A Case for Epistemology- and Context-Driven Accounts of Cognitive and Biological Functions

Philosophers tend to focus on the metaphysics of functions: establishing unifying theories employing general criteria for being a function, avoiding spooky backward causation, distinguishing functions from accidents, and correctly representing the functional structure of the world. We show that there is a need for localized, practice-sensitive accounts of the epistemology of functions—accounts that explain the identification, justification, and explanatory applications of function attributions in particular scientific contexts—and that this need is best met alongside a plurality of unifying metaphysical theories of functions. Our argument begins with the observation that unifying theories assume that (i) functions are discovered or justified by criteria for being a function and (ii) criteria identified in a theory are generally applicable across target systems. We then draw on examples from cognitive science and ecology to show that function ascriptions are sometimes presupposed or assigned based on highly contextual epistemic factors not described in unifying metaphysical theories.