20-2017, tome 114, 4, p. 637-658 - Adriana Soto Sebastián - Changements technologiques pendant l?€?Holocène dans le bassin de l?€?Èbre. Stratégies de production et de gestion de l?€?industrie lithique d?€?Atxoste (Álava, Espagne)

Les coquilles marines sont utilisées de façon récurrente par les populations mésolithiques de l’Europe en tant que parures. La reprise des fouilles de la grotte Rochefort (Saint-Pierre-sur-Erve, Mayenne) vient compléter le corpus existant. Un lot de 39 coquilles percées y a été découvert au côté de parures en dents animales et humaines, de restes osseux fauniques et de pièces lithiques. Les données stratigraphiques et le contexte géoarchéologique de la grotte Rochefort permettent d’attribuer les parures en coquillages au Mésolithique. Nos objectifs sont, à partir de leur détermination spécifique, de leur analyse taphonomique et de la description de leur stigmates de percement et d’usure, d’expliquer les raisons de la présence de ces coquilles dans les couches d’occupation de la grotte Rochefort, de savoir si ces coquilles ont été façonnées dans la grotte elle-même, de savoir si elles ont été portées, d’en définir les modes d’acquisition et de replacer ces parures dans le contexte du Mésolithique européen.

Leur détermination spécifique révèle une diversité importante des origines potentielles de ces artefacts et de leur mode d’acquisition : soit à l’état d’épaves échouées sur les plages atlantiques et méditerranéennes mésolithiques, soit à l’état de fossiles dans des gisements fossilifères. Leur opposition avec les coquillages consommés à l’époque souligne un mode d’acquisition spécifique dédié à cette utilisation. Elle souligne l’importance que ces populations mésolithiques leur prêtaient au sein de la sphère symbolique. Le fait que ces parures soient uniquement percées n’est sans doute pas anodin. En préservant leur forme, elles véhiculent l’image d’un environnement majeur dans le mode de vie de ces chasseurs-cueilleurs totalement dépendant des ressources naturelles. Le façonnage de ces parures n’a pas pu être mis en évidence au sein même de la grotte Rochefort. Elles semblent avoir été intensivement portées avant d’avoir été perdues sur le sol d’occupation de la grotte durant le Mésolithique. L’analyse des parures en coquillages de la grotte Rochefort montre de fortes similitudes avec celles du Mésolithique européen (mollusques utilisés, transformation des supports, localisation des stigmates d’utilisation…). La grotte Rochefort est un jalon supplémentaire, pour le Mésolithique, du témoignage de réseaux d’échange sur de longues distances aux multiples influences.

Mots-clés : parure, coquillage, Mésolithique, fossile, territoires d’acquisition, grotte Rochefort.

Marine shells were constantly used as ornaments by the Mesolithic populations of Europe. The resumption of excavations in Grotte Rochefort (Saint-Pierre-sur-Erve, Mayenne, France) helps complete the existing corpus. The scientific interest of these ornaments is also underlined by the distance of the site from the sea, over 150 km at the time of its occupation. A series of 39 shells was discovered in Grotte Rochefort alongside other ornaments made from animal and human teeth, animal bone remains and lithic artefacts. The stratigraphic data and the geo-archaeological context of the cave allow us to attribute these shell ornaments to the Mesolithic. A basic description of these ornaments is proposed: specific determination, taphonomic analysis and description of their perforation and use-wear stigmata. Our objectives include explaining the reasons for the presence of these shells in the occupation layers of Grotte Rochefort, discovering if these shells were perforated in the cave itself, finding if they were worn, defining acquisition methods of the raw material, and situating these shell ornaments in the context of the European Mesolithic.

Their specific determination (Dentalium sp., Ampullina sp ., Euspira nitida, Littorina littorea, Littorina obtusata, Neritidae (Neritinae) ind., Nucella lapillus, Trivia sp ., Cerithium gr. vulgatum) reveals an important diversity of the potential origins of these artefacts and their acquisition method. Some were collected after being washed ashore on beaches along the Mesolithic Atlantic or Mediterranean coasts, others were found in several fossil deposits. The dichotomy with shells consumed during the Mesolithic underlines a specific mode of acquisition dedicated to this use. In fact, most of the species used as ornaments in Grotte Rochefort were not consumed during the Mesolithic (e.g. Dentalium sp., Littorina obtusata or Trivia sp.), and several indications plead in favour of their collection once dead and already deprived of the animal’s flesh (breaks in prominent parts of the shell such as the siphonal canal or the apex, followed by their regular and uniform polishing, the regular polish of the whole surface, erasing in a uniform manner the ornamentation of the shell). This dichotomy underlines the importance attributed to them by these Mesolithic populations within the symbolic sphere. The fact that these ornaments were merely perforated is probably not innocuous; thus, they convey the mystery of the oceans or at least the image of a major environment in the lifestyle of these hunter-gatherers totally dependent on wild resources. It was not possible to confirm shaping of these ornaments within Grotte Rochefort. These shells seem to have been extensively worn before being lost on the Mesolithic occupation floor of the cave, as revealed by the break of the shell bridge between the aperture of the shell and the original perforation. The ornaments found in Grotte Rochefort doubtless represent only a tiny proportion of the shell ornaments worn by the people who frequented this cave. Only the smallest components of the Mesolithic shell ornaments seem to appear there. Analysis of the shell ornaments from Grotte Rochefort shows strong similarities with those of the European Mesolithic: the shell ornaments discovered are small in size, the main species determined are Dentalium sp ., Littorina obtusa, their shaping (reducing of tusk-shells, perforation localized on the last whorl of gastropods, perforation moved from the outer lip to offer resistance to the tensions linked to the tie, two perforations of the dorsal area of Trivia), and the way they were worn (deformation of the shape of the perforation associated with the polish around the perforation, levelling of the inner lip).

Grotte Rochefort thus represents an additional milestone bearing witness to important exchange networks with multiple influences during the Mesolithic.

Keywords: ornaments, shells, Mesolithic, fossil, acquisition territories, Grotte Rochefort.

Les études anglo-saxonnes sur l’économie du feasting nous ont rendu attentifs à l’existence, dans des sociétés non-étatiques à organisation hiérarchisée ou stratifiée, de formes d’élevage dans lesquelles l’utilisation rituelle des principaux animaux domestiques joue un rôle central. Les travaux les plus marquants ont concerné essentiellement des sociétés sud-est asiatiques (Hayden, 2001, 2003 et 2009 ; Adams, 2016).

Le cochon et le buffle y sont les principaux animaux domestiques. Ils sont considérés comme des bien rituels et ne sont abattus et consommés, sauf exception notable, qu’en contexte festif.

Les choix en matière de gestion des cheptels sont conditionnés par la nécessité de satisfaire aux besoins du rituel. L’abattage est saisonnier et la consommation sur l’année est donc irrégulièrement répartie. Les animaux abattus lors des fêtes sont fournis pour partie par l’organisateur et pour partie par ses parents biologiques et ses affins. Si l’essentiel de la viande est consommé sur place, une partie significative fait cependant l’objet d’un partage et d’une consommation « décentralisée ». La circulation centripète des bêtes sur pied et la circulation centrifuge des parts de viande se fait selon les lignes de la parenté. Le rituel primant sur toute autre considération, certains aspects considérés ailleurs comme cruciaux ne jouent ici qu’un rôle secondaire. C’est le cas, par exemple, des contraintes liées à la reproduction des cheptels.

Si les travaux ethnoarchéologiques cités nous livrent une bonne caractérisation générale de l’économie du feasting, ils n’évoquent que de manière indirecte les conséquences archéologiques de cette manière de concevoir le rôle des animaux domestiques. C’est cette lacune qui a motivé l’organisation de notre projet, dont l’un des objectifs est de mener une analyse détaillée de la circulation des bêtes sur pied et du circuit de la viande partagée dans le cadre de l’économie du feasting. Cet article présente une analyse détaillée d’une fête en particulier qui s’est déroulée en juin 2016 dans le village de Tarung (île de Sumba, Indonésie). Il débouche sur une amorce de réflexion épistémologique sur les conditions de production des interprétations dans le domaine de l’archéozoologie des sociétés de la Préhistoire récente.

Mots clés : archéozoologie, ethnoarchéologie, Sumba, feasting, sacrifice.

Since the 1990s English-speaking studies on the feasting economy have shed light on the existence, in non-state hierarchically organized or stratified societies, of forms of breeding in which the ritual use of the main domestic animals plays a major role. The most landmark studies have dealt mainly with societies from south-eastern Asia (Hayden, 2001, 2003 and 2009; Adams, 2016), with a specific focus on the populations from Toraja land (island of Sulawesi, Indonesia) and the island of Sumba (Indonesia). The fact that these areas have long remained isolated is a positive aspect, since the effective takeover by the Dutch colonizers and the arrival of the first religious missions occurred only at the beginning of the twentieth century. Despite the Christianization of most of the local population, traditional practices and in particular feasts with mass slaughters of buffaloes and pigs have persisted, and this is what motivated our decision to initiate an ethno-archaeological research project dedicated to the conditions of formation of bone assemblages within the frame of a feasting economy.

Pig and buffalo are the main domestic cattle involved. They are considered as ritual goods and, with notable exceptions, they are slaughtered and consumed only in feasting contexts. During the most significant ceremonies, namely those held on the occasion of funerals or for the reconstruction of the ancestor house, mass slaughters often involve several dozen animals. These ceremonies are always two-dimensional, both religious (sacrifice) and social (the slaughtered animals representing ostentatious signs of wealth and prestige). The amount, in terms of quantity, of goods consumed depends on social status, wealth and on how much the organizer is involved in social competition.

The choices presiding over cattle management are conditioned by the necessity of meeting the needs for the ritual. The main characteristics of the breeding system are the following: 1) Seasonal slaughter, since the major feasts take place mostly during the dry season, after the rice harvest. Consumption is thus irregularly distributed over the year. 2) Slaughtered cattle are provided partly by the organizer of the feast and partly by his biological relatives and community. As a result, the diversity of sources is great since external cattle may come from several settlements and, often, a large number of households. 3) While most of the meat is consumed in situ, in particular at the banquet held during the feast, a significant part of it is nevertheless shared, this “decentralized” consumption often concerning several hundred households. 4) The major feasts take place only in the main villages, in which there are one or several ancestor houses. Since cattle, buffaloes in particular, are usually bred in secondary villages, the breeding locations and the places where most of the consumption occurs are significantly dissociated. 5) The centripetal movement of live cattle and the centrifugal movement of the pieces of meat follow the lines of kinship. 6) Since the ritual stands above all other consideration, some aspects that may elsewhere be considered as crucial here play a secondary role. This is for instance the case for the constraints linked to livestock reproduction.

The ethno-archaeological studies mentioned above provide a good overall characterization of a feasting economy, but the archaeological consequences inferred by this conception of the role of domestic cattle are only indirectly addressed. This lack of data drove us to set up this project, which took the form of two missions in the field carried out in 2015 and 2016. One of the goals was to conduct a detailed analysis of the circulation of live cattle and of the meat portions shared within a feasting economy framework. The economic and social contexts of the areas under study as well as the theoretical model for the circulation of cattle and shared meat portions have been presented in a previous article (Jeunesse, 2016a). This second study presents the detailed analysis of one particular feast. It is based on an inventory and a systematic pinpointing of all the donors of live cattle and all the recipients of meat portions. The first part describes a traditional feast held in June 2016 in the village of Tarung (island of Sumba, Indonesia). Although it was a rather modest ceremony, 21 animals coming from about ten different settlements were brought in and more than 600 portions of meat were distributed and consumed in at least 30 different settlements distributed over an area of a hundred square kilometres. The bones of one single animal are thus likely to end up scattered among the organizer’s house, the other houses of the village in which he resides, and a various number of other houses located in several distinct settlements. The bones will thus be distributed among the main rubbish structure, the one directly linked to the banquet and the ceremony, and a large number of secondary rubbish structures.

The second part puts in perspective the main results obtained. The pattern elaborated by this study can be considered valid for at least two areas of the Indonesian Archipelago (Sumba and Toraja land), and a first review of the literature available allows us to consider that its main aspects could certainly be easily extended to all the “hill tribes” of south-eastern Asia. A concentric approach leads, at the end of our reflection, to the consideration of the possibility of extending this pattern to other present or recent contexts. It is obviously too early to seriously consider transposing it to recent European Prehistory. We hope nevertheless that our research on presumed exotic forms of relationships between man and animal are already likely to nourish epistemological reflection on the way interpretations are produced in archaeo-zoology.

Keywords: archaeo-zoology, ethno-archaeology, Sumba, feasting, sacrifice

Une fouille réalisée en 2013 à Cazan « le Clos du Moulin » Vernègues (Bouches-du Rhône) a permis de mettre en évidence une importante occupation du Chasséen récent en fond de vallon. Aucun niveau de sol n’a été retrouvé, mais plus de 600 structures en creux permettent de documenter un des premiers sites d’habitat structuré de cette période dans le Sud de la France. Les attributions chronologiques reposent sur huit datations au 14C comprises en 4100 et 3800 avant notre ère et sur un abondant mobilier archéologique qui concorde avec l’attribution à une étape récente du Chasséen. Le gisement livre plusieurs fosses, des puits et une concentration de structures à pierres chauffées, mais aussi et surtout de nombreux trous de poteau et fosses d’implantation qui permettent de restituer les plans d’environ quinze bâtiments. Ces derniers correspondent notamment à dix édifices standards ; trois ensembles architecturaux plus imposants s’insèrent au sein de cette organisation et traduisent sans doute la présence d’édifices socialement plus valorisés.

L’analyse du mobilier archéologique ne permet pas de déterminer assurément de chronologies relatives au sein de ces structures et aménagements, mais elle met en évidence plusieurs zones d’activités privilégiées en particulier pour le travail du silex et de l’obsidienne. Ces activités signent une occupation domestique a priori pérenne au sein d’un paysage largement anthropisé et maitrisé d’après les observations anthracologiques et malacologiques.

L’ensemble de ces résultats tend à prouver l’attachement des populations chasséennes à ce territoire où elles se sont installées et ont vécues. Ce site permet donc enfin d’approcher pleinement la notion de village pour le Chasséen récent.

Mots clés : Néolithique moyen, Chasséen, habitat, maison, bâtiment, puits, foyer, céramique, silex, obsidienne, tracéologie, malacologie, anthracologie, matériel de mouture, faune.

In 2013 excavations near Vernègues (Bouches-du Rhône), “Cazan-Le Clos du Moulin”, revealed an important occupation site dating from the Middle Neolithic (late Chassey culture). The village of Vernègues is located just south of the Massif du Luberon, approximately 60 km north-west of Marseille and the Mediterranean.  The area excavated, a little over one hectare, is located in the centre of a small valley which joins the valley of the Durance just to the north. This valley is remarkable for the density of Neolithic remains and “Cazan-Le Clos du Moulin” is situated in the centre of a cluster of sites. The different sites form a discontinuous network across the floor of the valley, which is undoubtedly linked to the network of watercourses which drain the valley.

While no occupation ground was uncovered during the excavation, over 600 archaeological features were preserved. These features record one of the first structured dwelling sites for this period in the south of France.  The occupation of this site has been dated by eight radiocarbon dates (4100 - 3800 BCE) and a large assemblage of archaeological finds, which confirm the attribution to the late Chassey culture.

The site contained many pits, five wells and a concentration of twenty burnt stone combustion features, which are typical of this period. Through analysis of these structures we observe how two pairs of wells seem to have functioned together, particularly in their secondary use as middens. The burnt stone combustion features are all circular and organised into four sub-parallel lines.  These features group two different types of hearths, which may reflect a chronological difference or differences in use.

However, the most notable aspect of this site resides in the discovery of many postholes and foundation pits, which trace the outline of at least fifteen different architectural structures. These are the earliest buildings recorded for the late Chassey in the south of France.

They notably consist of ten houses which were implanted on well-preserved postholes. These structures repeat the same overall plan, based around a dozen posts organised into two aisles, with a probable ridge beam extending beyond the ridge purlin. Two smaller posts in front of the southern gable end probably indicate the location of an entrance. The surface area of these buildings varies between 80 and 180 m². There are few comparable examples for these structures for this period and they are markedly distinct from other examples in France or in northern Italy, which are generally rectangular in plan and frequently have foundation trenches.

In addition, three other architectural structures have been identified, which are defined by a series of very large foundation pits. The pits are rectangular in plan with asymmetric profiles (at times triangular, occasionally trapezoidal): they can be up to 2 m deep with an over-cut at the base. We find even fewer archaeological parallels for these buildings. However their construction clearly demanded a significant investment of time and effort, which could indicate the coordination of a group or community of individuals.

All of these buildings follow a plan which was clearly collectively predefined; all the buildings are orientated NW-SE, which coincides with the prevailing wind in the valley.  Moreover, the distribution of these postholes and pits indicates that several houses were completely rebuilt in the same place, up to two to three times. Each new phase of construction was preceded by the partial deconstruction of the remains of the previous phase. We note that some of the posts had clearly begun to decompose in situ, prior to being burnt and/or removed. Finally, some of the pits/postholes from the previous phases were backfilled with stones and broken fragments of querns.

Analysis of the archaeological finds has not allowed us to establish a relative chronology between these different buildings or the diverse features. However, it has enabled us to define several zones of activity, specifically related to the working of flint and obsidian. The obsidian comes from Sardinia and we have observed elements associated with shaping and working nuclei as well as debitage from a knapping area which were found in a well.

Obsidian is not the only example of imported material found; it is accompanied by several ceramic and earthenware objects (including a loom weight) and a copper awl, all of which probably originate from northern Italy. Furthermore, analysis of the querns recovered from the postholes and analysis of the marks on the worked flints indicate agricultural activities related to grain farming. The faunal remains were almost exclusively recovered from the deposits infilling the wells. These remains record the importance of caprines in the livestock, indicating pastoral activities were also important at Cazan, as at many other sites of the same period.

These different activities imply a domestic occupation, seemingly permanent, in a predominantly anthropogenically modified landscape (based on malacological analysis). The preliminary analysis of the snail assemblage indicates an open environment during the late Chassey culture, and this observation seems to be confirmed by the charcoal analysis. Furthermore, the study of the charcoal from the postholes and hearths allows us to distinguish the selection of different species of wood: oak is used for the posts while strawberry tree (arbutus unedo) is used as fuel.

These results suggest the attachment of the Chassey population to this place and the surrounding landscape, which they occupied and reoccupied several times over 300 years. While an important part of their activity was based around a mobile lifestyle and pastoral management, this site allows us to consider the concept of a Neolithic village. These first results presented here renew the debate and research concerning the character of the occupation of this landscape during this period.

Keywords: Middle Neolithic, Chassey culture, dwelling site, house, building, well, hearth, ceramic,

La colline Saint-Charles à Marseille a fait l’objet d’opérations d’archéologie préventive de 2000 à 2008 au préalable à une réhabilitation immobilière. Les sondages préliminaires ont montré qu’elle était partiellement occupée durant le Mésolithique I et le Néolithique ancien (VIIIe et VIe millénaires av. J.-C) et probablement entièrement durant le Néolithique moyen et récent (IVe millénaire av. J.-C.). Toutefois, seuls quatre « sites préhistoriques » définis artificiellement par l’emprise des opérations immobilières et pour lesquels les niveaux stratigraphiques étaient bien conservés ont fait l’objet de fouilles. Des structures d’habitat de différentes natures (trous de poteaux, fosses, sols de circulation, plan d’habitat, murs en terre) et du matériel archéologique (lithique, céramique) sur l’ensemble des sites témoignent des occupations successives de cette portion de colline.

L’une des caractéristiques majeures de ces habitats et l’objet de cet article est la présence de malacofaune marine dans tous les niveaux archéologiques de la colline à l’exclusion de tous autres vestiges de faune terrestre sauvage ou domestique et de restes végétaux sauvages ou cultivés. L’étude permet d’avancer quelques hypothèses sur les raisons de cette présence exclusive.

Elle met en évidence une évolution sur le long terme des espèces en présence. Aux patelles (Patella sp.) et bigorneaux (Phorcus turbinatus) consommés au Mésolithique I, succèdent les coques (Cerastoderma sp.) et les bigorneaux (Phorcus turbinatus) au Néolithique ancien, puis les cérithes (Cerithium vulgatum) et les bigorneaux (Phorcus turbinatus) au Néolithique moyen et enfin les murex (Hexaplex trunculus) et les tritons (Charonia lampas et variegata) au Néolithique moyen-récent. Le choix des espèces et du biotope apparaît orienté pour chaque période et permet de supposer des options préférentielles de cueillette en relation avec une évolution du goût et des techniques de collecte tout au long de l’occupation de la colline.

Au Mésolithique et au Néolithique, la faible proportion de coquilles et leur dispersion au sol sur le site de Bernard-du-Bois ne conduisent pas à privilégier l’hypothèse d’un amas coquillier, mais plutôt celui d’un dépôt coquillier. La cueillette de la malacofaune marine du Mésolithique I et du Néolithique ancien semble intégrée de façon opportuniste aux stratégies d’exploitation du territoire et pourrait être rapportée sur la colline à la faveur de déplacements saisonniers en complément d’un autre type de diète alimentaire. Au Néolithique moyen et récent, il s’agit toujours d’un dépôt coquillier. Toutefois, les stratégies de collecte s’orientent vers des coquillages à plus fort rendement alimentaire dont la consommation pourrait compenser en partie une diète carnée d’origine terrestre.

Par ailleurs, à partir du Néolithique moyen, sous les effets de la remontée de la mer, la colline passe d’un statut de site « d’arrière-pays » à celui de site côtier. La fréquentation de la colline devient plus pérenne. Durant le Mésolithique I et le Néolithique ancien, la présence de malacofaune marine sur les sites du littoral et sur la colline laisse supposer des liens de complémentarité, internes au bassin de Marseille. Au Néolithique moyen, la colline est intégrée aux réseaux de circulation de matières premières lithiques qui prennent place en dehors du bassin de Marseille (silex bédoulien, roches vertes), mais la nature des contreparties échangées demeure inconnue. Les stratégies économiques mises en place par les groupes humains du Mésolithique et du Néolithique méridional pour la gestion de l’environnement côtier en relation avec les sites de l’arrière-pays au moment du passage à l’Holocène apparaissent variées.

Mots-clefs : France, Marseille, Néolithique, Mésolithique, coquillages, consommation, exploitation du littoral.

Saint-Charles hill in Marseille was the subject of preventive archaeology operations from 2000 to 2008 prior to a real estate rehabilitation project. The preliminary evaluations showed that it was partially occupied during the Mesolithic I and the Early Neolithic (8th-6th millennia BC) and probably completely during the Middle and Recent Neolithic (4th millennium BC). However, only four ‘prehistoric sites’, defined artificially by the different areas covered by the rehabilitation project and for which the stratigraphic levels were well preserved, were excavated. At the top of the hill, on the Bernard-du-Bois site, the stratigraphy yielded coherent archaeological assemblages and the levels, about one metre high, covered periods from the Mesolithic I to the Early, Middle and Recent Neolithic with a hiatus for the Mesolithic II. Lower down, the sites of Nédelec, Voie Nouvelle and Rue Longue-des-Capucins corresponded to an extension of the Bernard-du-Bois site during the Middle Neolithic. At Nédelec too, the stratigraphy was over a metre thick and delivered, in the best preserved plots of the site, nine accumulated levels of archaeological structures (postholes, pits) dating from the Middle and Recent Neolithic. Structures of various types (postholes, pits, open-air dwelling floors, house plan, earth walls, probably a kiln) and material (lithic artefacts, ceramics) on all the sites showed successive occupations of this portion of the hill.

One of major characteristics of these occupations and the subject of this article is the presence of marine malacofauna in all the archaeological levels on the hill, without any other remains of wild or domestic terrestrial fauna or wild and cultivated plants. Their number is however relatively small, just under 4 000 individuals for all four sites for the whole period. The study allows hypotheses to be proposed regarding the reasons for this exclusive presence.

The study highlights an evolution over the long term of the different species present. After limpets (Patella sp.) and winkles (Phorcus turbinatus) consumed during the Mesolithic I, cockles (Cerastoderma sp.), clams (Ruditapes decussatus) and winkles (Phorcus turbinatus) follow in the Early Neolithic, then cerithes (Cerithium vulgatum) and winkles (Phorcus turbinatus) in the Middle Neolithic and finally murex (Hexaplex trunculus) and triton or trumpet shell (Charonia lampas and C. variegata) in the recent Neolithic. This succession is particularly represented on the Bernard-du-Bois site where all these periods are stratigraphically accumulated. The various choices of gathering allow us to estimate the way the occupants of the hill exploited the surrounding environments, particularly in the Neolithic where several biotopes were accessible (rocky shore, sandy beach, brackish water lagoon, posidoniae).The choice of the species and habitat seems to have been deliberate for every period and allows us to presume preferential options in connection with an evolution of taste and gathering-collecting techniques throughout the period the hill was occupied.

In the Mesolithic and Early Neolithic, the small proportion of shells and their dispersal on the archaeological ground levels on the Bernard-du-Bois site lead us to favour the hypothesis of a deposit rather than that of a midden. Gathering or collecting of the main marine mollusc species of these periods seems to have been integrated opportunistically into the territorial exploitation strategy and they could have been brought back onto the hill during seasonal movements as a supplement to some other type of diet. In the Middle and Recent Neolithic, there were still shell deposits. However, gathering strategies turned towards shellfish with higher energy efficiency, the consumption of which could partially compensate a meat-based diet of terrestrial origin.

In addition, during the Middle Neolithic, due to the effects of higher sea levels, the hill moved from the status of a hinterland site to that of a coastal site. The hill was frequented by more people, over longer periods. This raises questions about the change of status of these various occupations and about their place in the economy of coastal resources in chrono-cultural contexts of the Mesolithic and Neolithic in the northern Mediterranean. During the Mesolithic and Early Neolithic, the presence of marine malacofauna on coastal sites and on the hill suggests complementarity, within the Marseille basin. In the Middle Neolithic, the hill was integrated into the raw materials circulation networks (Bedoulian flint, greenstone) of southern France but the nature of the exchanges with the inland sites remains unknown. This study thus contributes information on the apparently varied economic strategies organized by human groups of the southern Mesolithic and Neolithic for the management of the coastal environment in relationship with inland sites at the time of the passage to the Holocene.

In general, the composition of the Early Neolithic diet was still based on the exploitation of wild fauna and plants in addition to agro-pastoral production (Courtin, 2003; Binder and Sénépart, 2004; Sénépart, 2007). During the Middle and Recent Neolithic this tendency seems to have decreased considerably, and even to have disappeared. The situation of the sites on Saint-Charles hill, where the diet was still based on this type of foodstuffs, is thus paradoxical.

Keywords: France, Marseille, Neolithic, Mesolithic, seashells, food, consumption, exploitation of coastal environment.

Si les procédés de réparation, réutilisation et recyclage ont fait l’objet de recherches spécifiques en ce qui concerne les productions lithiques ou osseuses préhistoriques, et les métaux pour les périodes plus récentes, les poteries n’ont encore que rarement été considérées de ce point de vue, en particulier pour les époques protohistoriques. Sensibles aux chocs mécaniques et thermiques, et donc renouvelés fréquemment, les ensembles céramiques représentent un matériau de choix pour appréhender ces questions porteuses d’informations sur les systèmes techniques mais aussi sur le fonctionnement et les conceptions socio-économiques, voire symboliques, des sociétés considérées.

Centré sur la Corse protohistorique, cet article fait le point sur les techniques de réparation et de remploi et les modalités de recyclage, leur fréquence et leur distribution. Pour ce faire, nous avons combiné les observations à différentes échelles sur les récipients en céramique afin de tracer des indices de réparation (anciennes fissures ou fractures colmatées, trous de part et d’autre d’une fissure, etc.) ou de remploi (traces d’usure) et l’analyse chimique des matières adhésives conservées sur certains tessons afin d’en déterminer la nature. La question du recyclage a, quant à elle, été surtout abordée par l’étude des fragments de céramiques intégrés dans des structures, qu’il s’agisse de sols de circulation, de radiers de fours, de foyers ou de comblements. Les résultats présentés dans cet article résultent de l’étude d’un important corpus de vestiges céramiques : plus de 100 000 tessons issus de près de 40 000 vases (estimation NMI).

Deux procédés principaux de réparation ont été mis en évidence : la méthode traditionnelle de la suture et le collage. Pour ce dernier, c’est du brai de bouleau qui est majoritairement utilisé. Au-delà de l’importance de ce matériau pour prolonger la vie des céramiques, la découverte de brai de bouleau est une donnée nouvelle dans ce contexte chrono-culturel. Il s’agit en effet du témoignage le plus méridional de son emploi, montrant la perduration d’une substance qui a été fabriquée sur le continent dès le Paléolithique moyen, qui a été largement utilisée au Néolithique, mais dont l’exploitation n’était pas documentée jusqu’à présent pendant la Protohistoire au sud de la Loire. Par ailleurs, l’identification, dans certains échantillons, de cire d’abeille et de résine de pin, conjointement au brai de bouleau, témoigne de l’exploitation d’une large gamme de ressources naturelles.

Les modalités de recyclage, fortement diversifiées (pavage de sols, éléments de foyers, supports de découpe, éléments d’objets et ustensiles variés, etc.), témoignent d’une gestion raisonnée du matériau céramique répondant à des besoins spécifiques mais révélant également de choix dépassant les seules contraintes matérielles et mécaniques.

Au-delà de ces données sur les systèmes de réparation des poteries et de recyclage des tessons, ce sont les mécanismes d’acquisition des matières premières impliquées dans les réparations ainsi que les choix techniques et socio-économiques réalisés et l’évolution des pratiques qui sont discutés et intégrés au sein d’une réflexion plus globale sur les sociétés protohistoriques de Corse. Les descriptions proposées ici montrent que les savoir-faire développés par les insulaires sont variés et parfois indépendants des disponibilités du milieu. Elles illustrent également des mécanismes de discrimination permettant d’offrir une dimension sociale, générale ou particulière, aux phénomènes de suture ou recollage et de remploi.

Mots-clés : vaisselles céramiques, Protohistoire, Corse, Sardaigne, réparation, recyclage, brai de bouleau, cire d’abeille, résine de pin.

While the recycling, re-use and maintenance of tools and objects have been the subject of thorough investigation regarding prehistoric lithic or bone products, and metals for the most recent periods, pottery has still rarely been considered from this standpoint, especially for the protohistoric periods (Bronze and Iron Ages). Sensitive to mechanical and thermal shocks, ceramics are a material of choice to address questions related to technical systems but also the functioning and socio-economic patterns or symbolic representations of these societies.

Centred on protohistoric Corsica, essentially on contexts in the south of the island, this article considers re-use and maintenance techniques for ceramic materials, their frequency, their distribution and the astonishing diversity of recycling processes (floor paving, elements for hearths, cutting supports, re-use as different utensils such as spoons for instance, etc.). In order to assess the modalities of re-use and recycling of pots, we combined observations on different scales regarding ceramic vessels in order to highlight mending clues (old cracks or breaks that had been filled in or sealed, repair perforations, etc.), use-wear analysis and the chemical analysis of adhesive materials still preserved on some sherds. Recycling was investigated through the study of ceramic fragments involved in the construction of several structures such as floors, hearths, etc. This investigation is based on the study of a wide ceramic corpus of over 100,000 sherds (from nearly 40,000 vessels) from 91 Corsican protohistoric sites, but comparisons are also made with Nuraghic Sardinia.

The results obtained show that Corsican Bronze and Iron Age potters developed a large variety of solutions and techniques to meet their daily needs by extending the lifespan of pottery vessels and sherds resulting from breakage. Two models of repairs were highlighted: stitching and gluing. The former is part of a ubiquitous technical tradition with adaptations over time related to the tools used to drill the holes, since iron is more efficient for this task than bronze. The use of birch-bark tar as an adhesive for different damaged parts of a pot is more surprising for the chrono-cultural context considered here (Late Bronze Age and essentially Early Iron Age in Southern Corsica). This is indeed the most southern evidence of this substance. This adhesive, known elsewhere on the continent since the Middle Palaeolithic onwards, was widely used during the Neolithic in Europe, but there was previously no evidence of its use during the protohistoric periods south of the Loire River. Furthermore, the identification of beeswax and pine resin, together with birch-bark tar in a few samples, provides evidence regarding the exploitation of a wide variety of natural substances.

Interestingly, the materials involved in the maintenance of ceramic vessels provide evidence for various acquisition networks. Some of them were easily found in the local environment of the sites, such as lead in Sardinia, while others required external input. This is the case for the birch-bark tar that has been chemically identified on the site of Cuciurpula, that necessarily comes from the inland mountainous part of the island, where it is still present.

By comparing repair methods with the general characteristics of the pottery (shape, size, domestic versus luxury vessels, use of local or exogenous clay, etc.), it appears that ceramic vessels were not all submitted to the same treatment when they broke: fine, polished and bright ceramics made with exogenous clay (Corsican but probably produced in specialized workshops) were never repaired, probably because they would have lost their aesthetic value by such an operation. On the other hand, more common vessels, especially large locally made containers, were regularly repaired when broken. This provides evidence for a distinction of status between different classes of ceramic vessels.

These differences are also perceptible through the reintroduction of sherds from broken vessels within other functional systems in recycling operations. The diversity of recycling methods gives evidence of rational management of the ceramic material and waste to meet specific needs. This also reveals choices that do not result from material constraints.

Thus, the detailed study of repair procedures, re-use and recycling of ceramics is of prime importance to fully understand the economy of ceramics and to gain full knowledge of their lifespan. The descriptions proposed here show that the know-how developed by the islanders was varied and sometimes independent of local resources. They also illustrate mechanisms of discrimination which offer a social dimension, general or particular, to the phenomena of gluing and re-use.

Keywords: Ceramic wares, Protohistory, Corsica, Sardinia, repair, re-use, birch-bark tar, beeswax, pine resin.

A la suite des travaux de P.-R. Giot, réalisés à partir du milieu des années 1950, le tumulus sud de Barnenez, situé sur la côte nord du Finistère, est l'un des plus emblématiques parmi les mégalithes bretons. Depuis lors, les connaissances concernant ce monument de soixante-dix mètres de long n'avaient guère été renouvelées, si ce n'est par la publication de la monographie correspondant à ces travaux, à la fin des années 1980. Trente ans plus tard, ce fut notamment l'objet d'un programme de recherche international et d'une thèse à l'université Rennes 1. Il mêle l'étude de l'ornementation des espaces internes et celle de l'architecture. Parmi les résultats obtenus, la chambre H et son couloir montraient des évidences d'état antérieur à celui que l'on observe aujourd'hui. Nous avons donc décidé de réaliser un sondage pour rechercher des traces éventuelles de l'état antérieur au sol. Les résultats ont dépassé nos attentes par la bonne conservation du plan du monument précédent qui a été en grande partie volontairement démonté. Son couloir débouchait directement dans une chambre mégalithique dont certains éléments sont encore en place. L'antichambre que nous connaissons aujourd'hui pour le dolmen H n'existait pas. L'étude de la façade occidentale a également révélé une histoire architecturale complexe. Elle présente de nombreuses pierres dressées. Il s'agit d'un renfort du tumulus contre la pente sur laquelle est construite la masse tumulaire. Mais, il y a également une réelle mise en scène proche de celle d'un alignement de pierres dressées à l'air libre. De plus, des rapprochements ont été faits avec le rythme des orthostates d'autres dolmens du tumulus, nous permettant de dire que les pierres dressées de la façade occidentale pourraient également provenir du démantèlement d'un autre dolmen. Réserver un tel processus aux seuls cas particuliers du remploi de fragments de grandes « stèles » brisées, initialement dressées à l'air libre, fut probablement un contresens.

Mots-clefs : Mégalithe, dolmen à couloir, alignements de pierres dressées, architecture, Néolithique, Bretagne.

The southern tumulus of Barnenez, situated on the north coast of Finistère, is one of the most famous Breton megaliths. It was excavated during the 1950s by P.-R. Giot, and the results published in 1987. Since then, data concerning this well-preserved monument have not been updated with the new knowledge regarding megaliths gained over the last three decades. In 2010, an international research program began on this tumulus, integrating an ongoing PhD at Rennes 1 University. The first part of the study concerns the ornamentation inside the funerary areas, with the discovery of paintings and the redefining of the engravings. The second section is an architectural study through interpretation of the elevations thanks to the use of archaeological building methodology. Both revealed multiple phases inside the tumulus, unveiling a complex history. This article will focus on two structures of the tumulus which contain discreet evidence of that substantial history, before the monument attained its imposing final form, seventy metres long.

The first structure is passage grave H. Two registers of ornamentation have been found, revealing two distinct and successive phases inside the chamber. In the passage, the architectural study showed some ruptures in the constructional mode, revealing extensions of the passage. Passage grave H seems to have had a previous state, destroyed by the Neolithic builders themselves. We decided to do a survey to test if the previous monument had left marks on the ground of the current passage grave. The results exceeded our expectations, with conservation of the plan and of part of the cairn with, in several places, two courses in elevation. Not all the monument was levelled off, the orthostats of the end of the chamber remained in place, integrated into the current chamber, which explains the differences of ornamentation. The survey proves that a monument was dismantled and partially reused in the current dolmen H. It is one of the first times we can prove the reuse of stones taken from a dolmen to be integrated within another. We needed all the megalithic architectural elements, marks on the ground, on the elevations and on the ornamentation, to partially understand the architectural history of dolmen H.

The second part studied is the western façade of the tumulus. It contains a row of many raised stones, whose wider sides face the bay of Morlaix. This alignment is a kind of buttress used to block the external mass of the tumulus against the slope on which it was built. Furthermore, resemblances have been found between the stones of the façade and alignments of raised stones in the open air. The main result, however, lies in the similarity between the organisation of these stones and those inside the passage graves, with shared rhythms in the forms and the geology of the blocks as criteria. This western façade shows, as it were, a flat version or representation of the walls of an orthostatic passage grave. This alignment of raised stones is an architectural manual to explain the internal space of a dolmen and how to build it. The corresponding dolmens inside the tumulus bore marks of extensions of the passage with the addition of new slabs, matching the phases of the tumulus. The alignment includes these extensions but with differences in rhythm. It allows us to suggest that the raised stones came from the dismantling of another dolmen, with its own architectural history, different from the tumulus, but "embedded" or recorded in the western façade.

The southern tumulus of Barnenez provides evidence that the Neolithic builders did not avoid the destruction of a previous monument in order to build a new architectural project, where the stones could be reused. To restrict this process only to the large broken menhirs, initially raised in open areas, was probably a misinterpretation

Keywords:  megalith, passage graves, standing stone alignments, architecture, Neolithic, Brittany.

La Quina « aval » est un site de référence pour l’Aurignacien ancien du fait des importantes opérations de fouilles qui se sont succédées dans ce gisement depuis le début du xxe siècle. Bien que le matériel issu des fouilles Henri-Martin (père et fille) soit exceptionnel de par sa quantité et sa composition, le gisement de la Quina « aval » reste peu mobilisé dans les discussions concernant les modalités d’émergence du Paléolithique supérieur en raison des biais générés par les méthodes de fouille alors en vigueur. Les restes fauniques figurent parmi les vestiges les plus affectés par ces biais, ce qui rend ce matériel peu compatible avec les exigences d’une étude archéozoologique moderne. Les dernières opérations en date (par V. Dujardin, années 1990) ont permis de mettre au jour un grand nombre de restes fauniques très bien conservés. Ce matériel constitue ainsi une opportunité pour conduire une analyse fine des modalités d’acquisition et de traitement du gibier pratiquées à la Quina « aval » et, par extension, de mieux appréhender les modes de vie au début du Paléolithique supérieur. Grâce à l’application de nouveaux outils méthodologiques, l’étude détaillée des traces de boucherie permet de reconstituer la chaîne opératoire de traitement des carcasses et de mieux appréhender les activités pratiquées à la Quina « aval ». Dans les grandes lignes, les données issues de l’analyse des restes fauniques de la Quina « aval » font écho à ce qui est documenté pour cette période dans le nord du Bassin aquitain : le renne, chassé sur une grande partie de l’année, domine le spectre faunique. Toutefois si on se réfère au poids de viande fourni par l’ensemble des ongulés chassés, on voit que les grands ongulés tiennent une place majeure dans l’alimentation des occupants de la Quina « aval ». Les traces de boucherie sont très abondantes et illustrent les activités classiquement observées dans les sites aurignaciens : éviscération, retrait de la peau, décharnement, désarticulation, retrait des tendons et récupération de la moelle. À la Quina « aval », un grand nombre d’indices de fracturation sont présents sur les os à cavité médullaire de cheval et pourraient témoigner d’un intérêt particulier pour la moelle riche en acides linoléiques de ce taxon. La présence de brûlures distales indique que certains quartiers de viande ont été consommés grillés. Plusieurs éléments suggèrent également une récupération de la graisse contenue dans les tissus spongieux des ossements. Certaines pratiques plus originales ont aussi été observées, comme le prélèvement de la cervelle et des sabots, et potentiellement une consommation de bois et de velours. Les caractéristiques des traces laissées par le retrait de la peau, associées au type de population chassée et aux saisons de capture, convergent par ailleurs vers la conduite d’activités liées au traitement de peaux à la Quina « aval », hypothèse soutenue par les autres données de la culture matérielle. Ces nouvelles données contribuent à une meilleure compréhension des modalités d’exploitation du gibier à l’Aurignacien ancien et permettent d’intégrer pleinement le gisement de la Quina « aval » dans les discussions entourant l’émergence du Paléolithique supérieur et l’apparition de l’homme moderne en Europe.

Mots-clés : Paléolithique supérieur ancien, Aurignacien, archéozoologie, boucherie, subsistance, ongulés.

La Quina ‘aval’—located at Gardes-le-Pontaroux in the Charente region of France—is one of the key sites for the Early Aurignacian. Much of the research at the site was conducted during the 20th century, first by Dr. Léon Henri-Martin between 1906 and 1936, and subsequently by his daughter Germaine between 1953 and 1971. Large quantities of lithic tools, faunal remains, bone tools, objects of adornment, and human remains were recovered during these two sets of excavations. Despite this, La Quina ‘aval’ remains fairly marginal in discussions surrounding the emergence of the Upper Palaeolithic, mostly due to concerns about recovery distortion related to 20th-century excavation standards. The faunal remains, for example, were not fully recovered and are therefore unsuitable for modern zooarchaeological studies. V. Dujardin undertook new excavations in 1994, 1995, and 1998: the stratigraphic sections of two survey-pits from the Henri-Martin excavations were cleaned and a new trench was opened. A large quantity of faunal remains, very well preserved, was recovered during the recent interventions. This material provides the basis for examination of carcass-processing activities that took place at La Quina ‘aval’ and, more broadly, provides further insights into ways of life at the beginning of the Upper Palaeolithic. More than 5,300 remains have been analysed; in addition to recording in a standard database, each element that could be precisely positioned on a complete bone has been drawn on Illustrator templates along with any marks observed on the surfaces (carnivore damage, cut marks, impact notches, etc.), which allowed for a detailed analysis of carcass-processing.

Overall, the data obtained on the faunal remains from La Quina ‘aval’ are consistent with those documented for the Early Aurignacian elsewhere in the northern Aquitaine Basin. Reindeer (Rangifer tarandus) dominates the faunal spectrum (95.6% of the NISP), with smaller numbers of horse (Equus caballus), bovids (Bovinae), and red deer (Cervus elaphus) represented. The presence of carnivores and small game (Vulpinae, Canis lupus, Lepus sp., Mustela putorius, Bubo scandiacus) is also attested. The people who occupied La Quina ‘aval’ exploited all these species except the polecat and red deer, whose remains do not bear any butchery marks. One important observation is that the bovids and horses provided as much meat as the 19 reindeer identified, which means that large ungulates constituted a considerable component of the diet. The seasonal data indicate that hunting was practised during most of the year. It is possible that the different prey species were hunted on a seasonal basis to complement the diet in periods when the primary species were scarce, but because seasonality could not be determined for species other than reindeer, such seasonal complementarity could not be evaluated. Incomplete carcasses were brought to La Quina ‘aval’, and there is a clear preference for portions of the carcass that provided both large amounts of meat and bone marrow. The high number of cut marks (36.8%) and breakage evidence (29.4%) indicates that the butchery process was intensive. All the classic butchery steps are documented: evisceration, skinning, defleshing, dismembering, sinew removal and marrow extraction. Defleshing was especially intensive, as nearly half of the long bones bear cut marks; the meat acquired was sometimes grilled. Although articular extremities are rare in the assemblage, those that occur frequently show evidence of dismemberment. Except for some mesial phalanges, all bones with a medullar cavity show evidence of fracture for marrow extraction. Horse is the species that shows the highest frequency of impact-notches. Although particularly rich in linoleic acids, horse marrow occurs in small quantities and is difficult to extract without heating. This, along with the energy required to break the thick cortical bone, indicates that the Aurignacian people at La Quina ‘aval’ were particularly interested in horse marrow. Additional atypical activities, with regard to other Palaeolithic assemblages, have also been recorded. Cut marks observed on the internal face of a parietal bone indicate that the brain was removed with a sharp tool. Evidence of hoof-extraction has also been inferred from cut marks on some distal phalanges. Several fragments of reindeer antler bear cut marks that seem to indicate velvet removal; the ethnographic literature tells us that antlers can be consumed roasted or after fermentation, but can also used for medicinal purposes or as ornamentation. Several faunal elements can also be considered as evidence for exploitation of the grease contained in cancellous bone (crushing marks, impact notches on articular extremities, straight breakage patterns, the morphology of certain elements). However, these spongy elements might have had multiple uses, as some burnt remains are consistent with the use of these greasy parts as fuel. Numerous skinning marks have been observed, and their location indicates meticulous care in this activity. The skin has been removed in its maximum dimensions, including the skin from the skull. Evidence of delayed skinning of the lower legs has also been recorded. These data, along with the prey hunted (age and season), the presence of tools dedicated to skin-treatment, and the large quantities of ochre observed during the excavations, indicate that hide-processing was a major activity at La Quina ‘aval’. Skeletal parts were also used for making tools such as spear points, awls, smoothers, etc. The presence of waste products indicates that part of this equipment was produced on-site. As a rule, retouchers are the most frequent type of bone tool. Though bovid remains recovered at La Quina ‘aval’ are relatively few, most of them have been used as retouchers, indicating that the Aurignacians preferentially selected fragments with thick compact tissue. Aligned impact notches observed on several pieces may result from working methods aimed at controlling the morphology of the blank. A recurrence of the areas exploited as retouchers indicates that the blanks were carefully selected and not simply gathered opportunistically from food waste. Finally, the teeth and antlers of mammals were used to make ornaments and the presence of unfinished pendants indicates that at least part of the fabrication sequence was conducted at La Quina ‘aval’.

The new data obtained from La Quina ‘aval’ faunal remains contribute substantially to our understanding of game exploitation during the Early Aurignacian, and makes it possible to bring evidence from the well-known site of La Quina ‘aval’ into discussions of the earliest evidence of the Upper Palaeolithic and the arrival of modern humans in Europe.

Keywords: Early Upper Palaeolithic, Aurignacian, zooarchaeology, butchery, subsistence, ungulates.

La phase récente de l’Aurignacien a longtemps été caractérisée uniquement par quelques outils types dont les plus connus sont les burins busqués, ainsi que les grattoirs à museau et les lamelles torses Dufour sous-type Roc-de-Combe. Certains d’entre eux sont retrouvés de l’océan Atlantique jusqu’au Levant. Les grattoirs Caminade ont pourtant très rapidement été érigés comme marqueurs de l’Aurignacien récent puisqu’ils ont été retrouvés sur leur site éponyme (Caminade, Dordogne) dans un ensemble de vestiges comprenant également des burins busqués. Cependant il ne semble pas que ce type ait diffusé au-delà du Bassin aquitain. De plus, tous les grattoirs Caminade du site éponyme sont réalisés dans le même matériau, reconnu a posteriori comme étant le silex « Grain de mil » de Charente-Maritime. L’ensemble de ces observations a soulevé plusieurs questions : comment se définit le territoire associé aux grattoirs Caminade ? Pourquoi la tradition technique des lamelles torses a-t-elle diffusé à l’échelle pan-continentale tandis que celle des grattoirs Caminade semble être restreinte au Bassin aquitain ? Quels sont les mécanismes sociaux (transmission des savoirs, mobilité humaine, évolution des traditions techniques, etc.) qui ont entrainé une telle répartition géographique de ces vestiges jugés jusqu’à présents contemporains ?

Avant d’interroger l’association couramment admise des grattoirs Caminade et des burins busqués, une synthèse approfondie des grattoirs Caminade s’impose. Le type reste figé dans sa première définition basée sur les quelques outils très standardisés du site éponyme, alors que la description de grattoirs Caminade sur d’autres gisements laisse entrevoir une plus grande variabilité morphométrique. De même, bien que les nucléus à grattoirs Caminade aient été identifiés, l’ensemble de la chaine opératoire n’a pas encore été décrite : quels sont les modes de production des supports de nucléus ? Existe-t-il une production intégrée de supports de grattoirs Caminade et d’autres d’outils type grattoir ou burin ?). De même, les données pétroarchéologiques ne sont pas systématiques, voire restent complètement inédites pour certains gisements. Dans le cadre de cet article, nous présentons l’étude comparative de cinq gisements : en Dordogne, Caminade, le Flageolet I, la grotte Maldidier et la Grotte XVI et en Gironde, le Pigeonnier. Nous abordons dans un premier temps la question de la variabilité typologique, morphologique et technologique des grattoirs Caminade afin de proposer une vision plus large et générale de ces outils, permettant de discuter de leur association éventuelle avec les burins busqués. Dans un second temps, nous croisons ces données avec des données pétroarchéologiques dans le but de discuter non seulement des territoires d’approvisionnement, mais aussi des voies de circulation et des modes de transport des matériaux et ainsi de mieux comprendre le phénomène de régionalisation de la répartition des grattoirs Caminade telle qu’observée actuellement.

Mots-clefs : grattoirs Caminade, Aurignacien récent, Sud-Ouest de la France, typologie, morphométrie, territoire d’approvisionnement.

The Late Aurignacian has long been characterized by just a few typical tools including the busked burins and the Roc-de-Combe twisted bladelets that are known from the Atlantic Ocean to the Levant. Caminade endscrapers were rapidly established as a typical tool of the Late Aurignacian as they were found at Caminade (Dordogne) within an assemblage comprising some busked burins. It seems however that this type never spread beyond the Aquitaine Basin. Moreover, all the Caminade endscrapers from the eponymous site were made from the same raw material, later recognized as ‘grain de mil’ flint from Charente-Maritime. All of these observations lead to a series of questions: How is the territory associated with Caminade endscrapers defined? Why did the twisted bladelet tradition spread on a continental scale while Caminade endscrapers were apparently limited to South-Western France? What were the social mechanisms (transfer of knowledge, human mobility, development of technical traditions, etc.) that led to such a geographic distribution of these lithic remains, considered as contemporaneous?

In order to question the currently admitted association of Caminade endscrapers with busked burins, an in-depth synthesis of Caminade endscrapers seems necessary. The original definition of this type, still in use, is based on only a couple of very standardized tools from the eponymous site, while descriptions of Caminade endscrapers from other sites suggest that this type may present wider morphometric variability. Furthermore, even though Caminade endscraper cores have been identified and described, the whole chaine opératoire (production mode of the core blanks? presence of some blade tools?) remains to be studied. Finally, petro-archaeological data are most often lacking and there are no studies of the supply territory or raw material economy on a regional scale. In this article we present an inter-site study based on five archaeological sites from the north of the Aquitaine Basin: Caminade, Le Flageolet I, Grotte XVI and Grotte Maldidier (Dordogne), and Le Pigeonnier (Gironde). We first address the typological, morphometric and technological variability of Caminade endscrapers.  Secondly, we cross this information with petro-archaeological data. Finally, we discuss the association of Caminade endscrapers with busked burins in the light of these new data.

Our study concerns 381 Caminade endscrapers. The length ranges between 6.8 and 44 mm. The length/width ratio shows a very high standardization of blank proportions. There is a potential double use of Caminade endscrapers according to their calibre: the smallest ones may be included in the hunting sphere (projectile components) and the largest may belong to the domestic sphere as part of other composite tools. Observation of the retouch and edge morphology shows that the most important required criteria are 1) a straight left edge, 2) a distal truncation inclined to the right edge, and 3) the rounded shape of the proximal part. The oblique truncation means that the longest edge is situated on the left, which may represent the active edge. We also propose that the distal truncation and the roundness of the proximal part may be related to their insertion into the shaft.

The inter-site petro-archaeological study highlights two supply areas. For the Dordogne sites, the supply area focuses on the Dordogne river basin with imports of Bergeracois and Senonian flint. At Le Pigeonnier, the supply area is widely open towards the north-west and the Seudre and Charente river basins. There is no particular “Grain de mil” flint management associated with  Caminade endscraper production. It is however the only material to spread across both supply areas, making it the best marker of this double territory. On the Dordogne sites, only local materials or those from nearby are reduced on site; distant materials are imported as tools and represented by the largest ones. At Le Pigeonnier however all materials (even the most distant ones) are managed similarly to the local and nearby ones in Dordogne. This is a very specific characteristic of this site. Its geographical location, between the Seudre and Charente river basins to the north and the Dordogne river basin to the east, could explain this observation: Le Pigeonnier could represent a border site at the limit of the Caminade endscraper’s geographic range, leading to specific supply strategies.

The ‘chaine opératoire’ including the Caminade endscrapers seems to be complex and presents several production goals such as ‘Caminade knives’, projectile components or even scrapers. In this case, both ‘chaines opératoires’ for obtaining busked burins and Caminade endscrapers do not appear to be complementary but, rather, share very similar production goals. The association of these two ‘chaines’ could be of economic or functional interest. We may therefore question the validity of this currently admitted association. They could represent two technical traditions, one succeeding the other over time, but considered until today as coherent lithic sets because of excavation methods, taphonomic processes, or even the small proportion of Caminade endscrapers and our lack of knowledge about their global ‘chaine opératoire’.  There could also be a local tradition (with Caminade endscrapers) within a wider continental tradition (with busked burins). Based on current knowledge, both of these hypotheses are credible, even though their consequences are very different: the first one confirms a unilinear chronological model, while the second one points to a dendritic evolution of technical traditions within a large social network.

Keywords: Caminade endscrapers, Late Aurignacian, South-Western France, typology, morphometrics, supply territory.

Sylvain Ducasse, Caroline Renard, Jean-Marc Pétillon, Sandrine Costamagno, Pascal Foucher, Cristina San Juan-Foucher et Solène Caux — Les Pyrénées au cours du Dernier Maximum Glaciaire. Un no man’s land badegoulien ? Nouvelles données sur l’occupation du piémont pyrénéen à partir du réexamen des industries solutréennes de l’abri des Harpons (Lespugue, Haute-Garonne)

En France, selon certains scénarios, la zone pyrénéenne apparaît au début du Dernier Maximum Glaciaire (DMG : 23-19 ka cal. BP) comme un espace singulier dont l’histoire culturelle diverge sensiblement des cadres définis plus au nord. Extension solutréenne du domaine vasco-cantabrique avec lequel elles partagent une même identité typo-technologique, les Pyrénées montreraient une perduration de cette tradition culturelle jusqu’au cœur du DMG (circa 21 ka cal. BP), ceci parallèlement au développement du Badegoulien dans les territoires voisins. La rareté, voire l’absence d’occupations badegouliennes sur l’ensemble de la chaîne pyrénéenne traduirait donc, pour certains préhistoriens, l’existence d’une géographie culturelle perceptible à l’échelle du Sud-Ouest français. Pourtant, alors que les données radiométriques sur lesquelles repose une partie de cet argumentaire sont aujourd’hui largement discutées, la grande ancienneté des fouilles menées sur les séquences pyrénéennes contemporaines du DMG ainsi que l’impact des processus géomorphologiques sur la conservation de ces séquences suggèrent l’existence de biais importants au sein de la documentation disponible. Tenant compte de ces biais et guidé par certains indices jusqu’alors peu exploités, plusieurs collections clés – à l’image de la couche D des Harpons (Lespugue, Haute-Garonne) – ont fait l’objet d’une révision collective et interdisciplinaire dans le cadre du projet « SaM ». Fouillé par R. de Saint-Périer entre 1912 et 1930, cet abri livre l’une des séquences les plus longues des gorges de la Save, comprise, a minima, entre le Solutréen et l’Azilien. Au début des années 2000, les industries de la couche D, alors rapportées au Solutréen supérieur par le fouilleur, firent l’objet d’une première réévaluation menée par deux d’entre nous (P. F. et C. S. J.-F.). Cette étude apporta de nouveaux éléments de caractérisation du Solutréen pyrénéen, tout en démontrant l’hétérogénéité de l’assemblage. Or, tandis que cette hétérogénéité restait circonscrite au Solutréen sensu-lato, la description de plusieurs « raclettes typiques », couplée à l’existence de mesures 14C SMA peu ou prou compatibles avec celles du Badegoulien nord-aquitain, ont motivé la mise en place d’un projet de réexamen des collections de la couche D des Harpons. Celui-ci avait pour objectifs de 1) confirmer la présence d’éléments badegouliens au sein de la série, 2) conduire une approche technologique de la composante solutréenne afin de compléter les données déjà acquises, 3) rechercher d’éventuels indices de chacun de ces technocomplexes au sein de l’équipement osseux et, sur cette base, 4) offrir un cadre radiométrique « raisonné » à partir de la datation directe de déchets techniques caractéristiques. Ce travail nous a permis de confirmer l’hétérogénéité de l’assemblage et d’en élargir l’étendue : aux vestiges solutréens se mêlent des éléments magdaléniens, gravettiens et, de manière aujourd’hui indiscutable, badegouliens. Si l’état des séries, triées à la fouille par R. de Saint-Périer, reste un frein à la caractérisation précise de cette composante post-solutréenne, nos travaux permettent de contredire l’idée que les Pyrénées constituaient un no man’s land badegoulien, du moins pour ses phases récentes. Soutenu par de nouvelles datations mises en lien avec les données les plus fiables disponibles pour le Sud-Ouest français, un cadre archéostratigraphique alternatif est proposé, vieillissant le terme des occupations solutréennes régionales. Ces premiers résultats, appelant des prolongements et compléments évidents, nous conduisent in fine à réinterroger, en France, la nature et le rythme de la transition solutréo-badegoulienne hors de la zone de définition classique de ces deux traditions culturelles.

Mots-clés : Dernier Maximum Glaciaire, Pré-Pyrénées, Solutréen, Badegoulien, industries lithiques et osseuses, datations radiométriques, géographique culturelle.

According to some, the first part of the Last Glacial Maximum (LGM: 23-19 ky cal. BP) in the Pyrenees appears to be a singular moment when the cultural trajectory of hunter-gatherer groups diverged from the frameworks defined further north. Similar to the Upper Solutrean of the Vasco-Cantabrian area with which it shares several typo-technological specificities (e. g. concave-base points, Cantabrian-type shouldered points), the Solutrean of the Pyrenees would thus continue into the heart of the LGM (circa 21 ky cal. BP), parallel to the development of the Badegoulian traditions in the adjacent areas. The current scarcity or absence of Badegoulian sites throughout the Pyrenees fits well with this establishing of a cultural geography on the scale of South-Western France. However, the radiometric data used to support this model are currently questioned, while the age of the excavations on the LGM Pyrenees sequences and the impact of geomorphological processes on their degree of conservation can cause a skewed vision of this phenomenon. A collective and interdisciplinary reassessment of several key sites taking into account these biases was undertaken as part of the “SaM” project to discuss the reality of this palaeogeographical hypothesis on a new and firm basis.

In the wake of the recent work undertaken on Grotte des Scilles (Lespugue, Haute-Garonne) providing the first evidence of Lower Magdalenian settlement in the Pyrenees (second part of the LGM; circa 19.5 ky cal. BP), our research focused on the gorges of the Save. This canyon, carved in the limestone of the Pre-Pyrenees, has recorded a long-term and continuous human occupation thanks to several sequences from caves and rockshelters. In this rich Palaeolithic environment, Abri des Harpons (Lespugue, Haute-Garonne), excavated by R. de Saint-Périer from 1912 to 1930, attracted our attention. Besides layers from the historical periods, it reveals one of the longest Upper Palaeolithic sequences, comprised at least between the Early Solutrean and the Azilian. In the early 2000s the Upper Solutrean industries and faunal remains from layer D were reassessed by two of us (P. F. and C. S. J.-F.) providing new data on the Pyrenean Solutrean but demonstrating at the same time the heterogeneity of the assemblage on typological and radiometric bases. However, while this heterogeneity was circumscribed on a Solutrean scale (from the early to the upper/final phases), the discovery of several “raclettes typiques” raised questions, in addition to 14C dates more or less compatible with a Badegoulian chronology. In 2012, these elements led us to set up a new reassessment project with the aim of 1) evaluating the existence of a Badegoulian component in layer D, 2) carrying out a technological analysis of the Solutrean component to complete the typological and lithological data previously acquired, 3) searching for technological evidence of these two technocomplexes in the osseous industries and 4) renewing the radiometric framework through direct dating of several characteristic antler and bone waste products.

This work allowed us to confirm and enlarge the heterogeneity of the layer D assemblage: the Solutrean remains are clearly associated with Magdalenian, Gravettian and indisputably Badegoulian elements. Although the existing conditions of the Saint-Périer collections are a brake on the precise characterization of this Badegoulian component, these results allow us to contradict the idea that the Pyrenees were a no man’s land for the Badegoulian groups. If the question of the early expressions of this cultural tradition (i. e. Early Badegoulian) cannot be addressed from the Abri des Harpons data, we are now at least able to confirm the frequentation of the Pyrenees by holders of Recent Badegoulian technology as was suggested at Enlène cave (Montesquieu-Avantès, Ariège) in the 1980s through another mixed assemblage. Based on the comparison of the new AMS 14C dates with the most reliable data available in south-western France, we finally propose an alternative archaeostratigraphic framework of the Pyrenean LGM occupations that restores the role of the Badegoulian in regional palaeohistory and demonstrates a significant aging of the end of the Upper Solutrean (circa 23 ky cal. BP). These encouraging results should be supplemented by further analysis and will ultimately allow discussion of the nature and rhythm of the French Solutreo-Badegoulian transition outside the area of the classical definition of these two cultural traditions.

Keywords: Last Glacial Maximum, Pre-Pyrenees, Solutrean, Badegoulian, lithic and osseous industries, radiocarbon dating, cultural geography.

Moulin-Neuf (Saint-Quentin-de-Baron, Gironde) est un site du Magdalénien moyen ancien daté entre 18700 et 17800 cal. BP. Découvert en 1939, le site s'organise comme la succession latérale de deux abris et d'un gisement en pied de falaise. L'intérieur de l'abri 1, le plus vaste, a d'abord été en partie fouillé par R. Cousté puis par H. Péquart dans les années 1940. Entre 1976 et 1980, M. Lenoir entrepris d'en fouiller la terrasse qui a notamment livré un riche ensemble archéologique (industries lithique et osseuse, faune, matières colorantes, art mobilier) attribué alors au Magdalénien moyen. Moulin-Neuf se situe à l'interface géographique et chronologique des « faciès » à navettes, à pointes de Lussac-Angles et à lamelles scalènes. Pourtant, il n'a livré aucun de ces marqueurs classiques.

Nous nous proposons de croiser les approches pétro-archéologiques et technoéconomiques pour déterminer le mode d'exploitation de chaque ressource afin de mieux cerner les occupations de Moulin-Neuf et de les replacer dans l'espace géoculturel du Magdalénien moyen ancien.

Cette étude a permis de mettre en évidence à Moulin-Neuf une grande diversité dans le spectre de matières premières exploitées. Au-delà d'un approvisionnement avec la zone périgourdine, le réseau de circulation des matières premières s'élargit nettement vers la Chalosse et le Sud du Bassin parisien. Les blocs de silex du Bergeracois, « grain de mil » et de Chalosse ont fait l'objet d'une exploitation sur place. Cette arrivée de blocs fut complétée par l???apport de supports lamino-lamellaires et d'outils provenant des formations du Turonien inférieur du Berry, du Grand-Pressigny et du Turonien d'Écoyeux. La zone de diffusion des silex du Turonien inférieur du Berry se trouve alors élargie au Nord du Bassin aquitain. Ces résultats inédits ne montrent pas de frontières territoriales nettes entre Poitou et Pyrénées mais plutôt des des réseaux complexes de circulation des matières premières lithiques, tout en conférant à Moulin-Neuf la place de carrefour géographique majeur de ces réseaux. Le partage d'un « litho-espace » similaire entre les différents faciès du Magdalénien moyen ancien relance la question de la pertinence de l'utilisation de ces faciès.

Mots-clés : Magdalénien moyen ancien, archéostratigraphie, technologie lithique, pétro-archéologie, économie des matières premières, silex de Chalosse, Paléolithique supérieur, Aquitaine, Sud-Ouest de la France.

Moulin-Neuf (Saint-Quentin-de-Baron, Gironde) was discovered in 1939. The site consists of a succession of two rockshelters and a site at the foot of a cliff. Our study concerns the largest shelter (Abri 1) which was excavated by R. Cousté and then by H. Péquart in the 1940s. Between 1976 and 1980, M. Lenoir excavated its terrace which delivered substantial archaeological vestiges (lithic and osseous industries, fauna, colouring substances, portable art) attributed at the time to the Middle Magdalenian. New radiocarbon dates, obtained by the MAGDATIS project, place the site between 18670 and 17815 cal. BP, during the Early Middle Magdalenian (EMM). The EMM reflects a veritable geocultural mosaic structured by three facies based on osseous and lithic industries: the 'Lussac-Angles facies', the 'shuttle facies' and the 'scalene bladelets facies'. These facies overlap chronologically and geographically, thus leading to the questioning of the validity of exclusive cultures defined from such facies. Moreover, what is the place of EMM sites that have not delivered any of the classic markers? This is the case of Moulin-Neuf (Abri 1, Lenoir excavations) which is located at the geographical and chronological interface of the different facies. Furthermore, none of the classic markers have been identified on the site.

We suggest crossing petro-archaeological and techno-economical approaches. The objective is not only to complete our vision of a typological territory but also to provide a dynamic vision of the management of supply areas for lithic raw materials. We therefore propose to determine the origin of the flints and identify the methods of exploitation (inside or outside the site) in order to better understand the occupations of Moulin-Neuf and replace them within the EMM geocultural territory. The corpus takes into account all tools, microliths, cores, lamino-lamellar blanks and knapping waste (> 2 cm) from layers 2, 2st, 2a, 2b and 2c (4,282 artefacts). The taphonomic and archaeostratigraphic review of the deposit led us to consider all the EMM layers as a palimpsest of several occupations.

The lithic technical system revolves around the laminar/lamellar dichotomy. On the one hand the production of blades and bladelets takes place according to two different 'chaînes opératoires'; they are never intercalated. On the other hand, the economic vocation of the supports is different. The polyfunctional blades were selected to be turned into tools (mainly burins and retouched blades). Bladelets were produced in a variety of ways (prismatic or pyramidal cores, flakes knapped on the edge). The dimensions of the microliths are relatively standardized and, typologically, they are represented by several morphotypes (simple backed bladelets, retouched or truncated backed bladelets).

This study highlights the wide range of raw materials found at Moulin-Neuf. We identified ten different flint types: Senonian flints (n = 1,326), Maastrichtian from the Bergerac region (n = 757), Upper Maastrichtian from the Chalosse (n = 528), grain de mil- flint (N=280), Lower Turonian from the Berry region (n = 156), Tertiary from North Aquitaine (n = 55), Upper Turonian from Grand-Pressigny (n = 13), Tertiary from the Aurillac Basin (n = 12), jasperoid Infralias (n = 7) and, finally, Écoyeux Turonian (n = 4). The proportion of indeterminate artefacts is 26.6% (n = 1,144). Beyond a supply from local alluvia and the Périgord area, the circulation networks widen clearly towards the Chalosse region and the south of the Paris Basin. The blocks of Bergerac flints, Chalosse flints and « grain de mil » were knapped on the site. The arrival of these blocks was completed by the importing of lamino-lamellar blanks and tools of Lower and Upper Turonian flint, respectively from the Berry region and Le Grand Pressigny, and also of Écoyeux Turonian flint. Blades made of good-quality materials were more frequently selected for the manufacturing of tools. These were either produced locally or brought to the site (in the form of blanks or tools). For lamellar production and the manufacture of microliths, there is greater adaptability to the local material of variable quality and a smaller selection of materials from elsewhere.

The "litho-espace" (sensu Delvigne, 2016) of Moulin-Neuf is the same as that of Roc-de-Marcamps 2 (with scalene bladelets) and is similar to that of Combe-Cullier (also with scalene bladelets). The origins of the materials are much more diversified than for the nearby sites of Roc-de-Marcamps 1 and Saint-Germain-la-Rivière. From a "litho-espace" point of view, Moulin-Neuf seems to fit well into the EMM with scalene bladelets and presents obvious relations with the south of the Paris Basin and probably the sites belonging to the Lussac-Angles facies (La Marche, Roc-aux-Sorciers, Taillis des Coteaux) and the shuttle facies (La Garenne). These unpublished results do not show a clear territorial border between Poitou and the Pyrenees but rather complex networks for the circulation of lithic raw materials. Moulin-Neuf seems to be located at a major crossroads in these networks. The distribution of a similar "litho-espace" between the various entities of the Early Middle Magdalenian leads us to question the partitioning of the EMM groups into exclusive 'facies'.

Keywords: Early Middle Magdalenian, archaeostratigraphy, lithic technology, petro-archaeology, raw material economy, Chalosse flint, Upper Paleolithic, Aquitaine, South-Western France.

Dans le cadre des diagnostics d’archéologie préventive réalisés sur le contournement ouest de Bergerac, un secteur positif a permis de documenter une occupation azilienne repérée sur 450 m2. Elle était installée sur un léger relief de la plaine alluviale de la Dordogne, aujourd’hui invisible dans le paysage. La fouille a porté sur deux locus, comportant chacun un foyer de galets chauffés. L’opération a livrée 1 485 restes lithiques taillés. La datation du foyer localisé dans le locus B place l’occupation vers 11430 -11117 av. J.-C, soit dans la seconde partie de l’interstade de l’Allerød. Si la conservation du gisement n’est pas optimale en raison de la présence d’occupations néolithiques et historiques qui ont partiellement endommagé le site, l’assemblage lithique recueilli est homogène et permet de documenter pour la première fois les comportements techniques de populations aziliennes en plein air dans le Sud-Ouest français, dans un contexte stratigraphique bien maitrisé. La découverte du gisement des Pinelles permet ainsi de combler une lacune en la matière. L’assemblage lithique constitue un jalon inédit de l’occupation azilienne de la Dordogne. En dépit d’un contexte taphonomique a priori peu favorable (destruction de plusieurs secteurs du site par des structures historiques et naturelles ; présence d’indices d’occupations néolithiques), la cohérence technique des vestiges collectés permet de les attribuer aux groupes contemporains de la fin de l’Azilien. Les groupes aziliens ont exploité deux catégories de matériaux : des silex sénoniens locaux de qualité médiocre et des silex du Bergeracois de bien meilleure qualité, disponibles localement et à environ 5 km du gisement. Les productions réalisées à partir des silex sénoniens se distinguent par le caractère opportuniste et peu hiérarchisé des chaînes opératoires. Les blocs de silex sénoniens ont été exploités à partir d'un ou deux plans de frappe avec un percuteur dur. L’exploitation des silex du Bergeracois a, pour sa part, fait l’objet de débitages plus soignés ayant pour conséquence une meilleure régularité des supports produits. Cependant, la recherche d’éclats et d’éclats laminaires reste l’objectif premier des opérations de débitage. La présence de monopointes à dos courbes, fabriquées sur des supports variés et de régularité variable, constitue le seul type d’armature identifié dans la série. En l’absence de remontage sur de longues distances, la contemporanéité de l’ensemble des vestiges du gisement ne peut être démontrée. En revanche, si l’on se fie à la stratigraphie de référence pour l’Azilien du Sud-Ouest français – le Pont d’Ambon (Bourdeilles) – et que l’on considère la présence ou l’absence de certains types d’outils – et en particulier les couteaux à dos – on peut prudemment envisager plusieurs moments d’occupations sur le site des Pinelles. Plus généralement, l’industrie lithique des Pinelles s’inscrit dans la dynamique commune documentée dans la région et au-delà. L’exploitation des excellents silex du Bergeracois constitue cependant une originalité économique loin d’être anecdotique.

Mots-clés : Paléolithique final, Azilien récent, Dordogne, technologie lithique, campement de plein-air, géomorpho-logie, structures de combustion.

As part of the archaeological surveys carried out during work on the western by-pass of Bergerac, an Azilian site covering 450 m2 was investigated. It was established on a slight rise in the alluvial plain of the Dordogne river, no longer visible in the landscape. The excavations examined two loci, each with a burnt-stone hearth structure, and yielded 1,485 lithic artefacts. The Azilian from the Allerød interstadial is known from a dozen of archaeological settlements in the Dordogne area. This is a low number compared with the numerous sites for the whole of the Late Palaeolithic. Azilian open-air sites are quite rare in this area in comparison with occupations in caves or shelters (Niederlender et al., 1956; Jude, 1960; Champagne and Espitalié, 1970; Bordes, 1979; Célérier, dir., 1993; Séronie-Vivien, dir., 1995; Detrain et al., 1996; Ballista, 2006; Fat-Cheung et al., 2014). While the site of Les Pinelles (Prigonrieux, dordogne) is not lacking problems, in particular from a taphonomical point of view, it provides an interesting contribution to document the techno-economical behaviour of Azilian societies in South-Western France. With these criteria, we will try to establish comparisons with the main Azilian site of the south-western area, Pont d’Ambon, excavated by G. Célérier (Célerier, dir., 1993). We will focus on the lithic industry discovered in locus B. This area delivered a hearth structure, dated by 14C (Fy 80), in association with lithic remains. The measurement gave a date that allowed this structure to be attributed to the Late Azilian (ERL-18 582, 11359 ± 89 BP, 11430–11 117 cal. BC). We will furthermore try to demonstrate that most of the lithic remains discovered in other parts of the site are also contemporaneous with the Azilian settlement. In spite of more recent disturbances (modern, medieval and Neolithic), the lithic industry is coherent and for the most part contemporaneous with the late Azilian. Productions made from the local raw material was particularly opportunistic and not hierarchical from a technological point of view. Senonian flint, collected in the immediate environment, was exclusively exploited with a hard hammerstone on a uni or bipolar knapping surface. The best quality Bergeracois flint shown a more sophisticated chaine opératoire and a higher regularity of the blanks produced. For both raw materials, the production of flakes and laminar flakes dominated. The Bergeracois flint was brought in as already exploited blocks, as demonstrated by the lack of cortical flakes. The hearth structure is one of the originalities of the site: such structures are rare in an Azilian context. While the absolute contemporaneity of all the lithic remains cannot be proved—due to the lack of refitting among different loci of the site—most of it can be associated with the Azilian settlement. These data prove that Les Pinelles was probably a large site during the Allerød. This is rather similar to the few known Azilian northern examples such as Le Closeau, Rueil-Malmaison, Hauts-de-Seine, France (Bodu, dir., 1998) or Niederbieber, Neuwied, Central Rhineland, Germany (Gelhausen, 2011). These two sites were also occupied on several occasions during the Azilian. With respect to the data from Pont d’Ambon, Bourdeilles, Dordogne (Célérier, dir., 1993; Fat-Cheung et al., 2014), in particular the presence or the lack of several types of tools (backed knife), we can propose that the Les Pinelles site was occupied on several occasions during the Azilian. Finally, the lithic industry from Les Pinelles belongs to the same cultural trend as the closest sites but also those outside the south-western area. The exploitation of the excellent Bergeracois flint is one of the originalities of the site. The difference with the strictly local flint (Senonian) can be explained by the very low quality of the Senonian flint. For a better understanding from an economic point of view we hope for the future discovery of a better preserved sites in this region. The example of Les Pinelles suggests the homogeneity of Azilian technical behaviour in south-western France. In other contexts, the data available at present point towards a technological diversity that the chronological difference does not totally explain (Mevel and Bodu, in press). The corpus of known sites is clearly limited and new data should renew our perception of the ‘Azilianisation’ process in this geographical area.

Keywords: Final Paleolithic, Late Azilian, Dordogne, lithic technology, open-air settlement, geomorphology, hearth structure.

En Europe occidentale, la définition du Mésolithique en tant que période s’est surtout basée sur la typologie des industries lithiques et en particulier des microlithes géométriques érigés en véritables fossile-directeurs. En comparaison, l’outillage du fonds commun, moins typé et visiblement moins représenté, a été relégué au second plan. Si certains chercheurs ont supposé une utilisation importante des supports bruts pour expliquer ce déficit, la prise en compte des supports « utilisés » ou « retouchés a posteriori » apparait très variable selon les lithiciens, ce qui rend difficiles les comparaisons entre sites. Surtout, la typologie morphologique montre ici ses limites pour donner un sens tant économique que culturel à ces outils supposés. Dans le cadre d’une thèse récemment soutenue, l’approche tracéologique de plusieurs corpus du premier Mésolithique du Nord de la France et de Belgique a été l’occasion d’aborder cette question et d’estimer la place et la fonction de l’outillage brut dans le système techno-économique des derniers chasseurs-cueilleurs. L’analyse exhaustive des collections, concernant tout autant les outils typologiques que les pièces non modifiées, a confirmé une utilisation massive de supports bruts qui constituent près de trois quarts des pièces lithiques marquées par des traces fonctionnelles. Ce résultat, numériquement plus important qu’attendu, nous amène à interroger la pertinence des approches méthodologiques les plus répandues qui apparaissent peu adaptées pour reconnaître ces outils bruts et leur donner un sens. L’intégration de la tracéologie dès les premières phases d’étude permettrait d’affiner les typologies et de mieux saisir les objectifs technologiques des productions lithiques. Cette démarche permettrait également d’enrichir les reconstitutions palethnologiques en plein développement actuellement. À cet égard, les 501 zones utilisées étudiées dans le cadre de ce travail permettent de mettre en évidence une ambiance originale propre au Mésolithique. Si l’exploitation des matières animales semble moins intense qu’au Paléolithique récent, le travail des matières végétales occupe par contre une place importante et distingue clairement les sociétés du début de l’Holocène de celles qui les précèdent. Au-delà de ce panorama général, les différents gisements étudiés livrent des données fonctionnelles contrastées, tant du point de vue du nombre d’outils reconnus que des spectres d’activités et démontrent la pertinence de la tracéologie pour aborder l’organisation économique des groupes au sein des cycles de mobilité. Ces résultats, qui reposent principalement sur les supports bruts, encouragent à développer l’approche exhaustive des corpus lithiques appliquée ici, non seulement pour le Mésolithique mais également pour le Paléolithique et le Néolithique.

Mots-clés : outillage brut, tracéologie, premier Mésolithique, Bassin parisien, Belgique.

In Western Europe, the definition of the Mesolithic as a period has mainly been based on a typological approach of the lithic industry. In this perspective, geometrical arrowheads are considered as the best ‘fossiles-directeurs’ (markers) because of their great diversity in shape and size. In comparison, common tools appear to be less typical and poorly represented in the assemblages. Therefore, their role in cultural and economic debates has been under-represented for most of the time. To explain this scarcity, some researchers have identified numerous ‘used’ or ‘retouched’ bladelets and flakes, suggesting that a large number of unretouched blanks were actually used. However, this aspect has been unevenly discussed depending on the different specialists and it is quite difficult to compare the published data, primarily because morphological typology shows its limits when it comes to interpreting these assumed tools on a cultural and economic perspective. Our doctoral thesis (Guéret, 2013) focused on this issue by questioning the status of these potential unretouched tools through a use-wear approach. Our body of research included several sites from the Early Mesolithic located in the Paris Basin and Northern Belgium: Verrebroek ‘Dok’, C17/C57/C58 (Sandy Flanders, Belgium), Doel ‘Deurganckdok’ J/L, C2 (Sandy Flanders, Belgium), Rosnay ‘Haut de Vallière’ (Marne, France) and Noyen-sur-Seine ‘le Haut des Nachères’ (Seine-et-Marne, France). An exhaustive study of these assemblages, concerning both the common tools and the hundreds of unmodified blanks, has confirmed a massive use of unretouched tools: this category represents three quarters of the used objects. This result, with a higher proportion of used tools than expected, leads us to question the methodological accuracy of the classical approaches which seem inappropriate to identify and interpret these tools. Use-wear analysis should be integrated in the initial study phases in order to obtain a precise typology and a better understanding of the lithic reduction strategy. Concerning these typological aspects, it has been possible to identify some new tool types, only defined by specific use-wear traces and the morphology of their active edges. The curved knives, involved in plant material processing, and the ‘strike-a-lights’ are two examples of unretouched tools which can be highlighted by an integrated approach. Moreover, this methodology can contribute to the cultural and economic debates which are motivating Mesolithic archaeology in Northern France and Belgium. In this regard, the 500 used zones identified on the artefacts in the six assemblages studied allow us to identify the original management of natural resources during the Early Mesolithic. Of course, activities concerning the processing of animal materials are well-represented, including hide scraping, but the variability and the intensity of the use-wear stigmata are clearly limited. In particular, the low frequency of bone working, associated with the scarcity of dry hide cutting, could indicate a lower investment in these domains than during the Upper and Final Palaeolithic. This relative disaffection for animal materials could be explained by the increase of plant exploitation. The diversity of the use traces, identified in the four sites studied, indicates, without doubt, the existence of several ‘chaînes opératoires’, probably for basketry and weaving. This new investment in plant-based handicraft, allowed by the climatic warming at the beginning of the Holocene, is probably a major cause of the emergence of Mesolithic economies. Beyond this general overview, the data acquired from the different sites reveals a great variability in the number of tools and in the activities carried out by the last hunter-gatherers in each settlement. These results prove the efficiency of use-wear analysis to approach the economy of human groups in the landscape. In the future, we can be optimistic about identifying the function of settlements and the evolution of mobility patterns in the Mesolithic of Western Europe. These results are mainly based on the information provided by the unretouched tools and should thus encourage the generalisation of exhaustive approaches of the lithic assemblages, especially for the Mesolithic but also for the Palaeolithic and Neolithic periods.

Keywords: unretouched tools, use-wear analysis, Early Mesolithic, Paris Basin, Belgium.

Naujan est un site de l'Arctique de l'Est situé sur la côte nord-ouest de la baie d'Hudson, une région occupée au début du xxe siècle par les Aivilirmiut. Participant à la 5e expédition de Thulé (1921-1924) dirigée par Knud Rasmussen, Therkel Mathiassen fut missionné par le musée national du Danemark pour entreprendre les premières fouilles archéologiques dans l'Arctique canadien. Durant l'été 1922, il répertoria une vingtaine de maisons semi-souterraines sur le site de Naujan et ramena une riche collection d'environ 2 800 objets issus de douze habitations. Naujan, daté par Mathiassen autour de l'an mil, est généralement considéré comme le site de référence pour le début du Thuléen dans l'Arctique de l'Est. Une analyse approfondie du matériel provenant des maisons VI et VIII a été réalisée, soit environ 17 % de l'ensemble de la collection. Les modalités de sélection de la matière première ainsi que les techniques de fabrication et d'assemblage ont été analysées.
Notre étude a permis de différencier clairement les pratiques techniques des artisans ayant occupé les deux maisons, utilisation de forets métalliques et non plus uniquement lithiques pour réaliser les perforations des objets, par exemple. Les techniques de débitage (essentiellement par rainurage) et de façonnage (essentiellement par raclage, taille au canif, percussion lancée tranchante et abrasion) sont toutefois très similaires pour les deux maisons. Dans la maison VIII, le bois de caribou est la matière première la plus utilisée mais dans des proportions cependant proches de celles des défenses de morse et de l'os de cétacé. En s'appuyant sur des critères typologiques, un certain nombre d'objets paraît plus ancien que ceux de la maison VI. Ces derniers, qui se retrouvent dans les niveaux les plus profonds, ont la particularité d'être tous abrasés. Ceci pourrait constituer un nouveau critère de diagnose pour caractériser la phase ancienne.
Bien que de forme générale proche de la maison VIII lorsqu'elle est prise seule, la maison VI a la particularité de partager son tunnel d'entrée avec la maison V. L'ensemble correspond ainsi à une habitation bilobée, qui contraste avec la maison VIII composée d'une seule pièce. La maison V a peut-être servi d'espace de stockage séparé pour la maison VI. La maison VIII est moins profonde (60 cm) que la maison VI (150 cm) et le plus grand nombre d'objets trouvés dans la maison VIII (337 versus 151) pourrait également évoquer une gestion de l'espace distincte. Par ailleurs, il semble y avoir une plus grande diversité en termes de matières premières (et d'espèces) exploitées dans la maison VIII, avec plus d'objets décorés et d'éléments de transport, caractéristiques que l'on retrouve généralement dans les occupations thuléennes les plus anciennes. De nombreux objets de la maison VIII sont uniques (absents dans le reste de la collection). Certains sont faits à partir de matériaux rares (e. g. poterie, cuivre) ou ont des formes qui rappellent des objets trouvés dans l'Arctique occidental. Ils pourraient témoigner d'un transport d'objets sur de longues distances en relation avec l'installation pionnière des Thuléens dans l'Arctique de l'Est. Alors que la culture matérielle de la maison VI semble tout à fait homogène tant d'un point de vue typologique que technologique, la présence d'une brosse à dent d'origine japonaise ayant transité par les États-Unis pose de nombreuses questions, tant en termes de datation de certaines maisons de Naujan qu'en termes d'interactions avec les groupes alentours. Un contact direct avec les Européens et Euro-américains semble à exclure car les objets d'échanges traditionnels (perles, pipes, porcelaines) sont absents à Naujan. L'occupation peut-être très récente de certaines maisons pose toutefois la question de leur possible réemploi et de l'absence de contact jusqu'à des périodes extrêmement récentes. Notons cependant que depuis le xviie siècle, un certain nombre d'explorateurs ont navigué autour de la baie d'Hudson ; Christopher Middleton aurait atteint Repulse Bay, tout près du site de Naujan, en 1742 ; d'autres explorations, ainsi que la présence récurrente des baleiniers à partir de 1889, puis l'installation dans la région d'un poste de traite en 1903 par la Compagnie de la Baie d'Hudson, constituent autant d'occasions de contact avec les habitants de Naujan. Comment expliquer alors le peu d'objets industrialisés retrouvés sur le site s'il a été occupé jusqu'à la fin du xixe siècle ?
L'analyse technologique des maisons VI et VIII de Naujan fait émerger de nouveaux critères de diagnose tels que l'usage de l'abrasion et de forets lithiques ou métalliques dans les techniques de fabrication ou d'assemblage. La présence ou non de perforations, rainures ou, stries d'adhérence pourrait également devenir autant d'éléments clés de datation relative, tout comme l'emploi de matériaux spécifiques (i. e. défense de morse, os de cétacé, corne). Le réexamen des pièces collectées par Mathiassen révèle donc deux périodes d'occupation bien distinctes. Bien que pressentie à plusieurs reprises depuis la fin des années 1970, cette diachronie n'avait jamais été démontrée sur une période de temps aussi longue. Nos résultats montrent que le site de Naujan dans son ensemble ne peut plus être considéré comme la référence pour le début du Thuléen ; la chronologie d'occupation des maisons mérite d'être révisée dans son intégralité. Les critères typologiques et technologiques (matière première rare, technique diagnostique) établis par la présente étude permettront d'orienter la sélection des échantillons à faire dater de manière absolue pour confirmer l'écart de temps entre les différentes occupations.

Mots-clés : Thuléen, Arctique central, Canada, baie d'Hudson, Naujan, technologie osseuse, Therkel Mathiassen, 5e expédition de Thulé.

Naujan is located on the north coast of Repulse Bay in Eastern Arctic, a region occupied by the Aivilirmiut at the beginning of the 20th century. This site, the first professionally excavated in the Arctic, is generally considered as the "type-site" for the Eastern Arctic Thule culture. In 1922, during summer, Therkel Mathiassen from the National Museum of Denmark participated in the well-known 5th Thule Expedition led by Knud Rasmussen, and was in charge of the archaeological investigations. With the help of his Greenlandic assistant Jacob Olsen, he carried out excavations during six weeks and brought back a rich collection of artefacts. Of these about 2,800 artefacts were recorded from twenty structures. Houses VI and VIII were the first and most carefully excavated. All of the artefacts from these two features, which represent about 17% of the whole collection, were recorded and mapped. Information on spatial distribution as well as stratigraphic position were also carefully reported, but less precisely for the houses excavated later. Therkel Mathiassen used a continuous numeration for the whole site, the numbers assigned to artefacts corresponding to the order of their discovery. The osseous assemblages from Houses VI and VIII have been carefully analyzed. The present thorough technological study provides new information regarding the manufacturing processes for antler, bone and ivory artefacts. Grooving for the debitage sequence and whittling for the shaping dominate. The raw material selection, the techniques of manufacturing and hafting, as well as the perforations, clearly distinguish the technological traditions of the craftsmen who occupied House VI from those of House VIII. Our study leads to the conclusion that at least two distinct chronological stages of occupation can be identified. The Naujan site as a whole can thus no longer be considered as a reference site for a specific Thule phase of the Eastern Arctic occupation.

Keywords: Thule culture, Central Canadian Arctic, Hudson Bay, Naujan, bone technology, Therkel Mathiassen, Fifth Thule Expedition.

Claire Houmard
CNRS, université Paris-Nanterre,
UMR 7055 Préhistoire et technologie,
Maison Archéologie & Ethnologie
René-Ginouvès, 21 allée de l'Université,
92023 Nanterre cedex, France
clairehoumard@yahoo.fr

Bjarne Grønnow
Arctic Centre
at the Ethnographic Collections (SILA),
National Museum of Denmark,
Frederiksholms Kanal 12,
1220 Copenhagen, Denmark
Bjarne.Gronnow@natmus.dk

Dans le Midi de la France, de très nombreuses sépultures collectives datant du Néolithique final ont été recensées. Des inhumations individuelles en contexte domestique sont également attestées sur plusieurs gisements. Nous ajoutons à ce corpus trois sites provençaux : le Clos de Roque (Saint-Maximin-la-Sainte-Baume, Var), les Fabrys (Bonnieux, Vaucluse) et les Martins (Roussillon, Vaucluse). Le premier a fait l'objet d'une fouille récente, les deux autres opérations datent des années 1980. La documentation est inégale, notamment pour ces dernières, dans la mesure où l'opération a eu lieu avant le plein développement d'une réelle archéologie préventive et surtout de l'anthropologie de terrain. Le Clos de Roque a livré deux structures domestiques partiellement comblées avant l'apport de corps complets ou partiels. Deux inhumations superposées dans un creusement dont les limites ne sont pas visibles ont été découvertes sur le site des Fabrys. Sur les cinq inhumations individuelles du site des Martins, deux datent du Néolithique final. L'une d'entre elles était partiellement comblée avant le dépôt de l'individu. Dans la seconde, le sujet repose sur le fond de la fosse. Ces six inhumations présentent entre elles des points communs, notamment le type de structure ainsi que l'absence ou la quasi-absence de mobilier associé mais, aussi, une certaine variabilité : la localisation de l'individu dans la fosse, la position des sujets, la situation de la fosse par rapport à l'habitat auquel elle est ou semble associée, et leur place chronologique dans le Néolithique final. En faisant le point sur les inhumations en contexte domestique dans la région d'étude, nous proposons un lien étroit entre ces configurations, qui ne correspondent pas toutes à des sépultures, et celles observées au Néolithique moyen dans des contextes comparables. D'autres traitements des morts, hors des sépultures collectives, sont également connus au Néolithique final. Nous discutons de l'apport de ces découvertes pour la caractérisation des pratiques mortuaires de cette période, notamment la relation entre les inhumations individuelles, parfois plurielles, et les sépultures collectives, entre alternative et complémentarité.

Mots clés : Néolithique final, inhumation individuelle, pratique mortuaire, site domestique, sépulture collective, France méridionale.

IIn the Late Neolithic in Southern France, collective burials are very well documented. However, numerous single burials have also been discovered in domestic settlements. One of our purposes is to provide information regarding unpublished or partially unpublished single burials from three sites located in Provence: Le Clos de Roque, Les Fabrys and Les Martins. The second goal is to discuss the meaning of this mortuary practice whereas collective burials are the most frequent funerary treatment documented in this region.

The preventive archaeological excavation in 2011 of the Clos de Roque deposit (Saint-Maximin la Sainte Baume, Var), revealed pre-and protohistoric remains over 11,200 m2. Two silo pits attributable to the Late Neolithic delivered human remains. The feature FS 23 was partially filled before a young woman was placed in it. She was laid prone, without any goods. The pit was filled after the body was deposited. The feature ST 2247 was also partially filled before the human remains were put in it. The bones are disarticulated and the body of a young individual is incomplete. No goods are associated with the human remains.

The Les Fabrys site (Bonnieux, Vaucluse) was excavated in the 1980s, over an area of 1,200 m2. The burial is located at some distance from the domestic settlement. The limits of the pit were not visible in the subsoil. Two individuals are superposed. The first one, an old woman, was laid on her left side, upper and lower limbs flexed. The body decomposed in an empty space. The second body, incomplete, is separated from the first deposit by a layer of sediment. The young woman lies on her back, the left arm flexed, hand on shoulder. The lower limbs are extended but the thigh is flexed. No items were found.

The excavation at the site of Les Martins (Roussillon, Vaucluse) in 1986 revealed Middle and Late Neolithic remains. Five pits contained human remains. Two of them belong to the Late Neolithic. Both pits are circular. The body of an adolescent lay on the bottom of pit S 93. He was covered by a few stones. He lay on his right side, lower limbs flexed, with no goods. Feature S 104 was partially filled before the human remains were deposited. The position of the individual, an adult male, is singular. He was probably not slid into the feature, but deposited, in a seated position, the upper part of his body partially on the lower limbs. Both pits were filled in after the corpses were deposited.

In the South of France, Late Neolithic individual burials have been known since the beginning of the 20th century. They are however more numerous since the development of preventive archaeology. These discoveries shed new light on funerary customs at the end of the Neolithic. Collective burials are very well documented and provide a huge number of deceased compared with the Early and Middle Neolithic. These tombs were usually used over a very long period of time and led to commingled assemblages of bones. As a consequence, it is difficult to identify changes in the treatment of bodies over time and between individuals. Moreover, the relations between collective burials and settlements are not very well known. Recently, it has been argued that collective burial gives only the illusion of equality and that people gathered in death enjoyed different social status. It is also likely that only some of the community members had access to the collective tomb. We may, thus, wonder, where and how the others were treated. We propose that the individual burials associated with domestic contexts could give some clues on this issue. Most of the burials we list are installed in pits or associated with features related to domestic activities. Very few individuals are associated with grave goods and some of the subjects seem to have been slid down into the pit rather than deposited with care. Such configurations are very well known in the Middle Neolithic and considered as a particular form of treatment devoted to only a small part of the community. We propose that this treatment was, in the Late Neolithic, allotted to individuals who were not allowed burial in the collective tomb.

Remains of babies are rather rare in the Middle and Late Neolithic. However, in this period it seems that they were carefully disposed in close relation to residential features, as in Protohistoric periods and Antiquity, whereas, in the previous period, it is likely that their bodies were thrown into domestic pits or wells.

However, in some cases, the grouping of single tombs could have been an alternative to collective graves. Some cultural groups may have had a different ideology and chosen another way of dealing with their dead that probably reflects a different social organization.

Late Neolithic single burials also give new parameters for the discussion of single Bell Beaker graves. This culture has for a long time been associated with the reappearance of the fashion for individual burials. It is more complex than it seems. Firstly, single burials never disappeared in the Late Neolithic. Secondly, they are rather different from single burials containing Bell Beaker items. Grave goods and/or ornaments are very abundant in Late Neolithic collective tombs and absent in single burials, emphasizing the difference between the two customs. Funerary items such as daggers and pottery in Bell Beaker single tombs indicate a very different status of the deceased compared with those of the Late Neolithic. Individuality has neither the same value nor the same signification in the Bell Beaker Culture as in the other Late Neolithic cultures.

Human remains in domestic settlements are known on other areas of France, such as the West and South-West, in enclosure ditches. However, single burials in domestic pits are specific to Southern France and probably correspond to configurations known in the Middle Neolithic, reflecting very similar mortuary meanings.

Keywords: Late Neolithic, single burial, mortuary practice, settlement, collective burial, southern France.

Aurore Schmitt
UMR 7268 ADES,
Aix-Marseille Université, CNRS,
faculté de médecine nord,
boulevard Pierre-Dramard,
13344 Marseille cedex 15
aurore.schmitt@univ-amu.fr

Maxime Remicourt
UMR 5608 TRACES,
université Jean-Jaurès,
Maison de la recherche,
5 allées Antonio Machado,
31059 Toulouse cedex 9
m.remicourt@laposte.net

André D'Anna
UMR 7269, Laboratoire méditerranéen
de Préhistoire Europe Afrique,
Aix-Marseille Université, CNRS,
ministère de la Culture,
Maison méditerranéenne des sciences
de l'homme, 5 rue du Château-de-l'Horloge,
BP 647, 13094 Aix-en-Provence cedex 2
danna@mmsh.univ-aix.fr

Un diagnostic archéologique réalisé par le service départemental de l'archéologie de Vendée a mis au jour un dépôt de parures annulaires enfoui dans un contexte particulier, topographique d'une part, mais surtout archéologique, au lieu-dit le Vrignoux à Aizenay. En effet, des fosses associées à ce dépôt livrent pour certaines de la céramique, et pour d'autres des fragments de stèles, réemployés comme éléments de calage. Deux autres stèles de petite taille, mais complètes, ont été découvertes, l'une dans un fossé parcellaire subactuel, et l'autre, mise à bas à côté de sa fosse d'implantation. Cet article propose, après une présentation du contexte de découverte et des structures du site, une description des stèles, à la fois sur un plan technologique ce qui permet de démontrer qu'elles ont bien été façonnées et sur un plan typologique. En l'absence d'élément de datation direct du contexte originel d'implantation de ces stèles, leur mise en perspective dans un cadre plus large, à l'échelle notamment du domaine atlantique où les découvertes de ces petites stèles anthropomorphes se multiplient, permet d'envisager une origine néolithique. L'étude céramique, la réalisation de plusieurs datations 14C et l'analyse du dépôt de parures annulaires convergent pour attribuer l'enfouissement du dépôt et la remobilisation des stèles entre la fin du Bronze moyen 1 et le début du Bronze moyen 2, soit entre la fin du xvie et la première moitié du xve siècle avant notre ère. Si l'interprétation du contexte dans lequel le dépôt a été effectué est discutée, on retiendra surtout, outre la probable pérennité du site, l'originalité des pratiques étudiées. En effet, le contexte du dépôt, en lien avec des structures bâties, est inédit dans la région. Ces pratiques sont mises en relation avec les remobilisations mégalithiques à l'âge du Bronze. Qui plus est, aux objets métalliques sont associées de la céramique et des datations 14C. Celles-ci suggèrent une attribution légèrement plus ancienne que ce qui était envisagé jusqu'ici pour ce type de parure annulaire à tige grêle, à section quadrangulaire ou biconvexe et extrémités indifférenciées, dont la répartition correspondrait aux contours d'un groupe régional spécifique, que le dépôt d'Aizenay contribue à mieux caractériser.

Mots-clés : Néolithique, stèles, âge du Bronze, Bronze moyen, dépôts métalliques, parures annulaires-parures.

An excavation prior to the widening of a road in the north-west part of Vendée (France) has revealed a hoard of bronze annular ornaments in a very specific context. The site, set at the top of a hill, dominates a landscape open towards the west and the Vie valley (and an Early Bronze Age occupation which will soon be excavated). This site, called Le Vrignoux, is seriously eroded but several pits have been discovered. Pottery has been found in some of them and stelae fragments reused as wedges in others. Study of the pottery, even though the assemblage is modest and very fragmented, allows the ultimate occupation to be dated to the end of the Early Bronze Age or the Middle Bronze Age. Some recent discoveries give a glimpse of pottery production of this period (Viau, 2005 and 2010), although it is still under-researched. Chronological attribution is also based on carbon dating of charcoals discovered in the stelae pit, which means these events are broadly simultaneous.

The stelae were reused during the Bronze Age, but we sought the original date and context of their erection. The technological study (by E. Mens) of two complete stelae and parts of four or six others, which are all in granite (except one in gneiss), has demonstrated that they were shaped by man. The stelae are characterized by an apical rostrum over a shoulder, to a greater or lesser degree, which means that we can classify them as anthropomorphic stelae.

This general profile refers to many well-known examples in Atlantic Neolithic megalithic art, especially in megalithic architectures (dolmens). The unpolished aspect of the representation and the sculpture is a recurrent observation. The Aizenay stelae form part of a series of recent discoveries of small anthropomorphic megaliths in western France: Le Bois du Fourgon, Avrillé, Vendée (Benéteau-Douillard, 2012), Le Douet and Groah Denn in Hoëdic island, Morbihan (Large and Mens, 2008 and to be published; Large, 2014 and 2015), for example. Even if we can observe an important degree of individual variability, these monuments bring to mind famous examples from Switzerland, such as those of Yverdon (Voruz, 1990 and 1992). On the basis of these regional and extra-regional comparisons, the Aizenay stelae can be placed in the Neolithic period, and more precisely in the Middle Neolithic. The original architecture and structure of this set of stelae, whose geological features are varied, are unknown.

One of the most significant points of this excavation is the reuse of stelae at a period subsequent to their erection, during the end of the Early Bronze Age or, more probably, the Middle Bronze Age (between the end of the 16th and the first half of the 15th century BC). In addition to study of the pottery, the date is confirmed by a radiocarbon date from a stratigraphic layer occurring after the (most probably voluntary) toppling of a stela which was discovered under its foundation pit.

Therefore, Bronze Age populations recycled the stelae, probably according to iconoclastic practices (which could have multiple meanings), at the same time as they buried the hoard. It is impossible to date precisely the intentional destruction of some of them, but the loosening of one stela, and pottery items associated with stelae fragments (used as wedges) in several pits, shows that it undoubtedly took place during the Middle Bronze Age, together with the placing of the hoard of annular ornaments. The hoard (studied by M. Nordez) consists of a pile of eight annular bracelets fitted into one another. They are homogeneous in terms of typology. They have an open shape and their internal diameter is from 50 and 60 mm, allowing them to be considered as wrist ornaments. The ends are undifferentiated, abrupt or tapered. The section is rather thin (width is from 4 and 7 mm; thickness is from 3.5 to 5 mm), of quadrangular or biconvex shape. Although preserved surfaces are few in number, the absence of any decoration should be noted.

Several points of comparison can be provided, from the point of view of either typology or the mode of depositing (as a pile or heap). The most obvious is a hoard recently discovered at Saint-Lumine-de-Clisson (Loire-Atlantique), 40 km from Aizenay, which includes thirteen bracelets, typologically similar to those of Aizenay (Boulud-Gazo et al., 2012; Boulud-Gazo and Nordez, 2013). The Saint-Lumine hoard defined a new eponymous sort of annular ornament dated from the second Middle Bronze Age. In some hoards, this type co-exists with ribbed bracelets. Comprehensive studies of several hoards have shown this ornament type is characteristic of a restricted geographical area in Vendée, southern Loire-Atlantique and Maine-et-Loire, up to Cher, including the Malassis hoard, Chéry (Briard et al., 1969).

Despite its poor state of preservation, the Aizenay hoard provides some more interesting information. Firstly, it is the first regional context that associates a bronze hoard with built structures. Few contemporary examples occur. Three hoards from Ribécourt-Dreslincourt (Oise) could be mentioned, where each hoard is in association with a posthole (Guérin and Armbruster, 2015).

In addition, the ornaments confirm the archaeological reality of the Saint-Lumine type, which is characterized by slender stem bracelets, without decoration. The deposit modein a pile appears to be a specific and infrequent practice, just like the bracelets morphology. Except for the Malassis hoard, all discoveries are concentrated within a 60 km radius. Several well-dated comparisons, like the one from Les Ouches, Auzay, Vendée (Lourdaux and Gomez de Soto, 1998) allow us to date the Aizenay hoard to the Middle Bronze Age (and probably the second half of this period, according to the examples studied). Other artefacts and carbon dating confirm this hypothesis.

The Aizenay site suggests complex practices relating to Bronze Age metal deposits and older anthropomorphic stelae. This phenomenon could be put into perspective with the reuse of megalithic burial sites which are very numerous in our area.

Keywords: Neolithic, stelae, Bronze Age, Middle Bronze Age, metallic hoards, bracelets- annular ornaments.

Axel Levillayer
Grand Patrimoine de Loire-Atlantique,
département de Loire-Atlantique, Nantes,
chercheur associé à l'UMR 6566 « CReAAH »
axel.levillayer@loire-atlantique.fr

Marilou Nordez
doctorante, université Toulouse - Jean-Jaurès,
UMR 5608 « TRACES »
marilou.nordez@gmail.com

Emmanuel Mens
Archéo Atlantica, Batz-sur-Mer,
chercheur associé à l'UMR 5608 « TRACES »
emmanuel.mens@free.fr

La concentration exceptionnelle de dépôts métalliques du Bronze moyen 2 en Médoc est connue depuis le xixe siècle. Cette pratique n'est pas limitée au domaine atlantique, mais la concentration de cinquante-sept dépôts sur un territoire aussi restreint en fait un cas particulier à l'échelle de la France. Le Médoc regroupe à ce titre plus de la moitié des dépôts découverts en Aquitaine. Ces dépôts se caractérisent par des haches en bronze massives, principalement du type médocain, mais aussi de productions exogènes, telles que les haches à talon de types breton, normand et beaucoup plus rarement du Centre-Ouest. Au total ce sont plus de 660 kg de bronze qui ont été découverts en Médoc depuis le xixe siècle. Pourtant, cette région ne dispose d'aucun gisement métallifère. Les échanges avec l'ensemble du domaine atlantique sont bien attestés par les productions métalliques venues de ces régions et qui se retrouvent dans les dépôts du Médoc. La relative rareté des productions continentales, telles que les haches du type du Centre-Ouest ou d'Orléans, prouve que les sociétés médocaines étaient prioritairement tournées vers l'Océan.

La situation géographique, entre l'Océan et le fleuve Garonne, a contribué au développement des échanges à longue distance. La place stratégique qu'occupait le Médoc au Bronze moyen a donc permis aux populations de la péninsule de s'approvisionner en ressources métalliques. La question se pose alors de savoir quelles contreparties ces sociétés pouvaient proposer dans les échanges. Des indices de production de sel sont attestés plus au sud à la dune du Pilat pour le Bronze moyen, mais également sur le littoral médocain pour des périodes plus récentes. Il s'agit d'une des possibilités, mais les données sur l'habitat et l'artisanat manquent cruellement pour pousser la réflexion. La quasi disparation de la pratique des dépôts au Bronze final coïncide avec une modification de l'environnement. La structure économique a pu à cette occasion être ébranlée, ne permettant plus aux sociétés de l'âge du Bronze du Médoc de disposer des ressources nécessaires au maintien des échanges à longue distance et à leur approvisionnement en ressource métallique.

L'évolution de l'environnement autour de l'estuaire de la Gironde est en effet très importante durant l'âge du Bronze, avec des modifications importantes du fonctionnement des marais médocains. Nous pouvons mettre en évidence que la pratique des dépôts métalliques au Bronze moyen se concentre exclusivement autour des marais tidaux, dont nous savons grâce aux études paléoenvironnementales qu'ils étaient en eaux à cette époque. Ce lien fort entre les dépôts et les zones humides questionne sur la relation qu'ils pouvaient entretenir. Nous pouvons envisager un rôle symbolique, en lien avec les eaux divinisées, qui est attesté par ailleurs notamment au Bronze final et au premier âge du Fer, ou bien encore à une structuration pragmatique du territoire. Il semble qu'il y ait une récurrence de cette pratique sur des mêmes espaces, comme en témoignent certaines découvertes réalisées autour de Pauillac, ou encore de Soulac-sur-Mer, avec plusieurs dépôts enfouis à quelques mètres les uns des autres. Ce comportement répétitif et codifié connote une certaine forme de ritualisation des dépôts métalliques. Si l'essence même de cette pratique est difficilement compréhensible pour notre mode de pensée contemporain, le cas du Médoc permet d'entrevoir des pistes de réflexion qui devront être confrontée à l'avenir aux autres régions de la façade atlantique.

Mots-clefs : Aquitaine, Médoc, Bronze moyen, dépôt métallique, haches, échanges, paléoenvironnement.

We have known since the 19th century that the Médoc region offers an exceptional number of metal hoards dating from the Middle Bronze Age 2. Metal hoarding is not restricted to the Atlantic coastal area but a concentration of fifty-seven hoards - more than half of the Aquitaine hoards - in such a limited area clearly makes it exceptional. Many isolated objects have also been found in the Médoc region, but we do not know why they were deposited there. Were they a discreet practice of metal hoarding, or just lost objects? These Médoc hoards are made up of massive bronze axes, mainly of the Médoc type, but palstave axes from Brittany and Normandy, and more rarely from the Centre-Ouest region, can also be found. In total, over 660 kilos of bronze artefacts have been discovered in the Médoc region since the 19th century, even though the area had no copper or tin deposits. This is further proof of the trade that existed along the Atlantic coastline. The relative scarcity of productions from the Centre-West or from Orléans indicates that the Médoc societies were turned towards the ocean, and the location, between the ocean and the River Garonne, contributed to the development of long-distance exchanges. This strategic position, with the difficulty of navigating southwards in the Bay of Biscay and the proximity with the River Garonne, allowed the populations in the Médoc region access to metal resources, raw materials as well as finished products. Furthermore, some signs of bronze production have been discovered in Médoc, such as moulds of flanged axes and an anvil on the Lède du Gurp site (Soulac-sur-Mer). What did those populations give in exchange for such resources? Indications of salt production during the Middle Bronze Age were found south of the Dune du Pilat, as well as signs of later production along the Médoc coastline. This could answer the question, although little is known about domestic crafts and local activities.

The environmental change around the Gironde estuary during the Late Bronze Age led to important alterations to the Médoc marshlands. The disappearance of the practice of metal hoarding during the Late Bronze Age in Médoc coincides with this transformation of the environment. In fact, during the Late Bronze Age and the Iron Age, the Médoc landscape became increasingly humid. This change may have disrupted the economic structure of Middle Bronze Age societies, leading to an interruption in long-distance exchanges and consequently in the supply of metallic resources. This hypothesis may however mask the possibility of a rupture in the social structure of these societies. Metal goods were probably controlled by an elite hit by a power crisis, which led to a drop in metal supply and metal hoarding. At the same time, many hoards were deposited in other parts of the Atlantic area, such as Brittany, Normandy and the Loire Atlantique département. This difference between North and South demonstrates that practices and societies changed dramatically in the Médoc between the Middle and the Late Bronze Age.

It is possible to prove that metal hoards during the Middle Bronze Age were exclusively situated near tidal marshes which were not dry at the time, as is indicated by palaeo-environmental studies. There is a strong link between hoards and wetlands, but what was the connection between the two? We can envisage a symbolic role, linked to the sacred nature of water? a possibility also noted during the Late Bronze Age and the early Iron Age or more simply the result of a pragmatic delimitation of territory. It would seem that this was a recurrent practice in those areas: hoards have been discovered only a few metres apart, in Pauillac and Soulac-sur-Mer. This standardised behaviour implies a certain ritualization of metal hoarding. Even if the meaning or motivation behind such practices is difficult to apprehend for our modern way of thinking, the case study of the Médoc region can pave the way for new hypotheses. It would be interesting to try this model out on other sites along the Atlantic coastal area for the Middle Bronze Age.

Keywords : Aquitaine, Médoc, Middle Bronze age, metal hoarding, axes, trade, paleoenvironmental.

Florian Couderc doctorant UMR 5608 TRACES, université Toulouse 2 - Jean-Jaurès, équipe RHAdAMANTE florian.couderc@hotmail.fr

Cette étude s'intéresse à une variété de parure annulaire dont l'originalité se devait d'être soulignée, qu'il est proposé de nommer bracelets du groupe d'Orgon en référence au dépôt de Notre-Dame de Beauregard, dans cette localité des Bouches-du-Rhône, qui en a livré le plus grand nombre. Ce groupe est illustré par quarante objets provenant de douze sites répartis dans quatre départements du Sud-Est de la France. Il correspond à des bracelets ouverts, de tige mince à section en D aplati dont les extrémités sont ornées de bourrelets transversaux. Les contextes de découvertes de ces objets, de nature variée (habitats, dépôts et nécropoles) convergent vers une datation de leur production au Bronze final 3b et éventuellement lors de la phase de transition entre l'âge du Bronze et l'âge du Fer, aux ixe et viiie s. av. J.-C. Ces parures sont munies de décors couvrant la totalité du jonc et présentant une certaine variabilité en fonction de la technique utilisée et des motifs représentés. Leur combinaison permet de définir 3 types qui s'excluent géographiquement, la majorité des exemplaires de chacun d'entre eux se concentrant dans un cercle ne dépassant pas les 50 km de diamètre. Ces différences peuvent signaler l'existence de trois ateliers spécialisés possédant chacun un type d'ornementation spécifique. Plusieurs indices, comme un moule, témoignent en effet d'une production locale de ces objets. Les bracelets de même type provenant du dépôt de Notre-Dame de Beauregard et de la nécropole des Lauzières à Lourmarin dans le Vaucluse présentent par ailleurs des signatures élémentaires semblables, qui laissent penser qu'ils proviennent d'un même atelier. Une autre hypothèse expliquant la production de ces types est que ces décors furent les supports d'expressions identitaires et que leur zone de répartition principale corresponde à un territoire communautaire. Une bonne correspondance peut en effet être établie entre la répartition de ces objets et les styles céramiques du bassin inférieur du Rhône à l'extrême fin de l'âge du Bronze. Localisées à distance des ressources en cuivre des Alpes et du Sud du Massif central, ces sociétés ne sont néanmoins pas dépourvues d'une production métallurgique propre, dont les bracelets du groupe d'Orgon constituent les témoins les plus directs.

Mots-clefs : Bronze final, Sud-Est de la France, bracelets, parures annulaires, décors, typologie, territoires, ateliers.

This paper focuses on a variety of annular ornament whose originality should be emphasized, and which we propose to call Orgon group bracelets in reference to a hoard discovered in Notre-Dame de Beauregard (Bouches-du-Rhône) which contained the majority of the known objects. This group is illustrated by fourty objects, to which must be added a possible casting mould, from twelve sites spread over four departments of southeastern France: Bouches-du-Rhône, Alpes-de-Haute-Provence, Vaucluse and Ardèche (see table 1). It corresponds to an open bracelet provided with a thin stem with a flattened D-section the ends of which are decorated with transverse ribs. These bracelets come from varied contexts (human settlements, hoards and tombs; fig. 12) and can be dated to the Late Bronze Age 3b, or possibly to the transition phase between the Bronze Age and the Iron Age (i.e. between the 9th and 8th centuries BC). These ornaments probably are objects of everyday life, found in the form of fragments on settlements, but which can also be parts of much more ostentatious ornaments, as illustrated by the rich costume of the hoard of the Déroc cave. The tombs discoveries indicate that they may have been worn either singly or in series. The tomb of the Lauzières and the two hoards clearly illustrate this, with increasing lengths allowing the shape of the forearm to be matched (fig. 13). They are provided with decorations covering the whole bangle and presenting a certain variability according to the technique used and the patterns depicted. The combination of the latter makes it possible to define 3 types: the Lauzières type corresponding to the specimens supporting only a relief decoration, the Notre-Dame de Beauregard type type characterized by engraved decors of lines, crosses and triangles, finally the Déroc type, also determined by the presence of engraved ornamentations, which combines the cruciform decorations with ovoid patterns (fig. 10). These decorative types find some correspondence with the morphology of the objects, considering their maximum length and width (fig. 11). These three different types appear to be not only located in distinct geographical areas, but also concentrated in circles which do not exceed 50 km in diameter (fig. 14). These differences may indicate the existence of three specialized workshops, each with a specific type of ornamentation. Several indications testify to a local production of these objects, which could proceed from the evolution of the Pourrières type, dating back to the Late Bronze Age 2b-3a. First of all, we can evoke the presence of a casting mould (fig. 3, no. 3) coming from the tumulus of the Moullard sud in Lambesc (Bouches-du-Rhône), which was transformed in a second stage for the production of socketed axes. Secondly we can notice that the type of bracelets discovered in the hoard of Notre-Dame de Beauregard and in the Lauzières necropolis have similar elementary signatures, which suggest that they originated from the same workshop (fig. 15). They are moreover similar to those of other objects coming from the Southern Alps, where this group of bracelet is unknown (fig. 16). Its creation could therefore correspond to a desire to produce less massive, less metal-consuming annular ornaments in an area located further from copper resources. Another hypothesis explaining the production of these types is that these decorations express identity and that their main geographical distribution corresponds to the territory of a human community. A good correspondence can be established between the distribution of these objects and the ceramic styles of the lower Rhone basin at the end of the Bronze Age. However, regardless of the fact they are located quite far from the copper resources of the Alps and the south of the Massif Central, these societies have their own metallurgical production, of which the Orgon group bracelets constitute the most direct witnesses.

Keywords: Late Bronze Age, Southern France, bracelets, annular ornaments, geometric design, typology, territories, workshops.

Thibault Lachenal

chargé de recherche CNRS

UMR5140 Archéologie des sociétés

méditerranéennes (ASM),

univ. Paul-Valéry Montpellier, CNRS, MCC,

F-34000 Montpellier, France

thibault.lachenal@cnrs.fr

La couche 4 du Petit Cloup Barrat (Cabrerets, Lot) a livré une probable ébauche d’objet arciforme (ou « anneau ouvert ») en bois de cervidé. Ce morphotype n’était jusqu’à présent connu qu’au travers des dix exemplaires mis au jour dans la grotte du Placard (Vilhonneur, Charente) et de celui de la série dite de la « Cave à Endives » (collection André-David, Cabrerets, Lot). Sa datation directe par SMA (17720 ± 90 BP, soit 21770-21122 cal. BP), réalisée au laboratoire d’Oxford, semble renvoyer, selon les cadres chronologiques aujourd’hui disponibles pour le Sud-Ouest français, aux traditions techniques badegouliennes. Cette datation est par ailleurs cohérente avec la mesure d’âge obtenue au sein du même ensemble à partir d’un éclat de bois de renne débité par percussion lancée, déchet technique caractéristique de la méthode de production des supports au Badegoulien.

Au-delà d’un nécessaire travail de critique archéostratigraphique de la couche 4 qui semble, d’après les données typo-technologiques et radiométriques, témoigner d’une diachronie insoupçonnée jusqu’ici (Magdalénien moyen, inférieur, Badegoulien et Solutréen), la datation directe de l’ébauche d’anneau ouvert du Petit Cloup Barrat et son examen technique apportent un complément important à l’étude de ce type d’objet. Ces résultats, étendus aux autres objets arciformes considérés jusqu’alors comme magdaléniens, constituent une première étape vers la re-contextualisation chronoculturelle d’un morphotype aussi rare qu’original.

Mots-clés : objet arciforme, anneau ouvert, datation directe SMA, Paléolithique supérieur, Badegoulien, Magdalénien, industrie osseuse.

Le Petit Cloup Barrat cave (Cabrerets, Lot, France) is located on the hill of Pech Merle, 800 metres from the famous Palaeolithic art cave. It opens onto the edge of a collapsed doline of 500 m2. Since 2003, the different excavations (by Castel and Chauvière, then by Castel) have revealed an important Upper Palaeolithic stratigraphy, with several archaeological assemblages. More than ten millennia have been recorded discontinuously in the various sedimentary complexes, with occupations dating from the end of the Gravettian to the Middle Magdalenian.

Layer 4 of the deposit yielded an ‘arciform object’ rough-out (so-called ‘omega-like shape open ring’) made in antler. Until now, this morphotype was known only through the ten specimens discovered at the end of the 19th century in Le Placard cave (Vilhonneur Charente, France) and that of the so-called ‘Cave à Endives’ archaeological collection (André David collection, Cabrerets, Lot). The direct dating of this rough-out using the AMS radiocarbon method (17720 ± 90 BP or 21770–21122 cal. BP) was carried out in the Oxford laboratory. According to the chronological frameworks now available for south-western France, it seems to refer to the Badegoulian technical traditions. The implications of the absolute chronological positioning of this original object are important, first on the scale of Le Petit Cloup Barrat, but also for the Upper Palaeolithic ‘arciform objects’.

At Le Petit Cloup Barrat, Badegoulian occupations have been identified since 2003 inside the cave, following the discovery of lithic and osseous remains (respectively raclettes and antler flakes). They were discovered in layer c8a1. This layer has a remarkable sedimentary signature and was dated by the AMS radiocarbon method from a faunal remain (18595 ± 150 BP or 22876–22101 cal. BP). The direct 14C dating of the open ring fills a gap in the radiometric chronology of the outer area of the main excavation sector, confirming the presence of the Badegoulian, already suspected through the recurrent demonstration of lithic “markers” on the heart and bottom of layer 4. Moreover, the direct 14C dating of the ‘omega-like shape open ring’ is consistent with another AMS date, in layer 4, from an antler flake produced by knapping (17800 ± 80 BP or 21838–21292 cal. BP), technical waste characteristic of the Badegoulian. These two dates are older than one of a splinter fragment from the same layer, and produced by grooving (16,100 ±70 BP or 19629–19208 cal. BP). Taken together, they refer to the anteriority of the use of knapping in comparison to the groove and splinter technique, according to the general evolutionary model proposed elsewhere, without excluding, for the moment, the possible coexistence of other elements resulting from these two modes of blank production, at least on the scale of the sedimentary unit which contains them. The bone industry data are added here to other observations about typo-technological analysis of the lithic industry and radiometric data to confirm the diachrony in layer 4.

The direct radiocarbon date of the ‘omega-like shape open ring’ rough-out from Le Petit Cloup Barrat and its technical examination also provide a substantial complement to the study of this rare and original object. The results can be extended to the other ‘arciform objects’, previously considered as Magdalenian, and constitute a first step towards their chronocultural re-contextualization. Firstly, if the attribution to the Badegoulian of the ‘arciform object’ type were to be confirmed, it would increase a corpus constituted of elements which, for most of the currently available series, suffer from the disconnection between the production of blanks and the different typo-functional categories (eye needles, awls, chisels and projectile points). Secondly, the direct dating of the ‘omega-like shape open ring’” rough-out from Le Petit Cloup Barrat encourages a critical re-reading of the typological assemblages yielded by Le Placard cave, according to the radiometric framework from the Badegoulian assemblages of this famous Palaeolithic site. As these assemblages served to seriate the Magdalenian periods, they need reassessment, using the AMS radiocarbon method, in addition to ‘arciform objects’ with other morphotypes such as ‘spiked single bevelled points’ and ‘Lussac-Angles points’.

Keywords: ‘arciform object’, ‘omega-like shape open ring’, radiocarbon date (AMS), Upper Palaeolithic, Badegoulian, Magdalenian, bone industry.

La carrière-atelier de Beulin à Saint-Germain-le-Guillaume (Mayenne) est située au cœur d’une zone où les filons de dolérite abondent. D’après les onze datations radiocarbones obtenues l’affleurement a été dégagé et exploité dès le milieu du Ve millénaire BC et la carrière est restée ouverte et exploitée jusqu’au milieu du IVe millénaire BC. Tous les affleurements de dolérite de cette zone ne sont pas exploitables. Les affleurements font l’objet d’exploitation en front de taille et une utilisation du choc thermique est attestée à la fin de l’exploitation, mais a pu être mise en œuvre plus tôt. L’objectif de l’exploitation est de dégager des blocs présentant un plan de frappe supérieur à 90°. L’un de ces blocs tombé au pied du front de taille a même pu être replacé sur le négatif laissé dans le filon. L’emploi de percuteurs en dolérite et en grès pesant plus de 5 kg et pouvant aller jusqu’à 70 kg est attesté. La quantité de déchets d’extraction et de déchets de taille semble avoir conditionné une gestion des déchets très opportune, avec des tas mis en place au gré des besoins et déplacés plus tard si nécessaire.

Les blocs sont ensuite utilisés en nucléus par percussion lancée sur enclume avec la volonté d’obtenir des éclats épais et larges pour permettre le façonnage de lames de hache courtes. Une taille bifaciale est alors mise en œuvre afin d’égaliser les bords puis une deuxième série d’enlèvements permet de mettre le tranchant en forme de biseau et le talon en forme de légère pointe.

L’étude anthracologique vient compléter les observations sur l’emploi du choc thermique (stockage préalable du bois, calibres et essences de bois variables pour une forte montée en température). Enfin, quelques pistes sur la question de la diffusion de ces produits sont proposées.

Mot-clefs : Néolithique moyen, Massif armoricain, dolérite, carrière, atelier, hache, chaîne opératoire (lithique), paléoenvironnement.

Wishing to emulate the advances made by their British colleagues (Stone Axe Studies), during the 1960s P.-R. Giot and J. Cogné worked to identify the main sources of the rocks used to create the Breton axes. They had success in identifying one particular dolerite, Type A metadolerite, which was sourced to a work area at Sélédin in the commune of Plussulien (Côtes-d’Armor) by C.-T. Le Roux in 1965. Another dolerite extraction site was subsequently confirmed at Le Pinacle in Jersey but, following this, work on sourcing the axes found in the west of Armorica was largely abandoned. This article revives the search and presents the results of the 2010 and 2012 fieldwork carried out on the Beulin dolerite axe factory at Saint-Germain-le-Guillaume (Mayenne).

The north-western area of Mayenne is characterized by a swarm of dolerite dykes, which are orientated between NNW/SSE and N/S. Allowing for some uncertainty, the Beulin dyke dates from the Devonian/Carboniferous boundary. Geographically its site is on a crest line rising to 129.5 m NGF (Niveau Général Français, ± equivalent to OD), the relief of which reflects the presence of the dyke. The rock on which the dyke intruded is Alexain Cadomian age Granite. The Beulin dyke in the area as a whole often takes the form of small domes on the surface but with the rock itself nonetheless buried quite deeply in the soil profile. On digging down the dolerite is eventually encountered as rounded core stones embedded in a dark brown clayey loam. The larger core stones are sometimes quite far apart with smaller ones in between, or separated by joints. At the field site the dolerite dyke is some 8.5 m wide and the dominant joints cutting the dyke have a different orientation, running NW/SE. Within the dyke the least altered, freshest dolerite occurs as rounded masses with the rock becoming progressively more angular as more or less joint-bounded blocks occur.

The discovery of the site resulted from following up on a surface scatter of flakes. Known 20th century agricultural working of the land had left gouges in the dolerite at a fairly constant depth of some 12–13 cm. Gouges, probably from plough blades and suchlike toothed implements, can be as wide as 10-12 mm. Patterns of crossing marks quite possibly reflect the ploughing of these fields during the 1960s on N/S and E/W lines. The various traces had to be identified for what they are and separated out from those resulting from prehistoric workings.

Over the cleared surface seven depressed areas correspond to recent stripping of the surface and some feature a resultant scatter of fresh fragments of rock. The form of the stripped areas is variable depending on the size and shape of the removed blocks. The loose associated sediment is aerated, rich in brown loam and contains a number of dolerite fragments. In addition, the outcrop situated in AB9-10 reveals very clearly the negative form of the blocks removed.

Only part of the whole site, labelled 1, will be described here. One of the first layers in this section was US 9, which is located against the working face of the outcrop and is a very dark brown to black clayey loam. It contains numerous charcoal fragments, sometimes forming layers themselves, and also contains pieces of worked dolerite. This layer has evidence of use of the thermal shock method to extract the rock and is dated to 4815 ± 35 BP. Overlying US 9 is a dark brown loam containing centimetre-sized pieces with only the occasional larger element. This layer appears more complex since some squares had more broken fragments than others and, in addition, the dip of the lithic fragments indicated different accumulations. Elsewhere, at least one of these areas contained abundant charcoal with a grey-blue sandstone hammer. Carbonised fragments of probable apple or similar wood from this level gave a date of 4760 ± 40 BP  which was supported by three others: 855 ± 30 BP, 4800 ± 35 BP and 4835 ± 35 BP. Layer US 5 was a light brown loam containing rare, centimetre-sized, joint bounded fragments plus rounded lumps, seemingly residual pieces of the bedrock. US 4 is a loamy orange sand, partly resting on US 5, and containing rare small reject fragments from the extraction process and the occasional worked piece. This level is interpreted as the spread of debris from an earlier extraction event working towards exposing the dyke on its west flank; it is dated to 5660 ± 35 BP. US 2 covers the whole of the US layers. It is a brown loam with abundant fragments from the extraction process which is taken to represent the collapsed remains of this debris; it is dated to 4890 ± 35 BP. US 1 contains a higher proportion of worked fragments and resulting spalls than the debris from the extraction process. The fragments are embedded in a light brown loam and may be considered a pile of reject spalls, an interpretation supported by the finding of a nucleus, placed in a flat position.

Not all the dolerite outcrops were exploited. Numbers 1, 3, 5 and 6 are examples. It appears that the Neolithic people preferred the outcrops with a more widely spaced network of joints to those without. The favoured outcrops (numbers 2, 4, 7 and 8) were subjected to thermal shock treatment of their exposed faces—definitely in the closing stages of exploitation but almost certainly earlier on too. The aim was to enlarge the joints to allow the extraction of thicker blocks with a strike plane of 90° or more. One such block, fallen to the base of the working face, weighed about 150 kg and could be retrofitted to its original position. Carbonised fragments of probable apple or similar wood, discovered beneath this block, have a date of 5410 ± 30 BP, which places it neatly at the beginning of the Middle Neolithic II. Another block of about 200 kg was found and put aside for future analysis with the intention of obtaining pieces of dolerite that were thick and wide enough to allow the creation of short blades. From these first shapes, a bifacial form was fashioned by matching the edges, followed by trimming these to make the bevelled cutting edge and the narrow slightly pointed end. An accumulation of waste fragments, with a massive 66 kg nucleus at its centre, is the last stage of the sedimentation process and shows clearly that the fashioning of the axes occurred on site.

Going beyond the discussion centring on the use of the thermal shock process (prior collection of wood specially chosen to create the strongest rise in temperature), the anthracological study has added to the data already known for the Mayenne département and considers the presence of beech.

To complete the study, we open up the discussion to consider two issues. The first is the question of the diffusion of the axes, a line of enquiry that still has a long way to go, while the second focuses on the existence of the long rough-outs that probably served for the fashioning of the fusiform axes so well-known over the northern half of the Armorican massif.

Keywords: Middle Neolithic, Armorican massif, dolerite, quarry, workshop, stone axe, ‘chaîne opératoire’ (lithic), palaeo-environment.

L’analyse en lames minces d’une série de céramiques campaniformes de style pointillé géométrique issues des fouilles réalisées par J. Courtin et G. Sauzade en Avignon dans les années 1960-1970 a permis de déterminer les origines des terres exploitées, le plus souvent locales mais avec des vases étrangers aux contextes géologiques des sites, ainsi que la nature des dégraissants introduits dans ces terres au moment de la préparation des pâtes. Ces analyses complètent celles réalisées il y a 25 ans sur les céramiques de même style d’autres sites provençaux et plus récemment languedociens. Elles ont également été confrontées aux analyses réalisées sur des corpus de vases attribués aux cultures autochtones.

Même si macroscopiquement la céramique campaniforme de style pointillé géométrique (style 2 d’O. Lemercier) est nettement différente de celle des groupes culturels locaux (Fontbouisse, Vérazien), elle présente en fait une proximité marquée dans l’emploi des types de dégraissant, lorsqu’ils ont été introduits dans les pâtes, ainsi que des terres exploitées similaires. Ces convergences ne peuvent pas être le fruit du hasard et résultent vraisemblablement de relations étroites, de transferts et d’emprunts entre potiers de différentes traditions.

Nous proposons d’expliquer cette situation par la mise en place de relations précoces dans le Sud de la France entre les populations autochtones et des individus étrangers qui fabriquent en majorité localement des gobelets campaniformes. Une adoption probablement rapide (quelques décennies) de cette céramique par une partie des indigènes conduit à la création de nouvelles formes et à la diversification des décors comme cela est illustré sur les sites d’Avignon, mais également d’autres sites provençaux et languedociens.

Mots clés : Sud de la France, Néolithique final, Campaniforme, céramique, dégraissant, matière première, emprunt.

The thin section analysis of a series of Bell Beaker geometric dotted style ceramics and of a mixed beaker from the city of Avignon (Vaucluse) allowed us to consider the current state of our knowledge and to reflect in an innovative manner on the Bell Beaker implantation in the South of France. The starting point of the study showed that the majority of the clay resources are local and come from the exploitation of a natural mixture of Rhone alluvions and lateral carbonate contributions. Apart from the mixed beaker, ceramics made from these clays were all tempered only with crushed calcite. The places of origin of the clays of the other two families, represented by only one vase each, are farther away, without it being possible to be more specific. Moreover, these two vases were the only ones that were tempered with grog. These data have made it possible to establish that geometric dotted style productions present in Avignon included various practices implemented in different geographical areas; one of the two foreign vases, moreover, has a paste that also contained crushed carbonates. On the other hand, in Avignon, on the basis of the samples analysed, the practice of tempering ceramics with grog does not seem to have existed; only the presence of crushed calcite could be identified for locally manufactured geometric dotted productions.

Within the series analysed, two types of temper were therefore used. The first, the most commonly used, corresponds to the crushed calcites present in the majority of pastes of decorated Bell Beaker pottery as well as in that of the pre-oral rope vase. Grog, which is the second temper, was introduced only in the pastes of the two foreign vases. In western Provence, crushed calcite is a temper which has already been identified in the pastes of decorated Bell Beakers vases from Les Calades (Orgon, Bouches-du-Rhône) and Les Barres (Eyguières, Bouches-du-Rhône), which are stylistically similar to those of Avignon, as well as in the pastes of Rhodano-Provençal decorated production from Collet-Redon (Martigues, Bouches-du-Rhône). To the west of the Rhone, in eastern Languedoc, very few Bell Beaker ceramics of the same style as those from Avignon are known and none have been analysed. Very few petrographic analyses were carried out on beakers attributed to J. Guilaine’s early phase, which proved to be almost never tempered with crushed calcite. On the other hand, the use of crushed calcite is largely attested for the ceramics of the Rhodano-Provençal group and the associated common ceramics in eastern Languedoc.

Since the Early Neolithic, crushed calcite has been traditionally used as a temper by potters in Provence and in eastern Languedoc. A large part of the Late Neolithic productions from Provence contain it and it is omnipresent in the pottery from the other Late Neolithic sites (Fontbouisse) in the Gard and the east of the Herault département.

The Avignon series has especially brought results for local vases of the geometric dotted style that have been confronted with the data previously obtained on other corpuses from Provence and Languedoc.

The new data acquired over the past few years allow two hypotheses to be envisaged. The first is to consider that the Bell Beaker potters originally corresponded to individuals from the same substratum, possessing a unique and common (but no longer determinable) practice, or they may also correspond to geographically heterogeneous individuals, without a single practice concerning the use or otherwise of tempers. In both cases, when settling in, they would have adopted the cultural practices of the populations of the local substratum. This would involve the borrowing of clay preparation recipes from indigenous populations. The second hypothesis resumes a proposition already presented in a previous study. In order to explain the characteristics of the geometric dotted production of phase 2 of J. Guilaine’s model, we proposed that at the end of the Neolithic, after an initial phase (phase 1) of short duration (probably some decades) during which the beakers were made locally by individuals foreign to the substratum, part of the indigenous population copied these beakers and then quickly created new shapes while diversifying the decoration.

The first proposal implies that the Bell Beaker potters who wished to establish themselves systematically adopted local cultural practices concerning the temper for the preparation of the ceramic pastes of each population of the substratum present on the territory where they settled. This hypothesis is quite plausible because several ethno-archaeological studies have shown that potters displaced into another human group can modify their know-how under the influence of the traditions of the populations in which they settle. Such borrowing may concern the temper but also shaping or decoration techniques. The nature and duration of contacts appear to be important. The reasons given for such borrowing are the ease and speed of shaping as well as consumer demand. These types of borrowing do not seem to be retained in the case of the Bell Beaker potters since the forms and the decorations are radically different from those of the indigenous productions. Nevertheless, the adoption of a locally used temper may have helped spread the beakers among indigenous populations.

The second hypothesis that we have chosen to privilege and develop is the opposite of the previous one. It is not the newcomers who borrow traditions but the natives who adopt a new ceramic form, the beaker. This hypothesis explains why the Bell Beaker productions are identical to the ceramics of the substratum, from the point of view of the use of temper, in particular crushed calcite, since they were made by the same local cultural groups. Moreover, this hypothesis also explains the systematic presence on the Bell Beaker sites of substratum ceramics other than through acquisitions made with indigenous populations. Nevertheless, a specificity can differentiate Bell Beaker productions from other locally made ceramics because, on the left bank of the Rhone, at Calades and Barres, but also in Avignon in the case of the two allochtonous ceramics, grog is found, in association with the crushed calcite still used in most productions.

This hypothesis thus makes it possible to propose a renewed reading of modalities concerning the borrowing, reinterpretation and development of Bell Beaker ceramics by some of the potters of the indigenous cultures of the Late Neolithic following the first contacts with individuals foreign to the substratum. In particular, they highlight the variability of the behaviour of potters who use this pottery with, in particular, the manufacture of stylistically similar ceramics in different places according to distinct practices, but which are largely identical with local traditions. This situation can be observed in current traditional societies where borrowing mechanisms can be rapid, in the region of a few decades. Imitations can be the result of the arrival of a new potter with his own decors, which is very close to the situation envisaged for the Bell Beaker culture. On the other hand, the time required for the appearance of new ceramic decorations and forms is difficult to access by ethno-archaeology, but archaeological examples from lakeside sites which have yielded dilated stratigraphic sequences allow an order of magnitude of a few decades for this period.

Keywords : South of France, Late neolithic, Bell Beaker, ceramics, temper, raw material, borrowing.

En Gaule atlantique, 696 dépôts métalliques non funéraires du Bronze final et du premier âge du Fer ont été découverts en milieu terrestre, spécialement dans le Massif armoricain et ses marges. Ils matérialisent deux séquences bien distinctes et successives de pratiques de constitution des dépôts. Au Bronze final, les dépôts réunissent essentiellement des objets variés, entiers ou cassés, à divers stades de leur vie technique. Au premier âge du Fer, ils consistent en objets intacts dont la forme est celle d’une hache, mais qui sont inutilisables en tant qu’outil. L’hypothèse développée ici est que ces dépôts atlantiques fossilisent des formes d’échange et de thésaurisation de la richesse. Plus précisément, nous pensons que les deux séquences matérialisent deux conceptions différentes de la valeur et de l’usage du métal dans un registre économique prémonétaire : la première est fondée sur du bronze pesé en vrac, la seconde sur du bronze plutôt compté sous la forme d’instruments standardisés à cet effet. Ces échanges ayant pour support le métal ne recouvraient probablement qu’une partie des sphères de transaction de l’époque. Sans doute matérialisent-ils le paiement d’obligations sociales particulières.

Mots-clés : Bronze final et premier âge du Fer atlantiques, Gaule, dépôt métallique, « hache » à douille armoricaine, économies prémonétaires.

In Atlantic Gaul, 346 non-funerary metal hoards from the Late Bronze Age and 350 from the Early Iron Age have been discovered on land, especially in the Armorican Massif and its fringes. The hoards were constituted, respectively, of 18,666 and 40,638 metallic elements. A more precise chronological study has helped to envisage the main variations in intensity of hoarding practices from one period to another. The two periods of maximum hoarding were during the 9th c. BC (BFa 3 récent or Vénat horizon), and the 6th c. BC (1er Fer a. 2 and 3, or Trelly and Saint-James horizons). Independent of these variations in the number of hoards and the amount of hoarded objects, we can observe two very distinct and sequential practices relating to the formation of these hoards. During the Late Bronze Age these hoards were made up of various objects, intact or broken, in diverse stages of usage. The more we advance in the Late Bronze Age, the more the objects are broken and disseminated from their place of origin. During the Early Iron Age, the manner in which objects are hoarded is radically different, although less so in the Armorican Massif. The Early Iron Age hoards consist of intact objects all belonging to the same functional category. These objects are shaped like axe blades, but are not usable as tools or weapons. Throughout the Early Iron Age, these pseudo-axe blades decrease in size, to the point that they become miniatures.

The aim of this paper is to answer two main questions: How do we explain the great differences between the composition of hoards from the Late Bronze Age and those of the Early Iron Age, even though the objects found are made of the same metal (copper alloy)? What was the biography of the objects before and during their deposit in their respective hoards, and what does that tell us about their protohistoric economy?

In my attempts to answer these questions, I investigated the ‘chaîne opératoire’ with regard to the manipulation of objects. I also used the concepts of the use value and of the exchange value of the metal distinguished by 19th century economists.

During the Late Bronze Age, most objects were used before hoarding. The biggest ones were systematically broken after having been used, therefore losing their functional value. At the beginning of the Late Bronze Age, the fragments of large objects found in the hoards can often be pieced together into the original form. This is indicative of a short time between fragmentation and deposition. However, as the Late Bronze Age progressed, fragments of the same object were no longer found together. This is evidence of the dispersion of the fragments long before their inclusion in a given hoard. This shows that the handling of objects between their being broken and being deposited in the hoard became more complex and/or took place over a greater period of time. I interpreted this phenomenon as the circulation of metal, in the form of fragments or else in smaller, intact objects that had lost their functionality.

During the Early Iron Age, the pseudo-axes are hoarded in an unfinished state. Other clues (miniature size, improper alloy for tools, flimsy cutting edge) prove that they did not have any use value. However, these objects were not made to be hoarded. Even though they have no trace of being worn down the way a tool or weapon would be, they nevertheless show signs of wear and tear in the sense that they have passed through many hands. In some cases, the presence of ties (metal wire, strings) joining the axes to one another would have facilitated their transport. The diversity of typological assemblages of the pseudo-axes in most hoards also supports the idea of long-term circulation.

The hypothesis developed here is that the pseudo-axes of the Early Iron Age, having no functional value, were circulated uniquely for their exchange value. Thus, they would have served as a pre-monetary currency. This hypothesis is not new, but it was put aside for several decades. What is important to note is that these were the first objects conceived to be used solely as currency in Atlantic Europe. Other small objects used as currency are known in the world around this time, especially in China, and demonstrates that this was not an isolated phenomenon. Since the hoards of the Early Iron Age fossilized a pre-monetary exchange system, we could imagine that the hoards from the Atlantic Late Bronze Age also gather goods used essentially as currency and a store of pre-monetary wealth. To support this hypothesis, I note that the heaviest and longest objects were shattered and compacted over a long period of time, and the fragments were widely dispersed from their presumed place of production. My interpretation is that bronze objects were reduced to their exchange value, that is to say their metallic weight, serving as a currency for pre-monetary exchanges. The shattering of large objects would be first linked to successive transactions and to the need to have metallic pieces that were more easily manipulated. The identification of beams and weights used in scales discovered in Late Bronze Age settlements in Atlantic Gaul shows that it was possible to weigh metallic objects or pieces precisely. Historical and ethnographical parallels show the use of weighed metal as currency in pre-monetary exchanges.

To sum up, the Late Bronze Age and Early Iron Age Atlantic hoards represent two different, successive sequences of forms of exchange and hoarding of wealth. Specifically, I believe that the two sequences demonstrate two different concepts of value and the use of metal in a pre-monetary economic setting: the first is based on bronze weighed in bulk, the second on bronze tallied in the form of instruments standardized for exchanges. Such trade, with metal as an underlying currency, probably only represents a portion of the exchange systems at that time. In all likelihood, they demonstrate the payment of particular social obligations. Finally, the ideas developed in this paper do not contradict the hypothesis that hoards could reflect ritualistic practices, because hoards are very numerous and sometimes discovered in particular or remote places, such as marshlands. If that were the case, hoards would be the fossilization of exchanges during a ritual.

Keywords: Atlantic Late Bronze Age, Atlantic Early Iron Age, Gaul, hoard, Armorican axe, premonetary exchanges and hoarding.

La troisième livraison du Bulletin de la Société préhistorique française vient de paraître.