Armstrong’s Revenge

This chapter distinguishes two forms of functionalism: realizer functionalism, defended by Armstrong and Lewis, and role functionalism, dominant in the US. Realizer functionalism identifies mental states with the causally efficacious states that play functional roles, avoiding issues about mental causation that arise for role functionalism, where mental states are higher-order states of having a state that plays a role. The author modifies realizer functionalism by incorporating an account of dispositions from C. B. Martin, where dispositions are identified with their categorical bases. This modification treats mental states as first-order dispositions with profiles matching functional specifications, potentially supporting a straightforward identification of experiential states with neurological states without previous pitfalls.

Study at a glance

Abstract

This chapter distinguishes two species of functionalism: what Ned Block called, ‘functional specifier functionalism’ (now often called ‘realizer functionalism’) and ‘functional identity functionalism’ (now often called ‘role functionalism’). The first species of functionalism was defended by Armstrong and by David Lewis, and remains the dominant strain in Australian philosophy of mind. The second species became dominant in the US. A range of problems in the philosophy of mind over the last forty years stemming from the latter’s dominance are discussed. Most notably, issues about mental causation do not arise for Armstrong and Lewis because mental states literally are the causally efficacious states that play functional roles. In contrast, on the second view, mental states are higher-order states, such as ‘the state of having a state that plays a particular role’ the efficacy if which is dubitable. Heil argues that problems for the former view can be dealt with by modifying it to incorporate an account of dispositions traceable to C. B. Martin, according to which dispositions are identified with their categorical bases. Mental states are first-order states, not abstract states of being in some state that satisfies a particular functional specification, but dispositions with dispositional profiles that correspond to functional specifications. The chapter concludes by arguing that this modification has the potential to support a straightforward identification of experiential states with neurological states that avoids the pitfalls of previous attempts to do this.