Topic:  Kim’s “Multiple Realization and the Metaphysics of Reduction”

I.

Kim considers the claims “that MR proves the physical irreducibility of the mental”(p. 309) and “that psychology constitutes an autonomous special science”. (p. 311)

    --Putnam and Fodor were the first and most prominent philosophers to argue for non-reductive physicalism on the basis of
    MR.  In the 1960’s and 70’s, Putnam and Fodor wrote papers that broke with the old logical positivist’s tradition of a “unity
    of science”—a reductionistic hope that all true scientific claims could, in principle, be stated in the vocabulary of basic
    science (i.e., physics).

II.

As either an empirical, or conceptual, fact there is no 1-to-1 correlation between mental kinds and physical (e.g., neurological) kinds.  Rather, the relationship is 1-to-many:  each mental kind (because it is functional?) can be realized in multiple physical kinds.  (Putnam first made this point clear, in his 1967 paper “Psychological Predicates”, later retitled “The Nature of Mental States.”)

    --There might be realizer variation across species, individuals of the same species, or even within the same individual over
    time.  This suggests (to Kim) a relativized (i.e., “structure-restricted”) version of the correlation thesis. (p. 313)

    --LePore and Loewer hold that if P realizes F, then P is sufficient for F and P explains F.  Kim rejects the explanatory
    requirement, and also argues that realizers must be both sufficient and necessary for what they realize. (pp. 314-315)

     --Even with relativized mental kinds, the realization kind will be highly disjunctive.

III.

But, what’s so bad about reducing mental kinds to a disjunction of physical kinds?  Fodor answers by first noting that kinds are what appear in laws.  For example, mass is a kind because there are laws of physics that essentially refer to mass.  But, a disjunction of physical kinds is not suitable for appearing in laws.  And the mental kinds, if they are to be preserved as part of a science, must be reduced (if they are reduced at all) to something that can appear in laws.

IV.

An analogy:  jade, jadeite, and nephrite

Features of a law:

    i)  supports counterfactuals:  e.g., if such-and-such had been the case, then so-and-so.
    ii)  is projectible:  e.g., it is confirmed by observation of positive instances

Some disjunctive “kinds” cannot appear in laws because there would be a failure with regard to projectibility.  The positive instances that purportedly confirm the “law” might confirm a more restricted law dealing with only one (or some other subset) of the disjuncts.  It depends on what is in the sample.  Compare:  If this were acceptable, Kim claims we could then confirm “cheap” laws like “All things that are F or H are G”, for any random H, whenever we just sample Fs and they all turn out to be G. (p. 320)

But not all disjunctive “kinds” are unacceptable in this way, however:

    “There is nothing wrong with disjunctive predicates as such; the trouble arises when the kinds denoted by the disjoined
    predicates are heterogeneous, “wildly disjunctive”, so that instances falling under them do not show the kind of “similarity”,
    or unity, that we expect of instances falling under a single kind.” (p. 321)

Disjunctive predicates, unlike conjunctive predicates, do not guarantee real similarity amongst objects that satisfy such predicates.  This point alone suggests that they might not pick out real properties.

V.

Physical Realization Thesis:

    1. Certain physical conditions are necessary and sufficient for pain (or any other mental state), and
    2. The nomic relationships between mental states are determined and explained by the properties and nomic connections of
    the physical substrates. (p. 322)

If jade is untenable as a kind (because there are really two kinds here—jadeite and nephrite), pain is even worse off.  There are many heterogeneous realizations of pain.

Moral:  PRT (its second clause in particular) shows that there cannot be a science of psychology, because different laws can hold for those creatures with different physical realizations.  The generalizations of human psychology might not apply to Martians. (pp. 324-325)

    “We must, I think, take seriously the reasoning leading to the conclusion that pain, and other mental states, might turn out to
    be nonnomic.  If this turns out to be the case, it puts in serious jeopardy Fodor’s contention that its physical irreducibility
    renders psychology an autonomous special science.” (p. 325)

Psychology is not autonomous, but divides according to realization differences.

VI.

A plausible principle:  scientific kinds are individuated by their causal powers.

Different realizations have different causal powers, and because mental “kinds” only have the causal powers of their realizations, they do not have unified causal powers.  There is not a unified psychology, but instead psychologies relative to each realization-kind. (p. 327)

VII.

Psychology is still a science, but there are distinct psychological theories for each of the different structural or species-specific realizations—e.g., human psychology, reptile psychology, etc.  There will be structural or species-specific reductions of mental terms to these distinct realizations.

VIII.

For the functionalist, states that satisfy mental predicates do not need to have anything in common but for sharing that functional role. (p. 331)

For the functionalist, mental properties are second-order properties.  (Worry:  Are all second-order properties causally impotent?)  They are then identical to the disjunction of first-order properties that satisfy a functional characterization. (p. 332)  For those averse to such disjunctive kinds, the alternative is to be the sort of eliminativist about mental properties who accepts only mental concepts.  This kind of eliminativism is different from the eliminativism of Churchland, etc., however.  It is eliminativism is the same sense in which Kim is an eliminativist about jade as a kind/property. (pp. 333-334)