Chapter 8 From "Identity and Necessity" Saul Kripke
Let me turn to the case of heat and the motion of molecules. Here surely is a case that is contingent identity! Recent philosophy has emphasized this again and again. So, if it is a case of contingent identity, then let us imagine under what circumstances it would be false.
Now, concerning this statement I hold that the circumstances philosophers apparently have in mind as circumstances under which it would have been false are not in fact such circumstances. First, of course, it is argued that "Heat is the motion of molecules" is an a posteriori judgment; scientific investigation might have turned out otherwise. As I said before, this shows nothing against the view that it is necessary—at least if I am right. But here, surely, people had very specific circumstances in mind under which, so they thought, the judgment that heat is the motion of molecules would have been false. What were these circumstances? One can distill them out of the fact that we found out empirically that heat is the motion of molecules. How was this? What did we find out first when we found out that heat is the motion of molecules? There is a certain external phenomenon which we can sense by the sense of touch, and it produces a sensation which we call "the sensation of heat." We then discover that the external phenomenon which produces this sensation, which we sense, by means of our sense of touch, is in fact that of molecular agitation in the thing that we touch, a very high degree of molecular agitation. So, it might be thought, to imagine a situation in which heat would not have been the motion of molecules, we need only imagine a situation in which we would have had the very same sensation and it would have been produced by something other than the motion of molecules. Similarly, if we wanted to imagine a situation in which light was not a stream of photons, we could imagine a situation in which we were sensitive to something else in exactly the same way, producing what we call visual experiences, though not through a stream of photons. To make the case stronger, or to look at another side of the coin, we could also consider a situation in which we are concerned with the motion of molecules but in which such motion does not give us the sensation of heat. And it might also have happened that we, or, at least, the creatures inhabiting this planet, might have been so constituted that, let us say, an increase in the motion of molecules did not give us this sensation but that, on the contrary, a slowing down of the molecules did give us the very same sensation. This would be a situation, so it might be thought, in which heat would not be the motion of molecules, or, more precisely, in which temperature would not be mean molecular kinetic energy.But I think it would not be so. Let us think about the situation again. First, let us think about it in the actual world. Imagine right now the world invaded by a number of Martians, who do indeed get the very sensation that we call "the sensation of heat" when they feel some ice which has slow molecular motion, and who do not get a sensation of heat—in fact, maybe just the reverse—when they put their hand near a fire which causes a lot of molecular agitation. Would we say, "Ah, this casts some doubt on heat being the motion of molecules, because there are these other people who don't get the same sensation"? Obviously not and no one would think so. We would say instead that the Martians somehow feel the very sensation we get when we feel heat when they feel cold and that they do not get a sensation of heat when they feel heat. But now let us think of a Counterfactual situation.1 Suppose the earth had from the very beginning been inhabited by such creatures. First, imagine it inhabited by no creatures at all: then there is no one to feel any sensations of heat. But we would not say that under such circumstances it would necessarily be the case that heat did not exist; we would say that heat might have existed, for example, if there were fires that heated up the air.
Let us suppose the laws of physics were not very different: Fires do heat up the air. Then there would have been heat even though there were no creatures around to feel it. Now let us suppose evolution takes place, and life is created, and there are some creatures around. But they are not like us, they are more like the Martians. Now would we say that heat has suddenly turned to cold, because of the way the creatures of this planet sense it? No, I think we should describe this situation as a situation in which, though the creatures on this planet got our sensation of heat, they did not get it when they were exposed to heat. They got it when they were exposed to cold. And that is something we can surely well imagine. We can imagine it just as we can imagine our planet being invaded by creatures of this sort. Think of it in two steps. First there is a stage where there are no creatures at all, and one can certainly imagine the planet still having both heat and cold, though no one is around to sense it. Then the planet comes through an evolutionary process to be peopled with beings of different neural structure from ourselves. Then these creatures could be such that they were insensitive to heat; they did not feel it in the way we do; but on the other hand, they felt cold in much the same way that we feel heat. But still, heat would be heat, and cold would be cold. And particularly, then, this goes in no way against saying that in this Counterfactual situation heat would still be the molecular motion, be that which is produced by fires, and so on, just as it would have been if there had been no creatures on the planet at all. Similarly, we could imagine that the planet was inhabited by creatures who got visual sensations when there were sound waves in the air. We should not therefore say, "Under such circumstances, sound would have been light." Instead we should say, "The planet was inhabited by creatures who were in some sense visually sensitive to sound, and maybe even visually sensitive to light." If this is correct, it can still be and will still be a necessary truth that heat is the motion of molecules and that light is a stream of photons.
To state the view succinctly: we use both the terms 'heat' and 'the motion of molecules' as rigid designators for a certain external phenomenon. Since heat is in fact the motion of molecules, and the designators are rigid, by the argument I have given here, it is going to be necessary that heat is the motion of molecules. What gives us the illusion of contingency is the fact we have identified the heat by the contingent fact that there happen to be creatures on this planet—(namely, ourselves) who are sensitive to it in a certain way, that is, who are sensitive to the motion of molecules or to heat—these are one and the same thing. And this is contingent. So we use the description, 'that which causes such and such sensations, or that which we sense in such and such a way', to identify heat. But in using this fact we use a contingent property of heat, just as we use the contingent property of Cicero as having written such and such works to identify him. We then use the terms ÚåàÊ in the one case and 'Cicero' in the other rigidly to designate the objects for which they stand. And of course the term 'the motion of molecules' is rigid; it always stands for the motion of molecules, never for any other phenomenon. So, as Bishop Butler said, "everything is what it is and not
From 'Identity and Necessity" 43 another thing." Therefore, "Heat is the motion of molecules" will be necessary, not contingent, and one only has the illusion of contingency in the way one could have the illusion of contingency in thinking that this table might have been made of ice. We might think one could imagine it, but if we try, we can see on reflection that what we are really imagining is just there being another Iectem in this very position here which was in fact made of ice. The fact that we may identify this Iectem by being the object we see and touch in such and such a position is something else.
Now how does this relate to the problem of mind and body? It is usually held that this is a contingent identity statement just like "Heat is the motion of molecules." That cannot be.
It cannot be a contingent identity statement just like "Heat is the motion of molecules" because, if I am right, "Heat is the motion of molecules" is not a contingent identity statement. Let us look at this statement. For example, "My being in pain at such and such a time is my being in such and such a brain state at such and such a time," or, 'Tain in general is such and such a neural (brain) state."This is held to be contingent on the following grounds. First, we can imagine the brain state existing though there is no pain at all. It is only a scientific fact that whenever we are in a certain brain state we have a pain. Second, one might imagine a creature being in pain, but not being in any specified brain state at all, maybe not having a brain at all. People even think, at least prima facie, though they may be wrong, that they can imagine totally disembodied creatures, at any rate certainly not creatures with bodies anything like our own. So it seems that we can imagine deflnite circumstances under which this relationship would have been false. Now, if these circumstances are circumstances, notice that we cannot deal with them simply by saying that this is just an illusion, something we can apparently imagine, but in fact cannot in the way we thought erroneously that we could imagine a situation in which heat was not the motion of molecules. Because although we can say that we pick out heat contingently by the contingent property that it affects us in such and such a way, we cannot similarly say that we pick out pain contingently by the fact that it affects us in such and such a way. On such a picture there would be the brain state, and we pick it out by the contingent fact that it affects us as pain. Now that might be true of the brain state, but it cannot be true of the pain. The experience itself has to be this experience, and I cannot say that it is a contingent property of the pain I now have that it is a pain.2 In fact, it would seem that both the terms, 'my pain' and 'my being in such and such a brain state' are, first of all, both rigid designators.
That is, whenever anything is such and such a pain, it is essentially that very object, namely, such and such a pain, and wherever anything is such and such a brain state, it is essentially that very object, namely, such and such a brain state. So both of these are rigid designators. One cannot say this pain might have been something else, some other state. These are both rigid designators.Second, the way we would think of picking them out—namely, the pain by its being an experience of a certain sort, and the brain state by its being the state of a certain material object, being of such and such molecular configuration—both of these pick out their objects essentially and not accidentally, that is, they pick them out by essential properties. Whenever the molecules are in this configuration, we do have such and such a brain state. Whenever you feel this, you do have a pain. So it seems that the identity theorist is in some trouble, for, since we have two rigid designators, the identity statement in question is necessary. Because they pick out their objects essentially, we cannot say the case where you seem to imagine the identity statement false is really an illusion like the illusion one gets in the case of heat and molecular motion, because that illusion depended on the fact that we pick out heat by a certain contingent property. So there is very little room to maneuver; perhaps none.3 The identity theorist, who holds
that pain is the brain state, also has to hold that it necessarily is the brain state. He therefore cannot concede, but has to deny, that there would have been situations under which one would have had pain but not the corresponding brain state. Now usually in arguments on the identity theory, this is very far from being denied. In fact, it is conceded from the outset by the materialist as well as by his opponent. He says, “Of course, it could have been the case that we had pains without the brain states. It is a contingent identity." But that cannot be. He has to hold that we are under some illusion in thinking that we can imagine that there could have been pains without brain states. And the only model I can think of for what the illusion might be, or at least the model given by the analogy the materialists themselves suggest, namely, heat and molecular motion, simply does not work in this case. So the materialist is up against a very stiff challenge. He has to show that these things we think we can see to be possible are in fact not possible. He has to show that these things which we can imagine are not in fact things we can imagine. And that requires some very different philosophical argument from the sort which has been given in the case of heat and molecular motion. And it would have to be a deeper and subtler argument than I can fathom and subtler than has ever appeared in any materialist literature that I have read. So the conclusion of this investigation would be that the analytical tools we are using go against the identity thesis and so go against the general thesis that mental states are just physical states.4
The next topic would be my own solution to the mind-body problem, but that I do not have.
Notes
1. Isn't the situation I just described also Counterfactual? At least it may well be, if such Martians never in fact invade. Strictly speaking, the distinction I wish to draw compares how we would speak in a (possibly Counterfactual) situation, if it obtained, and how we do speak of a Counterfactual situation, knowing that it does not obtain—i.e., the distinction between the language we would have used in a situation and the language we do use to describe it. (Consider the description: “Suppose we all spoke German." This description is in English.) The former case can be made vivid by imagining the Counterfactual situation to be actual.
2. The most popular identity theories advocated today explicitly fail to satisfy this simple requirement. For these theories usually hold that a mental state is a brain state, and that what makes the brain state into a mental state is its 'causal role', the fact that it tends to produce certain behavior (as intentions produce actions, or pain, pain behavior) and to be produced by certain stimuli (e.g., pain, by pinpricks). If the relations between the brain state and its causes and effects are regarded as contingent, then being such-andsuch-a-mental state is a contingent property of the brain state. Let X be a pain. The causal-role identity theorist holds (1) that X is a brain state, (2) that the fact that X is a pain is to be analyzed (roughly) as the fact that X is produced by certain stimuli and produces certain behavior. The fact mentioned in (2) is, of course, regarded as contingent: the brain state X might well exist and not tend to produce the appropriate behavior in the absence of other conditions. Thus (I) and (2) assert that a certain pain X might have existed, yet not have been a pain. This seems to me self-evidently absurd. Imagine any pain: is it possible that it itself could have existed, yet not have been a pain?
If X = Y, then X and Y share all properties, including modal properties. If X is a pain and Y the corresponding brain state, then being a pain is an essential property of X and being a brain state is an essential property of Y. If the correspondence relation is, in bet, identity, then it must be necessary of Y that it corresponds to a pain, and necessary of X that it correspond to a brain state, indeed to this particular brain state, Y. Both assertions seem false; it seems clearly possible that X should have existed without the corresponding brain state; or that the brain state should have existed without being felt as pain. Identity theorists cannot, contrary to their almost universal present practice, accept these intuitions; they must deny them, and explain them away. This is none too easy a thing to do.
3. A brief restatement of the argument may be helpful here. If "pain" and "C-fiber stimulation" are rigid designators of phenomena, one who identifies them must regard the identity as necessary. How can this necessity be reconciled with the apparent fact that C-fiber Stimuhtion might have turned out not to be correlated with pain at all? We might try to reply by analogy to the case of heat and molecular motion; the latter identity, too, is necessary, yet someone may believe that, before scientific
From 'Identity and Necessity" 45
investigation showed otherwise, molecular motion might have turned out not to be heat. The reply is, of course, that what really is possible is that people (or some rational sentient beings) could have been in the same epistemic situation as we actually are, and identify a phenomenon in the same way we identify heat, namely, by feeling it by the sensation we call "the sensation of heat," without the phenomenon being molecular motion. Further, the beings might not have been sensitive to molecular motion (i.e., to heat) by any neural mechanism whatsoever. It is impossible to explain the apparent possibility of C-fiber Stimuhtions not having been pain in the same way. Here, too, we would have to suppose that we could have been in the same epistemological situation, and identify something in the same way we identify pain, without its corresponding to C-fiber Stimuhtion. But the way we identify pain is by feeling it, and if a C-fiber Stimuhtion could have occurred without our feeling any pain, then the C-fiber Stimuhtion would have occurred without there being any pain, contrary to the necessity of the identity. The trouble is that although ⅛eat' is a rigid designator, heat is picked out by the contingent property of its being felt in a certain way; pain, on the other hand, is picked out by an essential (indeed necessary and sufficient) property. For a sensation to be felt as pain is for it to be pain.
4. All arguments against the identity theory which rely on the necessity of identity, or on the notion of essential property, are, of course, inspired by Descartes' argument for his dualism. The earlier arguments which superficially were rebutted by the analogies of heat and molecular motion, and the bifocals inventor who was also Postmaster General, had such an inspiration; and so does my argument here. R. Albritton and M. Slote have informed me that they independently have attempted to give essen- Halist arguments against the identity theory, and probably others have done so as well.
The simplest Cartesian argument can perhaps be restated as follows: Let 'A' be a name (rigid designator) of Descartes' body. Then Descartes argues that since he could exist even if A did not, O (Descartes ≠ A), hence Descartes ≠ A Those who have accused him of a modal fallacy have forgotten that 'A' is rigid. His argument is valid, and his conclusion is correct, provided its (perhaps dubitable) premise is accepted. On the other hand, provided that Descartes is regarded as having ceased to exist upon his death. "Descartes ≠A" can be established without the use of a modal argument; for if so, no doubt A survived Descartes when A was a corpse. Thus A had a property (existing at a certain time) which Descartes did not. The same argument can establish that a statue is not the hunk of stone, or the congery of molecules, of which it is composed. Mere non-identity, then, may be a weak conclusion. (See D. Wiggins, Philosophical Reoieiot VoL 77 (1968), pp. 90 ff.) The Cartesian modal argument, however, surely can be deployed to maintain relevant stronger conclusions as well