Identifying the Identity Theory

While I was perusing the new entries over at PhilPapers yesterday I came across Tom Polger‘s forthcoming paper in Philosophical Psychology Are Sensations Still Brain Processes? The paper is very interesting (disclaimer: I have a special interest in this stuff; see for instance The Identity Theory in 2-D) and I thought I would summarize its main points and then say something about where we disagree towards the end.

The first part of the paper Polger identifies eight theses that Smart defended in his celebrated paper. These are,

1. Sensation reports are genuine reports

2. Sensation reports do not refer to anything irreducibly psychical

3. Sensations are “nothing over and above” brain processes

4. Sensations are identical to brain processes

5. The identity theory is a metaphysical theory, not a semantic proposal or an
empirical hypothesis

6. Metaphysical theories of the nature of the mind do not make competing
empirical predictions; so they should be evaluated by their theoretical virtues,
e.g., simplicity and parsimony

7. For any thing or kind x, there are “logically” necessary conditions for being a thing of that kind.

8. Sensation expressions are topic-neutral.

The first claim is simply an endorsement of realism about phenomenal consciousness. Claims 2-4 spell out commitments to physicaism and the identity theory. Claims 5 and 6 spell out Smart’s distinctive views about the identity theory. Claim 7  basically asserts that we can know a priori what pains are essentially. Claim 8 amounts to the idea that concepts like ‘pain’ etc do not entail a commitment to any kind of ontology all by themselves.

Polger goes on to argue that of these every one but 7 should be accepted by contemporary identity theorists. Claim 6 should be accepted but not interpreted too narrowly. The identity theory should be accepted for broadly ‘inference to the best explanation’ reasons. Parsimony and simplicity play a role in that inference but there are other things that also play a role; As Polger says, “There are also what Jaegwon Kim has called explanatory and causal arguments for the identity theory”. The reason that 7 should be rejected according to Polger is the kind of resources that Kripkean arguments give to the identity theorist. In place of 7 above Polger suggests 7*

7*. A Posteriori. The identity of sensations and brain processes is a posteriori

7* is then an updated version of Smart’s claim that mind/brain identities are to be construed as ordinary scientific identities. We now have a post-Kripkean understanding of these kinds of identities and the contemporary identity theory should reflect that.

In the second part of the paper Polger goes on to formulate a master argument against the identity theory that he thinks subsumes all arguments against it and then responds to the various particular objections. The master argument goes as follows;

(P1) If the identity theory is true, then there is a necessary one-to-one relation between sensations and brain processes.15 (necessity of identity)

(P2) If VARIATION then there is not a necessary one-to-one relation between sensations and brain processes. (definition of VARIATION)


(P4) There is not a necessary one-to-one relation between sensations and brain processes. (P2, P3)

(C2) The identity theory is false. (P1, P4)

The particular objections that we find spell out varieties of variation claims: actual, nomological, metaphysical, logical. Polger identifies one major figure and style of objection this way. So, Putnam’s worries about octopi and Fred’s pain at 6:00 v.s. Fred’s pain at 6:15 count as actual variation while Fodor’s worries count as nomological, Kripke’s modal argument is metaphysical, and Chalmer’s zombie argument is logical. All of these arguments are united by trying to show that there is or can be variation.

Polger has a lot of interesting things to say in response to each of these objections.  Against actual variation he argues that even if we grant, as we might not, that we find the very same psychological properties across species on Earth (that is to say, even if an octopus can feel the very same kind of sensation that I do when I experience pain) there is still very little reason to think that psychological properties are multiply realizable in a way that is threatening to the identity theory. Sure there may be differences between species but that is no reason to rule out similarities a priori! Some people cite neural plasticity as a possible source of trouble. To this Polger replies, “evidence from plasticity is compatible with the neurobiological variations being variants within a more general kind that is also neurobiological.” Lacking any reason to believe in actual variation we also have no reason to believe in nomological variation, what about metaphysical variation? Here Polger endorses type-b physicalism and argues that Kripke’s argument is question begging. If the mind-brain identities are true then they are necessarily true. This leads Polger to the last kind of variation which he calls logical variation. It is here that we find Polger’s discussion of Chalmer’s zombie argument. His main complaint is that the argument rests on an assumption about the nature of reduction that the type-b physicalist will reject.

In the final section of the paper Polger introduces two further claims which he thinks should be endorsed by contemporary identity theorists.

9. Variability. Sensation processes are multiply constituted.

10. Strong Physicalism. Physicalism is necessarily true; all worlds are physicalist worlds.

In defense of accepting 9 Polger argues as follows,

accepting that there is…variability in the world is a far cry from accepting that it is the kind of variability that would be problematic for identity theories. Identity theories claim that sensations are brain processes, but they do not take any stand on the nature of brain processes. In particular, the identity theorist need not suppose that the world is organized into homogeneous columns of organization so that there is a one-to-one relation between sensations and microphysical processes. The identity theorist identifies sensations with brain processes, not with molecular or subatomic processes that occur inside brains.

I have always been sympathetic to this kind of argument and have seen some of my own work as generally supporting it. But what about 10? Why ought we accept that? The basic reason is to avoid the following reductio of the identity theory;

C1. Sensations are identical to brain processes in all possible worlds. (identity theory)

C2. Physicalism is contingent; there are some non-physicalist worlds containing non-physical sensations. (contingent physicalism)

C3. There are some worlds in which sensations are not identical to brain processes. (from C2)

C4. The identity theory is false.

Polger’s answer to this argument is to give up C2 thereby blocking C3. This may seem dramatic and I take 10, together with 7*, to entail that there are strong necessities in Dave Chalmers’ sense, “but”, says Polger, “so it goes. Just as there are necessary a posteriori truths, there are necessary a posteriori falsehoods.”

But it is just at this point that the difference between the kind of identity theory that Polger has and one that is in 2-D. Once we start thinking in 2 dimensional semantics we can see an equivocation in the redictio. C1 should be modified as C1*

C1* The secondary intension of ‘Sensations are brain processes’ is necessary; the primary intension of ‘sensations are brain processes’ is contingent (identity theory in 2-D)

Once we do that we do not have the worry about the reductio. Adopting C1* is tantamount to a compromise between 7 and 7*. In effect we agree that there is an a priori knowable description or reference fixer and an a posteriori identified physical state. Given that we know independently that identities like this are 2-necessary in Dave Chalmers’ sense we can conclude that those identities are necessary in spite of possible worlds where the a priori knowable description picks out a non-physical property.

3rd Birthday

Tomorrow marks the third anniversary of my starting Philosophy Sucks! I started my blogging career over at Brains and had my first post on April 12, 2007. I had several posts there before I was compelled to start my own blog and as people may know I continue to contribute to Brains and am very pleased to have seen it grow in recent times. I continue to post here as well and limit my posts at Brains to ones that directly relate to philosophy of mind and consciousness.

In these three years I have had over 100,000 hits, nearly 350 posts, and almost 2,000 comments…and next week I will be hosting my third Philosopher’s Carnival (I hosted the 58th and the 50th); not bad! I have had some rough experiences adapting to online discussion (there are some crazies out there as people well know) but all in all the discussion has been extremely helpful and challenging. I have had two papers and numerous presentations (two at the apa Pacific) develop out of discussions that started here. So thanks to everyone and I hope it continues in the future!

The year is still young but here are the most viewed posts so far (see also the best of all time).

10. HOT Qualia Realism
9. Am I a Type-Q Materialist?
8. Why I am not a Type-Z Materialist
7. Consciousness, Consciousness, and More Consciousness
6. More on Identity
5. The Singularity, Again
4. HOT Damn! It’s a HO Down-Showdown
3. Attention & Mental Paint
2. Part-Time Zombies
1. The Identity Theory in 2-D

More on Identity

In his 2004 paper How Can we Construct a Science of Consciousness? Dave Chalmers says,

Where there is systematic covariation between two classes of data, we can expect systematic principles to underlie and explain the covariation. In the case of consciousness, we can expect systematic bridging principles to underlie and explain the covariation between third-person data and first-person data. A theory of consciousness will ultimately be a theory of these principles…

In the third-person facts are all of the functional characterizations typically appealed to by the Lewisian. In the first-person data we find such things as that blue is more like purple than it is like orange and that pain has a certain phenomenal feel, etc. What kind of fundamental principles can we expect? Chalmers gives us a couple of examples in his Facing Up to the Problem of Consciousness. First he postulates “the principle of structural coherence” which he characterizes as;

This is a principle of coherence between the structure of consciousness and the structure of awareness. Recall that “awareness” was used earlier to refer to the various functional phenomena that are associated with consciousness. I am now using it to refer to a somewhat more specific process in the cognitive underpinnings of experience. In particular, the contents of awareness are to be understood as those information contents that are accessible to central systems, and brought to bear in a widespread way in the control of behavior.

In short, then, phenomenal consciousness correlates with what states we access or are aware of ourselves as being in. The second principle is that of “functional invariance” which he characterizes as follows;

This principle states that any two systems with the same fine-grained functional organization will have qualitatively identical experiences. If the causal patterns of neural organization were duplicated in silicon, for example, with a silicon chip for every neuron and the same patterns of interaction, then the same experiences would arise.

These principles are not fundamental but can be explained by some more basic principle and here Chamers offers the dual aspect of information principle;

This leads to a natural hypothesis: that information (or at least some information) has two basic aspects, a physical aspect and a phenomenal aspect. This has the status of a basic principle that might underlie and explain the emergence of experience from the physical. Experience arises by virtue of its status as one aspect of information, when the other aspect is found embodied in physical processing.

At this point one may wonder why we shouldn’t identify the phenomenal properties with the physical properties…Chalmers, in the 2004 paper, says this;

…What would this entail about the relationship between physical processes and consciousness? The existence of such principles is compatible with different philosophical views. One might regard the principles as laws connecting two fundamentally different domains (Descartes 1641/1996; Popper and Eccles 1977). One might regard them as laws connecting two aspects of the same thing (Lockwood 1989; Chalmers 1996). Or one might regard them as grounding an identification between properties of consciousness and physical properties (Smart 1959; Papineau 2002). Such principles could also be combined with different views of the causal relation between physical processes and consciousness (see Chalmers 2002).

So Dave does acknowledge that one might take these correlations to ground an identity claim. Interestingly he seems to take the way that the identity claim will be established as via thinking about parsimony. I suppose that Ned will argue that asserting the identity allows for explanations that we would not have otherwise. Both of these seem like fine reasons to posit the identity. What will the 2-D theorist say? it seems to me that the principle of structural coherence is exactly what gives us the functional foothold that we need to identify the biological basis of experience. What we need then is to determine whether the principle of structural coherence gives us any reason to think that some kind of higher-order theory is right.

UPDATE: It just occurred to me one might use the above kind of considerations to argue that Maria will be able to make the deductions from P to Q (and vice versa) a priori…if so then that would be an independent argument for the identity of the two sets of properties.

The Identity Theory in 2-D

I plan on writing a series of posts discussing various themes that came up in discussion at the online consciousness conference.

I have long been a type-type identity theorist. There was a time when I thought that I would write my dissertation defending a version of identity theory (in fact the very first talk I gave at a professional meeting was what I thought of as a ‘pre-prospectus’ available here: Saying “I Do” to Identity. I presented this as a poster at the ASSC in Antwerp and as a talk at the SPP in Barcelona (I called this my “European Identity Tour”))…When I approached Michael Devitt about the idea he said that people used to be interested in the identity theory but that people had moved on…it turns out that people are getting re-interested in the identity theory in the wake of work by people like Tom Polger, Chris Hill, and Ned Block. One thing that came out very clearly in the discussion is the difference between the identity theory that Block holds from the kind that I hold. The main difference concerns how we will eventually come to discover the mind-brain identities. Broadly speaking there are two different camps.

It is useful to remind ourselves of what the originators of the identity theory held. In “Is Consciousness a Brain Process?” U. T. Place says,

The answer seems to be that we treat the two sets of observations as observations of the same event in those cases where the technical scientific observations set in the context of the appropriate body of scientific theory provide an explanation of the observation of the man in the street. Thus we conclude that lightning is nothing more than a motion of electric charges, because we know that a motion of electric charges through the atmosphere, such as occurs when lightning is reported, gives rise to the visual stimulation which would lead an observer to report a flash of lightning (p. 58 in Chalmers 2002)

J.J.C. Smart in “sensations and Brain Processes” writes,

Why do I wish to resist [the suggestion that qualia are irreducibly psychial]? Mainly because fo Occam’s razor. It seems to me that science is increasingly giving us a viewpoint whereby organisms are able to be seen as psyico-chemical mecanisms: it seems that even the behavior of man himself will one day be explicable in mechanistic terms…That everything should be explicable in terms of physics (together of course with the descriptions of the ways in which the parts are put together –roughly, biology is to physics as radio-engineering is to electro-magnetism) except the occurrence of sensations seems to me to be frankly unbelievable. Such sensations would be “nomological danglers,” to use Feigl’s expression

We can see here an emphasis on the notions of explanation and parsimony. 16 years later David Lewis and David Armstrong establish the alternative camp. Lewis puts it most clearly when he writes,

Psychophysical identity theorists often say that the identifications they anticipate between mental and neural states are essentially like various uncontroversial theoretical identifications: the identification of water with H2O, of light with electromagnetic radiation, and so on. Such theoretical identifications are usually described as pieces of voluntary theorizing as follows. Theoretical advances make it possible to simplfy total science by positing brdge laws identifying some of the entities discussed in one theory with entities discussed in another theory. In the name of parsimony, we posit those bridge laws forthwith. Identifications are made, not found.

In ‘An Argument for teh Identity Theory,” I claimed that this was a bad picture of psychophysical identification, since a suitable physiological theory could imply psychophysical identites –not merely make it reasonable to posit them for the sake of parsimony. The implication was as follows:

Mental state M=the occupant of causal role R (definition of M)
Neural state N=the occupant of causal role R (by the physiological theory)
Therefore Mental state M=neural state N (by transitivity of =)

Nor is this peculiar to psychophysical identifications. He goes on,

…the usual account is, I claim, wrong; theoretical identifications in general are implied by the theories that make them possible –not posited independantly. This follows from a general hypothesis about the meaning of theoretical terms: that they are definable functionally, by reference to causal roles (Psychophysical and Theoretical Identifications)

In a recent paper on functional reduction Ned Block targets the Lewisian view in favor of the Place/Smart view. Here is what he says,

If we want to know why water = H2O, freezing = molecular lattice formation, heat = molecular kinetic energy, temperature = mean molecular kinetic energy, etc, we have to start with the fact that water, temperature, heat, freezing and other magnitudes form a family of causally inter-related “macro” properties. This macro family is mirrored by a family of “micro” properties: H2O, mean molecular kinetic energy, molecular kinetic energy and formation of a lattice of H2O molecules. (Of course a given level can be micro with respect to one level, macro with respect to another.) The key fact is that the causal and explanatory relations among the macro properties can be explained if we suppose that the following relations hold between the families: that water = H2O, temperature = mean molecular kinetic energy, heat = molecular kinetic energy and freezing = lattice formation. For example, why does decreasing the temperature of water cause it to freeze? Why does ice float on water? Here is a sketch of the explanation: The oxygen atom in the H2O molecule has two pairs of unmated electrons, which attract the hydrogen atoms on other H2O molecules. When the kinetic energy of the molecules decreases, (i.e. the temperature decreases) each oxygen atom tends to attract two hydrogen atoms on the ends of two other H2O molecules. When this process is complete, the result is a lattice in which each oxygen atom is attached to four hydrogen atoms.Ice is this lattice and freezing is the formation of such a lattice, which is why decreasing temperature causes water to freeze. Because of the geometry of the bonds, the lattice has an open, less dense structure than amorphously structured H2O (viz., liquid water)–which is why ice (frozen water) floats on liquid water.

Suppose we reject the assumption that temperature is identical to mean molecular kinetic energy in favor of the assumption that temperature is merely correlated with mean molecular kinetic energy? And suppose we reject the claim that freezing is lattice-formation in favor of a correlation thesis. And likewise for water/H2O. Then we would have an explanation for how something that is correlated with decreasing temperature causes something that is correlated with frozen water to float on something correlated with liquid water, which is not all that we want. The reason to think that the identities are true is that assuming them gives us explanations that we would not otherwise have and does not deprive us of explanations that we already have or raise explanatory puzzles that would not otherwise arise. The idea is not that our reason for thinking these identities are true is that it would be convenient if they were true. Rather, it is that assuming that they are true yields the most explanatory overall picture. In other words, the epistemology of theoretical identity is just a special case of inference to the best explanation. (See Block, 1978a; Block, 2002; Block & Stalnaker,1999).

Block goes on to argue that the Lewis style view is incompatible with the metaphysics of physicalism. Block distinguishes between ontology and metaphysics. Ontological physicalism is just the claim that in our ontological commitment to the existence of qualia we commit ourselves only to physical entities (ontological dualists deny this). Metaphysical physicalism is the claim that qualitative properties are essentially or metaphysically physical. That is to say that all qualitative properties will share the same physical properties in so far as they are physical. the Lewis style physicalism is ontologically but nit metaphysically physicalist. This is because as it happens all of the realizers of mental states are physical but metaphysically pain is a functional state for Lewis and only contingently a physical state. Metaphysical physicalism –real physicalism in Block’s view– says that it is not contingent but necessary that pain is a physical state.

But if we adopt the 2-D framework and put the Lewisian claims in terms of it this is no longer a problem. On this kind of view the functional definition gives us the primary intension of ‘pain’ and the physical state gives us the secondary intension. This allows us to treat ‘pain’ just as we do ‘water’. ‘Water is H2O’ has a contingent primary intension and a necessary secondary intension. So we can update Lewis view that ‘pain’ isn’t a rigid designator as the claim that the primary intension of pain is contingent (just like ‘water’). ‘Pain’ is still a rigid designator in the ordinary sense that its secondary intension is necessary. In all worlds considered as counter-factual pain is a brain state. However we accommodate the conceivability of Martians and disembodied minds by noting that in some worlds considered as actual pain is not a brain state (just as in some worlds considered as actual water is not H2O). This does not threaten the identity; it is the usual way that theoretical identities work. Notice also that this 2-D identity theory is a metaphysical physicalism in Block’s sense and not merely an ontological physicalism.

Of course the real resistance to the 2-D Lewisian identity theory is that qualitative states are not supposed to be functionally definable. In fact Block and Chalmers often talk as though qualitative properties are definable as ‘the not functionally definable properties of experience’ (more on that later). If that is your view then you cannot do the Lewsian deduction of the identity. What are we to make of this? I will come back to this in the next post.