I will begin my talk with a quick introduction to causal modeling, which frames how I think about problems and their solutions in evolutionary biology. Then I will move to thinking about causal structures in group selection, and how these are typically represented. What is nearly universally held by group selection researchers is that group selection requires fitness affecting interactions between individuals. However, typically we choose to model these interactions with aggregative, group-level variables. So, aggregate fitness is modeled as a function of an aggregate trait variable. The primary aim of my talk is to assess how such a modeling strategy fares given certain predictive aims and given that the world really is governed by fitness affecting interactions rather than aggregative variables. To do so I create a simulation involving fitness affecting interactions among healthy and cancer cells. Cancer research is increasingly filled with theoretical models of a Darwinian and multi-level selectionist stripe. The bulk of my talk will be dedicated to describing the simulation and its behavior as we change the range of fitness affecting interactions. Finally, I will reveal some preliminary results of testing the common aggregative representational strategy against this simulated ``truth.” For certain predictive aims, the strategy does pretty well. For others, it is not only inaccurate, on average, but woefully imprecise. Specifically, if we care about ``regime shift” predictions or predictions that are the result of interventions (as is the case in cancer research), I think these preliminary results provide some reason to rethink how we model fitness affecting interactions.
Wes Anderson did his undergraduate and masters studies in philosophy at Portland State University and the University of Wisconsin, Milwaukee, respectively. He has always been interested in causation, specifically causal inference and representation, in the biological sciences, and he recently defended his dissertation on the causal theory of natural selection within Arizona State University’s history and philosophy of science program. He is now a postdoctoral researcher at the KLI, where he is extending this research. In particular, he has been thinking about certain causal structures like niche construction in a demographic setting and how to make inferences from observational data to inter-individual causation.