Navigation – Plan du site

AccueilNuméros153PREMIERE PARTIE : DOSSIER THEMATIQUE1. ArchaeologyThe Medieval Muslim Cemeteries of...

PREMIERE PARTIE : DOSSIER THEMATIQUE
1. Archaeology

The Medieval Muslim Cemeteries of Tigray (Ethiopia)

Preliminary Remarks on their Materiality (Architectural and Lapidary Aspects)*
Les cimetières musulmans médiévaux du Tigray (Éthiopie) : remarques préliminaires sur leur matérialité (aspects lapidaires et architecturaux)
مقابر المسلمين في العصور الوسطى في تيغراي (أثيوبيا). ملاحظات أولية عن ماديتها (المظاهر المعمارية والحجرية)
Simon Dorso et Anna Lagaron
p. 83-114

Résumés

La découverte (ou redécouverte) puis la fouille du cimetière musulman médiéval de Bilet, dans le Tigray en 2018, a permis d’étudier pour la première fois sur les hauts plateaux septentrionaux d’Éthiopie une architecture et des rites funéraires islamiques antérieurs au viiie/xive siècle. Les données récoltées à Bilet, précieuses pour la connaissance de l’Islam éthiopien d’avant les grandes conquêtes du xe/xvie siècle, peuvent désormais être comparées localement au corpus inédit des cimetières contemporains identifiés dans les environs d’Arra, qui ont livré un important matériel épigraphique mais n’ont pas fait l’objet de fouilles. Ce premier état des lieux des cimetières tigréens invite à envisager les sites dans un contexte régional plus étendu, en Éthiopie, mais également au-delà, dans l’ensemble de la Corne de l’Afrique et sur les rives de la Mer Rouge, espaces avec lesquels les épitaphes de Bilet attestent d’importants contacts. Cette contribution entend proposer des réflexions préliminaires sur la matérialité des cimetières tigréens à travers ses aspects lapidaires (sélection des supports, travail, gravure et mise en œuvre des stèles inscrites) et architecturaux (implantation des espaces funéraires, formes et organisation des marqueurs de surface, intégration des stèles inscrites aux marqueurs).

Haut de page

Texte intégral

  • * * This paper is part of a project that has received funding from the European Research Council (ERC (...)
  • 1 The important Muslim necropolis on Dahlak Kabir island (Schneider 1986, Oman 1976-1987, Insoll 2001 (...)

1Historians of early Islam in medieval inland Ethiopia are greatly indebted to a few dozen Arabic inscriptions which constitute the only endogenous Arabic written sources before the sixteenth century (Schneider, 1967 & 1970; Fauvelle-Aymar & Hirsch, 2008: 349-352; Bauden, 2011; Chekroun & Hirsch, 2020: 89-93) (Fig. 1-A).1 Apart from one fragmentary monumental inscription now preserved in the medieval rock-hewn church of Wəqro Čərqos (Smidt, 2009; Muehlbauer, 2021), a handful of artefacts like jeweller’s moulds (Fauvelle & Mensan, 2011), a few documented graffiti (Bosc-Tiessé et al., 2010), and imports such as precious metallic vessels or rare pieces of textile (Mordini, 1957), all the Arabic inscriptions from Ethiopia are epitaphs carved or incised on funerary stelae or grave markers. Although the epitaphs are rightfully regarded as precious documents, providing names, dates as well as many other information on religious beliefs, foreign connections, and long-distance contacts (Schneider, 2009; Loiseau, 2020; Loiseau et al., 2021), the stelae on which they are inscribed also represent key archaeological artefacts for the understanding of funerary practices and architecture.

Fig. 1 – Map showing the location of the medieval Muslims cemeteries with inscriptions in Ethiopia (A) and in south-eastern Tigray (B).

Fig. 1 – Map showing the location of the medieval Muslims cemeteries with inscriptions in Ethiopia (A) and in south-eastern Tigray (B).

S. Dorso, 2022

2Too often however, they have been published with emphasis on the text and only limited information as to their materiality or the context for their discovery. The lack of details about the petrographic nature of the epigraphic support, about its preparation and general dimensions (including thickness), but also on its position and its implementation within a grave marker do not allow for a broader study of the funerary sites they belong to. This statement is not restricted to Ethiopian inscriptions and suffice it to say perhaps that in two master-studies such as Madeleine Schneider’s book on the Dahlak Kabir necropolis (Schneider, 1983) or that of Paolo Fernando de Moraes Farias on Arabic Medieval Inscriptions from the republic of Mali (de Moraes Farias, 2003), like in recent books on the funerary inscriptions from the Meccan cemetery of al-Muʿallā (al-Rāshid, 2004), one will find only a single picture of a standing stele associated with its grave markers. If some publications tend to devote more attention to the materiality of inscribed stelae (Ory, 1989; or more recently Da’ādli, 2011; Imbert, 2021), a standardised and systematic documentation protocol is still awaited.

3We believe that this bias, oddly persisting despite the “material turn” in historical sciences or the development of philology which raised awareness on how much there is to be gained by looking at support for writing and writing processes contributing to the current state of documentation (for instance Juvin, 2010; Tsakos, 2019), can be overcome through an improved dialogue between specialists of subfields involved in the study of inscribed stelae and burials. Archaeologists attempting to reconstruct funerary practices, architects and epigraphists trying to reveal the meaning and purpose of inscriptions could benefit from a more thorough documentation of the stelae in their archaeological environment, that is, as an “integral culture product” (Lambourn, 2008).

4Several aspects which are not yet systematically recorded include for instance the visibility of the inscribed stele (dependent upon its position and insertion in the surface marker but also on the potential evolution of its environment) which gives a clue on its intended public/readership and on its perception by visitors in the burial ground. The distinctive size, shape or colour of the stone can also, while compared to the rest of a corpus (i.e., not in isolation) provide interesting hints on the social status of a deceased within the cemetery’s population. In the same way, although this is generally more obvious and taken into consideration, the sophistication of the ornamentation or the quality of the engraving can imply different levels of investment and technicity, often reflecting the wealth and social status of the deceased’s relatives.

5We are well aware that practical conditions for research in the field do not always allow for an ideal documentation process, and it is in fact our own shortcomings in the field which led to most of the questions discussed in this article. The latter focuses on the medieval Muslim cemeteries identified by the ERC HornEast team in Northern Ethiopia and is intended as a general reflection on their materiality. It presents preliminary remarks regarding the stelae and funerary markers and aims at laying a groundwork for further comparisons on a regional scale.

The medieval Muslim cemeteries in south-eastern Tigray: a brief overview of the recent discoveries

6Ethiopia, and its northern region of Tigray held a special place in the beginnings of Islam because it is where, according to Muslim tradition, more than eighty of the companions of the Prophet found refuge when they were persecuted in Mecca. The asylum and protection granted by the Christian King of Aksum, appearing as al-Najāshī in the Sira, would have led some of the first Muslims to settle in Abyssinia (bilād al-Ḥabasha). This episode, which came to be known as “the first hijra” or “al-hijra ʾilā al-Habasha”, would have occurred around 615 CE and is nowadays commemorated in eastern Tigray on the site of Nagāsh, where the tomb of al-Najāshī along with those of several of the Prophet’s companions have been displayed and venerated at least since the 10th/16th century (ʿArab Faqīh, Conquête: 419-420, Gori, 2007). However, the current state of the funerary complex at Nagāsh is the result of continuous rebuilding which, while attesting the lasting importance of the site as a pilgrimage destination since the sixteenth century, prevents identification of potential earlier structures. It is therefore some 45 kilometres south of Nagāsh that the oldest physical remains so far associated with Islam in the region are to be found.

7In Kwiḥa near Mekelle (Maqala), the existence of a Muslim cemetery has been reported since the nineteenth century (Pearce, 1831: 121-124; Conti Rossini, 1937) and several funerary stelae found scattered in the vicinity of the hill of Ǝnda Čərqos have been studied, most notably by Madeleine Schneider (Pansera, 1945; Schneider, 1967 & 2009; Smidt, 2004). In 2018, a French-Ethiopian team discovered new stelae in the area and was able to locate the remains of a burial ground near the place known as “Bilet” and to expose and excavate several graves (Loiseau, 2020; Loiseau et al., 2021) (Fig. 1-B). Radiocarbon dates obtained from funerary deposits and skeletal remains proved coherent with the dates preserved in the epitaphs and attest to the use of the cemetery between the late 4th/10th century and the 7th/13th century, with a probable peak of activity around the 5th/11th century.

Fig. 2 – General plan of the Bilet cemetery showing grave markers and pottery deposits

Fig. 2 – General plan of the Bilet cemetery showing grave markers and pottery deposits

© S. Dorso/Mission Bilet, 2022

  • 2 Beside the 40 stelae studied by our team, five more inscriptions reportedly found in Bilet and publ (...)
  • 3 49% of the stelae refer to deceased males while 17% are still undetermined.

8Despite having been exposed only partially, the cemetery of Bilet so far provides the largest corpus of Medieval Arabic inscriptions from Ethiopia (40 inscriptions), as well as the opportunity to study local funerary architecture (Fig. 2-A)2. The total extent of the cemetery remains unknown, and it is noteworthy that 23 of 40 stelae discovered in 2018 were found scattered within a 200 m radius, covering an area of nearly 24 000 m². However, following excavation and surface observation it can be estimated that the funerary area exceeded 875 m², of which 128 m² were fully exposed (Area 1 – Sounding 1) and an additional 32 m² partially exposed (Area 1 – Soundings 2 and 3) accounting for less than 20% of the minimal estimated extent of the site. The size of the cemetery’s population is not known either. A total of 20 graves were identified and exposed during excavations in 2018, one of which (lacking a surface marker) contained the skeletal remains of an individual in anatomical order and bones of another individual which had been disturbed, suggesting consecutive burials or post-burial management in the cemetery. The number of deceased recorded in the epitaphs from Bilet amounted to 35 (5 deceased benefitting from twin stelae), a greater number than the inventoried surface markers. Moreover, only two of the exposed surface markers can be firmly associated with inscribed stelae: one preserving its head stele (B 24) and the other presenting the negative imprint of a missing head stele. One should also stress that at least four of the fully exposed grave surface markers did not present any evidence of stelae and that at least two of the partially exposed markers presented an epigraphic head stelae. It can be assumed therefore, that many grave markers in Bilet did not incorporate stelae and that the number of known inscribed stelae should not be used to extrapolate the cemetery population. The inscriptions on the stelae as well as the anthropological examination of the human remains confirm that both men and women were buried in the cemetery and that the use of inscribed stelae was not restricted to one gender. In fact, 34% of the stelae can be attributed to deceased females3. The analysis of the epitaphs also indicates that several deceased from Bilet were associated through their nisba-s to foreign lands or cities in the Islamic world, in Egypt, the Arabic Peninsula or Persia. If this suggests that Bilet had long-distance connections with other regions of the Islamic world, the recurrence of names (ism-s) of non-Arabic origin also hints at a Muslim community embedded within a local Ethiopian environment (Loiseau, 2020; Loiseau et al., 2021). The use of both “local” and Arabic names in the same Muslim cemetery can be interpreted in different ways, as evidence of a recent or ongoing process of Islamisation and Arabisation, as evidence of the presence of foreigners or as a token of the cosmopolitanism prevalent in the region during the Middle Ages (Insoll, 2021).

9While excavating the Bilet cemetery, another group of Medieval Muslim cemeteries was identified further south thanks to information provided by Dr. Fesseha Abraham and the help of local Muslim communities. These previously unreported sites are located in the vicinity of the village of Arra, near Adi Gudem (Addi Gudäm), some 30 km south of Bilet along the so-called “Eastern-Tigray Road” (Fig. 1-B). As these sites have only just been surveyed, their dating is based solely upon dated funerary inscriptions.

10The Arra region epigraphic corpus is divided into four subgroups. Three inscriptions from the 6th/12th century originated in Arra itself but were not found in a cemetery per se since two of them were found in a private collection (A01 and A02) while the third inscription (QR01) currently appears on an isolated cairn, which could represent a one-of-its kind grave marker in the area (Fig. 3-A). The three other sites located east of Arra, in the depression leading to the Afar lowlands, are proper funerary sites, each comprising several graves. Two cemeteries, Tsomar and Meyda Zeylegat (Fig. 3-D and C) are located along a path at the foot of a hill facing Arra, while the third, Ḥabera, stands on top of the hill, at its southern end (Fig. 3-B). Despite their proximity (less than 2 km apart as a crow flies), the location of the three cemeteries differs, two of them being easily accessible along a natural path while the third is located on high ground requiring a climb of about 100 m on a c. 15˚ slope between Meyda Zelegat and Ḥabera. The choice of opposing topographical spots like the reason why Muslims would bury their dead in three different locations during the same period are not fully understood yet. So far, no settlement clearly associable to the cemetery has been identified and one cannot rule out the possibility that the community or communities using the cemeteries may have lived at some distance from the sites, whether sedentary or not.

Fig. 3 – General views of Qelqel Rway (A), Ḥabera (B), Meyda Zeylegat (C) and Tsomar (D)

Fig. 3 – General views of Qelqel Rway (A), Ḥabera (B), Meyda Zeylegat (C) and Tsomar (D)

© S. Dorso/Mission Bilet, 2019

11The three cemeteries also differ in size. In their current state Tsomar (Loiseau et al., 2022) and Meyda Zeylegat both extend over c. 1500 m² while Ḥabera is twice as large, covering at least 3000 m². Because these sites were surveyed recently and because of the uneven conservation of the surface markers, the estimation of the number of graves is only tentative. In Tsomar, where erosion and vegetation complicated the survey, between 50 and 90 grave markers have been identified; in Meyda Zelegat about a hundred have been identified and many more then that have been identified in Ḥabera. As yet, the internal organisation of the funerary spaces has only partially been understood. In Bilet, grave markers arranged in rows suggest that at least part of the cemetery could have been managed, and a partially excavated wall may also indicate a sectorisation within the cemetery. In Tsomar, no specific arrangement has been identified but circular markers seem to be concentrated in the east. In Meyda Zelegat groups of circular markers are clearly visible but the logic behind the grouping is unknown. In Ḥabera, the picture is even more complex, as the cemetery seems to be divided in several areas including some organised with alleys of graves.

12Excluding the three inscriptions from Arra dating from the 6th/12th century, the dates of the Arra region epigraphic corpus range between 614/1217 and 780/1378, meaning they are contemporary to the final phase of occupation of Bilet although they are mostly later in date. Interestingly, the cemeteries thus encompass two pivotal moments in the history of Ethiopia: the rise and fall of a Christian royal dynasty, the Zagwe (between the 5th/11th and the 7th/13th centuries, Derat 2018), but also the emergence of local Muslim polities opposed to the Zagwe’s successors, the Solomonids (Chekroun & Hirsch, 2020). The cemeteries represent important evidence that Muslim communities were present in the region regarded as the cradle of the Christian Kingdom of Ethiopia, where they maintained themselves, at least on its inner border, during periods of conflicts with the local sultanates. It is possible that these communities played a part on a much larger stage and recent studies have emphasised the complex policy carried out in northern Ethiopia by the Fatimids, especially under the vizirate of Badr al-Jamālī (d. 487/1094) who sought to appear as the protector of Muslims in the region and propagate the daʿwa through the erection of mosques (Bramoullé, 2012; Brett, 2019; Derat, 2020). This Egyptian involvement in local affairs, if it ever stopped, clearly resumed under the Mamluks, shortly after their seizure of power in the second half of the 7th/13th century. Diplomatic correspondence suggests that by the second half of the 8th/14th century, the Cairene sultans presented themselves as protectors of Ethiopian Muslims, especially in the context of King ʿAmda Ṣeyon’s campaigns (Loiseau, 2019). The latest dates from the Ḥabera stelae attest that the cemetery was still in use during the regency of (future) sultan Barqūq (1376-1381) and around the succession between kings Newāya Māryām (1371-1379) and Dāwit (1379/80-1413) before the troubled period marked by the succession conflict following Dawit’s death and the wars against the Awfāt sultanate.

13This historical context, although relatively fragmented, can help to understand how the four medieval Muslim cemeteries from Tigray developed in a region politically dominated by a Christian power. With the monumental inscription of Wəqro Čərqos, the 84 epitaphs and their surface markers are, so far, our only window on an Islamic presence in the northern highlands of Ethiopia. These open new avenue for the study of these communities’ organisation and their relations to one another but also to other regions of the Islamic world and their Christian environment.

Grave markers and graves: variability, monumentality, and related questions

14Beside the issue of their location, size and chronology, the medieval Muslim cemeteries of Eastern Tigray also present an important diversity in terms of funerary architecture. While in Bilet grave surface markers belong to a single type, i.e., rectangular drystone platforms, the grave markers from the region of Arra vary in shape and in size with circular markers, rectangular enclosures, or nave-shaped markers (Fig. 4-B to D, fig. 5).

Fig. 4 – Schematic typology of medieval grave surface markers in Tigray

Fig. 4 – Schematic typology of medieval grave surface markers in Tigray

© S. Dorso, 2022

15This diversity invalidates the idea of a unique local funerary culture. It invites one to look for possible influences or models, both locally and among distant regions as well as to consider the existence of micro-local traditions or an evolution of the regional funerary architecture between the 4th/10th and 8th/14th centuries. More generally, it questions the use of a distinctive funerary architecture vis-à-vis non-Muslim neighbours but also within local Muslim groups.

Fig.5 – Diversity of the grave markers in Tigray

Fig.5 – Diversity of the grave markers in Tigray

© S. Dorso/Mission Bilet, 2019

16The rectangular platforms in Bilet are built of small unhewn slabs arranged in one to five courses (Fig. 6-A, B). They measure between 2 m and 3 m in length and between 0,9 m and 1,9 m in width. It means that although they share a common appearance, they do not constitute a homogenous group, with some markers being twice as wide or one and a half times as long as the smallest ones. It is noteworthy that in the excavated area, two of the largest grave markers stood out from the rest in being surrounded by an outer stone circle. The fact that some of the graves were the object of particular attention and investment, probably reflecting the status of the deceased in the community is further emphasised by the difference in size among the inscriptions, a matter that will be discussed later. The special status of some of Bilet’s deceased, or at least of the devotion expressed by the living toward them, is also evident through the concentration of funerary deposits identified around one of the most massive graves (F1051) which benefited from a stone circle and a standing stele (B24, Fig. 7-A). Both funerary architecture and practices (like mourning banquets or funerary offerings) offer valuable clues for the reconstruction of social hierarchies within the community along with references to common ancestors in the epitaph like Ḥafṣ b. ʿUmar al-Yamāmī (Loiseau, 2020).

Fig. 6 – Grave markers in Bilet and Arra, Tigray

Fig. 6 – Grave markers in Bilet and Arra, Tigray

© S. Dorso, 2018

17Strictly speaking, the tombs of Bilet comply with Muslim legal norms, their orientation allowing the deceased bodies to face the qibla, the direction of the Kaʿba in Mecca (Loiseau et al., 2021). Whether inscribed or not, the stelae incorporated on one or both ends of the platform also link them to a widespread Islamic tradition (Rāġib, 1992; Diem & Scholler, 2004; Petersen, 2013; Gleize, 2022). In the region, they however differ from the medieval Muslim grave markers documented so far in Ethiopia which generally belong to four main types (and subtypes): rectangular enclosures (Beri-Ifāt, Harlaa, Hajjī Mansur, Massāl, Asbari, Gozé, Faqi Dabbis, Norā), nave-shape enclosures (Adulis) stone circles (Adulis, Faqi Dabbis) or monolithic rectangular markers (Abdul Rasul, Fiqar, Gachene, Gazé, Rāssā Gubā, Hajjī Mansur, Massāl).

18The closest parallels to the Bilet grave markers are found in modern Tigrayan cemeteries like in Kwiḥa Čərqos or Arra and belong to Muslim as well as Christian deceased (Fig. 6-C, D). Most of the time, though not always, these modern “platforms” incorporate an anepigraphic stele, usually crowning the marker at the head or in its centre and with divergent orientations. The contemporary use of similar markers, albeit with small differences, in the same sites and in funerary areas sometimes very close to one another invites us to reconsider the theorical necessity of a distinctive funerary architecture. In Arra for instance, a village currently inhabited by both Muslims and Christians, it is the orientation that differentiates the grave markers, not their appearance or architecture. It is also noteworthy that in Arra, both Christian and Muslim cemeteries host different types of grave markers believed by our informants to be contemporaneous.

19These recent practices invite caution when studying the ancient cemeteries in the area. In the vicinity of Ḥabera, Meyda Zeylegat and Tsomar, no less than five other pre-modern burial grounds have been identified, among which four present grave markers similar to those of the local medieval cemeteries in terms of architecture and sometimes orientation (Fig. 1- B). Without funerary inscriptions and in the absence of excavations however, it is impossible to confirm the oral traditions identifying them as Muslim graves or to ascribe them to the medieval period. This clearly constitutes a major bias for understanding the local funerary landscape because, providing that these sites are indeed contemporary with Ḥabera, Meyda Zeyegat and Tsomar (and Arra?), they would double the number of known cemeteries in the area, increasing considerably their distribution density and therefore impacting any assumption regarding the presence of Muslim communities in this territory.

20It is possible that the use of different types of grave markers relates to changing practices over time and that local traditions prevalent before the adoption of Islam were maintained or adapted by Muslims or mixed communities in the same burial ground. This would challenge the common postulate that people from different faiths would have chosen to be buried in spatially distinct spaces, but there are precedents (Lloret, 2011). Dated stelae however, attest that different types of grave markers were used at the same time, for instance nave-shaped and circular markers in Ḥabera and most likely in Tsomar. Interestingly, the cemetery of Meyda Zeylegat, in use during the mid- 7th/13th century (i.e., around the same time as Ḥabera and Tsomar), only hosts circular grave markers with central stelae, questioning the preference of one type of grave marker over another. Rather than a confessional or gender identity, could grave markers have expressed a collective identity (family, confraternity, etc…)? So far, contrarily to the belief of some of our informants and based on epitaphs found in situ with their markers, there is no evidence for correlation between the shape of a marker and the gender of the deceased.

21Eventually, while the monumentality of grave markers is an important criterion to consider at the scale of a given site, it is also interesting to take into account how it may or may not have affected the landscape and signalled the burial ground. No doubt that Bilet’s markers, especially those incorporating large stelae would have been visible from a distance with platforms outcropping from the ground and located in a small vale at the foot of a hill, at the edge of the flat of the Gembele river basin. In contrast, the graves from Arra’s surroundings do not stand as high and their architecture may be considered more modest if not rudimentary. Moreover, they are located in a much more rugged environment, among hills and seasonal riverbeds. It seems that nothing indicated the presence of a cemetery in the landscape and while they can be noticed from the top of the surrounding heights, the concentration of dozens of graves being visible during the dry season when vegetation is scarcer, the sites of Meyda Zelegat and Tsomar are usually discernible at a short distance only. Ḥabera, on the southern peak limit of the homonymous hill, faces the Arra volcanic ridge (likely to have been inhabited during the Medieval period) from which it is in visual range. In its current state however, it is impossible to identify the cemetery either from the village or from the bottom of Ḥabera hill. In Bilet, oral traditions documented at least since the 1930’s refer to a possible “Qubba” on the slope overlooking the cemetery. This domed monument designated by its Arabic name (interestingly in a currently Christian-dominated environment) would likely have served a funerary and/or memorial function, perhaps to signal an Islamic burial site. Unfortunately, the monument which still partially existed nearly a century ago was not preserved and excavations at its supposed location were insufficiently conclusive to corroborate its nature and date, although foundations of a square building were exposed.

22This leads us eventually to a final remark on a missing dimension. Indeed, one must note that in the current state of research, none of the medieval Muslim cemeteries of South-Eastern Tigray is decisively associated with a settlement or a place of worship. This obviously constitutes an obstacle for discussion about the interactions between the communities and their burial sites or the inclusion of the latter in its inhabited environment.

Inscribed stele and grave markers: displaying the text between normativity and practicality

23The study of inscribed stelae as part of a Funerary marker is in our sense essential for the general understanding of the sites. After all, grave markers are the first if not the most important element of the inscription’s context. Then, two parameters appear crucial for the study of newly discovered inscribed stelae: their position and visibility. In the following lines, we offer some remarks drawn from our own exploration of the Muslim cemeteries of Enderta. The Tigrayan epigraphic corpus of funerary stelae being currently divided into two different geographic areas, we will refer for clarity to the Bilet cemetery as Group 1 and to the sites from Arra’s area as Group 2.

Location and position

24Although epitaphs sometimes cover entirely the grave marker, like in the case of the Walasmaʿ necropolis (Fauvelle et al., 2017), they are more commonly limited to a stele placed either at the head, the foot or in the middle of the grave marker. In this regard, the Tigrayan stelae and especially those from Bilet belong to a well-known model. In fact, they are aesthetically reminiscent of many stelae from Dahlak and the Arabian Peninsula, two areas with which connections are attested by the nisba-s and genealogical information contained in Bilet’s epitaphs. In Dahlak, most of the in situ stelae were discovered standing over the head of the deceased, and at the head and foot in case of twin stelae.

25In our case, assessing the trend in stelae position is rather confusing. Indeed, the Bilet site became an agricultural field and most of the stelae from Group 1 were discovered reused in the bordering walls of different parcels. Fortunately, one stele (B24) was excavated in primary position, embedded in a grave marker, giving us an idea of the stelae implementation in Bilet (Fig. 7-A). Although they seem better preserved, the southern cemeteries of Group 2 have also experienced modifications over time but, at least in Tsomar and Ḥabera, many stelae were recorded standing vertically in their original position (Fig. 7-C). Interestingly, other stelae were found in more unusual positions without however, clear evidence of removal or displacement. Most of these “uncommon” stelae were found lying horizontally, barely protruding from the ground, partially buried with the text facing either the sky (Fig. 7-B) or the ground. A few others were also found buried vertically, sometimes isolated or inside a stone circle, with only one line or none of the inscription legible (2 in Tsomar, 3 in Ḥabera, 1 in Meyda Zeylegat), and on several occasions with the text inclined at 90° laterally (Fig. 7-D). Although it is not uncommon for stelae to be found partially buried, intentionally or due to taphonomic processes (sedimentation and granulation after the backfilling of the pit), one may wonder if some inscriptions with five lines or more illegible were not intentionally buried and hidden, at the time of the initial burial or later.

Fig. 7 – Various implementations of the stelae within the grave markers

Fig. 7 – Various implementations of the stelae within the grave markers

© S. Dorso/Mission Bilet, 2022

26This hypothesis prompted us to investigate further the processes leading to the current state of preservation of the cemeteries. Today and for an unknown period of time, the cemeteries have been part of annual pilgrimages or “visits” (ziyāra). They are regularly frequented and maintained by the neighbouring Muslim communities, especially by faithful from Arra and Adi Gudem. During our survey, we were able to witness the rebuilding of several venerated modern tombs in the countryside as well as newly discovered inscriptions being turned upside down so as not to attract attention.

27While studying the dispersion of stelae, it appeared that some had been gathered prior to our arrival, but also that some pairs of stelae had been displaced and relocated. For instance, at Habera and Meyda Zeylegat, twin stelae (inscribed with epitaphs for the same deceased) were found between 5-6 m and 15 m apart (Fig. 8-A). The most eloquent case for the disposition of twin stelae comes from Meyda Zeylegat, where two epitaphs referring to the same deceased were found in different places, one fully buried in the middle of a circular marker, and the other lying on the ground at some distance with evidence of having been formerly partially buried (the colour of the stone changes when it is exposed to the weather for some time and contrasts with the buried part, see for instance fig. 10-B).

Fig. 8 – Draft plans of the cemeteries in the Arra region (based on aerial photographs) showing the location of inscribed stelae at the time of their discovery

Fig. 8 – Draft plans of the cemeteries in the Arra region (based on aerial photographs) showing the location of inscribed stelae at the time of their discovery

© S. Dorso, 2022

28The gathering of stelae evidenced on site seem to respond to distinct logic. In Meyda Zeylegat, all the dated and complete stelae were found in the same area of the cemetery, while in Ḥabera only some of them were, others having been obviously moved beyond their original grave or deposited in circular graves without being buried. Tsomar is yet another case. To our knowledge, the cemetery is not included in the pilgrimage route. There, only two stelae were found complete, both standing with a fully or near fully legible inscription. Both belong to the same period and present the same orientation, supporting the idea that they were preserved in primary position. The rest of the epigraphic material from Tsomar is very fragmentary. The destruction of the stelae may be attributed to human activity, possibly even before the development of a dirt track which crosses the cemetery since about a decade. From this fragmented material, it was only possible to partially recompose two epitaphs (Loiseau et al., 2022). These different cases indicate that, contrary to our first impression, the southern cemeteries were also subjected to important changes over time.

Visibility and implementation of the stele in the grave marker

29Obviously, the visibility of a stele depends first and foremost on its size (height, width and length). It is therefore useful (but not so common) to present an epigraphic corpus on a single plate at a single scale. By so doing (Fig. 9), the rather wide size-range of the stelae from Bilet is evident. The majority of Bilet stelae (Group 1) are more than 40 cm high, some even reaching one metre, but none measure less than 30 cm (table 1). Without doubt, the tallest stelae would have stood out in the landscape indicating the existence of a cemetery and presumably identifying the graves of prominent deceased. The Stelae from Group 2 are smaller, most do not exceed 40 cm and a number of them measure less than 30 cm in height. The width of the stelae is also interesting (table 2). In Group 1, stelae widths vary between 13 and 60 cm although the majority ranges between 20 and 40 cm. In Group 2 the range is again smaller, stelae widths vary between 18 and 45 cm, most measuring between 20 and 30 cm. Comparatively, the stelae of the second group are therefore less imposing than those of Group 1.

Fig. 9 – Bilet’s stelae discovered by the HornEast project presented on scale and ordered chronologically (the grey frame indicates twin stelae)

Fig. 9 – Bilet’s stelae discovered by the HornEast project presented on scale and ordered chronologically (the grey frame indicates twin stelae)

© S. Dorso, 2022

Table 1: Height of the stelae from Tigray

HEIGHT

Group 1

Group 2

Site

Bilet

Meyda Zelegat

Tsomar

Ḥabera

Number of complete stelae/total number

25/40

5/9

2/7

17/25

Height range

31 to 100 cm

24 to 45 cm

24 to 26 cm

20 to 55 cm

Height brackets (cm)

30-40

40-60

60+

-30

30-40

40+

-30

-30

30-40

40+

Total

6

15

4

1

2

2

2

3

9

5

Table 2: Width of the stelae from Tigray

WIDTH

Group 1

Group 2

Site

Bilet

Meyda Zelegat

Tsomar

Ḥabera

Number of complete stelae/total number

33/40

5/9

4/7

18/25

Width range

13 to 60 cm

18 to 29 cm

23 to 42 cm

21,5 to 43 cm

Width brackets (cm)

-20

20-30

30-40

40+

-20

20-30

20-30

30-45

20-30

30-40

Total

4

12

11

6

1

4

1

3

11

5

30Comparing the overall sizes of the stelae can be misleading however, since part of them may have been buried in the ground. While some stelae are fully engraved, part of others remained unworked, a feature we refer to as “the epigraphic reserve” (Fig. 10). Looking at the “reserve” in relation with the thickness of a stele may help to identify its possible functional role. When the reserve represents over a quarter of a stele’s height, it may have served to be buried so as to have the stele stand vertically. The two standing stelae from Tsomar provide good examples of this (Figs. 7-C and 10-C.; Loiseau et al., 2022: 659, fig. 12). Conversely, the absence of epigraphic reserve would infer that either part of the text had been buried and was illegible, that the stele had been inserted or rested on the grave marker like for Bilet B24, or that the stele was laid down horizontally in order to be legible. In other words, even when an inscribed stele is found in secondary position or out of context on the surface, the epigraphic reserve can offer clues on its original implementation.

Fig. 10 – The “epigraphic reserve”

Fig. 10 – The “epigraphic reserve”

© S. Dorso, 2022

31Stelae from both corpuses were sorted into three main categories according to the size of their epigraphic reserve: more or less null (Fig. 10-A); less than one quarter (Fig. 10-B); and more than a quarter of the stele’s height (Fig. 10-C). Since the thickness of a stele sometimes allows it to stand on its own as an orthostat, it was also taken into consideration. Again, three groups were defined according to the thickness: less than 10 cm, between 10 and 20 cm, and more than twenty cm. We considered that 10 cm was the minimum thickness for an orthostat stele. While interpreting the results, it is important to bear in mind that 80% of Bilet’s stelae are complete or almost complete when in Group 2, only 60% of the stelae are complete (40% are fragmented).

32Table 3 which compares stelae from Bilet and Habera (the two sites with more than 20 stelae) shows that at Bilet where most of the stelae date back to the 5th/11th century, 80% have no reserve and 86% are thicker than 10 cm (avg. 17 cm), with several stelae being thicker than 20 cm. On the contrary, at Ḥabera, a cemetery essentially in use during the 7th/13th century, 50% of the stelae have also little or no reserve, with a thickness rarely exceeding 10 cm. The results for Bilet suggest that most of the stelae could have been implemented in the same way as B24 and inserted in the platforms marking the graves. However, it questions how stelae from Ḥabera would have stood vertically without supporting structures and if some were not initially in horizontal position. This allows to reconsider several stelae from Ḥabera found placed horizontally in the centre of a naviform marker and first regarded as displaced. Being too thin to stand vertically, they may be in their initial position. The same phenomenon of stelae with and without epigraphic reserve is recorded in the cemetery of Munessa, in the Arsi country (Azaïs & Chambard, 1931: pl. LXIII; Rossi, 1942: 281).

Table 3: The use of epigraphic reserve in stelae from Bilet and Ḥabera

RESERVE

THICKNESS

Site

Bilet

Ḥabera

Site

Bilet

Ḥabera

Number of complete stelae/total number

29/40

20/25

Number of sufficiently preserved stelae/total number

36/40

20/25

Ratio of the epigraphic reserve

~0

less than ¼

more than ¼

~0

less than

¼

more than ¼

Range in cm

5-10

10-20

20+

5-10

10-20

Number of stelae

15

9

5

8

2

10

Number of stelae

5

20

11

18

2

Total ratio

80%

20%

50%

50%

Total ratio

14%

86%

90%

10%

33Given that work on the epigraphic support is limited or inexistent, it is also possible that the shape and the treatment of the stelae were imposed by the material used. In Bilet, 90% of the stelae were inscribed on relatively thick drums of eroded local basalt columns which naturally provided flat faces for the stonecutter. In the southern cemeteries, only 40% of the stelae were made from basalt, the rest being of local sandstone or limestone.

The materiality of inscriptions: looking beyond the text

34Looking beyond the text sometimes demands a real effort for the epigrapher who is instinctively attracted by its meaning. However, much can be gained from the study of the inscription’s materiality which often provides information on the funerary economy as well as the structure and activities of the community under scrutiny. The investment and care devoted to the epitaphs can be measured through the work on the epigraphic support, its preparation and the tools employed for the carving as well as by the chosen script or the decoration. These elements also inform us on the stonecutter since the quality of work and skills required would identify a lettered craftsman or a specialist possibly working in a local workshop.

35In the Tigrayan cemeteries, work on epigraphic supports is almost absent. It seems that the choice fell on rocks which did not require faceting and genuinely offered smooth surfaces although some inscriptions were also executed on irregular surfaces. Only a few examples suggest a preparation. This includes possible circular polishing or partial smoothening of the epigraphic field. Interestingly, like in Khor Nubt (Sudan, KN85, Oman, 1998), or other cemeteries located near the Eastern Desert gold mines (Egypt and Sudan, Klemm & Klemm, 2013: 476) or Dahlak (Insoll, 2020: 421) the reuse of grindstones appears extraordinary, possibly on a unique instance.

36The preparation of the epigraphic field also presents great diversity, with inscriptions exhibiting a great regularity of line spacing, implying a preliminary drawing or ruling with a string system leaving a temporary trace (Fig. 11-D), and others being more chaotic suggesting no preparation or planification. It is also possible that the text itself underwent a preliminary outline. One stele (Sch. 09) presents two types of writing and likely represents an instance of an unfinished stele rather than one finished in a hurry. Regarding the engraving techniques, the Tigrayan stelae do not belong to the monumental type, none are in relief and most of them have the aspect of graffiti of variable quality. The finest examples were made with an engraving tool, through the connected dots technique (Fig. 11-A) or the one stroke technique (Fig. 11-C). Most of the others were made through simple scraping (B20) (Fig. 11-B).

Fig. 11 – Engraving techniques and evidence for the preparation of the epigraphic support (in green: preliminary ruling of the epigraphic field, in orange: asymmetric frame of the decoration)

Fig. 11 – Engraving techniques and evidence for the preparation of the epigraphic support (in green: preliminary ruling of the epigraphic field, in orange: asymmetric frame of the decoration)

© S. Dorso, A. Lagaron, 2022

37In terms of palaeography, most of the writing styles used in the stelae belong to the Kufic script or transitional script tending toward the Naskh. In Bilet, five group of Kufic inscriptions were identified:1. the ornamental Kufic used from the 4th/10th century to the middle of the 5th/11th century (characterised by regular letters and rising tails); 2. a Kufic style with an ornamental tendency (with bifid extremities and heart shape) appears in the first half of the 5th/11th century (Fig. 12-10); 3. a simplified Kufic style with straight characters is attested from the mid-5th/11th century and persists during the second part of the century (Fig. 12-11); 4. an ornamental Kufic with bezel characters visible attested in the first half of the 6th/12th century; 5. a group of rough inscriptions made of simplified Kufic with a cursive tendency from the middle of the 5th/11th century to the middle of the 13th/7th century. A few examples of calibrated Naskh are dated to the middle of the 6th/12th century. In the southern cemeteries, more than half of the inscriptions are in characteristic Naskh with a flexible writing, some are unclassifiable, and the others are engraved in transitional script between the simplified Kufic style with relatively angular characters tending toward the Naskh.

38Fig. 12 – Examples of decoration and palaeography in Bilet (Group 1), Meyda Zeylegat and Tsomar (Group 2)

© S. Dorso, A. Lagaron, 2022

39As previously mentioned, some stonecutters chose to keep a reserve in the lower part of the stele, but also on the upper part to allow decoration. In Bilet, 57% of the stelae present a decoration against only 20% of the stelae of Group 2. In Bilet, the most recurrent designs are herringbone and wavelet patterns (Fig. 12-1 to 6), but there are also some seals of Solomon during the 4th/10th century (Fig. 12-1 to 3) and numerous “ansa” during the 5th/11th century (Fig. 12-4 to 6). Although this ornamentation usually takes place around it, the text can integrate some decoration with characters taking the shape of a heart (Fig. 12-10). It is noteworthy that apart from one 4th/10th century stele which may have served as a source of inspiration, all the stelae displaying this kind of motif belong to women deceased during the 5th/11th century. Generally, decoration is less present and simpler in the southern cemeteries, often consisting of simple lines or a few fragmentary sketches and chevrons. Nevertheless, the motif of the seal of Solomon is attested there on three occasions (Fig. 12-7 to 9), twice in Tsomar engraved at the end of the basmala and once in Meyda Zelegat where three seals were carved at the end of the same stele.

Table 4 – Decors on the funerary stelae from Tigray

DECORS

Site

Bilet

Around Arra

Century (AH)

All

4th

5th

6th

7th

All

6th

7th

7-8th

Number of stelae

23

3

17

3

-

9

-

6

3

A workshop in Bilet?

40Our study of the different corpuses identified only a few evidently similar inscriptions while cemeteries in Mecca (Al-Rāšid, 2004; Juvin 2013), Aswan (ʿAbd al-Tawāb, 1977) or Dahlak (Schneider, 1983) show that, there, regularity of execution usually prevails among stelae, attesting undoubtedly to the production of a workshop. While the study on the lapicides around Arra is still in progress, we are able to present our conclusions about Bilet. There, the study of stelae (text and context) allowed us to identify the work of four craftsmen during the 5th and the beginning of the 6th centuries AH. In parallel with these lapicides, it appears that other “hands” also engraved stelae suggesting the contemporary presence of skilled craftsmen residing or passing through Bilet and amateur or occasional lapicides. The presence of stelae of radically different styles without any apparent echo in the corpus invites us to reflect on who might have executed the stelae (specialised lapicides with a repertoire of epitaphs? relatives?). In this respect, the cases of twin stelae are remarkable. Our comparative study reveals that the twin stelae among the other corpuses are very close stylistically. If they sometimes differ in their contents, they are visually quite similar (Schneider, 1983: 60). In contrast, in Bilet, if three of the five pairs look relatively similar with only minor stylistic or qualitative differences, the other two pairs have stelae that are radically different from each other. For instance, in the pair formed by stelae Sch 03 and Sch 07, Sch 03 is executed with an ornamental Kufic while Sch 07 is in simplified Kufic, a discrepancy which would qualify them as “false twins” (Schneider, 1967: pl. xlvii [j.e.3373] and xlix [j.e.3366]; and Fig. 12). In our view, it would appear that two distinct craftsmen worked for the benefit of the same deceased. One may wonder if this was prompted by a desire for a quick completion, as if time were running out. It is of course tempting to imagine a caravan stop, where the bodies of people who died along the way could not have been repatriated, and where the duration of the halt would have been dictated by the collective. As we have seen however, the Bilet site was occupied over a long period. In addition, the other example of “false twin” stelae was made for a female descendant of Ḥafṣ b. ʿUmar al-Yamāmī, the presumed founder of Bilet’s Muslim community whose family members were buried in the cemetery over more than four generations. It would appear therefore than rather than the lack of time hypothesis, the intervention of two lapicides could have been related to a matter of cost or some personal engagement in the realisation of the epitaphs.

Conclusions. Materiality, functionality, performativity: graves and epitaphs as memorial artefacts

41Looking at the cemeteries from the perspective of their materiality opens new avenues for research as it forces us to question facts often taken for granted like the functions of the inscribed stelae or the reasons behind the adoption of a specific grave marker. The diversity of funerary markers, and the many ways in which the Arabic epitaphs from Tigray have been implemented needs to be better understood. Beside the chronological argument, the distinctive traditions of grave markers highlighted in Bilet and the surroundings of Arra perhaps also reflect the different social, economic, and political status of the communities. In Bilet, a site well connected to Arabia and Egypt, as well as to other areas of the Islamic world, presumably through its integration into commercial networks, the funerary architecture may have been influenced by distant models and a mixed population. In any case, it seems that the Bilet cemetery was easily noticeable in its immediate landscape, with several tombs displaying a form of ostentation suggesting a relative prosperity and a peaceful environment at least between the 4th/10th and 5th/11th centuries. In contrast, the cemeteries from the region of Arra are characterised by a more modest architecture, and less remarkable locations (even if the landscapes are more dramatic), with a typology of markers known elsewhere in the region such as the circular ones commonly found in the coastal area, from the eastern desert of Egypt to Somalia. Added to the fact that the epitaphs from Group 2 (near Arra) do not present evidence of the same long-distance connections as Bilet, it may infer a more local dimension for the community or communities near Arra, potentially oriented more towards the lowlands than towards Christian highlands and possibly involved in other activities than commerce. Of course, this is only speculative and other factors may have influenced the locations and form of cemeteries like the growing hegemony of the Christian kingdom in the area by the 7th/13th and 8th/14th centuries, which could have relegated Muslim communities to frontier areas or imposed the adoption of a more distinctive funerary architecture as well as offset locations for their cemeteries. Only new fieldwork and surveys will allow us to elaborate on these issues.

42More universally, the study of Muslim funerary practices from medieval Ethiopia, a peripheral region of the Islamic world, may also prompt new discussion on Muslim norms and practices in general. Without Arabic inscriptions, the cemeteries of Bilet, Ḥabera, Meyda Zeylegat and Tsomar would not have been identified with certainty as Muslim burial grounds.

  • 4 We thank the anonymous reviewer who suggested the enlargement of the formula may in fact serve to h (...)

43Therefore, it is safe to assume that, as in the past, one of the purposes of these inscriptions was to identify the various structures as graves. Even if it is a bit provocative, one might consider that this was in fact their primary function. It is striking to note that on many stelae from Dahlak, it is not the name of the deceased but the formula “hadha qabr” (this is the tomb) which is extravagantly enhanced in a much larger script than the rest of the text (Schneider, 1983; Oman, 1976-1987), although it could be a mere stratagem to draw attention on the deceased’s name while complying with the Islamic prescription of humility in death4. Clearly, the inscriptions were also a key feature in proclaiming the faith of a deceased. The use of Arabic, of widely quoted formulae (including Quranic quotes) but also of evocative decoration and symbols (as effective as the text?) would have attached the deceased to the umma. In sum, in the context of the Northern Ethiopian Highlands during the 5th/10th to 8th/14th centuries, the carving of an epitaph would have been a strong identity marker both on an individual and a collective level, also participating in the “islamisation” of the landscape. In fact, this also would explain why today some stelae randomly found lying on the ground are re-buried by local visitors, sometimes upside down, for the sole purpose of identifying a grave as Muslim.

44From an internal perspective, one may wonder what those Arabic epitaphs meant in the community. Considering the widespread use of anepigraphic stelae (possibly on a majority of the markers) or the absence of any standing feature, the inscription would also have had a role in singularising the deceased’s grave. The fact that several inscribed stelae may have been only partly legible because they were originally half buried also questions their assumed purpose. Perhaps the mere fact of engraving the text was in the end more important than its legibility. Distancing ourselves from the materiality, is it too far-fetched to look at those stelae as “performative” artefacts rather than objects intended to defy the passage of time and to perpetuate the memory of an individual? It seems obvious that carving names and dates in the stone for others to remember or to signal a place of final rest implies that the place and the person’s remains would have been cherished by the living and that the latter could read it (in Arabic in our case). But an important dimension of the stelae may have rested in the performance of carving and in its symbolic (and apotropaic?) significance. In other times and in other places, similar epitaphs with a complete formulary have been written on easily disposable potsherds placed on modest grave markers (Mershen, 2004; Ibrahim, 2001). This would support the idea that the making of the inscription was in fact as important for the living as part of the mourning rites, than as a means of remembrance of the deceased for the whole community.

Haut de page

Bibliographie

SOURCES

ʿArab FaqĪh, René Basset (trad.), 1897-1909, Histoire de la conquête de l’Abyssinie (XVIe siècle), Paris, Ernest Leroux.

WORKING TOOLS

Thesaurus d’Epigraphie Islamique [Online], URL : [http://www.epigraphie-islamique.uliege.be/thesaurus/].

MODERN STUDIES OR ACADEMIC REFERENCES

‘ABD AL-TAWĀB, ‘Abd al-Raḥmān M., 1977, Stèles islamiques de la Nécropole d’Assouan, Le Caire, Institut français d’archéologie orientale.

Azaïs François, Chambard Roger, 1931, Cinq années de recherches archéologiques en Éthiopie : province du Harar et Éthiopie méridionale, Paris, Geuthner, 2 vols.

Bosc-tiessé Claire, DERAT Marie-Laure, Fritsch Emmanuel, Awad Abullif Wadi, 2010, “Les inscriptions arabes, coptes et guèzes des églises de Lālibalā”, Annales d’Éthiopie, 25, p. 43‑53.

BAUDEN Frédéric, 2011, “Inscriptions arabes d’Éthiopie”, Annales islamologiques, 45, p. 285‑306.

BRAMOULLE David, 2012, “The Fatimids and the Red Sea (969-1171)”, in Agius Dionisius A., Cooper John P., Trakadas Athena, Zazzaro Chiara (eds.), Navigated spaces, connected places, Oxford, Archaeopress, p. 127-136.

BRETT Michael, 2019, The Fatimids and Egypt, London/New York, Routledge.

CHEKROUN Amélie, HIRSCH Bertrand, 2020, “The Sultanates of Medieval Ethiopia”, in KELLY Samantha (dir.), A Companion to Medieval Ethiopia and Eritrea, Leyde/Boston, Brill, p. 86‑112.

Conti Rossini Carlo, 1937, “Necropoli musulmana ed antica chiesa cristiana presso Uogrì Haribà nell’Enderta”, Rivista degli studi orientali, 17, p. 399‑408.

DA’ĀDLI Tawfiq, 2011, “Mamlūk Epitaphs from Māmillā Cemetery”, Levant, 43/1, p. 78‑97.

DERAT Marie-Laure, 2018, L’énigme d’une dynastie sainte et usurpatrice dans le royaume chrétien d’Éhiopie, XIe au XIIIe siècle, Turnhout, Brepols.

DERAT Marie-Laure, 2020, “L’affaire des mosquées. Interactions entre le vizirat fatimide, le patriarcat d’Alexandrie et les royaumes chrétiens d’Éthiopie et de Nubie à la fin du xie siècle”, Médiévales, 79/2, p. 15‑36.

DIEM Werner, SCHÖLLER Marco, 2004, The living and the dead in Islam: studies in Arabic epitaphs, Wiesbaden, Harrasowitz, 2 vols.

Fauvelle-Aymar François-Xavier, HIRSCH Bertrand, 2008, “Établissements et formations politiques musulmans d’Éthiopie et de la Corne de l’Afrique au Moyen Âge : vers une reconstruction”, Annales islamologiques, 42, p. 339‑375.

Fauvelle-Aymar François-Xavier, MENSAN Romain, 2011, “Moules de coulées en pierre trouvés à Harlaa”, in Fauvelle-Aymar François-Xavier, HIRSCH Bertrand (eds.), Espaces musulmans de la Corne de l’Afrique au Moyen Âge, Paris, Centre français des études éthiopiennes - Éditions de Boccard, p. 88‑91.

Fauvelle-Aymar François-Xavier, HIRSCH Bertrand, CHEKROUN Amélie, 2017, “Le sultanat de l’Awfāt, sa capitale et la nécropole des Walasmaʿ. Quinze années d’enquêtes archéologiques et historiques sur l’Islam médiéval éthiopien”, Annales islamologiques, 51, p. 239-295.

GLEIZE Yves, 2022, “Islamic Burials. Muslim graves and graves of Muslims”, in KNÜSEL Christopher J., SCHOTSMANS Eline M. (eds.), The Routledge Handbook of Archaeothanatology, Abingdon/New York, Routledge, p. 377-394.

GORI Alessandro, 2007, “Nägaš”, in Encyclopaedia Aethiopica, Wiesbaden, Harrassowitz, vol. 3, p. 1107-1109.

Ibrahim Moawiyah M., 2001, “Tombs and their inscriptions from Nizwā and al-Ḥaymalī, Sultanate of Oman”, Proceedings of the Seminar for Arabian Studies, 31, p. 97‑113.

IMBERT Frédéric, 2021, “Les inscriptions arabes d’al‘Īss et de ses environs”, in ROUSSET Marie-Odile (ed.), Chalcis/Qinnasrin (Syrie) : De l’âge du Bronze à l’époque mamelouke. Qinnasrin II, Lyon, MOM Editions, p. 71‑98.

INSOLL Timothy, 2001, “Dahlak Kebir, Eritrea. From Aksumite to Ottoman”, Adumatu, 3, p. 39‑50.

INSOLL Timothy, 2020, “The Islamic archaeology of Ethiopia and the Horn of Africa”, in WALKER Bethany J., INSOLL Timothy, FENWICK Corisande (eds), The Oxford Handbook of Islamic Archaeology, Oxford, Oxford University Press, p. 417-445.

INSOLL Timothy, 2021, “The archaeology of complexity and cosmopolitanism in medieval Ethiopia: an introduction”, Antiquity, 95/380, p. 450‑466.

JUVIN Carine, 2010, “Les stèles du cimetière d’al-Maʿlā à la Mecque”, in AL-GHABBAN Ali Ibrahim, ANDRE-SALVINI Béatrice, DEMANGE Françoise, COTTY Marianne (eds.), Routes d’Arabie. Archéologie et histoire du royaume d’Arabie Saoudite, Paris, Musée du Louvre-Somogy, p. 491-499.

JUVIN Carine, 2013, “Calligraphy and writing activities in Mecca during the medieval period (twelfth-fifteenth centuries)”, Proceeding of the Seminar for Arabian Studies, 43, p. 153-156.

Klemm Rosemarie, Klemm Dietrich D., 2013, Gold and gold mining in ancient Egypt and Nubia: geoarchaeology of the ancient gold mining sites in the Egyptian and Sudanese Eastern deserts, Berlin/Heidelberg, Springer.

Lambourn Elizabeth, 2008, “Tombstones, Texts, and Typologies: Seeing Sources for the Early History of Islam in Southeast Asia”, Journal of the Economic and Social History of the Orient, 51/2, p. 252‑286.

Lloret GUTIÉRREZ Sonia, 2011, “Histoire et archéologie de la transition en al-Andalus : les indices matériels de l’islamisation à Tudmīr”, in VALERIAN Dominique (ed.), Islamisation et arabisation de l’Occident musulman médiéval (viie-xiie siècle), Paris, Éditions de la Sorbonne, p. 195‑246.

LOISEAU Julien, 2019, “The Ḥaṭī and the Sultan: Letters and Embassies from Abyssinia to the Mamluk Court”, in BAUDEN Frédéric, DEKKICHE Malika (eds.), Mamluk Cairo, a Crossroad for Embassies. Studies on Diplomacy and Diplomatics, Leiden/Boston, Brill, p. 638‑657.

LOISEAU Julien, 2020, “Retour à Bilet : un cimetière musulman médiéval du Tigray oriental (Inscriptiones Arabicae Aethiopiae 1)”, Bulletin d’études orientales, 67, p. 59‑96.

LOISEAU Julien et al., 2021, “Bilet and the wider world: new insights into the archaeology of Islam in Tigray”, Antiquity, 95, 380, p. 508‑529.

LOISEAU Julien et al., 2022, “To whom do the dead belong? Preliminary observations on the cemetery of Tsomar, eastern Tigray (Inscriptiones Arabicae Aethiopiae 2)”, in DURAND Caroline, MARCHAND Julie, REDON Bérangère, SCHNEIDER Pierre (ed.), Networked spaces: the spatiality of networks in the Red Sea and Western Indian Ocean, Lyon, MOM Editions, p. 647‑672.

Mershen Birgit, 2004, “Pots and tombs in Ibrāʾ, Oman. Investigations into the archaeological surface record of Islamic cemeteries and the related burial customs and funerary rituals”, Proceedings of the Seminar for Arabian Studies, 34, p. 165‑179.

MORDINI Antonio, 1957, “Un tissu musulman du Moyen Âge provenant du couvent de Dabra Dāmmò”, Annales d’Éthiopie, 2, p. 75‑79.

MORAES FARIAS Paolo Fernanado De, 2003, Arabic Medieval Inscriptions from the Republic of Mali. Epigraphy, Chronicles, and Songhay-Tuāreg History, Oxford, The British Academy & Oxford University Press.

Muehlbauer Mikael, 2021, “From Stone to Dust: The Life of the Kufic-Inscribed Frieze of Wuqro Cherqos in Tigray, Ethiopia”, Muqarnas, 38, p. 1‑34.

OMAN Giovanni, 1976-1987, La necropoli islamica di Dahlak Kebir (Mar Rosso), Naples, Istituto Universitario Orientale, 3 vols.

OMAN Giovanni, 1998, 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, Instituto Universitario Orientale, 2 vols.

ORY Solange, 1989, Cimetières et inscriptions du Ḥawrān et du Ǧabal al-Durūz, Paris, Recherche sur les civilisations.

Pansera Costantino, 1945, “Quattro stele musulmane presso Uogher Haribà nell’Enderta (fine sec.  IV Egira)”, in CONTI ROSSINI Carlo (ed.) Studi Etiopici racolti da C. Conti Rossini, Rome, p. 3‑7.

Pearce Nathaniel, 1831, The Life and Adventures of Nathaniel Pearce, written by himself during a Residence in Abyssinia, from the years 1810 to 1819, London, Henry Colburn and Richard Bentley, 2 vols.

PETERSEN Andrew, 2013, “The Archaeology of Death and Burial in the Islamic World”, in NILSSON STUTZ Liv, TARLOW Sarah (eds.), The Oxford Handbook of the Archaeology of Death and Burial, Oxford, Oxford University Press, p. 241‑258.

Rāġib Yūsuf, 1992, “Structure de la tombe d’après le droit musulman”, Arabica, 39/3, p. 393‑403.

AL-RĀSHID Sa‘ad b. ‘Abd al-‘Azīz (ed.), 2004, Aḥjār al-Mu‘allā al-Shāhidiyya bi-Makka al-Mukarrama, Riyadh, unknown ed.

Rossi Ettore, 1942, “Iscrizioni funerarie arabe musulmane del paese degli Arussi nel museo coloniale di Roma”, Rassegna di Studi Etiopici, 2/3, p. 277‑281.

SCHNEIDER Madeleine, 1967, “Stèles funéraires arabes de Quiha”, Annales d’Éthiopie, 7, p. 107‑122.

SCHNEIDER Madeleine, 1983, Stèles funéraires musulmanes des îles Dahlak (mer Rouge), Le Caire, Institut français d’archéologie orientale, 2 vols.

SCHNEIDER Madeleine, 2009, “Des Yamāmī dans l’Enderta (Tigre)”, Le Muséon, 122, p. 131‑148.

SMIDT Wolbert, 2004, “Eine arabische Inschrift in Kwiḥa, Tigray”, in Böll Verena, Nosnitdin Denis, Rave Thomas, SMIDT Wolbert G.C, Sokolinskaia Evgenia (ed.), Studia Aethiopica in Honour of Siegbert Uhlig on the Occasion of His 65th Birthday, Wiesbaden, Harrassowitz, p. 259‑268.

SMIDT Wolbert, 2009, “Eine weitere arabische Inschrift von der osttigrayischen Handelsroute: Hinweis auf eine muslimiche Kutstätte in der “dunklen Periode”?”, Aethiopica, 12, p. 126‑135.

TSAKOS Alexandros, 2019, “Materiality and Physicality of Medieval Manuscripts from Christian Nubia”, Cahiers d’études africaines, 236, p. 967‑992.

Haut de page

Notes

* * This paper is part of a project that has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (Grant agreement No. 72620).

1 The important Muslim necropolis on Dahlak Kabir island (Schneider 1986, Oman 1976-1987, Insoll 2001), today in Eritrea, is not discussed here because it has been studied elsewhere and developed in its own insular context.

2 Beside the 40 stelae studied by our team, five more inscriptions reportedly found in Bilet and published by previous scholars have not been located (Loiseau, 2020).

3 49% of the stelae refer to deceased males while 17% are still undetermined.

4 We thank the anonymous reviewer who suggested the enlargement of the formula may in fact serve to highlight the name of the deceased, which indeed always follows in the epitaphs.

Haut de page

Table des illustrations

Titre Fig. 1 – Map showing the location of the medieval Muslims cemeteries with inscriptions in Ethiopia (A) and in south-eastern Tigray (B).
Crédits S. Dorso, 2022
URL http://journals.openedition.org/remmm/docannexe/image/19303/img-1.png
Fichier image/png, 2,2M
Titre Fig. 2 – General plan of the Bilet cemetery showing grave markers and pottery deposits
Crédits © S. Dorso/Mission Bilet, 2022
URL http://journals.openedition.org/remmm/docannexe/image/19303/img-2.png
Fichier image/png, 1,4M
Titre Fig. 3 – General views of Qelqel Rway (A), Ḥabera (B), Meyda Zeylegat (C) and Tsomar (D)
Crédits © S. Dorso/Mission Bilet, 2019
URL http://journals.openedition.org/remmm/docannexe/image/19303/img-3.jpg
Fichier image/jpeg, 359k
Titre Fig. 4 – Schematic typology of medieval grave surface markers in Tigray
Crédits © S. Dorso, 2022
URL http://journals.openedition.org/remmm/docannexe/image/19303/img-4.jpg
Fichier image/jpeg, 100k
Titre Fig.5 – Diversity of the grave markers in Tigray
Crédits © S. Dorso/Mission Bilet, 2019
URL http://journals.openedition.org/remmm/docannexe/image/19303/img-5.png
Fichier image/png, 5,8M
Titre Fig. 6 – Grave markers in Bilet and Arra, Tigray
Crédits © S. Dorso, 2018
URL http://journals.openedition.org/remmm/docannexe/image/19303/img-6.jpg
Fichier image/jpeg, 349k
Titre Fig. 7 – Various implementations of the stelae within the grave markers
Crédits © S. Dorso/Mission Bilet, 2022
URL http://journals.openedition.org/remmm/docannexe/image/19303/img-7.jpg
Fichier image/jpeg, 576k
Titre Fig. 8 – Draft plans of the cemeteries in the Arra region (based on aerial photographs) showing the location of inscribed stelae at the time of their discovery
Crédits © S. Dorso, 2022
URL http://journals.openedition.org/remmm/docannexe/image/19303/img-8.png
Fichier image/png, 271k
Titre Fig. 9 – Bilet’s stelae discovered by the HornEast project presented on scale and ordered chronologically (the grey frame indicates twin stelae)
Crédits © S. Dorso, 2022
URL http://journals.openedition.org/remmm/docannexe/image/19303/img-9.png
Fichier image/png, 904k
Titre Fig. 10 – The “epigraphic reserve”
Crédits © S. Dorso, 2022
URL http://journals.openedition.org/remmm/docannexe/image/19303/img-10.png
Fichier image/png, 984k
Titre Fig. 11 – Engraving techniques and evidence for the preparation of the epigraphic support (in green: preliminary ruling of the epigraphic field, in orange: asymmetric frame of the decoration)
Crédits © S. Dorso, A. Lagaron, 2022
URL http://journals.openedition.org/remmm/docannexe/image/19303/img-11.png
Fichier image/png, 1,5M
Crédits © S. Dorso, A. Lagaron, 2022
URL http://journals.openedition.org/remmm/docannexe/image/19303/img-12.png
Fichier image/png, 196k
Haut de page

Pour citer cet article

Référence papier

Simon Dorso et Anna Lagaron, « The Medieval Muslim Cemeteries of Tigray (Ethiopia) »Revue des mondes musulmans et de la Méditerranée, 153 | 2023, 83-114.

Référence électronique

Simon Dorso et Anna Lagaron, « The Medieval Muslim Cemeteries of Tigray (Ethiopia) »Revue des mondes musulmans et de la Méditerranée [En ligne], 153 | 2023, mis en ligne le 07 juillet 2023, consulté le 29 septembre 2023. URL : http://journals.openedition.org/remmm/19303 ; DOI : https://doi.org/10.4000/remmm.19303

Haut de page

Auteurs

Simon Dorso

Aix Marseille Univ, Cnrs, Iremam, ERC HornEast, Aix-en-Provence, France ; simon.dorso[at]univ-amu.fr

Articles du même auteur

Anna Lagaron

Aix Marseille Univ, Cnrs, Iremam, ERC HornEast, Aix-en-Provence, France ; alagaron[at]hotmail.fr

Haut de page

Droits d’auteur

CC-BY-NC-SA-4.0

Le texte seul est utilisable sous licence CC-BY-NC-SA-4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Search OpenEdition Search

You will be redirected to OpenEdition Search