SimFeodal: an agent-based model to explore the combined effects of social and demographic changes on the hierarchy of rural settlement patterns in North-Western Europe during the Middle Ages

International audienceIn North-Western Europe, regional settlement patterns that were dispersed in 800 CE became much more concentrated and hierarchical in 1200 CE (Tannier et al. 2014). This phenomenon occurred in all regions but the resulting level of concentration and hierarchy of settlement patterns differed notably among the regions. Several processes jointly explain this major transition (Cura et al. 2017). •The dismantling of the Carolingian Empire and the dissipation of powers induced struggles among lords and thus a rise in violence. The result of this was the creation of castles as well as an increase of the need of protection for peasant households•The militarisation of the society mentioned previously also intensified the need of lords for money as well as their need of fighters. From this originated the so-called “feudal revolution” of the 11th century during which ever smaller lords appropriated various administrative, fiscal and judicial rights for themselves and their vassals. As a result, usage fees and rents paid by peasant households to the lords increased a lot.•Peasant households created villager communities to increase their productivity and counterbalance the power of lords. •The religious control of the society increased a lot (especially as a result of the Gregorian reform). From this resulted a great upwelling of piety and level of devotion, an increase of the number of parish churches, and the increase of usage fees and rents paid by peasant households to the Church.At a regional analysis level, how all these processes reinforce or counterbalance each other and finally increase the concentration and the hierarchy of settlement patterns remains still unknown. Our interdisciplinary group, made up of geographers, archaeologists and historians assumes that four variables may have a crucial impact on the level of concentration and hierarchy of a given regional settlement pattern: the number of inhabitants in 800 CE, the number and the size of the villages in 800 CE, the demographic growth during the whole time period, and the proportion of peasant households being so highly dependent on their lord that they can not leave his grounds (serfs, slaves).The agent-based model SimFeodal (https://simfeodal.github.io/) has been created with the GAMA platform (Taillandier et al. 2018) to explore the combined effects of these variables on the relocation of agricultural holdings (farms) in the course of time. A simulation begins in 800 CE and ends in 1200 CE. A simulation step represents a time length of twenty years, which corresponds to the average life duration of a generation at the time period under consideration, thus a simulation goes on for twenty simulation steps. The time length of a simulation (400 years) is long enough to be able to verify if simulated dynamics are lasting or not. The main simulation result is the progressive appearance of enduring population clusters located around the castles and the churches. Population clusters (hamlets, villages, small towns) become more numerous; many of them grow (i. e. their number of peasant households increases as the agricultural holdings concentrate within them) but some clusters grow more than others. The aim of our presentation is to show how the values of the four variables quoted above influene the simulation results, especially the rank-size distribution of population clusters. Thus we expect to better understand the combined effect of each variable on the spatial dynamics of rural settlement systems in North-Western Europe during the Middle Ages