References for semantic argument against identity theory?

There is a standard story about the emergence of the identity theory. Namely, before (say) 1956, because the concept of an experience was not synonymous with the concept of a brain state, we knew a priori that experiences could not be brain states. Let’s call this the semantic argument against identity theory.

It was eventually pointed out that you could have different concepts that refer to the same object (e.g., lightning and electrostatic discharge), at which point identity theory became viable again.

There are two references I hae that suggest the semantic argument was a real force, in analytic philosophy, that blocked the identity theory. First, UT Place, in ‘Is Consciousness a Brain Process?’, wrote:

The all but universally accepted view that an assertion of identity between consciousness and brain processes can be ruled out on logical grounds alone derives, I suspect, from a failure to distinguish between what we may call the ‘is’ of definition and the ‘is’ of composition.

Basically he is saying that even if it is not definitional of ‘consciousness’ that it is a brain process, that is not sufficient to show that it isn’t a brain process. As he goes on to say, he considers such identity claims to be “a reasonable scientific hypothesis, in the way that the statement, “Lightning is a motion of electric charges, ” is a reasonable scientific hypothesis.”

Putnam makes the point more explicitly in his essay ‘Mind and Body’ when recounting the history of the identity theory:

[An old-school philosopher] frequently regards talk of properties as interchangeable with talk of concepts. For such a philosopher, properties cannot be the same unless it is a conceptual truth that they are the same; in particular, the property of having a sensation with a certain qualitative character cannot be the same as the property of being in a certain brain-state, since the corresponding predicates are not synonymous

Was such a semantic argument, specifically against identity theory, ever clearly made in writing? I’m looking for references.

If both Place and Putnam hadn’t said what they did, I would find it unbelievable that people would have bought the force of the semantic argument. Long before Ryle, Frege’s claims about sense and reference already established that
co-referring terms don’t have to have the same meaning (e.g., morning
star evening star). Were Place and Putnam referring to actual people, or to the zeitgeist in the heyday of analytic philosophy?

Perhaps some might say the semantic argument was implicit in Ryle, who thought that concepts about mental states could be analyzed, without remainder, into claims about behaviors and behavioral dispositions. However, this seems a different argument: Ryle claimed that a behavioral analysis literally left nothing out, so there is nothing left for a neuronal story to do once the behavioral story was told. That, I think, was the “logical” error Ryle spoke of, the “category mistake” of thinking that there was more to the mind than behavior. Asking where the ‘mind’ is once behavior has been delineated is like asking where the ‘university’ is once all of the buildings, people, and other details of the university have been delineated. This seems to be different than the semantic argument as I have described it.

So, am I wrong about Ryle: did he actually use the semantic argument against identity theory? And regardless of Ryle, are there clear instances of the argument anywhere in the literature?

6 Comments

  1. Joshua Stern

    There is a standard story about the emergence of the identity theory. Namely, before (say) 1956, because the concept of an experience was not synonymous with the concept of a brain state, we knew a priori that experiences could not be brain states. Let’s call this the semantic argument against identity theory.

    It was eventually pointed out that you could have different concepts that refer to the same object (e.g., lightning and electrostatic discharge), at which point identity theory became viable again.

    But – it’s been known for millenia that you could token *words* for a thing, even an experience. What seems to have been lacking pre-1956 (or whatever) is the idea that brain states could be tokens of some kind, any kind, linguistic or otherwise. It has just alway seemed to me an odd oversight, a blindness.

    Different *concepts* coreferring, how about different word tokens, of the same or different words? I’m afraid the concept concept is weak, no matter how traditional.

    The a priori assertion has always been that “well, you can’t *seriously* propose that an experience in the human mind is like a written word!” Why not? Because, um, … well, in any case I’m afraid the baby went out with the bathwater, and the idea of tokening went out with the putative, tacit faults of a word as a memory.

    Now that we have words (bits) as memory (computers) all around us, this assertion is hard to credit. I’m afraid the intution (!) is still common, that “my quale/experience cannot be anything like a mere *word*”, but that is just not an argument.

    I’m afraid that’s about all I can say offhand about the origin of arguments against identity. The Ryleian arguments just leave my head spinning and jaw on the ground as I try to analyze the implications of what is being rejected (no physicalism) and of what is being advocated (neo-crypto-essentialisms).

  2. Eric Thomson

    Joshua: yes, as I mentioned I am a little surprised by the claims about the significance of the semantic argument against identity theory because the sense-reference distinction has been around at least since Frege (many years before Ryle and the other behaviorists).

    However, because I trust Place and Putnam,  I tend to not think the semantic argument is a straw man. The question is whether it was ever published in a prominent place, or perhaps did analytic philosophers consider it too obvious of a semantic point, and therefore unworthy of publication.

    Those analytic guys were probably very intimidating when they said you were mangling the “logic” of X.

    Perhaps Putnam conflated the semantic argument, as I’ve laid it out, with Ryle’s analytical behaviorism. It seems Place and Putnam (rightly) want to reject the analysis of consciousness-talk in terms of behavior-talk, but that is different from saying that Ryle rejects neural identity theory based on the semantic argument.

    In other words, Ryle would agree that consciousness-talk and brain-talk have different meanings, but because he is familiar with Frege he would never say that this is enough to show that consciousness is not a brain process.  Rather, his point is that he left nothing out when he did his behavioral analysis of conscious-state terms. That is, Ryle would say his analytical behaviorism leaves out brains because he disagrees with the hypothesis that conscious states are brain states. He would say that conscious states are behavioral/behavioral disposition states, and that’s it. There is no residual mysterious question that needs to be answered, except perhaps how brains produce behavior and embody behavioral dispositions (e.g., how to brains produce pains, where the latter are defined in terms of behavior and behavioral dispositions). That is, Ryle could be said to be offering an alternative theory of consciousness, based on a different understanding altogether of what it means to be conscious, but it isn’t obvious that this is an instance of the semantic argument simpliciter.

    Indeed, this analysis of the situation seems right when we see what Place says about Ryle in ‘Is consciousness a brain process?’ (italics added):

    Modern physicalism, however, unlike the materialism of the seventeenth and eighteenth centuries, is behavioristic. Consciousness on this view is either a special type of behavior, … or a disposition to behave in a certain way, an itch for example being a temporary propensity to scratch. In the case of cognitive concepts like ‘knowing’, ‘believing’, ‘understanding’, and ‘remembering’, and volitional concepts like ‘wanting’ and ‘intending’, there can be little doubt, I think, that an analysis in terms of dispositions to behave (Wittgenstein, 1953; Ryle, 1949) is fundamentally sound. On the other hand, there would seem to be an intractable residue of concepts clustering around the notions of consciousness, experience, sensation, and mental imagery, where some sort of inner process story is unavoidable (Place, 1954). It is possible, of course, that a satisfactory behavioristic account of this conceptual residuum will ultimately be found. For our present purposes, however, I shall assume that this cannot be done and that statements about pains and twinges, about how things look, sound, and feel, about things dreamed of or pictured in the mind’s eye are statements referring to events and processes that are in some sense private or internal to the individual of whom they are predicated.

    What seems to be a deeper place of disagreement is the verifiability theory of meaning of the positivists (and perhaps present implicitly in Ryle). Because everyday folk can talk about conscious states meaningfully, there must be some public criterion of verification of such states (he is wincing and holding his arm), so the meaning of statements about consciousness is literally the facts that would verify said statements.

    This seems in some ways to be the deeper point of disagreement that Place and Putnam don’t mention.

    Boy positivists believed some silly things.

  3. Hi Eric,

    I think there’s something close to what you’re looking for in C.I. Lewis’s “Some Logical Considerations Concerning the Mental”, J Phil vol. 38, 1941. e.g.:

    “If this is correct, then it may serve to explain the pertinence of the argument that mental facts cannot be identified with brain facts or facts of physical behavior because we directly inspect an are fully acquainted with the mental factuality but may be ignorant of the brain state or the behavior; or if we should be also aware of these latter, may still be ignorant of any connection between them and what is mental. Admitting this phenomenal meaning of language used to denote the mental, such argument is entirely sound, and proves its point.”

    and

    [on scientific identity claims] “There is also the consideration that whereas in natural science, which concerns itself exclusively with the existent, such definition by description [e.g. middle C = a vibration of 256 Hz] has its pragmatic justication, such justification is lacking in philosophy, whose concern is not that of establishing synthetic a posteriori truths. From this point of view, the behavioristic or brain-state theory of mind substitutes an hypothesis which only the future development of natural science can corroborate or disprove for our more apporpriate business of the analysis of meanings.”

    Sometimes, thought, he starts to sound a bit like Kripke…

  4. Eric Thomson

    Dan, that is a great example: thanks!

    He says (p 226) that ‘this point has been made often enough and will not need to be elaborated here’. I wish he had given references.

    Perhaps Bertrand Russell?

Comments are closed.