Navigation – Plan du site

A Magdalenian camp at Étiolles, Essonne (abridged version)

From activities to the micro-sociology of a habitat
Monique Olive, Nicole Pigeot et Olivier Bignon-Lau
Traduction de Louise Byrne
p. 47-51
Cet article est une traduction de :
Un campement magdalénien à Étiolles (Essonne) [fr]


L’ambition de cet article est d’appréhender la structure sociale d’une petite communauté de Magdaléniens à travers l’analyse palethnographique d’un niveau d’occupation d’Étiolles, le niveau U5-P15. Fouillé sur 700 m2, ce niveau est interprété comme une partie de campement, composée de deux habitations, U5 et P15, abritant chacune une unité familiale, et associées à quatre foyers annexes. L’analyse spatiale met en évidence l’organisation des activités dans l’habitat et les remontages de silex attestent une circulation importante des produits de la taille entre les unités, essentiellement des nucléus et des lames. Cette circulation traduit la mobilité des individus dans l’espace du campement et aussi dans le temps du séjour. De nombreux échanges entre les habitations révèlent une réalité sociale complexe qui se manifeste par des comportements d’entraide et de partage, et une certaine forme de hiérarchie entre les deux familles. En outre, une temporalité inégale des deux habitations pose la question d’une recomposition du groupe magdalénien durant son séjour à Étiolles.

Haut de page

Notes de la rédaction

This article is not quite a traduction but rather is a long summary of the French article « Un campement magdalénien à Étiolles (Essonne) ».

Texte intégral

1This article focuses on an occupation identified over a surface of 700 m2 with several settlement structures interpreted as dwelling units (U5 and P15), associated with adjacent units (G13, J18, N20 and S25). Numerous flint refits demonstrate the coherence of the settlement and the contemporaneity of all these structures and delimit the domestic area of each unit. The excavated surface only corresponds to part of the camp, which must initially have been much larger, but the analysis of this partial surface was sufficient to reveal a form of social organisation. This level has already been partially studied and published but the analyses focused on several units: the only two dwellings U5 and P15 (Pigeot 1987a, Olive 1988) and two of the small adjacent units (Olive 1992). Up until now, the two other adjacent units, N20 and S25, were still unpublished and no overall analysis had been carried out.

2First of all, this study aims to identify the type of each camp unit, and their domestic or annex status. Then, we examine the connections between the different units. The refits show an intensive circulation of objects and people in the settlement. This network of movements reveals the complementarity of the activity areas and connections between the residential units. The temporal parameter is also taken into account in order to understand how the camp developed throughout time. In conclusion, we attempt to decipher the socio-economic structuration of this Magdalenian community on the banks of the Seine from the spatio-temporal data of the site.

Installation of the camp and type of occupation units

3The paleotopography and the sedimentary study bring to light the choice to establish the camp on the convex bank of the former bed of the Hauldres stream, a tributary of the Seine. The presence of the latter clearly determined the distribution of the different units. The dwellings are implanted on the top of the bank and the slope while the four annex hearths are aligned at the base of the bank, along the channel (fig. 2).

Fig. 2 – The camp of dwellings U5 and P15

Fig. 2 – The camp of dwellings U5 and P15

CAD M. Olive and N. Pigeot

4The two dwelling units can be clearly differentiated on the basis of their organisation and the quantity and type of remains. We can clearly distinguish a main dwelling, U5, and a dwelling that could almost be described as subordinate, P15, based on qualitative and quantitative criteria.

5After more than 40 years of excavations, no comparable units to unit U5 have been unearthed at Étiolles (Pigeot 1987a). It is unique on account of its stony infrastructure and it is also one of the dwellings with the most lithic remains at Étiolles and by far the richest unit in level U5-P15. At the exterior, the domestic territory of this unit extends to the limit of dwelling P15 and comprises several flint concentrations and four hearth purges. Many elements point to a rather long occupation duration of unit U5. The diversity of domestic and technical activities carried out around the hearth and the identification of clumsy debitage evoke the residence of a family group comprising adults and children.

6Unit P15 can be clearly differentiated from the neighbouring U5 as it contains no stony structure, a much smaller quantity of flint, a less extensive domestic territory, and no hearth purges or heaps of flint, as observed in U5. The only common denominator between these two units is the presence of a large stony hearth. There is no direct proof of the presence of a shelter in P15 but this unit contains several criteria that appear to back up such an interpretation (type and organisation of activities, movements of objects…).

7We observe a similar diversity of manufacture and consumption activities between these two dwellings, but with very different densities. This suggests a shorter occupation duration of P15. Another difference clearly emerges between these two units: the type of debitage. The laminar production carried out in P15 is of a much poorer quality than the debitage in U5. In P15, there is no elaborate debitage supplying series of very long blades. In fact, simplified, often rather clumsy laminar debitage predominates in this unit, giving the production a rather singular character.

8The four units associated with the two dwellings are interpreted as open-air annex activity areas. They present similar characteristics: an identical topographic situation, a more limited quantity of remains than the dwellings and a simpler, less structured hearth than the large stony hearths U5 and P15.

Organisation of activities at the scale of the camp

9The comparison of all the units enables us to grasp how the Magdalenians used their living space and to clarify the status of these units in the camp.

10The poor preservation of the bone remains masks a whole sector of activities related to game processing, which must have played an important role in daily tasks. However, the exploitation of abundantly available flint of exceptional quality was extremely important and clearly structured the different areas of the camp. Other more discreet evidence, such as ochre and ornaments, also contributes to our knowledge of the camp.

11First of all, we note the absence of genuine bladelet cores in this level, except for a few doubtful cases, as bladelet production was carried out on burins. Most of them are relatively classically distributed at Étiolles, near the domestic hearth where activities related to repairing hunting arms were carried out. It is logical to find burins in the same place, with armatures, as they were sometimes used to work reindeer antlers, and sometimes as cores producing blanks.

12The preparation areas are distributed around the whole camp, but when we look a little closer, we observe different behaviours between the diverse units. These zones are relatively rare in U5, more numerous in P15, and predominant in the annex units, particularly in N20 and especially S25. This debitage phase is carried out outside, which is hardly surprising as it generates a huge quantity of waste, particularly at Étiolles where the flint nodules are very large.

13Conversely, the laminar debitage workstations are clearly clustered in unit U5, mainly around the hearth, but also in the workshops near the dwelling. Indeed, three-quarters of the blades were produced there. In contrast, few cores were knapped in dwelling P15 and in the four annex units. The best and most productive laminar production thus clearly took place in U5. The best knappers worked in this residence unit to obtain series of long blades. In most of the units, we also find debitages showing signs of awkwardness, to diverse degrees, apart from in S25. It is important to recall the relative quantitative and qualitative importance of such debitage in unit P15, and the contrast with laminar production in both dwellings.

14In this camp, the manufacture and repair of hunting arms also took place in dwelling U5. Nearly all of the backed bladelets, or 267 elements, which even after being refitted represent 46.9% of the useful production (Pigeot 1987a, p. 70), were found near the domestic hearth, and to a lesser extent, around the shelter. It is thus probable that hunting activities played a major role in the daily life of these Magdalenians, in spite of the extreme paucity of bone remains in this dwelling. In contrast, the maintenance of arms appears to have been a minimal activity in P15, which only yielded a few arms (less than ten). Finally, this operation left practically no traces near the annex hearths, as only one armature was found near hearth N20. Again, most of the retouched tools are concentrated in unit U5, but unlike armatures, the former are more broadly distributed around the camp and are found in all the other units, in particular in dwelling P15 which contained about fifty retouched tools. Tools were also found near the four annex hearths but in smaller quantities. They show that activities other than debitage took place in these small units.

15In spite of the small number of preserved bone remains (NRt=335), mainly due to an unfavourable taphonomic context, it is nonetheless possible to extrapolate some data related to the processing of faunal remans. The reindeer is the dominant animal species at the camp with 107 determined remains (97% of NRd) and a minimum of seven individuals (85% of the NMIc), followed by the bison, with only three osseous fragments. The representation of the skeletal parts of the reindeer and the variability of the left and right elements imply that some quarters are missing, either because they were not brought back to the units, or because they were processed beyond the limits of the excavation.

Circulation of objects, mobility of individuals

16Among the knapping products, two specific categories circulated in the camp: cores and blades, and some rare flakes.

17When core movements are sufficiently explicit, they reveal a diversity of situations; transport between preparation and laminar debitage, during preparation, during laminar debitage, after a knapping accident or not. The most frequent case is the retrieval of a core by a novice knapper. This latter situation confirms a rule ordering knapping activity in the habitat: apprentice workstations are generally distinct from those of competent knappers. We observe a reciprocity between habitations U5 and P15 for the circulation of depleted and reworked cores. This behaviour reveals a more subtle relationship between the occupants of these units than the marked quantitative and especially qualitative disequilibrium between the laminar productions carried out in each unit tends to imply. It is also important to note that the movement of cores is not limited to the excavated portion of the camp and mostly takes place beyond the excavation limits. This is shown by the considerable number of debitage sequences carried out “elsewhere”. These exchanges with zones beyond the limits of the excavation occur for all situations: cores brought in or out, already prepared, already knapped or already reworked, with strong potential or not, propitious for debitage or not, etc.

18Utilitarian blanks underwent multiple movements in the camp, and circulated more than cores. These consist mainly of unretouched, retouched or used blades, circulating from one unit to another, mainly between the habitations. These movements occur in both directions, but, are more clearly unbalanced in this case. The blanks knapped in U5 then transported to P15 are twice as frequent as those transported in the opposite direction (from P15 towards U5), and are not of the same quality. The dissimilar circulation of blades between the two habitations shows the marked quantitative and qualitative disparity of the debitage carried out in each unit. The introduction of retouched or non-retouched blanks in P15 supplements the low local production of long blades. About ten connections associate the two habitations with the annex hearths and most of the movements take place in a single direction, towards the annex units.

19In conclusion, the flint refits bring to light an important network of circulation of lithic products between the two dwellings on one hand, and between the dwellings and the satellite hearths on the other hand (fig. 60). Exchanges are clearly more intensive between the two dwellings than elsewhere, denoting close relationships between their respective occupants. The inter-unit connections also reveal the movement of group members between the dwellings and exterior hearths, preferentially hearths N20 and J18. Connections between the annex hearths are limited to several rare cases of cores transported during the course of debitage.

Fig. 60 – All the movements between the different camp units

Fig. 60 – All the movements between the different camp units

CAD M. Olive and N. Pigeot

Occupation duration

20Previous studies carried out on several dwellings of Étiolles (U5, Q31, A17, W11) showed that the contribution of flint refits is not limited to the demonstration of strict contemporaneity between the units, but rather brings to light their relative synchrony. This method shows at least the partial sequencing of debitage operations and retraces a fragment of the history of the occupation. This reconstructed chronology of the occupation brings to light a hierarchy of activities and reveals the consistent management of the living space throughout the duration of the occupation.

21The dwelling unit U5 is the only one combining all the criteria allowing us to propose a model of evolution for space and activities. In all the other units, including dwelling P15, the debitage deposits are not dense enough to interpret their sequencing. Thus, the interpretations of the internal history of the camp are based on the U5 model.

22In her study of dwelling U5, N. Pigeot discerned two main settlement phases marked by a radical change of behaviour in the domestic area (Pigeot 1987a, p. 109 sq.). During a first phase, the elaborate debitage of long blades mainly took place near the central hearth, in the southern sector of the shelter, while knapping waste was evacuated outside towards the waste heap O8-9, along the contiguous southern route. The hearth purges are also related to this early period, and some of them are near P15. In the second phase, the knappers’ behaviour changed: elaborate debitage now took place outside, in workshops AZ 8-9 (to the west), then T 8-9 (to the south), and only tools were brought inside, still near the domestic hearth but in the opposite sector, this time using a new entrance opposite the hearth to the west. The cores were placed near the entrance of the shelter without being brought inside. In contrast to the now functional western entrance, the previous southern entrance no longer seemed to be used and appears to be closed. The explanation advanced to interpret this change of attitude is that the same human group stayed at the site for several seasons of the year, in winter, spring and/or summer, for example (Pigeot 1987b). Improved climatic conditions between the two occupation periods could explain the choice to first of all knap in the dwelling (with, as a result the evacuation of waste), then outside. We cannot rule out the possibility that the observed rupture in the organisation of knapping activities may coincide with a relatively brief abandonment of the site, which did not require taking down the tent structure. Note that the sequencing from one phase to the other is all the more significant in that it also coincides with the renewal of siliceous materials.

23The first very surprising element concerning the temporality of P15 in relation to U5 is that it was only in operation during the early part of the occupation of U5. All the debitage showing exchanges between these two dwellings comes from the early occupation phase, including reworked cores and debitage supplying the blanks used in P15. Unit P15 would thus have been abandoned during the second occupation phase of U5, confirming the radical rupture that took place in U5 during the second phase.

24Most, if not all of the annex hearths were in activity from the early period of the camp onwards when the Magdalenians lived in both dwellings. The camp thus had the same layout as when it was discovered from an early stage onwards, apart from another possible unit below the road to the north of U5 and perhaps an extension towards the west, as the question remains open for S25. On the other hand, the situation is less clear for the recent period, following the abandonment of the dwelling P15. Unit N20, which is the richest of the annex units, was still occupied but the data are missing for the three other hearths. Were they abandoned, like P15, or only some of them?

An attempt to decipher the micro-sociology of a Magdalenian community

25The circulation of knapping products thus sheds precious light on the life of Magdalenians in their habitat. It establishes the exact status of the diverse occupied areas, dwelling units or annex units, and the duration of their functioning. In the background of the spatial organisation of certain activities and the interlacing of the route followed by several significant blades and cores, we discern individuals working and moving throughout the camp, and the social relations structuring this small community. During the duration of the occupation, we also perceive the flexibility of this nomadic group in its annual cycle of movement.

26The movements of blanks between dwellings U5 and P15 corroborate what had already been revealed by the comparison of the structures and their respective lithic productions. The best knapper (or the best knappers) worked in U5, on the best, largest and most difficult materials, as they were capable of extracting series of very long blades, some of which circulated in the camp. It is also in U5 that hunters repaired and made their weapons before hunting expeditions. These two dwellings clearly did not have the same status in the camp and their occupants did not have the same economic weight within the community. The annex units along the stream appear to be exterior living areas where the adults and children of families U5 and P15 spent time together.

27At first, we would expect an opposition between the tent representing a private space for its inhabitants, and outside, a free shared area for collective use. In reality, the use of space appears to be more subtle as the domestic area itself, in U5, is different to the community area of the central hearth and the personal peripheral zones. Yet, the same partition exists for the exterior area, where the immediate perimeter of the tents is a “private” space in a certain way, dependent on the activities and evacuation of materials of the inhabitants and not used for the collective movements of the camp. The exterior hearths cannot be interpreted in a univocal manner either. Although N20 and J18 are collective areas open to both families, this is less sure for G13, which could be a simple extension of P15. Unit S25 is excluded from movements around the camp.

28Finally, the role of children in the camp also contributes to our understanding of the social sharing of the living space. We find traces of their presence on the whole known habitat surface, except in S25, where they are more discreet. Children were not confined to a specific place and seem to have moved around freely. Let us recall, however, that unit U5 is an exception to this rule, where the inside of the tent was strictly divided between good knappers in the centre and young knappers around the periphery of the shelter.

29Although this reconstitution of the camp is truncated, it nonetheless reveals a complex social organisation. Numerous elements bring to light a hierarchy between the two family groups. The U5 family held a dominant position in the technical and economic sphere. The most experienced knapper(s) worked in this tent. Naturally, he worked the best and largest blocks of flint and obtained the longest series of blades for the whole group. The size of the blades absent from the refitted cores from U5 implies that this knapper (or these knappers) may have supplied good blanks to another family (or even several), as well as to the family living in P15. The competence of this knapper, resulting from a long apprenticeship, was probably a source of prestige in the whole community. It is also in U5 that the best hunter (the very good knapper or another person?) prepared and repaired hunting tools for the whole group. Unfortunately, we have no information concerning the sharing of game due to the poor conservation of fauna in this level. All of these data confer a specific status on the family in U5, or at least on one of its members. What can we say about the family living in P15? The fact that activities were less intensive in this dwelling may be related to a shorter occupation duration and perhaps also to a smaller number of occupants. But that does not account for the absence of very good debitage or the importance of clumsy individuals. Nor does it explain the scarcity of armatures abandoned near the hearth which reveals very poor investment in the preparation of hunting expeditions. On the other hand, the elaborate construction and regular maintenance of the hearth, in addition to several faunal elements, suggest the non-negligible importance of domestic tasks, including the onsite preparation and consumption of meals.

30A certain form of hierarchy thus existed in the Magdalenian community of Étiolles, along with mutual aid and sharing behaviours, in addition to the classical sphere of hunting and game sharing. The economic autonomy often highlighted for hunter-gatherers operates at a nomadic group level and not for the individual group components.

31The consideration of the duration of the occupation adds an additional element of complexity to our understanding of the Magdalenians of Étiolles, as the occupation of the main dwelling, U5, is divided into two main periods. The links established between the units show that the known part of the camp, with its two tents and annex hearths, was set up at the beginning, although the situation is less sure for hearth S25. The data are less explicit for the second period of the camp. The question arises as to whether the camp represents a continuous and long stay or whether the settlement was momentarily abandoned. The location chosen for dwelling U5, at the top of the bank, may have been sheltered from the highest floods and enabled the Magdalenians to remain at the site for several months. The hypothesis of a seasonal change to explain behavioural modifications, in particular knapping inside or outside, in relation with the transition between a rigorous and a more clement phase, does not seem to be in accordance with the use of open-air hearths from the beginning of the settlement onwards. In addition, the rare data on seasonality supplied by the fauna are ambiguous. Regardless of the occupation season or occupation continuity, a major change occurs during the stay: the abandonment of dwelling P15. This naturally raises the question of what happened to the occupants. Although we have no proof, the departure of part of the community is in keeping with numerous ethnographic observations highlighting the flexibility of nomadic societies characterised by a series of periods of division and aggregation during the annual nomadic cycle.

32Ultimately, the analysis of the camp of Étiolles thus generates new questions. However, the approach adopted here enhances our understanding of this Magdalenian community and enables us to grasp some of the structuring social relations. By multiplying comparisons with other occupation levels (Caron-Laviolette et al. 2018, Pigeot 2004), our knowledge of the Magdalenians of Étiolles is refined owing to the extensive habitat of this exceptional site.

Haut de page


Caron-Laviolette E., Bignon-Lau O., Olive M. 2018 – (Re)occupations: Following a Magdalenian group through three successive occupations at Étiolles, Quaternary International, 498, p. 12-29.

Olive M. 1988 – Une habitation magdalénienne d’Étiolles. L’unité P15, Paris, Société préhistorique française (Mémoire 20), 2 vol.

Olive M. 1992 –En marge des unités d’habitation d’Étiolles : les foyers d’activité satellites, Gallia Préhistoire, 34, p. 85-140.

Pigeot N. 1987a – Magdaléniens d’Étiolles. Économie de débitage et organisation sociale, Paris, CNRS Éditions (Suppl. Gallia Préhistoire 25), 168 p.

Pigeot N. 1987b – Éléments d’un modèle d’habitation magdalénienne (Étiolles), Bulletin de la Société préhistorique française, 84 (10-12 [E&T]), p. 358-363.

Pigeot N. 2004 – L’apport de l’unité Q31 dans l’élaboration de modèles culturels. De la palethnologie à la paléohistoire, in Pigeot N. (dir.), Les derniers Magdaléniens d’Étiolles : perspectives culturelles et paléohistoriques (l’unité d’habitation Q31), Paris, CNRS Éditions (Suppl. Gallia Préhistoire 37), p. 255-266.

Haut de page

Table des illustrations

Titre Fig. 2 – The camp of dwellings U5 and P15
Crédits CAD M. Olive and N. Pigeot
Fichier image/jpeg, 714k
Titre Fig. 60 – All the movements between the different camp units
Crédits CAD M. Olive and N. Pigeot
Fichier image/jpeg, 693k
Haut de page

Pour citer cet article

Référence papier

Monique Olive, Nicole Pigeot et Olivier Bignon-Lau, « A Magdalenian camp at Étiolles, Essonne (abridged version) »Gallia Préhistoire, 59 | 2019, 47-51.

Référence électronique

Monique Olive, Nicole Pigeot et Olivier Bignon-Lau, « A Magdalenian camp at Étiolles, Essonne (abridged version) »Gallia Préhistoire [En ligne], 59 | 2019, mis en ligne le 29 août 2019, consulté le 08 avril 2020. URL :; DOI:

Haut de page


Monique Olive

UMR 7041 Arscan, équipe Ethnologie préhistorique, Maison Archéologie et Ethnologie René-Ginouvès, 21 allée de l’Université, 92000 Nanterre —

Nicole Pigeot

UMR 7041 Arscan, équipe Ethnologie préhistorique, Maison Archéologie et Ethnologie René-Ginouvès, 21 allée de l’Université, 92000 Nanterre

Olivier Bignon-Lau

UMR 7041 Arscan, équipe Ethnologie préhistorique, Maison Archéologie et Ethnologie René-Ginouvès, 21 allée de l’Université, 92000 Nanterre —

Haut de page

Droits d’auteur

Gallia Préhistoire

Haut de page
  • Logo Maison des sciences de l'Homme Mondes
  • Logo Logo CNRS
  • Logo Ministère de la Culture
  • OpenEdition Journals