January 1292).
41l. 3. The correct spelling of the last word should be السلام.
Stele 3
42Location: found standing in place in Tsomar cemetery, preserved in the mosque of Adigudom under the number TS01.
43Material: sandstone.
44State of preservation: complete, 8 lines readable, the left upper corner missing.
45Forms of writing: transition style, Naskhi script incised, no diacritical dots.
46Decoration: a six-branch star at the left end of the first line.
47Dimensions: visible height: 26 cm; width: 34 cm; thickness: 5.5 cm.
48Previous publication: unpublished.
49Photography: fig. 12: 3.
١. بسم الله الرحمن الرحيم ✡
٢. *كل من عليها فان ويبقى وجه
٣. ربك ذو الجلال والاكرام* هذا قبر عبد الله
٤. بن عبد الرحمن بن محمد بن عيسة توفي يوم الا
٥. ربعا السادس والعشرون من شهر
٦. المحرم من شهور سنة احد وتسعين
٧. وستمائة وصلى الله على رسوله سيدنا محمد
٨. واله وصحبه وسلم تسليما
1. In the name of God, the Lord of Mercy, the Giver of Mercy ✡
2. *Everyone on earth perishes; all that remains is the Face
3. of your Lord, full of majesty, bestowing honour* (Qur’an, 55, 26‑27). This is the tomb of ‘Abd Allāh
4. b. ‘Abd al‑Raḥmān b. Muḥammad b. ‘Īsa. He died on
5. Wednesday the 26th of the month of
6. al-Muḥarram from the year 691 (18 January 1292)
7. May God bless His messenger, our lord Muḥammad,
8. his family and his companions, and grant them the strongest salvation.
50Remarks:
51l. 4: the name of the deceased’s ancestor reads عيسة (‘Īsa) whereas the expected spelling of the name should be عيسى (‘Īsā).
52l. 5‑6: Muḥarram 26th 691 did not fall on a Wednesday but on a Friday.
Stele 4
53Location: two fragments found out of context resting on the ground of Tsomar cemetery, 2 m apart from each other, preserved in the mosque of Adigudom under the numbers TS04 and TS05.
54Material: sandstone.
55State of preservation: the stele has been broken in several fragments; the right lower corner is missing, along with a lacuna in the middle of the stele; the two remaining fragments cover about two thirds of the original and include seven lines partly readable.
56Forms of writing: transition style, Naskhi script incised, no diacritical dots.
57Decoration: none.
58Dimensions: TS04, height: 11‑15 cm, width: 24 cm, thickness: 4 cm; TS05, height: 3‑17 cm, width: 2‑18 cm, thickness: 4 cm.
59Previous publication: unpublished.
60Photography: fig. 13: 4.
١. بسم الله الرحمن الرحيم *كل نفس
٢. [ذائـ]ـقة الموت انما توفون اجوركم
٣. [يوم القيامة] فـ[ـمن] زحزح عن النار واد
٤. [خل الجنة فقد فاز وما]
٥. الحيوة الدنيا] الا (sic) الغرور* هذا
٦. [قبر] [...] بن ابي بكر توفا يوم الاربع
٧. [...] الله قبره
٨. [...] محمد النبي واله جمـ[ـيعا]
1. In the name of God, the Lord of Mercy, the Giver of Mercy. *Every soul
2. will taste death and you will be paid in full only
3. (on the Day of the Resurrection). Whoever is kept away from the Fire and
4. (admitted to the Garden will have triomphed)
5. (The present world) is only an illusory (pleasure)* (Qur’an, 3, 185). This is
6. (the tomb of) […] b. Abī Bakr. He died on Wednesday
7. […] God his tomb
8. […] Muḥammad the Prophet and his whole family
61Remarks:
62l. 1-5. The Qur’anic quotation (Qur’an, 3, 185) can be restored on the basis of the extant. It is however misstated: on l. 5, the word متاع (“pleasure”) is missing between the words الا (“only”) and الغرور (“illusory”).
Stele 5
63Location: found out of context resting on the ground of Tsomar cemetery, preserved in the mosque of Adigudom under the number TS07.
64Material: sandstone.
65State of preservation: upper left fragment, 4 lines partly readable.
66Forms of writing: transition style, Naskhi script incised, no diacritical dots.
67Decoration: none.
68Dimensions: height: 14 cm; width: 23 cm; thickness: 5 cm.
69Previous publication: unpublished.
70Photography: fig. 13: 5.
١. [بسم] الله الرحمن الرحيم
٢. * [كل من عليهـ]ـا فان ويبقى
٣. [وجه ربك ذو] الجلال واكرام *
٤. [...] الى رحمة
1. (In the name of) God, the Lord of Mercy, the Giver of Mercy
2. *(Everyone on earth) perishes; all that remains
3. (is the Face of your Lord), full of majesty, bestowing honour* (Qur’an, 55, 26‑27)
4. […] to the mercy of
71Remarks:
72l. 2‑3. The Qur’anic quotation (Qur’an, 55, 26‑27) can be restored on the basis of the extant letters.
Stele 6
73Location: found out of context resting on the ground of Tsomar cemetery, preserved in the mosque of Adigudom under the number TS08.
74Material: sandstone.
75State of preservation: right upper fragment, few letters readable in the beginning of the three first lines.
76Forms of writing: transition style, Naskhi script incised, no diacritical dots.
77Decoration: none.
78Dimensions: height: 10 cm; width: 13 cm; thickness: 5 cm.
79Previous publication: unpublished.
80Photography: fig. 13: 6.
١. بسم ا[لله الرحمن الرحيم]
٢. *كل من [عليها فان ويبقى]
٣. و[جه ربك ذو الجلال والاكرام]*
1. In the name of God, (the Lord of Mercy, the Giver of Mercy)
2. *Everyone (on earth perishes; all that remains is)
3. the Face (of your Lord, full of majesty, bestowing honour)* (Qur’an, 55, 26‑27).
81Remarks:
82l. 1. The basmallah can be restored on the basis of the extant letters.
83l. 2‑3. The Qur’anic quotation (Qur’an, 55, 26‑27) can be restored on the basis of the extant letters.
Stele 7
84Location: found out of context resting on the ground of Tsomar cemetery, preserved in the mosque of Adigudom under the number TS06.
85Material: sandstone.
86State of preservation: left upper fragment, one word readable on l. 1, few letters on l. 2.
87Forms of writing: transition style, Naskhi script incised, no diacritical dots.
88Decoration: a six-branch star at the left end of the first line.
89Dimensions: height: 11 cm; width: 9.5 cm; thickness: 5 cm.
90Previous publication: unpublished.
91Photography: fig. 13: 7.
١. [بسم الله الرحمـ]ـن الرحيم ✡
٢. [كل من عليها] فان
1. (In the name of God, the Lord of Mercy), the Giver of Mercy ✡
2. *(Everyone on earth) perishes (Qur’an, 55, 26)
92Remarks:
93l. 2. The restoration of the Qur’anic quotation (Qur’an, 55, 26) is hypothetical.
Stele 8
94Location: found out of context resting on the ground of Tsomar cemetery, preserved in the mosque of Adigudom under the number TS03.
95Material: sandstone.
96State of preservation: upper fragment, few letters partly readable on the first line.
97Forms of writing: transition style, Naskhi script incised.
98Decoration: plain lines sketching the form of a miḥrāb on the top of the stele.
99Dimensions: height: 8 cm; width: 18 cm; thickness: 5 cm.
100Previous publication: unpublished.
101Photography: fig. 13: 8.
١. [بسم الـ]ـله الـ[ر]حـ[من] الـ[رحيم]
1. (In the name of) God, the (Lord of Mercy), the (Giver of Mercy)
102Remarks:
103l. 1. The basmallah can be restored on the basis of the extant letters.
Stele 9
104Location: found out of context resting on the ground of Tsomar cemetery, preserved in the mosque of Adigudom under the number TS09.
105Material: sandstone.
106State of preservation: lower left fragment, two letters readable.
107Forms of writing: transition style, Naskhi script incised.
108Decoration: none.
109Dimensions: height: 11 cm; width: 6 cm; thickness: 5 cm.
110Previous publication: unpublished.
111Photography: fig. 13: 9.
١. [...] [السلـ]ام
1. […] salutation
112Remarks:
113The suggested restoration of the text is based on the position of the two readable letters (ام) at the very end of the epitaph.
The Arabic funerary inscriptions of Tsomar cemetery: preliminary remarks
114The 9 Tsomar inscriptions are part of a larger corpus of 41 epigraphic stelae, or fragments of stele, collected in three different Ḥabera Mountain cemeteries in 2018‑2019 (Ḥabera, Meyda Zelegat and Tsomar). Since study of the whole corpus still is in progress and in the absence of external points of comparison, the following remarks cannot be but preliminary. Tsomar cemetery, however, likely developed on its own for specific reasons at the northern base of the mountain. It might be therefore of interest to consider its epigraphic stelae as a sub-corpus and to compare the data they provide to each other.
Material, decoration, script
115All the epigraphic stelae or fragments of stele collected in Tsomar use local sandstones as material. Unlike the neighbouring cemeteries of Ḥabera and Meyda Zelegat, there is no evidence in Tsomar of the use of basalt blocks for epigraphic purpose. By way of comparison, 93 percent of the 10th‑13th‑century epitaphs in Bilet cemetery (30 km north of Arra) were engraved on basalt blocks: the Bilet stelae include only three fragments on sandstone.20 Sandstone’s relative fragility, along with deliberate or accidental destruction during road clearance, likely explains the poor preservation of Tsomar stelae, most of which are in fragments, with the sole exception of the two stelae found still standing. No tool marks have been observed: sandstone blocks were probably not prepared but simply selected from the surface for their smooth face, suitable for engraving.
116The Tsomar stelae are noticeable for their decoration when compared with the whole corpus of Arabic inscriptions from Tigray. The plain lines on the top of Stele 8 may indicate a miḥrāb form, though the missing upper portion precludes a definitive assessment (fig. 13: 8). Its closed shape is a contrast to the arch, or ansa, that crowns several stelae from Bilet cemetery and can also be identified on a single fragment from Ḥabera cemetery. Stelae 3 and 7 feature another decoration worthy of note (fig. 13: B): a six-pointed star engraved at the end of the basmallah opening the epitaph (left end of the first line). On Stele 3 (dated to 691/1292), the six-pointed star is 4 cm high and 3 cm width. On Stele 7 (with no date preserved), it is 3.5 cm high and 3 cm width. The six-pointed star is attested in another case within the Ḥabera Mountain corpus: a stele (dated to 658/1260) from Meyda Zelegat cemetery (2 km south-east of Tsomar) is adorned with a line of three irregular six-pointed stars engraved below the epitaph. The six-pointed star pattern has been used in the Muslim funerary context in Ethiopia since at least the 11th century. Three stelae from Bilet cemetery (one dated to 399/1009, the two others most likely belonging to the 5th/11th century) are adorned with a six-pointed star, in two cases framed by an arch heading the epitaph, in one case engraved within the epitaph and visually separating the Qur’anic quotation on one hand from the name of the deceased on the other.21 It was still in use in the 14th century, as evidenced by the so-called “Tomb of Ḥajji Mänsur” (near Kedä Bura, Ifat), a quadrangular monolith whose four corners are engraved with a six-pointed star.22 The variation in position and number of six-pointed stars on Ethiopian Muslim funerary monuments should not overshadow the fact that, in both cases in Tsomar, the pattern is engraved only once and at the same place, ending the basmallah at the left end of the first line.
117Islamic culture has long linked the six-pointed-star pattern to the “seal of Solomon” (khātam Sulaymān), whom the Qur’anic narrative casts as master of the jinns. The “seal” was the tool used by the king-prophet to maintain his power over his supernatural assistants. Considered a talisman, the “seal of Solomon” usually adorned amulets and drinking cups. Engraved on funerary monuments, it would help the deceased to face judgement and be spared tomb’s torments.23 The earliest attestation of the six-pointed star in an Islamic context comes from a graffito dated to 92/710 written on a wall in the Umayyad palace of al‑Kharrāna, Jordan.24 It is found on Muslim funerary stelae from the second half of the 3rd/9th century onwards in Egypt and Nubia, and is particularly common (in 1 up to 7 copies) on the 10th‑12th‑century stelae of the mining site of Khor Nubt, in the eastern desert of present-day Sudan.25 Though the pattern of the six-pointed star on funerary stelae once belonged to the broader north-eastern African Islamic area, it is noteworthy that Ethiopian Christian people have also long known and used the “seal of Solomon”, here as an eight-pointed star, in both magical and funerary contexts.26 At this stage of the study, however, one can assume neither mutual influence between Ethiopian Muslim and Christian peoples nor cultural borrowing from one to the other.
118Tsomar stelae all display plain, if not crude, Arabic script, without diacritical dots. It can be described as a cursive transition style belonging to the Naskhi script, although the carving technique by incision often results in quite angular letters. However, several hints, such as the cursive form of the letters sīn and shīn deprived of teeth or the cursive superposition of lām and jīm, point to a regular use of writing by the lapicides. Stele 5 script is particularly crude. The exception is Stele 2 (likely dated to Muḥarram 691/24 December 1291-22 January 1292), written with care, though without diacritical dots (fig. 13: A). Letters also are larger and more regular on Stele 2 than on Stele 3, dated to Muḥarram 691/January 1292. If one presumes that the suggested datation of Stele 2 is correct, this would mean that two different hands were active at the same time in relation with Tsomar Muslim community. One cannot exclude either that the making of Stele 2 could have been ordered elsewhere than in the immediate surroundings of Ḥabera Mountain. No systematic comparison can be undertaken with the stelae from the neighbouring cemeteries of Meyda Zelegat and Ḥabera, the study of which is still in progress. However, it is worth noting that the corpus of Arabic funerary inscriptions from Tigray do not include so far any sample of champlevé, but only incised inscriptions, a plain carving technique pointing to local craftmanship.27 The only example found in Tigray of a monumental Arabic inscription sculpted in champlevé is a fragment discovered in Wuqro (65 km north of Arra), which likely belonged to a mosque’s foundation inscription.28
Epitaphs
119Arabic inscriptions on Tsomar funerary stelae all include elements pertaining, with few variations, to the same type of Muslim epitaph, which is also found on the stelae in Bilet cemetery and which was a norm in the entire Islamicate world in the Middle Ages. Each epitaph opens with the basmallah, followed by a quote from the Qur’an, or in rare cases verses of poetry. The quotation is followed by a formula introducing the name of the deceased, then the death date (in most cases, to the day), and a final eulogy calling salvation upon Prophet Muḥammad.
Words of Islam in Tsomar
120The use of the basmallah (“In the name of God, the Lord of Mercy, the Giver of Mercy”) as epitaph’s opening is present on all the Tsomar stelae whose first line is preserved. The basmallah is ended in two instances by a six-pointed star (fig. 12: 3; fig. 13: 7, B).
121The Qur’anic quotation, the second element of the epitaph, usually begins on the second line. In Stele 1, the last third of the first line after the basmallah was left empty, and the Qur’anic quotation begun on the second line (fig. 12: 1). Stele 4 is the exception: the first two words of the Qur’anic quotation were engraved at the left end of the first line.
122A Qur’anic quotation is at least partly legible on 6 of the 9 stelae. Two different quotations are evidenced in Tsomar, both of which with a clear funerary meaning: Qur’an, 3, 185 (“Every soul will taste death and you will be paid in full only on the Day of Resurrection. Whoever is kept away from the Fire and admitted to the Garden will have triumphed. The present world is only an illusory pleasure”) and Qur’an, 55, 26‑27 (“Everyone on earth perishes; all that remains is the Face of your Lord, full of majesty, bestowing honour”). Qur’an, 3, 185 is found on Stelae 1 (dated to 688/1289) and 4 (no date preserved). Qur’an, 55, 26-27 is found on Stelae 3 (dated to 691/1292), 5, 6 and 7 (no date preserved). Qur’anic quotations in Tsomar are quite stereotypical when compared to the epitaphs in Bilet cemetery. Qur’an, 55, 26‑27 and Qur’an, 3, 185 appear in Bilet on 40 percent of the 32 stelae which have conserved a Qur’anic quotation. However, 13 different quotations are evidenced on Bilet stelae, among which several have no explicit funerary meaning.29 This is especially the case of Qur’an,112 (“Say, ‘He is God the One, God the eternal. He begot no one nor was He begotten. No one is comparable to Him’.”) which was found on seven occasions in Bilet (22 percent of the total quotations). Note that Qur’an, 112 is also found on several stelae from the cemeteries of Meyda Zelegat and Ḥabera neighbouring Tsomar.
123In addition, Qur’anic quotations are often misstated on Tsomar stelae, which was barely ever the case in Bilet. Though nothing much can be inferred from highly fragmentary epitaphs (Stelae 5, 6 and 7), it is noteworthy that the quotation of Qur’an, 3, 185 is incorrect on both Stelae 1 (two words missing in two different places) and 4 (one word missing). Conversely, the shorter quotation of Qur’an, 55, 26-27 is correctly engraved on Stele 3.
124The final eulogy closing the epitaph is preserved in Tsomar on 5 of 9 stelae (including Stele 9 on which only the two final letters are readable, which makes the restoration of the text hypothetical). In the other 4 instances, epitaphs conclude by calling God’s salvation or the reader’s salutation (salām, in both cases) upon the Prophet Muḥammad, sometimes along with his family and companions, and/or upon the whole community of the Muslims. In one case (Stele 1), the final eulogy is preceded by a prayer for the deceased (“May God forgive her and all the Muslims, Amen!”).
Arabic names in Tsomar
125The name of the deceased is (at least partly) preserved on 3 of the 9 stelae (Stelae 1, 3 and 4). In two cases (Stelae 3 and 4), the name is introduced by the formula “This is the tomb of” (“Hadha qabr”). The epitaph on Stele 1, however, introduces the name of the deceased with the sentence “Was called back to the mercy of Almighty God” (“Tuwufiyyat … ilā raḥmat Allāh ta‘ālā”). The same may be restored on Stele 5, which does not preserve the name of the deceased. Both formulas are standard in the Islamicate world.
126In one case out of the three stelae for which the deceased’s gender is known, the deceased is a woman (Stele 1). This can be compared with the Bilet stelae, which include 12 women out of 29 documented cases. The woman of Stele 1 was named Zahra, a common female Arabic name meaning “flower”. One of the Prophet Muḥammad’s daughters, Fāṭima, was traditionally nicknamed al‑Zahrā’ (“the Shining one”), an epithet belonging to the same root (Z.H.R). The personal name of the deceased of Stele 3, ‘Abd Allāh (“the Servant of God”), belongs to the repertoire of the most common Arabic Muslim names, a reference to the Prophet Muḥammad’s father.
127The deceased’s genealogical name (his/her nasab) is preserved in the three instances. The deceased of the Stele 4 was named “[…], son of Abū Bakr”, another common Arabic Muslim name, a reference to the famous companion of and first successor to the Prophet Muḥammad, according to the Muslim tradition. The patrilineal genealogy of the deceased of Stele 3 is evoked over three generations: ‘Abd Allāh was “son of ‘Abd al‑Raḥmān son of Muḥammad son of ‘Īsa”. The deceased, or the patron of the stele, whoever he/she was, hence claimed that the former belonged (at least) to a fourth generation of Muslims in his family. The tentative reading of his ancestor’s personal name (‘Īsa) could be related to ‘Īsā (spelled with a final ā), the name the Qur’an gives to Jesus (spelled Yasū‘ in Christian Arabic). In this case, it is not impossible that the personal name ‘Īsa, provided the reading is correct, might have been a token of conversion from Christianity to Islam. As for Zahra, the female deceased of Stele 1, her genealogy includes only one generation. Her father’s name reads عںره, a ductus which we suggest to read عُبرة (‘Ubra), a male personal name actually recorded in Arabic dictionaries, but that is, to our knowledge, quite rare.30
128Nevertheless, the Tsomar epitaphs only preserve Arabic personal names. In contrast to the Bilet stelae, neither relation names (nisba) indicative of (geographical, ethnic or social) origin, nor local non-Arabic names are found. The corpus is too narrow, however, to come to any conclusions about the community’s origins, whether foreign or local.
Hijra dates and datation of Tsomar cemetery
129Dates are critically important to understand the Tsomar cemetery. Without excavation, its datation so far relies solely on the death dates provided by the epitaphs. Death dates, given to the day, are fully preserved on Stelae 1 and 3 and partly preserved on Stelae 2 and 4. The epitaph on Stele 1 specifies that the death date is given according to the Hijra calendar (“min hijrat al‑nubuwwa”). The use of month names belonging to the Hijra calendar leaves no doubt either for Stelae 2 and 3.
130Stele 4 only preserves the name of the day (“Wednesday”) on which the deceased died. Stele 2 preserves more critical elements: the month (“al‑Muḥarram”), the year and the decade (91); the century, however, is missing. Comparison of script with 5th‑6th/11th‑12th‑century stelae from Bilet cemetery, as well as with 6th/12th‑century stelae from Arra, precludes the possibility that Stele 2 might be dated to 491/1097 or 591/1194. The cursive script featured on Stele 2 most likely belongs to the 7th‑8th/13th‑14th centuries. So far, the latest date in the Ḥabera Mountain corpus comes from a Ḥabera cemetery stele dated to 782/1380. Though it is not impossible that Stele 2 may date to 791/1389, the dates provided by Stele 1 (688/1289) and Stele 3 (691/1292) in its immediate vicinity argue for 691/1291‑1292. In this case, Stele 2 would have been dated to Muḥarram 691/24 December 1291-22 January 1292.
131If that is the case, Tsomar stelae provide three dates very close to one another. Three deceased, whose tombs were adorned by epigraphic stelae, would have been buried in the same area in the southern part of the cemetery in the span of two or so years, between November 1289 and January 1292, and two of them during the same month of Muḥarram 691 (24 December 1291-22 January 1292). It is noteworthy that the months of November, December and January coincide in the area with the harvest season. The death dates provided by these three stelae does not necessarily mean that most of burials in Tsomar cemeterey occurred in the late 13th century. However, it does suggest that around 1290, Tsomar cemetery witnessed several burials of Muslim individuals whose social status was reflected in the making of an epigraphic stele. A remarkably similar situation is observed in the neighbouring cemetery of Meyda Zelegat: three deceased, whose tombs were adorned with epigraphic stelae, were buried in the same area between 657/1259 and 658/1260. In addition, it should be stressed that, in Tsomar, the burials marked by Stelae 1 and 3, respectively dated to 688/1289 and 691/1291‑1292, are located within 1.5 m of each other (fig. 11), even though the deceased did not belong to the same patrilineage.
Conclusion
132Although unevenly preserved, the 9 epigraphic stelae of Tsomar are of critical importance to the investigation of the history of the cemetery. Incised with Arabic inscriptions, they evidence the burial of Muslim men and women in Tsomar around 1290, three times during the harvest season. Epitaphs reveal a community bearing Arabic names, using Arabic Naskhi cursive script, Qur’anic quotations and the Hijra calendar, even if misstatements in the use of the latter, as well as in the quotation of the Qur’an, suggest that they enjoyed uneven connections only with the wider Islamicate world. Neither Arabic names nor Muslim faith and burial practices are sufficient evidence to assume that this community was of foreign origin. Conversion to Islam of local inhabitants may have led as well to the adoption of the ones and the others. However, as long as no remains of an important settlement are found in its vicinity, it cannot be excluded that Tsomar cemetery was used by non-permanent settlers, such as nomads trading between the Danakil depression and the highlands. Be that as it may, a Muslim community was living in Tsomar’s vicinity, or was using its cemetery, located in a central area of the Christian kingdom, at the same time when the latter was undergoing major political changes. Indeed, Tsomar epigraphic stelae are dated about two decades after the advent of the Solomonic dynasty, which brought about a disruption in the balance of power in Tigray.31
133Tsomar, however, is not an isolated case but part of what one might call a “mountain of the dead” in the Ḥabera area including several cemeteries, among which at least two other graveyards used by Muslims in the 13th century. The relation between these contemporary cemeteries is still unclear. Their location also raise several questions: why did these three cemeteries developed separately in a 2 km radius as the crow flies? Why did the cemeteries of Tsomar and Meyda Zelegat developed on low grounds seven decades after the earliest burial in Ḥabera, located on high ground and still in use until the late 14th century after an interruption of almost one century? It is likely that the ongoing study of the other cemeteries’ layout, inscriptions and tomb surface markers, will help to better understand Tsomar’s own development and the discrepancies observed between its northern and southern parts, nowadays bisected by the modern motor track. The relation of the Ḥabera Mountain cemeteries to their surroundings is also still unclear either. Evidence of permanent settlements, consistent in scale with the number of graves, is still missing in their immediate vicinity. Were the cemeteries used by distinct communities living permanently in remote parts of the area or by non-permanent settlers traveling long distances? In both cases, what was their relation with the nearby village of Arra, located about 3 km south-west as the crow flies? The three epigraphic stelae preserved or erected today in Arra suggest the presence of a Muslim community in the vicinity during the second half of the 12th century: however, two of them were found displaced in a private collection and the structure associated with the third is the only one of its kind. Were these three stelae recently displaced from Ḥabera Mountain and reused in Arra? Or do they rather suggest that a Muslim community was living earlier in Arra, whose presence may have attracted the Muslim people who buried their dead in the 13th century in the Ḥabera Mountain cemeteries? On a broader scale, can it be assumed that the Muslim people evidenced in Ḥabera Mountain cemeteries were part of a larger network of Muslim communities scattered in Tigray? Was there any connection with the Muslim community of Bilet, about 30 km north of Arra, whose cemetery was still in use around 1260?
134These preliminary observations on Tsomar cemetery raise more questions than they provide answers. A better understanding of the site implies a return to the field and test excavations. The appropriation of the Ḥabera Mountain cemeteries in the collective memory of present Muslim communities cannot conceal the fact that they also belong to an older stage of Muslim presence in the area. However, the question remains to figure out to whom the dead initially belonged.
135Muehlbauer 2021: M. Muehlbauer, “From Stone to Dust: The Life of the Kufic-Inscribed Frieze of Wuqro Cherqos in Tigray, Ethiopia”, Muqarnas 38, 2021, pp. 1‑34.
Bibliographie
Ancient sources
Ibn Mākūlā, ‘Alī b. Hibat Allāh, Al‑Ikmāl fī raf‘ al‑irtiyāb ‘an al‑mu’talif wa l‑mukhtalif fī al-asmā’ wa l‑kunā wa l‑ansāb, 8 vol., ed. N. al‑‘Abbās, Cairo, Dār al‑kitāb al‑islāmī, 1962.
Qur’an, transl. M.A.S. Abdel Haleem, Oxford/New York, Oxford University Press, 2004 (new translation).
Stelae TS01 to TS10 preserved in the mosque of Adigudom (Ethiopia, Tigray).
Modern sources
Ayenachew 2020: D. Ayenachew, “Territorial expansion and administrative evolution under the ‘Solomonic’ dynasty”, in S. Kelly (ed.), A companion to medieval Ethiopia and Eritrea, Leiden/Boston, Brill, 2020, pp. 57‑85.
Bauden 2011: F. Bauden, “Inscriptions arabes d’Éthiopie”, Annales Islamologiques 45, 2011, pp. 285‑306, https://www.ifao.egnet.net/anisl/045/15 (accessed 22/01/2021).
Chekroun, Hirsch 2020: A. Chekroun, B. Hirsch, “The sultanates of medieval Ethiopia”, in S. Kelly (ed.), A companion to medieval Ethiopia and Eritrea, Leiden/Boston, Brill, 2020, pp. 86‑112.
Conti Rossini 1937‑1939: C. Conti Rossini, “Necropoli musulmana ed antica chiesa cristiana presso Uogrì Haribà nell’Enderta”, Rivista degli studi orientali 17, 1937‑1939, pp. 399‑408.
Dege, Smidt 2011: S. Dege, W.G.C. Smidt, “Ṭälsäm”, in S. Uhlig (ed.), Encyclopaedia Aethiopica IV. O‑W, Wiesbaden, Harrassowitz, pp. 850‑852.
Derat 2020: M.‑L. Derat, “Before the Solomonids: crisis, renaissance and the emergence of the Zagwe dynasty (seventh-thirteenth centuries)”, in S. Kelly (ed.), A companion to medieval Ethiopia and Eritrea, Leiden/Boston, Brill, 2020, pp. 31‑56.
Derat et al. 2020: M.‑L. Derat, E. Fritsch, C. Bosc‑Tiessé, A. Garric, R. Mensan, F.‑X. Fauvelle, H. Berhe, “Māryām Nāzrēt (Ethiopia): the twelth-century transformations of an Aksumite site in connection with an Egyptian Christian community”, Cahiers d’études africaines 239, 2020, pp. 473‑507.
ERC Becoming Muslim 2019: ERC Becoming Muslim, “2019 fieldwork”, ERC project Becoming Muslim site, 2019, http://www.becomingmuslim.co.uk/2019/03/12/2019-fieldwork (accessed 01/10/2020).
Fauvelle‑Aymar et al. 2006: F.‑X. Fauvelle‑Aymar, B. Hirsch, L. Bruxelles, C. Mesfin, A. Chekroun, D. Ayenachew, “Reconnaissance de trois villes musulmanes de l’époque médiévale dans l’Ifat”, Annales d’Éthiopie 22, 2006, pp. 133‑175, https://www.persee.fr/doc/ethio_0066-2127_2006_num_22_1_1486 (accessed 23/01/2021).
Fauvelle‑Aymar, Hirsch 2004‑2010: F.‑X. Fauvelle‑Aymar, B. Hirsch, “Muslim historical spaces in Ethiopia and the Horn of Africa: a reassessment”, Northeast African Studies 11/1, 2004‑2010, pp. 25‑54.
Fauvelle‑Aymar, Hirsch, Chekroun 2017: F.‑X. Fauvelle, B. Hirsch, A. Chekroun, “Le sultanat de l’Awfāt, sa capitale et la nécropole des Walasma‘. Quinze années d’enquêtes archéologiques et historiques sur l’Islam médiéval éthiopien”, Annales Islamologiques 51, 2017, pp. 239‑295, https://www.ifao.egnet.net/anisl/051/12 (accessed 22/01/2021).
Gaastra, Insoll 2020: J.S. Gaastra, T. Insoll, “Animal economies and Islamic conversion in eastern Ethiopia: zooarchaeological analyses from Harlaa, Harar and Ganda Harla”, Journal of African Archaeology 18/2, 2020, pp. 181‑208, https://brill.com/view/journals/jaa/18/2/article-p181_3.xml (accessed 23/01/2021).
González‑Ruibal et al. 2017: A. González‑Ruibal, J. de Torres, M.A. Franco, A.M. Ali, A. Shabelle, C. Martínez, K. Aideed, “Exploring long distance trade in Somaliland (AD 1000‑1900): preliminary results from the 2015‑2016 field seasons”, Azania 52/2, 2017, pp. 135‑172.
Gutiérrez Lloret 2011: S. Gutiérrez Lloret, “Histoire et archéologie de la transition en al‑Andalus: les indices matériels de l’islamisation à Tudmīr”, in D. Valérian (ed.), Islamisation et arabisation de l’Occident musulman médiéval (viie‑xiie siècle), Paris, Éditions de la Sorbonne, 2011, pp. 195‑246, https://books.openedition.org/psorbonne/2514 (accessed 23/01/2021).
Hagos, Smidt, Rashidy 2013: T. Hagos, W.G.C. Smidt, M. Rashidy, “The two Arabic inscriptions of Mekelle Museum: a further contribution to the history of the eastern Tigrayan trade route (IV)”, ITYOPIS – Northeast African Journal of Social Sciences and Humanities 3, 2013, pp. 150‑153.
Hirsch 2018: B. Hirsch, “À la recherche des stèles musulmanes de l’Endertā. Les sites de Wager Ḥaribā et Kwiḥā”, ERC HornEast site, 2018, https://horneast.hypotheses.org/451 (accessed 01/10/2020).
Imbert 1995: F. Imbert, “Inscriptions et espaces d’écriture au palais d’al‑Kharrāna en Jordanie”, Studies in the History and Archaeology of Jordan 5, 1995, pp. 403‑416, http://publication.doa.gov.jo/Publications/ViewChapterPublic/1224 (accessed 22/01/2021).
Insoll 1999: T. Insoll, The archaeology of Islam, Oxford/Malden, Wiley‑Blackwell, 1999.
Insoll et al. 2017: T. Insoll, N. Khalaf, R. MacLean, D. Zerihun, “Archaeological survey and excavations, Harlaa, Dire Dawa, Ethiopia (January‑February 2017). A preliminary fieldwork report”, Nyame Akuma 87, 2017, pp. 32‑38, https://www.africanistarchaeology.net/s/NA-87-4-Insoll.pdf (accessed 22/01/2021).
Kaplan 2011: S. Kaplan, “Solomon”, in S. Uhlig (ed.), Encyclopaedia Aethiopica IV. O‑W, Wiesbaden, Harrassowitz, 2011, pp. 687‑688.
Loiseau 2020: J. Loiseau, “Retour à Bilet. Un cimetière musulman médiéval du Tigray oriental (Inscriptiones Arabicae Aethiopiae 1)”, in F. Imbert (ed.), Nouveaux itinéraires épigraphiques d’Orient et d’Occident – années 2018‑2019, published in BEO 67, 2020, pp. 59‑96.
Loiseau, Chekroun, Hirsch 2018: J. Loiseau, A. Chekroun, B. Hirsch, Archaeological survey around Igre Hariba (Tigray, Ethiopia). Fieldwork preliminary report (8‑15 March 2018) presented to the ARCCH (Addis Ababa), 2018 (unpublished).
Loiseau et al. 2019: J. Loiseau, S. Dorso, Y. Gleize, D. Ollivier, A. Chekroun, B. Hirsch, D. Ayenachew, Preliminary report. Excavations and surveys. Bilet (Tigray, Ethiopia). 1‑20 December 2018, presented to the ARCCH (Addis Ababa), 2019 (unpublished).
Loiseau et al. 2021: J. Loiseau, S. Dorso, Y. Gleize, D. Ollivier, D. Ayenachew, H. Berhe, A. Chekroun, B. Hirsch, “Bilet and the wider world. New insights into the archaeology of Islam in Tigray”, published in T. Insoll (ed.), Cosmopolitanism in medieval Ethiopia, Antiquity 95/380, 2021, pp. 508‑529, https://doi.org/10.15184/aqy.2020.163 (accessed 25/04/2022).
Oman, Grassi, Trombetta 1998: G. Oman, V. Grassi, A. Trombetta, The book of Khor Nubt. Epigraphic evidence of an Islamic-Arabic settlement in Nubia (Sudan) in the III‑IV centuries A.H./X‑XI A.D., Naples, Istituto Universitario Orientale, 1998.
Redlak 2008: M. Redlak, “Ornaments on funerary stelae of the 9th‑12th centuries from Egypt – Joseph Strzygowski’s publication anew”, Polish Archaeology in the Mediterranean 20, 2008, pp. 561‑574, https://pcma.uw.edu.pl/wp-content/uploads/pam/PAM_2008_XX/PAM_20_Redlak_561_574.pdf (accessed 22/01/2021).
Schneider 1967: M. Schneider, “Stèles funéraires arabes de Quiha”, Annales d’Éthiopie 7, 1967, pp. 107‑122, https://www.persee.fr/doc/ethio_0066-2127_1967_num_7_1_867 (accessed 23/01/2021).
Schneider 1983: M. Schneider, Stèles funéraires musulmanes des îles Dahlak (mer Rouge), 2 vol., Cairo, IFAO, 1983.
Schneider 2009: M. Schneider, “Des Yamāmī dans l’Enderta (Tigré)”, Le Museon 122/1‑2, 2009, pp. 131‑148.
Smidt 2004: W.C. Smidt, “Eine arabische Inschrift in Kwiḥa, Tigray”, in V. Böll, D. Nosnitsin, T. Rave, W. Smidt, E. Sokolinskaia (ed.), Studia Aethiopica. In honour of Siegbert Uhlig on the occasion of his 65th birthday, Wiesbaden, Harrassovitz, 2004, pp. 259‑268.
Smidt 2010: W.G.C. Smidt, “Another unknown Arabic inscription from the eastern Tigrayan trade route. Indication for a Muslim cult site during the ‘Dark Age’?”, in W. Raunig, A. Asfa‑Wossen (ed.), In memoriam Peter Roenpage. Juden, Christen und Muslime in Äthiopien – ein Beispiel für abrahamische Ökumene, published in Orbis Æthiopicus. Beiträge zur Geschichte, Religion und Kunst Äthiopiens 13, 2010, pp. 179‑191.
Tafla 2014: B. Tafla, “Yoḥannǝs IV”, in A. Bausi (ed.), Encyclopaedia Aethiopica V. Y‑Z, Wiesbaden, Harrassowitz, 2014, pp. 73‑80 (in cooperation with S. Uhlig).
Walker, Fenton 2012: J. Walker, P. Fenton, “Sulaymān b. Dāwūd”, in P. Bearman, T. Bianquis, C.E. Bosworth, E. van Donzel, W.P. Heinrichs (ed.), Encyclopaedia of Islam, s.l., Brill Online Reference Works, 2012 (2nd ed.), http://dx.doi.org/10.1163/1573-3912_islam_SIM_7158 (accessed 23/01/2021).
Notes de bas de page
1 This paper is part of the project HornEast that has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Grant agreement no. 726206). The authors thank the Authority for Research and Conservation of the Cultural Heritage (ARCCH, Addis Ababa), its Cultural Heritage Research Directorate’s director Demerew Dagne and its experts Yared Assefa, Bruk Jifara and Dawit Tibebu; the Tigray Culture and Tourism Bureau (TCTB, Mekelle), its former director Dawit Hailu, its present director Brekti Gebremedhin, its deputy director Zenebu Halefom and its expert Guesh Tsehaye; the Institute of Ethiopian Studies (IES, Addis Ababa University), its former director Ahmed Hassen Omer and its present director Takele Merid; the Centre français des études éthiopiennes (CFEE, Addis Ababa) and its director Marie Bridonneau; Fesseha Abraham (Addis Ababa University); Shaykh ‘Abd al‑Raḥmān ‘Abd al‑Qādir, imām of Adigudom, and Muḥammad ‘Alī Kahsay, his deputy in Arra; Shaykh Muḥammad Muḥammad Saraj, imām of Godjabele. They also thank Nicolas Baker (CNRS Images) for the drone views; Frédéric Imbert (Aix-Marseille University) for the rereading of the Arabic inscriptions’ edition and interpretation, and for his suggestions regarding their paleography; and Mathilde Montpetit for the final English editing of this paper. Finally, they thank the anonymous reviewers of this paper, whose remarks and advices proved to be highly fruitful. The mistakes and misinterpretations that may remain fall under the authors’ responsability.
2 Chekroun, Hirsch 2020, pp. 87‑89.
3 Conti Rossini 1937‑1939.
4 Fauvelle‑Aymar, Hirsch, Chekroun 2017; Insoll et al. 2017; Gaastra, Insoll 2020.
5 Derat 2020, pp. 41‑49.
6 Loiseau, Chekroun, Hirsch 2018; Loiseau et al. 2019.
7 See also Loiseau 2020; Loiseau et al. 2021.
8 Tafla 2014.
9 Conti Rossini 1937‑1939; Schneider 1967; Smidt 2004; Hagos, Smidt, Rashidy 2013; Hirsch 2018; Loiseau, Chekroun, Hirsch 2018; Loiseau 2020; Loiseau et al. 2021.
10 On the inscriptions of Dahlak Kabīr, see Schneider 1983. For an overview of medieval Arabic inscriptions from Ethiopia prior to the latest findings in Bilet, Arra and Ḥabera Mountain, see Fauvelle‑Aymar, Hirsch 2004‑2010, pp. 38‑41. Note that the British-Ethiopian archaeological mission working in Harlaa, eastern Ethiopia, under the leadership of T. Insoll, announced the finding of “several new Arabic inscriptions” during its 2019 campaign: ERC Becoming Muslim 2019.
11 Gutiérrez Lloret 2011.
12 Note that archaeological survey and excavations have been recently undertaken by another French-Ethiopian mission in the site of Nāzrēt, 17 km as the crow flies south of Arra, which underwent important transformations in the 12th century. See Derat et al. 2020.
13 Loiseau et al. 2021.
14 Drone photographs were taken in October 2019 by Nicolas Baker (CNRS Images).
15 Similar structures and patterns have recently been recorded in Somaliland by the Spanish team led by Jorge de Torres: González‑Ruibal et al. 2017.
16 Loiseau et al. 2021.
17 Insoll 1999, pp. 166‑200.
18 Qur’an (transl. Abdel Haleem 2004).
19 Ibn Mākūlā, Al-Ikmāl (ed. al‑‘Abbās 1962, vol. VI, pp. 296‑302).
20 Loiseau 2020.
21 Schneider 1967, p. 119, pl. LXVII; Smidt 2004, p. 261; Hagos, Smidt, Rashidy 2013, p. 153; Schneider 2009, p. 148.
22 Fauvelle‑Aymar et al. 2006, pp. 151‑152 and 172, photo 21; Bauden 2011, pp. 286‑293 and 304.
23 Walker, Fenton 2012.
24 Imbert 1995.
25 Redlak 2008, p. 568; Oman, Grassi, Trombetta 1998, pp. 175‑178.
26 Kaplan 2011; Dege, Smidt 2011.
27 Loiseau 2020.
28 Smidt 2010; Muehlbauer 2021.
29 Loiseau 2020.
30 Ibn Mākūlā, Al‑Ikmāl (ed. al‑‘Abbās 1962, vol. VI, p. 299).
31 Ayenachew 2020.
Auteurs
-
Julien Loiseau
ORCID : 0000-0002-4655-0329
Aix-Marseille Université, CNRS, IREMAM, ERC project HornEast
-
Simon Dorso
Aix-Marseille Université, CNRS, IREMAM, ERC project HornEast
-
Hiluf Berhe
Mekelle University; Collège de France
-
Deresse Ayenachew
Aix-Marseille Université, CNRS, IREMAM, ERC project HornEast
-
Amélie Chekroun
CNRS, IREMAM, Aix-en-Provence
-
Bertrand Hirsch
Université Paris 1 Panthéon-Sorbonne

Le texte seul est utilisable sous licence Creative Commons - Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International - CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.
The Iranian Plateau during the Bronze Age
Development of urbanisation, production and trade
Jan-Waalke Meyer, Emmanuelle Vila, Marjan Mashkour et al. (dir.)
2019
Les vaisseaux du désert et des steppes
Les camélidés dans l’Antiquité (Camelus dromedarius et Camelus bactrianus)
Damien Agut-Labordère et Bérangère Redon (dir.)
2020
Chantiers et matériaux de construction
De l’Antiquité à la Révolution industrielle en Orient et en Occident
Anne Baud et Gérard Charpentier (dir.)
2020
Multidisciplinary approaches to food and foodways in the medieval Eastern Mediterranean
Sylvie Yona Waksman (dir.)
2020
On salt, copper and gold
The origins of early mining and metallurgy in the Caucasus
Catherine Marro et Thomas Stöllner (dir.)
2021
Chalcis/Qinnasrin (Syrie)
De l’âge du Bronze à l’époque mamelouke. Qinnasrin II
Marie-Odile Rousset (dir.)
2021
Kition-Bamboula VIII
Le port de guerre de Kition
Olivier Callot, Sabine Fourrier et Marguerite Yon (dir.)
2022
Networked spaces
The spatiality of networks in the Red Sea and Western Indian Ocean
Caroline Durand, Julie Marchand, Bérangère Redon et al. (dir.)
2022
Kition-Bamboula IX
Les cultes des Salines à Kition : étude des terres cuites d'époque classique
Pauline Maillard
2023
Les espaces funéraires autour d’Apamée de l’Euphrate (iiie siècle avant-viiie siècle après J.‑C.)
Catherine Abadie-Reynal (dir.)
2023
Les théâtres antiques et leurs entrées
Parodos et aditus
Djamila Fellague et Jean-Charles Moretti (dir.)
2024