This post is a stab at a summary and evaluation of Panayot Butchvarov's "Metaphysical Realism and Logical Nonrealism" which is available both online and in R. M. Gale, ed., The Blackwell Guide to Metaphysics (Oxford: Blackwell 2002), pp. 282-302. Page references are to the Blackwell source. The ComBox stands open if readers have some informed commentary to offer. ('Informed' means that you have read Butchvarov's paper, and my response, and you have something pertinent to contribute either in objection to or agreement with either Butchvarov or me.)
Summary
1. The thesis of Butchvarov's article is that metaphysical realism and logical nonrealism are inconsistent: ". . . if logical nonrealism is true, then metaphysical realism is largely false." (283) This is a very interesting result if true. For most of us would be inclined to suppose that logical nonrealism is quite compatible with metaphysical realism. Butchvarov maintains that logical nonrealism is true and is a tenable version of metaphysical nonrealism. But it all depends on how these terms are being used.
2. Metaphysical realism, roughly, is the view that the existence or at least the nature of things is independent of our cognition. Metaphysical nonrealism denies this either by denying that both the existence and nature of things are independent of us (independent of our language, conceptual frameworks, forms of sensibility, forms of understanding, etc.) or by denying that the nature of things is independent of us. On this definition one could be a metaphysical nonrealist if one were a sort of Kantian who admitted the existence of things in themselves, but held that their nature depends on our cognition. It will turn out that Butchvarov is a sort of Kantian subject to a couple of caveats.
3. Logical nonrealism is the view that nothing in reality corresponds to such logical words and phrases as 'and,' 'or,' 'It is not the case that . . .,' 'all,' 'some,' 'no,' 'is' in its various uses as expressing predication, identity, existence, set-inclusion, and so on. Moreover, logical nonrealists also deny that there is anything in reality corresponding to whole declarative sentences thereby eschewing both Fregean and Russellian propositions. Consider 'Tom is tall but Paul is not.' (My example.) The logical nonrealist will take only 'Tom,' 'Paul,' and perhaps 'tall' to have extramental or extralinguistic correlates. For these words are 'nonlogical' or 'categorematical' in an older terminology. The logical nonrealist will deny that 'is,' 'but,' 'not' and the sentence itself have anything in reality corresponding to them.
The logical nonrealist, then, holds that no logical expression has a worldly (extramental, extralinguistic) correlate. They are one and all syncategorematical. Equivalently, he denies that there are any 'logical objects.' Thus there is no object corresponding to 'and' in the way there is an object corresponding to 'Obama.' The logical realist maintains the negation of this, namely, that some logical expressions have worldly correlates. The logical realist holds that there are some 'logical objects.'
The logical realist need not hold that all logical expressions have worldly correlates, for he may hold that the putative reference of some logical expressions can be paraphrased away in terms of others whose reference cannot be paraphrased away. For example, given that 'All Fs are Gs' is equivalent is to 'a is F & b is F & c is F & . . .' one might hold that there is no logical object corresponding to 'all' while holding that there is one corresponding to '&.' (Not that this is a good view to hold.)
4. Butchvarov's thesis, again, is that logical nonrealism is not consistent with metaphysical realism. To embrace logical nonrealism is to embrace a version of metaphysical nonrealism. But what exactly is his argument for this intriguing claim? Getting clear about his argument will help us get clear about the exact sense of his thesis and the exact nature of the metaphysical nonrealism Butchvarov is proposing. Here is what I take to be the argument in bare outline:
A. The world must have a logical structure.
B. The logical structure of the world cannot be independent of language.
Therefore
C. The logical structure of the world must depend on language and thus on us as language-users.
(A) seems unproblematic. I take it to mean that the world and everything in it is subject to the laws of logic. Accordingly, the laws of logic are not mere laws of thought, but laws of being as well: every being qua being is subject to the laws of logic. Butchvarov quotes Aristotle to this effect: metaphysics, the science of being qua being, begins with a study of the principles of the syllogism. (284) Butch refers us to Metaphysics 1005b7-35. The gist of this passage is that "the surest principle of all," the law of noncontradiction (LNC),is not merely a law of thought, but a law of being as well. Or at least that is the way I read the passage.
What logic studies are the relations between and among propositions (declarative sentences, statements, judgments) that hold in virtue of their logical form. The words indicative of logical form are the logical expressions we have already mentioned, 'and,' 'or,' 'is,' etc. Without these expressions the world cannot be described. "To say that the world has a logical structure is to say that any description of it employs such expressions." (284) Since we have no conception of a world that cannot, in principle, be described, it follows that we have good reason to hold that the world must have a logical structure.
So far, so good. Premise (A), as I have explicated it, seems plainly true.
But from (A) alone it does not follow that the world's logical structure is something we supply. For it could be that this structure is inherent in the world in itself and so independent of us and our conceptual and linguistic abilities. That, I think, is what Aristotle would say. Although Butchvarov cites Aristotle, the former's position is importantly different from the latter's. Aristotle was no Kantian; Butchvarov is. There is also an epistemic possibility, one which I doubt Butchvarov takes seriously, namely, that the world's logical structure has a divine origin. If that is the case, then logical nonrealism is true without it being the case that the world's logical structure is imposed by language or by humans using language. For both of these reasons, (A) alone does not entail metaphsyical nonrealism. Butchvarov is of course well aware of this.
(B) is the crucial and controversial premise in Butchvarov's overall argument. (B) is true if and only if logical nonrealism is true. Or at least that is what I take him to be saying. (284) For if there are no logical objects, then whatever job these logical objects perform must be done by parts of language. For example, if there are no logical objects, then there are no Fregean propositions and no truth-making facts. Some of us of a realist bent think that if a sentence such as 'Tom is tall' is true, then there must be something in the world that makes it true. We tend to think that at least some truths need truth-makers, ontological grounds of their truth. Some of us introduce concrete facts to play the truth-making role, the concrete fact, e.g., of Tom's being tall which is distinct both from the sentence 'Tom is tall' and from the Fregean proposition expressed by the sentence on a particular occasion of its use. Butchvarov, however, maintains that such truth-making facts "are merely hypostatized sentences, shadows that sentences cast upon things." (286) Butchvarov's logical nonrealism implies that "the logic of the world is not distinguishable from the logic of words." (286, emphasis added)
I interpret him to mean, not that there is an Aristotelian isomorphism between being and thought or between world and language, but that the logical structure of the world is supplied by language and is reducible to, and therefore identical with, the logical structure of language. Now the world is not "a mere assemblage of items. . . ." (286) But although the world is "a totality of facts not of things" (Wittgenstein), it is not a totality of language-independent facts; it is a totality of things whose fact-like structuration is provided by us as language-users. There is, in other words, no fact-structure independent of us as language-users. The world (reality) is not, in itself, a totality of facts. In itself, the world does not have a propositional, or proposition-like, structure. In itself, the world is nothing like a declarative sentence. The world in itself has no 'logic.' The world in itself is not the embodiment of a language-independent Logos. We, as language-users, supply the logic. We impose it on the things in themselves; we do not find it or discern it in the things in themselves.
Perhaps we can sum up the argument for (C) as follows. There is no world without a logical structure. A logical structure is a sentential structure given that logical nonrealism is true and there are no such 'logical objects' as Fregean propositions and truth-making facts. There is no sentential structure without language. There is no language that is not either a human language or at least translatable in principle into a human language. (286) There is no human language without humans. Therefore, any world that could count as a world for us must be "ours," must be "human," must "depend on us. (286-287) The upshot is "a restrained version of contemporary metaphysical nonrealism and an heir of Kant's transcendental idealism." (284)
The picture is thus broadly Kantian except that the structuring is linguistic rather than mental, and is limited to logical structure. As in Kant, there is a distinction between things-in-themselves and things-for-us. Butchvarov's position is not "conceptual or linguistic creationism." (300) For Wittgenstein, the world is a totality of facts, not of things. For Butchvarov, the world for-us is a totality of facts, but the world in-itself is a totality of things.
Towards an Evaluation
Butchvarov's article is very rich, and I have touched upon only some of its riches. My main concern in this post is to understand what exactly Butchvarov's version of metaphysical nonrealism amounts to and whether it has a clear and unambiguous sense. So I won't question his logical nonrealism. I will instead assume it to be true. I will asume that nothing in reality corresponds to the propositional connectives, the quantifiers, the various uses of 'is,' whole declarative sentences, modal and other operators upon whole sentences, etc. We ought to agree, I think, that whether or not logical nonrealism is true, it is plausible. In 'Tom is tall,' for example, 'is' is plausibly taken to be syncategorematic, to have nothing in reality corresponding to it. It is plausible to maintain that talk of Being -- with a majuscule 'B' -- as we find it in Heidegger and others is what thinkers as diverse as Adorno and Frege said it was, an hypostatization or even an apotheosis of the copula. (See my The Copula: Adorno Contra Heidegger.)
I won't even question the inferential move from logical nonrealism (There are, in reality, no logical objects) to premise (B) according to which the logical structure of the world cannot be independent of language. One could question this inference by saying, "Yes indeed, there is none of this Platonica that Fregeans and others posit; but surely it doesn't follow straightaway that language or language-users introduce logical structure into the world; it could be that a Transcendental Mind does that. If Butchvarov asserts that truth-making facts or states of affairs "are merely hypostatized sentences, shadows that sentences cast upon things" (286) then with apparently equal justice we can tax him with a hypostatization of Language and an ascription to it of world-constituting powers it cannot possess; and if he says that it is not Language, but human language-users who import logical structure into the world, then we can wonder how such measly members of a zoological species could be up to such an onerous task." Perhaps in a separate post I will consider whether there is anything to this line of critique. But for now I set it aside.
What I want to question is the coherence of the main conclusion Butchvarov draws:
Therefore, insofar as we can conceive of the structure of the world and thus of the world itself as a world, they are "ours," "human," they "depend" on us. Of course, that the only conception of logical structure we have is that of sentences, of language, does not entail that the world does not have that or some other kind of structure independently of language. But it is a very good reason for reaching such a conclusion. For it does entail that our cognition of the world, insofar as it involves logical concepts, depends on language, and that so does the world insofar as it is cognized by us. (286-287, emphases in original)
Butchvarov is arguing in effect that (T1) below entails (T2), and that while (T2) does not entail (T3), (T2) is yet a "very good reason" for (T3).
T1. The only conception of logical structure we have is that of sentences, of language.
T2. The world insofar as it is cognized by us depends on language, at least in respect of its logical structure.
T3. The world depends on us, at least in respect of its logical structure.
(T2) appears to be a tautology. For the cognized world, qua cognized, cannot be cognized without cognizers. Butchvarov speaks (283) of propositions like (T2) as "virtual tautologies." (T3), however, is not a tautology, but a substantive metaphysical thesis that rules out metaphysical realism. How can a tautology be a good reason to accept a non-tautology? Obviously, if (T2) is a genuine tautology, then it cannot be a reason to accept either (T3) or its negation.
For (T2) to be a reason for (T3), it has to be more than a tautology, it has to 'say something.' By use of the slippery phrase "virtual tautology" Butchvarov appears to signal that it does say something. But (T2), by my lights, is just a plain old tautology.
So my first objection is that (T2) does not gives us a good reason, contrary to what Butchvarov says, for accepting (T3).
A second worry is as follows. Butchvarov is committed to things in themselves. But given his denial of logical realism, they must lack all logical structure. This implies that things in themselves are not subject to the laws of logic. Consider the law of noncontradiction: nothing is both F and not F at the same time, in the same respect, and in the same sense. On Butchvarov's scheme this principle can be true only of things-for-us, things as conceptually and linguistically processed by us. And the same goes for the other laws of logic. But that to which logic is not applicable is not thinkable, and so the very notion of a realm of things in themselves becomes empty and unthinkable. We cannot say that this realm is or is not, is one or many, is dependent on us or not, etc. Under dialectical pressure it appears to evaporate. And yet Buchavarov will not deny "The need for something like Kant's distinction between things-in-themselves and things-for-us." (300) Nor can he lest he fall into "conceptual or linguistic creationism." (Ibid.)
Butchvarov is entangled in familiar difficulties. As Jacobi said about Kant, and I quote from memory, "Without the Ding an sich one cannot enter Kant's system, but with it one cannot remain there." If there is structuring from the side of the subject, whether mental or linguistic, then there has to be something that gets structured. Otherwise, talk of structuring is incoherent. If cognition is processing, if to know is to impose forms, then there must be that upon which the forms are imposed, something which, in itself is bare of the forms imposed by the subject. But philosophical experience has shown that this scheme-content, or form-matter distinction is hard to uphold. Kant gave way to Fichte, and this sort of move, this liquidation of the Ding an sich, has been reenacted in recent years.
Butchvarov's view strike me as more radical than Kant's involving as it does the very logical form of things. Kant, in some passages at least, held that the realm of noumena was thinkable, albeit in an empty sort of way, using logical categories, whose reach was not restricted to objects of sensible intuition. His point, at least in those passages, was that the Ding an sich could be thought but not known, knowledge requiring the data of sensibility. After all, noumena are intelligibilia, understandables. Butchvarov's view is more radical inasmuch as the in-itself becomes unintelligible due to the fact that the very logical forms of things are imposed by language-users.
Metaphysical realism, of course, has its own set of difficulties.
Comments