Searle identifies seven "foundations of modern materialism" and criticizes each. Indented paragraphs are my own comments.
1. Consciousness is unimportant.
Much work in cognitive science, and until recently in the philosophy of mind, as assumed or argued that consciousness is a relatively unimportant phenomenon. It has seemed possible to understand cognition in information-processing terms, and to understand information processing in turn as a kind of computation. Since computers aren't conscious, it has seemed that much of what is most important about human cognition doesn't require consciousness.
Searle argues that, on the contrary, consciousness is the defining characteristic of the mental. To be a mental state at all, a state must either be conscious or at least have the potential to be conscious. There cannot be a mental state that is "in principle inaccessible to consciousness" (p. 19).
This view is not defended in this chapter, however; the defense is in chapter 7.
2. Science is objective.
Searle suggests that we need to distinguish between two different senses of objectivity. (1) epistemological sense: to be objective is to remove biases or prejudices from the search for truth. (2) ontological sense: to be objective is to be independent from any point of view. This is closely related to the idea that an objectively real phenomenon should be accessible to any observer. If I can see a blue spot over there and you can't, that's good evidence that the blue spot is not part of the world. Much of science seems to be about constructing a conception of the world that is observer-independent in this sense. (Even relativity theory, which holds that such properties as speed and duration are relative to a frame of reference, is precisely an attempt to explain the features of the world that are observer-independent, and then explain how observer-dependent properties are the result of observer-independent features together with the situation of the observer.)
Searle's view is that, while epistemological objectivity is a necessary part of science, it is a mistake to think that everything that is scientifically respectable is objective in sense (2), the ontological sense. Conscious experience is irreducible subjective, accessible from one and only one point of view.
An interesting contrast here is with Daniel Dennett, who in Consciousness Explained constructs his method of "heterophenomenology" precisely because he insists that for consciousness to be scientifically investigated, we must deal only with observer-independent evidence.
3. Methodology must be objective.
Searle thinks that insisting on an objective methodology leads to the idea that cognitive science can only study behavior and its causal basis. Thus an objective methodology leads necessarily to something similar to behaviorism.
By contrast, Searle's view is that the study of consciousness is necessarily the study of an essentially first-person phenomenon.
4. We know about the mental phenomena of others only by observing their behavior.
This is what behaviorism holds, of course, but it also is implied by (e.g.) the Turing Test for machine intelligence. It also seems to be assumed in many discussions of the "problem of other minds."
Searle: I know others are conscious not because of the way they behave, but because I know that they have a biological makeup similar to mine.
5. Behavior and causal relations to behavior are the essence of the mental.
Searle: behavior is contingent, not essential. Mental states don't require behavior at all.
I'm not sure why Searle focuses on behavior at the expense of perception: functionalism stresses the importance of both perceptual "inputs" and behavioral "outputs."
It's true that a brain in a vat, e.g., will not actually behave at all (or perceive at all, if perception requires sense organs). But it still seems plausible to me that what mental states it is in has essentially to do with the inputs they would be caused by and the behavior they would cause if the brain were more normally connected to the world. (Of course, for many mental states the connections are very indirect; your beliefs about astrophysics might only affect your behavior if someone asks you about them.)
6. Everything is in principle knowable.
Searle: some things are in principle unknowable. (For example, it might be impossible to know what it's like to be a bat.)
I'm not sure why S thinks the tradition is committed to view # 6. I guess it depends on what's meant by "in principle," but on anyone's view things beyond our event horizon, for instance, are unknowable. Moreover, I think most people would agree that there may be mathematical truths too complex for the human mind to grasp. And Goedel's Incompleteness Theorem suggests that there may even be true mathematical propositions that we can understand, but cannot prove, and therefore may never know to be true. (The theorem does not directly imply anything about human knowledge, just about axiomatized theories, but it's suggestive.)
7. Everything is physical, in the Cartesian sense.
Searle: everything is physical (for instance, consciousness is a biological phenomenon), but not in the Cartesian sense. The Cartesian sense of "the physical" is hopeless. For one thing, being physical does not imply being nonmental and being mental does not imply being nonphysical. For another, D's criterion for being physical (having extension) is inconsistent with modern physics.
Searle says that modern physics treats electrons as extensionless points. I'm not certain this is correct -- materials I have looked at say that the diameter of an electron is too small to be measured, i.e. less than 10-18 m. But saying the diameter is too small to be measured seems a lot different than saying that there isn't one. --Not that this really matters. It does seem fishy to make extension the defining characteristic of the physical, as though if physics did discover an extensionless particle it couldn't be a physical thing. The weird conception of the physical as extensionless is also partly responsible for Leibniz's weird notion of the monad. [I can't resist bragging here that I have driven on Monad Road in Billings, Montana. No connection with Leibniz, unfortunately, but still pretty cool.]
But I don't think contemporary philosophers accept the Cartesian conception of the physical anyway, so this seems a bit of a red herring. I would suggest something like this as a definition of the physical with respect to which the mind-body problem can be coherently posed. The definition is recursive.
- Any object or property that plays a role in basic physics is a physical object or property.
- Any object that is constructed entirely out of physical objects is a physical object.
- Any property that supervenes on physical properties is a physical property.
(Actually 3 requires some amplification, since for an epiphenomenalist dualist mental properties would naturally supervene on physical properties without thereby being physical. I have in mind supervenience with respect to all metaphysically possible worlds, not merely with respect to all nomologically possible worlds.)