Functional Analysis: Some Recent Philosophical History

1. Introduction. A standard treatment of functional analysis ( functionalism ) among philosophers of science has recently come under attack by a new variety of functionalist in the philosophy of psychology.1 Jerry Fodor, the chief proponent of this new approach, challenges the view that functional statements can be eliminated in favor of causal (nomological) statements. He supports his position by providing the outlines of a functional analysis of mental concepts which he believes illuminates the mind-body problem. In particular, Fodor holds that his version of functionalism clarifies how a materialist may avoid adherence to the reductionist thesis that mental concepts can be eliminated in favor of physical concepts.2 Fodor\u27s criticism of the standard view of functional analysis is blunted, however, in two ways. (a) The parallel between the standard treatment and Fodor\u27s variant breaks down in a crucial manner. Consequently, even if he establishes that functional statements as he construes them cannot be eliminated in favor of causal statements, he has not yet shown that functional statements as construed by the standard view are ineliminable. (b) Many of the virtues of Fodor\u27s account, particularly as it avoids reductionist solutions to the mind-body problem, can be obtained without recourse to the troublesome concept of a function. 2. Fodor\u27s Account. According to Fodor, a complete explanation in psychology consists of two parts: a functional analysis and a mechanistic or casual analysis.3 A phase one theory provides a functional characterization of internal states of an organism such as memories, motives, needs, drives, desires, strategies, beliefs, etc. solely in terms of the way in which they function in producing behavior. Such theories attribute only those properties and degree of complexity to internal states of an organism necessary to account for some part of its behavior and make no reference to neurophysiological conditions or structures.4 An example of such a phase one theory is the use of concepts like memory trace, long-term memory and short-term memory in order to account for human memory evincing behavior.5 Such a theory does not provide a causal explanation, it is asserted, although it may provide the basis for predicting human behavior given sufficient knowledge about stimulus conditions.