Rationality and the argument for anomalous monism

Philosophical Studies

Volumn 87, Issue 3, 1997, Pages 235-258

  Yalowitz, Steven ^a  

^a UNIVERSITY OF CALIFORNIA   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 52549111467     PISSN: 00318116     EISSN: 15730883     Source Type: Journal    
DOI: 10.1023/A:1004200832034     Document Type: Article

References (35)

1. In Essays on Actions and Events (New York: Clarendon Press, 1980); hereafter, 'ME'.
2. My positive proposal is worked out in detail in 'Causation in the Argument for Anomalous Monism', unpublished manuscript; hereafter, 'CAAM'.
3. 2'.
4. 'Psychophysical Laws' in E. LePore and B. McLaughlin (eds.), Actions and Events: Essays on the Philosophy of Donald Davidson (New York: Blackwell, 1986); hereafter, 'PL'.
5. 'Anomalism, Uncodifiability, and Psychophysical Relations', The Philosophical Review 102, 2 (April, 1993); hereafter, 'AUPR'.
6. 'Virtue and Reason', The Monist 62, 3 (July, 1979);
7. 'Functionalism and Anomalous Monism', in Actions and Events: Essays on the Philosophy of Donald Davidson.
8. I find no mention of these issues in Kim's paper, but the argument clearly rests upon the view that deductive-nomological derivation is sufficient for explanation of some sort.
9. Kim says nothing like this explicitly, but I am confident that this way of unpacking his views captures the spirit of the argument at PL, 380-381.
10. One might wonder if Kim's argumentative structure and strategy could be recycled without reliance on the distinction between normative and descriptive. The answer is clearly "no". Recall that Kim's strategy is to characterize the psychological and physical domains in (1) and (4) respectively and then show how bridge laws would lead, via derivational relations, to the disruption of each domain's constitutive essence. Therefore "laws" playing the role of (1) could not count as strict predictive (i.e. descriptive) laws in an argument purporting to show that there can be no such laws (whether of co-instantiation or succession).
11. This issue is discussed in greater detail in CAAM, where the relationship between laws of succession and co-instantiation is examined in more detail. What about a hybrid argument using (1*) and (4), going only in one direction (up)? Then Kim's complaint against (2) and (3) would read as follows: "there would now be a purely physical, and thus contingent, law explaining a normative ceteris paribus psychological one." There are a number of problems with this, some of which mirror those already mentioned in the text with respect to (1*) and (4*); but I want to mention a different one. It is obvious that bridge laws cannot license the derivation of a strict from a ceteris paribus law. But it is also unclear how they can license the derivation of the latter from the former; it would have to be a ceteris paribus law for which, as Jerry Fodor puts it, "'all else' is always equal" ('Making Mind Matter More', Philosophical Topics 17, 1 (Spring, 1989), 75); and that is no ceteris paribus law at all. Only if (2) and (3) were ceteris paribus would the derivation of (1*) from (4) be even potentially valid, and Davidson explicitly has no objection to such hedged bridge laws (ME, 216). (Even in the latter case, problems arise: if the ceteris paribus conditions for (2) and (3) differ (and why shouldn't they?), they might not conjoin in a way that makes (1*) coherent, much less true. Thus, the truth of (4) (or (4*)), (2*) and (3*) on their own would not entitle one to derive (1*). The point here is that what count as 'normal conditions' are interest-relative, and those interests may not be mutually supportive. Fodor's discussion of 'completers' and non-random exceptions for ceteris paribus laws
12. ('Hedged Laws and Psychological Explanation', Mind 100, 397 (January, 1991)) is relevant here. How this general worry about the relation between the D-N model and ceteris paribus bridge laws applies to psychology depends, I suspect, on debates concerning holism and localism about content. I hope to elaborate on these points elsewhere.)
13. 'Thought and Talk', in Inquiries into Truth and Interpretation (New York: Clarendon Press, 1984), 157. This point is made in a slightly different way at 216-217 of ME.
14. 'Incoherence and Irrationality', Dialectica 39, 4 (1985), 352.
15. This is Kim's example (PL, 382) of a strict normative psychological law.
16. However, this sort of holism is not itself the basis of Davidson's thesis of mental anomalism, as he himself notes at ME 222, as well as 'Belief and the Basis of Meaning', in Inquiries into Truth and Interpretation, 154. The point is that this sort of holism afflicts physical concepts and theories as much as psychological ones. See related discussion in §V. Holism at best provides "clues" and "hints" (ME, 217) for the basis of mental anomalism. In CAAM (§§III-1V), I argue that a different sort of holism about dispositional properties is the basis for mental anomalism.
17. Though it is a side issue, it is perhaps worth pointing out that Davidson's own distinction (ME, 219) between homonomic and heteronomic laws is extremely problematic, and does not coincide with that between strict and ceteris paribus generalizations (contrary to Brian McLaughlin's claim in 'Anomalous Monism and the Irreducibility of the Mental', in Actions and Events: Essays on the Philosophy of Donald Davidson, 342). This point is elaborated on in CAAM.
18. This point is clear from ME, 209: "a version of the identity theory of the mental and the physical... shows how the three principles may be reconciled".
19. Establishing the causal closure of the physical actually requires more than what Davidson is entitled to by his argument for monism. (That closure is not presupposed by Davidson is argued for in note 21 below.) The monism established in ME only establishes that all (causally interacting) mental events are physical (Davidson illegitimately extends his 'monism' to all events at 214 of ME). Closure requires that (1) all (causally interacting) events whatsoever are physical, and (2) all such events are covered by strict, 'pure vocabulary' laws (that vocabulary, whatever it is, just is the 'physical' vocabulary). (I am putting aside the possibility of overdetermination-see note 21 below.) (1) could be established by running the same argument for all 'special science' events that Davidson runs for psychological events; assuming that all such former events are causally defined (on which, see my CAAM and §VII below), those arguments will be unproblematic. (2) would kick in once (1) is established in that way, and the PNCC is applied. It is important to realize that assuming the PNCC is not equivalent to assuming physical closure. For more on this point, see note 21 below.
20. Putting aside, of course, the possibilities of massive overdetermination or parallelism; but it is safe to assume that these are not taken seriously by Davidson (see note 21 below).
21. This unified project is explored in detail in CAAM; for related discussion, see note 17 above and notes 20 and 21 below. Kim's focus on psychophysical bridge laws, and failure to raise the issue of psychophysical laws of succession (see §1 above), partly explains his inability to see the problematic nature of his attribution of the assumption of causal closure to Davidson. The relations between these sorts of laws is explored in CAAM.
22. This explains Kim's otherwise inexplicable failure to see a unified rejection of any strict laws (bridge or succession) incorporating psychological predicates in Davidson's argument (on which see §VII). Kim briefly suggests (384) that the place to look for an independent argument for psychological anomalism is in Davidson's discussions of confirmation problems for decision theory, particularly in 'Psychology as Philosophy' (in Essays on Actions and Events). I am not entirely sure what Kim has in mind here, but I would have thought that those discussions merely reiterate and elaborate on a point Davidson makes in ME (216-217) concerning the failure of definitional behaviorism due to the holism of the mental. This holism is responsible for the confirmation problems that Davidson believes decision theory faces. As I have already noted, however, holism cannot itself justify psychological anomalism. See note 14 above and §V.
23. Davidson remarks on the open character of the mental at 224 of ME. Notice that to hold that the mental is an 'open' system is not yet to claim that the physical is 'closed'. That conclusion can be drawn only once monism is derived from mental anomalism (together with the PCI and PNCC), and so cannot be brought in to support it (see note 17 above on how closure can be derived from monism). The point is discussed at length in CAAM, but some remarks are in order here about the relationship between monism and causal closure of the physical domain. Davidson never assumes closure, and the assumption of the PNCC is not equivalent to that of closure, for otherwise there would be far less point in arguing for mental anomalism as a means for deriving (as Davidson explicitly does) monism. For closure guarantees that every physical event has a physical cause/explanation. But since (PCI) some mental events causally interact with physical events, either those mental events (at least the ones which cause physical events) must then be physical (i.e. a substantial portion of monism can be established without mental anomalism); or, many (if not all) physical events are causally overdetermined. Since Davidson never takes overdetermination of this massive kind to be a serious option (if he did, he could not derive monism from the PNCC, PCI and PAM, since dualism along with overdetermination also squares these three principles), and explicitly argues for monism from mental anomalism, I conclude that he cannot be assuming physical closure. I take it that a larger part of Kim's reading, in particular his refusal to give Davidson's concerns about monism a central role, is due to Kim's independent views, argued for extensively in a number of papers, that 'nonreductive materialism' is not a stable or coherent position (for an overview of the relevant issues, see 'The Myth of Nonreductive Materialism', Proceedings of the American Philosophical Association 63, 3 (1989))
24. . According to Kim, the only genuine contending positions are dualism, reductionism and eliminativism. One part of the argument for this claim is that 'nonreductive materialism' is committed to the epiphenomenalism - the causal inefficacy - of the mental. It may be that, assuming this point, Kim rejects the PCI and the subsequent argument for monism in Davidson to which it contributes. Kim thus thinks that the only coherent 'realist' view that is compatible with mental anomalism is dualism; and, in particular, a dualism on which mental states explain, not by causation, but by rationalization. Hence, the mental can have its own, autonomous 'laws'. Kim's reconstruction of the argument for mental anomalism, then, is tailor-made for the overall position into which Kim holds it must fit, since the reconstruction does not require either monism or mental causation. Since Davidson attempts to rebut the charge that anomalous monism leads to epiphenomenalism, and reaffirms the centrality of the concerns with monism in his overall picture ('Thinking Causes', in J. Heil and A. Mele (eds.) Mental Causation (New York: Oxford University Press, 1992)), an adequate reading should take this into account. I am thus suggesting that Kim's reconstruction, well-motivated as it is, blinds us to the actual argument for mental anomalism in Davidson; an argument which has independent merit and might well help, in the end, with the very problems that Kim argues 'nonreductive materialism' is plagued with (the question whether causality is a relation that obtains between events in extension or properties/tropes is crucial to the epiphenomenalism issue).
25. The issue is of course importantly complicated by the problems of conceptual change. While I cannot go into the disputes that have arisen about the degree of belief variance that is compatible with stability of reference, it is surely the case that Davidson rejects the kind of causal theories of reference associated with Kripke and Putnam. I assume that, for Davidson, some overlap or invariance of belief is therefore a necessary condition for stability of reference.
26. This example is in one way somewhat misleading. Part of the point of the alternative reconstruction I sketch in §VII depends on the concepts invoked in formulations of strict laws of succession being 'non-causally' defined. What such concepts are is anybody's guess (for a relevant discussion, see Nelson Goodman, Fact, Fiction, and Forecast, 3rd. ed., (Cambridge: Harvard University Press, 1983) 41ff.). But clearly, given what I say here, 'electron' is not one of them, since it is defined causally. Since strict law concepts cannot be defined causally, there will indeed be a distinction between them and psychological (as well as all other) concepts. This does not vitiate the point being made in the text, however, since for Kim the distinction between normative and descriptive laws and concepts is orthogonal to that between strict and non-strict (or non-causal and causal) laws and concepts: Kim's discussion commits him to the possibility of non-strict but descriptive laws (laws covering electrons being a case in point). My point in the text is that what makes psychological laws 'normative' for Davidson and Kim makes laws about electrons 'normative' as well.
27. The Cement of the Universe (Oxford: Clarendon Press, 1974), 62.
28. There is of course some threshold here, after which usefulness goes down relative to rise in risk; virtus dormitiva cases come to mind.
29. 'Hempel on Explaining Action', in Essays on Actions and Events, 274.
30. In CAAM (§V) I try to account for Davidson's appeals to the normativity of rationality in his discussions of mental anomalism by distinguishing between reasons for thinking psychology to be irreducible and reasons for thinking it to be ineliminable. I explore how the centrality of the concept of error in Davidson's thinking might be construed as an attempt to ground anomalism in an aspect of rationality, and why it fails, in 'Normativity, Externalism and Status of Content'; and the general connections between anomalism and Davidson's requirement of social interaction for thought and meaning in 'Davidson's Social Externalism', both unpublished manuscripts.
31. 'Actions, Reasons and Causes', in Essays on Actions and Events, 9-11.
32. This is especially clear in 'Problems in the Explanation of Action', in P. Pettit, R. Sylvan, and J. Norman (eds.), Metaphysics and Morality (Oxford: Basil Blackwell, 1987), 42.
33. McLaughlin, in 'Anomalous Monism and the Irreducibility of the Mental', 345, 348, makes a similar suggestion.
34. Davidson hints at an argument for the PNCC in his 'Reply to Strawson' in B. Vermazen and M. Hintikka (eds.) Essays on Davidson: Actions and Events (New York: Oxford University Press, 1985), 227.
35. Material from earlier versions of this paper was presented to colloquia at University College, Dublin and the University of California at Riverside, and I am grateful to those audiences for discussion. For valuable advice on earlier drafts, I am especially indebted to David Brink and Pat Kitcher. I would also like to thank Adrian Cussins, Pierre Keller, Philip Kitcher, Noa Latham, Jim Levine, Sandy Mitchell and Gila Sher for helpful comments.