- Home
- Burgess, John P. ; Burgess, Alexis G.
Truth (Princeton Foundations of Contemporary Philosophy) Page 8
Truth (Princeton Foundations of Contemporary Philosophy) Read online
Page 8
Linguistic evidence for the reality of the distinction includes the fact that (2b) would be pronounced with a distinctive stress (indicated by our italics) not present in (2a), and the fact that in (3b) “not intelligent” cannot be contracted to “unintelligent” as it can be in (3a).
The two-negations theorist is in effect answering the question “Have you stopped?” by saying “Neither yes nor no would be an appropriate answer,” and would answer the follow-up question “Well, would it be true to say that you have stopped?” in exactly the same way. Giving the same answer to both questions is, of course, just what one would expect of a defender of the equivalence principle. If there is any problem here it is that both answers are what lawyers call nonresponsive. The question whether “You have stopped” is true or false or neither or both does not go away simply because some theorist declines to answer it. In short, the two-negations approach looks less like a solution than an evasion.
Dependency. A somewhat less unresponsive answer than a mere “I don't say yes and I don't say no” would be “It all depends.” It may be suggested that in our examples of indeterminacy, the problematic question might appropriately be answered “Yes” in some contexts and “No” in others, and that it is only on account of its context-dependence that it cannot be answered either way in isolation.
For instance, with the color chips, it might be suggested that it is natural to call the borderline chip “red” in some contexts (say where one is working with pure red, orangish red, borderline, and pure orange, but no reddish orange) and to call it “orange” in others (say where one is working with pure orange, reddish orange, borderline, and pure red, with no orangish red). Less plausibly perhaps, it might be suggested that “Have you stopped?” is in some contexts most naturally taken as “Do you agree that if you ever did, you have since stopped?” and answered “Yes” (understanding the conditional as in mathematics, so that it counts as true if its antecedent is false), while in other contexts it would be most naturally taken as “Do you agree that you used to, but have now stopped?” and answered “No.”
The view that the paradox of the horned one involves some kind of context-shift between one premise and the other may not be especially common or inviting, but a view that context-shift is crucial to the paradoxes of the heap and the bald one is fairly widely held. More generally speaking, “contextualism” of one sort or another is an approach to the problems of vagueness that is of growing popularity among philosophers of language, though like almost any philosophical ‘ism it also has its detractors.
Defeatism. Another course would be to admit defeat, to surrender the equivalence principle and try to reconstruct one's theory of truth so as to retain its most important features while giving up whatever features committed it to the now-abandoned position that the equivalence principle holds as an exceptionless law. For a deflationist, reconstruction would presumably involve (i) coming up with some amended equivalence principle; (ii) revising the thesis that the equivalence principle is all there is to the meaning of the truth predicate so as to refer to the equivalence principle as amended; (iii) retaining the thesis that an account of the meaning of the truth predicate is all that is required of a theory of truth.
If presupposition failure were the only problem to worry about, an amended equivalence principle might say that saying something and saying it is true are equivalent except that presuppositions of the former become entailments of the latter. It would be left to the theory of meaning, not of truth, to identify what the presuppositions of a given sentence are. This amended equivalence principle implies that if the presupposition of the question whether p is that q, then it is true that p iff q and p, and false that p iff q but not p, and neither iff not q. This proposed adjustment to the deflationist position in order to accommodate presupposition failure is painless enough that one might hope that the problem of borderline cases could be handled by assimilating it to presupposition failure. Perhaps “Chip #500 is red” can be construed as having the failed presupposition that chip #500 is either noticeably closer in color to one of the paradigms than to any of the foils, or else noticeably closer to one of the foils than to any of the paradigms.
4.5 RELATIVITY
Our discussion thus far, needless to say, does not settle the question of the bearing of indeterminacy phenomena on the theory of truth; nor can the extensive literature on the issue be said to resolve it decisively. A question mark, or rather, a cluster of them, hangs over the subject. Hartry Field, one of the most prolific writers on deflationism, urges its sympathizers to adopt it only as a methodological stance, something to be held onto provisionally until forced to give it up. The same would go for most of deflationism's rivals, to the degree that they, too, are committed to the equivalence principle.
Our discussion has been not only inevitably incomplete in its treatment of the vast topics of presupposition and vagueness, but also incomplete in that it addressed only those two types of alleged indeterminacy. We should, and in this section do, address at least one further type, a certain kind of alleged relativity that has inspired the development by John MacFarlane and others of a recently much-discussed view called by its sympathizers “truth relativism,” and by the less sympathetic “new-age relativism.”
The kind of relativity in question is claimed to affect judgments of probability, of morality, and of a number of other matters. We will concentrate on moral relativity, but let us first, by way of contrast, consider the less problematic case of legal relativity. Consider the following exchange:
(4) Abortion is legally prohibited.
(5) That's true in Chile, but not in China.
Everyone accepts legal relativism, the view that legality is relative to legal systems, which vary with time and place. (4) is in the present tense, indicating the pertinent time is the present, but it gives no indication of the pertinent place, leaving an opening for (5). But do we really want to say that there is a single item, the proposition that abortion is legally permissible, that is true in some places and not others?
The more usual view is that what (5) conveys is not that, but something like the following:
(5’) It's true that abortion is legally prohibited in Chile, but not that abortion is legally prohibited in China.
On this view, considered in isolation (4) fails to express a specific proposition, while uttered in a context where Chilean (respectively, Chinese) law is understood to be at issue, it expresses a true (respectively, false) one. If two legal experts have been discussing the legal situation in different countries for so long that they have lost track of which country is currently under discussion, and one who thinks they are still on Chile says (4), while one who thinks they have moved down the alphabet to China says
(4’) Abortion is legally permitted.
then the two have miscommunicated rather than disagreed.
Considered out of context, the form of words “The proposition that abortion is legally prohibited” fails to denote anything, while taken in a context where some specific legal system is tacitly understood to be intended, it denotes the proposition that abortion is legally prohibited under that system. As with “I'm glad to meet you,” the context of utterance must supply something not supplied by the semantic sentence type to get us from uttering the sentence to expressing a specific proposition. There is nothing visible or audible comparable to the indexical pronouns “I” and “you” to mark this context-dependence, but philosophers speak of hidden indexicality in (4) or (4’).
Contrast this situation with a situation where a religious fundamentalist and a radical feminist say, respectively,
(6a) Abortion is morally prohibited.
(6b) Abortion is morally permissible.
Of course the proponent of (6a) is judging by religious fundamentalist standards, and is aware of doing so, while the proponent of (6b) is judging instead by radical feminist standards, and again is aware of doing so. But each thinks that he or she is doing something more than just judging by certain
moral standards or a certain moral system: Each thinks that his or her moral system is right while the other's is wrong. (Or at least, this is certainly the attitude of the religious fundamentalist and probably the attitude of the radical feminist.)
In this case there seems to be genuine disagreement, not miscommunication, and so it seems there is a single proposition that abortion is morally prohibited that the two are disagreeing about, which the form of words “Abortion is morally prohibited” suffices to express, without any help from any context of utterance. The opposite attitudes of belief and disbelief on the part of the fundamentalist and the feminist towards one and the same proposition may explain their opposite actions, say of demonstrating and counterdemonstrating, on one and the same occasion. But does this proposition have a truth value, or does it exemplify a further type of indeterminacy, beyond failed presupposition and borderline cases?
For moral absolutists, who take there to be a single objectively correct moral standard or system, the proposition expressed by (6a) is either true or false, according as that objectively correct moral standard or system does or does not prohibit abortion. For moral relativists, who hold the opposite view, the proposition expressed by (6a) may exhibit a truth-value gap.
For one version of moral relativism, the error theory, ordinary people today still tend to be implicit moral absolutists, as five hundred years ago ordinary people tended to be implicit geocentrists; but today as five hundred years ago, ordinary people are just making a mistake, since absolute morality is as nonexistent as phlogiston. For the error theorist, moral relativity provides examples of indeterminacy, but examples of a familiar kind, raising no new issues: essentially, cases of presupposition failure.
But many philosophers have shown themselves equally reluctant to endorse moral absolutism or to accuse ordinary people of making some systematic error akin to belief in phlogiston. Consequently a large literature on the supposed tension between deflationism and “nonfactualism,” especially about morality, has developed. So-called truth relativism is one fairly recent and rather influential position on this issue.
Often, as in the legal example, something that can only be supplied by the context of utterance is needed to get us from a semantic sentence token to a proposition. According to “truth relativism,” sometimes something that can only be supplied by a “context of assessment” is needed to get us from a proposition to a truth value. The anthropological or sociological claim
(7) Abortion is prohibited according to religious fundamentalist morality, but not according to radical feminist morality.
should be no more controversial than the claim
(8) Abortion is prohibited under Chilean law, but not under Chinese law.
But “truth relativism” in effect construes the uncontroversial (7) as supporting a more controversial theoretical claim:
(9) The proposition that abortion is morally prohibited is true in a religious fundamentalist context of assessment, but not in a radical feminist context of assessment.
Whether there is any need or use for the apparatus of contexts of assessment is at present still hotly debated, which is to say that the reality of the phenomenon of “truth relativity” is hotly debated. So much so, indeed, that comparatively little attention has been given to the follow-up question whether, if there really is such a phenomenon, it poses a significantly different and perhaps more serious threat to the equivalence principle than do the phenomena of presupposition and vagueness. At present “truth relativism” is another question mark in the cluster of question marks hanging over the center of deflationism (and over the periphery at least of most of its rivals).
4.6 LOCAL VS GLOBAL
The views that truths about law and morality that we have just been discussing, to the effect that they are both in different ways relative, are examples of “local” relativisms, often contrasted with “global” relativism, the view that all truth is relative. Global relativism is sometimes considered another theory of truth of the same sort as the correspondence, coherence, utility, deflationist, and so on, but it does not seem to be a theory that any significant number of analytic philosophers have ever been seriously tempted to adopt. Our brief discussion of it will therefore be something in the nature of a digression.
Global relativism is often claimed to be self-defeating. One argument claims that if the thesis of global relativism were true, then according to itself it would only be true relative to something, so presumably false relative to something else. This objection is not compelling. After all, though all legality is relative, breathing happens to be lawful under every law code (though perhaps heavily taxed under some). Might it not be that though all truth is relative, the thesis that this is so is true relative to every whatever-it-is that truth is supposed to be relative to?
A perhaps better argument is due to Paul Boghossian. Suppose all truths were only true relative to something or other. Then the truth that grass is green, for example, would only be true relative to something. But if it is true that grass is green relative to something, then that too would only be true relative to something: It would only be true relative to something that it is true relative to something that grass is green. But if this is true, then…. This regress shows that the only genuine truth in the vicinity is the “infinite” proposition that we can only express in English using ellipsis: “It is true relative to something that it is true relative to something that…it is true relative to something that grass is green.” Surely the idea that these sorts of propositions are the only truths is absurd.
Boghossian has attempted to parlay this argument against global relativism into a challenge to various local relativisms as well. He assumes that every local relativist owes us an account of the things to which he or she takes truth in the pertinent locality to be relative—let us call them “systems,” as in the legal case, just to have a word for them, without reading too much into the word—and then poses a dilemma. Are these systems themselves true or false absolutely or only relatively? On the one hand, if systems are only relatively true, a regress argument like that against global relativism looms: Each system is true relative only to some system itself true only relative to some system, and so on. And saying that each is true relative to itself only trades regress for circularity, leaving systems equally baseless. On the other hand, if one system is absolutely true and the others absolutely false, then we have after all a notion of absolute truth for the local area where truth was supposed to be merely relative, namely, truth relative to the system that is absolutely true.
As we have crudely formulated it, the argument proves too much: A general refutation of all local relativisms is impossible, since whatever may be the status of moral relativism, legal relativism is a truism. The best response of the local relativist is presumably to claim that systems just aren't the sorts of things that can be evaluated for truth or falsehood. Legal and moral systems, for instance, invite being thought of as systems of imperatives, not declaratives. A charge of “baselessness” may still be raised, but depending on the “locality” in question, the local relativist may not be afraid of it. (Boghossian's special target is local relativism about “epistemic justification,” where a charge of baselessness may be less easy to shrug off than in some other areas.)
Whatever may be the case with other local relativisms, there is one that seems to raise special problems because it threatens to collapse into global relativism. The local relativism in question is relativism about judgments, not of legality or morality, but of truth. Relativism about this kind of judgment is what “truth relativism” would naturally be taken to mean, were that label not already in use for MacFarlane-type views, but the point is that, call it what you will, it threatens to collapse into global relativism: If all the truths about truth are only true relative to something, then it seems all truths are only true relative to something.
For consider an arbitrary truth, say the truth that grass is green. This truth is not about truth, in any straightforward sense, bu
t it seems equivalent to the truth that it is true that grass is green, and this second truth is about truth. If truths about truth are only relatively true, this second truth is only relatively true, and then presumably the first truth, that grass is green, is also only relatively true—though obviously this little argument assumes the equivalence principle, which as we have seen has been subject to various objections. But it is time to draw this digression to a close, and turn to deflationism's inflationist rivals.
CHAPTER FIVE
Realism
EVEN IN TRADITIONAL PHILOSOPHY, the label “realism” had multiple uses whose connections with each other were anything but clear. There was, for instance, realism about universals (properties and relations), as opposed to conceptualism and nominalism, but also realism about material objects, as opposed to idealism. In present-day analytic philosophy the terminological situation is so bad as to have led one important contributor to our subject to say that a philosopher who announces without further explanation that she is a realist has accomplished no more than to clear her throat. With other terms it is reasonable to demand that authors use them in their accepted senses. Since “realism” has no accepted sense, all that can reasonably be demanded is that authors explain the sense in which they propose to use it.
We are not here concerned with realism about universals or about material objects, but only about truth. What we mean by “realism” about truth has been indicated in §1.4: A realist is someone who holds that truth involves an appropriate relation between a truthbearer and some portion(s) or aspect(s) of reality. We will successively consider a variety of theories that are realist in this sense, beginning with correspondence theories and proceeding to others more exotic. We also consider objections different kinds of realists raise against rival views or themselves face.