Bulletin de la Société préhistorique française 2023

La grotte de la Fosse Marmandrèche à Port-d'Envaux (Charente-Maritime, France)  : des céramiques du Bronze ancien, mais des restes humains du Bronze final Sépultures et autres dépôts humains du Bronze final en Poitou et Charentes

Jacques Gachina, Bruno Boulestin, José Gomez de  Soto, Céline Trézéguet

Résumé  : La grotte de la Fosse Marmandrèche s'ouvre sur le territoire de la commune de Port-d'Envaux, en Charente-Maritime. C'est au cours de son exploration, en 1983, que des spéléologues y recueillirent quelques ossements animaux et humains, des tessons de céramique et de rares artefacts lithiques. Le mobilier céramique, très fragmenté et constitué uniquement de récipients à paroi épaisse, paraît globalement cohérent avec une série homogène datable du Bronze ancien saintongeais, sans que l'on puisse exclure qu???une partie puisse dater du Bronze moyen. Les restes humains, par contre, sont datés du Bronze final par le radiocarbone. Ils se rapportent à au moins deux individus, un adolescent et un sujet adulte ou de taille adulte.

L'assemblage humain de la Fosse Marmandrèche est actuellement le seul attesté pour le Bronze final en cavité en Saintonge. Dans le reste du Centre-Ouest, les vestiges humains datant de cette période sont également peu nombreux, et dans leur grande majorité ils ont été ramassés dans des cavités karstiques charentaises. Leur attribution chronologique repose essentiellement sur des datations radiocarbone, du mobilier contemporain leur étant rarement associé. Ces dépôts humains pourraient pour partie être de nature funéraire, mais le statut de certains demeure ambigu. En dehors des cavités, les sépultures du Bronze final sont rares en Centre-Ouest, qui ne sont connues que sur trois sites.

La grotte de la Fosse Marmandrèche apporte donc une documentation nouvelle à la fois sur le Bronze ancien saintongeais et sur les dépôts humains du Bronze final en Centre-Ouest, deux sujets dont nous ne savons presque rien. Elle rappelle en outre que les associations chronologiques apparentes entre mobiliers et vestiges humains ne sont pas toujours réelles. Elle donne enfin l'occasion de souligner l'intérêt des datations radiocarbone des vestiges humains, à la fois pour la raison précédente et pour documenter les pratiques mortuaires et leur chronologie.

Mots-clés  : Bronze ancien, Bronze final, Centre-Ouest de la France, pratiques mortuaires, dépôts humains en cavités, dates radiocarbone.

Abstract: The cave of the Fosse Marmandrèche is located near the village of Port-d'Envaux, in the French department of Charente-Maritime. This small cave about forty metres wide, is only accessible today through an opening in the roof and is largely filled in with fallen rocks. Upon its exploration in 1983, speleologists collected animal and human bones, ceramic sherds and rare lithic artefacts.

The very fragmented ceramic material consists only of thick-walled vessels. The forms that can be identified correspond to barrel-shaped vases, a vase with a conical upper part, another with a constricted opening and subvertical neck, and a jar. The decorations identified are mainly digital impressions, pustules, cords, cord impressions and pellet decoration. Vases decorated with pellets appear in Central West France in Beaker contexts and become more numerous from the Early Bronze Age. In inland Central West France, they disappear during the Middle Bronze Age with the development of the Duffaits culture, but they persist at this time in the maritime zone, as well as in Aquitaine, where they are also attested from the Early Bronze Age. In the Central West France maritime area, the barrel-shaped vases can also be attributed to the Early or Middle Bronze Age, as can digital decorations and cords arranged in a net pattern. On the other hand, the cord-impressed sherds and the vase with a closed orifice, probably a biconical vessel, are characteristic of the Early Bronze Age. Finally, the ceramic material as a whole is coherent with a homogeneous series that can be dated to the regional Early Bronze Age. Nevertheless, this homogeneity is only conjectural, as some of the ceramics may also date to the Middle Bronze Age.

The human remains, numbering a little under thirty, are from a minimum of two individuals: an adolescent aged 15 to 20 years and an adult or adult-sized individual. The results of the radiocarbon dating of two bone fragments date the remains to the Late Bronze Age. The significant difference in the two dates points to the successive deposits of a collective burial, if these deposits are in fact funerary, which, given the data at our disposal, is not entirely clear.

The human assemblage of the Fosse Marmandrèche is currently the only one attested for the Late Bronze Age in caves in Saintonge, where burials from this period are only known at the Borderie site, at Saint-Pierre-d'Oléron. In the rest of Central West France, deposits of human remains dating from the Late Bronze Age are also rare, and for the most part very poorly known, being only documented by speleological discoveries or collectings made during surveys in the karstic cavities of the Charente department. In these cases, human remains seem rarely associated with contemporary material, the only apparent exceptions being the La Tour cave at Vouthon and the Licorne cave at La Rochefoucauld-en-Angoumois, discovered in 2021. The chronological attribution of these remains is therefore largely based on radiocarbon analyses, with all the dates obtained covering the entire Late Bronze Age. Funerary deposits in the karst cavities of Central West France are known from the Middle Neolithic, well documented for the Early and Middle Bronze Age, and still attested in the Hallstatt D and La Tène A periods. Thus, some of the Late Bronze Age deposits could also be funerary in nature, expressing a continuity in traditional regional practices. However, it is uncertain whether all the human deposits in underground environments belong to the funerary sphere, the status of those found in karstic wells in Charente, at the Four du Diable in Bunzac and at Aven Victor in Torsac, or of the isolated skull in the Rancogne cave, being particularly ambiguous. Except for those found in caves, human assemblages from the Late Bronze Age are very rare in Central West France. In addition to the burials from Saint-Pierre-d'Oléron, only two inhumations are known at La Viaube 1 in Jaunay-Clan, in the Vienne department, and a cremation at Luxé, in the Charente department. It should also be remembered that the ditched enclosures, too systematically presented as funerary structures and which in some cases date back to the Late Bronze Age, have not yielded any human remains from this period.

In fact, the contribution of the Fosse Marmandrèche cave is three-fold. Firstly, it has provided a collection of ceramics attributable to the Early Bonze Age, a period about which we know practically nothing for the Saintonge area. Secondly, it offers the only human assemblage in an underground environment from the Late Bronze Age in this same region, human deposits dating from this phase of the Bronze Age being extremely rare in the entire Central West France. Finally, the Fosse Marmandrèche provides an opportunity to recall that human remains and material discovered in close proximity to each other and which may at first sight appear to be chronologically associated are not necessarily contemporary. This is particularly true for finds in caves, but it is also true for open-air sites. This highlights the importance of the radiocarbon dating of human remains, both for the above reason and to document mortuary practices and their chronology.

Keywords: Early Bronze Age, Late Bronze Age, Central West France, mortuary practices, human deposits in caves, radiocarbon dates.

Le dolmen de Peyre Dusets (Loubajac, Hautes-Pyrénées) Mégalithisme, terre crue et pierre consacrée du Néolithique récent ouest-pyrénéen

Pablo Marticorena, Emmanuel Mens, Marylise Onfray, Francis Bichot

Résumé  : Dans le cadre d'un projet collectif de recherche (PCR) portant sur le mégalithisme des Pyrénées nord-occidentales, une opération de sondage a été effectuée sur le dolmen de Peyre Dusets (Loubajac, Hautes-Pyrénées). Ces travaux ont révélé une structure de type caussenarde datée du Néolithique récent régional (3 800-3 300 BC), construite autour d'un menhir préexistant et placée au centre d'un tertre de terre massive. En outre, l'exploration du pied de la dalle de chevet a révélé la présence d'une petite stèle anthropomorphe.

Ces résultats, replacés dans une vision plus globale, viennent participer au renouveau des problématiques sur le mégalithisme et permettent d???offrir pour notre région, restée trop longtemps à l'écart des grands pôles de recherche, un autre regard sur les sociétés néolithiques. Ainsi au-delà de la découverte de cette stèle qui apparaît comme un élément de premier ordre faisant écho à des sites prestigieux du mégalithisme ouest-européen, l???importance du travail de la terre, le soin et l'investissement apporté à l'édification du monument, participent à l'émergence d'une image renouvelée du mégalithisme et subséquemment du Néolithique de notre région. Il apparaît dès lors que la vision proposée par une large frange de la bibliographie disponible sur notre secteur, d'un territoire parsemé de dolmens simples de chronologie tardive et aux dépôts modestes, parcouru par de petits groupes de pasteurs transhumants à l'écart de tout et (surtout) des phénomènes socio-économiques en jeu ailleurs aux mêmes moments, est davantage le reflet d'un état de la recherche que d'une réalité historique. Aujourd'hui, il semble ressortir au contraire l'image de sociétés aux organisations territoriales et sociales fortes, bien moins en marge des grands phénomènes sociaux-économiques ou symboliques visibles à des échelles plus larges et bien plus dynamiques que ce qui avait pu être évoqué jusqu'à présent.

Mots-clés  : Néolithique récent, mégalithisme, dolmen, architecture, matière première, terre crue, technologie, géoarchéologie, Hautes-Pyrénées.

Abstract: The North-Western Pyrenees offers a remarkable archaeological potential for the Neolithic period in a geographical key sector of Western Europe, between the Atlantic Ocean and the Pyrenees, with natural passages to the Iberian Peninsula. However, the region has been excluded from major research with a noted absence from national overviews.

In 2015, we developed a Collective Research Project (PCR) on the theme of Considering megalithism as a gateway, with the aim to improve our understanding of regional megalithism, its forms, its symbolism and its chronology, through multi-method and multi-scalar approaches and the Neolithic and recent Prehistory of the north-western Pyrenees. Our study focused on the emergence of megalithic building and its development on a more global scale.

Known as Pierre d'Uzès, the dolmen of Peyre Dusets is mentioned on the cadastral maps of Lourdes (1812) and Loubajac (1813) as a boundary marker between the two municipalities. E. Dufourcet describes the first excavation and a survey in a paper published in 1876. The dolmen is mentioned subsequently, but no other excavations seemed to have taken place.

The first study of the monument mentions a megalithic chamber composed of four monoliths including a cover slab and three orthostats, one of them can be interpreted as the chevet stone, facing the East. The four slabs are all of "Cauterets" granite, probably of the same origin from the Mindel moraine that outcrops to the south of the site. The architectural and geological studies, as well as a geophysical survey of the monument (electrical map and tomography) raises questions as to the chamber's architecture, its position and the composition of the tumulus. A survey in 2019 aimed to answer these questions. Our work used an ???evaluation??? type approach that is minimally invasive and adapted to our research. We have made three surveys on the Peyre Dusets site.

We were not able to document the sepulchral levels, due to the excavation. We did however survey the inside of the chamber that seems to have been slightly over-deepened, forming an elongated hollow in the direction of the length of the chamber. The excavation also provided details on how the monoliths were put into place, by demonstrating that blocks no  2 and no  3 were put in first. The orthostat simply stands on the local clay substrate, while the bedside slab is wedged in a pit dug into the clay. The pit has a double-layered filling. The lower part is made up of a dense array of different rocks (schist, granite and sandstone) almost exclusively yellow kneaded clay. The upper layer contains balls of yellow and white clay clearly visible during excavation with an average diameter of 8 to 10 cm and a few scattered stones.

The radiocarbon dating places the monument's construction in the recent Neolithic period (4735 ± 30  BP) and several clues suggest that the bedside slab was first a menhir erected before the building of the dolmen. The date is consistent with the known data for this type of monument in the area.

A shale slab with a particular shape was discovered during the excavation of the chevet stone wedging pit. The slab's form and the context of its discovery suggest that it is an anthropomorphic stele.

The excavation of the tumulus revealed an earthen feature preserved to a height of nearly 1.80 m and built using the massive dirt technique. The preparation of the ground, the implementation of a foundation layer and the finely prepared character of the materials are all indicators of the investment made in the building of the monument.

These results, placed in a more global setting renew research into megalithism in our region giving it its rightful place as a major reference for Neolithic studies.

Beyond the discovery of the stele that mirrors the prestigious megalithic sites of Western European, the importance of the mound, the care and the investment brought to the building of the monument, renew our knowledge of megaliths and of the Neolithic in our region. Until recently, the vision shared and transcribed in archaeological literature, reflected more a history of the research carried out in the area than a historical reality. It depicted a landscape dotted with simple dolmens, walked by shepherds in small transhumant groups, away from everything and (especially) from the socio-economic phenomena that were playing out elsewhere at the time. Recent work on West Pyrenean megaliths, whether it falls within the framework of our research project that focuses on the dolmen of Peyre Dusets, or within the framework of preventive excavations as on the site of the CM10 in Lannemezan, offer a vision quite far from this "image d'Épinal". They all show that the building of monuments in the area happened within strong technical and symbolic systems that led to major changes in the landscape, these dynamics being completely synchronous with those in play in the rest of Western Europe.

Keywords: Late Neolithic, Megalithism, dolmen, architecture, raw material, earthen construction, technology, geoarchaoeology.

Les chasseurs néandertaliens d'aurochs de La Borde (Livernon, Lot) : apport de l'archéozoologie

Malika Rivière, William Rendu, Jacques Jaubert

Résumé  : L'organisation des activités au sein du territoire des Néandertaliens est au cour de la recherche sur leurs sociétés. Le site de La Borde, présentant un spectre monospécifique centré sur l'aurochs, a été l'un des premiers pour lesquels les hypothèses de chasse collective et de son utilisation comme site d'abatage en masse ont été évoquées. Cependant, ces interprétations reposaient principalement sur une interprétation générale sans que l'assemblage faunique, issu initialement d'une fouille de sauvetage conduite en 1971, ait réellement été étudié du point de vue des comportements de subsistance, limitant notre perception des activités conduites localement. Par une reprise des collections osseuses du gisement, incluant restes déterminés et une sélection des indéterminés, nous proposons ici une nouvelle étude taphonomique et archéozoologique. Les résultats permettent une réévaluation du nombre d'individus et de confirmer en la précisant l'origine anthropique de leur accumulation. Bien que le profil squelettique s'avère biaisé par les conditions de fouille, les spécificités des modalités d'exploitation des carcasses témoignent de leur consommation expédiente et de l'abandon sans traitement de certaines. L'accumulation serait le résultat de multiples évènements de chasse d'un nombre d'individus suffisamment élevé pour permettre aux chasseurs d'être sélectifs dans leur exploitation. Le site aurait ainsi servi de lieu d'approvisionnement en matière carnée via des chasses spécialisées à l'aurochs tout au long de l'année pour les Néandertaliens fréquentant les causses du Quercy.

Mots-clés  : La Borde, Pléistocène moyen, archéozoologie, taphonomie, subsistance, aurochs, industrie lithique.

Abstract: Recent research on Neanderthal subsistence strategies in southwestern Europe identify the development, at the end of the Middle Paleolithic, of task specific locations dedicated to the capture of a large number of prey (Delagnes et Rendu, 2011). One of the main characteristics of these sites is their monospecific faunal spectrum centered on large mammals. However, it has been proposed that these task specific locations evidence the emergence of specialized hunting (Costamagno et al., 2006) at the end of the Mousterian era (starting during MIS4). Earlier specialized faunal spectra have been identified in the region (Jaubert et al., 2005 ; Discamps et al., 2011; Discamps et Royer, 2017) and the question of the development of such hunting specialization in earlier periods of the Middle Paleolithic can be raised.In this context, the site of La Borde (Livernon, France) is of prime interest to discuss the development of specific hunting strategies by Neanderthal populations. Localized in the Quercy and excavated by M. Lorblanchet in the 70???s, La Borde is a skinhole, which yielded a large faunal assemblage almost exclusively composed of aurochs remains associated with a denticulate Mousterian (Jaubert et al., 1990). The carcass accumulation was interpreted as resulting from communal hunting conducted by Neanderthal on the large bovine (Jaubert and Brugal, 1990). However, very limited zooarchaeological data were available, limiting our perception of the hunting strategies developed by the Neanderthals. Thus, we propose here a reevaluation of the site based on a combine taphonomic and zooarchaeological approach. The study was realized in two steps: firstly, the collection of La Borde stored at PACEA (CNRS ??? Bordeaux University) was analyzed in the framework of a Master 2 thesis (Rivière, 2018). This collection was constituted of identified bones selected by Slott-Moller (1990) for his paleontological analysis. A selection of the Cabrerets collection, which includes the faunal remains considered as unidentifiable, was included to the analysis in order to precise the faunal spectrum and to correct the MNI. During our study, 5770 bones remains and 497 teeth were analyzed. The aurochs largely dominates the faunal spectrum (with a   %  NISP of 95    %), which is characterized by a high rate of identification (28  %) due to the over-representation of large remains, a direct consequence of the excavation strategies. New species were identified in our revision with the discovery of a hyena canine and five remains of bear. Based on the faunal spectrum and its association with the denticulate Mousterian (Jaubert, 1990), an attribution of the site to the MIS7 is proposed. The absence of digested bones, the scarcity of carnivores and the absence of their young individuals conducted to reject these agents for the origin of the bone accumulation. On the opposite, the human modifications on the material associated with the large number of silex, sustain the hypothesis of a human accumulation. The taphonomic analysis of the material confirms the important alteration of the cortical surface of the bone mainly due to the weathering and some calcite depositions. However, these modifications seem to have only a limited impact on the representation of the human modifications and cannot explain their relative scarcity. The reexamination of the aurochs bone assemblage leads to a large increase of the previously proposed MNI to now 63 individuals (based on the second upper molar). The skeletal profile has been largely biased during the rescue excavation and no significant correlation were identified between the frequency of the anatomical parts and their density or their nutritive utility (tested by the FUI and the MUI). The skull and the belts are over-represented compared to the long bones. The phalanges are found complete with no evidence of human intervention. The limited human impact (8,1  % of the remains of the Bordeaux collection) is characterized by cut marks, notches and retouchers. Disarticulation and filleting are attested but bone fragmentation for recovery of the marrow and grease was limited. It results from it a limited number of green fractures (observable on only 54  % of the remains) and the prevalence of long bone articular extremities attests that they were not exploited for the extraction of grease. Simultaneously, the phalanges are found complete with no evidence of human intervention. All these elements suggest a limited investment of the butchery activities. The seasonal indication obtained by Slott-Moller (1990) based on tooth eruption sequence indicate multiple periods of death around the year. Based on our results and by taking into account the age profile established by Brugal (1995), we conclude that the accumulation is the consequence of multiple events of mass slaughtering of aurochs resulting from a specialized hunting on this animal. The site would have been secondarily scavenged by carnivores. Thus, our study brings some quantify data confirming the first interpretation of the site made by Jaubert and colleagues (1990), but simultaneously it adds some nuances. The hunters of La Borde aurochs took advantage of the topography of the landscape to bring down and trap the animals, and then processed the carcasses to obtain the resources they needed. The limited human impact on the carcasses supports the hypothesis of their large number and a quantity of food to process larger than what the group needed or was able to transport. The presence of all skeletal elements but in imbalance proportion and the low rate of butchery marks could correspond to one of the following exploitation modalities: 1) either the carcasses were only slightly or not at all invested, the first individuals that fell in the trap being neglected because they were buried under their fellow animals, or 2) some animals may have been the object of a more selective exploitation. The site was probably used recurrently during the year (Slott-Moller, 1990), and no evidence of differed consumption have been identified. Thus, the question of a residential mobility may be here raised. The topography characteristic of the site and the possible presence of aurochs in large number all along the year is probably the reasons why humans decided to invest this specific location, finding there the assurance to have always at their disposition a large resource of food. The conclusion of our study highlights the originality of La Borde in the Middle Paleolithic record of Western Europe with a monospecific faunal spectrum resulting in the slaughter of a large number of aurochs at least 130 ky ago. This study helps to document the diversity of the subsistence strategies developed by Neanderthal during the Pleistocene.

Keywords: La Borde, Middle Pleistocene, zooarchaeology, taphonomy, subsistance, aurochs, lithic industry.