Digital history & historiography

Using environmental predictive settlement choice models as input data for settlement pattern simulations

Inductive models of archaeological site locations have been successfully used for predicting archaeological potential of places in landscapes. These models are mostly based on currently observable environmental information. To reduce environmental determinism and increase both explanatory and predictive power several variables like visibility of locations have been interpreted as social factors of settlement locations.
In the current paper we explore the possibilities of using inductive environmental models as input to simulation models. Although similar to models created for predictive purposes they need to be designed with different considerations.
We present a study where we use inductive models of archaeological site locations to describe the spatial configuration of space environmentally suitable for residence. To do so we develop a conceptual agent-based model of residential choice based on discrete choice theory and theories of residential choice used in multiple fields from archaeology to contemporary urban studies.
We discuss the role of environmental influences as perceived in archaeological data and how they relate to social influences and historical processes leading to emergence of settlement patterns. We argue that spatial structures of the inductive models of specific settlement patterns can inform us about the causal processes behind them when experimented with agent-based simulations.
We present case study using inductive models of settlement locations from different periods of the Stone Age of Estonia. The differences of inductive settlement choice models and the ways of comparing them are discussed. The spatial configurations of the models of economic modes have different structures. For example region where settlements of water connected hunter-gatherers can be found has a different spatial structure than that of early agrarian communities. Those differences give insights on socioeconomic histories and can be used in explaining settlement pattern formation processes.

Show this publication on our institutional repository (