Argument from Reason

There are, however, a number of serious objections to this proposal [the Computational/Representational Theory of Thought, or CRTT]. Consider first the implications of taking mental states to be states of a computer program whose causal efficacy derives entirely from their implementation in electrochemical processes in the brain.When you type “2,” “+,” “2,” and “=” on the keyboard of an electronic calculator, various electrical signals are sent through the device which ultimately cause the symbol “4”to appear on the display screen. But that that symbol signifies to us the number 4, and that the other symbols signify the number 2, the function of addition, and the relation of being equal, plays no role whatsoever in the causal process.If we decided to change the meanings of these symbols – for instance, by using the sequence “2 + 2 =” to mean “Please display the message that it is raining”and the symbol “4”to mean “it is raining”– this would have no effect on how the device operates.Nor would it have any effect if we all forgot the meaning of the symbols, and came to regard calculators merely as toys that displayed different shapes whenever one pressed their keys.The meanings of the symbols are,in short,completely irrelevant to their causal efficacy, for they would have the same causal properties whatever meanings they had,or even if they had no meanings.

If this is true of the symbols processed by a calculator it would be true also of the symbols “processed”by the brain – it would be true, that is to say,of the contents of our thoughts as they are characterized by the CRTT.If your thought that “Socrates is a man”is identical with a neural process instantiating a sentence in Mentalese which has the meaning or content that Socrates is a man, then that meaning per se plays absolutely no role in causing whatever events the neural process,and thus the thought,causes.The causal properties of the neural process/thought would be just as they are even if it had instead the meaning that “it is raining,” or even if it had no meaning at all.And that entails that the fact that your thought has the content that “Socrates is a man”plays absolutely no role whatsoever in causing you, for example, to say or write the sentence “Socrates is a man.”You would have written or uttered the same sentence even if your thought had been about the rain or even if it had had no meaning at all.The electrochemical properties of the neural process implementing the thought are all that matter to its causal efficacy, just as the electronic properties of the symbols in a calculator are all that matter to their causal efficacy.

What this seems to mean is that distinctively mental properties turn out in the materialistic CRTT to be no less epiphenomenal than they do with property dualism. Nor is the CRTT the only materialist theory to have this consequence; indeed, any theory that takes mental states to have whatever causal efficacy they have only because of their identity with or supervenience upon physical states seems destined to have the same result: the physical properties of such states end up doing all the causal work, with the mental properties being an irrelevant, epiphenomenal extra.Epiphenomenalism would thus appear to threaten materialist theories no less than it does dualist ones – in which case the claim of materialist theories to be better able than dualist ones to account for the causal relations between mind and body seems to dissolve.

The problem, however, seems especially poignant for the CRTT,given its claim to provide a materialistic explanation of our capacity for rational thought. If the content or meaning of thoughts has, in the CRTT, no causal influence on behavior, neither does it have any causal influence on other thoughts.That your thoughts have the content that Socrates is a man and that all men are mortal can have no influence whatever on producing the thought that Socrates is mortal, for that last thought would have been caused by the others even if those others had instead had the content that Fido is a dog and all fish have fins, or even if they had no content or meaning at all. The electrochemical properties of the neural processes with which the thoughts are associated are entirely sufficient to bring about whatever effects they do bring about.The meaning or content of the thoughts is irrelevant.

That this result is as counter-intuitive as it is is bad enough, but the problem goes deeper. It is only in virtue of the meaning or content of thoughts that they can serve as a rational justification for other thoughts:your thoughts that Socrates is a man and that all men are mortal are a rational justification for believing that Socrates is mortal only because they have the meaning they do, and they would not serve as a rational justification for the latter thought if they meant instead that Fido is a dog,etc.Yet if the meaning or content of a particular thought plays absolutely no role in bringing about any other thought, it would seem to follow that it can provide no rational justification for any other thought.You’d have exactly the same beliefs you have now whatever the content had been of the further beliefs you appeal to in justifying them.In that case, however, your beliefs would seem to have no rational justification at all. But surely this cannot be right – surely you do have a rational justification for at least many of your beliefs. Yet the CRTT,it seems,cannot account for this – ironically enough,given that its very rationale was to account for our capacity for rational thought.Even worse,advocates of the CRTT obviously think they have a rational justification for their own belief in the CRTT; but if the theory is correct, it would seem that they can’t! The theory appears to undermine itself.

The CRTT defender might appeal to evolution as a guarantee of the reliability of our thought processes:wouldn’t natural selection ensure that our brains are wired in such a way that the thoughts we generate are, for the most part, true? Wouldn’t we have died out long ago if things were otherwise? One quick reply to this would be to suggest that it is question-begging: for it assumes that we can be rationally justified,in the CRTT,in believing the Darwinian evolutionary story (or believing anything else) in the first place, which is precisely what is at issue.Another reply would be to note that what natural selection tends to maximize is the capacity of an organism to survive and reproduce, and there is no reason to assume that having a true system of beliefs really is what is most conducive to survival: maybe our environment is such that we have been able to survive and reproduce as well as we have only because we have developed a mostly false system of beliefs, a kind of elaborate fantasy world that shields us from certain truths, the knowledge of which would tend toward our destruction (perhaps because they would be too horrifying for us to bear).But there appears to be an even deeper problem.The general truth or falsity of a system of beliefs can only be affected by natural selection if that system of beliefs has, by virtue of its truth or falsity, some causal influence on behavior – that is, if the truth or falsity per se causes behavior which is either adaptive or maladaptive, and which will tend therefore to get either selected for or selected out. But a belief ’s being either true or false is bound up with its having the particular content it has, and as has been suggested, there seems to be no way, in the CRTT (or perhaps in any materialistic account of thought), for the content or meaning of a thought to have any causal influence on behavior.The purely neurophysiological properties which,according the CRTT,instantiate the thought are the only ones that can have any causal relevance. So there is no way for the truth or falsity of a belief to have any effect on behavior, and thus natural selection cannot affect in any way the general truth or falsity of a system of belief. But in that case, if the CRTT (or any purely materialistic account of thought) is true, evolution cannot account for the reliability of our thought processes.