J.Fodor. Individualism and Supervenience, from Psychosemantics.

Presented: September 27, 2001

The task from the end of chapter one

Common sense psychology would be vindicated if a theory of the min proved to be committed to entities which are both semantically evaluable and etiologically involved.  RTM looks like such a theory; so, if RTM is true, common sense is vindicated.  However, RTM needs to make an empirical case.  That is, we need a good accounts, independently confirmed, of mental process as causal sequences of transformations of mental representations.  For Fodor, modern cognitive psychology is devoted to confirming such accounts.  What he wants to do in the remainder of the book is deal with the doubts about RTM that turn on its semantic assumptions.

Preamble: Putnam and Burge

1.  Two stories, one aim. The Putnam and Burge stories are offered up as having morals that tend to undermine the notion of content and thereby raise problems for propositional-attitude-based theories of mind.

2.  Intuitions

     a.  Twin-Earth Intuitions: the intuitions that we are to share are: 1) the form of words ‘water is wet’ means something different for Twin-Me than it does for me.  2) The content of the thought that Twin-Me has when he thinks that water (XYZ) is wet is different from the content of the thought that I have when I think that water (H₂0) is wet.

     b.  Brisket Intuitions:  the intuitions that we are to share are: 1)  If a butcher can bear attitude A toward the proposition that brisket is F, so too can Oscar; that is, Oscar has the concept BRISKET.  2)  Oscar2 doesn’t have brisket-attitudes, for he does not have the concept BRISKET; he has the concept BRISKET2.

3.  Ground Rules (assumptions).

    a.  For the sake of argument, grant the intuitions.

    b.  The Burge story shows that if the Putnam story raises any problems for the notion of content, then the problem are completely general and affect all content-bearing mental states.

    c.  Putnam and Burge appear to suggest that the attitudes are in some sense individuated with respect to their relational properties.

4.  Terminology

    a.  Standards of individuation are relational if my beliefs differ in content from my Twin’s.

    b.  Operative standards are nonrelational if attitudes are individuated in such a fashion that my beliefs and my Twin’s are identical in content

    c.  Standards of individuation according to which my Twin and I are in the same mental state are individualistic.

Supervenience

1.  The Problem posed by supervenience: Although my twin and I have different propositional attitudes, we are in identical brain states.  So it seems that the Burge and Putnam stories show that propositional attitudes don’t supervene on brain states.  But supervenience seems to be indispensable, for it seems to be the only plausible explanation so far about how mental state causation is possible.  The moral: we can’t make respectable science out of the attitudes as commonsensically individuated.

2.  Reactions to the problem raised by supervenience

    a.  Concession: We need identity conditions for mental states other than those that common sense prefers.

    b.  So what’s the problem?  To get a violation of supervenience, you need not just the relational individuation of mental states; you also need the nonrelational individuation of brain states.  The Burge and Putnam examples imply only the former.  That is, my brain states are type-identical to my Twin’s only if it is assumed that relational properties do not count for the individuation of brain states.  The question is: why should we assume that?  If we don’t assume it, then supervenience is simply not an issue.  Fodor’s response: the suggestion that brain states should be relationally individuated is plain silly.

    c.  Revisionist account.  We proceed with principles of individuation that violate supervenience, and accept the loss of supervenience.  Fodor’s response: Such a proposal cannot be sustained.  The considerations that require the nonrelational individuation of mental states are no different from the ones that require the nonrelational individuation of brain states.

Strategy

1.  First, it will be shown why we think that brain states and the like should be individuated nonrelationally.  This involves developing a sort of metaphysical argument that individuation in science is always individualistic.

2.  It will be shown how the contrary opinion could ever have become prevalent.

3.  It will be made clear what notion of content survives this shift in criteria of individuation

Causal Powers

1.  The Initial argument.

     a.  The stage: We define ‘is an H-particle at t’ so that it is satisfied by a particle at t iff my dime is heads-up at t.  Correspondingly, we define ‘is an T-particle at t’ so that it is satisfied by a particle at t iff my dime is tails-up at t.

    b.  ‘is H at t’ and ‘is T at t’ legitimately used as predicates

    c.  ‘is H at t’ and ‘is T at t’ wrongfully used in particle physics

        i.  H-particles and T-particles have identical causal properties.  Hence, the properties of being H or T are taxonomically irrelevant for purposes of scientific causal explanation.

        ii.  Similarly, mutatis mutandis, for the properties of being H and T brain states.

        iii.  Similarly, mutatis mutandis, for the properties of being H and T mental states.

    d.  Conclusion: If you’re interested in causal explanation, it would be mad to distinguish between Oscar’s brain states and Oscar2’s; their brain states have identical causal powers.  That’s why we individuate brain states individualistically.  And if you are interested in causal explanation, it would be mad to distinguish between Oscar’s mental states and Oscar2’s; their mental states have identical causal powers.  But common sense deploys a taxonomy that does distinguish between the mental states of Oscar and Oscar2.  So the commonsense taxonomy won’t do for the purposes of psychology.

2.  Objections

    a.  First objection:  The causal powers of my water-utterances do, after all, differ from the causal powers of my Twin’s ‘water’-utterances.

         Reply: Context is important

    b.  Second Objection:   First, we must consider the behavioral consequences of the mental states of Oscar Oscar2.  Oscar’s desires might eventuate in his saying such things as that he prefers brisket to other things; Oscar’s thought may lead to his evincing brisket-eating preferences and brisket-purchasing behavior.  However, what does this have to do with Oscar2?  Well, Oscar may also say that he likes brisket2, he may also evince brisket2 preferences, and he may behave brisket2-purchasingly.  And when Oscar2 says and does things with brisket2 in mind, he may very well produce precisely the same bodily motions as his counterpart produces when he says and does the corresponding things with brisket in mind.  However, all this shows is that behaving isn’t to be identified with moving one’s body.

     Second, there is a strong motivation for denying that Oscar’s and Oscar2’s behavior are, in general, type identical.  Behavior is the result of mental causes, and hence different mental causes should eventuate in correspondingly different behavioral effects.  By assumption, the twin’s attitudes differ a lot, so if these very different sorts of mental causes nevertheless invariably converge on identical behavioral effects, that would seem to be an accident on a very big scale.  The way out is to deny that the behavioral effects are identical; that is, to insist that the commonsense way of identifying behaviors goes out into the world for its principles of individuation, and that it depends essentially on the relational properties of the behavior.

        First reply: If this argument shows that my mental state differs from my Twin’s, it’s hard to see why it doesn’t show that our brain states differ too.

        Second reply:  First, we must consider if there is a way to distinguish H-particles from T-particles.  One proposal: being H rather than T does affect causal powers after all; for H-particles enter into H-particle interactions, and no T-particle does.  What does this do to a causal theory?  We can save classification by causal powers by changing the criteria for event identity.  But does this not lead to a dead end?

It looked like the notion of taxonomy by causal powers gave us a sort of a priori argument for individualism and thus put some teeth into the idea that a conception of mental state suitable for the psychologist’s purposes would have to be interestingly different from the commonsense conception of a propositional attitude.  But now it appears that the requirement that states with identical causal powers ought ipso facto to be taxonomically identical can be met trivially by anyone prepared to make the appropriate ontological adjustments.

3.  Salvaging Causal powers: You can affect the relational properties of things in all sorts of ways.  But for one thing to affect the causal powers of another, there must be a mediating law or mechanism.  It’s a mystery what this could be in the Twin or Oscar cases; not surprisingly, since it’s surely plausible that the only mechanisms that can mediate environmental effects on the causal powers of mental states are neurological.  The way to avoid making this mystery is to count the mental states—and, mutatis mutandis, the behaviors—of Twins (Oscars) as having the same causal powers, hence as taxonomically identical.

4.  Conclusion:

     Methodological point: Categorization in science is characteristically taxonomy by causal powers.  Identity of causal powers is identity of causal consequences across nomologically possible contexts.

     Metaphysical point:  Causal powers supervene on local microstructure.  In the psychological case, they supervene on local neural structure.

Mental Content

    Preamble:  There is a difference between the way psychology individuates behaviors and mental states and the way common sense does.  For Fodor, there seems to be little reason to doubt this.  However, psychology needs the commonsense notion of mental content.  So what are we to make of this?  Does the notion of mental content survive the transition from the layman’s categories to the scientist’s?

1.  The Twin-Earth problem. We have made the following assumptions:

    a. We have assumed a typology for mental states according to which my thoughts and my Twin’s have identical contents.

    b. We have assumed a typology according to which the physiological identity of organisms guarantees the identity of their mental states.

a) and b) are entailed by the principle that the mental supervenes upon the physiological, together with the assumption that mental states have their contents essentially, so that typological identity of the former (i.e. mental states) guarantees typological identity of the latter (i.e. contents). [1]   However, the identity of the contents of mental states does not ensure the identity of their extensions: two tokens of the same thought can have different truth conditions.  Hence, if mental state supervenes upon physiology, then thoughts don’t have their truth conditions essentially; if thoughts are in the head, then content doesn’t determine extension.

            However, Fodor claims that the Twin-Earth problem isn’t really a problem: it is simply a handful of intuitions together with a commentary on some immediate implications.  What connects the intuitions and their implications with the proposal that we give up on propositional-attitude psychology is a certain diagnosis.

2.  The Diagnosis:  If we want to have mental contents, we owe some sort of account of their individuation.  Prior to Twin-Earth, there was some sort of account of their individuation.  It was at one point held that identity of content depends on identity of extension.  That is, inferences from differences of extension to difference of content used to bear almost all the weight of propositional-attitude attribution.  But it seems as if Twin-Earth shows us that it simply is not true that differences of extension implies difference of content.  The price of supervenience is thus evident: the consequence of the psychologist’s insistence on preserving supervenience is that we now have no idea at all what criteria of individuation for propositional attitudes might be like

3.  A Mistaken Diagnosis: This diagnosis rests on a trivial mistake: the Twin-Earth examples do not break the connection between content and extension; they just relativize it to context.

4.  Contents and truth conditions:

    a.  Prior to Twin-Earth, contents were understood to be functions from thoughts to truth conditions.  But even if all the intuitions about Twin-Earth are right, and even if they have all the implications that they are said to have, extensional identity still constrains intentional identity because contents still determine extensions relative to a context.

    b.  Extensional identity criterion for mental contents: Two thought contents are identical only if they effect the same mapping of thoughts and contexts onto truth conditions.

    c. narrow vs. broad content

etiology  n. 1. the assignment of a cause  2. the science of causes or origins.

nomology n. 2. the branch of a science that investigates and formulates the principles          governing its phenomena.

taxonomy  n. 1. the science of classification; laws and principles covering the classifying   of objects.

(added by chris)

Fodor

• Genearal approach: A. There is a Twinearth problem. The moral is that attitudes are relationally individuated (i.e. ‘wide’). B. It seems relational individuation violates supervienience because then all physical states can be identical by the attitudes can be different. C. Furthermore, all scientific explanation relies on non-relational individuation, because scientific explanations are causal and causal powers don’t differ in the relation individuation case. D. So for psychological explanations, we need a non-relational notion of individuation of mental states. E. This makes Fodor distinguish methodological individualism from methodological solipsism (42). F. So scientific and common sense psychological individuations are different (the former ‘wide’, the latter ‘narrow’). G. So is commonsense content relevant? Notably, content won’t determine extension if we go with scientific psychology, and it seems we won’t be able to individuate content as we did before! H. Solution: relativize individuation/truth conditions to a context (This violates the original intuitions in the Twinearth problem and needs a good explanation of context individuation) J. So, what’s meaning then? A: It’s radically inexpressible because trying to express it relies on a language that has to have wide content. Narrow content is content ‘only potentially’.

Pitt

• Error at the bottom of page 7.

<course home page>