1 The issue largely turns on the nature of properties and their relation to predicates. See Mark, Wilson ‘What is This lbing Called “Pain“? The Philosophy of Science Behind the Contemporary Debate,’ Pacific Philosophical Quarterly 66 (1985) 227-67Google Scholar, for an interesting discussion of this problem.

2 I shall assume the falsity of type physicalism only with respect to mental events (states, etc.) of an intentional sort (thoughts, desires, etc.)- the only sort of mental events of concern in this paper. Whether type physicalism is also false for qualitative events (pains, sensations, etc.) is a question we may leave open; see my ‘Materialism, Functionalism, and Supervenient Qualia,’ Dialogue 32 (1993) 475-92 for a discussion of the issue.

3 See Donald, Davidson ‘Mental Events,’ reprinted in Essays on Actions and Events (Oxford: Oxford University Press 1980), 207-25.Google Scholar

4 Davidson also takes type-physicalism to be inconsistent with his mental anomalism (and not just factually false). However, his conception of properties is decidedly nominalistic, so on his account there is likely to be no substantive difference between type-physicalism and reductive materialism. This issue, however, need not concern us here.

5 The objection has been raised by several philosophers. See Ted, Honderich ‘The Argument for Anomalous Monism,’ Analysis 42 (1982) 59-64Google Scholar and ‘Smith and the Champion of the Mauve,’ Analysis 44 (1984) 86-9; Ernest Sosa, ‘Body-Mind Interaction and Supervenient Causation,’ Midwest Studies in Philosophy 84 (1984) 630-42; Jaegwon, Kim ‘Epiphenomenal and Supervenient Causation,’ Midwest Studies in Philosophy 9 (1984) 257-70Google Scholar and ‘The Myth of Nonreductive Materialism,’ Proceedings and Addresses of the American Philosophical Association 63 (1989) 31-47; Frederick Stoutland, ‘Davidson on Intentional Behavior,’ in Ernest LePore and Brian, McLaughlin eds., Actions and Events: Perspectives in the Philosophy of Donald Davidson (Oxford: Basil Blackwell 1985), 44-59Google Scholar; Fred, Dretske ‘Reasons and Causes,’ Philosophical Perspectives 3 (1989) 1-15Google Scholar; Louise, Antony ‘The Causal Relevance of the Mental: More on the Mattering of Mind,’ Mind and Language 6 (1991) 295-327Google Scholar; and others.

6 See Jaegwon, Kim ‘Explanatory Realism, Causal Realism, and Explanatory Exclusion,’ Midwest Studies in Philosophy 12 (1988) 225-39Google Scholar, at 234ff. for a discussion of these principles.

7 See Davidson, ‘The Individuation of Events,’ in Rescher, N. ed., Essays in Honor of Carl Hempel (Dordrecht: D. Reidel 1969), 216-34CrossRefGoogle Scholar.

8 See citations in n.5.

9 ‘Concrete universals,’ ‘tropes,’ and the like, are spatiotemporal occurrences of properties: they are particulars, albeit particulars of an ‘abstract’ sort.

10 Ernest, LePore and Barry, Loewer ‘Mind Matters,’ Journal of Philosophy 84 (1987) 630-42Google Scholar, at 633. Of course this point assumes a Davidsonian scheme of events, where one and the same event can be individuated under different types. But it is precisely under such a scheme of events that the Davidsonian account of the causal efficacy of the mental is being evaluated. It would be question begging to criticize such an account by showing that it does not succeed under a different scheme of events (e.g. Jaegwon Kim’s ‘Events as Property Exemplifications,’ in M. Brand and D. Walton, eds., Action Theory (Dordrecht: D. Reidel1976), 159-77, where events as Fs tum out to be necessarily distinct from events as Gs, when F ot G, and may thus differ in their causal powers). As an editor of this journal pointed out to me, there is indeed ‘some controversy whether it is right to say that it is an event a that causes b, or whether it is the fact that a is an instance of F that is properly the cause’; and on the latter account, as he also points out, the epiphenomenalist objection may be harder to deal with than on the Davidsonian account. I agree, but I would regard this as a reason for preferring the Davidsonian account.

11 Jaegwon Kim, ‘Explanatory Exclusion and Mental Causation,’ in Villanueva, E. ed., Information, Semantics, and Epistemology (Oxford: Basil Blackwell 1990), 37Google Scholar

12 An explanandum, properly speaking, is an event taken under a description. On Davidson’s conception of events, one and the same event can thus’ generate’ distinct explananda, depending on how it is described.

13 A confusion between causation and causal explanation may underlie Kim’s repeated reference to singular causal events by means of locutions of the form ‘c’s being an event of kind R (or c’s having property R) caused B,’ where B is a token event like the rising of George from the couch (’Explanatory Exclusion and Mental Causation,’ 39ff.). But on Davidson’s account of events and event causation (which Kim does not call into question in his discussion of mental causation), what caused B was c, not c’s being R (or c’s having R). (Note that on the Davidsonian account of events, where cis a mental event of type Rand c• a neural event of type N, ‘c=c•’ may be true even if ‘R=N’ is not true. But if ‘R=N’ is not true, then, if c’s having R were a cause, it could not be the same cause as c•’s having N: property dualism would entail that mental events cannot be neural events!;) This confusion may account for the following puzzling remark by Kim (where cis a mental event, c• a neural event identical to c, R a mental property and N a neural property): ‘Given [Davidson’s rejection of the psychoneural identity theory], the identification of Rand N, and hence of the two causes, c’s having Rand c’s having N (assuming c=c•), is out of the question for Davidson, or as part of a defense of Davidson’ (ibid., p. 42, my emphasis). But for Davidson the causes to be identified are c and c•, not c’s having Rand c’s having N! (The exemplifying of a property by an event is not itself an event, hence not a cause.)

14 I have belabored the point because of Kim’s insistence in recent writings (esp. in ’Explanatory Realism, Causal Realism, and Explanatory Exclusion’ and ‘Explanatory Exclusion and Menatal Causation’) on what he calls ‘the principle of explanatory exclusion,’ according to which there can be no more than one complete and independent explanation of a given event. Though there is much in Kim’s defense of this principle that merits careful discussion, here I merely observe that his principle may well be correct given his conception of events, which requires each event to have a privileged characterization in terms of its essential, ‘constitutive’ property, the property that makes the event the event that it is. So on his account each event is an event qua F (or qua G), where F (or G) is its constitutive property. Consequently, on his but not on Davidson’s account, each event generates exactly one explanandum (the event qua F), and we are not disputing that for each explanandum there may only be one complete and independent explanation.

15 The distinction between event epiphenomenalism and property epiphenomenalism is similar to Brian McLaughlin’s distinction between token epiphenomenalism and type epiphenomenalism in his ‘Type Epiphenomenalism, Type Dualism, and the Causal Primacy of the Physical,’ Philosophical Perspectives 3 (1989) 109-35. However, while on this account of type epiphenomenalism mental-event types are said to have no causal powers (110), on my account of property epiphenomenalism mental-event types (properties) are said to have no causal (explanatory) relevance. The appearance of Davidson’s ‘Thinking Causes,’ in Heiland, J. Mele, A. eds., Mental Causation (Oxford: Oxford University Press 1993) 3-18Google Scholar and of papers by J. Kim, B. McLaughlin, and E. Sosa in that same volume has made it clear that the causal relevance of properties, while distinct from the causal efficacy of events, is ontologically, and not merely epistemically (explanatorily) significant, as the foregoing remarks might suggest. For events cause what they do because of the kinds of events they are, i.e., because of the properties they have and not all properties of an event may be relevant to what an event causes. The question whether mental properties are causally relevant, then, amounts to the question whether they make a difference to what mental events cause; and to the extent that they do, they may be relevantly employed in causal explanations.

16 See, e.g., LePore and Loewer, and McLaughlin. Davidson acknowledged the possibility of nonstrict psychological generalizations as early as his ‘Mental Events,’ written in 1969; their function, however, is to alert us of the nearby presence of a strict physical law. A recent explicit acknowledgment of the existence of nonstrict laws, in psychology and in the ‘special sciences’ in general, occurs in Davidson’s ’Thinking Causes.’

17 David, Lewis Counterfactuals (Oxford: Basil Blackwell 1973Google Scholar); Robert Stalnaker, ‘A Theory of Conditionals,’ in Rescher, N. ed., Studies in Logical Theory (Oxford: Basil Blackwell 1968)Google Scholar

18 It’s unlikely that neural properties figure in strict laws, if strict laws are only possible for closed systems like basic physics. But for our purposes it doesn’t matter: let the properties in question be a relevant subset of the neural event’s physical properties.

19 Compare Sosa, 278.

20 There are other (more subtle) responses: see, e.g., Antony or McLaughlin.

21 This corresponds essentially to Kim’s definition of strong supervenience in ‘Concepts of Supervenience,’ Philosophy and Phenomenological Research 45 (1984) 153-76, with the following qualifications: (1) since Kim’s definition is more generic than mine (applying to other domains besides the psychophysical one), he leaves the interpretation of the modal operators unspecified, whereas I shall (provisionally) ascribe them nomological force; (2) my definition applies specifically to (Davidsonian) events, while his may apply to other types of individuals as well. In virtue of (2) it can be seen at a glance that my definition entails token-physicalism as well as the possibility of the multiple (variable) physical realization of mental properties. As Kim has pointed out to me (private correspondence), these entailments would not hold for broader definitions of supervenience which do not assume that the same individuals instantiate both supervenient and subvenient properties (see Kim’s ’Explanatory Realism, Causal Realism, and Explanatory Exclusion’). The literature on supervenience is vast; besides the works by Kim cited in this paper, see Terence Horgan, ed., ‘The Spindel Conference 1983: Supervenience,’ The Southern Journal of Philosophy 22 Supplement (1984) for other classic papers, and, of recent vintage, Kim’s ‘Supervenience as a Philosophical Concept,’ Metaphilosophy 21 (1990) 1-27.

22 See the papers in Horgan’s ‘The Spindel Conference’ for a good sampling of such objections and responses, and Cynthia, Macdonald Mind-Body Identity Theories (London and New York: Routledge 1989Google Scholar), ch. 6 and William, Seager Metaphysics of Consciousness (London: Routledge 1991)Google Scholar, ch. 4 for recent overviews. In view of the ’Twin Earth’ objections, (SS) may only hold forM-properties individuated in terms of a more abstract notion of content than ‘wide’ (truth conditional) content: e.g. in terms of the notion of ‘psychological content’ (as distinct from ‘social content’) proposed by Brian Loar, ‘Social Content and Psychological Content,’ in R.H. Grimm and D. Merrill, Contents of Thought ﹛Tucson, AZ: University of Arizona Press 1988) 99-110. It is also worth stressing that it is not the burden of this paper to argue that nonreductive materialism (inclusive of [55] or otherwise) is actually true, but only that it is consistent with the possibility of mental causation.

23 Kim (’Supervenience as a Philosophical Concept’) has recently questioned that (55) actually captures the sense of metaphysical dependence generally associated with the notion of supervenience, pointing out that (55) merely expresses a relation of lawful functional covariance of physical with mental properties. For a discussion of some implications of this fact see my ‘Supervenience and Reductionism: An Odd Couple,’ Philosophical Quarterly 43 (1993) 215-22. It should be noted that a relation of supervenience as metaphysical dependence would have to be asymmetrical, in which case if psychophysical supervenience is true, type physicalism would have to be false, since identity is a symmetrical relation.

24 In ‘Concepts of Supervenience.’ Weak supervenience only requires supervenience conditionals to hold within, but not across, worlds. But as Seager has pointed out, even weak supervenience may be too strong for Davidson insofar as he denies the possibility of any nomological connection between mental and physical properties: for under anomalous monism, ‘two people with identical brain states could have distinct experiences in this world …. The doctrine of weak supervenience rules out this possibility’ (William, Seager ‘Weak Supervenience and Materialism,’ Philosophy and Phenomenological Research 48 [1988)697-709Google Scholar, at 704).

25 In other words, the token-physicalist conclusion of Davidson’s argument could still be derived if the Principle of the Nomological Character of Causation were restated so as to make its intended causal sense explicit (’events that are causally related must have descriptions that instantiate strict causa/laws’), and if the Principle of the Anomaly of the Mental were weakened so as to only exclude the possibility of strict psychophysical causal laws. (The resulting form of anomalous monism would of course be correspondingly weaker.)

26 I quite realize that Davidson would hardly be happy with this interpretation of anomalous monism, given the kinds of arguments he offers for the Anomaly Principle (in terms of rationality constraints on mental ascriptions and the like). But those arguments are notoriously dense and unconvincing, and establish at best that psychological and psychophysical laws are not exceptionless the way basic physical laws are. This deficiency, as many have pointed out, applies to the laws of all the special sciences, not just to the laws of psychology or psychophysics; so it’s a deficiency psychology can certainly live with, for no one would deny that the laws of the special sciences, nonstrict though they be, have counterfactual force and can thus be employed in explaining and predicting phenomena. As Antony has persuasively argued, the thrust of Davidson’s arguments for the Anomaly principle is really to establish that psychological explanations are rationalizations of behavior and not (despite what he also wishes to claim) causal explanations, for the latter must surely advert to psychological causal laws.

27 Jerry, Fodor ‘Making Mind Matter More,’ Philosophical Topics 17 (1989) 59-79Google Scholar

28 Kim also considers the suggestion (which he credits to Barry Loewer) of weakening the relation of supervenience to ‘ceteris paribus supervenience,’ and seems to find it promising (’Explanatory Exclusion and Menatl Causation,’ 51). For a recent debate on the issue of ceteris paribus laws see Stephen, Schiffer ‘Ceteris Paribus Laws,’ Mind 100 (1991) 1-17Google Scholar and Jerry, Fodor ‘You Can Fool Some of the People All of the Time, Everything Else Being Equal: Hedged Laws and Psychophysical Explanations,’ Mind 100 (1991) 19-34Google Scholar.

29 Let me stress, however, that if we adopt a supervenience condition as strong as (SS), we have to affirm the possibility of psychological causal laws as strict as physical causal laws for, as indicated earlier, the former could be derived from the latter by the mediation of strict supervenience conditionals and would thus inherit whatever strength physical laws have. Given (SS), the distinction between strict and nonstrict laws no longer reflect a distinction between physical and nonphysical laws.

30 ‘Psychophysical Supervenience and Nonreductive Materialism,’ Synthese 95 (1993) 275-304; ‘Supervenience and Reductionism: An Odd Couple’

31 In effect, psychophysical supervenience enables us to view mental causation as an instance of what Kim has called supervenient causation: mental events cause other events by virtue of their instantiating mental properties which supervene on physical properties, whose instantiations cause the events in questions (’Epiphenomenal and Supervenient Causation’). However, for Kim, supervenient causation fails to avoid even event epiphenomenalism because, on his theory of events, mental events turn out to be necessarily distinct from physical events, even when their respective properties stand in the supervenience relation to one another, so no causal work is left for them to do since all of it is done by the physical events. But this is not so on the Davidsonian scheme of events: an event instantiating a supervening mental property may be identical to an event instantiating a subvening physical property, in which case whatever causal work is done by the latter is thereby done by the former. Furthermore, according to Kim, supervenient causation is not an option for the nonreductive materialist since it requires a supervenience relation to obtain between mental and physical properties, and this for Kim entails the possibility of reducing the former to the latter. But, as I have indicated, this is not the case: neither supervenience nor supervenient causation poses any threat of reductionism.

32 The second statement, for reasons presently to be seen, actually provides a very inadequate explanation of the event, given that I might have got a drink of water even if that (type of) neural event had not occurred. On the other hand, Kim has proposed a thesis of ‘explanatory realism’ on which ‘c caused e’ and ‘c*; caused e’ provide the same explanation of e’s occurrence as long as c = c*; (’Explanatory Realism, Causal Realism, and Explanatory Exclusion’). This strictly extensionalist account of explanation seems not only counterintuitive, but it makes it difficult to see how Kim could accuse anomalous monism of being vulnerable to epiphenomenalism: if ’mental event m caused e’ and ‘neural event n caused e’ have the same explanatory content, when m = n, how could mental events, under anomalous monism, be said to play no explanatory role?

33 I am assuming, of course, that (non-finitely specifiable) disjunctive causes won’t do; even if they exist, they can’t be represented in explanatory laws. See my ‘Psychophysical Supervenience and Nonreductive Materialism’ on this.

34 An analogous point is made by William, Demopoulos ‘On Some Fundamental Distinctions of Computationalism,’ Synthese 70 (1986) 79-96Google Scholar, with reference to explaining the distinction between analogue and digital systems. (A digital system is one whose computational description applies to a class of systems that are distinct under a physical ‘analogue’ description.)

35 See, e.g., William, Lycan ‘Psychological Laws,’ Philosophical Topics 12 (1981) 9-38Google Scholar. Louise Antony (private correspondence) also regards this issue as a red herring.

36 Terence Horgan and Michael Tye, ‘Against the Token Identity Theory,’ in Lepore and McLaughlin, eds., Actions and Events, 427-43; Drew, Leder ‘Troubles with Token Identity,’ Philosophical Studies 47 (1985) 79-94Google Scholar

37 Jaegwon Kim, ‘Events as Property Exemplifications’; Alvin, Goldman A Theory of Human Action (Englewood Cliffs, NJ: Prentice-Hall 1970Google Scholar); Lawrence, Lombard Events: A Metaphysical Study (London: Routledge and Kegan Paul 1986)Google Scholar

38 An earlier and shorter version of this paper was presented at a symposium on Mental Causation held at the Canadian Philosophical Association meetings held in Charlottetown, PEl, in May 1992. A version of this paper was also presented, in Italian, to the department of philosophy of the State University of Pisa, Italy. I wish to thank all those philosophers (too numerous to mention here) whose writings have provided the inspiration for this paper. Support from the Social Sciences and Humanities Research Council of Canada is also gratefully acknowledged.