Is Metaphysics Possible?
It may also be that there is no internal unity to metaphysics. More strongly, perhaps there is no such thing as metaphysics—or at least nothing that deserves to be called a science or a study or a discipline. Perhaps, as some philosophers have proposed, no metaphysical statement or theory is either true or false. Or perhaps, as others have proposed, metaphysical theories have truth-values, but it is impossible to find out what they are. At least since the time of Hume, there have been philosophers who have proposed that metaphysics is “impossible”—either because its questions are meaningless or because they are impossible to answer. The remainder of this entry will be a discussion of some recent arguments for the impossibility of metaphysics.
Let us suppose that we are confident that we are able to identify every statement as either “a metaphysical statement” or “not a metaphysical statement”. (We need not suppose that this ability is grounded in some non-trivial definition or account of metaphysics.) Let us call the thesis that all metaphysical statements are meaningless “the strong form” of the thesis that metaphysics is impossible. (At one time, an enemy of metaphysics might have been content to say that all metaphysical statements were false. But this is obviously not a possible thesis if the denial of a metaphysical statement must itself be a metaphysical statement) And let us call the following statement the “weak form” of the thesis that metaphysics is impossible: metaphysical statements are meaningful, but human beings can never discover whether any metaphysical statement is true or false (or probable or improbable or warranted or unwarranted).
Let us briefly examine an example of the strong form of the thesis that metaphysics is impossible. The logical positivists maintained that the meaning of a (non-analytic) statement consisted entirely in the predictions it made about possible experience. They maintained, further, that metaphysical statements (which were obviously not put forward as analytic truths) made no predictions about experience. Therefore, they concluded, metaphysical statements are meaningless—or, better, the “statements” we classify as metaphysical are not really statements at all: they are things that look like statements but aren’t, rather as mannequins are things that look like human beings but aren’t.
But (many philosophers asked) how does the logical positivist’s central thesis
The meaning of a statement consists entirely in the predictions it makes about possible experience
fare by its own standards? Does this thesis make any predictions about possible experiences? Could some observation show that it was true? Could some experiment show that it was false? It would seem not. It would seem that everything in the world would look the same—like this—whether this thesis was true or false. (Will the positivist reply that the offset sentence is analytic? This reply is problematic in that it implies that the multitude of native speakers of English who reject the logical positivists’ account of meaning somehow cannot see that that sentence is true in virtue of the meaning of the word “meaning”—which is no technical term but a word of ordinary English.) And, therefore, if the statement is true it is meaningless; or, what is the same thing, if it is meaningful, it is false. Logical positivism would therefore seem to say of itself that it is false or meaningless; it would be seem to be, to use a currently fashionable phrase, “self-referentially incoherent”.
Current advocates of ‘metaphysical anti-realism’ also advocate a strong form of the thesis that metaphysics is impossible. Insofar as it is possible to find a coherent line of argument in the writings of any anti-realist, it is hard to see why they, like the logical positivists, are not open to a charge of self-referential incoherency. Indeed, there is much to be said for the conclusion that all forms of the strong thesis fall prey to self-referential incoherency. Put very abstractly, the case against proponents of the strong thesis may be put like this. Dr. McZed, a “strong anti-metaphysician”, contends that any piece of text that does not pass some test she specifies is meaningless (if she is typical of strong anti-metaphysicians, she will say that any text that fails the test represents an attempt to use language in a way in which language cannot be used). And she contends further that any piece of text that can plausibly be identified as “metaphysical” must fail this test. But it invariably turns out that various sentences that are essential components of McZed’s case against metaphysics themselves fail to pass her test. A test-case for this very schematic and abstract refutation of all refutations of metaphysics is the very sophisticated and subtle critique of metaphysics (it purports to apply only to the kind of metaphysics exemplified by the seventeenth-century rationalists and current analytical metaphysics) presented in van Fraassen 2002. It is a defensible position that van Fraassen’s case against metaphysics depends essentially on certain theses that, although they are not themselves metaphysical theses, are nevertheless open to many of the criticisms he brings against metaphysical theses.
The weak form of the thesis that metaphysics is impossible is this: there is something about the human mind (perhaps even the minds of all rational agents or all finite rational agents) that unfits it for reaching metaphysical conclusions in any reliable way. This idea is at least as old as Kant, but a version of it that is much more modest than Kant’s (and much easier to understand) has been carefully presented in McGinn 1993. McGinn’s argument for the conclusion that the human mind is (as a matter of evolutionary contingency, and not simply because it is “a mind”) incapable of a satisfactory treatment of a large range of philosophical questions (a range that includes all metaphysical questions), however, depends on speculative factual theses about human cognitive capacities that are in principle subject to empirical refutation and which are at present without significant empirical support. For a different defense of the weak thesis, see Thomasson 2009.