an NSF supported program

Village Ecodynamics Project


The late A.D. 1200s depopulation of the Mesa Verde region of the American Southwest is one of the great mysteries of American archaeology. Deserted Cliff DwellingsMany mechanisms have been proposed to account for this rapid out-migration of regional populations. Most suggest increasingly severe resource imbalances across a densely populated landscape. Some accepted research, however, shows that potential maize production was sufficient to support the estimated populations of the time. If these populations emigrated due to resource scarcity, then scarcity of other resources must have contributed to decisions to leave. On the other hand, there are hints of important changes in sociopolitical organization just prior to the depopulation.

This famous depopulation is one of the riddles that the Village Project addresses. The project was undertaken to examine the interaction of simulated agrarian households with their natural environment taking in to account the production and consumption of various natural resources essential for everyday life. By evaluating the possibility of crises in factors such as potable water, woody fuels, and protein, this research will help determine whether resource factors were in fact critical in these decisions, or whether social factors may have largely influenced the exodus.

See our Research Plan for other problems we are addressing.

Turkey domestication in VEP model

Please note that this content is not current. It is being maintained for historical purposes.

Kyle Bocinsky, a first-year graduate student working as a research assistant to Tim Kohler, has been busy building turkey use into the VEP simulation. Modeling turkey has two components: first, Kyle is generating wild turkey on the landscape, using diffusion algorithms congruent with those already in place for deer. Turkey have a different preferred browse than deer, rabbit, or hare, and will hypothetically choose to exist in different habitats than the other fauna. This component should be completed in early May.

The second component is generating a mechanism by which agents in the simulation may choose to keep and breed turkey (i.e., turkey domes–tication). A first step of generating this module—an in-depth look at costs and payoffs associated with keeping turkey—was completed by Kyle in Fall 2008. The next phase consists of exploring how to model the transmission and adoption domestication practices. This may be done in two primary ways: either agents are always aware of domestication as a strategy, with costs and payoffs, and simply choose that strategy once other strategies prove less effective, or domestication practices (and knowledge) spreads through populations via a transmission algorithm, such as Kobti and Reynold's Cultural Algorithms, which are already implemented in other parts of the model. The use of these for directing the transmission of domestication practices is currently under investigation.

Of course, the mechanisms for spread of turkey domestication practices are hardly understood in the Southwest, and the VEP stands to make great contributions in this regard.  Perhaps both transmission and adoption mechanisms will be modeled; these could be toggled as a parameter, thus allowing VEP researchers to explore whether one matches evidence of turkey domestication practices in the archaeological records with more fidelity.

Kyle is also in the tedious process of documenting the v2.8 VEP simulation codebase, and preparing it for public release, as well as reassessing hunting in v2.8 for the forthcoming VEP I final report.

Agent-based Modeling

Since the 1990s there has been a marked increase in interest in computational approaches—including simulation—by social science researchers. This appears to be driven both by a cross-disciplinary interest in the sciences of complexity and the ever-increasing computational capacity at our disposal.

In the past, due to the complexity of the phenomena involved, we have been forced to use simplistic world models. Today we are able to study a world in which most important phenomena emerge from the non-linear interaction of many agents (physical, biological, or social) in systems that are rarely at equilibrium.

This vision promotes a method—agent-based modeling—that provides a computational environment in which the behaviors of such systems can be studied.

Read more

Department of Anthropology, PO Box 644910, Washington State University, Pullman WA 99164-4910, 509-335-3441, Contact Us