Putting History Back into Mechanisms

Abstract

Mechanisms, in the prominent biological sense of the term, are historical entities. That is, whether or not something is a mechanism for something depends on its history. Put differently, while your spontaneously-generated molecule-for-molecule double has a heart, and its heart pumps blood around its body, its heart does not have a mechanism for pumping, since it does not have the right history. My argument for this claim is that mechanisms have proper functions; proper functions are historical entities; so, mechanisms are historical entities, too. This thesis runs against the mainstream new mechanist way of thinking about mechanisms, where mechanisms are generally thought of in an ahistorical way. After arguing for this thesis, I draw out some consequences for philosophy of science and metaphysics.

The social epistemology of introspection

“…On mechanistic explanation in cognitive science, seeBechtel (2007),Craver (2007), andGlennan (2017). On the notion of function in mechanistic explanation, seeGarson (2013Garson ( , 2017Garson ( , 2022 andNeander (2017). For more discussion on the relevance of both to pragmatic competence, seeUnnsteinsson (2022).4 This is a slight modification of the standard terminology in relevance theory(Carston, 2002;Sperber & Wilson, 1986;Sperber & Wilson, 1995;Wilson & Sperber, 2012), where "effective" is "informative" and "signaling" is "communicative".…”