MICHAEL V. ANTONY
DAVIDSON’S ARGUMENT FOR MONISM*
(appears in Synthese 135(1), 2003, 1-12)
ABSTRACT. Two criticisms of Davidson’s argument for monism are presented. The first is that there is no obvious way for the anomalism of the mental to do any work in his argument. Certain implicit premises, on the other hand, entail monism independently of the anomalism of the mental, but they are question-begging. The second criticism is that even if Davidson’s argument is sound, the variety of monism that emerges is extremely weak at best. I show that by constructing ontologically “hybrid” events that are consistent with the premises and assumptions of Davidson’s argument, but entail ontological dualism.
My guess is that if you want to get a lot of physicalism out [of Davidson’s argument], you’re going to have to put a lot of physicalism in.
—Jerry Fodor 1989, 159
Over three decades ago Donald Davidson introduced his celebrated argument for the token-identity theory. He dubbed his view ‘anomalous monism’—“monism, because it holds that psychological events are physical events; anomalous, because it insists that events do not fall under strict laws when described in psychological terms” (PP, 231). The argument has three main premises:
P1. The Principle of Causal Interaction. At least some mental events interact causally with physical events.
P2. The Principle of the Nomological Character of Causality. Events related as cause and effect must fall under strict laws.
P3. The Anomalism of the Mental. There are no strict laws on the basis of which mental events can be predicted and explained. (ME, 208)
From those premises, Davidson proposed, “we can infer the truth of a version of the identity theory” (ME, 209):
Suppose m, a mental event, caused p, a physical event; then, under some description m and p instantiate a strict law. This law can only be physical, according to the previous paragraph [in which P3 is said to have been established]. But if m falls under a physical law, it has a physical description; which is to say it is a physical event. (ME, 224)
The class of mental events to which Davidson’s argument applies is restricted in two important ways. First, only mental events that causally interact with physical events are addressed. To establish a more general monism, Davidson tells us, “it would be sufficient to show that every mental event is cause or effect of some physical event; I shall not attempt this” (ME, 224). This paper thus primarily concerns mental events that causally interact with physical events, and the term ‘monism’ is restricted accordingly. Second, Davidson’s argument is limited to mental events that exhibit the kind of full-blown intentionality characteristic of propositional attitudes:
…my arguments are limited in application to branches of psychology that make essential reference to ‘propositional attitudes’ such as belief, desire, and memory, or use concepts logically tied to these, such as perception, learning, and action. (Some of these concepts may not always show intensionality, and in such cases are also exempt.) (CR, 240; see also ME, 210 ff.; PP, 229-230)
Excluded, then, are phenomenal or qualitative events like pains or itches, if they lack such intentionality.
What Davidson took to be most distinctive about his argument was that the anomalism of the mental (P3)—the principle that there are no strict laws on the basis of which mental events can be predicted and explained—played a central role in establishing the identity theory. That, he pointed out, was unusual, since traditionally it had been assumed that support for the identity theory could come only from the discovery of laws or correlations linking mental and physical event-types: “If there is a surprise,…it will be to find the lawlessness of the mental serving to help establish the identity of the mental with that paradigm of the lawlike, the physical” (ME, 223; see also ME, 212-213).
In this paper I present two criticisms of Davidson’s argument for monism. The first exposes difficulties with the argument itself, and the second focuses on the variety of monism that is supposed to emerge from it. Neither criticism threatens Davidson’s argument for P3. In the first criticism I argue that there is no way for P3 to contribute to the derivation of monism, Davidson’s claims to the contrary notwithstanding. Other of Davidson’s assumptions do lead to monism, but independently of P3. The route from such assumptions to monism is very direct, however, and the assumptions insufficiently defended. To the extent that such assumptions are crucial to his argument, therefore, his argument begs the question. The second criticism is independent of the first. Even if Davidson’s argument is sound, questions still remain as to the precise nature of his monism. I argue that the monism that emerges is extraordinarily weak, so much so that it is consistent with various forms of ontological dualism.
1. DAVIDSON’S ARGUMENT
What role is P3 meant to play in a Davidson’s argument for monism? With respect to any mental event m that causally interacts with any physical event p, Davidson’s strategy is always to establish that there is a strict physical law covering m and p. That done, given that he holds that an event’s being covered by a strict physical law entails that it has a physical description (ME, 215, 224), and that an event’s having a physical description suffices for its being a physical event (ME, 211, 224), it follows that m is a physical event. But how does Davidson show that there is a strict physical law covering m and p? Since, by P1, m and p are causally related, it follows from P2 that some strict law or other must cover the case. But why must the law be physical? Here, presumably, is where P3 is meant to come in. Davidson often seems to argue by a simple process of elimination: “By Premise three, the laws are not psychophysical, so they must be purely physical laws” (PP, 231; see also ME, 224). But surely that is too fast. For why suppose that the only alternative to psychophysical (or psychological) laws are physical laws? Why think there are just two options?
Here is a bad reason for so thinking. For Davidson, “laws are linguistic; and so events can instantiate laws…only as those events are described in one way or another” (ME, 215; cf. Davidson 1993, 6 ff.). Since we are interested in strict laws for connecting m, a mental event, and p, a physical event, only mental descriptions and physical descriptions can be relevant. And given that mental laws, and hence mental descriptions, have been eliminated, only physical descriptions, and hence physical laws, remain. The reasoning is confused, however. For just as m’s being a mental event does not bar it from having a physical description, it also does not bar it from having indefinitely many other descriptions, including those that are neither mental nor physical; and similarly for p. But then, for all that has been said, there might be a strict law covering m and p that is composed of predicates that are neither physical nor mental. Davidson himself is clear about this:
Given the endless possibilities for redescribing events (or anything else) in non-equivalent terms, it is clear that there may be no clue to the character of an appropriate law in the concepts used on some occasion to characterize an event. (1995, 265)
So even if on some occasion m is described as mental and p as physical, that says nothing about the character of any strict law(s) covering m and p. The possibilities thus remain wide open, even after psychophysical and psychological strict laws have been eliminated by P3.
Others have noticed this point. Mark Johnston (1985), for example, writes:
…imagine there is some other sort of strict law besides physical law, say pure laws of the ectoplasm or laws relating states of the ectoplasm to physical states. Then it could be that some true psychological or psychophysical singular causal statements are not backed up by physical law but by laws of ectoplasm. So there need be no physical redescriptions of the mental events figuring in such singular causal relations, but only redescriptions employing predicates figuring in the laws of ectoplasm. (411-412; cf. McLaughlin 1985, 338; Stanton 1983)
Johnston concludes that Davidson’s needs something stronger than P3 to reach monism, specifically, “the…principle that there are no strict laws except physical laws.” (412) Call this principle ‘P4’:
P4. There are strict laws only in physics
It indeed seems that Davidson needs something like P4 to infer that the strict law connecting m and p must be physical, since it is hard to see how “ectoplasmic” or other kinds of non-physical strict laws could otherwise be eliminated. The trouble, however, is that nowhere does Davidson explicitly say that the law connecting m and p must be physical because there are strict laws only in physics. That he believes P4 is clear, but does P4 play a role in his argument, even implicitly?
A recent formulation of his argument in which he appeals to the causal closure of physics appears to contain P4 implicitly, albeit quite close to the surface:
It is plausible that there is a set of concepts (perhaps there are many such sets) which lend themselves to the formulation of a closed causal system. Let us call these concepts the concepts of physics. In this case, for any two events related as cause and effect, there will be a strict law, i.e., a physical law, covering the case. Since mental concepts are not amenable to inclusion in a closed system, the strict laws covering singular causal relations expressed in (at least partly) mental terms must also be expressible in physical terms. Hence events described in mental terms must also be expressible in physical terms: in ontic language, mental events are identical with physical events. (Davidson 1995, 266)
The first thing to notice is that Davidson seems to be saying that any concepts out of which a closed causal system can be formulated are physical concepts: physics is the closed system. Second, notice the tight connection between closure and strict laws. Davidson moves from closure to strict laws as if the former entails the latter. Indeed, he appears to hold that a system is closed if and only if strict laws apply to it. Taking the two points together, it follows that there are strict laws only in physics. P4 is thus implicit in the above formulation of Davidson’s argument. Davidson, of course, nowhere explicitly employs P4 to rule out strict laws that are neither physical nor psychological. But that P4 is implicit in the text may leave many readers with the impression that excluding such possibilities is unnecessary.
P4 is thus something Davidson believes, appears to need, and it is implicit in at least one formulation of his argument. He can appeal to P4, however, only at the cost of radically altering his argument. For if all strict laws are physical, then since all causal interactions require subsumption under strict law (by P2), it follows that all causal interactions, including m’s causing p, involve only events that are physical. So m is physical. We have reached monism from P4 and P2! Such an argument, however, leaves no work for the anomalism of the mental (P3) in deriving the identity theory, which is proved before P3 is even mentioned. But that P3 plays a key role in establishing the identity theory, we saw, is what Davidson took to be unique about his argument.
The special role of P3 is also jeopardized by Davidson’s appeal to the causal closure of physics. Though Davidson does not characterize closure, presumably he has in mind at least this: physical events causally interact only with events that are themselves physical (cf. Mclaughlin 1985; Stanton 1983). Now if Davidson assumes closure, notice how quickly monism follows. All he needs is P1—the premise that at least some mental events interact causally with physical events. For suppose m, a mental event, caused p, a physical event. By closure, all events with which p causally interacts are physical. So m is physical, and we have reached the token-identity theory without P3 (or P2 for that matter, if closure turns out to be independent of P2). Again, P3 has no work to do in Davidson’s argument for monism.
Davidson’s argument is thus in trouble. He needs something like P4 to rule out the possibility that non-physical and non-mental strict laws might cover m’s causing p. However, if he appeals to such a principle, or to closure, he gets monism without P3. In that case Davidson’s argument for P3, while of considerable interest in its own right, is irrelevant to his argument for monism. Of course, Davidson could still establish anomalous monism by deriving monism from P4 or closure, and then affixing P3 to it, but that would be a very different argument. But would that be so bad? Even if P3 plays no role in deriving monism, after all, Davidson would still have a defense of anomalous monism. Notice that if an argument for monism is to rest on P4 or closure, those assumptions must be very well-defended. For in the context of assuming psychophysical causation, and arguing against ontological dualism, it is clearly question-begging to simply assume that physical events interact only with physical events (closure). Similarly, when assuming that singular psychophysical causal relations must be covered by strict laws, it is equally question-begging to assume that all strict laws link up physical events (P4). The trouble, however, is that nowhere does Davidson defend P4 or closure. So if his argument depends on those assumptions, it begs the question; if it does not, it unclear how Davidson can establish the identity theory at all.
Above it was argued that because P3 cannot rule out strict laws that are neither mental nor physical, Davidson needs something stronger, like P4. However it might be thought that non-mental and non-physical strict laws could be excluded without P4. One way of doing that, for proposals like Johnston’s laws of ectoplasm, for example, might be simply to ridicule the proposals, deem them silly, not worth taking seriously. Another approach would be to insist that such proposals are so undeveloped and confused that it is unclear what they amount to. Thus one might hope to make it plausible that after intentional strict laws have been eliminated by P3, nothing else is worth considering besides purely physical laws. On this suggestion, neither closure nor P4 would play a role in Davidson’s argument, but P3 would. Such responses are illegitimate, however. For it is irrelevant how (im)plausible or (un)clearly formulated a certain proposed dualist alternative is. As an argument for the identity theory, Davidson’s argument is directed against all ontologically dualist proposals regarding mental events, not just worth-taking-seriously or well-worked-out proposals. Davidson’s argument itself, therefore, must rule out strict laws involving ectoplasm, events in Cartesian souls, and so forth; otherwise it is unsound.
A related strategy for excluding certain dualist proposals is to argue that such proposals conflict with other elements of Davidson’s philosophy—for example, regarding the individuation of events, the subjectivity of the mental, etc. To illustrate, suppose one held that propositional attitudes are token-identical with non-physical ectoplasmic events (which lack physical descriptions), and that ectoplasm has no spatial location. One might hope to exclude ectoplasmic events on the grounds that they are inconsistent with Davidson’s theory of events. Davidson holds that “events, like physical objects, are identical if they occupy the same places at the same times” (1985, 175). If he also believes they are identical only if they occupy the same spatiotemporal locations—and it seems he does—that immediately rules out any view according to which distinct propositional attitudes can be simultaneously realized by distinct ectoplasmic events. Similar points could be made about how Davidson’s earlier criterion for event individuation, in terms of same causes and effects (see IE), excludes or restricts dualisms that are epiphenomenalistic; or how his eliminativism about qualia, sense data, etc. (Davidson 1989) rules out various forms of dualism.
The trouble with such responses, again, is that they try to do the work that Davidson’s argument was intended to do. However, if all forms of dualism must be ruled out before we even get to Davidson’s argument, the argument is otiose. Now I grant that certain features of Davidson’s philosophy must be assumed to be in place in order to run the argument. But given how holistic and tightly woven Davidson’s overall philosophy is, it is hard to know just how much must be presupposed. The policy I adopt is to allow anything needed for understanding the premises and assumptions of his argument, but to be wary of anything that comes close to entailing monism or ruling out common forms of dualism, so that we might see what the argument itself is able to deliver.
2. DAVIDSON’S MONISM
Let us assume now that Davidson’s argument is sound. Perhaps P3 plays a special role after all, maybe Davidson has a non-question-begging argument for closure or P4, etc. Questions still remain, however, concerning the precise nature of his monism. What does it amount to?
Davidson typically glosses his monism by saying that mental events are identical to physical events (e.g., ME, 214; PP, 231). In such cases we have a single event describable in the vocabularies of psychology, of physics, and any number of other vocabularies; we have an event monism coupled with a descriptive pluralism. Davidson often shies away, however, from characterizing his monism as materialistic or physicalistic, either because his view is non-reductionist (ME, 214), or simply because “if some mental entities are physical events, this makes them no more physical than mental. Identity is a symmetrical relation” (1987, 453). Of course, other non-reductive materialists (e.g., most functionalists) maintain that one can both be a materialist and preserve what is distinctive about the mental; they assume that while it is false that mental events are “nothing but” physical events (cf. ME, 214), they are in some sense physical “at bottom.” With this Davidson seems to agree. He often points to the privileged status of the physical, either in terms of supervenience (ME, 214; 1993), or by more directly claiming that physics is ontologically basic:
Anomalous monism says that mental entities add nothing to the furniture of the world that is not treated in physics (1999b, 654)
my argument for anomalous monism…is designed to show that the ontology of any science that is not reducible to physics shares its ontology with physics. (1999a, 619)
In a reasonably clear sense, then, Davidson’s monism is meant to be physicalistic.
I wish to argue, however, that the monism that emerges from his argument is much too impoverished to be properly considered a physicalistic monism, except in the most misleading sense. Others have noted the weakness of Davidson’s monism (e.g., Kim 1998), but I shall attempt to bring out the point differently.
To do that, I appeal to the notion of an ontological hybrid—an entity composed of parts (or properties) from more than one ontological category. More specifically, what will concern us are hybrid events with physical and non-physical parts (subevents). Since for Davidson an entity is physical if and only if it has a physical description (ME, 211, 224), such hybrids must contain both physically describable subevents and subevents that lack physical descriptions. Now notice that a partly physical hybrid can be given a wholly physical description. To take a fanciful example, imagine a war fought by ghosts using real guns, where the latter but not the former are physical (i.e., the guns but not the ghosts have physical descriptions). There will be a physical description of the physical subevents of the war, but that physical description will also apply to the war itself (‘the event with such and such physical characteristics’). It is of course not a “full” description of the war (whatever that would be) , but it does what Davidson requires of it: it picks out the war in physical terms. Since the war has a physical description, it must count as a physical event by Davidson’s lights. But it clearly is not physical in the sense required for a physicalistic monism, for only part of the war is physical. The war’s ghosts add substantially to the furniture of the world not treated in physics.
In a similar fashion we can construct hybrid events that are identical to intentional events (e.g., propositional attitudes). Assume that phenomenal or qualitative conscious events are non-physical, in that they lack physical descriptions. They can be events in ectoplasm, Cartesian souls, etc., or they can be sui generis. Assume also that they are not intentional in Davidson’s sense. Our hybrids will have as subevents both phenomenal events and physical (e.g., neurophysiological) events. Intentional events like propositional attitudes—at least occurrent attitudes—we shall assume are token-identical with such hybrids. Neither neurophysiological events nor phenomenal events on their own will suffice for complete propositional attitudes. Notice that this is not an outlandish suggestion. Chalmers’s (1996) property dualism, for example, involves a closely related view regarding occurrent attitudes. Now just as the not-entirely-physical war has a physical description, so do these not-entirely-physical hybrids have physical descriptions (‘the event with such and such physical characteristics’). So they too must count as physical for Davidson. But they are obviously not physical in the sense required for a physicalistic monism: they are ontologically dualistic if anything is. If Davidson’s argument allows for such ontological hybrids, therefore, his monism is of an exceedingly weak sort.
I shall now show that his argument does allow for such hybrids. I do that by constructing hybrids similar to those described above that are consistent with Davidson’s premises P1-P3, as well as with P4 and closure. It will follow that such hybrids are consistent with, and hence comprehended by, Davidson’s monism.
Consider first the causal closure of physics. For hybrids to be consistent with closure they must be constructed in such a way that their non-physical, phenomenal components do not enter into causal relations with physical events. That means that all interactionist dualisms must be excluded, as well as any versions of epiphenomenalism that allow for physical-to-non-physical causation. Parallelisms between physical and non-physical substances, however, will do (sustained, perhaps, by non-causal fundamental laws), as would certain forms of property dualism if properties were admitted into our ontology. Here, as with Johnston’s laws of ectoplasm, many will judge that such dualisms are not to be taken seriously. However they are consistent with the causal closure of physics, and the question to be asked here is whether Davidson’s argument rules them out. We shall see that it does not.
Let us turn now to P2 and P4, which state, respectively, that events related as cause and effect must fall under strict laws, and that there are strict laws only in physics. If P2 and P4 are assumed, then if m causally interacts with p, the strict law(s) covering m and p must be physical. Could hybrids enter into causal relations in accordance with P2 and P4? Since hybrids have physical subevents, each hybrid, as we saw, will have a physical description formulated in terms of the physical descriptions of those subevents. It would thus seem, prima facie at least, that strict physical laws could subsume such hybrids under their physical descriptions. Consider a strict physical law P1ÞP2, and token hybrids, each of which satisfy the physical description ‘P1’. Given that our hybrids’ non-physical subevents are causally unrelated to their physical subevents, the former cannot affect the latter. Consequently each token hybrid satisfying ‘P1’ will be followed by a token event (hybrid or purely physical) satisfying ‘P2’. Hybrids, then, could be lawfully related to other hybrids or to purely physical events, and so enter into causal relations in accordance with P2. Notice that it does not matter how hybrids’ phenomenal subevents vary across instances of the same type of physical subevents. If they vary lawfully, then since the law need not be causal, it need not conflict with closure or P4. If they vary non-lawfully, there obviously will be no such conflicts. Our hybrids thus respect P4 as well.
Are there reasons for thinking that hybrids could not fall under strict physical laws? That they include subevents with no physical descriptions might be thought to be such a reason. If it is, however—and it is far from obvious that it is—an argument is required, and Davidson has not provided one. That being so, merely to stipulate that an event’s possessing non-physical subevents excludes it from subsumption under strict physical law is ad hoc. Taking a different tack, one might claim that only hybrids’ physical subevents, and not hybrids themselves, fall under strict laws, since hybrids are not genuine events—much as this morning’s rainfall conjoined with the Andromeda Galaxy’s most recent supernova might not be considered a genuine event. Intuitively, however, hybrids differ from such arbitrary, gerrymandered events; they seem to be perfectly unified, complex events, in spite of their internal dualistic nature. Now if hybrids’ phenomenal subevents causally influenced the physical realm, that might prevent hybrids from being subsumed under strict physical laws. However, our assumption of no physical-phenomenal causation ensures that the nomic relations holding between hybrids’ physical subevents and other physical (sub)events go undisturbed. Hybrids thus appear consistent with P2 and P4 after all.
Consider next P3, the principle that there are no strict intentional laws. Here propositional attitudes make their appearance. Occurrent propositional attitudes, recall, are being assumed to be token-identical with hybrids: the belief that P, for example, will be identified with tokens of different physical-phenomenal types across cases. For hybrids to be consistent with P3, they must not involve any strict intentional laws—neither intentional-hybrid nor intentional-phenomenal—since there can be no strict intentional laws of any kind according to P3. So long as there are no strict intentional-physical laws (by P3), however, intentional-hybrid laws would appear to be impossible. For if no strict law connects propositional-attitude type M with physical-event type P, M could hardly be nomically related to an event type with the same physical subevents as P, and some additional non-physical subevents. If there were a law relating M to such a hybrid, M would be nomically sufficient for the realization of certain physical (sub)events, and that is ruled out by P3. There also can be no strict intentional-phenomenal laws so long as phenomenal events lack intentionality in Davidson’s full-blown sense. For the same considerations Davidson employs in arguing against intentional-physical laws (rationality constraints, charity, etc.) can be marshaled against intentional-phenomenal laws if phenomenal events lack intentionality. P3 thus seems perfectly compatible with hybrids.
Turn finally to P1, the principle that at least some mental events interact causally with physical events. Since we are assuming that occurrent propositional attitudes are token-identical with hybrids, and we established that hybrids can interact causally with physical events, it follows that occurrent attitudes can interact causally with physical events.
Hybrids are thus consistent with P1-P3, as well as with closure and P4. If Davidson’s argument for monism is sound, it follows that such ontologically dualistic hybrids are consistent with his monism. However, any monism that admits mental events that are partly constituted by non-physical ectoplasmic events (events in Cartesian mind-stuff, etc.) is a monism in name only. On such views the mental does not share its ontology with physics. Now to exclude hybrids, Davidson requires a premise to the effect that all subevents of physical events are themselves physical. Such a premise, however, would prevent him from maintaining that an event is physical if it has a physical description, since hybrids have physical descriptions. It would then be unclear how Davidson could infer that m is a physical event.
Chalmers, D. (1996). The Conscious Mind. Oxford: Oxford University Press.
Davidson, D. (1980). Essays on Actions and Events. Oxford: Clarendon Press.
Davidson, D. (1985) ‘Reply to Quine on Events’, in E. Lepore and B. McLaughlin (eds.) Actions and Events. Oxford: Basil Blackwell.
Davidson, D. (1987). ‘Knowing One’s Own Mind’, Proceedings and Addresses of the American Philosophical Association 60, 441-458.
Davidson, D. (1989). ‘The Myth of the Subjective’, in M. Kransz (ed.), Relativism: Interpretation and Confrontation. Notre Dame: University of Notre Dame Press.
Davidson, D. (1993). ‘Thinking Causes’, in J. Heil & A. Mele (eds.), Mental Causation. Oxford: Oxford University Press.
Davidson, D. (1995). ‘Laws and Cause’, Dialectica 49, 263-279.
Davidson, D. (1999a). ‘Reply to Bjørn Ramberg’, in L. Hahn (ed.), The Philosophy of Donald Davidson. Peru, Illinois: Open Court.
Davidson, D. (1999b). ‘Reply to Bruce Vermazen’, in L. Hahn (ed.), The Philosophy of Donald Davidson. Peru, Illinois: Open Court.
Dretske, F. (1995). Naturalizing the Mind. Cambridge: MIT Press.
Fodor, J. (1989). ‘Making Mind Matter More’, Philosophical Topics 17, pp. 59-80.
Johnston, M. (1985). ‘ Why Having a Mind Matters’, in E. Lepore and B. McLaughlin (eds.) Actions and Events. Oxford: Basil Blackwell.
Kim, J. (1998). Mind in a Physical World. Cambridge: MIT Press.
McLaughlin, B. (1985). ‘Anomalous Monism and the Irreducibility of the Mental’, in E. Lepore and B. McLaughlin (eds.) Actions and Events. Oxford: Basil Blackwell.
Quine, W. (1985). ‘Events and Reification’, in E. Lepore and B. McLaughlin (eds.) Actions and Events. Oxford: Basil Blackwell.
Stanton, W. (1983). ‘Supervenience and Psychophysical Law in Anomalous Monism’ Pacific Philosophical Quarterly 64, 72-79.
Tye, M. (1995). Ten Problems of Consciousness. Cambridge: MIT Press.
Department of Philosophy
University of Haifa
* I am grateful to Hanoch Ben-Yami, Joseph Berkovitz, Ruth Weintraub, and anonymous referees for helpful comments on earlier drafts.
 I use the following abbreviations for essays in Davidson 1980: ‘ME’ for ‘Mental Events’, ‘PP’ for ‘Psychology as Philosophy’, ‘MM’ for ‘The Material Mind’, ‘CR’ for ‘Comments and Replies’, ‘IE’ for “The Individuation of Events’.
 Recently it has been argued that qualia are intentional (e.g., Dretske 1995; Tye 1995). However, the relevant notion of intentionality differs from Davidson’s (it is not subject to the same kinds of rationality constraints, to charity, etc.). Such qualia thus remain untouched by Davidson’s arguments.
 “I made clear that what I was calling a [strict] law…was something that one could at best hope to find in a developed physics…[and] could…be viewed as treating the universe as a closed system….I allowed that there are not, and perhaps could not be expected to be, laws of this sort in the special sciences.” (Davidson 1993, 8-9; see also 1995; ME)
 In one place Davidson states P2 in terms of closure: “The second premise is that when events are related as cause and effect, then there exists a closed and deterministic system of laws into which these events, when appropriately described, fit” (PP, 231). See also note 3 above, and ME, 219.
 Quine (1985, p. 167) takes events to be identical if and only if spatiotemporally coextensive, and Davidson (1985, 175) endorses Quine’s criterion.
 In accordance with Davidson’s extensionalist leanings, I usually avoid talk of properties in what follows.
 Compare Chalmers’ dualism according to which fundamental psychophysical laws do not interfere with the causal closure of physics (Chalmers 1996, 127).
 See note 7. I assume here that for Davidson strict laws are causal laws.
 Many other kinds of dualistic hybrids would be possible if P4 or closure were not assumed, or if non-physical, non-intentional properties were admitted into our ontology.