A massacred village community? Agent-based modelling sheds new light on the demography of the Neolithic mass grave of Talheim

Abstract

The virtual experiments presented below reveal the counterintuitive archaeological demography of the Neolithic mass grave of Talheim and underline the importance of distinguishing between the demographic structures of living and dead populations, as well as between attritional and catastrophic mortality patterns. We utilise a new agent-based modelling approach called Population & Cemetery Simulator based on the NetLogo programming language and the Behaviour Composer of the modelling4all project, which allows us to extrapolate from dead to living populations and vice versa. Contrary to received opinion, we argue that the population of the Neolithic mass grave holds specific demographic information only, as it represents a pure catastrophic mortality pattern, i.e. a living population at a single point in time rather than the population of a conventional cemetery. The first experiments illustrate why the published demographic data (e.g. mortality, life expectancy, mean age at death) is misleading. It is illogical to utilise mortality tables devised for conventional (attritional) cemeteries in the case of living populations. Modelled populations with the published mortality rates of the massacre site are, furthermore, unable to stand up to plausible human demographic circumstances. In the second part, we evaluate the actual demographic information content of the Talheim sample. Comparative modelling illustrates that the Talheim population appears to be similar to possible living populations based on the mortuary record of Schwetzingen, an isochronal site of the Linear Pottery Culture (LBK), and Bärenthal, a site which dates back to the early medieval period (7th to 10th centuries). It is therefore very likely that the Talheim population is a representative sample of a living population in the LBK and might even represent a massacred village community in its entirety.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 12 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.