- Against A Posteriori Functionalism*
There are two constraints on any functionalist solution to the Mind-Body Problem construed as an answer to the question, 'What is the relationship between mental properties and relations (hereafter, simply mental properties) and physical properties and relations?' The first constraint is that it must actually address the Mind-Body Problem and not simply redefine the debate in terms of other, more tractable, properties (e.g., the species-specific property of having human-pain). Such moves can be seen to be spurious by the very multiple-realizability intuitions that motivate functionalism in the first place. For, according to those intuitions, it is possible for a being to experience pain, have beliefs, etcetera, and yet not only to be of a different species, but to have an entirely different material constitution from human beings. Such intuitions imply that our ordinary mental concepts are not species-restricted.1
Second, in order to be properly a functional solution, it must in some way incorporate the idea that mental properties display some characteristic [End Page 83] pattern of relations that are both necessary and sufficient for their individuation.2 This characterization of functional solutions to the Mind-Body Problem is sufficiently broad to capture most traditional theories, including machine state functionalism (Putnam 1960), both American and Australian versions of Ramsified functionalism (Shoemaker 1981; Lewis 1972), language of thought functionalism (Fodor 1987), and non-reductive functionalism (Bealer 1997; Shoemaker 1999).
As I will use the term, a posteriori functionalism is the doctrine that the characteristic pattern of mental properties that is used as the basis for their functional definition will essentially involve a posteriori truths. This is not to say that the a posteriori functionalist must eschew a priori truths or a priori methods altogether, but only that a posteriori investigation is necessary for establishing the truth of at least some of the principles that figure into the individuating pattern. In this paper, I develop and expand a familiar (though underappreciated) argument against a posteriori functionalism (Jackson and Pettit 1993; Jackson & Braddon-Mitchell 1996). (For ease of exposition and because of its familiarity, I will focus my discussion on Ramsified functionalism, but the argument can be easily generalized to cover all other genuine forms of functionalism. The reason for this generality is that the focus of the argument is the modal status of characteristic pattern and, as noted above, the delimitation of this pattern is essential for any genuinely functionalist theory; the argument, thus, gets in 'on the ground floor.') The argument turns on the requisite modal status of the principles contained in the base psychological theory on which functionalists Ramsify. In order for the resulting functional definitions to be counterexample-free, these principles must be necessary in the sense that they must hold necessarily for every sentient creature at the requisite level of cognitive functioning. At the same time, we have ample reason to believe that the results of a posteriori scientific investigation will yield a significant number of non-necessary (contingent) principles. Consequently, we cannot simply carry over the results of scientific investigation unfiltered for the purposes of giving functional definitions of the mental; rather, we must have some way of sorting the core psychological principles (on which we may Ramsify) from the peripheral ones (on which we may not). Unless they are able to do this, a posteriori functionalists will not have an adequate account of the multiple-realizability of the mental. [End Page 84]
Of course, it is widely accepted that Kripke has cleared the way for just such a theory by showing that the necessary and the a priori are conceptually independent. And, at least prima facie, the existence of necessary a posteriori truths provides the perfect framework for defending a posteriori functionalism. Surprisingly, however, there is no generally acceptable means of selecting necessary truths from the results of cognitive science that is consistent with their status as a posteriori truths, or so I shall argue. Consequently, the necessary a posteriori provides no comfort for the functionalist. Thus the a posteriori functionalist faces the following dilemma: either she must adopt the results of cognitive science wholesale (in which case her Ramsified definitions...