Kripke on the distinctness of the mind from the body

Jeff Speaks

April 25, 2005

At pp. 144 ff., Kripke turns his attention to the mind-body problem. The discussion here brings to bear many of the results from earlier in the book, including rigid designation, the necessity of identity, the possibility of necessary a posteriori truths, and explanations of the illusion that certain a posteriori necessary truths are contingent.

1 Three identity theories

An identity theory of some mental phenomenon is a theory which says that that phenomenon (state, property) is identical to some physical phenomenon. Kripke argues that no identity theory can be correct. In doing so, he distinguishes between three different identity theories:

“Identity theorists have been concerned with several distinct types of identifications: of a person with his body, of a particular sensation (or event or state of having the sensation) with a particular brain state (Jones’s pain at 6:00 was his C-fiber stimulation at that time), and of types of mental states with the corresponding types of physical states (pain is the stimulation of C-fibers).” (144)

These are not so much three varieties of identity theory as they are three different topics which one can be an identity theorist about: substances which have mental properties, the mental properties themselves, and instances, or tokens, of those mental properties. In each case, we have an opposition between identity theories and dualist theories (along with, perhaps, some intermediate positions): substance identity theory/dualism, property (type) identity theory/dualism, and token identity theory/dualism.

2 Three arguments against identity theories

Kripke presents arguments against all three sorts of identity theories. The arguments themselves, as Kripke says, have been around at least since Descartes; the main original contribution of Kripke’s discussion is to show that certain apparently convincing replies to the Cartesian arguments are unsuccessful.

The form of the Cartesian argument is to point out the possibility of some mental phenomenon without the physical phenomenon alleged to be identical to it by the identity theorist, and to conclude from there that the mental and physical phenomena are actually not identical. The three versions of this argument which Kripke gives are:

Argument against person-body identity theories:

Possibly, this person /= this body.

This person /= this body.


Relevant scenarios: I exist without this body, or this body exists without me.

Argument against token identity theories:

(Following Kripke (p. 146), let ‘A’ be the name of a particular pain sensation, and let ‘B’ name the brain state with which A is claimed to be identical. Think of ‘A’ as meaning ‘this pain.’)

Possibly, A /= B.

A /= B.


Relevant scenarios: this very pain sensation exists without this very brain state, or this brain state exists without this pain sensation.

Argument against type (property) identity theories:

Possibly, pain /= C-fiber firing.

Pain /= C-fiber firing.


Relevant scenarios: some creatures experience pain without their C-fibers firing, or some creatures have C-fibers which fire but no pain.

In each case, the premise is that some non-identity is possibly true. We establish this by considering a possible world in which one of the terms of the identity, but not the other, does not refer. (These possible worlds are the ‘relevant scenarios’ listed alongside the arguments.) The identity statement in question is thus false with respect to this possible scenario, and hence possibly false. This is what each of the premises of the Cartesian arguments say. We then conclude from the fact that the identity possibly fails to hold that it actually fails to hold. (More on this transition below.)

3 Materialist responses

Faced with these arguments, the materialist has two options: claim that the argument is invalid, or deny the premise. The material earlier in Naming and Necessity provides arguments which count against either move.

3.1 Contingent identities

Suppose that we take the first option, and deny that the arguments are valid. Then we must deny that the relevant identities are necessary if true since, if they were necessary if true, their possible falsehood would show their falsehood simpliciter. But don’t we come across contingent identities all the time -- for example, ‘The first Postmaster General of the United States = the inventor of bifocals’? Given examples like this, it might well seem that the Cartesian owes the materialist an explanation of why the identities posited by the materialist cannot just be contingent identities.

But recall our discussion of Kripke’s treatment of identity sentences. There we were given a convincing argument for the necessity of identity, and an explanation of apparent contingent identities as involving non-rigid designators. So, if this response to the Cartesian is to be convincing, it must be the case that some of the terms which figure in the relevant identities are non-rigid. Are they?

Consider the first two arguments, since the third raises its own difficulties. It seems that none of the four terms which figure in these two identity claims are non-rigid; ‘this person’ and ‘this body’ seem to rigidly designate me and my body, respectively, and ‘A’ and ‘B’ were supposed to be introduced as names for particular sensations and particular brain states/events. So it is plausible that each of these four terms is a rigid designator. If so, each of the identities will be necessary if true, and this is enough for the Cartesian arguments to be valid.

But it is important to see that there is a sense in which it doesn’t matter much whether the particular terms used in these arguments are rigid designators; for, if they are not, we could replace them with rigid designators. All the Cartesian needs to run his arguments against the identity theorist is that there be some rigid designators or other which refer to the relevant substances and events/states. Accordingly, the anti-Cartesian proponent of contingent identities must show that we cannot rigidly designate the items in question. But it is difficult to see why this should be so.

Consider how the third argument, having to do with mental properties. This presents extra complications because the class of statements to which Kripke assimilates the view of type identity theorist presents extra complications:

“The final kind of identity the type-type sort of identity exemplified by the identification of pain with the stimulation of C-fibers. These identifications are supposed to be analogous with such scientific type-type identifications as the identification of heat with molecular motion, of water with hydrogen hydroxide, and the like.” (148)

Kripke thinks, as we have seen, that these scientific identifications are necessary if true; so he thinks that the claim of the type identity theorist is also necessary if true.

Here we run into a problem with the evaluation of Kripke’s argument. In our discussion of theoretical identities, we found Kripke’s claim that theoretical identifications are necessary if true to be plausible, but found his argument for this claim to be unconvincing. The problems with the argument were that (i) it rests on an under-specified extension of the term ‘rigid designator’ from singular terms to predicates, and (ii) it rests on treating theoretical identities as identity sentences, whereas in many cases (‘Cats are animals’) it is clear that they cannot be treated in this way, but must rather be treated as universally quantified conditionals and biconditionals.

One way around this problem is as follows. Even though we cannot treat all of the statements with which Kripke was concerned as identity statements, we can treat the claim of the type identity theorist as an identity statement. It is simply an identity involving a property rather than one involving an object. So, in this context, ‘pain’ is to be thought of as functioning not as a predicate, but as a singular term which refers to a property. Construed this way, the claim that ‘pain’ is a rigid designator is perfectly well defined: it is a rigid designator just in case it refers to the same property with respect to every possible world. Same goes for ‘C-fiber firing.’ Then we can block the present reply to the Cartesian argument as above.

A reply by the identity theorist: ‘pain’ is not a rigid designator. A reply on behalf of the Cartesian.

3.2 A posteriori necessities and the illusion of contingency

So we have headed off one line of reply to the Cartesian arguments: the claim that the arguments are invalid. There’s only one other possible reply: that the arguments are valid, but unsound. Since the arguments have only one premise, the only way to go here is to argue that the claims about possibility employed as the premises in these arguments are false. But, since these premises would clearly be true if the ‘relevant scenarios’ listed above were possible, this reply amounts to the claim that these seemingly possible scenarios are not really possible.

Why, without argument, this sounds implausible; a link between conceivability and possibility, and the conceivability of the relevant scenarios.

The anti-Cartesian might back up his claim by using Kripke’s own work. After all, Kripke has shown that there can be a posteriori necessities, and has shown that in the case of a posteriori necessities there will usually be an illusion that the claims in question are contingent rather than necessary. But we can explain this intuition away, as Kripke has shown us. Let’s recall how this works.

Let’s focus on the claim that gold is the element with atomic number 79. Initially, this seems contingent, since we can imagine, as we put it, ‘it turning out that gold has a different atomic number.’ But if we can imagine this, can’t we clearly conceive of it’s being the case that gold does not have atomic number 79? Kripke thinks not. He thinks that what we imagine when we imagine it’s turning out that gold does not have atomic number 79 is a situation qualitatively identical to our actual one, in which someone discovers that some stuff with has the same superficial identifying marks as gold is a compound rather than an element. But, if we think carefully about this situation, we will see that we are not really inclined to say that this is a scenario in which gold is a compound. Rather, we are inclined to say that it is a scenario in which the stuff which looks like gold is not really gold. Thus we have a proposition which seemed clearly to be possible -- that gold is a compound rather than an element -- which turns out on closer inspection not to be possible.

Now it seems that we have given the anti-Cartesian all that he could want. We have explained that certain initial intuitions about possibility can be false, and have even offered a plausible explanation of why these illusions of contingency arise. Can’t the Cartesian just say, at this point, that our intuitions that I could exist without my body, or this sensation without this brain state, or pain without C-fiber firing, are just like this? On this view, the identities posited by the identity theorist are genuine a posteriori necessities, and our intuitions that they are possible false are illusions, to be explained in terms of a confusion of qualitatively identical scenarios with genuine counterinstances.

Kripke thinks not:

“Now I do not think it likely that the identity theorist will succeed in such an endeavor. ...What was the strategy used above to handle the apparent contingency of certain cases of the necessary a posteriori? The strategy was to argue that although the statement itself is necessary, someone could, qualitatively speaking, be in the same epistemic situation as the original, and in such cases a qualtatively analogous statement could be false. ...

Now can something be said analogously to explain away the feeling that the identity of pain and the stimulation of C-fibers, if it is a scientific discovery, could have turned out otherwise? I do not see that such an analogy is possible. In the case of the apparent possibility that molecular motion could have existed in the absence of heat, what seemed really possible is that molecular motion should have existed without being felt as heat, that is, it might have existed without producing the sensation S, the sensation of heat. In the appropriate sentient beings is it analogously possible that a stimulation of C-fibers should have existed without being felt as pain? If this is possible, then the stimulation of C-fibers can itself exist without pain, since for it to exist without being felt as pain is for it to exist without there being any pain. Such a situation would be in flat out contradiction with the supposed necessary identity of pain and the corresponding physical state. ...The trouble is that the identity theorist does not hold that the physical state merely produces the mental state, rather he wishes the two to be identical. ...

The trouble is that the notion of an epistemic situation qualitatively identical to one in which the observer had a sensation S simply it one in which the observer had that sensation. The same point can be made in terms of the notion of what picks out the reference of a rigid designator. In the case of the identity of heat with molecular motion the important consideration was that although ‘heat’ is a rigid designator, the reference of that designator was determined by an accidental property of the referent, namely the property of producing in us the sensation S. It is thus possible that a phenomenon should have been rigidly designated in the same way as a phenomenon of heat, with its reference also picked out by means of the sensation S, without that phenomenon being heat and therefore without its being molecular motion. Pain, on the other hand, is not picked out by one of its accidental properties; rather it is picked out by the property of being pain itself, by its immediate phenomenological quality. Thus pain, unlike heat, is not only rigidly designated by ‘pain’ but the reference of the designator is determined by an essential property of the referent.” (150-153)

Standard explanations of the illusion of contingency rely on there being a qualitatively identical situation in which one of the key terms -- e.g., ‘Hesperus’, ‘gold’, ‘heat’, ‘water’ -- does not refer to what it actually does. But consider a qualitatively identical situation to one in which you are in pain. Could that situation be such that you are not in pain? No. And this is what makes the case of mind-body identities different in principle from other scientific identifications.