Groundwork of the Metaphysics of Models

Abstract

In the context of evolutionary biology, reciprocal causation is the increasingly popular view that adaptive evolution is a bidirectional process. Proponents of reciprocal causation argue that natural selection and the actions of individual organisms are causally interdependent. During cycles of reciprocal causation, two different types of processes interact: the process of natural selection occurs when causation flows from the environment to the organism, and processes such as the construction or modification of one’s own niche, for example, occur when organisms causally impinge on their environment. In this way, the processes of natural selection and niche construction are both causes and effects of each other.Using the tools of “experimental philosophy,” my dissertation surveys the views of practicing biologists on models of reciprocal causation in evolutionary biology. I present quantitative survey data from faculty members in biology departments at universities across the United States to evaluate the model of reciprocal causation. The survey data indicate that most of the participants do not agree (i.e., most either disagree or neither agree nor disagree) that reciprocal causation confers a larger advantage on research practices than its conceptual competitors. However, most of the participants agree that reciprocal causation more accurately represents the structure of the biological world. These results demonstrate that the explanatory merits of a conceptual framework and its practical utility can come apart in interesting and informative ways. After considering the quantitative data, my dissertation tackles reciprocal causation using the more traditional tools of analytic metaphysics. I argue that reciprocal models of adaptive evolution ultimately face an unforgiving dilemma. The dilemma is due to what I call the “emergent character of natural selection.” By this I mean that natural selection is an emergent process—it plays out in ensembles, not individuals. An evolving population is a complex biological system. It is an aggregate that exhibits unique patterns, behaviors, and processes. Furthermore, these emergent processes are multiply realizable by infinite combinations of causal interactions between organisms and their environment. For instance, a population undergoing natural selection may change in predictable ways, and these predictions may be afforded even when we do not have epistemic access to events that occur at the individual level. I argue that the emergent character of natural selection threatens the concept of reciprocal causation—specifically the claim that the concept captures a causal interdependency between selection and the behaviors of individual organisms. This is especially salient, given the increasingly accepted view that organisms act as causal “agents” in their own evolution. If selection is an emergent phenomenon, then reciprocal causation depicts individual-level processes (such as the construction or modification of one’s evolutionary niche) as causes and effects of a population-level process (natural selection). The problem is that individual-level processes and population-level processes do not stand in a causal relation, but a compositional one. Because natural selection emerges from the underlying actions of organisms, their relationship fails to meet an important criterion for causal relationships—namely, that causes are spatiotemporally distinct from their effects. To say that natural selection and niche construction co-cause one another is therefore to commit a category mistake.