In his 1959 article on mind-body identity, J.J.C. Smart considered an objection (objection 3) that he says he thought was first put to him by Max Black. In my view, this argument, the Property Dualism Argument, is one of the major arguments for mind-body dualism. It has been mentioned in one way or another often in the literature but there is no adequate refutation of it. This paper is aimed at refuting the Property Dualism Argument.
Stephen White (1983, forthcoming) has done more than anyone to sharpen the argument and John Perry’s (2001) book argues that the machinery he uses to defeat modal arguments against dualism and Jackson’s Knowledge Argument also vanquishes the Property Dualism Argument. Brian Loar’s (1990 and 1997) papers are also immersed in the territory of the argument, although not explicitly about it.
I will sketch an account of phenomenal concepts, arguing that certain features of the account allow the Property Dualism Argument to be refuted. I hope that the account of phenomenal concepts and its utility in refuting the Property Dualism Argument will be mutually reinforcing, though those who accept the Property Dualism Argument may fairly regard them as mutually destructive. I will describe White’s and Perry’s approaches as well as an approach inspired by Loar, showing how mine avoids the pitfalls of the others.
Consider a specific type of phenomenal feel, Q, e.g. the feel of the pain I am having right now. (If a type of pain just is a type of feel, then Q is a type of pain.) The physicalist says, let us suppose, that Q = cortico-thalamic oscillation of such and such a kind. (Using this as a paradigm phenomenal-physical identity claim in this paper, I will drop the last six words.) This is an a posteriori claim. Thus the identity depends on the expressions on either side of the ‘=’ expressing distinct concepts, for if the concepts were the same, the identity would be a priori.
We can also see that the concepts must be different by noting that for any true a posteriori identity (e.g. ‘Mark Twain = Samuel Clemens’), we could have been in in qualitatively the same epistemic situation (as the one we are actually in) in which our concepts (counterfactually) picked out different things. For example, Samuel Clemens might have chosen a different pen name and a different author might have adopted the name ‘Mark Twain’.
Here is a rendition of the Property Dualism Argument: If ‘Q= cortico-thalamic oscillation’ is true and a posteriori, then there must be two concepts of the referent, one mentalistic the other physicalistic. The mentalistic concept conceives of the referent under a mental aspect of it. But according to physicalism, that aspect is also physical. Calling that aspect ‘M’, the physicalist would appear to require another identity thesis, M= brain property B53. The original identity thesis raises the issue of another. And this might be said to be the first step in a regress. Is the regress vicious? Consider a closely related problem involving a case of rational error. Suppose someone thinks, rationally though wrongly (according to the physicalist) that he has Q but does not have cortico-thalamic oscillation. (He has been given misleading evidence that Q and all other phenomenal properties are non-physical.) We cannot explain his error without saying how he conceives of Q (and also how he conceives of cortico-thalamic oscillation). But his conception of Q involves a property of Q that is also mental, M. And the physicalist is required to say that the latter mental property is physical too, that M= brain property B53. So the rationalizing explanation of how someone could rationally be wrong about Q requires an appeal to M, about which the very same issue arises. So, it might be said, there is a series of explanatory loans taken out, the filling of each one requiring the next item in the series.
My solution will be one that avoids regresses altogether, so there will be no need to discuss the issue of whether these regresses are genuinely vicious.
The notion of a mode of presentation will play a major role in this paper. “Mode of presentation” is a technical term which can be used to refer to different things given different purposes. The purpose of the notion that I will be assuming is explaining rational error, such as the belief that Mark Twain wrote “The Gilded Age” but Samuel Clemens did not. The belief just mentioned errs by presenting one person under one name that is associated with “The Gilded Age” and also under another name that is not associated with that book title. The notion of explanation here is not just causal explanation but rationalizing explanation. There are two major theoretical options for types of ‘mode of presentation’-- White refers to them as “the side of language” and “the side of the referent” On one way of thinking of modes of presentation, both types of modes of presentation are required to explain rational error. The mode of presentation on the side of the referent is the metaphysical mode of presentation, the property of the referent in virtue of which the word or concept in question picks out that referent. The cognitive mode of presentation is what the thinker or speaker grasps that presents that metaphysical mode. (White calls these the representational and non-representational senses.) The cognitive mode of presentation is what philosophers more normally think of as a mode of presentation, and it is as good a candidate as any for what a concept is. (I shall usually identify concepts and cognitive modes of presentation in what follows.) The Property Dualism Argument involves both of these ideas. I shall argue that the Property Dualism Argument is best formulated in terms of metaphysical modes of presentation rather than cognitive modes. I shall go through the argument in terms of metaphysical modes, then point out why cognitive modes do not have the same relevance to it..
Of course the two kinds of ‘mode of presentation’ are tightly linked, though exactly how is not something I will try to explore here. (According to White, the relation is a priori.) In my account of phenomenal concepts, I ascribe cognitive modes of presentation and metaphysical modes of presentation that are sufficiently tightly linked by any reasonable criterion.
How can a true identity be informative? In the terms used here, the answer would be that although everything is necessarily self-identical, an a posteriori identity depends on two different properties of the referent (metaphysical modes of presentation) and two different cognitive modes of presentation corresponding to them. It is the different cognitive modes which make it the case that the terms of the identity are not a priori equivalent.
In ‘Hesperus=Phosphorus’, the cognitive modes of presentation have something to do with both the meanings and syntax of ‘the evening star’ and ‘the morning star’. That something in the vicinity of meanings are involved in cognitive modes of presentation is suggested by the fact that ‘my bank=my bank’ is far from trivial if the first ‘bank’ means the river kind and the second the financial kind. That syntax (including word identity) is required is suggested by the fact that ‘a fortnight=14 days’ can be informative to someone who knows the meaning of ‘fortnight’ but is always a bit unconfident about the meanings of words. (“I was right”, he says, “a fortnight is 14 days.”) That something else cognitive is required is suggested by the example of ‘Paderewski= Paderewski’, which can be informative to someone who has two uses of ‘Paderewski’ which he takes to denote different people but which actually denote one person. We could imagine that the subject has forgotten where he learned the two words and remembers only of each ‘Paderewski’ that it is the name of a famous person. We could name the cognitive difference by saying that the subject has two “mental files” corresponding to the two uses of ‘Paderewski’. I will assume for the purposes of this paper that meaning, syntax and mental files are physicalistically unproblematic to the extent that they do not involve phenomenality. The rationale for this assumption is that the Property Dualism Argument is concerned with the physical nature of phenomenality, not with these other matters insofar as they do not involve phenomenality.
Cognitive modes of presentation allow one to pick out the referent, but they do so in virtue of some properties of the referent rather than other properies of the referent. These are the metaphysical modes of presentation. We can take the metaphysical modes of presentation involved in ‘Hesperus = Phosphorus’ to be the property of rising in the evening and the property of rising in the morning, respectively. If the cognitive mode of presentation of ‘Hesperus’ is something like the interpreted mental representation that corresponds to “the evening star”, and the metaphysical mode is the property of being the evening star, then in this case at least, the intimate relation between the cognitive mode of presentation and the metaphysical mode of presentation is the relation between a description and the property it ascribes.
As Loar (19xx) notes, Paderewski type situations can arise for general terms, even in situations where the subject associates the same description with the two uses of the general term. An English speaker learns the term ‘chat’ from a monolingual French speaker who exhibits cats, and then is taught the term ‘chat’ again by the same forgetful teacher as applying to the same creatures. The student thinks that they must be different creatures in some hidden way even though they seem exactly the same. Then the student forgets all the specific facts about the learning of the two words except that he believes that things that fit ‘chat’ in one sense do not fit it in the other. We can imagine that the student retains two separate mental files for ‘chat’, each of which says that chats in that sense are not the same as chats in the other. So if he learns ‘my chat = my chat’ where the first ‘chat’ is one use and the second is the other, that will be informative even though there will be no difference in anything that could reasonably be called the meaning of the two uses of ‘chat’. Is ‘my chat = my chat’ then a case in which there are two cognitive modes of presentation, but only one metaphysical mode of presentation? I don’t think there is any correct answer here. Those who would like to see Frege cases explained always in terms of both differing cognitive and metaphysical modes of presentation can say that there are two metaphysical modes of presentation: one metaphysical mode of presentation has something to do with one use of ‘chat’ --e.g. the property denoted by use1 of ‘chat’-- the other metaphysical mode being the property denoted by use2 of ‘chat’. The Property Dualism requires this way of construing the relation between cognitive and metaphysical modes of presentation.
Now let us return to the statement of the Property Dualism Argument. If ‘Q = cortico-thalamic oscillation’ is to be informative, we require different cognitive and metaphysical modes of presentation for both sides. So a metaphysical mode of presentation of the left hand side, M, is required (M for metaphysical, mental and mode) that is different from the mode of the right hand side. We could take the metaphysical mode of presentation of the right hand side to be something to do with electromagnetic properties of the brain property that allow it to be detected and measured. Let us take the right hand side to be unproblematic; whether or not it really is unproblematic, its problems aren’t the same as the problem that is the spring of the Property Dualism Argument. So we must confront the issue of the ontological status of M, the metaphysical mode of presentation of Q. But it looks as if any issue about the metaphysical nature of Q will be duplicated for M. Even if we were to agree for the sake of argument that Q is physical, we can avoid the conclusion that M is non-physical only by supposing that M is a physical property too, say B53. But then that identity, being a posteriori as well will require different metaphysical modes of presentation of both sides. As before, the issue arises as to whether the regress is vicious. As before, there is a way of seeing viciousness in terms of explanatory loans that cannot all be paid back.
There is considerable overlap between my understanding of the Property Dualism Argument and that of Perry (2001) who describes the argument saying: “…even if we identify experiences with brain states, there is still the question of what makes the brain state an experience, and the experience it is; it seems like that must be an additional property the brain state has…There must be a property that serves as our mode of presentation of the experience as an experience...” Later in discussing the Knowledge Argument, Perry considers Jackson’s famous neuroscientist Mary who is raised in a black and white room (which Perry calls the Jackson Room) and learns all that one can learn about the experience of red without ever seeing anything red. While in the room, Mary uses the term ‘QR’ for the sensation of red, a sensation whose neurological character she thinks she understands but has never herself had. Perry says: “If told the knowledge argument, [Max] Black “might say, ‘But then isn’t there something about QR that Mary didn’t learn in the Jackson room, that explains the difference between ‘QR is QR’ which she already knew in the Jackson room, and (5) [QR is this subjective character], which she didn’t?’ There must be a new mode of presentation of that state to which ‘QR’ refers, which is to say some additional and apparently non-physical aspect of that state, that she learned about only when she exited the room, that explains why (5) is new knowledge.” (81) Note that Perry uses ‘mode of presentation’ to denote the metaphysical mode rather than anything cognitive or linguistic, and he sees Black’s problem as having to do with the physicality of the mode of presentation in that sense (metaphysical) of the phenomenal property picked out by our sample phenomenal term. (I will return to Perry’s view later.)
It is important to note that Q must be referred to under a phenomenal concept of it for the problem to even get off the ground. Suppose in the original identity claim we allowed any old concept of Q-- e.g. “the property whose onset of instantiation here was at 5 PM” or “the property whose instantiation causes the noise “ouch””. There is no special problem having to do with phenomenality for the physicalist about the properties in virtue of which such concepts could pick out the phenomenal feel (though in the case of the second example, issues of the physicality of intentionality may arise). That is, the modes of presentation of these properties raise no issues of the metaphysical status of phenomenality. If the original paradigm of phenomenal/physical identity were “the property whose onset of instantiation here was at 5 PM = cortico-thalamic oscillation”, the mode of presentation of the left hand side would not be a special candidate for non-physicality. It would be the property of being instantiated here starting at 5 PM. The Property Dualism Argument depends on an identity in which a phenomenal concept is involved on the mental side. As we will see, a number of putative solutions to the Property Dualism Argument fall afoul of this point.
Note also that in the form in which I am discussing it, Q, M, etc. are properties. The mind-body identity that is at issue is a property identity.
To frame the Property Dualism Argument, we need to use a contrast between deflationism about consciousness and phenomenal realism (or “inflationism”, a term that I do not favor, but which seems inevitable). In its strong form, deflationism is conceptual reductionism concerning concepts of consciousness. More generally, deflationism says that a priori or at least armchair analyses of consciousness (or at least armchair sufficient conditions) can be given in non-phenomenal terms, most prominently in terms of representation, thought or function. The rationale for the terminology can be seen by comparing eliminativism and deflationism. The eliminativist says phenomenal properties don’t exist. The deflationist says phenomenal properties do exist, but that commitment is deflated by a conceptual analysis that makes the commitment less than meets the eye. For present purposes, we can take phenomenal realism to be just the denial of deflationism. The conclusion of the Property Dualism Argument is that physicalism and phenomenal realism are incompatible: the phenomenal realist (“inflationist”) must be a dualist and that the physicalist must be a deflationist.
The renditions of the Property Dualism Argument given above depended on evaluating the viciousness of a regress. There is another version, closely related, that does not have this overall form (though regress arguments come in). It depends on listing all the leading candidates for the nature of the metaphysical mode of presentation of the mental side. Recall that the phenomenal side of the identity is Q and its metaphysical mode of presentation is M. There are five options for the nature of M. M might be (1) mental, (2) physical, (3) non-physical, (4) topic-neutral or (5) non-existent, i.e. the reference is “direct”. Smart’s (1959) and White’s (1983) argument is that (1) and (2) are ruled out for reasons given below. (3) is incompatible with physicalism, so the only remaining option for the physicalist is (4). (They don’t consider (5).) White argues that (4) is deflationist. The argument is this: Topic-neutral properties are functional properties. If M, the mode of presentation of Q, is a functional property, then that could only be because the phenomenal concept has an a priori functional analysis. E.g. the concept of pain might be the concept of a state that is caused by tissue damage and causes certain reactions including interactions with other mental states. But an a priori functional analysis is deflationist, by definition. Their conclusion is that a physicalist must be a deflationist, or, alternatively, that a phenomenal realist (“inflationist”) must be a dualist.. (White (1983) used the argument in pursuit of deflationism but White (forthcoming) uses it to argue for dualism.). The point of view of this paper is phenomenal realist and physicalist, though see Block (2002) for a different kind of doubt about this combination.
I must say something about what a physical property, a mental property, etc., are supposed to be. As Hempel (19xx) noted, there is a serious problem of obscurity for physicalism. Physicalism about properties could be put as: all properties are physical. But what is a physical property? Hempel noted a dilemma (that has been further elaborated by Chomsky): Horn 1 is: we tie physicalism to current physics, in which case physicalism is unfairly judged false, since there are no doubt entities and properties which are not countenanced by current physics. These entities and properties would be counted as non-physical by this criterion, even if the physics of next week will acknowledge them. Horn 2 is: we define physicalism in terms of future physics. But what counts as physics? We cannot take physics as given in an inquiry about whether physicalism can be unproblematically defined. And we surely don’t want to count as physics whatever is done in academic departments called ‘Physics Departments’. For if theologians hijacked the name ‘Physics’, that would not make God physical.
But not all philosophy concerned with physicalism can be about the problem of how to formulate physicalism. For some purposes, physicalism is clear enough. In particular, the debate about the Property Dualism Argument seems relatively insensitive to issues about what exactly physicalism comes to. (If not, that is an objection to what follows.)
I will take the basic notion to be used in defining physicalism to be that of a physicalistic vocabulary, and I will take that as given. A physical property is a property expressible in physicalistic vocabulary. For example, the property of being water is a physical property because that property = the property of being H2O. The predicate ‘___is H20’ is a predicate of physics (or anyway physical science), the property of being H2O is expressed by that predicate, and so is the property of being water, since they are the same property. A mentalistic property is a property expressible in mentalistic vocabulary. A non-physical property is a property that is not expressible in physicalistic vocabulary. (So physicalism dictates that mental properties are expressible in both vocabularies.)
Note that the relation of “expression” is distinct from referring. ‘___is a pain’ is a mentalistic predicate and thus expresses a mental property (that of being a pain). But the fact that I can refer to H2O (or the property of being H2O) with a mentalistic expression like “my favorite property” does not make being H2O a mental property.
Smart said that a topic-neutral analysis of a property term entails neither that the property is physical nor that it is non-physical. It would not do to say that a topic-neutral property is expressible in neither physical nor non-physical terms, since if physical terms and non-physical terms are all the terms, there are no such properties. The key kind of topic-neutral property for present purposes is a functional property, a second order property that consists in the having of certain other properties that are related to one another (causally and otherwise) and to inputs and outputs, all specified non-mentalistically. One could say that a topic-neutral property is one that is expressible in terms of logic, causation and non-mentalistically specifiable input-output language. This raises the issue of whether these terms are to be counted as part of physical vocabulary or not. Since the issue of dualism in the context of the Property Dualism Argument is whether mental properties are non-physical, it would be best for current purposes to count logic, causation and the non-mentally specifiable input-output terms as physicalistic. So I count a topic-neutral property as a kind of physical property. (This decision will require me to move from a version of physicalism that says that all properties are physical to one that says that all properties are micro-physical at some points below.)
As I said, M (the metaphysical mode of presentation of Q) might be (1) mental, (2) physical, (3) non-physical, (4) topic-neutral or (5) non-existent. I will briefly sketch a point of view on each of these options that suggests that that option will not do. This point of view is put in my terms but is intended as the argument of my opponent, the advocate of the Property Dualism Argument. As noted, my version of the Property Dualism Argument is keyed to metaphysical modes of presentation. Later, I will consider whether putting the issue in terms of cognitive modes of presentation will help my opponent. In the end, I will opt for versions of the mental, physical and direct reference options, arguing that the difficulties advanced below for those options can be avoided.
If M is mental, then the same issue of physicalism arises for M, the metaphysical mode of presentation of Q, that arises for Q itself. So we have gotten nowhere. Option 1 is apparently useless. (I will argue below that Option 1 is true and not problematically so.)
The regress arguments mentioned earlier would come in here. However, they are not needed if the consideration of the last paragraph is enough to rule out the physical option.
The problem with this version of a direct reference (in the sense of no mediation) view is that a mental representation can be triggered without itself—apart from the property that does the triggering--involving any phenomenality. Thus the triggered representation need not itself be a phenomenal concept. Recall that it is easy to frame a concept of a phenomenal property that does not motivate the Property Dualism Argument, e.g. the concept of an event that occurs here at 5 P.M. This variant of the direct reference proposal in effect substitutes a non-phenomenal concept for the phenomenal one, thereby failing to refute the Property Dualism Argument. (Here I shift from expositing the Property Dualism Argument to rejecting a false solution to it.)
The same point shows that the triggering notion of a phenomenal concept does not suffice to reply to Jackson concerning the “Knowledge Argument”. On a triggering conception of a phenomenal concept, it cannot explain at the most basic level what Mary learned--since what Mary learned is phenomenal. On seeing the red thing, Mary has a thought. That thought constitutes what she learns, so it must contain a phenomenal concept.
So I reject any version of the direct reference view that does without truly phenomenal concepts. However, as I will argue later, there is a version of the direct reference view that does not have this defect.
Construed thus, “being the sensation attended to by Mary” could be a concept of a zombie and is certainly not a phenomenal concept. So Perry’s suggestion as I read it has the same problem as the “triggering” version of the direct reference view canvassed above—that it changes the subject by substituting a non-phenomenal concept for a phenomenal one. “The property whose instantiation causes me to buy aspirin” is a topic-neutral concept of Q, but the existence of such a topic-neutral concept of Q shows nothing about whether a genuine phenomenal concept is topic-neutral. The view that the concept of a property whose instantiation causes me to buy aspirin is a phenomenal concept is just a version of Smart’s original proposal--which is a form of deflationism. The phenomenal realist (“inflationist”) will find Perry’s proposal as defective as Smart’s proposal.
Smart (followed by Lewis) takes the upshot to be contingent physicalism. That is, their view is that ‘pain’ contingently picks out a physical state, for ‘pain’ is a non-rigid designator whose sense is the item with such and such functional role. But the view that stands behind this picture is that the nature of the mental is given a priori as functional as White recognizes. ‘Pain’ is a non-rigid designator, but ‘having pain’ rigidly picks out a functional state. So the view is a version of deflationism.
The attentive reader may have noted a strange shift in the topic of discussion concerning the topic-neutrality option. As I emphasized earlier, Perry takes the Property Dualism Argument to concern the status of M, the metaphysical mode of presentation of Q, the property of Q in virtue of which the phenomenal term of the phenomenal/physical identity picks it out. He appears to opt for the topic-neutral option. But then he shifts to the topic-neutral option for the mode of presentation in the other sense of the term, the cognitive mode of presentation. The shift happens in the passage quoted above. “This is an informative identity; it involves two modes of presentation. One is the scientifically expressed property of being B52, with whatever structural, locational, compositional and other scientific properties are encoded in the scientific term. This is not a neutral concept. The other is being a sensation that is attended to by Mary. This is a neutral concept; if the identity is true, it is the neutral concept of a physical property.” The passage starts with metaphysical modes of presentation (the underlined phrases express metaphysical modes of presentation) and then shifts to the topic-neutrality of concepts, which are modes of presentation in the cognitive sense. The view he actually argues for is: “We need instead the topic-neutrality of demonstrative/recognitional concepts.” (159) Given that the Property Dualism Argument requires phenomenal concepts, this proposal would have to be the proposal that phenomenal concepts are topic-neutral concepts, which would make Perry a deflationist, despite his vehement protestations to the contrary. If Perry defends himself by saying that he wasn’t talking about genuinely phenomenal concepts, but only a certain group of demonstrative concepts that could be used to pick out a phenomenal property, then the charge is that by shifting away from genuinely phenomenal concepts, he has failed to come to grips with the Property Dualism Argument.
To sum up, the Property Dualism says that in the identity ‘Q=cortico-thalamic oscillation’, the metaphysical mode of presentation of Q must be either mental, physical, non-physical, topic-neutral or “direct”, in which case there is no metaphysical mode of presentation. The mental option is supposed to be useless and possibly viciously regressive. The physical option is supposed to be ruled out because there is no a priori available physicalistic description of Q. The “direct reference” option appears to be ruled out by the fact that the concept of Q needed to get the argument off the ground is a phenomenal concept with a phenomenal metaphysical mode of presentation. So the only options are non-physical and topic-neutral. The topic-neutral option involves a form of deflationism. So the phenomenal realist (“inflationist”) cannot be a physicalist.
Recall that concepts in my sense are a kind of mental representation used in thought. This usage contrasts with a sense of ‘concept’ in which a concept is something more like a meaning—something that two different representations might share. However, concepts in my sense involve meanings—they are “interepreted” representations. We could frame a fine-grained notion of a concept as follows: X and Y are instances of the same concept if and only if they are instances of the same representation, have the same meaning, and are the same in non-descriptional (e.g. phenomenal) features. Suppose for the sake of an example that natural languages are the languages in which we think, so ‘water’ and ‘H2O’ are terms in a language of thought, mental representations. Then, since ‘water’ and ‘H2O’are instances of different representations, they are instances of different concepts. Further, ‘water’ and ‘eaux’ are instances of different representations, so they are instances of different concepts too, despite having the same meaning. Schiffer (1998) notes that such a concept of concept differs from the ordinary one, since we ordinarily take speakers of different languages to nonetheless share concepts, given that they share meanings.
Schiffer considers a proposal that x and y are instances of the same concept provided that x and y have the same meaning and are instances of mental representations that have relevantly similar inferential roles. This proposal may work well for descriptive concepts but the theory I will suggest is that experiential concepts contain a non-descriptional element. The proposal that Schiffer considers strains to accommodate them, since the non-descriptional element doesn’t have meaning in any straightforward sense. Two concepts might have different non-descriptional phenomenal elements—and nonetheless might have descriptional elements with the same meaning and the same inferential role. For example, this would be the case if the traditional “inverted spectrum” hypothesis obtains. Your concept of red (the concept you apply to red things) might function in the same way as my concept of red (the concept I apply to red things) yet be phenomenally different because of a difference in the non-descriptional phenomenal element. If we regard that as showing that your concept of red differs from mine, and we regard meaning of the descriptional parts of our concepts as what is relevant to whether or not they have the same meaning, then the proposal Schiffer considers doesn’t work for phenomenal concepts. (That is, the proposal doesn’t work if it is understood as a biconditional. It is much less interesting construed as merely a sufficient condition of same concept.)
The idea of my account of phenomenal concepts is simply that a phenomenal feel can serve in thought to represent itself. (Loar says something similar, but his view of phenomenal concepts combines this idea with something else, that phenomenal concepts are recognitional concepts.) This notion seems similar to what Chalmers (forthcoming) calls a direct phenomenal concept. The thought that the phenomenal property I am experiencing now is the best ever would have the form “X is the best ever”, where X is the phenomenal property itself. The corresponding past tense thought would be “X was the best ever” in which X is a faint image of the phenomenal property. The “mental sample” of the phenomenal property in both cases functions in the most basic case without any concept of it.
If we identify the concept with the cognitive mode of presentation of it, then the cognitive mode of presentation of the phenomenal property is the phenomenal property itself. Further, the metaphysical mode of presentation, the property of the referent in virtue of which the phenomenal property is picked out will be the very same phenomenal property. So the heart of my account as it affects the Property Dualism Argument is that Q=M, and M = the cognitive mode of presentation (in the most basic case).
I believe that this bare bones view is enough of an account of phenomenal concepts to be used to refute the Property Dualism Argument, so the rest of this section is strictly speaking unnecessary for the train of thought of the paper. But I have some worries about the bare bones view, and those worries suggest a slightly different view. The first worry is purely introspective: it seems to me that there is something demonstrative about my phenomenal concepts, and the bare bones view does not have room for that. Second, we can have phenomenal concepts of things other than our own phenomenal states, and the bare bones view has no room for that. Let me explain.
The expressions ‘this sudden involuntary muscle contraction’ and ‘this [experience] thing in my leg’ are two expressions that pick out the cramp I am now having in my leg. (These are versions of examples from Loar, op.cit.) In ‘this [experience] thing in my leg’, attention to an experience (of the cramp) functions so as to pick out the referent. (That is the meaning of the bracket notation.) The first way of thinking about the cramp is an objective concept of the cramp. The second is a subjective, indeed a phenomenal, concept of the same thing—there is a phenomenal mode of access to the thing picked out. Just as we can have both objective and phenomenal concepts of a cramp, we can also have objective and phenomenal concepts of a cramp feeling. Importantly, the same experience type could be part of—though function differently in—both phenomenal concepts, the phenomenal concept of the cramp and the phenomenal concept of the cramp feeling.
Assuming physicalism, we could have an objective neurological concept of a cramp feeling, e.g. ‘the phased locked 40 Hz cortico-thalamic oscillation that is occurring now’. And we could have a phenomenal concept of the same thing, ‘this [experience] feeling.’ These are concepts of the sort that figure in a mind-body identity claim.
A few more examples: One could have two concepts of a color, an objective concept (“the color of fire engines”) and a subjective phenomenal concept (“this [mental sample] color”, where the demonstrative is linked to a phenomenal quality exemplified in a mental sample. (I assume for purposes of this and other examples that follow that colors are objective properties of objects.) And one could rationally apply one but not the other. One could also have two different phenomenal concepts of a single shape, one keyed to a visual experience of that shape, the other keyed to a tactual experience of that shape. One could rationally apply one but not the other—e.g. if one feels the edge but does not see it. So it would be a reasonable use of the term ‘concept’ to take phenomenal difference as sufficient for difference in concept. Assuming that we want a uniform notion of concept, which uses the same criteria both intrapersonally and interpersonally, if our spectra are all shifted somewhat relative to one another, we all have somewhat different concepts of red.
Notice that in ‘this [tactual experience] shape, the demonstrative refers to the shape, not the experience. But attention to the experience is part of the route to the referent. In the following two concepts, the demonstrative refers to the same thing, but via attention to different experiences: ‘this [visual experience] shape’ and ‘this [tactual experience] shape’. The words in brackets are words of the attributor of the concept and need not be words that the person whose concept it is understands. This is not to say that the contents are non-conceptual, but only that that the attribution abstracts from the issue of what the subject’s concept is, if any.
The concept ‘this [experience] phenomenal property’ requires anyone who grasps it to understand what a phenomenal property is. But someone who does not have that concept might nonetheless think about their pain in a phenomenal way. There is a simpler and more basic form, ‘this [experience]’ which is closer to the bare bones view mentioned above and does not have this defect. Suppose the experience is the cramp feeling mentioned earlier. Why does ‘this [experience]’ function in thought so as to denote the cramp feeling rather than the cramp? I don’t have an answer—except to say that there just seems to be a basic use of experience-plus demonstrative in thought that functions this way. What makes the experience of scarlet function in thought so as to denote that experience rather than the experience of red? Again, I don’t have an answer except that there does seem such a primitive function.
A single sample can be used to determine reference in different ways, so it makes sense to allow the words within brackets to indicate more than just attention to a sample. Here is an example that illustrates that point using David Austen’s (19xx) two tubes case. Imagine the subject to be looking at a red circle through two tubes whose exact orientation he does not know. The subject can rationally wonder whether “That [red circle seen via the left eye] is the same as that [red circle seen via the right eye]”. The words ‘red circle seen via the left eye’ indicate the route of reference but without attribution of those concepts to the subject. Of course, for the route of reference to be specified by those words, the subject will have to have some appreciation of the route or at least the difference, but the appreciation need not be explicit.
The phenomenal color concept just mentioned is more straightforward than some other phenomenal color concepts one could have. For example, a phenomenal color concept could exploit experiences of colors other than the one the concept is a concept of: Here are two examples: (1) ‘the color between these [two visual experiences] colors’ (Hume’s missing shade of blue), (2) ‘the complementary of this [visual experience] color’. These phenomenal concepts would seem to require the subject to have concepts of the experience, for example the concept of complementarity as it applies to experiences. But basic phenomenal concepts require no conceptualization of the experience attended to in my view. Thus it is a primitive fact that the sample is a sample of a certain property. In what follows, I will ignore such complicated methods of determining reference, concentrating on basic phenomenal concepts that could be expressed as ‘this [experience]’, in which the phenomenal property of the attended mental sample is the same as the phenomenal property that is referred to. (Sometimes I will talk of ‘this [experience] phenomenal property’ for explicitness.)
The basic phenomenal concepts I am talking about involve a simultaneous occurrent phenomenal element, a mental sample, e.g. an image or an experience. There are other, less basic phenomenal concepts, that involve a phenomenal element only dispositionally. In my view, we could not have dispositional phenomenal concepts unless we could have occurrent phenomenal concepts. More relevantly to the argument of this paper, were I to use a more liberal notion of phenomenal concept that allowed dispositional phenomenal concepts, the advocate of the Property Dualism Argument could reasonably complain that his argument depends on the fact that the application of phenomenal concepts necessarily brings in the very phenomenality that the concepts are about, and so the argument works only if we assume basic phenomenal concepts.
Note that a creature can have a state with phenomenal properties even if it has no concept of those phenomenal properties—even if it cannot have a concept of those phenomenal properties.
To summarize the sketch of the basic phenomenal concepts as they pertain to the property Dualism Argument, there are four main features:
Non-descriptivity: Phenomenal concepts are partly non-descriptive, involving a mental sample.
No concept: The mental sample functions in a basic phenomenal concept without any concept of it.
The cognitive mode of presentation of Q = Q itself. In thinking the thought that Q is nice using a basic phenomenal concept of Q, the cognitive mode of presentation of Q is just Q.
M=Q: The property that presents the referent of the mental side of the mind-body identity is just the referent itself. Note that this claim obliges me to slightly change the definition of a metaphysical mode of presentation. I said that the metaphysical mode of presentation of the referent is a property of it in virtue of which the referring term picks out the referent. But Q is not a property of itself. So the definition of a metaphysical mode of presentation should be: the property associated with the referent in virtue of which the referring term picks out the referent, where “association” comprises both predication and identity.
Are phenomenal concepts of the sort that I am talking about natural kind concepts of the sort discussed by Putnam and Kripke? Yes and no. They are natural kind concepts in that they purport to pick out objective kinds. But they are not in that no Twin Earth cases are possible for them. If Twin Earthers are phenomenally like us but physically different, then if physicalism is true, there is some higher order physical property in common or else the phenomenal property in question is a disjunctive physical property. (See Block 2002 for more on this issue.)
Armed with this treatment of phenomenal concepts, let us return to the Property Dualism Argument. My solution rehabilitates versions of all of options 1, 2 and 5, though in the case of 1 and 5, not exactly the versions criticized above. Recall that the Property Dualism Argument starts with the physicalist claim that phenomenal feel Q = cortico-thalamic oscillation. I said that in order for the identity to be a posteriori, ‘Q’ and ‘cortico-thalamic oscillation’ must pick out the property referred to (viz., Q) via different properties of that property. I called the property of Q via which ‘Q’ picks it out ‘M’. M is the metaphysical mode of presentation of Q. What is the nature of M? In particular, how do we avoid supposing it is non-physical? Option 1, you will recall was that M is mental. But, the Property Dualism Argument says, this option is one we can dismiss quickly: If M is mental, then the same issue of physicalism arises for M that arises for the original property that is the referent, Q. So we have gotten nowhere.
However, as should now be obvious, there is a way around this reasoning. The phenomenal concept refers not exactly via a mental property of the referent but via a mental property that is the referent itself. So the putative regress is stopped at the first step: M = Q. (Also, M¢ = M = Q.) The concept expressed by ‘this [mental sample] phenomenal property’ uses a sample of a phenomenal property to pick out that very phenomenal property. The phenomenal property does double (indeed triple, indeed quadruple,…) duty: it is the referent but it is also the metaphysical mode of presentation of the referent: it presents itself.
Does this response to the Property Dualism Argument show that option 1 is true and harmless or does it show rather that there is a new option closely related to option 1 that was omitted earlier that is harmless? I think one could go either way in answering this, which is why I described my proposal as rehabilitation of option 1 rather than an argument for option 1.
So whether or not the putative regress would be vicious, we can see that there is no regress to be vicious. Option 1 or a variant of it is true and harmless. If Q is physical, so is M (since M=Q). But we have not yet replied to the objection that M cannot be physical.
Consider White’s argument that “there is no physicalistic description that one could plausibly suppose is coreferential a priori” with ‘Q’ or ‘that feeling’. White is claiming that if the mode of presentation is physical, then there must be a physical description that is coreferential a priori with ‘Q’. (Recall that we are considering an identity, ‘Q= cortico-thalamic oscillation’.) To see what is wrong with this claim, it will be useful to take an example outside the mind. Consider the demonstrative concept “that very wet thing”. The thing is picked out via the metaphysical mode of presentation of being wet, i.e. at least partially covered (or saturated) with water. The property that is that metaphysical mode of presentation is in fact physical, indeed micro-physical. Being covered with water = being covered with H2O, and that is a physical property. But must the user of the concept “that very wet thing” have a priori access to the physical description or know a priori that the physical description is coreferential with the original description? Obviously not. The user of the concept “that very wet thing” need never have heard of oxygen and hydrogen. The metaphysical mode of presentation can be physical even if there is no a priori available physical description. A metaphysical mode of presentation can be physical because identical to a physical property even if the subject has no a priori access to the physicality of it.
What the subject does have a priori access to is the cognitive mode of presentation. Without saying exactly what a cognitive mode of presentation is, we can take the cognitive mode of presentation connected with “that very wet thing” to involve the words “that very wet thing” or a mentalese version of them, and the meanings of these words. We agreed at the outset that we would not take the physicality of words and meanings to be problematic, given that the issue of this paper is the physicality of phenomenality. So the cognitive mode of presentation is not a problem for physicalism either.
White distinguishes between “thin” and “thick” properties. Thick properties, e.g. water (or the property of being water) have hidden essences. Thin properties do not. Examples of thin properties are mathematical properties and functional properties. According to White, the explanatory purpose of metaphysical modes of presentation preclude thick properties serving as modes of presentation. For, he says, it is not all of a thick property that explains rational error but only an aspect of it.
But an aspect is just a property. If we are willing to allow properties to explain rational error, then why not the original property? Further, functional properties also have aspects. Consider the functional property of being caused by tacks and causing “Ouch!”. An aspect of this property is being caused by tacks. Another aspect is causing something.
If we accept that metaphysical modes of presentation must be thin, the conclusion of the Property Dualism Argument would follow quickly by a route different from the ones mentioned earlier. For the only candidates for such “thin” metaphysical modes of presentation would be (1) functional properties, in which case deflationism would be true, and (2) phenomenal properties that are non-physical, in which case dualism is true. Such a claim would have to be a conclusion of a powerful argument against phenomenal realist physicalism, not a postulation.
To sum up, the Property Dualism Argument says that in the identity ‘Q=cortico-thalamic oscillation’, the metaphysical mode of presentation of Q must be either mental, physical, non-physical, topic-neutral or “direct”, in which case there is no metaphysical mode of presentation. The mental option is supposed to be useless and possibly regressive. The physical option is supposed to be ruled out because there is no a priori available physicalistic description of Q. The “direct reference” option appears to be ruled out by the fact that the concept of Q needed to get the argument off the ground is a phenomenal concept with a phenomenal metaphysical mode of presentation. So the only options are non-physical or topic-neutral. The topic-neutral option involves a form of deflationism. So the phenomenal realist (“inflationist”) cannot be a physicalist. I have argued that both the mental and physical options are fine, so the argument collapses. In addition, my way out involves a notion of a phenomenal concept that has some affinities with the “directness” story in which there is no metaphysical mode of presentation at all. The doctrine that the metaphysical mode of presentation is the same as the referent says there is no metaphysical mode of presentation over and above the referent.
In the rest of this section, I will respond to a series of objections stimulated by a critique in White (forthcoming) of an earlier version of this chapter. White’s paper is not published, so I don’t want to ascribe anything very specific to him. The following renditions should be thought of as inspired by his paper rather than explicitly stated in it.
Objection: if phenomenal property Q is a physical property, then it can be picked out by a physical—say neurological—concept that identifies it in neurological terms. But those neurological identifications—e.g. in terms of radiation emitted by the brain processes that affect instruments that detect cortico-thalamic oscillation-- are irrelevant to first person phenomenal identifications, showing that the first person phenomenal identification depends on one aspect of the phenomenal property—its “feel”—rather than another aspect—its neurologically identifying parameters. That is, since the effect of cortico-thalamic oscillation on instruments is not part of the first person route by which we pick it out, it follows that not every aspect of the physical property is relevant to the first person route. Therefore (the objection continues) the identity ‘Q = cortico-thalamic oscillation’ is one in which the terms pick out a single referent via different properties of it, different metaphysical modes of presentation. And so the Property Dualism Argument has not been avoided.
Reply: It is true that the phenomenal concept and the neurological concept use different metaphysical modes of presentation (as well as different cognitive modes). But there is no problem in this. The metaphysical mode of presentation used by the phenomenal concept, viz., M, is nothing other than the phenomenal property itself. (M=Q). M is not a mere aspect of Q; M is Q. The neurological concept picks out Q (i.e. M) via a neurological property of it that is not Q itself, but rather an aspect of Q—e.g. the property of emitting electromagnetic radiation that affects instruments in a certain way. Q is identical to a neurological property and also has neurological properties in virtue of being a neurological property. The neurological concept picks it out via an aspect of Q that is not identical to Q, but the phenomenal concept picks it out via Q itself. So there are two metaphysical modes of presentation, and they are harmless. We can rationally wonder whether the thing that emits radiation affecting instruments in such and such a way has M. And we can wonder whether what has M emits radiation affecting instruments in such and such a way. So rational error in both directions is possible within the constraints set up in the account of phenomenal concepts.
It may seem that this reply puts too much emphasis on the neurological side of the identity. It might be thought to rely implicitly on something on the order of a malfunctioning fMRI machine that gives misinformation about the brain state, as suggested by Boyd (1980). This complaint seems to me to be an allusion to a form of argument very different from the Property Dualism Argument. Consider a mind-body identity claim in which both terms have metaphysical modes of presentation that are identical to the referent. Using ‘P’ to represent the physical side, let us represent it as ‘Q=P’. Using ‘M’ as before, M= Q, and since Q=P, it follows that the metaphysical mode of presentation of the ‘P’ side of the identity is also M. The physical side has the same metaphysical mode of presentation as the mental side. But if the metaphysical modes of presentation involved in both sides of the identity are the same, then (arguably) the identity cannot be a posteriori. Since the only viable non-aposteriori identity is a deflationist functional analysis, the upshot would be the same as that of the Property Dualism Argument: that the only physicalism is deflationist physicalism, and a posteriori physicalism and phenomenal realism are incompatible. I don’t have the space here to adequately discuss this argument, but my quick response is that what the argument shows is something that is independently plausible, namely that physicalistic concepts do not have metaphysical modes of presentation that are the same as the referent. Certainly the thought that Boyd’s point is defective should not be taken to show that we require such peculiar concepts.
Objection: But it does not make sense to suppose that cortico-thalamic oscillation is its own mode of presentation. At best, an aspect of cortico-thalamic oscillation might be a phenomenal mode of presentation.
Reply: the objection is question-begging. The physicalist claim is that Q = cortico-thalamic oscillation. Q is its own metaphysical mode of presentation and by Leibniz’s Law, cortico-thalamic oscillation is its own metaphysical mode of presentation—though as pointed out above, this is not the mode of presentation involved on the right side of the identity. The mode of presentation involved on the right hand side is not the property referred to (i.e. Q, i.e. cortico-thalamic oscillation) but an aspect of it. An aspect is the mode of presentation of the right hand side; the property itself is the mode of presentation on the left side. Note that the idea is not that on the left side, the property is presented under all aspects of it but rather that it is presented via itself.
Admittedly, there is an odd sound to these claims. Philosophers should remember the embarrassing history of ordinary language philosophy of the 1960s when it was commonly said as an objection to the claim that water = H2O that there is an odd sound to talk of a glass of H2O. However, I believe that the odd sound derives at least in part from the tendency—as we saw above—to switch from metaphysical modes of presentation to cognitive modes of presentation. If the switch is coupled with thinking of physical cognitive modes of presentation as explicitly physicalistic (see below), it can seem as if the mode of presentation of Q could not be physical.
Objection: To explain rational error, there must be a possible world in which the subject is in exactly the same epistemic situation as in the actual world but there is no error. However, your account does not allow for such a possible world. Of course, there is a world that contains Q but not some appearance of cortico-thalamic oscillation (as in Boyd’s point). But we can put that world aside by stipulating that there is no error on the scientific side. Still, one can imagine rationally believing that cortico-thalamic oscillation is present but Q not or conversely.
Reply: The physicalist cannot allow that rational error is always explained by a genuine possible world. Phenomenality without its physical basis and that physical basis without phenomenality are both imaginable, and even if the physicalist is right that phenomenality is its physical basis, it is possible to be rationally mistaken about the presence of one with the absence of the other. To require that these imaginable situations reflect genuine possibilities would immediately lead to dualism by Kripke’s and Chalmers’ familiar form of argument. I am not denying that this point may be the basis of an argument for dualism. Rather, my point is that the Property Dualism Argument would then be smuggling in another argument for dualism, whereas the reason for being interested in it is that it purports to be an argument for dualism that is distinct from Kripke’s and Chalmers’ arguments.
Objection: As the two tubes example shows, demonstrative concepts allow different routes to the referent. But different routes to the referent require different metaphysical modes of presentation. So in principle a subject could rationally wonder whether that [mental sample] phenomenal property is identical to that [mental sample] phenomenal property even if the mental sample is the same, showing that metaphysical mode of presentation cannot be identical to the phenomenal property of the mental sample. For since the mental sample is the same in both cases, if the metaphysical modes of presentation were the same in both cases, there could not be two different routes of reference and the subject could not rationally wonder whether the properties are one and the same.
Reply: We must be clear about the application in the two tubes type of cases of the type/token distinction. The two token demonstratives occur at different times, involving different token cognitive and metaphysical modes of presentation, the two metaphysical modes being different token time slices of the same mental sample. So the presupposition of the objection is mistaken, at least in the normal case.
Perhaps it is possible to think ‘that [mental sample] phenomenal property’ twice at one time. But if so, there are still two token cognitive modes of presentation. And as mentioned earlier, it is easy to see how the two token cognitive modes could be leveraged into two token metaphysical modes.
Readers may feel that I have so far focused on the easy part of the problem, concentrating as I have on metaphysical modes of presentation.
One problem in discussing the issue has to do with the vagueness of the notion of physicality that tends to be used in these discussions, and the tendency to lump two classes of physical properties together, the properties of physics and topic-neutral properties. To remind the reader of the connection between physicalism and physics, I will temporarily substitute the notion of micro-physical for the notion of physical. On the notion of physical that I have been using, a physical property is one that is expressible by a physical predicate. A micro-physical property is expressible by a micro-physical predicate. This slight narrowing down of the class of what we consider physical could not change the logic of the argument.
Note that there are two quite different things that could be meant by saying that the cognitive mode of presentation is micro-physical.
First, one might have an ontological thesis in mind—that the cognitive mode of presentation is identical to a micro-physical entity or property. On this understanding, there is no support for the idea that anyone who grasps the mental side of the mind-body identity must have an a priori available micro-physical understanding of the phenomenal property.
The second interpretation of the claim that a cognitive mode of presentation is micro-physical is that it is itself explicitly micro-physical or a priori analyzable in micro-physical terms, that is that the vocabulary or conceptual apparatus involved in the cognitive mode of presentation is micro-physical or micro-physically analyzable a priori. (For brevity, I will sometimes leave out the analyzability, focusing on the explicit physicality.) In this sense, ‘that very wet thing’ is not micro-physical. And in this sense, no phenomenal cognitive mode of presentation could be micro-physical. ‘That [mental sample] phenomenal property’ is not explicitly micro-physical because the term ‘phenomenal’ is not a term of microphysics (and not analyzable micro-physically) and because the mental sample is not a term at all (and not analyzable at all). If a cognitive mode of presentation is explicitly micro-physical, then White’s point applies: there is an a priori available coreferential description in micro-physical terms. And in this second sense, he is right that the mode of presentation (in the sense of cognitive mode of presentation) associated with a phenomenal concept is not micro-physical. But what reason is there to suppose that the cognitive mode of presentation must be explicitly micro-physical? The physicalist is committed only to its being micro-physical, not explicitly micro-physical. Physicalism is not the doctrine that all descriptions are in micro-physical vocabulary or are analyzable in micro-physical ideas-- or that all ways of picking out properties use micro-physical vocabulary or micro-physical ideas. The issue of explicitly micro-physical cognitive modes of presentation is irrelevant to physicalism.
As noted earlier, “cognitive mode of presentation” is a technical term, as is “metaphysical mode of presentation” and “mode of presentation”. Is there anything in the purpose we have for cognitive modes of presentation that requires that if the cognitive mode of presentation is physical, then it must be explicitly physical, or that it have no nature that is hidden from the subject who grasps it. The purpose of cognitive modes of presentation is explaining rational error. One can appeal to the subject’s thought about a phenomenal property using that phenomenal property to represent itself, even if the subject has no idea what the ultimate nature of that phenomenal property is. It can do its job without the subject knowing anything about its physicality or micro-physicality.
The upshot I think is that there is a sense in which if a cognitive mode of presentation is physical, there must be an a priori available equivalent physical description. But that sense (explicitly physical) is not relevant to any version of the Property Dualism Argument that has the right form to show that phenomenal properties are non-physical.
In my view, the considerations involved in the Property Dualism Argument have little to do with other prominent arguments for dualism: Jackson’s “Knowledge Argument”, Kripke’s modal argument and the modal arguments given by Chalmers and Jackson. But Loar, Perry and White disagree.
Loar 1997 locates the flaw in Jackson’s “Mary” argument and Kripke’s modal argument in a certain principle and Loar (1999) extends this analysis to Chalmers’ and Jackson’s modal arguments. (White, forthcoming, argues for a version of the principle and for its relevance to the Property Dualism Argument.) The principle is that if ‘Q=G’ is a posteriori, then at least one of the terms must refer via a contingent property of the referent. I agree with Loar that this principle (including White’s version) is false, but I think it is false independently of the mind-body problem and it has little to do with the Property Dualism Argument. To see the falsity, note that the person formed by a certain sperm = the person formed by a certain egg. This identity is a posteriori (the egg and sperm might have been named by a doctor prior to fertilization), yet both terms pick out their referents via essential properties of it. Call the sperm and egg that formed George W. Bush ‘Herbert’ and ‘Barbara’ respectively. The person formed from Herbert= the person formed from Barbara. “The person formed from Herbert” does not pick out George W. contingently, nor does “The person formed from Barbara”, yet one could explain rational error in the normal way. Of course there is some contingency in the vicinity. Herbert might have mated with a different egg. But this does not save the principle.
Even if Kripke is wrong about the necessity of origins, the logic of the example remains. One thing can have more than one necessary but insufficient property, and thus the terms in a true identity can pick out that thing, each term referring by a different necessary property as the mode of presentation. Since neither of these necessary properties need entail the other a priori, one might know one but not the other. So rational error is possible, whether or not the modes of presentation are necessary properties of the referent.
Even if we allow that M is a necessary property of Q, the Property Dualism Argument goes through as well or as badly. We can see this by going through each of the five options to see if the issue of whether M is a necessary property of Q makes any difference. (a) The issue of whether it is useless to postulate that M is mental doesn’t depend on whether M is or is not a necessary property of Q. (b) Next consider the issue of whether if M is physical, there would have to be a physical description that is a priori equivalent to ‘M’ or a priori coreferential with ‘Q’. Since the necessity in question is not a priori necessity, it is hard to see why that should make a difference. (c) Similarly, the issue of whether M is topic-neutral doesn’t depend on this matter. If I am right that Perry’s proposal substitutes a non-phenomenal concept for a phenomenal one, then that is so independently of whether M is a necessary property of Q. (d) If M is non-physical, then dualism is true whether or not M is a necessary property of Q. (e) If M does not exist, then M cannot be either a necessary or contingent property of Q. In sum, none of the five options have anything to do with whether M is a necessary property of Q.
The idea of Kripke’s, Chalmers’ and Jackson’s modal arguments for dualism is that we can conceive of phenomenal properties without brains and brains without phenomenal properties. If these scenarios are genuine possibilities, by a familiar form of argument, phenomenal properties cannot be identical to physical properties. I see the physicalist response as one of resisting the move from conceivability to possibility, or more concessively: whatever kind of possibility it is that zombies and disembodied minds have, it is not the kind that defeats mind-body identity. This point has little to do with whether M is a necessary property of Q.
The leading idea of Perry’s book is that a thought has a variety of “reflexive contents” that have the same truth conditions as the original thought. One reflexive content of my belief that Perry smokes is that the person who I am now attending to (assuming Perry is here) is in the extension of the property that is the object of my concept of being a smoker. A mind-body identity claim like Mary’s utterance of “Thisi sensation = B52” has as one of its reflexive contents that the sensation attended to by Mary has such and such scientifically specifiable properties. Perry recognizes that the concepts in a reflexive content may not be ones that the subject actually has, but he argues persuasively that they may be psychologically relevant nonetheless if the subject is “attuned” to these concepts in reasoning and deciding. “Attunement and belief are different kinds of doxastic attitudes…”(107)
Though the idea of reflexive contents is an interesting and useful contribution, I doubt that it does much work in dealing with the problems about phenomenality that Perry applies it to, namely the Property Dualism Argument, Jackson’s “Mary” argument for dualism, and the modal arguments. I agree with Perry’s solution to the “Mary” problem at one level of description—that what Mary learns is a new concept of a physical property that she already had a physical concept of in the black and white room. She acquires a phenomenal concept—which we can express as ‘thisi sensation’ of a property which she formerly conceptualized as ‘B52’. But as one can see from the fact that Mary actually has the phenomenal concept, there is no need to introduce Perry’s reflexive contents here. She actually has the phenomenal concept so we needn’t suppose that the situation is one in which she (a) doesn’t have the concept but (b) is merely “attuned” to it.
Suppose that when Mary sees red for the first time, she is sophisticated enough to have as an explicit content that thisi phenomenal property is the phenomenal character of the sensation of red that she already knew about in the black and white room under its physical description. If what sophisticated Mary learns is not a non-physical fact, then what naïve Mary (who doesn’t actually have the phenomenal concept but is only “attuned” to it) learns is not a non-physical fact either.
What is the application of Perry’s apparatus to the modal arguments? Why do we have the illusion that “Thisi sensation = B52” is contingent, given that (according to physicalism) it is a metaphysically necessary truth? Perry’s answer is that the necessary identity has some contingent reflexive contents such as: that the subjective character of red objects appears like so and so on an autocerebroscope, is called ‘B52’, and is what I was referring to in my journal articles. This idea has some initial plausibility, but that plausibility begins to dissipate when one asks oneself “Couldn’t I identify the brain state by its essential properties and still wonder whether I could have that brain state (so identified) without thisi phenomenal property?
It seems that the contingency that Perry was supposed to have explained away just reappears. More generally, the problem with Perry’s approach to Kripke’s problem is that it isn’t enough to find some contingent reflexive contents for the necessary mind-body claim. What is necessary, rather, is for Perry to make it plausible that the appearance of contingency fades once these contents are firmly distinguished from the necessary mind-body identity claim. It is this that Perry has not done and which I have given reason to doubt can be done.
In sum, the Property Dualism Argument can be avoided but not by the same means that serve to avoid Jackson’s “Knowledge Argument” or the modal arguments given by Kripke, Chalmers and Jackson.
 Deflationism with respect to truth is the view that the utility of the concept of truth can be explained disquotationally and that there can be no scientific reduction of truth. (Paul Horwich, Truth, Blackwell: Oxford, 1990. Second edition 1998, Oxford University Press: Oxford; Hartry Field, “Deflationist Views of Meaning and Content”, Mind 103, 1994: 249-285.) Deflationism with respect to consciousness in its most influential form is, confusingly, a kind of reductionism—albeit armchair reductionism rather than substantive scientific reductionism--and thus the terminology I am following can be misleading. I think I introduced this confusing terminology (in my review of Dennett in The Journal of Philosophy, p. 181-93, 1993.).
 I have suggested (1978) defining physicalism as the view that everything is decomposable into particles of the sort that make up inorganic matter. This definition would get the wrong result if pan-psychism obtains (electrons are conscious) and so it fails as a sufficient condition. It doesn’t capture the meaning of ‘physicalism’ (and it doesn’t even try to define ‘physical property’), but it does yield a necessary condition that is not without interest, especially if there simply is no full characterization of physicalism See also Montero, 1999.
 The quotation is from the 1990 version of Loar’s “Phenomenal States”, op.cit. p. 87. This picture is abandoned in the 1997 version of Loar’s paper in which he retains talk of triggering and the direct reference terminology, but with a new meaning, namely: refers but not via a contingent property of the referent. The view common to both the 1990 and 1997 paper is that a theoretical concept of, e.g. neuroscience might pick out a neurological property “that triggers a given recognitional concept, and so the two concepts can converge in their reference despite their cognitive independence…” (88)
Loar (1990, p. 98; 1997, p. 603) considers an analog of blindsight in which the recognitional concept is triggered by a physical property of the brain but without any phenomenality. (Presumably, he is not thinking of the recognitional concept as triggered by a brain property that is a phenomenal property, since then phenomenality would be instantiated.) The conclusion is that one cannot define ‘phenomenal concept’ as ‘self-directed recognitional concept’.
 The rationale for the functionalist understanding of this point of view is spelled out in Block, 1980 and in less detail in Block, 1994a. Lewis, 1981 adopts a more complex mixture of functionalism and physicalism.
 Instances in my terminology could be tokens or types, but I normally have types in mind. I will often just say ‘concept’ when I mean an instance. For example, consider the following—“Your concept of water and my concept of water may be instances of the same concept.” When I say your concept and my concept I am talking about your instance of that concept and my instance of that concept.
 Here I disagree at least on the surface with David Rosenthal (1997) and William Lycan (1990). Rosenthal holds that “[A] mental state is a conscious state when and only when it is accompanied by a suitable higher order thought”. Lycan says “I cannot myself hear a natural sense of the phrase ‘conscious state’ other than as meaning ‘state one is conscious of being in’.” I regard these claims as pertaining to consciousness in some sense other than phenomenality.
 Loar (1990, 1997) suggests a version of M=Q, but not Non-descriptivity or No Concept.