Truth (Princeton Foundations of Contemporary Philosophy) Read online

  5.1 REALISM VS DEFLATIONISM

  Traditional realists would insist, against traditional idealists and pragmatists, that whether the proposition that snow is white is true depends on whether snow really is white, and not on whether the thought that it is fits comfortably with our other ideas or is convenient to adopt in practice. Deflationists cannot be accused of holding that the truth of the proposition depends on our comfort or convenience, but some realists complain that deflationism cannot account for a clear commonsense intuition according to which it is snow's being white that makes the proposition true—and not vice versa. For according to deflationism, to say that snow is white and to say that the proposition is true are equivalent by virtue of meaning, and such equivalence is a symmetric relation, whereas dependence is not. The deflationist seems to have no more justification for giving the answer “Because snow is white” to the question “Why is the proposition that snow is white true?” than for giving the answer “Because the proposition that snow is white is true” to the question “Why is snow white?”

  Deflationists answer that even where a biconditional holds by definition, there generally remains the asymmetry that the term whose definition is involved occurs on one side but not the other, and that is enough to explain and justify a corresponding asymmetry in colloquial language. Thus we may say that if a woman's husband is dead, that makes her (count as) a widow, but not that if she is a widow, that makes her husband (count as) dead. In a similar sense, we may say that if snow is white, that makes the proposition that snow is white (count as) true, but not that if the proposition is true, that makes snow (count as) white. Inflationists may insist that asymmetry between definiens and definiendum is not a genuine explanatory asymmetry in any philosophically weighty sense, but deflationists can reply that while there may be a commonsense intuition that snow's being white in some sense “makes” the proposition that snow is white true, it is not clear that it is a philosophically weighty sense.

  It is worth mentioning that Googling turns up many examples of questions of the form “Why is it true that p?” in areas from pop psychology (“Why is it true that matters of the heart and wallet are the areas where we are most susceptible to self-deception?”) to differential geometry (“Why is it true that if the charts of a smooth structure overlap in only one connected component then the manifold is orientable?”) and that the answer offered by online experts is never the simple “Because p.” The construction “Why is it true that…” or “Why is it that…” seems in practice to be used mainly as a convenient device for turning an assertion into a why-question, without having to insert the auxiliary verb “do” and so forth—an observation entirely in harmony with deflationism.

  5.2 CORRESPONDENCE THEORIES

  Russell and Moore were each for a time tempted by the view that facts are just true propositions. That view, sometimes dignified with the title of the identity theory, leaves no room for the criticism of deflationism just discussed, that facts make propositions true in some serious sense of “making,” and it was soon enough abandoned for the view that a true proposition corresponds to a fact. Variant formulations of the correspondence theory may take the truthbearer to be, not a proposition, but a sentence or a belief, and may take that to which a truthbearer corresponds to be, not a fact but a state of affairs or situation (though “states of affairs” may be the same as “facts” not conceived of as just true propositions). But the most important differences among correspondence theorists are over the nature of correspondence. The main division, which goes back to Russell and Moore, is between congruence and correlation theorists.

  A congruence theorist holds that a truthbearer and what it corresponds to are both structured complexes, and that when one corresponds to the other, there is likeness of structure, and correspondence of components to components. A version of this view taking sentences as truthbearers and facts or states of affairs as what they correspond to might hold that the true sentence “Snow is white” has two components, the subject “snow” and the predicate “is white,” while the fact or state of affairs of snow's being white has two components, the substance snow and the property whiteness. The sentence is true because it corresponds to the fact or state of affairs, with the subject corresponding to the substance it denotes and the predicate corresponding to the property it connotes. Such a view suggests analyzing truth in terms of denotation and connotation.

  Congruence theorists subdivide over the requirement of “likeness” of structure, with some demanding identity of structure and others only similarity. The two groups are labeled with Greek-derived terms borrowed from mathematics, being called isomorphism and homomorphism theorists. But the differences between the two are subtleties we will not pursue.

  A correlation theorist simply holds that a true truthbearer as a whole corresponds to a fact or state of affairs or situation as a whole, with no further analysis. The main complaint about correlation theories is that they add nothing but rhetoric to the deflationist account on which it is true that things are some way iff things are that way. Austin's variant does posit descriptive conventions relating a sentence type to a general kind of situation and demonstrative conventions relating a sentence token to a specific historic situation, with a token being true iff its specific situation is of its type's general kind. That's telling us something, but it's not telling us much.

  There are serious worries about congruence theories, however. A first worry, often cited by correlation theorists, concerns atomic sentences, ones that like “Snow is white” are not logically compounded out of simpler ones. The problem is that though it may be clear what the components, snow and whiteness, of the corresponding fact are supposed to be, it is not so clear what is supposed to unite them, to bind them together to produce the fact. Some say that snow instantiates whiteness, but that only raises further questions. For saying that instantiation glues snow to whiteness seems to make instantiation itself a component of the fact; that then seems to call for some kind of meta-instantiation to glue snow and whiteness to instantiation; and that then seems to call for some further adhesive; and so on. A regress threatens. (Many see a similar difficulty on the side of truthbearers if these are taken to be propositions.)

  A second worry is that for many logically compound truths, including heterogeneous disjunctions such as “Snow is white or grass is green,” and especially negative existentials such as “Hippogriffs do not exist,” it is not clear what the corresponding fact is supposed to be, or even that there is one. We may say colloquially that the fact that hippogriffs do not exist explains why hippogriff-hunters always come home empty-handed, and Ramsey explained such colloquial speech in terms of a deflationist or lightweight theory of facts, paralleling his theory of truth. But metaphysically inclined theorists have a weightier conception of facts, as structured complexes of objects and properties or whatever. It is not clear what structured complex if any could plausibly be identified with the heavy-duty fact that hippogriffs do not exist. Are we to recognize the whole wide world as an object, and freedom from hippogriffs as a property it instantiates? That sounds so far-fetched that many metaphysically inclined theorists have been reluctant to recognize negative existential facts. Some hesitate to recognize even disjunctive facts.

  5.3 TRUTHMAKER THEORIES

  Reluctance to posit such facts eventually led Russell to the view (also seen in early Wittgenstein) that the truth of compound propositions derives from facts only indirectly, by way of their logical relations to atomic components whose truth derives directly from correspondence with facts. This logical atomism represents a serious departure from the correspondence theory.

  A not dissimilar departure is found in the writings of many metaphysically inclined theorists today, who put less emphasis on correspondence, and more on each truth having some thing(s) that make(s) it true, in some philosophically weighty sense of making. Whereas the word “correspondence” may suggest a one-one relation, many today allow that a single truthmaker may make more than
one truthbearer true, and a single truthbearer may be made true by more than one truthmaker. The fact that snow is white may make the two propositions true, that snow is white and that either snow is white or grass is green, which latter may also be made true by the fact that grass is green. As the example suggests, truthmaker theory does away with any need for disjunctive facts, such as a fact that snow is white or grass is green.

  But there is still the problem of negative existentials. The late David Lewis, for instance, though sympathetic to motivations underlying the positing of truthmakers, suggested that only affirmatives have them, while taking negatives to be made true, not by the presence of a truthmaker, but by the absence of any falsehood-maker. The truthmaker principle asserts that for any two possible world-states, if something would have been true if the world had been in one of them that would not have been true if the world had been in the other, then something would have existed in the former case that would not have existed in the latter. Lewis would amend this by adding “or something would not have existed in the former case that would have existed in the latter,” thus obtaining the (weaker) principle of supervenience of truth on being.

  Debates continue over simple cases as well, not only about what unites the components of the fact or state of affairs, but also over whether it is facts or states of affairs that are the truthmakers at all. Some say that it is not the fact or state of affairs of Frosty the Snowman's being white that makes the proposition that Frosty the Snowman is white true, but rather a feature of Frosty, Frosty's whiteness. Various unattractive labels (“tropes,” “abstract particulars,” “qualitons”) for such features compete with each other in the literature. Terminological differences combine with substantive disagreements to make truthmaker theory one of the murkier areas of metaphysics.

  Fortunately we are excused from having to penetrate into the murk by Lewis's observation that truthmaker metaphysics may not really have very much to do with truth. For the metaphysicians' view is not that, even if snow is white, that is not sufficient to make the proposition that snow is white true. Their view is not that, without the existence of the fact of snow's being white (or “trope” or “abstract particular” or “qualiton,” snow's whiteness), snow could still somehow or other manage to be white, but the proposition that snow is white couldn't manage to be true. Rather, their view is that snow couldn't be white without the fact (or feature) existing. “Things being some way necessitates the being of some thing” might be their slogan. This is perhaps not so catchy a slogan as “No truth without a truthmaker,” but in the catchier formulation “truth” may be serving as little more than a device of generalization of the kind deflationism emphasizes.

  5.4 PHYSICALISM

  Molière mocked physicians of his day who “explained” why opium is capable of making people sleepy by positing that it has a sleepy-making capability (virtus dormitiva). The logical positivists, who (like Nietzsche before them) liked to cite that example of “meaningless metaphysics,” would have taken an attitude like Molière's towards metaphysicians debating which of the two it is, snow's being white or snow's whiteness, that makes it true that snow is white. For the positivists, the very notion of truth was suspect because it seemed metaphysical. Tarski convinced many positivists that his definition showed that acceptance of a notion of truth was compatible with physicalism as the positivists understood it. Decades later Hartry Field questioned whether the definition really establishes compatibility with physicalism (as he understands it).

  What did Field want that Tarski did not supply? Tarski's recursive definition of truth depends on a list of base clauses like the ones for the toy language considered in §2.3 that tell us “0” denotes zero and “1” denotes one. (If there were more constants in the language, there would be a longer list.) Field in effect wanted a single, uniform, general, physical account of the relation

  (1) Symbol ______ denotes object ______ in the language used by person _______.

  and ultimately a single, uniform, general, physical account of the relation

  (2) Symbol-sequence _______ is true in the language used by person ________.

  This is a very ambitious goal indeed. Any account of (2) in physical terms would stand to metaphysical accounts of truth (such as “homorphism with a fact” or “correlation with a situation”) much as present-day physiological explanations of the action of opiates on the nervous system stand to the “explanations” offered by Molière's contemporaries. It would make identity, isomorphism, homomorphism, correlation, atomist, truthmaker, supervenience, and any other metaphysical theories all seem like empty verbiage by comparison. An account of (2) in physical terms would also accomplish tasks that deflationism insists explicitly (and that metaphysical theories seem to assume implicitly) belong to the theory of meaning rather than the theory of truth, beginning with the task of determining whether or not emitting given sounds or displaying given marks amounts to utterance of a proposition-expressing or “truth-apt” sentence at all. It would refute deflationism by showing that there is enormously more to be said about what truth is than what the equivalence principle tells us.

  But why would anyone have imagined that a task so ambitious could be accomplished? Two factors were at work. On the one hand, Tarski's work (which Field commended for going as far as it did and condemned for not going further) seemed to show a good deal about how to get to (2) from (1). On the other hand, work of Saul Kripke on proper names (along with related work by him and by Hilary Putnam on natural-kind terms), which commentators often labeled a “causal theory of reference,” seemed to show a good deal about how to provide a causal, natural, physical account of (1). The hope seemed to be that a cross between Tarski's theory of truth and Kripke's causal theory might breed a causal theory of truth.

  Critics, notably Scott Soames, pointed out that Tarski did not really supply as much help in getting from (1) to (2) as it may initially have appeared. For the further clauses Tarski adds to the ones about “0” denoting zero and “1” denoting one in order to complete his recursive definition assume that one knows the meaning of negation, conjunction, disjunction, and so forth. Tarski makes no pretence of providing a physicalist account of “symbol _______ expresses disjunction in the language used by person _______,” for instance. There seems also to have been a misunderstanding about the aims and claims of the so-called causal theory of reference.

  Kripke's aim was to show how a proper name or natural-kind term could refer to an individual or kind even though not associated with any description thereof. Grossly simplified, his account, which he called “the chain of communication picture,” ran as follows. Jay may describe an individual or kind and introduce a name by stipulating that it is to refer thereto, as in “Let ‘Aldebaran' denote that orangish star [pointing],” or “Let ‘tigers' denote that striped kind of thing [pointing].” So long as in subsequent uses of the word he intends—and the intention need not be a conscious thought—to refer to whatever he has been using it to refer to, it will continue to do so, even after he has forgotten what description he used and everything else about how he originally introduced it. Kay may hear Jay use the word and decide to use it herself to refer to whatever Jay uses it to refer to. So long as in subsequent uses of the name she intends it to refer to whatever she has been using it to refer to, it will continue to do so, even after she has forgotten when, where, why, how, and from whom she originally picked it up. Then Ella may pick it up from Kay, Emma from Ella, and so on.

  It was a mistake for commentators to call this a “causal theory of reference.” There is no requirement of causal contact between the first introducer of the name and the object named, which might be a number rather than a star. And not only does the chain of communication picture not reduce the notion of reference to causal notions, it does not offer any reductive analysis of reference at all. Reread the summary account of the picture in the preceding paragraph, and you will see that it simply takes for granted a notion of intention to refer. Kripke explicitly an
d emphatically disavowed any aim to provide a reductive analysis. His work may raise doubts about the deflationist claim that there are no substantive questions about reference, simply by being an example of a substantive “theory” about reference, but if his work makes trouble for deflationism, it offers very little help to physicalism.

  Once it is realized how very little Tarski and Kripke offer in the direction of a physicalist account of truth, discouragement may set in. (Field himself eventually abandoned physicalism for a qualified defense of disquotationalism.) Discouragement is reinforced by certain nagging questions. The question how a naturalistic, causal theory could apply to truth in mathematics or ethics might be dismissed by physicalists hardheaded enough to question whether there is any truth in mathematics and ethics, but how could a general, uniform account hope to cover even the physical sciences, considering the great variety of objects, from quarks to quasicrystals to quasars, with which such sciences are concerned, and the diversity of our relations to them? The more fully one understands what the physicalist project would have to accomplish, the more it comes to seem that the project is unrealistic—in the everyday, nonphilosophical sense of “realistic.” It is a pipe dream.

  5.5 UTILITY

  If physicalism survives today, it is less as an active program than as an inchoate feeling that a physicalist account is needed, even if no one will for the foreseeable future be in a position to provide one, for certain explanatory purposes, and especially for explaining the utility of truth. Traditional pragmatists explained the utility of truth by defining truth as utility, but with that no longer considered a live option, it may be thought that a genuine explanation would require a naturalistic, causal account of truth: that a metaphysical theory could only provide a pseudo-explanation (of the virtus dormitiva type), and a deflationist theory not even that.