ANR Obresoc

UMR Trajectoires was a partner of the ANR Obresoc, with Jean-Paul Demoule and Jérôme Dubouloz

OBRESOC (A retrospective observatory of an archaeological society: The trajectory of the LBK Neolithic), is a project funded by the ANR (Agence Nationale de la Recherche, Convention ANR-09-CEP-004-01/OBRESOC) as part of the ANR- CEP (Global Environmental Change) programme. It brings together seven partner teams in a transdisciplinary framework, including archaeobotany, archaeozoology, palaeodemography, palaeoenvironment, palaeoecology, cultural archaeology, palaeoclimatology, economics and computer modelling).

The expansion of the agricultural system from Anatolia to the Balkans has long been an important research topic. Classical archaeological approaches have enhanced our understanding of the course and timing of this expansion, but they do not shed light on anything that is not directly observable in the archaeological record: namely, the socio-economic structure of a prehistoric agricultural society. In this thesis, a multi-agent model was used to explore elements that are invisible to archaeology. This model, called BEAN (Bridging European and Anatolian Neolithic), is an adaptation of the OBRESOC (Un OBservatoire REtrospectif d’une SOCiété archéologique) model. OBRESOC was created to simulate the expansion of LBK farmers in Central Europe, and has been modified to adapt it to the Balkan archaeological context. The expansion of early Balkan farmers is simulated by combining archaeological data with ethnohistorical and palaeodemographic inferences. A realistic environment has been modelled, where optimum agricultural areas are determined by assessing meteorology and soil fertility. Each agent corresponds to a domestic household; agents interact in this environment by following partial socio-economic intermediary models that determine the rules of their behaviour. For example: houses with nuclear families; intensive small field agricultural system with complementary hunting and gathering; expansion determined by village scalar stress; solidarity networks between relatives; food shortages and famines caused by weather events. In this way, the model simulates the functioning and geographical expansion of a Neolithic agricultural society. Many simulations were carried out, varying the most important parameters, which were identified with sensitivity analysis. The adhesion between archaeological data and simulated data was measured mainly with geographical criteria and the simulation generating the simulated expansion pattern that best coincides with archaeological expansion is considered the best. Specific procedures were created to handle the huge quantity of simulated data produced by the model. Observation of these data led to the exploration of certain aspects that are invisible in archaeology; for example, the model helped to question archaeological beliefs based on assumptions that were not otherwise testable. The model also enabled us to explore other topics, such as the comparison between the pioneer settlement front and areas of former occupation, and the influence of meteorology on the expansion of the agricultural system. The model produced expansion patterns that are geographically and chronologically consistent with the expansion suggested by the archaeological record. The exploration of socio-economic outputs has resulted in new hypotheses that could not be made merely on the basis of what is found in the archaeological record. Even when there is a large gap between what is observed in archaeology and what is produced by the model, this multi-agent modelling approach opens up new questions, adding new ideas and perspectives to current research.