The proximate-ultimate distinction and evolutionary developmental biology: causal irrelevance versus explanatory abstraction

Mayr's proximate-ultimate distinction has received renewed interest in recent years. Here we discuss its role in arguments about the relevance of developmental to evolutionary biology. We show that two recent critiques of the proximate-ultimate distinction fail to explain why developmental processes in particular should be of interest to evolutionary biologists. We trace these failures to a common problem: both critiques take the proximate-ultimate distinction to neglect specific causal interactions in nature. We argue that this is implausible, and that the distinction should instead be understood in the context of explanatory abstractions in complete causal models of evolutionary change. Once the debate is reframed in this way, the proximate-ultimate distinction's role in arguments against the theoretical significance of evo-devo is seen to rely on a generally implicit premise: that the variation produced by development is abundant, small and undirected. We show that a "lean version” of the proximate-ultimate distinction can be maintained even when this isotropy assumption does not hold. Finally, we connect these considerations to biological practice. We show that the investigation of developmental constraints in evolutionary transitions has long relied on a methodology which foregrounds the explanatory role of developmental processes. It is, however, entirely compatible with the lean version of the proximate-ultimate distinction