|
Dual-Use Neural Systems and Theories of Mental Representation Theories of mental representation, particularly in philosophy, but also in cognitive science generally, have been strongly influenced both by a largely tacit assumption and by an explicit, often rehearsed argument. I’ll call the assumption "the univocality assumption" or UA. UA asserts that an adequate theory of mental representation must assign univocal, task-independent representational contents to mental states. I’ll call the argument "the Stich-Fodor argument" or SFA. SFA is widely held to demonstrate that one theory of mental representation--functional role semantics--cannot correclty assign representational contents to mental states. The difficulty alleged by SFA seems inescapable: Functional role semantics asserts that a mental state represents an object, property, or event in virtue of the causal relations (functional roles) that that state has with other states and/or the world. SFA alleges that functional role semantics cannot correctly assign representational contents because there is no nonarbitrary means to distinguish content-relevant from content-irrelevant functional roles. In this paper I argue that the brain has numerous "dual-use systems" including the visual cortex (vision and mental imagery) and the sensory-affective-motor mirror neuron system (sensory-motor-affective planning and motor action prediction and interpretation, empathetic response). The brain employs these dual-use systems to perform distinct but relatively isomorphic tasks, task involving different objects, properties, and/or events. Explaining, for example, how humans solve a rotation problem using vision as opposed to mental imagery requires that the brain use elements of these systems to represent distinct objects/properties with the exact same neuronal resources. I argue that current understanding of such systems reveals that the univocality assumption is not, therefore, universally true. Likewise, since the brain seems to mediate between representational contents for these systems in terms of the functional roles between system elements and other brain areas, the Stich-Fodor argument proves unsound. Specifically, it appears that nonarbitrary content-relevant functional role specifications are discoverable through empirical investigation of task-specific functional roles of brain regions. I suggest that the error in SFA lies in its tacit presuppositions: (1) Nonarbitrary content-relevant functional roles are universal and task-independent. (2) Content-relevant functional roles can only be discovered at conceptual level through intuitive or apriori methodology. |