Habitat du Néolithique ancien et nécropoles du Néolithique moyen I et II à Vignely « la Porte aux Bergers », Seine-et-Marne

International audienceThe site of Vignely “ la Porte aux Bergers ”, Seine-et-Marne, was exca-vated in advance of gravel quarrying in 1993 and 1994. It is located on the edge of the lower terrace of the Marne, which widens downstream of Meaux and forms vast meanders with extensive alluvial plains.Environmental studies were undertaken on a long section cut into the former bed of the Marne at Vignely “ la Corvée ”. As early as 9240-8790 cal. BC, peat forms in the channel and the landscape is dominated by pine woods (Preboreal), then hazel (Boreal). Between 7516 and 6435 cal. BC (Early Atlantic) detritic activity increases and mixed oak forest develops, even though hazel remains abundant. After an incision, a small channel forms and starts to fill just before 2470-2030 cal. BC. A broad alder wood covers the banks of the Marne, with oak and beech on the valley sides. Bronze and Iron Age people frequent the area, regularly clearing the riparian woodland and carrying out farming activities. The early and middle periods of the Neolithic are not documented on the section. Nevertheless, the abundant pollen data from this area do provide some environmental evidence for these periods, with quite open oak-lime forest on the valley sides and a gradual appearance of alder in the valley bottom. It remains impossible, however, to characterize the activities and environmental impact of the inhabitants of Vignely “ la Porte aux Bergers ”.The Early Neolithic settlement at Vignely “ la Porte aux Bergers ” consists of ten house units and two graves. Four house units have inter-pretable ground plans, two were only partially uncovered as they were on the edge of the excavation, and one was considerably disturbed by an overlying Middle Neolithic funerary monument. The three other units are more hypothetical. The reconstruction of house 10 as part of a project for an archaeological park provided an opportunity to combine architectural and archaeological considerations, relating to the layout and volume of the structure, as well as its framework, walls and roofing.Settlement features produced over 4,000 sherds for a total weight of 37,6 kg. The pottery is unequally distributed between houses, with quan-tities varying from 1 kg to 18 kg. Despite the quite low numbers of deco-rated vessels per house, the pottery provided a chronological sequence from the end of the Seine basin late Linear Pottery Culture (RFBS) to the final stage of Blicquy/Villeneuve-Saint-Germain (BVSG). The earliest settlement phase, as represented by houses 195 and 90, provided evidence for a stage previously poorly known in the lower Marne valley. This covers the end of RFBS and the very beginning of BVSG, as shown by bands filled with crossed lines, large triangles filled with impressions, and fish-bone motifs combined with decoration inherited from Limburg pottery. The middle stage of BVSG is only represented by two small and poorly defined assemblages from houses 150 and 315, although it is very well attested on other settlements of the lower Marne valley. Lastly, the end of BVSG is divided into two stages, on the basis of the distinction between the fine cordons of late BVSG (houses 10 and 320) and thick cordons which are more frequent in final BVSG (feature 152). Two graves dating to the Early Neolithic contained one or several vessels. Grave 155, with its association of two small inflection vessels and a flask, all undecorated, corresponds to the standard shapes placed in RFBS and BVSG burials.Petrographic analysis of a sample of fifteen vessels revealed consider-able diversity in clay resources. While some materials come from the other bank of the Marne, though located nearby, other more distant sources were also exploited. Most of the materials were used at the same time and vessels made from different materials were found in the same house assem-blage. The same clay resources were used throughout the occupation sequence.Grog is the only temper that was used for over half the vessels right through the occupation sequence. A house assemblage can contain vessels with or without grog. These results are comparable to those obtained on other BVSG series from the lower Marne valley. It is noteworthy that there is no evidence for use of crushed bone, a temper that is sometimes found in pottery from other sites in the area.The flint industry includes some original traits compared to other sites in the Marne valley. This is especially the case with the main raw material used, Cretaceous flint, which occurs throughout the occupation sequence. The local raw material is Bartonian flint so the Cretaceous variety must have been imported. These imported nodules were used for blade produc-tion that was diffused to other settlements in the valley, while blades in Bartonian flint produced elsewhere were brought in. The houses with the most Cretaceous flint artefacts tend to be located in the north/north-west part of the settlement and it is suggested that these differences reflect complementarity between houses in each occupation phase. These obser-vations also raise questions about the socio-economic status of the inhab-itants of Vignely “ la Porte aux Bergers ”, as this could explain its isolation in relation to access to local raw material.The macrolithic industry of Vignely “ la Porte aux Bergers ” consists of only sixty-two tools or tool fragments, as well as nine flakes and seven fragments. The finds come from seven of the house units and two unat-tributed features (features 80 and 305). The highly siliceous sandstone for the grinding tools and the ferruginous sandstone for the abraders come either from the alluvial terraces of the Marne or from local outcrops. A few hammerstones, cutting tools and flakes are in quartzite. As the numbers of artefacts are low, it is not really possible to discuss differences between house units in the composition of tools. However, they seem as diversified as on most BVSG settlements. The tools mostly reflect grinding activities and are only roughly shaped, unlike the artefacts known from other sites such as Rungis, Ocquerre or Luzancy. Paradoxically, the various distinctive criteria seem to occur throughout the house units, whatever their chrono-logical attribution. Like the flint industry, the macrolithic artefacts under-line the rather particular status of the inhabitants of Vignely “ la Porte aux Bergers ” within the network of BVSG sites in the lower Marne valley.With points on abraded metapodia, fine proximally shaped points, bevel-ended tools on bovine metapodia, red-deer antler punches and bevel-ended tools on roe-deer antler, the bone industry combines the main character-istics of BVSG sites in the Marne valley. However, one notes the absence at Vignely “ la Porte aux Bergers ” of certain types that are quite common in the Marne valley, such as heavy antler tools, points on complete ulnas, points on metapodia of large ruminants, as well as bevel-ended tools on ribs.Ornaments are made of stone or shell. Despite the small numbers of stone ornaments, and the narrow range of materials employed, this series underlines use of white and grey Tertiary limestone, local raw material, for the manufacture of bracelets, from the early phase of BVSG. The exploitation of these new local outcrops is probably a response to the disappearance of the networks that had previously supplied sites at the end of the RFBS with bracelets in white limestone from Vocontian Cretaceous deposits in the Vaucluse. The surface of the Vignely “ la Porte aux Bergers ” bracelets is nonetheless of comparable aspect to those of the RFBS. The majority of shells are of local fossil origin, even though a certain number come from the Mediterranean, the Channel or the Atlantic. Through the raw materials used and the types of object manufactured, the site of Vignely “ la Porte aux Bergers ”, like other contemporary sites, shows cultural continuity with the RFBS and particularly with that of the Aisne valley. Vignely “ la Porte aux Bergers ” shared with other communities occupying the region a set of cultural traits, but stands out however through its very diversified procurement of raw materials. This again raises the question of the site’s status within the lower Marne valley.The faunal assemblage from Vignely “ la Porte aux Bergers ” is broadly comparable to other Early Neolithic sites in the Paris basin. The fauna mostly consists of domestic animals, with cattle in first position, although hunting is not totally abandoned. In terms of the trends observed in other BVSG sites with faunal remains, Vignely “ la Porte aux Bergers ” represents an exception. Although cattle are the main source of meat throughout the period, sheep and goat predominate over pig, despite the fact that the latter normally predominate at the end of BVSG. The interpretation of this phenomenon is difficult as the settlement is only partially excavated, but these differences are perhaps related to the variability of social structures and/or the village space.Two graves were also found at Vignely “ la Porte aux Bergers ”, as well as isolated human bones in the settlement pits. The orientation of the burials towards the east and their position on the back with the legs flexed to the left are characteristic of the Early Neolithic. On the contrary, burial 155, on the back with flexed limbs and raised knees, is the only burial in this position known for BVSG. However, this positioning of the body is attested in RFBS contexts. With the notable exception of a quartzite tool, the grave-goods are fully representative of BVSG funerary practices, in terms both of rules of placement in the grave and in relationship to the person buried.In sum, all the lines of evidence gathered at Vignely “ la Porte aux Bergers ” converge to make this a particular site amongst the Early Neolithic settlements in the lower Marne valley. The settlement was occu-pied over a long period of time and the various categories of find display differences with the other contemporary sites in the valley. The heritage of the RFBS seems more pronounced at Vignely “ la Porte aux Bergers ”, thus raising the question of the site’s founding role in the Neolithic settle-ment of this part of the valley.After the Early Neolithic and until the Iron age, only funerary remains are found at Vignely “ la Porte aux Bergers ”. There is a Late Neolithic collective grave, at least two Bronze age burials and a cremation, as well as an Iron age grave, but none of these are included in this volume. Most of the funerary remains are Middle Neolithic, with a total of twenty-six graves. The term cemetery might logically be applied here, although this is debatable. The twenty-two Middle Neolithic I graves extend over a surface area of almost 0,5 ha and there is no evidence that they were originally larger numbers of graves. Their low density in the investigated area indeed leaves room for possible domestic structures, although the total absence of finds from this period outside the graves argues against this hypothesis.The attribution of graves to the Middle Neolithic I or II is firstly based on typology, then on radiocarbon and lastly on associated finds. The diver-sity of many of the graves makes dating difficult, with the notable excep-tion of the predominant Balloy type, with thirteen graves. Identified on ten or so sites in the central Paris basin, this type is generally assigned to the Cerny culture. Radiocarbon dates confirm this association and as at Passy show that this architecture occurs from the very beginning of Cerny. For the graves with flexed burials, the distinction between Middle Neolithic I and II is essentially based on radiocarbon. The 2003 attributions have thus been revised, and logically so in view of the siting of the graves. The attribution to Middle Neolithic II of the monument and associated grave, as well as the two burials with flint axes, can be made without much difficulty, but ultimately the actual position within the long period of time involved can only be established with radiocarbon.Thus half the graves at Vignely “ la Porte aux Bergers ” are of Balloy type. The site has indeed provided much new information about this type. A vault is built in a “ bath-shaped ” pit of sizeable dimensions and consid-erable depth (up to 0.80 m below the excavation surface at Vignely “ la Porte aux Bergers ”). There are no traces of the construction material, but one can envisage that wood or earth were used. The corpse was lowered into the grave in a rigid, open container. After burial, the grave is closed but not back-filled. In at least one case at Vignely “ la Porte aux Bergers ”, the grave was re-opened and a second burial was added after back-filling to cover over the initial burial (feature 130) ; this results in a so-called grave “ with storeys ” (“ à étages ”). Less common on the site, the other grave types have few parallels in Neolithic contexts, with the exception of a small wooden cist grave which evokes similar 5th millennium struc-tures in central France and around the Alps. Amongst these rare burial types, one can mention a grave for which only the foundation slot for the wall was preserved (feature 273) ; comparable features at Passy can thus be reinterpreted as graves that were completely or partially above-ground. Two graves contain a corpse wrapped in a flexible cover.The burial positions are closely linked to the type of grave. In the Balloy-type graves, the corpse is on the back, with all limbs stretched out. The head is raised. With the flexed burials, left-sided positions predominate in Middle Neolithic I and are the only position observed in Middle Neolithic II. In the same manner, all the corpses in flexed position have the head to the east (with a wide range), while in the Balloy-type graves the head is either to the east or the west.The Middle Neolithic graves contain thirty-seven individuals. Leaving aside the poorly documented feature 137, there are thirty-one Middle Neolithic I individuals, nineteen of which are adults although all age-classes are represented. The age-at-death pattern is nevertheless incompatible with a natural distribution. The youngest individual from a Balloy-type grave is between nine and twelve years old. Were young children – few in number – buried in a manner that is less likely to have been preserved ? In contrast, individuals between ten and nineteen are over-represented (five). Lastly, adult males are only found in Balloy-type graves. The anthro-pometric traits of the small series reflect the picture of the Cerny popula-tion in the Paris basin : adults are relatively small in stature, with a gracile skeleton and quite unpronounced muscle attachments. Most of the differ-ences suspected here between men and women are clear at population level. Three pathological cases are worth underlining : the woman 72-2 was trepanated, the child 148A had scaphocephaly, and the woman 301, deceased at over sixty, exhibits a bone lesion evoking Pott’s disease, an expression of osteoarticular tuberculosis. Nevertheless the state of health of Cerny individuals from Vignely “ la Porte aux Bergers ” is generally good, as is the case with populations known for the same period in the Paris basin. Statistically, the six post-Cerny individuals from the site show more signs of physiological stress. Lastly, some specific features suggest genetic proximity between the Middle Neolithic I individuals.Half of the Middle Neolithic individuals are associated with grave-goods. For the Cerny, taking into account the composite nature of some of these (mostly ornaments), forty-seven items accompany the deceased. They are not uniformly distributed in relation to age and gender. Although more numerous, only 40 % of adult females have grave-goods, compared to 80 % of adult males or immature individuals under twenty. The differ-ences become greater when one considers the number of items present in the graves : women with grave-goods have on average fewer items than men and children.The situation varies according to the category of find. Pottery is rare: it occurs in two Middle Neolithic I graves and one later grave, but in all three cases the pottery was outside the actual burial. The deposit with grave 257 laid on the edge of the pit, up against the grave structure: four small vessels with point or comb decoration characteristic of Cerny-Vide-lles. The lithic finds are difficult to distinguish from those of the BVSG settlement. This can even cause problems of interpretation for the graves the most disturbed by animal burrows: blades knapped by indirect percus-sion or tranchet axes could have been incorporated by chance, either in the fill or through disturbance of the bottom of the grave. Well-represented in other Paris basin Cerny cemeteries, a single arrowhead was found, associated with individual 130A. This arrowhead, an asymmetrical pointed type classic in BVSG contexts, raises once again the problem of identifying actual placement of lithic artefacts, as well as the tendency to include “ exotic ” arrowheads in Cerny graves. The bone industry makes good use of wild species. Both for tools and ornaments, bones and teeth of large mammals were used: red-deer scapula and aurochs metatarsal, red-deer antler and canines associated with a probable wolf vertebra, bear molar, etc. Although most items are unique at Vignely “ la Porte aux Bergers ” and rare elsewhere (as often in cemeteries), particular mention can be made of a roe-deer leg, the only case of a possible food offering in a Cerny context in the central Paris basin, as well as the three unworked pig phalanxes found in the thorax area of three individuals, items of enigmatic function that are now known from several cemeteries. Lastly, ornaments represent the most abundant finds category in terms of numbers of items. Apart from the tooth and bone artefacts already mentioned, there are also half a dozen shell species, of both fossil and marine origin, used for manu-facture of beads, in addition to limestone.In the absence of “ Passy-type features ”, a group of four pairs of single graves holds an important position, perhaps central, in the cemetery. Despite characteristics very similar to Passy, Balloy or Gron, the absence of Passy-type monuments distinguishes the Vignely “ la Porte aux Bergers ” cemetery and contributes to the necessary deconstruction of the Cerny funerary system: types of grave, artefact assemblages, monuments, struc-turing with preeminent individuals, monuments-Cerny culture relationship, are all elements that are not necessarily associated at the beginning of the Middle Neolithic. In fact there is one Middle Neolithic funerary monument, located as at Balloy on the ruins of an Early Neolithic house, but this monument is later than Cerny. The first of a series of similar discoveries, the monument is problematic as far chronological and cultural attribution is concerned. The monuments at Saint-Julien-du-Sault, Beaurieux or Passy/Véron are assigned to different periods of the Middle Neolithic, none of which are synchronous with the radiocarbon dates from the burials asso-ciated with the Vignely “ la Porte aux Bergers ” monument.Twenty-five years after its excavation, the Vignely “ la Porte aux Bergers ” site remains of key importance, both for the Early and Middle Neolithic. The succession in a single area, without interruption according to the dates, of a BVSG settlement and a Cerny cemetery suggests a form of filiation, underlined by the links between the two cultures, that cannot be considered as linear in the absence of an identified settlement for the second period