A Successful Theory of Mental Representation

2355 Words5 Pages

A Successful Theory of Mental Representation A successful theory of mental representation must answer two related questions: (1) How does representation work? (2) How is misrepresentation possible? These questions are related because answering (1) is a requisite for answering (2): in order to explain how it is is possible that cognitive systems produce erroneous representations (i.e. representations that do not correspond to their "proper" meanings), we must first explain how is it possible that cognitive systems produce representations at all. Moreover, in the contemporary philosophical scene, it is an additional requirement that both questions must be given naturalistic answers, i.e. answers which ultimately account for intentional phenomena in nonintentional terms. In this essay, I will briefly expound and contrast the theories advanced, in order to answer these questions, by three main contemporary philosophers of mind: Fred Dretske, Jerry Fodor, and Ruth Millikan. In his essay "Theory of Content I" Jerry Fodor presents the antecedents of his own theory of representation by referring to the work of the behaviourist psychologist B.F. Skinner. According to Skinner's theory, and following standard behaviourist dogma, semantical facts are to be explained in terms of behavioural dispositions. Thus the fact that, e.g. the (mental or linguistic) representation 'dog' expresses the property being a dog can be reformulated as a subject's production of the representation 'dog' is under the control of a certain type of discriminative stimu... ... middle of paper ... ... would mean mouse, not mouse-or-shrew, and 'mouse' representations tokened by shrews would be asymmetrically dependent on 'mouse' representations tokened by actual mouses. As Robert Cummins has shown, this can be expressed in the respective truth and falsity of the following counterfactuals: (1) If mice didn't cause 'mouse's, then shrews wouldn't either. (True) (2) If shrews didn't cause 'mouse's, then mouses wouldn't either . (False) But, as Cummins argues, if mice and shrews are indistinguishable, there is no way for statement (1) to be true while (2) is false: the same state of affairs that makes (1) true makes (2) also true. In this way, shrew-tokened 'mouse's cannot be regarded as asymmetrically dependent on mouse-tokened 'mouse's, and thus cannot see 'mouse' as meaning mouse (and not mouse-or-shrew).

Open Document