notas of kim.doc

7/29/2019 notas of Kim.doc

1/6

1 In my presentation of Kims argument I rely chiefly on Kim 1989a; 1989b; 1990; 1992;1993a; 1993b; 1993c; and 1998. 1989a, 1989b, 1992 and 1993b are reprinted in Kim1993d. Page references to these three papers are to the 1993d volume.2 If I understand Kim correctly, his point is that since the instantiations of P and M aresimultaneous, it is better to regard them as being related to each other by a realizationrelation, not by causation; so P is a nomologically sufficient condition for P* with justthe same proximity to it in the causal chain as is enjoyed by M; hence P has at least asstrong a claim to being efficacious with respect to P* as M does. Cf. Kim 1993b, 354.3 The weaker version of the principle of closure is presented when Kim says, If aphysical event has a cause at t, it has a physical cause at t. (Kim 1993c, 360.)Consider also the following: If we trace the causal ancestry of a physical event, weneed never go outside the physical domain. (Kim 1989b, 280.) Cf. Kim 1990, 39 and45. He expresses the stronger closure principle when he says, No causal chain willever cross the boundary between the physical and nonphysical. (Kim 1998, 40.) Kimsalternation of these two versions of the closure principle is described in Putnam 1999,215 n. 17; cf. Clarke 1999: 312. E. J. Lowe documents a similar vacillation by otherauthors in his 2000: 573-4. Cf. Montero 2003. I have confined myself to distinguishingamong versions of the closure principle solely on the basis of whether it is taken to ruleout a mental items figuring in the causal history of a physical effect. Sven Walter haspointed out to me that Kims denial that causal chains ever cross the boundarybetween the physical and nonphysical implies an even stronger principle, one thatrules out a physical causes production of a mental effect.4This version of the principle differs from the version of it that appears in Kims earlierwork. According to the earlier version, No event can be given more than one completeand independent explanation. (Kim 1989a, 239, italics in the original.) Again, Acause, or causal explanation, of an event, when it is regarded as a full, sufficient causeor explanation, appears to exclude other independent purported causes or causalexplanations of it. (Kim 1989b, 281, italics in the original.) The importance of thesubstitution of distinct for independent in the more recent version of the exclusionprinciple is clear when we consider Kims statement that supervenient properties aredistinctfrom the properties on which they supervene but are not independentof them(Kim 1989a, 251). In view of this, only the more recent version of the exclusionprinciple precludes the efficacy of the supervening mental properties. In fact, the

adoption of this 22 more recent version of the exclusion principle is what led Kim toabandon his earlier model of supervenient causation (Kim 1993c, 361).5 Strong supervenience relations hold at least across the nomologically possible worlds.Kim says that the modality at issue in strong supervenience is either nomological,metaphysical, or logical, but does not further specify it (Kim 1998, 10; cf. Kim 1998,39). However, in allowing for worlds in which M occurs without being grounded in aphysical property, he seems to rule out the stronger of the modalities (at least in thecase of the strong supervenience that links mental to physical properties).6This objection was presented in an earlier version of this paper at the 2000 meeting ofthe Wester Canadian Philosophical Association, where I received helpful commentaryfrom Bob Bright and Jeff Foss (commentary which gave rise to the following paragraph).Similar criticisms of Kim can be found in Crisp and Warfield 2001: 314; and in Loewer2001: 319-20.7 Kim poses similar questions in his 1993a, 26; 1993b, 354-5; 1993c, 361; and 1998,37.8 I have used e where Kim uses E in this quotation. Kim says that while it is reallyinstances of properties (i.e., events) that are causes and effects, he sometimes, for thesake of brevity, speaks of properties as the causal relata. (Kim 1993b, 354 n. 33;1993c, 361; 1998, 41-2.) I think this leads him on occasion to use capital letters(usually reserved for properties) where lower-case letters would make his meaningclearer. (The lower-case m and n have been preserved, unaltered from the quotedtext.)


2/6

9 I have changed the lettering in this passage.10 In this connection, Ausonio Marras says, But dependent, or ìnherited, causalefficacy is causal efficacy nonetheless: why suppose that it is merely àpparent orèpiphenomenal? (Marras 2000, 148.) My criticism of Kim is more general. To wit, myconcern is that Kims reasons for denying the efficacy of dependent causes threaten toprove too much, for they would, if sound, establish the inefficacy of all apparentlyoverdetermining causes (dependent or not), thereby rendering all overdeterminationimpossible.11 For Davidsons model of events, see Davidson 1967a and 1967b.12The lettering in this quotation has been changed. Kim uses m instead of c for theevent that has both M and P. 13 In a more recent passage, Kim adopts Davidsons model and speaks of the sameevent (m) as falling under both the mental and physical kinds, M and N (Kim 1998, 33-4). In a footnote to that passage, Kim maintains that even on this model one stillconfronts the exclusion problem, since, No causal work is left for M [the mentalproperty]. (Kim 1998, 126 n. 6.)14 Following here Ned Blocks suggestion in Block 1990, 163.15 I develop this line of thought more fully in Raymont 1999.16 One may question whether the explanandum property counterfactually dependsupon the causes mental feature. However, if the cause had lacked its mental feature,it would also have lacked th physical property on which it supervenes (since thesubvening physical property is sufficient for the supervening property); in which case,the explanandum property would not have been instantiated, assuming that thesubvening physical feature really is causally relevant to it. Granted, it seems that if thesubvening physical feature were present in the absence of the supervening mentalfeature, the explanandum property would still have been instantiated. But, given thatthe subvening feature is held to be at least nomologicallysufficient for the superveningproperty, this last bit of counterfactual supposing takes us beyond the nomologicallypossible worlds; hence it is of dubious relevance to questions of causal-explanatoryrelevance in the actual world. Cf. LePore and Loewer 1987 and 1989.17 I think this suspicion of properties that are always explanatorily superfluous is one ofthe concerns at issue in Kims discussions of the Causal Inheritance Principle. (1992,326-7; 1993b, 355; 1993c, 362-3; Kim 1998, 54-6.)18The example is borrowed from Brodbeck 1963: 310.

19 I allow that physical properties might causally account for intentional actionsonly because I think it likely that the former will meet any plausible standard forcausally explaining the latter, and not because of the principle of closure. The closureprinciple requires only that each physical explanandum (viz., each case of an eventshaving a physical property) be causally explainable in physical terms. While intentionalaction properties certainly depend upon physical features for their realization (e.g., anarm-raising depends on an arms rising), it does not follow that they are themselvesphysical properties. Hence an events possessing an intentional action property neednot be regarded as a physical explanandum. Cf. Marras 1998. 23

20 Kim nicely elucidates this sort of explanation in Kim 1984b, but sees it as a mode of

understanding actions which is not causal-predictive . (Kim 1984b: 316, italics in theoriginal.)

21This simply results from the fact that physical properties can meet the standard forcausal relevance, but do not rationalize explananda; therefore, the standard for causalrelevance cannot have built into it any requirement of rationalization.

22 New Fact proponents typically rule this out. E.g., Marras (1998) seems to endorse akind of parallelism, in which mental properties are causally relevant only to intentional


3/6

action explananda, while physical properties are relevant only to physical explananda.Cf. Thomasson 1998: 191-2.23Thanks to Brian Garrett, Danny Goldstick, Bill Seager, Sonia Sedivy, Sven Walter, andespecially Ausonio Marras for comments and conversations that contributed to thispaper. I would also like to thank Jaegwon Kim, who was the commentator when Ipresented parts of this paper at the Eastern Division meeting of the AmericanPhilosophical Association in 2001.

Notes1 Throughout this paper, I shall at times speak of mental and physicalproperties and at other times of mental and physical events. None of thearguments in this paper depend on focusing on one rather than theother.2 Versions of this argument can be found in Malcolm (1968), Sosa(1984), MacDonald and MacDonald (1986), Honderich (1988), Yablo(1992), Kim (1993, 1996, and 1998), and Van23 Gulick (1993). This particular formulation is a substantially modified

version of the (CEA) found in Yablo (1992).3 I should say here that this premise will be discussed in greater detaillater in this paper.4 It should be emphasized that it is a token event that is at issue in thisexample. It is easy to imagine another event of this type, that is, anotherwindow breaking, being caused by something other than my throwing abaseball at it. But this doesnt show that the token event of my throwingthis baseball at a certain speed at T1 is not causally sufficient for thisbreaking of the window at T2.5 Burge (1993, pp. 115-16).6 See Kim (1997 and 1998).7 Baker (1993, p. 93).8 It is not entirely clear whether proponents of the (PEA) intend forexplanatory efficacy to be both necessary and sufficient for causalefficacy or merely sufficient. For ease of exposition, I shall focus on thestronger reading; however, the arguments in this paper apply to theweaker reading as well.9 See Kim (1993 and 1996).10 I should note that both Burge and Baker briefly mention the role thatcounterfactuals play in determining the success of explanations. Forconvincing arguments showing the problems with invokingcounterfactuals in this context, however, see Kim (1998).11 Kim (1994).12 Following Kim, I shall use lower-case letters, e, g, etc. to stand forindividual events and the corresponding upper-case letters, E, G,etc. for statements affirming that these events occurred.13 Kim (1994, p. 57).24


4/6

14 Proponents of explanatory internalism include Hempel (1965),Friedman (1974), and Kitcher (1981).15 Kim (1994) himself actually calls this view explanatory realism.However, in order to avoid the potential implication that proponents ofthe opposing view are irrealists about explanation (and ultimately about

causation) and, moreover, to be consistent with Kims use of the termexplanatory internalism, I shall instead call this view explanatoryexternalism.16 See Lewis (1986).17 See Salmon (1984).18 Other proponents of explanatory externalism include Miller (1987) andRuben (1992).19 I just wish to emphasize that what I mean here by explanations iswhat I earlier called because statements of the form E happenedbecause of G (to be distinguished from the descriptive statements of Gand E).20

Kim (1994, p. 64).21 For instance, Hilary Putnam defends a view of explanation that David-Hillel Ruben calls explanatory pragmatism. Such a view holds that explanation has to be partly a pragmatic concept and thus does notseem to be a clear example of either explanatory internalism,externalism, or a hybrid thereof. [Putnam (1978, pp. 41-2).] However, fora convincing argument showing that such an approach unjustifiablyconflates the analysis of explanation with the pragmatics of givingexplanations, see Ruben (1992, pp. 21-3, emphasis added). For asimilar distinction, see Lewis (1986).22 For a very nice discussion of much of the literature on explanation,

see Ruben (1992).23 Kim has put forth an exclusion argument regarding explanation in his(1993) and (1998). However, he does not specifically claim that such anargument applies even25 when explanatory efficacy is construed on purely internalist grounds,as I intend to show.24 As I shall argue, the mere fact that our current scientific theoriesare unable to specify the precise physical properties underlying ourmental properties does not contribute to discriminating betweenexplanations such as (E4) and (E5). Thus, the lack of specificity in (E5)should not affect the arguments in this paper in any way.25

Burge (1993, p. 116).26 A point of clarification: proponents of the (PEA*), such as Burge, tell usthat mental and physical explanations, such as (E4) and (E5), need notcompete with one another. Now, a natural interpretation of this claim isthat such explanations need not compete because both can beaccepted. Strictly speaking, this may not mean that we are being told toaccept conjunctive explanations (such as (E6)) for every action, wherethese are understood as being single explanations in some sense.


5/6

However, proponents of the (PEA*) are at least committed to us havingboth (E4) and (E5) in our belief systems. Thus, any belief system thatincludes both of these explanations will be subject to the difficulty that Iwill raise for the conjunctive (E6).27 Pace Baker who claims that [s]ystematic explanatory success, in

either science or everyday life, stands in no need of metaphysicalunderpinning. Baker (1993, p. 94).28 Baker (1993, p. 93).29 I am grateful to Hilary Putnam and especially, Baron Reed, for veryhelpful comments on earlier versions of this paper.ReferencesBaker, Lynn Rudder. 1993. Metaphysics and Mental Causation, in Heiland Mele


6/6

26(1993): 75-95.

notas of kim.doc

Documents