Biological functions are causes, not effects: A critique of selected effects theories.

The theory of Selected Effects (SE) is currently the most widely accepted etiological account of function in biology. It argues that the function of any trait is the effect that past traits of that type produced that contributed to its current existence. Its proper or etiological function is whatever effect was favoured by natural selection irrespective of the trait's current effects. By defining function with respect to the effects of natural selection, the theory claims to eschew the problem of backwards causality and to ground functional normativity on differential reproduction or differential persistence. Traditionally, many have criticised the theory for its inability to envisage any function talk outside selective reproduction, for failing to account for the introduction of new functions, and for treating function as epiphenomenal. This article unveils four additional critiques of the SE theory that highlight the source of its critical problems. These critiques follow from the fact that natural selection is not a form of work, but a passive filter that merely blocks or permits prior functioning traits to be reproduced. Natural selection necessarily assumes the causal efficacy of prior organism work to produce the excess functional traits and offspring from which only the best fitted will be preserved. This leads to four new incapacities of the SE theory, which will be here analysed: (i) it provides no criterion for determining what distinguishes a proper from an incidental function; (ii) it cannot distinguish between neutral, incidental, and malfunctioning traits, thus treating organism benefit as irrelevant; (iii) it fails to account for the physical work that makes persistence and reproduction possible, and (iv) in so doing, it falls into a vicious regress. We conclude by suggesting that, inspired by Mills and Beatty's propensity interpretation, the aporia of backward causation implicit in anticipatory accounts of function can also be avoided by a dispositional approach that defines function in terms of work that synchronously counters the ubiquitous tendency for organism entropy to increase in the context of far-from-equilibrium thermodynamics.