Introduction
The site of Ephesos, located in the south of İzmir in western Türkiye, near the modern city of Selçuk, is an eminent Hellenistic-Roman city in Asia Minor (Figure 1). In addition to being one of the most active harbours in the Mediterranean in the Greco-Roman period (c. eighth century BCE to fifth century CE), the city is renowned for its grand sanctuary to the Greek goddess Artemis (Figure 1) and for being the capital of the Roman province of Asia. The Church of Mary, situated south of the Olympieion and north-east of the Roman harbour (Figure 1), functioned as the episcopal church of Ephesos from the early fifth to sixth centuries CE. Probably built on the foundations of a disused Roman stoa for the Ecumenical Council of 431, it was later reconstructed as a double church—perhaps in the sixth century CE—before possibly being destroyed in the seventh century CE. Despite losing its episcopal status, the church and its immediate surroundings were used as a burial ground during the Middle and Late Byzantine periods. This period generally refers to the seventh to fifteenth centuries CE (Heberdey Reference Heberdey1907; Reisch Reference Reisch1932; Vetters Reference Vetters1986, Reference Vetters1988; Karwiese Reference Karwiese1989, Reference Karwiese and Karwiese1998, Reference Karwiese and Pillinger1999; Langmann Reference Langmann1991; Karwiese et al. Reference Karwiese, Beyll and Karwiese1994; Karwiese & Beyll Reference Karwiese, Beyll and Karwiese1994; Karwiese & Turnovsky Reference Karwiese and Turnovsky2000).
Figure 1. Location of the town of Selçuk in Western Türkiye (left) and location of the Church of Mary on the site (right) (© OeAW-OeAI/; figure by B. Danthine).
A series of 46 cist graves—mostly marble- or limestone-lined and occasionally tile-lined graves—found both inside and outside the church, were partially excavated between 1984 and 1986, and again in 1990, 1993, 1997, 2000 and 2001 (Vetters Reference Vetters1986, Reference Vetters1988; Karwiese Reference Karwiese1989, Reference Karwiese and Karwiese1998; Langmann Reference Langmann1991; Karwiese et al. Reference Karwiese, Beyll and Karwiese1994; Karwiese & Turnovsky Reference Karwiese and Turnovsky2000; see also online supplementary material (OSM) nos. 1 & 2). Twenty-four graves were identified in the part of the cemetery located to the north of the church, 11 within the church and 11 in the part of the cemetery located to the south of the church. Eleven of these graves (three in the northern cemetery, three in the church and five in the southern cemetery) were noted as not excavated (Karwiese Reference Karwiese1989). The 35 excavated graves consist either of individual primary burials or of burials containing two or three individuals, exhibiting successive deposition and post-depositional manipulation. Individuals found in primary positions (i.e. those whose remains were discovered in their original and final place of deposition) showed extended supine placement of the body, while those in secondary positions (i.e. intentionally moved from their primary location) appear to have undergone ‘reduction’ (reorganisation of elements to occupy a reduced space in the grave; Knüsel et al. Reference Knüsel, Gerdau-Radonić, Schotsmans, Knüsel, Gerdau-Radonić and Schotsmans2022).
Despite the numerous excavations conducted at Ephesos and the Church of Mary, uncertainties remain regarding the particularities of local mortuary practices, cemetery management and the dynamics of grave reuse. Thus, greater clarification of the characteristics of the burials—arrangements, post-depositional events and the relative chronology of deposits and subsequent interventions—is needed. Excavations in 2023 aimed to address this need through the investigation of intact burials using modern archaeothanatological methods (Duday et al. Reference Duday, Courtaud, Crubezy, Sellier and Tillier1990; Duday Reference Duday2009; Knüsel & Robb Reference Knüsel and Robb2016). The two main objectives of the campaign were: to generate new data on the mortuary chaîne opératoire (i.e. Valentin et al. Reference Valentin, Rivoal, Thevenet and Sellier2014) in relation to the biological profiles of the deceased; and integrate these findings into a broader reflection of practices at the site, following a retrospective archaeothanatological approach based on earlier excavation reports.
Material and methods
Between 10 and 28 July 2023, an area of 36m2 was excavated in the southern cemetery (Figure 2a). Three funerary structures (graves 1/23, 2/23 & 3/23) were identified (Figure 2b–d) and excavated following archaeothanatological principles. Archaeothanatology, or archaeology of death, aims to study the taphonomy of the human body and to reconstruct the original organisation of the grave in order to interpret mortuary behaviour and draw social conclusions (Duday et al. Reference Duday, Courtaud, Crubezy, Sellier and Tillier1990; Duday Reference Duday2009). The structures had not been identified during previous campaigns because this specific area had not been excavated.
Figure 2. Area excavated in 2023 (a & b) and the location of the graves (c & d) with regards to the Church of Mary. Scale: 0.5m (© OeAW-OeAI/; figure by C. Partiot & B. Danthine).
Grave 1/23 was first to be discovered, located against the south wall of the church, oriented north-west/south-east and superimposed on grave 3/23 (Figure 2d). Grave 2/23 is approximately 0.5m south of graves 1/23 and 3/23 with a similar orientation (Figure 2d). Two clusters of human bone were also discovered on the margins of graves 1/23 and 3/23, one on the north side (SE1003) and the other on the south side (SE1028) (see OSM no. 3).
Each of the three graves contained one individual: two adults and one immature individual. The goals of the biological study were to determine the number of individuals, reassociate bones, identify individual profiles and examine the health status of the individuals (Buikstra & Ubelaker Reference Buikstra and Ubelaker1994). Biological sex of the adult individuals was assessed probabilistically (using the Bruzek et al. (Reference Bruzek, Santos, Dutailly, Murail and Cunha2017) revision of the Murail et al. (Reference Murail, Bruzek, Houët and Cunha2005) method) and age at death was estimated based on the sacro-pelvic surface (Schmitt Reference Schmitt2005). Stature was calculated following Trotter and Gleser (Reference Trotter and Gleser1952, Reference Trotter and Gleser1958) and Cleuvenot and Houët (Reference Cleuvenot and Houët1993). For the immature individual, age at death was based on dental mineralisation and eruption (AlQahtani et al. Reference AlQahtani, Hector and Liversidge2010) and on stature assessment using logarithmic formulae (revised from Fazekas & Kósa (Reference Fazekas and Kósa1978) by Sellier and cited in Schmitt & Georges (Reference Schmitt, Georges and Charlier2008), validated on individuals up to four years old, see Partiot (Reference Partiot2018)). The presence of carious lesions in the dentition was assessed macroscopically and scored based on severity (0 = absent; 1 = minimal localised lesion; 2 = moderate lesion; 3 = extensive crown involvement) and location on the tooth. The presence of dental calculus was evaluated macroscopically and recorded as either present or absent. The presence of enamel hypoplasia was investigated macroscopically on the upper and lower incisors under natural light and with the aid of a magnifying lamp (×1.75). Dental wear was scored for severity (Smith Reference Smith1984). A macroscopic palaeopathological analysis was performed for each individual, and CT scans were carried out in one specific case at the Selçuk State Hospital.
To contextualise the findings in relation to previous excavation results, a cross-analysis of information from published reports was conducted, complemented by an archaeothanatological re-examination of this documentation, particularly drawings and photographs.
Regarding the results from the new excavation, and for clarity, an examination of the undisturbed individual burial (grave 2/23) is presented first, followed by the two superimposed burials (graves 1/23 & 3/23). The archaeothanatological re-examination of previously excavated graves is available in the OSM.
Grave 2/23 (SE1034)
Grave 2/23 is an individual cist burial, measuring 1.1 × 0.6m externally and 0.9 × 0.3m internally, with the upper part of the individual covered by a stone slab (Figure 3a). Inside the grave, the body of a young immature individual was laid out in a generally extended supine position, with the head oriented to the west and the face turned eastward. The overall anatomical integrity of the skeleton was preserved, with the exception of the right femur, tibia and ulna, which are missing (see OSM no. 4a, based on the recording sheets from Bocquentin & Anton Reference Bocquentin and Anton2021), and a fragment of the left radius, which was found above the feet at the top elevation level of the burial (see OSM no. 5a). The right upper limb was semi-extended, with the presence of phalanges and metacarpals in the pelvic cavity indicating that the hand had been placed on the abdomen, while the left upper limb was extended along the north wall of the grave, beside the body (Figure 3b). The lower limbs were extended; the right limb represented only by the proximal ossification centre of the tibia and the distal half of the fibula, both in anatomical proximity. The lumbar and lower thoracic spine, as well as the cervical spine, were either articulated or lying in anatomical proximity. Two beads were found near the cervical vertebrae.
Figure 3. Grave 2/23: a) in situ photograph, dashed line shows the placement of the slab above the upper body; b & c) details of archaeothanatological analysis. Scale: 0.5m (© OeAW-OeAI/; figure by C. Partiot).
The overall anatomical integrity of the skeleton and the preservation of epiphyses in anatomical connection indicate that this was a primary deposition, while the displacement of vertebral elements between the lower limbs suggests that the body was deposited in an empty space. The migration of these light spongious bones may have been caused by natural processes, including the release of bodily fluids, or possibly by the infiltration of water into the cist (Duday & Guillon Reference Duday, Guillon, Schmitt, Cunha and Pinheiro2006). Although there is no patterning in the sediments to support this latter hypothesis, local climate characteristics, alternating between very dry and very wet seasons, make water infiltration highly likely during the rainy period. These taphonomic factors are also likely responsible for the migration of a left metacarpal to the left ilium, beyond the extent of the delimitation effect discussed below (Figure 3). The dislocation of the right humerus and clavicle/scapula further suggests that the burial remained empty throughout the decomposition process.
Several indicators suggest that the body was placed directly against the bottom of the structure in a perishable soft and tight container, which maintained lateral and supero-inferior pressure at a distance from the structure’s walls (Figure 3; see also OSM no. 5). This containment resulted in delimitation effects, including the strict alignment of the left rib cage, left ilium and left lower limb. The position of the bones from the left foot suggests that the foot was held in a 90° anatomical position until joint decay caused the metatarsal to shift superiorly. This indicates that the perishable container also enclosed the feet. Both clavicles showed verticalisation. The craniofacial skeleton, including the mandible, was held against the chest by posterior support, as demonstrated by the support effect identified and the very limited left lateral rotation of the cranium with respect to the mandible.
Considering the proximity of the cranium to the north wall of the structure and the indications of pressure applied directly to the body, it is improbable that the perishable container was a wooden coffin. It is more likely that the body was wrapped in a shroud before being placed in the empty cist. No evidence of post-depositional anthropic disturbance was identified. The migration of the left ischium and a proximal fragment of the right radius to the upper filling level, along with the absence of the right femur and tibia, are likely attributable to the activity of a burrowing bioturbator following the completion of the body decomposition process—though no trace of a tunnel was observed in the sediment.
The age at death of the grave 2/23 individual is estimated to be between 1.5 and two years based on dental mineralisation and eruption, and approximately one year and nine months based on stature assessment (0.80±0.02m). No skeletal or dental pathologies were identified (see OSM no. 4b).
Grave 1/23
Grave 1/23 is a stone-lined individual burial, measuring 2.5 × 1m externally and 1.9 × 0.5m internally (Figure 4a). The north-east wall of the grave is relatively straight, while the south-west wall curves outward (Figure 4b). The body was laid out in a generally extended supine position with the head oriented westward and facing south. The upper limbs were slightly flexed, likely with the hands on the abdomen. The left lower limb was extended, with the foot positioned against a stone of the north-east wall, while the right lower limb was slightly flexed. The facial skeleton is not preserved. The primary nature of the deposition is evidenced by the overall anatomical integrity of the skeleton (excepting the facial skeleton, the carpal bones and part of both forearms, see OSM no. 6) and the fact that several labile joints (those that are usually considered to disarticulate more rapidly; Knüsel et al. Reference Knüsel, Gerdau-Radonić, Schotsmans, Knüsel, Gerdau-Radonić and Schotsmans2022: 670), are in anatomical connection, particularly in the feet.
Figure 4. Grave 1/23: a) in situ photograph; b) details of archaeothanatological analysis. Notes in green indicate evidence of decomposition in an empty space; in blue, post-depositional disturbance; in yellow, possible localised desiccation. Scale: 0.5m (© OeAW-OeAI/; figure by C. Partiot).
Various pieces of evidence, including bone displacements beyond the original volume of the body, suggest that the body decomposed in an empty space (Duday Reference Duday2009) before the grave was secondarily infilled (Figure 4b). Anatomical connections in the feet, between the metatarsals and phalanges, were preserved, possibly due to localised desiccation (Maureille & Sellier Reference Maureille and Sellier1996; Sellier & Bendezu-Sarmiento Reference Sellier and Bendezu-Sarmiento2013) or the presence of a perishable covering, such as a shoe. The migration of light spongious bones, seen in the superior shift of the right calcaneus and the displacement of the first metatarsal on the right foot (with proximal phalanx still attached) toward the left foot, may have been caused by natural processes, as discussed for grave 2/23, although again there is no sedimentary pattern supporting the hypothesis of water infiltration.
Extensive displacements and alterations were observed in the thoraco-abdominal region (displacement of thoracic vertebrae and ribs, superior flipping of the right first rib toward the mandible) and in the distal portion of the upper limb (postmortem fracturing of the right and left radius and ulna and of the left humeral diaphysis, postero-proximal tipping of the left radius found exposed with the distal extremity oriented proximally) (Figure 4b). These disturbances could be attributed to the impact of falling stones and/or a reopening of the grave when the body was at least partially skeletonised, possibly to retrieve jewellery (the two hypotheses are not mutually exclusive). In contrast, disturbances in the lower limbs were much less pronounced. There is no definitive evidence indicating the presence of a perishable container, such as a coffin or a shroud. The disconnection of the sacrum from both ilia, which might otherwise be interpreted as an indicator of lateral compression, is more plausibly attributed to the southeast-northwest tilting observed in the pelvic area and caused by an uneven underlying surface (see OSM no. 5b).
The individual from grave 1/23 is identified as a female who died after the age of 40, likely between 50 and 60 years old. Her stature is estimated at 1.56±0.05m. Extensive antemortem tooth loss (six teeth on the mandible, the maxillary bone was not preserved, see OSM no. 6b), with complete or ongoing healing of the alveolar bone, is observed, along with substantial wear to the enamel of the remaining teeth (scored at 7 out of a possible 8). No calculus or enamel hypoplasia are identified. Degenerative lesions consistent with age are present in the spine, including cervical, thoracic, lumbar and sacral arthrosis. Additionally, a supernumerary rib has fused with the seventh cervical vertebra on the right side. Although indicative of developmental anomalies in infants (Partiot et al. Reference Partiot, Guillon, Peressinotto, Castex and Maureille2020), the presence of a supernumerary rib in adults is generally non-pathological and typically asymptomatic.
The main palaeopathological finding is exhibited by the right lower limb. Macroscopic examination revealed extensive periosteal appositions on the postero-lateral surface of the proximal half of the tibial diaphysis (Figure 5f), and in a corresponding location on the fibula, resulting in the thickening and enlargement of diaphyseal diameter (Figure 5a). CT scans reveal substantial intracortical abscesses, known as ‘Brodie abscesses’. One is more than 110mm in length along the lateral side of the tibia (Figure 5g–j), another is approximately 40mm in length on the medial part of the fragmented fibula, displaying two external cloacae (holes for the drainage of pus) (Figure 5b–e). The presence of these intracortical fistulae strongly suggests chronic osteomyelitis, with fistulous tracts connecting to the interosseous space. This condition is most often associated with Staphylococcus aureus infections, which typically arise from wounds or hematogenous dissemination. The tibia is a common site for such infections (van Der Naald et al. Reference van der Naald, Smeeing, Houwert, Hietbrink, Govaert and van der Velde2019). Chronic osteomyelitis was not necessarily fatal before the development of modern medical interventions—because the regular evacuation of pus through cloacae, sometimes extending to the skin surface, can mitigate the abscess—but the condition likely contributed to a compromised general health status.
Figure 5. Comparison of the left (top) and right (bottom) fibulae (a) reveals pathological changes, including two cloacae (c), on the right fibula; vertical CT-scan cross-sections of the right fibula (b) show the intracortical ‘Brodie’ abscess and transversal CT-scan cross-sections through the right fibula (d & e) identify the cloacae. Periosteal appositions are visible on the right tibial diaphysis (f) and in CT scans (h); vertical (g & j) and transversal (i) CT-scan cross-sections show an underlying intracortical abscess. Scale: 50mm (© OeAW-OeAI/; figure by C. Partiot & F. Boursier).
Grave 3/23 and surrounding remains from SE1028 and SE1003
Removal of the interstitial sediment revealed that the base for grave 1/23 was the top of the underlying grave 3/23 (Figure 6a). This process also exposed a cluster of human bones in a secondary position at the centre of grave 3/23 (stratigraphic unit SE1025), and two distinct bone discard zones in the immediate surroundings (stratigraphic units SE1003 & SE1028; see Figure 6a & b and OSM no. 5b). The first, located on the northern side (SE1003), included four metatarsal fragments that were of adult size and maintained anatomical proximity, suggesting potential localised desiccation or the temporary preservation of a shoe (see OSM no. 8a). The second discard zone, situated on the southern edge at the level of the top of Structure 1 (SE1028, OSM no. 5b), contained 32 bone fragments from various anatomical regions of an adult individual (see OSM no. 8b & c).
Figure 6. a) The bottom of grave 1/23 and the top of grave 3/23, also showing discard zone SE1003; b) top layer of the reduction inside grave 3/23 (SE1025) and of the discard zone SE1028. Scale: 0.5m (© OeAW-OeAI/; figure by C. Partiot & E. Baudouin).
The upper layer of grave infill contained dispersed, disarticulated remains not in anatomical proximity, while the lower layer contained a central organised ‘reduction’, with the left humerus, right femur and right tibia organised as a bundle against the north-eastern and south-western walls of the grave, the left and right fibula, left femur and left tibia arranged obliquely, the left os coxa lying flat, and the cranium in the centre (Figure 7). The ‘wall effects’ identified on the western (proximal epiphysis of the right femur, distal epiphysis of the left femur) and eastern (distal epiphysis of the left humerus, proximal epiphysis of the right femur, antero-lateral side of the left coxal, distal epiphysis of the right fibula, distal epiphysis of the right tibia) sides of the central cluster (Figures 7 & 8). These suggest that the remains may have been placed in a perishable rigid container, such as a wooden box, deposited within the empty space of the grave.
Figure 7. Archaeothanatological analysis of grave 3/23: a) upper layer of the reduction with discarded elements on the north (SE1003) and south sides (SE1028); b) lower layer of the reduction. Scale: 0.5m (© OeAW-OeAI/; figure by C. Partiot).
Figure 8. Detailed view of the bone elements from the upper and lower layers of the central cluster in grave 3/23, which are involved in the wall effects on its western and eastern sides. The distal end of the left femur, not visible in the drawing of the upper layer in Figure 7, is shown in dashed lines (© OeAW-OeAI/; figure by C. Partiot).
The decomposition of this container and substantial taphonomic events—possibly including anthropogenic intervention—likely resulted in the partial collapse of the reduction and a dispersal of the remains into the western and eastern parts of the structure. The representation, morphology and taphonomic condition of the elements from SE1025, SE2028 and SE1003, are consistent with their being derived from a single individual. Six additional metacarpal and carpal bones likely percolated into grave 3/23 and have been reassociated with the individual from grave 1/23. This movement would indicate that grave 3/23 had not yet filled with sediment when grave 1/23 was established (see OSM no. 7a).
The individual from grave 3/23 is identified as male, and died between the ages of 20 and 39. His stature is estimated at 1.69±0.04m. Only three upper teeth are preserved, and two others were lost antemortem. Enamel wear is considerable (between stages 5 & 8), which prevents the observation of potential enamel hypoplasia or calculus (see OSM no. 7b). Aside from poor dental health, no specific pathology was identified in the skeletal remains.
Archaeothanatological cross-analysis and burial reuse practices
Re-analysis of reports from previous campaigns faced limitations, as data on the age at death and sex of buried individuals were rarely available, and catalogues provided in situ photographs and drawings for only a few individuals. Previous publications indicate that among the 35 excavated graves, 21 appear to have contained a single individual, 11 contained two individuals and three contained three individuals, for a total of 52 individuals. A quarter of the individuals (13/52) are children, just under three-quarters (39/52) are adolescents or adults. No immature individuals are reported as being buried inside the church and only about a quarter of the single burials (6/21) contain children. Of the 15 burials that contain two or three individuals (‘sépultures plurielles’; Knüsel et al. Reference Knüsel, Gerdau-Radonić, Schotsmans, Knüsel, Gerdau-Radonić and Schotsmans2022), immature individuals are present in two of the 11 double burials (one case an adult and a child; the other, two children) and in two of the three triple burials (one case with one child among two ‘reduced’ adults and the other with three children buried in a single grave). There does not appear to be any preferential selection for age at death in relation to the type of deposition.
A total of 19 graves excavated in previous campaigns are represented by photographs or drawings and could be re-examined from an archaeothanatological perspective (see OSM nos. 10 & 11). Seven are single graves that appear to be mostly primary deposits (preservation of labile connections and presence of small bones: graves 2, 11, 20, 26, 29 & 36), with one reduction (grave 27). Within these burials, decomposition occurred in an empty space for graves 20, 27 and 36 (bone displacements outside the body volume, pubic symphysis disjunction, flattening of the ossa coxae), with the individual from grave 2 possibly wrapped in a shroud and an unclear environment or possible rapid infilling for graves 11, 26 and 29.
The other 12 graves are sépultures plurielles and appear to be collective burials (successive deposits) rather than multiple burials (simultaneous deposits) (Boulestin Reference Boulestin2019; Schmitt & Déderix Reference Schmitt and Dederix2019; Knüsel et al. Reference Knüsel, Gerdau-Radonić, Schotsmans, Knüsel, Gerdau-Radonić and Schotsmans2022). This is, however, less clear for two burials (7 & 24), where the second individual appears to be represented only by a cranium, which may have been deposited at the same time as the primary individual or later. In most of the collective burials, subsequent deposition(s) occurred after the previously buried individual(s) had fully decomposed. Previous occupants of graves underwent reduction, with their bones either pushed along the northern and southern walls, bundled at the eastern end or placed simultaneously at both the western and eastern ends of the burial. It appears that precautions were taken to prevent the mixing of individuals during secondary manipulations. One apparent exception to this careful handling is noted by Karwiese (Reference Karwiese1989) in the seemingly carelessly discarded bones left outside of grave 24, in a manner similar to grave 3/23.
Within each grave, the individuals in primary positions were placed supine, with hands at the level of the chest or abdomen, and lower limbs extended. One individual from the southern cemetery (grave 44) exhibits unusual positioning, with the upper limbs flexed on each side of the thorax. While this position could reflect a deliberate choice during burial, it could also result from autolytic processes and the swelling of the abdomen as gas accumulates during decomposition—migration of upper limbs has been reported in such cases (Schotsmans et al. Reference Schotsmans, Georges-Zimmermann, Ueland, Dent, Knüsel, Gerdau-Radonić and Schotsmans2022). Alternatively, the individual may have been placed in a loosely wrapped shroud, with the arms initially crossed over the chest, in this case the upper limbs could have shifted position as the body was transported from the preparation site to the burial ground (e.g. Bonnabel & Carré Reference Bonnabel and Carre1996). Markers, such as the verticalisation of clavicles, suggest that the individuals found in primary positions were deposited within soft, tight and perishable envelopes, probably shrouds. Small items, such as ceramic fragments or coins, were occasionally found in the burials (Karwiese Reference Karwiese1989).
The coexistence of primary inhumation rites for both adults and immature individuals involving placement of the body, with or without a shroud, into an empty space, alongside secondary burial practices concerning reopening, reduction and discard, suggests that the results of the 2023 excavation campaign are in broad agreement with those of previous campaigns. The reduction of an individual within a single burial, without the addition of a new deposit, as observed in grave 3/23, is also documented in earlier reports (grave 27, see OSM), though in that case the bones were pushed to one side of the structure, while they were shifted to the centre of the burial and possibly positioned in a perishable container in grave 3/23. Discard outside the structure is mentioned only for graves 12 and 24, although no photographs or drawings are available. The new analyses do, however, offer a more comprehensive investigation of the mortuary chaîne opératoire and the sequence of burial reuse.
Possible interpretations for funerary treatment evidence in graves 1/23 and 3/23 highlight practices previously undocumented at the site, including the vertical reuse of the burial structure and the centralised reduction of remains with the possible use of a wooden box. Several scenarios can be proposed based on the observed contrast between the careful reduction within the structure and the careless discard of remains outside it (Figure 9).
Figure 9. Mortuary chaîne opératoire of graves 1 and 3 under hypotheses H1, H2 and H3. In yellow, events associated with an in situ primary deposition; in blue, events associated with an in situ secondary deposition and reintervention; in orange, events associated with an ex situ secondary deposition; in green: taphonomic events and bioturbations (© OeAW-OeAI/; figure by C. Partiot).
The first (H1) and second (H2) scenarios both involve the primary deposition of a body in grave 3/23 with subsequent decomposition in an empty space before the reopening of the grave. This reopening may have been for the purposes of performing the reduction before looting could occur (H1) or looting, with concomitant discard, may have been the primary driver (H2), with the reduction of the rest of the bones in a wooden box occurring afterward, before the construction of grave 1/23. In both scenarios, it is then possible that anthropogenic and/or natural disturbances occurred after the deposition of the individual in grave 1/23, with such disturbances typically affecting localised areas of the skeleton while leaving others relatively intact (i.e. Gleize et al. Reference Gleize2016).
A third option (H3) is also possible: the individual from grave 3/23 may have decomposed elsewhere, undergone reduction and then been transported for secondary deposition in the wooden box, placed centrally in the grave. Anthropogenic intervention, causing the external discarding and internal dispersion of the bones, happened later, potentially coinciding with the construction of the upper structure. This type of secondary deposition, involving the relocation of remains from another site, has no precedent in earlier excavation reports from the Church of Mary in Ephesos.
Conclusion
Application of archaeothanatological methods in the excavation of three new graves and of an archaeothanatological lens to the reanalysis of excavation reports from previous campaigns at the Church of Mary has provided new insights into the complexities of burial reuse and secondary burial in Byzantine Ephesos. The identification of vertical reuse of grave structures and of the central placement of reduced remains within a possible perishable container offers a nuanced perspective of the funerary chaîne opératoire. These results demonstrate the potential contributions of detailed archaeothanatological analysis to understanding burial arrangements and the spatial dynamics of funerary space, both at the level of an individual burial and in a site-wide survey. Together with osteobiographical analysis, these findings contribute a deeper understanding of the life histories and funerary customs of this population during the Middle and Late Byzantine periods.
Acknowledgements
We offer our warmest thanks to the team from the Austrian excavations in Ephesos, as well as to the workers and the hospital imaging service from Selçuk. We also thank Sacha Kacki (CNRS UMR5199 PACEA), Mélie Le Roy (Bournemouth University), Géraldine Sachau-Carcel (CNRS UMR7268 ADES) for their archaeothanatological opinions, Brigit Danthine and Bettina Schwarz (Austrian Archaeological Institute (OeAI) at the Austrian Academy of Sciences (OeAW)) for the archives, and the reviewers for their useful comments. This study was carried out in Selçuk within the framework of the Department of Historical Archaeology at the Austrian Archaeological Institute led by Martin Steskal, and of the Bioarchaeology Lab – Archaeological Sciences co-ordinated by Andreas G. Heiss and Alfred Galik.
Funding statement
The study was funded by the Austrian Archaeological Institute at the Austrian Academy of Sciences.
Conflict of interest
The authors have no conflict of interest to declare.
Data availability statement
The data related to the archaeothanatological and biological study of the individuals are available from the authors upon reasonable request. Skeletal material is curated at the excavation house of the Austrian Archaeological Institute (OeAI) in Selçuk (Türkiye).
Online supplementary material (OSM)
To view supplementary material for this article, please visit https://doi.org/10.15184/aqy.2025.10249 and select the supplementary materials tab.
Author contributions: CRediT Taxonomy
Caroline Partiot: Conceptualization-Lead, Data curation-Lead, Formal analysis-Lead, Investigation-Lead, Methodology-Lead, Visualization-Equal, Writing - original draft-Lead. Elise Baudouin: Conceptualization-Supporting, Formal analysis-Supporting, Investigation-Supporting, Methodology-Supporting, Validation-Supporting, Visualization-Equal, Writing - original draft-Supporting. Fabian Hermann: Investigation-Supporting, Writing - original draft-Supporting. Frédéric Boursier: Formal analysis-Supporting, Investigation-Supporting, Validation-Supporting, Writing - original draft-Supporting. Martin Steskal: Conceptualization-Supporting, Data curation-Equal, Formal analysis-Supporting, Funding acquisition-Lead, Investigation-Equal, Methodology-Supporting, Project administration-Lead, Resources-Lead, Supervision-Lead, Validation-Supporting, Writing - original draft-Supporting.
Open access
