1The many stone-heating combustion structures in the Seine-Eure confluence region (Normandy) are used here as a guiding thread in a study of Neolithic occupations of riverbanks and valley bottoms. In these environments, preventive archaeology has revealed sites with features unknown in settlements, over extensive spatial and temporal scales. These include the absence of architectural remains, the scarcity of pottery and certain types of tools, numerous windfalls, stone-heating structures and lithic industries. Most of the data presented here come from excavations carried out at Louviers “Avenue Winston Churchill” and Alizay “Le Postel”. These two sites comprise hundreds of stone-heating structures, also identified in a wide variety of other contexts in north-western France, such as settlements, riverbanks, coastlines, megaliths, and so on. They provide data often considered to be of limited value (variable conservation, few or no artefacts). However, the successive study of the different components of these structures reveals forms of Neolithic social organisation that are rare or unknown in settlements. They indicate a range of more or less specialised activities, organised in keeping with the natural environment and the social and cultural context, where stone-heating structures appear to be central elements.
2This work is based on two main lines of research:
- Investigating the status and function of stone-heating structures in Neolithic societies, by successively examining the intra-site structural features (conservation, stratigraphy, associated objects, anthracological data, residue analyses) and macroscopic features (distribution, contexts, associated sites) that govern how they operate.
- Defining the chronocultural context(s) in which these structures appear and evolve, backed up by an assessment of radio-chronological measurements to multiply the scales of observation over the long term: first on a local scale of the Louviers and Alizay sites, then on a scale of the microterritory formed by the Vaudreuil loop, then on a much wider scale extending over Normandy and north-western France.
3In this study area, Neolithic stone-heating structures present several common denominators; a circular or oval plan, generally between one and two metres in size (fig. 1), a flat-bottomed bowl profile, rare or no archaeological material, and the use of stones collected within a radius of one kilometre, whatever their quality. Morphological variations of the stones are due to their geological context (type of rocks used) or their state of abandonment and conservation. In the Seine-Eure confluence area, they are often spatially associated with abundant lithic industries, in stark contrast to the scarcity of pottery. There is no evidence of any permanent architectural structures. The flint material comes from simple productions, related to temporary activities linked to the exploitation of riverbank resources: flint knapping, clearing or collecting wood, leatherworking, and possibly meat or fish processing. These activities appeared to involve moving a group of individuals from the settlement for a short period of time, and also account for the use of a structure intended for the preparation of a meal eaten together during this collective work. In fact, the various analyses carried out at the Alizay and Louviers sites suggest that the stone-heating structures were used for culinary purposes, probably for stewing food, particularly the meat of land-based animals.
Fig. 1 – Planimetric representations of a selection of six stone-heating structures from Louviers “Avenue Winston Churchill” (Eure, Normandy).
CAD V. Théron
4Archaeologists who have worked on the subject over the last fifteen years in north-western France have therefore reached a relative consensus regarding the function and operation mode of these structures, backed up by a substantial and well-documented corpus. These interpretations can also be compared with models documented by ethnography, particularly in Oceania, where Polynesian ovens operate in a similar way (Ghasarian 2004, Serra Mallol 2014, Hart 2024).
5On an intra-site scale, whether at Louviers or at the majority of the Neolithic occupations considered, the distribution of stone-heating structures does not reflect any particular organisation and was probably dictated by expedient logic, or possibly physical landscape constraints (vegetation?) at the time of their installation. However, concentrations of stone-heating structures are sometimes observed near buildings, at Saint-Vigor-d’Ymonville (Marcigny et al. 2002), or at Pléchâtel (Juhel 2014), for example.
6Nevertheless, the chronocultural distribution of these structures is telling (fig. 2): it reflects a trend that clearly began at the end of the Early Neolithic, from 4800 BC onwards, and gradually developed over the following half millennium, i.e., during the first half of the Middle Neolithic. The peak of their use more or less coincides with the transition between the Middle Neolithic I and II, around 4300 BC. Their number then seems to decline gradually until around 3800 cal BC, and these structures become rare or minor after that date. These are therefore clearly transchronological and transcultural features: from the Blicquy/Villeneuve-Saint-Germain, to the Cerny and Northern Chassaean, or ancient and recent Castellic in Armorica, they are long-term structures, occurring in diverse cultural entities.
Fig. 2 – Correlations between climate and cumulative and weighted curves of 14C dates obtained on stone-heating structures in Normandy and Brittany.
CAD M. Magny, modified C. Marcigny, after Stuiver et al. 1998.
7This ubiquity does not apply to archaeological contexts. Stone-heating structures may occasionally be associated with dwellings or monumental contexts, but such examples are not systematic in Brittany, and are rare in Normandy, where most structures are identified at some distance from the most emblematic sites of the period. They are mainly linked to the occupation of areas with abundant multiple resources that are not naturally suitable for dwellings, such as coastlines and valley bottoms.
8As a result, although most of them appear to be spatially disconnected from the other major events of the period, in chronological terms their development is perfectly synchronised with that of enclosures or monumentalism in general, but also with mining, for example (Charraud 2013), or the practice of silage making (Charraud and Ghesquière 2023). From 4700 BC onwards, the development of such multiple evidence of collective social structures coincided with a phase of recomposition of cultural entities (Marcigny and Mac Fadyen 2016) and the regionalisation of exchanges (Augereau 2005). This general dynamic, which did not occur at the same time everywhere and probably emerged quite late in the west (Bostyn et al. 2019, Charraud 2022), is undoubtedly the sign of restructuring communities around the in-depth exploitation of the resources of the newly-cleared lands in which they were established.
9Our understanding of Neolithic societies is therefore not confined to settlements, which yield the most palpable archaeological evidence. On a macro-economic scale, in a logic of complementary occupations, stone-heating structures bear witness to the exploitation of specific environments by Neolithic communities, which were not permanently occupied, but are nonetheless part of the same overall process of Neolithisation.
10Within this general movement, the status implications perceptible through the study of stone-heating structures appear to consolidate this macroeconomic model of the use of valley bottoms. From a technical point of view, the lithic assemblages from Alizay and Louviers display features specific to these environments, already observed at Muids and Bouafles (Prost 2015). They clearly correspond to occasional (seasonal?) work, either for acquiring resources (water, wood, flint, hunting, etc.) or for carrying out activities causing disturbances, for which it was preferable to work at some distance from settlements (flint knapping, leatherworking, etc.).
11Owing to their number and density, stone-heating structures hold a central place in these temporary occupations, and are probably evidence of the collective meals that accompanied these activities. In the various Oceanic models detailed in this work, groups of ovens comparable to those observed in the Neolithic corpus are linked either to seasonal practices (harvesting, hunting migratory animals), which can be more or less spread out over time, or to very large festive gatherings during which several dozen structures can be used simultaneously, at trade markets or customary festivals.
12We do not wish to simplistically transpose an ethnographic model to an archaeological case, but it is important to recall that these structures are part of a complex social framework. In all the examples documented by ethnography, the social principle of harnessing resources (whatever they may be) is never a matter of individual action; it is always based on collective efforts (Pétrequin and Pétrequin 1993). This cooperation can take place within the same village, where all available forces can be put to work during a collective “chore”.
13The stone-heating structures around the fire, and the communal meal prepared there to nourish the group and reinforce social bonds, probably characterise forms of Neolithic social organisation specific to particular environments and unheard of in other contexts. These environments provide a glimpse of an entire section of the Neolithic technical system that developed at the crossroads of a global territorial management macroeconomic model, and a complex interplay of social and environmental interactions.