Abstract

The following describes one distinct sense of ‘mechanism’ which is prevalent in biology and biomedicine and which has important epistemic benefits. According to this sense, mechanisms are defined by the functions they facilitate. This construal has two important implications. Firstly, mechanisms that facilitate functions are capable of breaking. Secondly, on this construal, there are rigid constraints on the sorts of phenomena ‘for which’ there can be a mechanism. In this sense, there are no ‘mechanisms for’ pathology, and natural selection is not a ‘mechanism of’ evolution, because it does not serve a function.