Philosophy 257 Behavior and Other Minds

Asst. Prof James Pryor Dept. of Philosophy

Strawson

He will also argue that to be able to understand such self-ascriptions, one has to recognize the existence of criteria for ascribing experiences to others.

Strawson begins by formulating two questions:

Q1. Why are one's states of consciousness ascribed to anything at all?
Q2. Why are they ascribed to the very same thing that material properties (weighing 150 lbs., being in the study, etc.) are ascribed to?

A necessary condition of states of consciousness being ascribed at all is that they should be ascribed to the very same things as certain corporeal characteristics, a certain physical situation, etc. That is to say, states of consciousness could not be ascribed at all, unless they were ascribed to persons, in the sense I have claimed for this word. (p. 102)

• On the no-owner view, experiences can't be correctly ascribed to any subject of experience at all. There is no subject which "owns" the experiences one is aware of.

• On the Cartesian view, experiences are correctly ascribed to a subject, but that subject is an immaterial soul, and so is not something that material properties (like weighing 150 lbs., or being in the study) can be correctly ascribed to.

  We should distinguish two versions of the Cartesian view. The radical Cartesian will say that the only sorts of properties that subjects of experience have are psychological properties. This is the position that Strawson seems to introduce as "the Cartesian view" on p. 94. The sensible Cartesian will say that subjects of experience have no intrinsic material properties or states, but allows that subjects of experience stand in certain relations to material objects.

On the other hand, we have views which hold that experiences are correctly ascribed to the same things that material properties are correctly ascribed to:

• On Williams' view, subjects of experience are just material objects. Among all the material objects, some of them have thoughts and experiences.

• Then there's Strawson view, that subjects of experience are persons, where this is a special, "logically primitive" concept (see pp. 101-103). This concept is primitive in the sense that it's not a compound concept of a body and some immaterial consciousness which is somehow connected to that body. It's a concept of a single thing which is an intrinsic bearer of both material and psychological properties.

  Unlike Williams, Strawson seems to think that persons are not themselves material objects.

  • One reason to think he thinks this is that he says the concept of a person is primitive. But if a person were just a special kind of material object--the thinking and feeling kind--then it's unclear whether he'd be entitled to say that.
  • Another reason is that he says that persons constitute "a major logical type or category of individuals." These words strongly suggest that he doesn't think that persons are just a subclass of material objects.
  • A third reason is that he allows for the possibility of disembodied persons (see pp. 103 and 115-16). Surely there's no such thing as a disembodied body.

  If Strawson does in fact think that persons are not themselves material objects, then we can think of him as holding one of the following two views:

  0. Persons coincide with material bodies (they have no parts over and above their bodies' parts) but they're not identical with those bodies. They stand in the same relation to their bodies as statues stand in to the lumps of clay that make them up: coincidence without identity.
  1. There is no material object occupying exactly the space a person occupies. Where persons are, there are no material objects. However, persons presumably have material objects as parts (hands and feet and the like).
  View (ii) is much less plausible than view (i). I will assume that view (i) is Strawson's view.

Strawson's Argument against the Cartesian

On p. 100 he writes:

To put it briefly. One can ascribe states of consciousness to oneself only if one can ascribe them to others. One can ascribe them to others only if one can identify other subjects of experience. And one cannot identify others if one can identify them only as subjects of experience, possessors of states of consciousness.

Let me rehearse briefly the stages of the argument. There would be no question of ascribing one's own states of consciousness, or experiences, to anything, unless one also ascribed, or were ready and able to ascribe, states of consciousness, or experiences, to other individuals of the same logical type as that thing to which one ascribes one's own states of consciousness. The condition of reckoning oneself as a subject of such predicates is that one should also reckon others as subjects of such predicates. The condition, in turn, of this being possible, is that one should be able to distinguish from one another, to pick out or identify, different subjects of such predicates, i.e., different individuals of the type concerned. The condition, in turn, of this being possible is that the individuals concerned, including oneself, should be of a certain unique type: of a type, namely, such that to each individual there must be ascribed, or ascribable, both states of consciousness and corporeal characteristics.

1. One can ascribe experiences to oneself only if one can ascribe them to others.

   Call the following Strawson's Principle 1 (SP1):

   It is a necessary condition of one's ascribing states of consciousness, experiences, to oneself in the way one does, that one should also ascribe them, or be prepared to ascribe them to others who are not oneself. (p. 99)

   Strawson modifies this principle slightly in his later discussion. We also see what I'll call SP1+:

   There would be no question of ascribing one's own states of consciousness, or experiences, to anything, unless one also ascribed, or were ready and able to ascribe, states of consciousness, or experiences, to other individual entities of the same logical type as that thing to which one ascribes one's own states of consciousness (p. 104; my underlining)

   And what I'll call SP1++:

   There is no sense in the idea of ascribing states of consciousness to oneself, or at all, unless the ascriber already knows how to ascribe at least some states of consciousness to others. (p. 106; my underlining)

   There's some difficulty in interpreting these claims. On one reading, claims about whether one is able or ready or prepared or knows how to ascribe experiences to others are claims about whether one would be able to justifiably ascribe experiences to others. This reading is supported by the first part of the footnote on p. 99. There Strawson says that being able or ready to ascribe experiences to others entails having a conception of what the appropriate conditions for ascribing those experiences to others would be. (I'm understanding "appropriate conditions" here to mean "conditions which warrant the ascription.")

   On the other hand, we might understand SP1 and its variants just to be claims about whether one would understand an ascription of experiences to others. This reading is supported by what Strawson says in the latter part of the footnote on p. 99. There he says that SP1 (and presumably its variants) are supported by the idea that "a predicate is correlative with a range of distinguishable individuals of which the predicate can be significantly, though not necessarily truly, affirmed" (p. 99). This premise would not support SP1 on the first reading.

   I cannot tell which of these readings Strawson intended. I suspect he did not clearly distinguish them in his mind while writing this chapter. So we'll have to keep both readings in mind. Let's take SP1+ as the canonical statement of Strawson's first premise. Then the two readings come out as:

   SP1a. To be able to understand [and justifiably believe?] ascriptions of experience to oneself, one must have some conception of what would justify ascriptions of experiences to other things "of the same logical type" as oneself.

   SP1b. To be able to understand ascriptions of experience to oneself, one must be able to understand ascriptions of experience to other things "of the same logical type" as oneself.

   SP1b sounds plausible. SP1a, on the other hand, sounds a lot like the criteria theorist's claims about what's required to understand concepts like "pain." We haven't yet seen convincing arguments for those claims. As we discussed, someone who thinks we know about others' mental states by analogy or by inference to the best explanation (or by perception of their mentally-loaded behavior, for that matter) will see no incoherence in the possibility that a person knows what pain is without yet knowing what would justify the ascription of pain to someone else. So an argument for SP1a would have to show that these other, non-criterial epistemologies are defective.

2. One can ascribe experiences to other subjects of experience only if one can identify those other subjects.

   This is Strawson's Principle 2 (SP2). He first states it as:

   Surely there can be a question of ascribing only if there is or could be a question of identifying that to which the ascription is made. (p. 100)

   I interpret this as:

   SP2. It's possible for you to make or understand a particular ascription of experience to another subject only if you can identify that other subject.

   When Strawson first states SP2, he doesn't make it clear that it's a principle he endorses, but the summaries of his argument on p. 100 and p. 104 make it clear that this is a principle he accepts and relies on.

   What does Strawson mean by saying that one "can identify" another subject? Well, here are some definitions and claims from earlier in the book:

   • A hearer identifies the particular a speaker referred to iff he knows what particular the speaker was talking about (p. 16)
   • A speaker makes identifying reference to X iff (roughly, and not in Strawson's own terms) the speaker asserts some singular proposition about X (p. 16)
   • A speaker identifies X iff the speaker makes identifying reference to X, "on the strength" of which his hearer identifies the particular he was referring to (p. 16)
   • It's possible for us to identify Xs only if it's possible for us sometimes to make each other understand which X we're talking about (p. 16)
   • A sufficient condition for a hearer identifying the particular referred to is that he be "able directly to locate the particular referred to," where that means he can pick it out by sight or touch etc., knowing that it is the particular referred to (pp. 18-19)
   • A person reidentifies a particular iff he knows of it that it is the same individual that he encountered on another occasion, or the same individual as was referred to in another context (p. 31)
   • A person has to be able to identify who's in pain when he is himself in pain, in the sense of being able to refer to himself, and attribute pain to himself; he doesn't have to identify who's in pain in the sense of "figuring out" who the pain belongs to (p. 100). It's the second kind of identification that Shoemaker argues is absent in the case of self-ascriptions of experience.

   The relevant claims here seem to be the ones in red. However, if we understand SP2 in the way those claims suggest, then it will be a necessary condition for you ever to make or assess an ascription of experience to another subject that you have an audience when you do so, or that you be someone else's audience. That seems very implausible. Presumably it's enough if one is able to pick out or distinguish one other subject from other things; one doesn't in addition need an audience. This weaker reading is supported by Strawson's restatement of SP2 on p. 104:

   The condition, in turn, of this being possible [viz., of reckoning others as subjects of experience], is that one should be able to distinguish from one another, to pick out or identify, different subjects of such predicates, i.e., different individuals of the type concerned.

3. One cannot identify other subjects of experience if subjects of experience are Cartesian subjects.

   Strawson claims that it's impossible to identify Cartesian subjects. His argument goes as follows:

   But how is it that one can ascribe [states of consciousness] to others? Now one thing here is certain: that if the things one ascribes states of consciousness to, in ascribing them to others, are thought of as a set of Cartesian egos to which only private experiences can, in correct logical grammar, be ascribed, then this question is unanswerable and this problem insoluble. If, in identifying the things to which states of consciousness are to be ascribed, private experiences are all one has to go on, then...there is...no question of telling that a private experience is another's. (p. 100)

   Two comments. First, I cannot fit the ellided passages into any reasonable reconstruction of the argument. Hence I omitted them, so that we're not distracted by them. Second, we face the question: Why does Strawson conclude it is not possible to tell that a certain experience is someone else's? Is this because he takes it that having only private experiences to go on in identifying others makes it impossible to identify others? Call that suggestion 1. Or is it because he takes it that it's impossible to know of a Cartesian soul what experiences it's having? Call that suggestion 2. So far we've seen no argument for suggestion 2. Hence charity demands that we interpret Strawson in way 1. The argument then becomes:

   It's impossible to identify Cartesian subjects, because the only properties one can use to identify Cartesian subjects are their psychological properties, and it would be impossible to identify another subject solely by its psychological properties.

   Let's grant for the sake of argument that it would be impossible to identify another subject solely by its psychological properties. Why does Strawson think the only properties one can use to identify Cartesian subjects are their psychological properties?

   Recall our earlier distinction between the radical Cartesian and the sensible Cartesian. The radical Cartesian says that the only sorts of properties that subjects of experience have are psychological properties. The sensible Cartesian says that subjects of experience have no intrinsic material properties or states, but allows that subjects of experience stand in certain relations to material objects. Strawson might be assuming in this argument that a Cartesian has to be a radical Cartesian. However, he has no right to assume that. On the other hand, Strawson might instead be assuming that the only properties we can use to identify a particular are its intrinsic properties. Hence, since all Cartesians will agree that the only intrinsic properties a subject of experience has are psychological properties, they'll all have to say that we can only use psychological properties to identify other subjects of experience. However, it's unclear why we should only be allowed to use intrinsic properties to identify particulars. (Indeed, if you look at Chapter 1 of Strawson's book, you'll see that his examples of identifying particulars often employ relational properties.) So it's unclear to me how Strawson's argument here is supposed to work.

   Hence, Strawson has to explain why a Cartesian isn't permitted to say we identify other subjects of experience by reference to the relations they stand in to material bodies. He does entertain such an objection on behalf of the Cartesian:

   There is no difficulty in distinguishing bodies from one another, no difficulty in identifying bodies. Does not this give us an indirect way of identifying subjects of experience, while preserving the Cartesian mode? Can we not identify such a subject as, for example, "the subject that stands to that body in the same special relation as I stand in to this one," or, in other words, "the subject of those experiences which stand in the same unique causal relation to body N as my experiences stand in to body M"? (p. 100-101)

   Strawson responds to this objection as follows:

   This suggestion is useless. It requires me to have noted that my experiences stand in a special relation to body M, when it is just the right to speak of my experiences at all that is in question. That is to say, it requires me to have noted that my experiences stand in a special relation to body M; but it requires me to have noted this as a condition of being able to identify other subjects of experiences, i.e., as a condition of my having the idea of myself as a subject of experience, i.e., as a condition of thinking of any experiences as mine. (pp. 101)

   What's going on here? I don't understand this argument.

   Is Strawson arguing that the Cartesian is committed to the claim that noting that my experiences stand in certain relations to my body is a necessary condition of being able to understand ascriptions of experience to myself? There doesn't seem to me to be anything incoherent in that claim.

   Or is Strawson arguing that the Cartesian is committed to the claim that noting that my experiences stand in certain relations to my body is a precondition of being able to understand ascriptions of experience to myself? That claim may very well be incoherent; but I can't see why the Cartesian is committed to it. Strawson never argued that being able to identify others is a precondition of being able to understand self-ascriptions of experiences; he just argued that it necessarily has to accompany the ability to understand self-ascriptions. So I don't see why the Cartesian has to say the incoherent thing Strawson here attributes to him.

Evaluating the Argument

SP1b. To be able to understand ascriptions of experience to oneself, one must be able to understand ascriptions of experience to other things "of the same logical type" as oneself.

SP2. It's possible for you to make or understand a particular ascription of experience to another subject only if you can pick out or distinguish one other subject from other things.

3. It's impossible to identify subjects of experience unless those subjects are the intrinsic bearers of some material properties.

I'm not sure it does. The first premise says that you have to be able understand what it would be for ascriptions to others to be true. The second premise talks about what's required to understand particular ascriptions to others. I don't see why that's relevant to the requirement imposed by the first premise. Suppose we believed that poltergeists exist, even though we lacked the means ever to pick out or identify a particular poltergeists. Couldn't we nevertheless understand what it would be for a poltergeist to have certain properties (e.g., being benign, being crafty, etc.?) If we then ever acquired a means of picking out or identifying particular poltergeists, we could then make and understand ascriptions to particular poltergeists. But we needn't then be learning for the first time what it means to say that a poltergeist is benign, or crafty, or whatever.

A second worry I have is that Strawson's discussion seemed more to concern my conception of myself than what my nature in fact is. So I'm not sure that Strawson has an argument that I have to in fact be "of the same logical type" as certain other things. Perhaps he only has an argument that I have to think of myself as being "of the same logical type" as those other things.

In any case, even if Strawson's argument here did establish that understanding self-ascriptions required one to be "of the same logical type" as things which are intrinsic bearers of both psychological and material properties, this would fall short of establishing Strawson's announced conclusion in two respects:

0. Strawson does not argue for the claim that subjects of experience are subjects of the sort he describes (i.e., things which coincide with material bodies but are not themselves identical to any material body), as opposed to subjects of the sort Williams describes (i.e., material bodies which happen to have thoughts and experiences).

1. Even if we grant Strawson that the only things which could be intrinsic bearers of both psychological and material properties are subjects of the sort he describes, that doesn't yet establish that I am a subject of that sort. It only establishes that I am "of the same logical type" as a subject of that sort. Does that rule out my being a Cartesian subject?

   Strawson may not be entitled to say it does. For he says that it's possible for there to be a disembodied subject (pp. 103 and 115-16), and that such a subject can still make and understand ascriptions of experience to himself and to other, embodied subjects. Presumably, then, such a subject is "of the same logical type" as other embodied subjects. But such a subject does not then bear any intrinsic physical properties. So what's to rule out the possibility of there being a subject who never bore any intrinsic properties and who is also "of the same logical type" as embodied subjects? Strawson does not offer any convincing account of why disembodied subjects can be "of the same logical type" as embodied subjects, while Cartesian subjects can not be.

Strawson's Argument for Criteria

Clearly there is no sense in talking of identifiable individuals of a special type, a type, namely, such that they possess both M-predicates and P-predicates, unless there is in principle some way of telling, with regard to any individual of that type, and any P-predicate, whether that individual possesses that P-predicate (p. 105)

In any case, I'll assume that Strawson intended SP3 in the third, weakest sense. So understood, it serves merely a rhetorical function and plays no role in his argument for criteria.

His argument then proceeds as follows:

0. If we didn't have criteria for the ascription of experiences to others, then the only way we could tell whether others had those experiences would be on the basis of behavioral signs.

1. The only way we could tell that certain behavior counts as a sign of an experience in others would be by observing a correlation between the behavior and the experience.

   We've already seen Chihara and Fodor attack this assumption.

2. We could only make such observations in our own case.

3. But by SP1++, it's impossible to ascribe experiences to oneself unless one already knows how to ascribe them to others.

   So Strawson's argument for criteria relies on a version of SP1. Does it rely on SP1a or on SP1b? Well, SP1b merely says that understanding ascriptions of experience to oneself requires one to be able to understand ascriptions of experience to others. That wouldn't show that it's impossible to figure out what justifies one in ascribing experiences to others by observations of one's own case. So SP1b doesn't provide good reason to accept this tep in Strawson's argument. So Strawson must be appealing to SP1a here, instead. SP1a says that understanding ascriptions of experience to oneself requires one to have some conception of what would justify ascriptions of experiences to others. That principle would serve Strawson's purposes here, and it seems to be the principle he is relying on. However, as we've already said, SP1a sounds like the claims criteria theorists make, and I don't see why anyone who hadn't already accepted criteria theory would find SP1a plausible. Hence Strawson's argument for criteria is unconvincing.