Eliminativist positions in philosophy are a variety of anti-realism, which is in turn typically contrasted with realistand reductionist positions. A realist account of some phenomenon takes it to be both real and essentially what it appears to be. A reductionist account of some phenomenon takes it to be real but not what it appears to be. An eliminativist view of some phenomenon would take it to be in no way real, and something we ought to eliminate from our account of the world altogether. Instrumentalism is a milder version of anti-realism, where an instrumentalist view of some phenomenon holds that it is not real but nevertheless a useful or even indispensible fiction.
So, for example, a realist account of the mind would hold that it is both real and (just as it appears to be) irreducible to anything material; a reductionist account of the mind would hold that it is real but “really” just “nothing but” something material; and an anti-realist position would be that the mind is not real at all and should either be regarded merely as a useful fiction or eliminated altogether from our account of human beings and replaced by concepts derived entirely from physical science. A realist account of free will would hold that it is both real and (just as it appears to be) incompatible with causal determinism; a reductionist account would hold that free will is real but compatible with determinism; and an anti-realist position would be that it is in no way real. And so forth.
Some forms of anti-realism might seem at least coherent, whether or not they are true. For example, someone who takes an anti-realist position in ethics -- that is, who denies that moral notions like “good” or “right” name any real features of the world -- is, arguably, not taking a self-defeating position, even if he is taking an incorrect position. The same might seem to be true with respect to anti-realism about the existence of God, i.e. atheism.
In fact, I think, things are not quite that simple. At least given an Aristotelian-Thomistic metaphysics of the good, on which the true and the good qua transcendentals are convertible with one another, you cannot coherently affirm that it is truethat there is no such thing as goodness. (See the relevant sections of chapters 3 and 5 of Aquinas.) Nor, I would say, can you consistently affirm that the world is intrinsically intelligible while denying that there is something that is actus purus rather than a compound of act and potency, or ipsum esse subsistens rather than having merely derived existence. And in that case at least certain forms of atheism will ultimately be incoherent. (I addressed the incoherence of denying that the world is intrinsically intelligible in a couple of earlier posts, hereand here.) However, it obviously takes a fair bit of work to establish such claims about the good and God. The incoherence (as opposed to mere incorrectness) of denying their reality is certainly not obvious or blatant.
Blatant incoherence is more commonly attributed to eliminativist views about consciousness or thought. Even here there might seem to be wiggle room. The eliminativist vis-à-vis consciousness can claim that what he denies is not consciousness per se but only the existence of qualia -- those aspects of conscious experience that are accessible only from the first-person point of view of the subject of the experience. The eliminativist about thought can claim that what he denies are merely propositional attitudes like belief, desire, and the like, but not that there are other information-bearing states in the brain that need to be understood in terms of neuroscience rather than commonsense psychology.
In both these cases I think the incoherence is only disguised rather than avoided. With respect to qualia, one problem is that it is dubious at best whether there is anything left to consciousness when qualia are entirely subtracted from it; another is that the motivation for denying qualia is often supposed to be scientific, but to deny their existence would be to undermine the evidential base of science itself. (This is a paradox which, as I’ve pointed out before, has been noted by thinkers like Democritus and Schrödinger, whose respectability from the point of view of scientism can hardly be denied.)
In the case of thought, the trouble is that the motivation for eliminativism here is the difficulty of accounting for the intentionality, “aboutness,” or directedness of thought in terms of a modern, mechanistic, anti-Aristotelian conception of matter, on which matter is inherently devoid of finality, directedness, or teleology of any kind. Getting rid of beliefs, desires, and the like only eliminates one kind of intentionality. But some kind of intentionality must be affirmed if notions like theory, concept, model, evidence, inference, truth, and the like -- which are central to the very notion of, and practice of, science itself -- are to be affirmed, or even reconstructed in some more scientistically “respectable” way. The notion of “information” seems to do the trick only because it is systematically ambiguous. If meant in something like the technical, Claude Shannon sense, it is itself prima facie compatible with scientism, but irrelevant to reconstructing inherently intentional notions like theory, concept, truth, etc. in materialist-friendly terms. If meant instead in the ordinary sense, it is relevant, but then smacks of intentionality of just the sort the advocate of scientism was supposed to be explaining away. (I’ve discussed these sorts of problems with eliminativism about intentionality in several places, such as here.)
But I would say that all of this is secondary to what I take to be the two areas in which eliminativism reaches its absolute, undeniable limits in principle: formal or abstract thought; and change. The first is what James Ross, in an argument I defend at length in an article in the latest ACPQ, notes is essentially determinate in a way material properties and processes cannot be in principle. As Ross argues, to deny that our thought processes are ever really determinate -- to deny, for example, that there is ever a fact of the matter about whether we add, square, reason in accordance with modus ponens, etc. -- is doubly incoherent. For one thing, it entails that none of our arguments -- including the arguments that purportedly support the denial that we ever have thoughts of a determinate form -- is valid. For another, even to deny that we ever really add, reason in accordance with modus ponens, etc. requires that we grasp what it would be to do these things, and that requires having thoughts that are determinate in the ways in question.
That denying change cannot coherently be done has been obvious since Parmenides and Zeno first tried to do it. Even to entertain their sophistical arguments requires that one work through their premises and, if one is to come around to their view, that one be convinced that their reasoning is sound -- all of which involves change. Modern, Einstein-inspired attempts to deny the reality of change face a similar incoherence if pushed through consistently, as I argued in my recent paper on motion and inertia.
Now it is the reality of formal or abstract thought that, in the view of classical philosophers, provides the chief reason why our intellectual faculties cannot possibly be entirely accounted for in material terms. (See my defense of Ross for the full story.) And the reality of change is the foundation of the Aristotelian theory of act and potency, which is in turn the key to the chief Aristotelian-Thomistic proofs of the existence of God. New Atheist types in love with the ad hominem will no doubt be quick to conclude that this must be the reason why some philosophers insist that change and formal thought cannot coherently be eliminated. But it is rather obvious why someone might agree that there is something fishy in denying the reality of change or formal thought processes even if he is not inclined either to theism or dualism. What is much harder to see is why anyone would for a moment take seriously eliminativism about change or formal thought unless he was motivated to try to avoidtheism and dualism. As is so often the case, the person quick to fling an ad hominem will soon find he has thrown a boomerang.
More interesting, perhaps, is the question why eliminativism about change and formal thought does not these days get the attention that eliminativist views regarding consciousness, intentionality, and the like do -- especially given that, as I would claim, the existence of change and formal thought processes ultimately pose the gravest challenge to naturalism, scientism, and related views. Part of the answer is the general ignorance of the arguments of classical (Platonic, Aristotelian, Scholastic) natural theology and philosophical psychology that prevails today, and about which I so often complain. When the modern reader hears talk of arguing from the world to God, he thinks of Paley and Leibniz, of “irreducible complexity,” Sufficient Reason, and the like -- not of the theory of act and potency. When he hears talk of the immateriality of the mind, he thinks of qualia or perhaps of intentionality understood as mere directedness on to an object -- neither of which have much to do with Aristotelian or Thomistic arguments for the immateriality of thought.
A more remote cause, I would speculate, lies in the two epistemological doctrines that first vied to replace the Aristotelian-Scholastic conception of knowledge -- rationalism and empiricism. The Scholastics affirmed the principle of causality, according to which any actualized potency must be actualized by something already actual. This is a claim about objective reality, part of the theory of act and potency, whose foundations lie in the philosophy of nature and the analysis of how change as a feature of the objective world is possible. The rationalists pushed this aside in favor of the “Principle of Sufficient Reason,” which is a purported “law of thought” rather than a thesis about objective, empirically knowable reality. Change per se as the starting point for arguments in natural theology dropped off the “mainstream” radar screen, and failed to return even after the desiccated rationalist versions of the old proofs were dealt their supposed death blows by Hume and Kant.
Meanwhile, the empiricists crudely conflated conceptual thought with mental imagery, thereby obscuring that aspect of the mind that the Scholastics regarded as truly distinctive of human beings and the obvious mark of immateriality. Even though later philosophers would see through the empiricists’ sophistries on this particular score, Locke, Berkeley, and Hume had succeeded in kicking up enough dust that the debate over materialism would no longer focus primarily on conceptual thought but instead on secondary issues (again, qualia and intentionality understood as mere directedness on to an object -- neither of which are essentially incorporeal on an Aristotelian-Scholastic view).
(I said more about the role modern rationalism and empiricism have played in obscuring the arguments of classical and Scholastic writers in a post on the philosophy of nature some months back.)
In any event, a failure to see their theistic and dualistic implications is surely at least one reason why change and formal thought do not show up in the contemporary eliminativist’s crosshairs as frequently as (say) intentionality or consciousness do. One way to avoid seeing the obvious is to try to convince yourself that your eyes are lying to you. Another is just to look in the wrong direction.