Several other commentators capably articulate and defend an important objection to Christie, Brusse, et al.: the claim that the existence of a trait is entirely, rather than partially explained by the effects for which it was selected is stronger than the selected effects theorist needs or seeks to defend. Nonetheless, Christie, Brusse, et al.’s cases do draw our attention to a point about the explanatory relation between selected effects functions and their bearers that has not been emphasized by selected effects (...) theorists, although some have quietly accommodated it. This is the idea that, for a trait T to be for performing F, T’s ability to perform F must figure in an explanation of T’s existence in a particular way. It is not T’s absolute propensity to produce F as an effect but its comparative propensity, as compared with alternatives to T, that must figure in the explanation. The pattern of explanation sought by selected effects theorists to underwrite function attributions therefore requires a correct choice of comparison class. When we spell this idea out, we arrive at a hypothesis about what is bothering Christie, Brusse, et al. about the cases they find problematic. Consider their Gouldian finches. Focusing on the role of comparative propensities, the difficulty becomes clear: the claim that the red-heads’ hormonal levels have persisted in the finch population because they have had the effect of dislodging competitors more reliably than the black-heads’ hormonal levels is incorrect. The red-head phenotype doesn’t stick around by outdoing the black-heads at dislodging, since the black-heads can’t be outdone at something they are not in the business of doing in the first place. However, this point is not in the end a serious problem for selected effects theories because the problem comes from a wrong choice of comparison class: namely, black-heads, rather than inferior red-heads. If the hormonal trait of today’s red-heads reached or persists at its present frequency because of some historical selection against other, perhaps quite similar, hawkish variants, this trait will be for dislodging competitors, since it is around because it outdid those alternatives at dislodging. (shrink)

Christie, Brusse, et al. argue that selected effects are insufficient to explain the prevalence of traits when selection is heterogeneous. One could object that it’s useful to ground functions in selected effects so long as selected effects are necessary to explain the prevalence of traits. This raises a challenging question: what justifies singling out selected effects from other factors that are necessary to explain the prevalence of traits when selection is heterogeneous? I consider three answers: selected effects are the only (...) factors that explain the prevalence of traits and (a) can plausibly ground functions; (b) enable biological generalization; and (c) enable representational explanation. I argue that none of these answers is satisfactory. Only the third answer might be able to justify singling out any selected effects at all (when selection is heterogeneous): the small share of selected effects that ground representations of causal particulars. Thus, I agree with Christie, Brusse, et al. that heterogeneous selection is a serious problem for selected effects theories of function. (shrink)

In the target article for this issue, Christie, Brusse, et al. [2022a] argue that Selected Effects Functions (SEF), at least as currently articulated, often do not explain biological traits. In res...