J ulie Y oo analytic truth, not an empirical one (a problem the mid-century philosophers like Ryle, Anscombe, and Melden had long pointed out) If, on the other hand, they are cashed out in different vocabulary, say, in lower-level vocabulary (this is Fodor’s view), then they lose their identity as full-fledged generalizations of folk psychology since they would ultimately have to be framed in physical vocabulary, making Fodor’s strategy just collapse into Davidson’s original proposal The second is LePore and Loewer’s proposal (1987), where the relation between M and P is a relation of counterfactual dependence A mental property M of event c is causally relevant to property P of event e if c causes e, c has M and e has P, M and P are metaphysically independent (satisfying Hume’s condition), and e’s having P counterfactually depends upon c’s having M, that is, if c did not have M, then e would not have had P The main problem with this approach is that counterfactual dependencies may not be enough (Braun 1995; Kim 1998; McLaughlin 1989) Fires give off both heat and smoke Only the heat is relevant to warming a room However, there is a counterfactual dependency of the heating upon the smoke because smoke and heat reliably co-occur when a fire is lit Thus, there are misleading counterfactual dependencies, and for all we know, the counterfactual dependency of bodily motion upon mental properties is misleading One has to wonder whether the approaches by Fodor and LePore and Loewer miss the point of the principle of the anomalism of the mental Davidson’s insistence on there being no laws couched in the vocabulary of folk psychology is not a mere empirical observation that no laws have yet been formalized It goes back to insights made by Ryle, Wittgenstein, and Anscombe We make sense of people’s minds and actions, not by invoking a universal law and initial conditions, but by appreciating why an action was performed given the agent’s distinctive constellation of beliefs and desires set within a unique, non-repeatable, set of circumstances (McDowell 1984; Child 1993) In brief, we gain a better understanding of an agent’s reasons in terms of a narrative setting, not a scientific one where individual peculiarities are whitewashed in the interest of capturing (possibly nonexistent) generalities (von Wright 1971) 4.2 The exclusion problem 4.2.1 A formulation of the exclusion problem Kim is to be credited with introducing the the exclusion problem as it is known today (1984, 1989, 1998, 2005) Kim’s formulation is a reworking of a problem raised in by Norman Malcolm (1968), who asked whether it is possible for a segment of behavior to have both a ‘mechanistic,’ or physicalistic explanation as well as a mentalistic or a ‘purposive’ one This raises the question of whether it is possible to give multiple complete explanations for a single explanandum The answer given by Malcolm was ‘no’; a complete explanation makes any other superfluous Thus, once an explanation for an action has been furnished by an 184