A causal interpretation of selection theory

The following dissertation is an inferentialist account of classical population genetics. I present the theory as a definite body of interconnected inferential rules for generating mathematical models of population dynamics. To state those rules, I use the notion of causation as a primitive. First, I put forward a rule stating the circumstances of application of the theory, one that uses causal language to pick out the types of entities over which the theory may be deployed. Next, I offer a rule for grouping such entities into populations based on their competitive causal relationships. Then I offer a general algorithm for generating classical population genetics models for such populations on the basis of what causal influences operate within them.;Dynamical models in population genetics are designed to demystify natural phenomena, chiefly to show how adaptation, altruism, and genetic polymorphism can be explained in terms of natural rather than supernatural processes. In order for the theory to serve this purpose, it must be possible to understand, in a principled fashion, when and how to deploy the theory. By presenting the theory as a system of ordered inferential rules that takes causal information as its critical input and yields dynamical models as its outputs, I show explicitly how classical population genetics functions as a non-circular theoretical apparatus for generating explanations. The generalization of the theory achieved by presenting it using causal vocabulary shows how the scope of the theory of natural selection extends beyond its traditional domain of systems of individuals distinguished by genetic variations.