A reader, recently deployed to Afghanistan, finds time to raise an objection that I will put in my own words to make it as forceful as possible:
You endorsed William Lycan's Moorean refutation of eliminative materialism, but then you criticized him for thinking that Moorean appeals to common sense are also effective against standard idealist claims such as Berkeley's thesis that the objects of ordinary outer perception are clusters of ideas. You maintained that there is a crucial difference between the characteristic claims of eliminativists (e.g., that there are no beliefs, desires, intentions, pleasures, pains, etc.) and the characteristic claims of idealists (e.g., Berkeley's thesis just mentioned, McTaggart's thesis of the unreality of time, Bradley's of the unreality of relations.) The difference is that between denying the existence of some plain datum, and giving an account of a plain datum, an account which presupposes, and so does not deny, the datum in question. In effect, you insisted on a distinction between identifying Xs as Ys, and denying the existence of Xs. Thus, you think that there is an important difference between identifying pains with brain states, and denying that there are pains; and identifying stones and physical objects generally with collections of ideas in the mind of God and denying that there are physical objects. But in other posts you have claimed that there are identifications which collapse into eliminations. I seem to recall your saying that to identify God with an unconscious anthropomorphic projection, in the manner of Ludwig Feuerbach, amounts to a denial of the existence of God, as opposed to a specification of what God is. Similarly, 'Santa Claus is a fictional character' does not tell us what Santa Claus is; it denies his very existence.
Now why couldn't Lycan argue that this is exactly what is going on in the idealist case? Why couldn't he say that to identify stones and such with clusters of ideas in the mind of God is to deny the existence of stones? Just as God by his very nature (whether or not this nature is exemplifed) could not be an anthropomorphic projection, so too, stones by their very nature as physical objects could not be clusters of ideas, not even clusters of divine ideas.
It seems you owe us an account of why the reduction of physical objects to clusters of ideas is not an identification that collapses into an elimination. If you cannot explain why it does not so collapse, then Lycan and Co. will be justifed in deploying their Moorean strategy against both EM-ists and idealists. They could argue, first, that idealism is eliminationism about common sense data, and then appeal to common sense to reject the elimination.
I think it is clear that someone who identifies God with an anthropomorphic projection simply denies the existence of God. This putative identification collapses into an elimination. You are not telling me what God is when you tell me he is an anthropomorphic projection, you are telling me that there is no such being. Same with felt pains. A putative identification of a felt pain with a brain state collapses into an elimination of felt pains. For a felt pain simply cannot be identical to a brain state: it has properties no brain state could possibly have. But an identification of a physical object with a cluster of items such as Berkeleyan ideas or Husserlian noemata, items that exist only mind-dependently, does not collapse into an elimination, the reason being that there is nothing in the nature of physical objects as we experience them that requires that such objects exist in splendid independence of any mind. I just located my coffee mug on my desk, and now I am drinking from it. There is nothing in my experiential encounter with this physical thing that requires me to think of it as something that exists whether or not any mind exists. And so I am not barred from the idealist interpretation of the ontological status of stones and coffee cups and their parts (and their parts . . .). Nor does the meaning of 'coffee cup' or 'physical object' constrain me to think of such things as existing in complete independence of any mind. Neither phenomenology nor semantics forces realism upon me. There is nothing commonsensical about either realism or idealism; both are theories.
Neither a realist nor an idealist interpretation of the ontological status of physical objects can be 'read off' from the phenomenology of our experiential encounter with such things or from the semantics of the words we use in referring to them and describing them. Only if realism were built into the phenomenology or the semantics would an identification of a physical thing with a cluster of mind-dependent items collapse into an elimination of the physical thing. For in that case it would be the very nature of a physical thing to exist mind-independently so that any claim to the contrary would be tantamount to a denial of the existence of the physical thing. The situation would then be exactly parallel to the one in which someone claims that God is an anthropomorphic projection. Since nothing could possibly count as God that is an anthropomorphic projection, any claim that God is such a projection amounts to a denial of the existence of God. But the cases are not parallel since there is nothing in the nature of a physical thing as this nature is revealed by the phenomenology of our encouter with them to require that physical things exist in sublime independence of any and all minds.
Of course, one could just stipulate that physical objects are all of them mind-independent. But what could justify such a stipulation? That would be no better than Ayn Rand's axiomatic declaration, Existence exists! What Rand means by that is that whatever exists exists in such a way as to require no mind for its existence. But although that may be true, it is far from self-evident and so has no claim to being an axiom.
In sum, token-token-identity theory in the philosophy of mind collapses into eliminativism about mental items. As so collapsing, it is refutable by Moorean means. The identitarian claims of idealists, however, do not collapse into eliminativist claims, and so are not refutable by Moorean means.