Introduction

1From 2002 to 2012 an excavation was conducted around the “Saint-Jean-de-Rousigue” chapel in Laudun-l’Ardoise, Gard. According to the archival sources, it was the priory church of Saint-Jean de Todon (Ardagna et al. 2010). It is located on part of a vast limestone plateau overlooking the Rhône valley, a site that also includes an 18-hectare protohistoric and Roman settlement named on the cadastral plan as “Camp de César.” By the end of the 10 excavation seasons, the church and its cemetery had been fully excavated. As well as the funerary area, several other sectors yielded traces of occupation from periods before the Medieval settlement. They include the remains of a late Republican temple, a building, and two funerary areas, one of which was complex and displayed an arrangement of human and animal remains that seems to be unique in the literature. In the south-western part of the site, there is an area consisting of a large cist tomb (SP 21628) and two children’s burials, one buried in the body of an amphora (SP 21674) and the other in a diaclase covered with tegulae. Tomb SP 21628 may resemble a “monumentalized” burial. It is the first tomb in this small area and is located in the bedrock. It is a composite cist comprised of large limestone slabs on the base and the walls (two slabs on the north side, one monumental one on the south side, and another thick one on the west side) as well as tegulae, made of lime mortar and pebbles, at the ends. It is the roof slab that suggests the monumentalization of the whole. Altogether, the three structures contained one adult and four juveniles. In terms of dating, the typology (the gabled construction and the amphora) is similar to that of other finds to the north of the church (walls and two gabled tombs containing very young juveniles), and thus indicates funerary use during Late Antiquity.

1. An unusual funerary arrangement from Late Antiquity

2Other than these three structures, this funerary area also contained a burial (SP 21229) used for a very young juvenile (fig. 8). The placement pit for this burial was not found. It consists of a cist tomb made of large shards of tegulae (parallelipedal cist, with one side leaning or gabled). Located in a foundation trench for a structure that has been totally looted, this area has been repeatedly disturbed. Because the pit and the entire northern part of this burial have been looted, the original form of the cist cannot be reconstructed.

Fig. 8. Photogrammetric reconstruction of SP 21086.

3Radiocarbon analysis dates this burial to between the 3rd and 5th century AD Following the excavation of the pit of this tomb, a separate funerary deposit was found (SP 21086). Excavation of this assemblage showed that SP 21086 consisted of a cluster of seven human crania along with animal bones, particularly bovine. Their layout suggests deliberate arrangement. The pile of bones was “demarcated” by animal bones and, on the western side, by a carved block placed on its side. Three of the seven crania were found right next to this stone block (US 21292, US 21296, US 21298). The whole northern edge of the assemblage may have been looted and disturbed by the digging and installation of SP 21229. Thus, the crania were placed next to this carved stone on top of an arrangement of disparate animal bones laid directly on the cut bedrock. This first group of animal bones comprises unconnected bovine long bones and an equine sacrum, while the anatomical connections already seen at the site suggest that the second group of animal bones correspond to a bovine hock.

4The idea that the human crania were deliberately arranged in a semicircle is also supported by a mark engraved in the rock over 4 to 5 cm. Five crania were joined together in a ring around a central cranium; only the westmost cranium was not in contact with the central cranium. Four of the crania were found with their superior face visible, while the initial deposit position of the others was uncertain. Likewise, the orientation of the faces was not completely homogenous. Four crania were facing south-east while three were facing north-west. None of the objects, such as the carved block or animal bones, polarises the others. Within this assemblage, the animal bones do not seem to have imposed any particular restrictions on the crania, although the hock may have served to mark the southern and eastern limit.

5As things stand, it is difficult to ascertain the post-depositional movement of the bones from this assemblage. Some elements may have tipped over and the northern edge has been disturbed by the placement of SP 21229. Nevertheless, the position of the crania remains compatible with the hypothesis that the assemblage was probably not deliberately filled with sediment. The analysis of the crania was difficult, particularly because of their rather poor state of preservation. There was significant fragmentation caused by the placement of burial SP 21229 and none of the crania can be deemed fully or partially complete. Nevertheless, we can identify three adult crania, two of individuals in late adolescence (15–19 years old), and one cranium that is in all likelihood juvenile. Likewise, examination of sex characteristics, although necessarily uncertain, suggests the presence of two males and one female. There was no dental pathology of note observed. Likewise, there was no trace of traumatic injury on the crania (blows, cuts) and no other signs of injury. Meanwhile, the condition of the animal bones shows that the alterations to them were not caused by sedimentary action but were made before the bones were deposited in the cist containing the crania. The presence of probably ancient fractures, the chewed extremities, and the cut marks found on several of the bones show that they form a random collection.

6Even with the aid of six radiocarbon dating analyses, doubts persist regarding the dating of the assemblage. Burial SP 21229 is dated by radiocarbon analysis to between the 3rd and 5^th centuries AD (between 263 and 427). The same applies to three sampled crania (between 231 and 418) and one of the animal bones (US 21298). This striking homogeneity is, however, called into question by the results of the dating of the bovine hock, which placed it between the 1st century B.C. and the 1st century AD. This is consistent with the comparative placement of the assemblage in the context of the chronological development of bone dimensions in Languedoc and Western France (Forest, Rodet-Belarbi 2010). The fact that some of the animal remains were in contact with the crania thus suggest either a more complicated or a more random hypothesis, associated with the multiple rearrangements of this area over time.

2. Origin of this unusual funerary arrangement

7Thanks to the radiocarbon dating, we can confirm that the deposit of crania does not correspond (at least directly) to pre-Roman cultural and ritual practices involving “severed heads”, known to have existed in Provence (Dedet 2011). The ancient authors, Greek and Roman, associated severed heads with the Celtic barbarians. Nevertheless, archaeology shows that the Romans also had a certain appetite, even a particular inclination, for decapitation (Voisin 1984) as well as for “headhunting.”

8In his analysis of accounts of this practice in the ancient authors, Voisin points out that the tendency to automatically attribute severed heads to Celtic activity obscures disciplinary practices among Gallic legionaries or borrowings of certain aspects of other religions in order to obtain “the support of the western provinces” (Voisin 1984: 244). More broadly, it is possible that headhunting and, at the very least, practices involving severed heads existed in the Roman army without being institutionalized. In that case, it would represent the activity of “uncivilized legionaries” among the auxiliaries. Moreover, “headhunting” was not exclusive to the Celtic world. The Romans also had a liking for the objects of power embodied by the head of the vanquished. In the British Isles, there are several archaeological examples of deposits of isolated crania from the Romano-British period (Tucker 2012: 80).

9In the case of Laudun, as well as this unusual assemblage of human crania, the excavation also uncovered a small series of well-defined burials.

10New ¹⁴C dating and more extensive research into rituals and practices in Narbonnaise (Aycard et al. 2001) are needed given that the complexity of the cult and cultural connotations of the “head” always requires caution. Does this example, unique as far as we know, represent an isolated autochthonous or allochthonous practice? The dates and general historical data point toward migrant populations. Auxiliary troops, merchants, invaders from another part of the Empire? Central or Eastern Europe? In any case, their presence must have been long-term enough to leave behind a funerary site with unique aspects. During our study, it became clear that their influence, particularly the potential symbolic and ritual impact of animal bones, must also be reconsidered. In conclusion, the layout of this secondary deposit and its spectacular arrangement are atypical for the period concerned. Both its origin and its relationship to the other structures from the ancient occupation of the site remain difficult to determine. Moreover, the presence of a tomb of a very young juvenile above it raises further questions regarding the simultaneity of the deposits or the possibility that a particular memory informs the topographic continuity of the site.