(version papier) BSPF2024-2

Modélisation des interactions en Italie du Nord au premier âge du Fer : de la circulation de parures aux réseaux d'influences culturelles

Veronica Cicolani, Thomas Huet, Lorenzo Zamboni

Résumé : Dans les approches traditionnelles fondées sur les sources écrites, le plus souvent postérieures et reflétant un point de vue extérieur, les cultures archéologiques de l'Italie du Nord sont souvent définies comme étant des entités autonomes (peuples), caractérisées par des productions spécifiques et surtout délimitées par des frontières géographiques et culturelles clairement établies. Dans le cadre du projet ANR JCJC Itineris (https://itineris.huma-num.fr/), consacré à l'étude des pratiques artisanales en Italie nord-occidentale au premier âge du Fer, le réexamen critique de la documentation et l'analyse des nouvelles données permettent d'aborder sous un autre angle de vue les relations entre entités culturelles et territoriales, dépassant l'interprétation dichotomique et hiérarchique de centre-périphérie et les approches culturalistes. Il est alors possible d'interpréter les processus d'interaction entre entités dites marginales à travers une modélisation des données archéologiques. Cet article propose donc de réexaminer les rôles des populations décrites comme périphériques dans les réseaux d'interactions et de transferts technoculturels avec les grands « centres » : domaines culturels de Golasecca, ligure, étrusque et Este. Le corpus retenu rassemble plus de 2000 objets en alliage à base cuivre, essentiellement des parures et des déchets liés à leur fabrication, issus tant des habitats que des contextes funéraires, bien documentés et localisés dans le Piémont méridional (communautés indigènes de Ligurie intérieure) et en Émilie occidentale (communautés indigènes du secteur occidental de la plaine du Pô). Par l'application de modèles statistiques (programmation R et analyse des réseaux), l'objectif a été de visualiser et ordonner ces données, complexes et hétéroclites, afin de dégager une première typologie dynamique de réseaux et de sous-réseaux connectés, reliant entités périphériques et grands domaines culturels selon des géométries variables. Pour cela faire, le corpus, qui fait l'objet de publications et révisions récentes, a été structuré sur la base de descripteurs hiérarchiques et sa caractérisation (typologie, contexte et quantification) complétée par l'ajout d'une variable qualitative, ici nommée « style ». Ce critère, issu de l'étude morphostylistique de chaque objet, a pour rôle de traduire en termes factuels les goûts/adaptations locales de modèles et types suprarégionaux, permettant ainsi de modéliser les rapports d'influence culturelle entre communautés. La modélisation, fondée sur la théorie des graphes et l'application d'algorithmes de détection de communautés (ici edge.betweenness.community, du package "igraph" de R), permet d'évaluer le degré d'importance et de proximité entre les groupes par segmentation hiérarchiques des liens, aboutissant ainsi à l'identification d'ensembles cohérents (communautés de caractères). Appliquée pour chaque période, les résultats de cette modélisation sont ensuite représentés sur une carte pour faciliter la lecture spatiale et l'interprétation géographique des résultats obtenus. Cette démarche a mis en évidence le rôle majeur des fibules en tant qu'élément vestimentaire et ornement le plus adopté et adapté, tant sur le plan morphologique que stylistique et technique. Néanmoins, sa présence diffuse dans l'ensemble de l'Italie du Nord finit par masquer la reconnaissance et la circulation de certaines connexions plus spécifiques, soulevant le problème de l'usage de ce marqueur en tant qu'indicateur privilégié de mobilité ou d'identité culturelle. Si l'influence de Golasecca est attestée le long de toute la période ici considérée, des interactions plus localisées et centrées sur des classes spécifiques d'objets dévoilent un jeu subtil d'influences qui traduit des relations à géométrie variable, jamais univoques et unidirectionnelles, et où productions locales, adaptations, créativité et importations décrivent des réalités sociales en mouvement et d'une complexité croissante.

Ainsi, cette étude comparative des types et des styles de vêtements anciens et leur mobilité permet d'explorer plus en profondeur la nature de leurs interactions, tout en clarifiant, par la même occasion, l'impact de leur diffusion à courte et moyenne échelle en Italie du Nord.

Mots-clés : Âge du Fer, Italie du Nord, réseaux, modélisation des interactions, identités matérielles, parures, influence culturelle, périphéries.

Abstract: In traditional approaches based on written sources, most often posthumous and reflecting an external point of view (etic), Italian archaeological cultures are often defined as autonomous entities (peoples), characterized by specific productions and sharing clearly established geographical boundaries. Within the framework of the ANRJC Itineris (https://itineris.huma-num.fr/), focusing on the study of craft practices in north-western Italy during the first Iron Age, the critical re-examination of the documentation and the analysis of new data allows us to approach the interactions between cultural entities from another angle, going beyond the dichotomous and hierarchical interpretation of centre-periphery, and thus to be able to interpret the phenomena of cultural interactions between so-called marginal entities through a modelling of the archaeological data. This paper therefore proposes to re-examine the roles of italic populations described as peripheral within the networks of interaction and technocultural transfers that developed in the early Iron Age in northwestern Italy. By applying statistical and spatial data modelling (R program and Graph theory), it is possible to visualise and order complex and heterogeneous data, such as archaeological records, as well as to identify a dynamic typology of connected networks and sub-network. The selected data set includes more than 2000 copper-based alloy objects, mainly ornaments and the waste products associated with their production, from both settlements and funerary contexts, well documented and located in southern Piedmont (indigenous communities in Inner Liguria) and western Emilia (indigenous communities in the western sector of the Po valley). By applying statistical models (R programming and network analysis), the aim was to be able to examine and classify this complex and heterogeneous data, in order to identify a first dynamic typology of connected networks and sub-networks, linking neighbouring entities and major cultural areas according to variable geometries. This data set, which has recently been published and revised, has been structured on the basis of hierarchical descriptors, and its characterization (typology, context and quantification) has been completed by adding a qualitative variable, here called 'style'. The role of this factor, based on the morphostylistic study of each object, is to translate into factual terms the local tastes/adaptations of supra-regional models and types, thus making it possible to model the relationships of cultural influence between communities. The modeling, based on graph theory and the application of community detection algorithms (here edge.betweenness.community, from 'igraph' R package) enables the degree of importance and proximity between groups to be assessed by hierarchical segmentation of links, in order to identify coherent clusters. By breaking the links between sub-groups of nodes, we can identify groups of nodes that share more with each other than they share with the other nodes in the graph (communities of attributes). Applied for each period, this segmentation is represented on a map by assigning the color of the groups (clusters) to the sites, thus facilitating spatial reading and geographical interpretation of the results. The visual analysis of the graph, or the statistical analysis of the network, is in fact both a model and a powerful formalism, but it is no substitute for the archaeological and historical interpretation of the data. This approach has highlighted the major role played by fibulae as the most widely adopted and adapted item of clothing and ornament, in terms of morphology, style and technique. Nevertheless, their widespread occurrence across the whole of northern Italy ultimately masked the identification and circulation of more specific connections, raising the issue of the use of this marker as a key indicator of mobility or cultural identity. While the influence of Golasecca is attested throughout the period under consideration here, more localised interactions based on specific classes of objects reveal a sophisticated interplay of influences that reflects relationships with variable geometry, never unambiguous or unidirectional, and where local production, adaptation, creativity and imports describe social realities on the move and of growing complexity.

Starting from this point of view, the comparative study of clothing item types and styles has allowed us to explore more deeply their interactions, clarifying, by the same way, the impact of their dissemination in a medium scale: Northern Italy. Highlighting these dynamic, non-exclusive links based on creativity and adaptation, rather than on predefined geographical or cultural boundaries, illustrates how these boundaries are socially negotiated and dynamically recomposed over time, defying the traditional, unidirectional perception of a hierarchical structure of dependence between center and periphery.

The goal is to produce operational distinctions between local productions, imports, imitations, and local transformations. By modelling the fashion tradition through the analysis of shared clothing items, this paper would like to propose a critical review of the markers of cultural identity, mobility, interaction traditionally used looking for alternative theoretical interpretation and social narrative of the role of indigenous communities in the North Italic trade et craft activities.

Keywords: First Iron Age, Northern Italy, networks, interaction modeling, material cultures, ornaments, buffer zones, cultural influence.

Architectures domestiques ???et activités économiques au Néolithique final 

L???habitat de Rebreuve-Ranchicourt (Pas-de-Calais)

Élisabeth Panloups, Emmanuelle Bonnaire, Justine Cadart, Kai Fechner, ???Élodie Lecher, Cécile Monchablon, Marylise Onfray, Thibaud Paulmier, Aurélie Salavert

Résumé : Nouvelle découverte d'habitat structuré dans la sphère d'influence du complexe culturel de Deûle-Escaut, le site de Rebreuve-Ranchicourt a livré les vestiges de trois bâtiments diachroniques, dont une partie de la chronologie a pu être retracée. La grande particularité de cette occupation est la découverte d'un mobilier riche et varié qui a permis d'engager une réflexion sur la spatialisation des activités au sein et autour de la maison. Cet habitat vient également compléter les informations pour la chronologie interne du Néolithique final. Ainsi le site de Rebreuve-Ranchicourt semble pouvoir se placer dans une étape moyenne du Néolithique final, tel qu'il est actuellement défini, notamment par l'absence d'éléments campaniformes. Ces résultats se heurtent toutefois à des dates chronométriques plus tardives, concordant avec une étape terminale de la fin du Néolithique. D'autres questions, d'ordre géographique, se posent sur l'homogénéité culturelle du Deûle-Escaut. A l'échelle du Nord-Pas-de-Calais, les sites livrant un ou plusieurs plans de bâtiment apparaissent groupés sur deux zones distinctes. Les principaux habitats Deûle-Escaut sont regroupés, comme leur nom l'indique, le long des vallées de la Deûle et de l'Escaut, incluant les vallées secondaires de la Scarpe et de la Sensée. Les sites situés au nord des vallées de la Lys de l'Aa marquent une seconde concentration d'habitats située à environ 40 km au nord-est des sites les plus proches de la vallée de Deûle. La découverte de Rebreuve-Ranchicourt occupe une position géographique un peu particulière, à mi-chemin de ces deux pôles. La comparaison de l'ensemble de ces habitats, localisés sur un très large périmètre, laisse entrevoir des différences au niveau de la culture matérielle, comme les décors céramiques ou la part des matériaux siliceux exogènes, questionnant sur les sphères d'influences et les contacts entre les communautés de la fin du Néolithique.

Mots-clés : France du Nord, Néolithique final, chronologie, habitat, bâtiment, micromorphologie, céramique, lithique, macrolithique, carpologie, anthracologie, phosphore.

Abstract: At the scale of Nord-Pas-de-Calais, the main habitats of Deûle-Escaut are grouped together, as their name suggests, along the valleys of the Deûle and the Scheldt, including the secondary valleys of the Scarpe and the Sensée. This sector includes more than a dozen sites. To the north of the Lys basin and at the junction with that of the Aa, three other sites mark a second concentration of habitats located about 40 km northeast of the Deûle valley sites.

The discovery of Rebreuve-Ranchicourt occupies a somewhat particular geographical position, halfway between these two poles. This occupation includes three diachronic buildings, to which two pits can potentially be attached. The multidisciplinary approach put in place has made it possible to address several aspects of this occupation, from the construction of the buildings to their abandonment, focusing on the domestic, agricultural, and craft activities that took place inside and around the dwellings.

The two large buildings of Rebreuve-Ranchicourt find elements of easy morphometric comparison in the Deûle-Escaut cultural sphere, with a certain homogeneity in the length of the buildings, generally between 17 and 26 meters long. Building 1000, of smaller size, is rarer.

One of the major interests of the Rebreuve-Ranchicourt habitat site is its building 200. The construction methods are unusual, with oversized foundations that remain difficult to interpret. A reconstruction phase of part of the building could be considered, but has not been highlighted through the various stratigraphic and micromorphological studies. Similarly, a simple abandonment of the building does not explain the impressive quantities of discovered artifacts and in such good condition. In fact, this material data is rare in dwellings. Generally, it is not the foundation structures that yield artifacts, but the pits and occupation levels associated with the houses. For us, the field data and studies conducted on the site converge towards the highlighting of a complete destruction of the building, probably accompanied by a recovery of timber. The homogeneity of the archaeological material and the absence of alterations testify to conditions of rapid burial compatible with a deliberate leveling action. One plausible explanation for this leveling may be related to the erection of a new building nearby, building 700.

Thanks to the destruction of this building, many traces of the daily life of its occupants have been preserved over several millennia. While the entire lithic corpus fits perfectly into the known acquisition, management, and production patterns of flint tools for the Deûle-Escaut cultural group, the scarcity of finished products in exogenous materials raises questions about its integration into networks of circulation of highly valued objects. The discovery of a Bartonian flint blade during the nearby excavation of Houdain could indicate an insertion of these populations into networks of object circulation that may be broader than suggested by the assemblage from the buildings. The general composition of the series and the macrolithic tooling is comparable to the productions of sites in the Deûle-Escaut group with building plans. Quantitatively, it is a small series, characterized by fragmentation and significant recovery of supports. The types of tools are also in line with the regional Late Neolithic assemblages: hammerstones, burins, and intermediate pieces. While the cup-shaped grinding stone does not stand out in the Late Neolithic productions, this tool is less frequently encountered than rimmed grinding stones. It raises questions about the possible significance of this morphological difference: functional, chronological, or other. The ceramic assemblage finds its best comparisons with sites located further north in the Lys valley, with the same decorative registers of smooth or digitate cords placed under the rim or in rupture of the body. These decorations exist in the Deûle valley corpus, but in a more discreet manner. The absence of short-necked and high-keeled vessels at Rebreuve-Ranchicourt, and generally in the sites of the Lys valley, constitutes another notable difference with the eastern Deûle-Escaut occupations. The presence of decorated ceramics and the absence of bell-beaker pottery sherds invite placing this corpus in the middle phase of the Deûle-Escaut cultural complex.

Thus, the cross-referencing of the furniture studies agrees to integrate the Rebreuve-Ranchicourt site within the Deûle-Escaut cultural complex. The sites located north of the Lys valley find comparative parallels sometimes more relevant with the Rebreuve-Ranchicourt occupation than the sites located to the east. Furniture studies have also focused on the spatialization of activities carried out inside and outside of this dwelling, with debatable results. Projecting a functional determination near the place of discovery of the movable remains is tempting, but the exploitation of the results obtained lacks convincing parallels for the interior of dwellings. Furthermore, it remains difficult to clearly attribute the furniture to the occupation of the building itself. Finally, the characterization of actions laden with meaning, such as breaking a complete millstone and placing it in a former post hole, remains difficult to grasp and could only be interpreted from a functional point of view.

This habitat also complements the information for the internal chronology of the Late Neolithic. Thus, the Rebreuve-Ranchicourt site seems to be able to be placed in a middle stage of the Late Neolithic, as it is currently defined, notably by the absence of bell-beaker elements. However, these results are confronted with later chronometric dates, coinciding with a terminal stage of the Late Neolithic.

Keywords: Northern France, Late Neolithic, chronology, habitation/dwelling site, building, micromorphology, ceramics, lithics, macrolithics, carpology, anthracology, phosphorus.

Des croissants au pays des « livres de beurre » à la fin du Néolithique

Véronique Brunet, Laure-Anne Millet-Richard, Juliette Durand, Renaud Gosselin, Romana Blaser, avec la collaboration d'Ève Boitard-Bidaut

Résumé : La découverte récente en contexte préventif de croissants en silex en Île-de-France, en dehors de la Touraine où ils ont été identifiés la première fois, est l'occasion de s'interroger sur cet outil si particulier et rare. L'objectif de notre recherche était de recenser cet objet dans la bibliographie à l'échelle nationale et hors des frontières de l'Hexagone, d'en déterminer le nombre, les contextes de découverte, sa fonction et la place qu'il occupe au sein des productions de la fin du Néolithique.

Les croissants sont des objets plutôt rares, une quarantaine est connue à ce jour dans toute la France. On les trouve essentiellement dans deux régions, en Touraine et en Île-de-France. On les rencontre le plus souvent dans les habitats, mais ils apparaissent également en contexte minier, d'atelier de taille du silex et même sépulcral.

Ils se caractérisent par une morphologie particulière avec l'aménagement de deux pointes. Peu de dissemblances sont lisibles dans les dimensions de l'outil, mais l'aménagement des bords diffère d'une région à l'autre, il est systématique en Touraine à l'inverse des exemplaires franciliens. Les pointes des croissants tourangeaux sont bien dégagées, tandis que les franciliennes apparaissent avec une convexité plus accentuée.

De nouvelles recherches sur leur fonction ont permis de compléter les premières analyses tracéologiques des croissants tourangeaux qui n'avaient pas été concluantes en raison des états de surface. Les résultats tendent à montrer que les traces d'usage ne sont généralement pas observées dans l'encoche à l'exception de l'une d'entre elles qui atteste d'un contact avec une matière dure d'origine animale ou végétale. Le bilan de l'analyse fonctionnelle est faible mais permet de s'interroger sur le traitement des pointes, dont un certain nombre s'avère cassé et/ou ébréché ce qui pourrait être compatible avec un contact plus ou moins violent avec une matière dure animale ou végétale.

Mots-clés : croissant, Néolithique récent, Néolithique final, Le Grand-Pressigny, Touraine, Île-de-France.

Abstract: A recent discovery of flint croissants in a preventive excavation in Île-de-France, outside the area where they were first discovered in Touraine, provided an opportunity to investigate this rather particular artefact. A search was made in the literature published both within and outside France, in order to determine the numbers of croissants and their find contexts, as well as to examine their function and role in later Neolithic flint production.

Croissants are quite rare objects, as about forty are currently recorded for the whole of France. Occurring in two regions, Touraine and Île-de-France, they are most often found in settlements, but also appear in mining, workshop and even burial contexts.

Of the thirty or so croissants known in the Grand-Pressigny area, twenty-two were discovered in the commune of Abilly (Indre-et-Loire), thirteen of which come from excavations on the Foulon site and one other from La Madelone, about 1 km to the north-west. In the Île-de-France region, croissants are still few and far between, with six examples and only one per occupation. Two come from sites in the north of the Seine-et-Marne département, located in the river Marne valley and less than 5 km apart. There is one example in the Loiret département, another from the Val d???Oise and lastly there are two from the Yvelines.

This tool is generally made from local raw materials. The morphology is particular, due to the creation of two tangs. Although there is little difference in tool dimensions between the two regions, the shaping of the edges differs from one region to another. The edges are systematically retouched in Touraine, while this is not always the case in Île-de-France. The Touraine croissants have clearly defined tangs, whereas the Île-de-France croissants appear less concave.

While selected flakes may have been specifically produced by hard direct percussion, some flakes from the Grand-Pressigny area are derived from the preparation or maintenance of "pound of butter" (livre de beurre) cores. The croissants were most often produced along the axis of the flakes, with retouched tangs on each of the two lateral ends. The retouch was done by hard hammer. The lengths croissants vary from 32 to 65 mm in Touraine and from 25 to 90 mm in Île-de-France. In Touraine widths are between 29 and 93 mm and in Île-de-France between 33 and 146 mm. The Touraine croissants are mostly between 10 and 19 mm thick, whereas in Île-de-France thickness ranges from 6 to 23 mm.

While the croissants from Touraine and western Île-de-France can be grouped in the same family because of their similarity in shape, the eastern Île-de-France croissants are more difficult to include as they deviate from the norm. This suggests that they are probably different objects. In fact, the morphology of the croissant from Coupvray in Seine-et-Marne has little in common with the broad family of croissants. Nevertheless, it does show similarities in function with the Touraine examples, as indicated by similar use-ware flaking observed on the barbs of the tool.

The initial use-wear analyses of the Touraine croissants were not conclusive, due to the surface conditions. The new study undertaken here generally shows that no traces of use can be observed in the concavity, with one exception attesting to contact with a hard material of animal or vegetable origin. The result of the functional analysis is thus limited but does raise questions about the treatment of the tangs, some of which are broken and/or chipped, which could be compatible with more or less violent contact with a hard animal or plant material.

In the Île-de-France region, croissants are frequently associated with flaked axes. Other associated tools vary greatly from one area to another. In Touraine, fourteen different categories of tools are recorded, mainly daggers, micro-denticulates, notched side-scrapers, retouched blades, scrapers etc.

In Île-de-France, the tool categories are more numerous, with up to twenty different categories in the west and a dozen in the east. This is probably due to the context in which the tools were found, with the composition of the associated tools differing between workshops and settlements, with a majority of bifacial pieces with active cutting edges in the former and scrapers in the latter.

The croissants from Île-de-France may possibly be older than their counterparts in Touraine, since the latter are associated with occupations dated to the Final Neolithic. In Île-de-France, croissants are found in the Late Neolithic and perhaps even as early as the Middle Neolithic at Adon (Loiret).

Since only two regions have so far produced evidence for croissants, in different proportions, this raises the question of the origin of the tool type. Was it mainly produced in Touraine, spreading to Île-de-France with the Pressignian phenomenon, or the reverse from east to west?

Keywords: croissants, late Neolithic, final Neolithic, Grand-Pressigny, Touraine, Île-de-France.

Une sépulture du Néolithique moyen liée à la pratique de la crémation sur le site de Piechegu (Bellegarde, Gard)

Aurore Schmitt, Marion Gasnier, Marilyne Bovagne, Sabine Negroni, Adrien Reggio, Samuel van Willigen et la collaboration de Blandine Lecomte-Schmitt

Résumé : Le site de Piechegu (Bellegarde, Gard) fouillé par l'Inrap en 2016, a livré, entre autres, 140 structures s'échelonnant entre le VIe et le IVe millénaire avant notre ère. Parmi les vestiges funéraires néolithiques, qui témoignent surtout de la pratique de l'inhumation, une sépulture secondaire à crémation a été découverte ainsi qu'une trentaine de fosses datées du Néolithique moyen 2. La structure 3614, datée entre 4250 et 4000 avant notre ère, contenait des restes humains calcinés accompagnés d'un mobilier constitué de onze récipients en céramique, neuf armatures, deux lames de hache et un objet en bois. Si les différents types d'objets représentés ne se démarquent pas de ce qui est connu par ailleurs dans la région, leur nombre est plus surprenant et fait de FS3614 une des plus riches sépultures du Néolithique moyen du Midi de la France, tous types de traitement des défunts confondus. De plus, la disposition des objets dans la sépulture n'est pas aléatoire et témoigne d'une véritable mise en scène impliquant des tris sur le bûcher.

La pratique de la crémation est attestée dans le Néolithique moyen méditerranéen depuis les découvertes du Vallon de Gaude à Manosque et du Camp del Ginebre à Caramany dans les années 1990. Avec une trentaine d'occurrences, ce type de traitement des corps peut être considéré actuellement comme l'une des composantes du système funéraire du Néolithique moyen méditerranéen. Cela nous conduit à poser la question du statut des défunts qui ont bénéficié de ce traitement à une période où peu de morts bénéficient d'un enfouissement.

Mots-clés : crémation, Néolithique moyen, midi de la France, céramique, lithique, objet en bois, restes osseux brûlés.

Abstract: The Piechegu site (Bellegarde, Gard, fig. 1) excavated by Inrap in 2016 yielded a structure (3614) located in a group of fifteen features dating from the late 5th millennium (fig. 3) and attesting the practice of cremation. The circular pit had a diameter of one meter, a bowl-shaped profile and a preserved depth of 30 cm (fig. 4). An oval, wooden object was placed on the bottom of the pit (fig. 4, 12 and 13), with a laddle in a bowl with an internally thickened, everted rim (fig. 4, 5 nos. 3 and 6, fig. 13). The remains of the pyre containing some of the deceased's bones, as well as arrowheads and a few ceramic fragments, were then thrown into the pit, forming a layer of fifteen centimeters. A level of sediment a few centimeters, devoid of archaeological material, seems to have sealed off this first deposit. Most of the archaeological material (arrowheads, axe blades and ceramics) was discovered on top of this layer (fig. 4 and 14). The pit was then backfilled by about ten centimeters of carbonaceous sediment.

Only one adult-sized represented by 487 g of mostly calcinated white bones, was identified. The burnt remains of the individual were not entirely removed from the crematory structure. The ceramics consist of a minimum of eleven vessels : two vases with S-shaped profile (fig. 5, nos. 1 and 2), two vases with spherical or elliptical bodies (fig. 7, nos. 2 and 3), two plates with decorated rims (fig. 6, nos. 1 and 2), the lower part of a vase support with a decorated base (fig. 7, no. 1), a vase with a low inflection, a straight, closed upper part and a perforated cordon (fig. 5, no. 5), a bowl with an internally thickened, everted rim (fig. 5, no. 3) and, finally, fragments of two ladles (fig. 5, nos. 4 and 5). At least some of the ceramics show evidences of high-temperature exposure: flaking surfaces and eroded breaks. The series as a whole is characterized by high quality production, showing the craftsmanship of fine and robust ware. Nine sharp flint arrowheads and two axe blades made of tenacious rock (fig. 11) were uncovered in the fill of structure 3614. The geometric bitroncatures have been burnt intensely, causing very severe weathering. The small wooden vessel is made of deciduous oak. Such vessels are relatively rare for the Neolithic period, due to the special conditions required for their preservation.

The radiocarbon date of a charcoal (Poz-90710 5310 ± 40 BP) places the feature 3614 to the last third of the 5th millennium BC. The ceramic corpus is comparable to the range of forms and decorations attested in several Middle Neolithic Mediterranean assemblages. They correspond to a cultural group described by different authors as Early Chassean (see, for example, Georjon and Léa, 2013) or Chassey-type Middle Neolithic (van Willigen et al., 2020).

Considered an anecdotal funerary practice since the 1990s, after the discovery of the Caramany and Vallon de Gaude sites, cremation is now attested by 33 structures on at least ten sites. For the most part, these are secondary cremation deposits reserved to adult-sized individuals. These deposits are not associated with burials and, like burials, are either isolated (as at Piechegu) or form groups of around ten structures. Bones are highly fragmented and are partially collected at the pyree before being transferred and deposited in a receptacle (pit or, in one case, ossuary vessel). In most cases, secondary cremation deposits are accompanied by objects (sherds, arrowhead, axe blades, fragments of tools made in animal bones) that went through high temperatures. The practice of cremation belongs therefore to the funerary system of southern France. Structure 3614 shares several caracteristics with other cremations: the combination of arrowhead and axe blades, the partial removal of the burnt human remains and the deposition of burnt objects. However, the structure 3614 has several peculiarities. The deposition of certain objects is not random, and clearly shows that the burnt objects were sorted and placed in the grave in a precise order. Moreover, with at least eleven ceramic vessels, one wooden vessel, nine arrowhead and two axe blades, this feature at Piechegu is one of the richest Middle Neolithic burials in the region.

Keywords: cremation, Middle Neolithic, South of France, ceramic, lithic, wooden object, burned human remains.

SIGMENT : une base de données bibliographique critique pour estimer la répartition géographique des unités archéologiques lithiques du Gravettien moyen et récent en France

Anaïs Vignoles, Arnaud Caillo, William E. Banks, Laurent Klaric

Résumé : Cet article présente la base de données SIGMENT (Sites du Gravettien Moyen et récent), dont l'objectif est d'estimer de manière fiable et reproductible la distribution géographique des unités archéologiques lithiques du Gravettien moyen et récent en France et dans ses marges. Nous avons collecté des données sur 254 collections archéologiques provenant de 166 gisements situés en France, en Belgique, en Espagne, en Italie et en Allemagne, à partir de 196 références bibliographiques, ainsi que sur base de l'examen de plusieurs collections. Nous avons ensuite établi une grille d'analyse qualitative permettant d'évaluer la fiabilité de la présence des unités archéologiques étudiées dans chaque localité. Le résultat de cette évaluation montre qu'une partie importante des occurrences n'est pas totalement fiable, et nous conseillons donc de ne pas les intégrer dans des modèles qui se focalisent sur la distribution géographique de ces unités archéologiques. A partir de l'étude historiographique et archéologique de plusieurs collections du sud-ouest de la France (Le Fourneau du Diable, le Facteur, la Rochette, Laussel, les Jambes, les Artigaux, la Roque Saint-Christophe, Lespaux, Tourtoirac, Grotte XVI, Combe-Saunière, le Flageolet I, Solvieux), nous discutons ensuite des facteurs qui ont pu influencer la sur- ou sous-estimation de la présence de ces unités archéologiques lithiques à partir d'un recensement strictement bibliographique, en particulier la variabilité des approches analytiques, des méthodes de fouille et des méthodes de conservation ainsi que la façon dont les ensembles archéologiques sont décrits dans la littérature. La base de données critique qui résulte de ce travail pourra être utilisée par des chercheur·euse·s souhaitant explorer la répartition spatiale d'éléments typo-technologiques recensés pour les unités archéologiques du Gravettien moyen et récent ou souhaitant implémenter la structuration de cette base de données à leurs propres inventaires de sites.

Mots-clés : Gravettien, base de données, industrie lithique, recensement, histoire des fouilles, Paléolithique supérieur.

Abstract: The SIGMENT database was created in the framework of A. Vignoles' doctoral thesis, which examined ecological dynamics with respect to the evolution of lithic archaeological units during the Middle and Recent Gravettian in France. This work combines typo-technological examinations of archaeological assemblages with Eco-cultural Niche Modeling (ECNM) to test the potential relationships between cultural and environmental changes in this chrono-geographic framework. ECNM employs a variety of methods aimed at estimating and comparing the eco-cultural niches of different archaeological units (i.e. the range of environments occupied by populations using said units). In practice, it consists of defining a mathematical relationship between the geographic distribution of archaeological sites where the archaeological unit is observed and environmental variables that characterize the conditions in which the populations operated. In so doing, it is paramount to ensure the quality of input data in order to avoid situations of "Garbage-in, garbage-out" in which models based on low quality data will inevitably have less interpretive power.

The evaluation of occurrence data quality was conducted via a critical review of the literature organized in a relational database: SIGMENT. The aims of this database are to: 1) produce an exhaustive list of sites that can be attributed to the Middle and Recent Gravettian in France (and its margins to a lesser extent); and 2) evaluate the reliability of the occurrence of the archaeological units at each locality via the combination of bibliographic descriptions and direct personal observations. This database allows for a reliable estimation of the geographic distributions of Middle and Recent Gravettian archaeological units that can be employed in large-scale studies.

The SIGMENT database is comprised of 21 tables centered around two principal tables: 1) "site", which contains information about the archaeological locality; and 2) "ensemble" (translated as assemblage), related to a collection of artefacts contained stratigraphically and/or spatially within the site. Each bibliographic reference was scrutinized to extract a variety of information at the scale of the site or the assemblage???examples being taphonomic integrity, presence of other material culture elements and other archaeological cultural complexes, the dates of excavation and diverse data concerning lithic industries (mainly typo-technological data). Finally, we evaluated the presence of each archaeological unit at each locality by applying a qualitative reliability grid comprised of five levels.

To ensure the exhaustiveness of this survey, we systematically analyzed the highest possible number of references for each site and made an effort to review the princeps publications and primary data in order to avoid posterior interpretations by subsequent authors. We started the survey with recent critical bibliographies that served to identify older references. In total, 196 references were analyzed. Their publication dates vary between 1910 and 2023, but most sources were published after the 1950s. It covers 166 sites and 254 assemblages, mainly located in France. To enhance the reliability of some sites, we personally reviewed or examined nearly 20 collections from sites located in South-western France: Roque Saint-Christophe, Facteur rock-shelter, Fourneau du Diable, Tourtoirac, Flageolet I, Grand Abri of Laussel, Grotte XVI, La Rochette, Combe-Saunière, Solvieux, Lespaux rock-shelter, and Les Artigaux.

As a result of these examinations, we retained 75 sites for the Noaillian, 26 sites for the Rayssian and 15 sites for the Recent Gravettian. The geographic distribution of each of these curated data sets is slightly different from the total number of sites evaluated in the database for each archaeological unit. However, this approach is susceptible to enhance the results of macro-scale modeling procedures, since it allowed for the elimination of sites from the data sets for which the cultural attribution is not certain. Using these vetted data sets in archaeological modeling will allow for more robust and reliable models.

We followed the FAIR principles defined by Wilkinson et al. (2016). These principles are: Findable (public server, linked to a DOI), Accessible (Creative Commons 4-BY License), Interoperable (data format and SQL language) and Reusable. Reusability of the data is facilitated by the fact that the data model was developed with the open-source software LibreOffice Base 7.5, with a description for each field. The present article also contributes to data reusability by describing the database's scope and the methodology used to create it.

We provide a discussion in order to alert future users to possible biases in these data and their potential implications. The data set is likely biased by the very nature of the methodology used to create it, i.e. a bibliographic review. The review demonstrated that publications yield unequal amounts of information from one site to another, and sometimes even for the same site or assemblage. The critical review of collections served to identify additional factors that may distort the geographic estimation of the Middle and Recent Gravettian archaeological units. Namely, unprecise excavation methodologies used in the early 20th century, and those still in use during the second half of the century sometimes significantly impacted the recovery of discreet fossiles directeurs. Finally, post-excavation and conservation activities also underwent methodological changes during the 20th century, which contributed to masking the presence of certain archaeological units.

This discussion serves as a reminder that surveys based solely on the literature may be biased to a certain extent. It is our hope, however, that the integration of historical and archaeological reviews of older collections will allow us to establish better the geographic estimations of archaeological units by integrating formal data quality controls into examinations of archaeological data.

Keywords: Gravettian, database, lithic industry, census, history of excavations, Upper Palaeolithic.

Les Agals, Villarzel-Cabardès (Aude)  :

un nouveau jalon pour la connaissance

de l'artisanat textile au Néolithique en Languedoc

Jean Vaquer, Régis Aymé

Résumé  : Le site des Agals qui a été détruit par un labour profond a livré des vestiges typiques du Néolithique final vérazien datables du 3e millénaire avant notre ère. Le mobilier recueilli en surface comporte deux éléments assez rares relevant de problématiques distinctes qui nous ont paru justifier des analyses et d'établir des comparaisons dans une zone de compréhension plus étendue.

Le premier est un tesson de fond de récipient ayant conservé une croûte carbonisée qu'il a été possible d'analyser en chimie organique et qui a révélé un caramel de cuisson comportant des résidus de plusieurs sortes de préparations alimentaires (cf annexe de N. Garnier). Il s'agit de corps gras d'animaux, de viande, de produits laitiers, de traces d'huiles végétales ou de graines broyées, des traces de légumes feuilles et des boissons fermentées de fruits de Rosacées autres que la pomme. C'est un premier pas dans la connaissance des pratiques culinaires et de l'alimentation au Néolithique en zone nord occidentale de la Méditerranée.

Le second élément remarquable est un poids de métier à tisser en argile cuite qui appartient au type réniforme ou en croissant. Il dénote l'utilisation d'un type de poids tendant en même temps les deux nappes de fils de chaîne et ayant pu fonctionner par basculement. Ce dispositif se caractérise d'abord par une économie de moyen qui a pu être appréciable : le métier vertical à poids permet de se passer d'une barre de lisse par rapport au métier horizontal et le peson allongé à attache bipolaire divise par deux le nombre de modules à mettre en place au bas du métier. Un autre avantage par rapport aux pesons à mode de suspension unique et centré a pu concerner les productions. Les dispositifs à pesons compacts et à attache centrée ont pu servir à la fois pour des réalisations d'étoffes cordées et des tissus sur des métiers à deux rangs de poids. Les pesons à système d'attache bipolaire autorisaient l'enchevêtrement automatique des nappes pour des tissages en armures diverses (toile et sergé) d'après les expérimentations faites par A. W. Lassen. Cet avantage technique aurait permis l'adoption de ce type de peson à partir de la fin du 5e millénaire avant notre ère en Italie, puis sa généralisation sous des formes diverses au 4e millénaire au nord des Alpes et dans le Midi de la France et au 3e  millénaire dans la péninsule Ibérique.

Mots-clés  : Néolithique final, Vérazien, caramel alimentaire, étoffes cordées et tissus, poids de métier à tisser, peson réniforme ou en croissant, armure toile, armure sergé.

Abstract: The Agals site, that was destroyed by deep ploughing, yielded remains typical vestiges of the late Neolithic Verazien style dating from the 3rd millennium BC. The artefacts collected on the surface include two rather rare elements addressing  two separate issues that justify analyses and set comparisons in a wider area of understanding.

The first is a pot bottom potsherd that has retained a charred crust that has been analyzed in organic chemistry and reveals  residues of several kinds of food preparations (Annex by Nicolas Garnier). These include fats from animals, meat, dairy products, traces of vegetable oils or crushed seeds, traces of leafy vegetables and fermented beverages from Rosaceae fruits, other than apples. It is a first step in the knowledge of culinary practices and food in the Neolithic in the north-western area of the Mediterranean.

The second remarkable element is a cooked clay loom weight that belongs to the kidney-shaped or crescentic type. It denotes the adoption of a mode of operation by rocking of weights tending the two layers of the warp threads and authorizing the weaving of fabrics with plain weave, or even with more complex canvas according to the experiments made by A. W. Lassen or by K. Grömer. A survey was carried out to find out when and how this kind of crescentic weights appeared and developed in the north western Mediterranean. Generally speaking, cooked clay loom weights are attested to in the Neolithic but they are infrequent and come in various forms and weights. The oldest were found in peninsular Italy in cardial-epicardial context in the 6th millennium BCE and in context Diana, Ripoli, VBQ in the 5th millennium BCE. These are center suspension system weights that could be used in a single row for twining fabrics or in two rows for plain weave fabrics.  The first crescentic loom weights or increasing to two opposite suspension systems appeared in northern Italy towards the end of the 5th millennium and developed amply at the beginning of the 4th millennium in the Po valley in Lagozza culture contexts, but also in the Southern of France in the recent Chasséen and even north of the Alps during late Neolithic. They coexisted with other types of loom weights until the end of the late Neolithic, as the example of the Agals proves. In the Iberian Peninsula the loom weights of cooked clay are known in the south from the end of the 4th millennium first in the form of weights in centered suspension mode and then, in the 3rd millennium in the form of elongated weights with opposite double perforations (“cuernecillos”). Some crescent shapes are known, but most are rectangular or shaped horns very arched and very narrow, which gives them a certain originality. We can link them to some fabric finds that are linen and plain weave canvas.

The loom weight of Agals datable during the first half of the 3rd millennium BCE is part of the group of loom weights whose opposite apical perforations allowed to stretch the wires of the two layers (even and odd wires) of the warp on the same weights arranged in line at the base of the loom.

This system is characterized first of all by an economy of means which could have been appreciable : the vertical weight loom saves a bar of rails compared to the horizontal loom and the elongated bipolar load cell halves the number of modules to be installed at the base of the loom. Another advantage over the single-mode and centered load cells may have been the productions. The devices with compact scales and centred fasteners could be used both for making rope fabrics and fabrics on two-tier looms. The scales with bipolar fastening system allowed the automatic entanglement of the tablecloths for weaves in various armor (canvas and twill) according to the experiments made by A. W. Lassen.

These advantages can make it possible to better understand the adoption of this kind of loom weight not only in the late Neolithic in Italy but also in the South of France and their enormous success in the southern Iberian Peninsula during the late Neolithic and Chalcolithic. It is therefore possible to envisage that fabrics with twill weave canvas or chevron patterns and diamond points appeared long before protohistory as is generally envisaged.

Keywords: late Neolithic, Verazien, food caramel, twining fabrics and plain weave fabrics, loom weight, kidney shaped or crescentic loom weights, canvas weave, twining, twill weave.

La grotte Huchard (Saint-Martin-d'Ardèche, Ardèche)

Approche chronologique intégrée d'une grotte ornée

paléolithique à la sortie des gorges de l'Ardèche

Julien Monney, Jacqueline Argant, Benjamin Audiard, Pierre-Antoine Beauvais,

Jean-Yves Bigot, Laurent Bruxelles, Stéphane Jaillet, Dominique Gasquet,

Nicolas Lateur, Magali Rossi, Aurélien Royer, Nadine Tisnerat-Laborde

Résumé  : Une approche intégrée de la grotte Huchard, ou grotte du Ranc pointu no1, a été menée en 2020 et 2021. Elle a eu pour objectif d'investiguer les éventuelles complémentarités d'usages fonctionnels et/ou symboliques ayant pu exister entre les différentes grottes ornées du «  complexe de sites  » du Ranc pointu, dont fait partie la grotte Huchard. Ceci a impliqué d'aborder au préalable la question de la contemporanéité d'une ou de plusieurs phase/s de fréquentation de ces cavités. Pour ce faire, plusieurs opérations ont été réalisées, à savoir  : un enregistrement numérique  3D de la cavité, une étude des lambeaux de remplissage laissés par les fouilles de 1957 et des vestiges qu'ils comprennent, ainsi qu'une approche de la géométrie des remplissages et de leur position vis-à-vis des gravures pariétales.

Les âges 14C obtenus sur charbons dans ces placages sédimentaires se placent entre 16330 et 13250  cal.  BP. Globalement cohérents avec le reste des études, ces résultats viennent confirmer les attributions chronoculturelles proposées antérieurement sur la base du matériel lithique (Magdalénien supérieur), tout en les étendant significativement à l'Épipaléolithique. La position altitudinale des gravures pariétales est, par ailleurs, compatible avec des sols de circulation dans la continuité des niveaux archéologiques de l'US  1.

Du point de vue des relations entre les différentes grottes ornées du méandre du Ranc pointu, la rareté des traces d'activité humaine attribuables au Paléolithique supérieur récent dans les zones d'obscurité des cavités du secteur ressort nettement. Ce constat rend difficile de soutenir l'hypothèse d'interactions intenses et/ou régulières au Magdalénien supérieur entre des espaces situés dans l'obscurité totale (Deux-Ouvertures, Tête-du-Lion) et des espaces de résidence dotés d'une ornementation en entrée de grotte (Huchard, Chabot, Figuier). Au contraire, ceci amène à s'interroger sur le caractère anté-magdalénien des gravures de la grotte Huchard et, par conséquent, sur une possible disjonction chronologique entre l'âge des vestiges retrouvés dans les dépôts sédimentaires et l'âge des gravures pariétales.

Mots-clés  : art pariétal, grotte ornée, Paléolithique supérieur, Magdalénien, stratigraphie, numérisation  3D, géomorphologie, chronologie, datation 14C, Sud-Est de la France.

Abstract: Research was carried out at Huchard Cave in 2020 and 2021 as part of the 'Cave Art Dating' project. The aim of this project, since 2008, is to investigate the chronology of human and animal activities in the Palaeolithic rock art caves of the Ardèche river gorge, and the ways prehistoric human groups made sense and symbolically inscribed subterranean landscapes, as well as the socio-cultural contexts in which Palaeolithic cave art was created.

In 2016, a visit to Huchard Cave highlighted its chronological potential and the interest in resuming research at the site.

Huchard Cave, also known as Grotte du Ranc Pointu no1 or Grotte du Squelette, is located in the last meander of the Ardèche river gorge (Saint-Martin-d'Ardèche; Ardèche). In 1908, P.  Raymond reported deep engravings, attributed to the Upper Palaeolithic in its penumbra zone, between 2.5 and 3.5  m above the current ground.

The position of Huchard Cave among a series of seven rock art caves in the 'Ranc Pointu site complex' raises the question of the possible complementarity of their - functional and/or symbolic - uses. The proximity of Deux-Ouvertures and Huchard Caves, and the latter's position as a compulsory passageway to the former raise the question of whether they should be considered as a single site. Are there contemporaneous phases of use of these two caves, as well as the other rock art caves in the Ranc Pointu meander? And, if so, what reciprocal interactions could have existed between them at an anthropological and sociological level?

While the chronology established in the Deux-Ouvertures and most of the rock art caves of the Ranc Pointu meander provide an overview of human activity in the sector, Huchard Cave had not yet benefited from an in-depth study. Since the end of the 19th  century, various archaeological operations have been carried out at Huchard Cave. But the chronology of human activity in the cave during the Upper Palaeolithic remained imprecise.

In 2016, we identified against the walls some remains of the sedimentary levels excavated in 1957 by P.  Huchard. These observations prompted an archaeological operation in 2020 and 2021. The aim was to obtain an age for the archaeological remains contained in these deposits and to test the chronocultural attribution previously proposed on the basis of lithic material. It was also the opportunity to check whether other chronological phases were present and to discuss the morphology of the floor during the Upper Palaeolithic.

To achieve this, we focused on: (1)  a 3D digital recording of the cave, the engravings and the deposits remaining against the wall, (2.1)  a geoarchaeological study of the sedimentary deposits, (2.2)  an analysis of the archaeological palaeontological and paleoenvironmental material they contained, (3)  a study of the natural and anthropogenic traces on the walls, and (4)  an approach to the geometry of past sedimentary fillings of the cave and their chronological and spatial positioning in relation to the parietal engravings.

In the end, we identified 6  siliceous lithic pieces, 56  microcharcoal, 4  elements of red colouring matter, 3  pebbles and 3  remains of microfauna on the surface of the residual sedimentary deposits of the diverticulum. These remains were mainly concentrated in, on the upper 12 to 17  cm of the deposits (SU  1). Radiocarbon measurements yielded seven dates ranging from 16,330 to 13,250  cal.  BP. This chronological range is compatible with the Upper Magdalenian attribution previously proposed by J.  Combier on the basis of the lithic industry. However, the 14C ages obtained suggest that human presence also took place during the Epipaleolithic. The geoarchaeological observations and the results of the lithic and anthracological studies are also consistent with the cultural and paleoenvironmental data available for the Upper Palaeolithic in the region. Although not providing precise chronological information, the presence of colouring matter, pebbles and faunal remains is compatible as well with an Upper Palaeolithic age. The only divergent data comes from palynology. The long use of the cave as a sheepfold during the Holocene period, the circulation of water in the sediment of SU  1 and the exposition of the stratigraphy to the open air for more than 60  years, are all possible sources of pollution that could explain these discrepancies.

The altitude of the engravings is compatible with Palaeolithic circulation floors at the level of the archaeological layer (SU  1). The engraved wall would then be within the manual range of a medium-sized individual (1.70  m). However, we identified a discordance between SU  1 and SU  2. This gap covers a period from the Early Dryas to potentially the Middle Pleistocene. Although there is no archaeological evidence to substantiate it, we cannot dismiss the creation of the engravings during this pre-Magdalenian period.

Regarding the socio-cultural interplay between the various rock art caves of the Ranc Pointu meander, we identified archaeological and spatial similarities between Huchard Cave and the other engraved caves (Chabot, Figuier and, to a lesser extent, Sombre Caves). These four rock art caves are the manifestation of phenomena that seem chronologically related (pre-Magdalenian) and anthropologically similar (dwelling sites), albeit with possible complementarities. Apart from Figuier Cave, no evidence of recent Upper Palaeolithic and/or Epipalaeolithic occupation, contemporary with that of Huchard Cave, is known in these caves. It may be due to the early nature of the excavations carried out there, or it could indicate an older age for Huchard Cave engravings. In fact, both in Huchard and Figuier Caves, the presence of Magdalenian remains may simply reflect the relatively marked presence of this chronological period in the Ardèche river gorge.

As for potential connections with the three caves of the Ranc Pointu meander whose rock art is located in total darkness (Tête-du-Lion, Deux-Ouvertures, and possibly Saint-Marcel), there is a chronological dissimilarity between the periods of human activity in these caves and the dates of occupation of Huchard Cave. It is therefore difficult to support the hypothesis of regular and/or intensive Magdalenian interactions between interior spaces located in total darkness and residential spaces with rock engravings in the semi-dark zone of Huchard, Chabot and Figuier Caves. This observation calls into question the pre-Magdalenian nature of the engravings of Huchard Cave and, consequently, suggests a possible chronological disjunction between the age of the remains found in its sedimentary deposits and the age of its parietal engravings.

The Ranc Pointu meander thus emerges as an ensemble in its own that needs to be understood as a whole. The results obtained at Huchard Cave represent an important milestone opening up further prospects for research on the scale of this complex of sites.

Keywords: rock art, cave art, Upper Paleolithic, Magdalenian, stratigraphy, 3D  digitalization, geomorphology, chronology, 14C dating, South-East of France.

Pourquoi l’homme sans arc devrait-il chercher des flèches  ?

Discuter une hypothèse non parcimonieuse  :

le cas de l’arc à la Grotte Mandrin (Paléolithique supérieur initial)

Laurent Klaric, Sylvain Ducasse, Mathieu Langlais

Résumé  : En Préhistoire, au delà de trouvailles inattendues faites dans des contextes clairs, il arrive que les découvertes ou résultats d’étude «  extraordinaires  » soient sujet à controverses. Cela reflète en général la difficulté qu’ont les archéologues à s’accorder sur le degré de précision et la construction de leur argumentation. De manière concomitante, une certaine surmédiatisation vise parfois à informer et convaincre le public de l’importance d’une découverte jugée majeure par ses auteurs, sans que, pour autant, le résultat ne remporte une large adhésion de la communauté scientifique. Constituant un exemple de la dérive sensationnaliste évoquée, une étude récente (Metz et al., 2023) consacrée aux pointes et micro-pointes du niveau E de la Grotte Mandrin (Malataverne, Drôme) vient de conclure à l’usage de l’arc il y a 54  000 ans par un groupe Homo sapiens en vallée du Rhône. Ce résultat proclame ainsi le vieillissement de près de 40  000 ans de l’usage reconnu de l’arc en Europe. Une lecture attentive de cette étude ne nous a pas convaincu de la validité de ce résultat «  extraordinaire  » qui nous semble a priori incompatible avec le principe de parcimonie. C’est sur la base du raisonnement et de la logique argumentaire que nous proposons une réfutation du mode de propulsion inféré qui est en grande partie fondée sur un «  postulat d’efficacité  » au détriment d’autres données archéologiques pourtant disponibles sur le site. Nous exposons une réfutation basée sur  : 1) les données présentées dans l’étude ainsi qu’à l’occasion d’autres travaux antérieurs sur les armes préhistoriques, 2) des observations ethnographiques issues de trois continents et relatives à la chasse et son apprentissage par les enfants, 3) une régularité générale touchant à l’apprentissage dans les sociétés d’Homo sapiens. À l’issue de notre réflexion, nous proposons que les pointes microlithiques de Mandrin puissent correspondre à des parties d’armes miniatures potentiellement attribuables à l’activité des enfants.

Mots-clés  : Paléolithique supérieur, industrie lithique, armement, chasse, apprentissage, enfants, microlithisation, miniature, argumentation, ethnographie.

Abstract: In Prehistory, beyond unexpected finds made in clear contexts, “extraordinary” discoveries or study results are sometimes the subject of controversy. The latter generally reflect the difficulty archaeologists often have in agreeing on the degree of precision and construction of their arguments. From the age of humankind to the settlement patterns of the Americas, the “modernity” of Homo sapiens and the role of women in prehistory, bold hypotheses and startling conclusions follow one another in the wake of media fashions. At the same time, there may be a certain amount of media hype aimed at informing and convincing the public of the importance of a discovery deemed major by its authors, although the announced result may not be widely supported by the academic community. A recent study (Metz et al., 2023) devoted to the points and micro-points of level E of the Mandrin Cave (Malataverne, Drôme) is just one example of the sensationalist drift mentioned above. It concludes that the bow was used 54,000 years ago by a group of Homo sapiens in the Rhône Valley. The spectacular result proclaims that the recognized use of bow and arrow in Europe is circa 40,000 years older than previously known. However, a careful reading of this study did not convince us of the validity of this “extraordinary” result, which seems to us a priori incompatible with the principle of parsimony. Indeed, the article by Metz and colleagues would require further comment on the technical aspects of the study (e.g. criteria for morpho-typological and productional distinction of flint points, reality of the existence of a nano-point component, criteria about the functional analysis, details of the parameters of the Initiarc experiment). However, we have chosen to accept as valid the hypothesis that a large proportion of the points analysed correspond to axial tip of hunting projectiles. It is therefore on the basis of reasoning and argumentative logic that we propose a refutation of the use of bow and arrow 54,000 years ago on the Mandrin site. Indeed, the conclusion of the study by L. Metz and colleagues is essentially based on a “postulate of effectiveness” that ignores certain other archaeological data available on the site. We therefore first propose a rebuttal based on 1/ the data presented in the study and in other works about to prehistoric weapons, 2/  ethnographic data from three continents relating to traditional hunting and the place of children in this activity, and 3/ a general regularity concerning learning in present-day Homo sapiens societies.

Firstly, we analyse the arguments used by L. Metz and colleagues, presenting them from the point of view of argumentative logic, using the categories explained in C. Plantin's dictionary of argumentation (Plantin, 2021). This analyse leads us to distinguish between the main arguments in the Metz et al. demonstration (archaeological impact fractures similar to those used in the experiment, the “rule” that the diameter of the projectile shaft is necessarily less than the width of the point, the use of the bow for maximum penetration efficiency of a light projectile, calculation of the TCSA) and the premises on which the demonstration is based. We show not only the weakness or artificiality of some of the arguments/premises presented, but also the “argumentation by default” nature of the demonstration put forward by L.  Metz and colleagues, since no new decisive argument is proposed that would allow a direct and differential diagnosis of bow use (as opposed to spearthrower or bare-handed use).

The entire demonstration is based on the necessity of using the bow to ensure optimal penetration of projectiles armed with micro- and nano-points. This reasoning by default implies a condition of refutability: the demonstration is only valid if the small lithic points were necessarily intended to be effective from the point of view of the projectile penetrating into its target.

We then show that there are many examples among traditional hunters of the ethnographic register of miniature children's weapons, more or less faithful replicas of those used by adults, which constitute children's play or hunting training equipment. These children's weapons are not necessarily intended to be lethal or as effective as those used by adults (i.e. penetration of the tip into a living animal target is not a major requirement, since the main targets are often inanimate for beginners), but they are often functional enough for shooting, practising and playing. In fact, they can reasonably sustain damage during these uses. This simple explanation, which does not imply that the bow was used 40,000 years earlier than archaeologically attested, has not been considered or tested in Metz and colleagues study. At the end of our analysis, we propose that the Mandrin microlithic points of level E could correspond to tip of miniature weapons that could potentially be used by children. It would, of course, be necessary to ensure that such miniatures could break in the same way as adult weapons when used by children on inanimate target. However, it seems to us that, given the current state of knowledge, this hypothesis is undoubtedly more parsimonious than that of the use of bows and arrows 40,000 years earlier than currently documented in the archaeological record at Stellmoor in Germany. Through this example, we finally underline the excesses of a certain number of sensational studies and publications with a wide international audience, where a critical reading of the arguments would often prevent the promotion of non parsimonious hypotheses as solidly established scientific results.

Keywords: Upper Palaeolithic, lithic industry, weaponry, hunting, apprenticeship, children, microlithism, miniature, argumentation, ethnography.