Maverick Philosopher

Suppose one tries to deny that there are truths. One asserts: there are no truths. But to assert any proposition is to assert it as true. So one who asserts that there are no truths is asserting it as true that there are no truths. The upshot is performative inconsistency: the act or performance of asserting 'contradicts' the content asserted.

Performative inconsistency ought to be distinguished from strictly logical inconsistency. The latter is a relation between propositions, which for present purposes we can think of Frege-style as the senses of indexical-free sentences in the indicative mood. Two propositions, p, q, are logically consistent just in case they can both be true, and logically inconsistent just in case they cannot both be true. Performative inconsistency, however, is not a relation between propositions, but one between a mental-linguistic performance, a speech-act or else an episode of unverbalized thinking, and a proposition. Thus the very act of asserting, or of thinking to oneself, that there are no truths 'contradicts' the propositional content asserted or thought, namely, that there are no truths.

There can be logical inconsistency even if there are no thinkers or speakers. All you need for that are two propositions, one the negation of the other, and the other the negation of the one. But there cannot be performative inconsistency unless there are 'performers,' i.e., thinkers and thoughtful language-users. This fact allows for skepticism about the probative reach of retorsive arguments such as the one given above. One can imagine a skeptic giving the following speech:

Your retorsive argument for the ineluctability of truth rests on the contingent fact that there are thinkers and language-users. Thus at the very most it establishes that truth is a transcendental presupposition of thought, discourse, and inquiry. Your argument from performative inconsistency does not establish that truth exists in itself, apart from beings like us. You have not excluded the possibility that truth depends for its existence on beings like us, and is merely something the presupposing of which we find unavoidable. So if the existence of truth is its existence apart from minds, then you have not established that truth exists.

The skeptic is saying in effect that the following propositions are logically consistent:

1. There is no truth in itself apart from us.

2. It is impossible for us to deny that there is truth.

If (1) and (2) are logically consistent, then (2) does not entail the negation of (1). The skeptic's point is that the ineluctability of truth for us, the unavoidability of our presupposing it, has no tendency to show that there must be truths whether or not we exist. If so, then our retorsive thinking above does not reach reality; at best it reaches a transcendental condition of our thinking about reality.

But are (1) and (2) logically consistent? If they are then it is possible that (1). But if it is possible that (1), then it is possibly true that there is no truth in itself. But this is a logical contradiction. For it cannot be possibly true that there is no truth in itself.

So although it seems that a skeptical wedge can be driven between truth-as-transcendental-presupposition and truth as existent in itself, this is a mere appearance. Truth, absolute and existent in itself, is not only undeniable by us, it is undeniable in itself.

Indeed I will go so far as to say that not even God can deny the existence of truth in itself. Suppose God were to assert

1*. There is no truth in itself apart from Me!

Asserting (1*), God is asserting that it is true that there is no truth apart from him. But then there is a truth apart from him, namely, the truth that there is no truth apart from him. The content of the divine assertion 'contradicts' God's asserting of it. So not even God can deny the existence of truth in itself.

If I am told that the divine aseity cannot allow for God's being subject to anything external to him, not even the aboslute truth in itself, then a classical reply would be that God is the absolute truth in itself. In God, truth-as-transcendetal-presupposition and truth as existent in itself are one and the same. God's inability to deny the truth in itself is then his inability to deny himself.

These speculations of Professor Shanon raise some interesting questions. I take Shanon to be saying that Moses on Sinai (i) really did have an unusual experience, and that therefore there is nothing legendary about the report in Exodus of this experience, but that (ii) this experience was not supernaturally caused, but caused by Moses' ingestion of a psychotropic drug. Thus God did not reveal the Torah to Moses on Sinai; Moses had a drug-induced nonveridical experience of God revealing the Torah to him, or, to be maximally punctilious, a nonveridical experience as of God revealing the Torah to him.

One question concerns the validity of the inference from

1. Subject S under the influence of drug D experiences that p

to

2. S's experience that p is nonveridical.

Surely this is a non sequitur. Right now, under the influence of caffeine, I note that my coffee cup is empty. This is consistent with the perceptual experience of the cup's being empty being veridical, which it is. So from the mere fact that a subject is 'on drugs,' it does not follow that that any of the subject's experiences are nonveridical. Now caffeine is a very mild drug. But suppose I was I was on a combination of caffeine, nicotine, marijuana, and methampehtamine. Even then one could not infer that the perception in question was nonveridical. Even on a dose of LSD-25 most of one's perceptual experiences remain veridical. In the case of Moses, from the fact, if it is a fact, that he was under the influence of a psychotropic drug while on Sinai, it does not follow that his experience of being addressed by God and being given the Decalogue was nonveridical.

Of course, if you begin by assuming that there is no God, then it is plausible to explain the Sinai experience by saying that it was drug-induced. But that simply begs the question against the theist. The crucial point is that a subject's being on drugs is logically consistent with the veridicality of his experiences; therefore, one cannot infer from the fact, if it is a fact, that Moses was under the influence of a psychedelic or psychotropic drug that his experince was nonveridical.

So even if Shanon could prove that Moses and the people around him were under the influence of powerful drugs, that would not support his contention that nothing supernatural occurred on Sinai. It would not because it is consistent with theism. How does Shanon know that the drugs Moses supposedly took did not open "the doors of perception" (in Aldous Huxley's phrase) allowing him access to the transcendent, as opposed to shutting him up among figments of his own imagination?

There is a second question whose full discussion should be reserved for a subsequent post. It is clear that mathematical and other truths can be grasped whether one is awake or dreaming, sober or drunk, on drugs or not. If a valid proof occurs to a mathematician in a dream, it is no less valid because the mathematician is dreaming. Why should not the same hold for moral truths? If it is true that it is morally obligatory not to kill human beings, then the experiencing of this truth is veridical whether or not the subject is awake or sleeping, sober or drunk, on drugs or not. So even if it could be proven that Moses was under the influence of powerful drugs on Sinai, what relevance would that have? At the very most it might cast doubt on the veridicality of Moses' perception of God, but not on the veridicality of his experiencing of the content of the Decalogue. If it is true that one ought not kill, then it is true whether or not God exists. And if it is true that one ought not kill, then one's intuiting that it is true is veridical whether one is awake of dreaming, sober or drunk, on drugs or not, or a brain in a vat as opposed to a brain in a skull.

On a standard analysis of 'knows,' where propositional knowledge is at issue, subject S knows that p just in case (i) S believes that p; (ii) S is justified in believing that p; and (iii) p is true. This piece of epistemological boilerplate is the starting-point for much of the arcana (Gettier counterexamples, etc.) of contemporary epistemology. But its pedigree is ancient, to be found in Plato's Theaetetus.

It is obvious that on the standard analysis mere belief is inferior to knowledge since if I believe what is false I don't have knowledge, and if I believe what is true without justification I don't have knowledge either. How then can mere belief be a form or type of knowledge? It is rather a necessary but not sufficient condition of knowledge. Or so it seems to the modern mind.

Another puzzle has to do with certainty. Whether or not knowledge entails certainty, it seems to the modern mind that belief definitely does not entail certainty: what I believe but do not know I cannot be certain about since if I believe but do not know, then either truth is lacking or justification is lacking or both. How then can mere belief be said to be certain? And yet we read in Aquinas that "It is part of the concept of belief itself that man is certain of that in which he believes." (Quoted from Pieper, Belief and Faith, p. 15).

Is easy to understand how one who believes but does not know that p can be subjectively certain that p; but it is difficult to understand how such a person can be objectively certain that p. Objective certainty, however, alone has epistemic value.

We now turn to the remarkable Edith Stein (1891-1942), brilliant Jewish student of and assistant to Edmund Husserl, philosopher, Roman Catholic convert, Carmelite nun, victim of the Holocaust at Auschwitz, and saint of the Roman Catholic church. In the 1920s Stein composed an imaginary dialogue between her two philosophical masters, Husserl and Aquinas. Part of what she has them discussing is the nature of faith.

One issue is whether faith gives us access to truth. Stein has Thomas say:

. . . faith is a way to truth. Indeed, in the first place it is a way to truths — plural — which would otherwise be closed to us, and in the second place it is the surest way to truth. For there is no greater certainty than that of faith . . . . (Edith Stein, Knowledge and Faith, tr. Redmond, ICS Publications 2000, pp. 16-17)

Now comes an important question. What is it that we as philosophers want? We want the ultimate truths about the ultimate matters. If so, we should take these truths from whatever source offers them to us even if the source is not narrowly philosophical. We should not say: I will accept only those truths that can be certified by (natural) reason, but rather all truths whether certified by reason or 'certified' by faith. Thus Stein has Aquinas say:

If faith makes accessible truths unattainable by any other means, philosophy, for one thing, cannot forego them without renouncing its universal claim to truth. [. . .] One consequence, then, is a material dependence of philosophy on faith.

Then too, if faith affords the highest certainty attainable by the human mind, and if philosophy claims to bestow the highest certainty, then philosophy must make the certainty of faith its own. It does so first by absorbing the truths of faith, and further by using them as the final criterion by which to gauge all other truths. Hence, a second consequence is a formal dependence of philosophy on faith. (17-18)

But of course this cannot go unchallenged by Husserl. So Stein has him say:

. . . if faith is the final criterion of all other truth, what is the criterion of faith itself? What guarantees that the certainty of my faith is genuine? (20)

Or in terms of of the distinction made above between subjective and objective certainty: what guarantees that the certainty of faith is objective and not merely subjective? The faiths of Jew, Christian, and Muslim are all different. How can the Christian be sure that the revelation he takes on faith has not been superseded by the revelation the Muslim takes on faith?

Stein's Thomas replies to Husserl as follows:

Probably my best answer is that faith is its own guarantee. I could also say that God, who has given us the revelation, vouches for its truth. But this would only be the other side of the same coin. For if we took the two as separate facts, we would fall into a circulus vitiosus [vicious circle], since God is after all what we become certain about in faith. [. . .]

All we can do is point out that for the believer such is the certainty of faith that it relativizes all other certainty, and that he can but give up any supposed knowledge which contradicts his faith. The unique certitude of faith is a gift of grace. It is up to the understanding and will to draw the practical conseuqences therefrom. Constructing a philosophy on faith belongs to the theoretical consequences. (20-22)

So there you have it. There are two opposing conceptions of philosophy, one based on the autonomy of reason, the other willing to sacrifice the autonomy of reason for the sake of truths which cannot be certified by reason but which are provided by faith in revelation. It looks as if one must simply decide which of these two conceptions to adopt, and that the decision cannot be justified by (natural) reason.

My task, in this and in related posts, is first and foremost to set forth the problem as clearly as I can. Anyone who thinks this problem has an easy solution does not understand it.

And so it ought to come as no surprise that skepticism can aid the cause of fideism, which in turn grants carte blanche to dogmatism and fanaticism. Skeptical reason, having proven its own powerlessness, is powerless to prevent nonrational fixation of belief. In other words, why, if reason is impotent, accept (A) over (B)?

A: Because reason is impotent to fix belief, and the arguments for any given position are not compelling, one must respect and tolerate competing positions.

B: Because reason is impotent to fix belief, and the arguments for any given position are not compelling, one must simply believe and obey and oppose the infidels who refuse to believe and obey.

Simply put, my point is that skepticism can lead to intolerance as easily as it can lead to tolerance.

On a draconian understanding of sufficient evidence, sufficient evidence for a proposition p is a distinct proposition q (perhaps conjunctive)such that q entails p, where q entails p if and only if there is no possible world in which q is true and p is false. (Entailment is thus the necessitation of material implication.) Correspondingly, insufficient evidence for a proposition p is a proposition q such that q does not entail p. I take it that this is not what is meant by sufficient and insufficient evidence in discussions of the ethics of belief.

On a liberal understanding of sufficient evidence, sufficient evidence for a proposition p is evidence that is enough to tip the balance in favor of p when p is compared to its negation ~p. For example, Capital punishment deters crime versus Capital punishment does not deter crime. Suppose the evidence for each member of this contradictory pair is equal. Then, on the liberal understanding, one who believes either proposition believes on insufficient evidence. For he believes on the basis of evidence that does not suffice to make one member of the pair of contradictories more credible than the other.

An ethics of belief premised on Cliffordian evidentialism would then require that one suspend judgment, i.e., take no position on the question whether capital punishment is an effective deterrent. For if one assiduously apportions belief to evidence, then in a case where the evidence is equal on both sides of the issue, the apportioning would not result in affirming one side over the other.

Question One: Is that (the liberal understanding) what we ought to mean by 'insufficient evidence' in discussions of the ethics of belief?

Question Two: If that is what we mean, ought we accept Clifford's view that it is wrong to believe on insufficient evidence, and that in ebvery such case we ought to suspend belief?

Phil Philologos commented:

The concept of “insufficient evidence,” if I’m not mistaken, is a borrowing from legal theory. For an offense to be regarded as proven, statutory law lays down a set of material elements that be must be proven by the plaintiff or prosecution, proven either beyond a reasonable doubt or by a preponderance of the evidence, depending upon the nature of the offence. There is also a legal standard of evidence for bringing an indictment. Something like, does the evidence taken as a whole cause a reasonable person to believe an offence may have been committed? A finding of “insufficient evidence” in the latter case mandates no indictment, in the former case, acquittal of the charges.

The first point is, a finding of insufficient evidence in the criminal law does not lead to any suspension of belief, but to a directed verdict of no charge or acquittal, the burden of proof not having been met.

The second and more relevant point for the “evidential thesis” is that we lack a sense of what it means to say “it is wrong to believe anything on insufficient evidence” where no Rules of Evidence lay down what is a standard of proof, and who has the burden of proof, and what material elements must be proven. For this reason “There is insufficient evidence to believe in God’s existence” strikes me utter nonsense.

Phil's first point seems clearly correct. But the ethics of belief is not like the positive law. The latter is posited and determinate. But the ethics of belief is little more than a philosophical project that some of us would like to achieve and others of us will reject as absurd (perhaps because it presupposes the truth of limited doxastic voluntarism which they reject). What we ought to mean in the ethics of belief by 'insufficient evidence' need have little to do with what the positive law, civil or criminal, means by it. The purposes are different, after all.

As for Phil's second point, I agree that it doesn't mean much of anything to say 'There is insufficient evidence to believe in God's existence' unless 'insufficient evidence' is explained and probative standards laid down. But these are for philosophers to work out and cannot be modelled on the positive law.

As I understand it, in the criminal law, the probative standard for guilt is 'beyond a reasonable doubt,' while in civil cases the standard is less demanding: 'preponderance of the evidence.' Now in terms of the ethics of belief it is permissible for me to believe that a man is a murderer if the preponderance of the evidence points in that direction; but if I were a juror in his trial, then it would not be permissible for me to pronounce him guilty merely on the preponderance of evidence.

In Giving Dualism Its Due, we read:

Being a philosopher, of course I would like to think that my [materialist]stance is rational, held not just instinctively and scientistically and in the mainstream but because the arguments do indeed favor materialism over dualism. But I do not think that, though I used to. My position may be rational, broadly speaking, but not because the arguments favor it: Though the arguments for dualism do (indeed) fail, so do the arguments for materialism. And the standard objections to dualism are not very convincing; if one really manages to be a dualist in the first place, one should not be much impressed by them. My purpose in this paper is to hold my own feet to the fire and admit that I do not proportion my belief to the evidence.

In sum:

1. The arguments for dualism and the arguments for materialism both fail.
2. The standard objections to dualism are not very convincing.
3. It is rational to be a materialist.

In my opinion (1)-(3) is a consistent triad. If so, what does 'rational' mean? It cannot have the Cliffordian meaning according to which one apportions one's belief to the evidence. For that would require suspension of belief on the issues that divide dualists and materialists given the truth of (1) and (2). But Lycan does not suspend belief; he remains a committed materialist. He believes beyond the evidence in that he believes on insufficient evidence. The evidence is insufficient because it is counterbalanced by the evidence for the position he disbelieves. However we define 'insufficient evidence,' it seems clear that if the evidence for p and the evidence for ~p are equal, then the evidence for either is insufficient.

Lycan's is an interesting case because it doesn't display all of the Jamesian marks. The issue is live for Lycan and for the people here present, but is it forced and momentous? An issue is forced in the sense of William James if it is such that one's remaining theoretically agnostic about it is tantamount to deciding it in a particular way. James gives the example of a man who hesitates to get married. "It is as if a man should hesitate indefinitely to ask a certain woman to marry him because he was not perfectly sure that she would prove an angel after he brought her home. Would he not cut himself off from that particular angel-possibility as decisively as if he went and married someone else?" (Will to Believe, p. 26) The man who refuses to commit himself to marriage commits himself to bachelorhood nolens volens.

But surely dualism versus materialism is not a forced option in the Jamesian sense. For one thing, one might reject both in the manner of the idealist. The positions are not logical contradictories of each other but logical contraries: they can't both be true, but they can both be false. Second, it is not the case that a suspension of judgment is tantamount to an opting for one side. If you take no position on dualism versus materialism, how does that commit you to one side or the other? On the God question, if one takes no position on whether or not God exists, then it it strongly arguable that one is a practical atheist: the agnostic lives as if God does not exist. And similarly for the immortality of the soul: to take no position is to live as if the soul is mortal. Or at least this is plausibly arguable. But the dualist need not be a substance dualist, and if he is not a substance dualist, then it is very difficult to see how the dualism versus materialism option is forced. And even if the dualist is a substance dualist, one might be a substance dualist without being committed to the immortality of the soul or mind.

A momentous option is one in which "We are supposed to gain, even now, by our belief, and lose by our nonbelief a certain vital good." (WB, 26) But I think it would be a stretch to think that the rather technical and abstruse issues that divide materialists and dualists are momentous in James' sense.

All this notwithstanding, the Lycan quotation above illustrates how rationality needn't require apportioning one's belief to the evidence. Or will you argue that Lycan is irrational in remaining a materialist despite his newfound insight that the arguments for it are not compelling?

Let's consider an example simpler than the one I gave yesterday. (This simplification may allow me to sidestep Phil's objection.) A man is dying of thirst in a desert. He knows that some of the water sources in the vicinity are lethally poisonous and some are potable, but he doesn't know which are which. He comes upon a water source, but the evidence available to him is insufficient to make This source is potable more credible than This source is poisonous. Assume that both of these propositions are equally credible, equally worthy of belief, given the scant evidence at the man's disposal, his weak state, his inability to perform tests, etc.

What is the rational course of action for this man dying of thirst? Clearly, to drink is rational and not to drink is irrational. For if he does not drink, the probability of his dying is 1, whereas if he drinks the probability of his dying is <1. In plain English, if our man does not drink, he will certainly die; but if he drinks there is some chance that he will survive. It is therefore irrational for him not to drink since it is irrational for a man who wants to live to refuse to perform a simple action that has a chance of saving his life.

So far I have been addressing the rationality of action, but what about the rationality of belief? The thesis that the man acts rationally by drinking I take to have been established by the example. But it it a less obvious thesis that I am concerned to defend, namely, that there are cases in which one believes rationally even if one's believing goes beyond the evidence. In terms of our example, what I want to say is that the man dying of thirst believes rationally in believing that the water source he has come upon is potable even though the evidence at his disposal is insufficient to show that potability is more probable than the opposite.

One way to rebut this is by arguing that the man ought to suspend belief concerning the potability of the source but drink from it anyway. On this approach the man would be irrational in believing the source to be potable, but rational in drinking from it. Now imagine the following dialogue transpiring between the man (A) and a philosopher (B) who happens upon the scene:

A. I take no position on the question whether this water is potable but I will drink it anyway.

B. So you will drink the water for no reason? Is it not irrational to do something for no reason?

A. Not for no reason: I want to live and it is more likely that I will live if I drink than if I refrain from drinking.

B. So you believe that there is some nonnegative likelihood that the water is potable?

A. Yes.

B. But then you are not just acting, you are acting on a belief, a belief for which you have no sufficient evidence! For all you know, the water is poisonous! If you go by evidence alone, you are not justified in making any affirmation with respect to the water's potability. So if you suspend belief, then you cannot act rationally. For if you suspend belief with respect to the water's potability, then you refuse to affirm any of the following: The water is potable, The water is poisonous, The water is more likely to be potable than poisonous, etc. Therefore, if you just drink the water, then you act for no reason at all, which is to say: you act irrationally. But I thought you agreed that to drink the water is the rational thing to do. Therefore your acting must have a reason and that reason is your belief that the water is potable.

A. And what does that mean?

B. That means that your believing that the water is potable is as rational as your acting to drink the water, and this despite the fact that your believing goes beyond the evidence.

W. K. Clifford is famous for his evidentialist thesis that "It is wrong always, everywhere, and for anyone, to believe anything on insufficient evidence." On this way of thinking, someone who fails to apportion belief to evidence violates the ethics of belief, and thereby does something morally wrong. Although Clifford had religious beliefs in his sights, his thesis, by its very wording, applies to every sort of belief, including political beliefs and the belief expressed in the Clifford sentence lately quoted!

Consider the case of a man dying of thirst in a desert. He comes upon two water sources. He knows (never mind how) that one is potable while the other is poisonous. But he does not know which is which, and he has no way of finding out. Should the man suspend belief, even unto death, since he has insufficient evidence for deciding between the two water sources?

On one way of looking at the matter, suspension of belief would clearly be the height of irrationality. The desert wanderer must simply drink from one of the sources and hope for the best. Clearly, by drinking from one (but not both) of the sources, his chances of survival are one half, while his chances of survival from drinking from neither are precisely zero. By making a leap of faith, he maximizes his chances of reality-contact, and thereby his chances of survival. Surely a man who wants to live is irrational if he fails to perform a simple action that will give him a 50-50 chance of living when the al;ternative is certain death.

Cases like this are clear counterexamples to theories of rationality according to which rationality requires always apportioning belief to evidence. In the above case the evidence is the same for either belief and yet it would be irrational to suspend belief. Therefore, rationality for a human agent (as opposed to rationality for a disembodied transcendental spectator) cannot require the apportioning of belief to evidence in all cases, as Clifford demands. There are situations in which one must decide what to believe. Will I believe that source A is potable? Or will I believe that source B is potable? In Jamesian terms, the option is live, forced, and momentous. (It is not like the question whether the number of ultimate particles in the universe is odd or even.) An adequate theory of rationality, it would seem, must allow for believing beyond the evidence. It must return the verdict that in some cases, to refuse to believe beyond the evidence is positively irrational.

Let us consider another such case. What evidence do I have that my new neighbors are decent people? Since they have just moved in, my evidence base is exiguous indeed and far from sufficient to establish that they are decent people. (Assume that some precisifying deinition of 'decent' is on the table.) Should I suspend judgment and behave in a cold, skeptical, stand-offish way toward them? ("Prove that you are not a scumbag, and then I'll talk to you.") Should I demand of them 'credentials' and letters of recommendation before having anything to do with them? Either of these approaches would be irrational. A rational being wants good relations with those with whom he must live in close proximity. Wanting good relations, he must choose means that are conducive to that end. Knowing something about human nature, he knows that 'giving the benefit of the doubt' is the wise course when it comes to establishing relations with other people. One must assume the best about others at the outset and adjust downwards only later and on the basis of evidence to the contrary. But note that my initial belief that my neighbors are decent people -- a belief that I must have if I am to act neighborly toward them -- is not warranted by anything tnat could be called sufficient evidence. Holding that belief, I believe way beyond the evidence. And yet that is the rational course.

So again we see that in some cases, to refuse to believe beyond the evidence is positively irrational. A theory of rationality adequate for the kind of beings we are cannot require that belief be always and everywhere apportioned to evidence.

Various objections and questions can be raised with respect to the foregoing. They will be explored in further posts.

1. For Hume, a belief is an idea to which we assent. Hume is thinking of occurrent beliefs, not latent or dispositional beliefs. To believe that God exists, for example, is to assent to the idea (concept) of God. Thus as I read Hume, believing is not a propositional attitude in Russell's sense: the believer in believing is not related to a proposition but to a non-propositional object, an idea. In every occurrent belief one must distinguish the act of believing from the content believed; but it is not obvious that this content must be a proposition.

In the case of singular existential beliefs or judgments (that God exists, that Vulcan does not exist, etc.) it is very tempting to say that the belief-content is a non-propositional object. For if the object of the existential belief were propositional, then it would have to have the subject-predicate logical form a exists when it is not at all clear that 'exists' is a first-level predicate or existence a first-level property. As Hume points out, employing one of his signature 'riffs,' "The idea of existence is nothing different from the idea of any object. . . ." (Selby-Bigge, 94) If I conceive of God as existent, I "in reality make no addition to or alteration on our first idea."

To conceive of X and to conceive of X as existent are the same: existence makes no addition to X's concept. This is an important point and seems wholly correct, and it doesn't matter what 'X' denotes. Consider the unique X so defined that X's existence is included in its concept. There is no incoherence in maintaining that this object does not exist. Or compare the golden mountain and the existent golden mountain. Neither exist. 'The existent golden mountain does not exist' is not a contradiction in the way 'The snow-covered golden mountain is not snow-covered' is a contradiction. And yet there is a huge difference between a golden mountain's existing and its not existing. One should conclude from this, not that there is 'no such thing' as existence, but that existence is not part of anything's concept. If you were to tell me that existence is part of the concept existent golden mountain, then I would reply that the existence that is includable in concepts is not the genuine article; it is not the existence we mean when we say that such-and-such exists. Nothing seems clearer than that existence, the genuine article! (stamp the foot, pound the table, release the safety) is not includable in anything's concept. Surely Hume and Kant and Brentano and Aquinas and Santayana and a host of others are right about this.

But if so, if, to put the point in an alternative way, existence is not a first-level property, then there is no atomic proposition God exists or atomic state of affairs God's existence, and to believe that God exists is not to stand in the belief relation to a proposition.

2. So what are we doing when we make existential judgments?

(A) When I think of God, think of him as existent, or believe him to be existent, "my idea of him neither encreases [sic] nor diminishes." (94) But (B) "there is a great difference betwixt the simple conception of the existence of an object, and the belief of it. . . ." And because (C) ". . . this difference lies not in the parts or the composition of the idea, which we conceive . . ." it follows that (D) ". . . it must lie in the manner, in which we conceive it." (Selby-Bigge 94-95)

To believe that Peter exists is not to believe that Peter has a certain property, the property of existence: there is no such property. With a sidelong glance in the direction of Brentano we might say that to believe that Peter exists is to accept Peter, and to believe that he does not exist is to reject Peter. Positive belief (acceptance) and negative belief (rejection) are mental acts external to the belief-content. The belief-content is non-propositional. Brentano thinks that this is the case for all contents including those that are not obviously existential. For Brentano, to believe that Peter is sick is to accept sick-Peter, and to believe that Peter is not sich is to reject sick-Peter. The object or accusative of the act of believing is structured, but not propositionally: the accusative is sick-Peter, not Peter is sick.

This is not quite what Hume means, but I would say that Hume in on the way to the Brentano view in that Hume holds that the contents of belief (in some cases at any rate) are non-propositional. Unlike Brentano, Hume does not see judging as a category of act distinct from conceiving, but as a conceiving with increased "force and vivacity":

So that as belief does nothing but vary the manner, in which we conceive any object, it can only bestow on our ideas an additional force and vivacity. An opinion, therefore, or a belief may be most accurately defin'd, A LIVELY IDEA RELATED TO OR ASSOCIATED WITH A PRESENT IMPRESSION." (Selby-Bigge, p. 96, emphasis in original)

There is much that is murky here. For example, how can a belief in God be a lively idea related to an impression when there is no impression of God?

Thus it is quite natural to suppose that such deontological predicates as 'permissible,' impermissible,' and 'obligatory' are legitimately applied to acts of belief. (An act is an occurrent mental state.)

But such predicates are applicable only if beliefs and cognate attitudes are under voluntary control. The point could be put as follows: Epistemic deontology is a viable project only if doxastic involuntarism is false.

I am coming to the view that doxastic involuntarism is false (thereby joining the excellent company of Augustine, Aquinas, Descartes and other great philosophers) and that at least some beliefs are under voluntary control. But of course the logically prior question is: What is belief? Whether belief is under voluntary control depends on what belief is.

One cannot understand belief except in relation to other mental states. So let's consider how believing and knowing are related. They are obviously different, and yet they share a common element. Suppose we say that what is common to S's knowing that p and S's believing that p is S's acceptance of p. I cannot (occurrently) believe that Oswald acted alone unless I accept the proposition that Oswald acted alone, and I cannot (occurrently) know that he acted alone with accepting the very same proposition. To accept, of course, is to accept-as-true. It is equally obvious that what is accepted-as-true might not be true. Those who accept that the earth is flat accept-as-true what is false. Now one could analyze 'S knows that p' as follows:

a) S unconditionally accepts-as-true p
b) p is true
c) S is justified in accepting-as-true p.

This is modelled on, but diverges from, the standard justified-true-belief (JTB) analysis of 'know' the locus classicus of which is Plato's Theaetetus.

And one could perhaps analyze 'S believes that p' as follows:

a) S unconditionally accepts-as-true p
d) S does not know that p.

These analyses accommodate the fact that there is something common to believing and knowing, but without identifying this common factor as belief. The common factor is acceptance. A reason for not identifying the common element as belief is that, in ordinary language, knowledge excludes belief. Thus if I ask you whether you believe that p, you might respond, 'I don't believe it, I know it!' Do I believe the sun is shining? No, I know the sun is shining. Do I know that I will be alive tomorrow? No, but I believe it. That is, I give my firm intellectual assent to the proposition despite its not being evident to me. Roughly, belief is firm intellectual assent in the absence of compelling evidence.

Surely this is what we mean by belief in those cases that clearly count as belief. Lenny the liberal, for example, believes that anthropogenic global warming is taking place and is a dire environmental threat. Lenny doesn't know these two putative facts; he believes them: he unconditionally accepts, he firmly assents to, the two propositions in the absence of compelling evidence. And it seems clear that an element of will is involved in our boy's belief since the evidence does not compel his intellectual assent. He decides to believe what he believes. His believing is in the control of his will. This does not mean that he can believe anything he wants to believe. It means that a 'voluntative surplus' must be superadded to his evidence to bring about the formation of his belief. Without the voluntative superaddition, he would simply sit staring at his evidence, so to speak. There would be no belief and no impetus to action. Beliefs typically spill over into actions. But there would not be even a potential 'spill over' unless there were a decision on Lenny's part to go beyond his evidence by superadding to it his firm intellectual assent.

"But aren't you just using 'believes' in an idiosyncratic way?" It is arguably the other way around. Someone who says he believes that the sun is shining when he sees that it is shining is using 'believes' in an idiosyncratic way.

"But then isn't this just a terminological quibble? You want to substitute 'accepts' or 'accepts-as-true' for 'believes' in the standard JTB analysis of 'knows' and you want to reserve 'believes' for those cases in which there is unconditional acceptance but not knowledge." The question is not merely terminological. There is an occurrent mental state in which one accepts unconditionally propositions that are not evident. It doesn't matter whether we call this 'belief' or something else. But calling it 'belief' comports well with ordinary language.

Let me now elaborate upon this account of belief, or, if you insist, of Aquinian-Pieperian belief.

1. Belief is a form of acceptance or intellectual assent. To believe that p is to accept p, and to disbelieve that p is to reject p. One may also do neither by withholding p.

2. If acceptance is the genus, then knowing, believing, and supposing are species thereof. In knowing and believing the acceptance is unconditional whereas in supposing it is conditional. It follows that believing is not common to believing and knowing as on the JTB analysis. To think otherwise is to confuse the genus (acceptance) with one of its species (belief).

3. What distinguishes believing and knowing is that the believer qua believer does not know, and the knower qua knower does not believe. Both, however, accept. But here I must confess to a difficulty. What I just wrote appears objectionably circular. It seems to boil down to this: what distinguishes believing and knowing is that they are distinct!

If believing and knowing are species of the genus acceptance, what is the specific difference whereby the one is distinguished from the other? Believing that p and knowing that p are not distingusihed by the common propositional content, p. Nor are they distinguished by their both being modes of unconditional acceptance. Can we say that they differ in that the evidence is compelling in the case of knowing but less than compelling in the case of believing? That is true, but then the difference would seem to be one of degree and not of kind. But if knowing and believing are two species of the same genus, then we have a difference in kind. Perhaps the difference is that belief is based on another's testimony whereas knowledge is not.

4. It is essential to belief that it involve both a proposition (the content believed) and a person, the one whose testimony one trusts when one gains access to the truth via belief. To believe is to unconditionally accept a proposition on the basis of testimony. If so, then there are two reasons why it makes no sense to speak of perceptual beliefs. First, what I sense-perceive to be the case, I know to be the case, and therefore, by #3 above, I do not believe to be the case. Second, what I sense-perceive to be the case I know directly without need of testimony.

On this approach, the difference between believing and knowing is that believing is based on testimony whereas knowing is not. Suppose p is true and that my access to p's truth is via the testimony of a credible witness W. Then I have belief but not knowledge. W, we may assume, knows whereof he speaks. For example, he saw Jones stab Smith. W has knowledge but not belief.

1. There is this much agreement: both of us are doxastic voluntarists. We agree that it in our power to believe G, disbelieve G, or 'practice epoche' with respect to G, i.e., suspend judgment as to the truth or falsity of G.

2. Thus we both endorse the possibility of an ethics of belief, though we differ as to its content, as to what it prescribes, proscribes, and permits. My view is that in the O.J. case, given that prob(G/E) >> prob(~G/E), it is permissible that a juror believe that G. (Let us set aside the question whether it is also obligatory. Everything obligatory is permissible, but not conversely.) I take TV man's view to be that it is impermissible for a juror to believe that G. (If it is permissible (impermissible) for a juror to believe that G, then a fortiori it is permissible (impermissible) for a non-juror to believe that G assuming the non-juror has access to the same evidence.)

3. Consider the following two immediate inferences:

A. Prob(G/E) >> prob (~G/E)
-----
B. It is permissible that one believe that G.

A. Prob(G/E) >> prob (~G/E) -----
C. It is impermissible that one believe that G.

If the first argument is invalid, then so is the second. So if TV man is accusing me of reasoning incorrectly, then I can easily return the compliment. I would say that both arguments are invalid — how does one extract a deontic conclusion from a non-deontic premise? — and that the conditionals corresponding to the two arguments are synthetic, i.e., neither narrowly logical nor analytic. In other words, I am not saying that (B) follows from (A), I am saying that If (A), then (B) is a synthetic necessary truth. Correspondingly, TV man can be interpreted as saying that, even though the probability that O. J. killed his ex-wife, given the available evidence, is much greater than the probability that he did not given the same evidence, it is nevertheless impermissible to believe that O.J. killed his wife, and that this is a necessary synthetic truth.

4. TV man's justification for his conditional is that we could be wrong about O.J. After all, it was a circumstantial case. There were no eye-witnesses. Thus TV man invokes the legitimate epistemic desideratum, Avoid error! But there is a competing desideratum, namely, Acquire truth! In the case before us, the truth is not merely theoretical but has important practical consequences. I don't think that there can be any reasonable dispute that in a case like this it is very important that we get to the truth even if our means of getting there is not absolutely sure.

5. If I am a juror, and so must take a position, then I must avoid convicting an innocent man, but I must also avoid acquitting a guilty man. Either way the consequences are momentous. If convicted, imprisonment or execution. If acquitted, an injury to the social fabric in the form of loss of respect for the rule of law, not to mention the injustice to the person murdered and those who care about her.

6. It seems to me obvious that a juror should vote to convict, and thus that it is permissible to believe that O.J. is gulty as charged. For the evidence stacked against him is overwhelmingly strong. (See Part I, Part II and Part III of an interview with Vincent Bugliosi.) The fact that the evidence does not raise the probability of his guilt to 1 (= objective certainty) is not a good reason not to convict. TV Man seems to ignore completely the negative consequences (the damage to society, the undermining of the rule of law, the setting of a terrible precedent, the allowing of race to be injected into any case regardless of its relevance, etc.) of freeing a man who is guilty beyond a reasonable doubt.

Indeed, if you think about it, TV man's position reduces itself to absurdity. For if the probative standard is set so high that "What if we are wrong?" is allowed to trump evidence as detailed and massive as that brought against Simpson, then the criminal justice system would have to be dismantled.

7. But what about the NT command: Judge not lest ye be judged? Such an invocation is a clear example of the misuse of Scripture as I explain in Immigration, the Catholic Bishops, and the Misuse of Scripture.

The absence of any reference to belief in my statements on knowledge and knowing will be immediately noticed. The absence is intended, and though rare today in discussions of knowledge it is by no means unique in the history of the theory of knowledge.14 Even such a resolute Naturalist as Roy Wood Sellars specifies the nature of knowledge without reference to belief.15

Belief I understand to be some degree of readiness to act as if such and such (the content believed) were the case. Everyone concedes that one can believe where one does not know. But it is now widely assumed that you cannot know what you do not believe. Hence the well known analysis of knowledge as "justified, true belief." But this seems to me, as it has to numerous others, to be a mistake. Belief is, as Hume correctly held, a passion. It is something that happens to us. Thought, observation and testing, even knowledge itself, can be sources of belief, and indeed should be. But one may actually know (dispositionally, occurrently) without believing what one knows.

Whether or not one believes what one represents truly and has an appropriate basis for so representing, depends on factors that are irrelevant to truth, understanding and evidence. It depends, one might simply say, on how rational one is. Now I do not think that this point about belief in relation to knowledge is essential to the rest of this paper, but I mention it to indicate that the absence of any reference to belief in my general description of knowledge is not an oversight. Belief is not, I think, a necessary component of knowledge, though one would like to believe that knowledge would have some influence upon belief, and no doubt it often does.

In addition, it seems to me that specification of knowledge in terms of belief is a harmfully tendentious characterization, favoring the naturalization of knowledge. This is because belief has an essential tie to action, and is therefore easily located in the natural world--say as a mere tendency of the physical organism to behave in certain ways. I suspect that it is the almost overwhelming Empiricist--and in that sense Naturalist--tendency of thought in our time that has created the general presumption that knowledge must be some kind of belief. Hence we must here at least question that presumption; and, I believe, when questioned it will not prove to be obvious or, finally, sustainable.

I am not sure how deep the Pieper-Willard similarity on this topic runs, but it is worth exploring. One caveat: Willard above agrees with Hume that belief is a passion, "something that happens to us." This does not appear to comport with Pieper's doxastic voluntarism.

I am also puzzled by this sentence of Willard's: "But one may actually know (dispositionally, occurrently) without believing what one knows." The dispositional and the occurrent are opposites. Why does Willard bracket them together? Furthermore, if I actually know something, if my knowing is 'in act' as a Scholastic might put it, then my knowing is occurrent, and there is nothing potential or dispositional about it. Finally, what would be an example of Willard's point?

2. It makes sense to apply deontological predicates to actions because they are under voluntary control. Thus it makes sense to say that one ought to feed one's children because (apart from unusual circumstances) it is within one's power to feed one's children. So if it makes sense to apply deontological predicates to beliefs and related propositional attitudes, then they too must be under voluntary control.

3. This brings us to the issue of doxastic voluntarism: Are any of our believings under our direct voluntary control as regards their coming into existence? To introduce some terminology of my own:

Extreme doxastic voluntarism: ALL beliefs are such that their formation is under one's direct voluntary control.
Limited doxastic voluntarism: There are only SOME beliefs over the formation of which one has direct voluntary control.
Doxastic involuntarism: There are are NO beliefs over the formation of which one has direct voluntary control.

Note that the issue concerns the formation of beliefs, not their maintenance, and note the contrast between direct and indirect formation of beliefs. Roughly, I form a belief directly by just forming it, not by doing something else as a means to forming it.

4. Alston's project in the pages I am covering is, first, to show that we do not have "effective voluntary control over believings." (p. 60), and second, to infer that the deontological triad "required, permitted, or forbidden" has no application to believings and such related propositional attitudes as withholdings. Alston, then, is a doxastic involuntarist.

Alston's thesis, to be precise, is not just that we do not, but that we cannot, have direct voluntary control over belief formation. He means the 'cannot' psychologically, not logically. His point, then, is that it is psychologically impossible to believe at will.

You can see why his book is entitled Beyond "Justification." Alston doesn't think that there is any place in epistemology for justification if this term is taken deontologically. The sneer quotes suggest that there is no such thing as epistemic justification. Thus, if I understand Alston, it makes no sense on his view to say that a belief or a believing is justified since that is tantamount to saying that it is permissible. But no believing is permissible or impermissible since no believing is under the direct control of the will. Therefore, for Alston, justification interpreted deontologically has no place in epistemology. There is no such thing as epistemic justification.

5. Alston's examples in support of his thesis fall into three categories.

a) One cannot believe at will what is obviously false. It is obviously false that the Romans continue to exercise hegemony over Western Europe, and it is hard to disagree with Alston when he concludes that no one has the power to believe this falsehood at will, just by deciding to believe it. (p. 63)

b) One cannot not believe what is obviously the case. It is obviously the case that this thing in front of me is a computer monitor. Can I disbelieve this perceptual deliverance? No. Seeing is believing. It is a more subtle question whether I can suspend judgment in the manner of Husserl's phenomenological epoche. But this is a topic for a separate post. For now I am happy to concede to Alston that one cannot disbelieve at will what is obviously the case.

c) The matter becomes much more difficult when we turn to propositions from religion, philosophy, science and elsewhere that are neither obviously true nor obviously false. It is not obviously true that God exists, but neither is it obviously true that God does not exist. It is not obviously true that doxastic voluntarism is true, but neither is it obviously true that it is not true.

Suppose I am concerned with the freedom of the will, study the issue thoroughly, but am torn between libertarianism and compatibilism. It is surely not obvious that one or the other is true. If the positions strike me as equally well-supported, then nothing at the level of intellect inclines me one way or the other. Must not will come in to decide the matter? Or consider the weightier question of the existence of God. Suppose the arguments pro et contra strike me as equally probative so that, at the level of intellect, I am not inclined one way or the other. If the issue is to be resolved, must I not simply decide to believe one way or the other? But Alston will have none of this: "How could we do that any more than, lacking any reasons at all for one alternative rather than another, we decide to believe that the number of ultimate particles in the universe is even rather than odd?" (p. 65)

This response packaged in a rhetorical question strikes me as very weak. No one cares what the number of particles is let alone whether it is odd or even. Indeed, it is not clear that the question even makes sense. (How could one possibly count them?) The God question is toto caelo different. In Jamesian terms, the God question is live, forced, momentous, and not intellectually decidable. A live issue is one that matters to us and seems to need deciding. Whether the number of ultimate particles is odd or even is certainly not live. A forced issue is one that is compulsory in the sense that we cbnnot not take a stand on it: to remain agnostic or uncommitted on the God question is practically to live as an atheist. There is nothing forced about the particles question. A momentous issue is one about which it matters greatly which position we adopt. The particles question is clearly not momentous. An intellectually undecidable qurestion is onr which, if it is to be decided, must be decided by an act of will.

So what I would say to Alston at this point in my study of his book is that in some cases -- those that fit the Jamesian criteria -- the will does in fact come into play in the formation of beliefs and indeed legitimately comes into play. To the extent that it does, a limited doxastic voluntarism is true.

Note that the issue concerns the formation of beliefs, not their maintenance, and note the contrast between direct and indirect formation of beliefs. Roughly, I form a belief directly, or 'at will,' by just forming it, not by doing something else as a means to forming it. Suppose the year is 1950 and you are a young person, sincere and idealistic, eager to consecrate your life to some cause higher than a bourgeois existence of consumption in suburbia. You have vibrant stimulating friends who are members of the CPUSA. They tell you that the Revolution is right around the corner. You don't believe it, but you want to believe it. So you go to their meetings, accept Party discipline, toe the Party line, and soon you too believe that the Revolution is right around the corner. In this example, the formation of belief is indirect. You do various things (go to the meetings, repeat the formulas, toe the line, etc.) in order to acquire the belief. But then in 1956 you learn of Krushchev's denunciation of Stalin and your belief in the glorious Revolution and its imminence suddenly collapses to be replaced by an opposing belief. The formation of the opposing belief is direct.

My correspondent Terra Vitae Man (hereafter, TV man) supplies an example of the third case, that of suspending belief:

Suspending belief. Sometimes in the face of good or strong evidence that p, I refuse to believe that p or again that not-p. I suspend any opinion on p.

This has always been my attitude on OJ and the murders he was charged with. Recently I talked with someone who had been teaching OJ knife-fighting in conjunction with a Commando-style TV show that never got launched. His evidence was excluded from the trial. Even in the face of this new evidence that OJ was competent with a knife, I do not form a opinion as to whether or not OJ killed his wife. (This is close to the classical skeptical epoche, except I do not bother to inquire and try to build a counterbalancing case for the opposite belief. Pyrrhonian skepticism says I always can build such a case and the result will be spontaneous cessation of belief.)

In both these cases [I omitted TV Man's first example] I think it's clear that what I believe (or don't believe ) is a function of what I will or wish to believe, trumping the evidence on hand and any reasonable induction therefrom. Hopefully, in both case it is also a principled refusal on my part to buy into beliefs that condemn other people. To believe my wife a poisoner or OJ a killer, I require evidence several parsecs beyond a reasonable doubt. You can say that that standard is too liberal [too stringent?], but I can choose to live (and die) by it and it is for several grounds an attractive ethics of belief.

This addresses my concern about the possibility of an ethics of belief. TV man suspends judgment, holds no opinion, on the guilt or innocence (as charged) of O.J. Simpson. By suspending judgment, he deliberately impedes or rather prevents the formation of two beliefs, the belief that O.J is guilty and the belief that O.J. is not guilty. I find that I have the same power of doxastic abstention, except that in this particular case I assent to the proposition that O. J. is guilty since I judge the evidence that he is guilty as charged to be overwhelming, and the notion that 'racism' played any part in this case utterly absurd. (My ethics of belief is perhaps less stringent than TV man's; but we both have an ethics of belief.)

Our question does not concern the content of an ethics of belief, but the very possibility of one whatever its content. Since 'ought' implies 'can,' if I ought to withhold judgment in some cases — and surely there are some cases in which I ought to withhold judgment — then I can withhold judgment. I have the power to withhold judgement; hence my epoche is voluntary. So here seems to be a case in which believing/disbelieving is under the direct control of the will: I decide to neither believe nor disbelieve. And from this it follows that the application of deontological categories is legitimate. "You ought not believe that your neighbor Jones is a homosexual on the basis of such flimsy evidence as that he is unmarried and has a Martha Stewart-like interest in home furnishings! You ought to suspend judgment!"

My provisional conclusion is that our manifest ability to suspend judgment in some case shows that we do have direct voluntary control over some of our believings. I have no control over my believing that a naked woman is standing in front of me if in fact such is the case (in good light, etc.) And I have no control over my believing that the Imperium Romanum no longer rules Western Europe (to adapt an example from Alston, Beyond 'Justification', p. 63). But I do have direct voluntary control over my believing that neighbor Jones is a homosexual.

The central problem of logic, Charles Sanders Peirce tells us, "is the classification of arguments, so that all those that are bad are thrown into one division, and those which are good into another . . . ." (Collected Papers vol. II, Elements of Logic, p. 119, sec. 14) On this characterization of logic's task, which I find attractive, it is a normative discipline: logic is not concerned to describe how people reason and argue as a matter of fact, but concerned to prescribe how they ought to reason and argue if they are to attain truth.

But talk of how people ought to reason, talk of what they ought to do and leave undone in matters of argument, presupposes that at least some belief formation comes under voluntary control. No one will contend that all our beliefs are subject to the control of the will, but if none of them are, then it seems there would be no possibility of reasoning, strictly speaking. And if there is no possibility of reasoning, strictly speaking, then there is no possibility of that normative study of reasoning called logic, which of course itself involves reasoning, namely, reasoning about correct and incorrect reasoning. Peirce:

Now, that which is uncontrollable is not subject to any normative laws at all; that is, it is neither good nor bad; it neither subserves an end nor fails to do so. But it is only the deliberate adoption of a belief in consequence of the admitted truth of some other proposition which is, properly speaking, reasoning. (p. 120)

The following argument suggests itself on the basis of this passage:

1. Reasoning is the deliberate adoption of beliefs on the basis of other beliefs.
2. Deliberate actions are under the control of the will.
Therefore
3. Reasoning is under the control of the will.
Therefore
4. Reasoning is subject to agential oughts and ought nots.
Therefore
5. There is room for an 'ethics of belief.'

For example, I open my mailbox and I see that it is empty. The belief that it is empty is not a belief I acquire at will; I can't help believing that the box is empty. Now consider four inferences I might draw: (a) The mailman is coming but has not yet come; (b) the mailman has come but had no mail for me; (c) today is a day on which mail is not delivered; (d) the mailman died on the way here. Reflecting on these inferences I realize that none of them are (deductively) valid and so I deliberately refuse to adopt any of the corresponding beliefs. But I do form the higher-order belief that none of the above beliefs is deductively justified by the perceptual deliverance. I also form the higher-order belief that it would be epistemically impermissible to think that the perceptual deliverance necessitates the truth of any one of (a)-(d).

I wonder what Alston would say about this case. Does it not show that some beliefs are formed at will?

The deflationist questions the presupposition. He suspects that truth has no nature. He suspects that there is no one property that all truths have, a property the having of which constitutes them as truths. His project is to try to account for our truth-talk in ways that do not commit us to truth's having a nature, or to truth's being a genuine property. Of course, we English speakers have and use the word 'true.' But the mere fact that we have and use the predicate 'true' does not suffice to show that there is a property corresponding to the predicate. (Exercise for the reader: find predicates to which no properties correspond.) So if we can analyze our various uses of 'true' in ways that do not commit us to a property of truth, then we will have succeeded in deflating the topic of truth and showing it to be metaphysically insubstantial or 'lightweight.'

The most radical approach would be one that tries to dispense with the predicate 'true' by showing that everything we say with its help can be said without its help (and without the help of any obvious synonym such as 'correct.') The idea here is not merely that truth is not a genuine property, but that 'true' is not even a genuine predicate. Consider two assertions. I first assert that snow is white, and then I assert that it is true that snow is white. The two assertions have the same content. They convey the same meaning to the audience. This suggests that 'It is true that ___' adds nothing to the content of what is asserted. And the same goes for the predicate '___ is true.' Whether we think of 'true' as an operator or as a predicate, it seems redundant, or logically superfluous.

In "Facts and Propositions" (1927), Frank Ramsey sketches a redundancy or logical superfluity theory of truth. This may be the first such theory in the Anglosphere. (Is there an historian in the house?) For Ramsey, "there really is no separate problem of truth but merely a linguistic muddle." Ramsey tells us that ". . . 'It is true that Caesar was murdered' means no more than that Caesar was murdered, and 'It is false that Caesar was murdered' means that Caesar was not murdered." (F. P. Ramsey, Philosophical Papers, Cambridge UP, 1990, ed. D. H. Mellor, p. 38)

But what about a case in which a proposition is not explicitly given, but is merely described, as in 'He is always right'? In this example, 'right' has the sense of 'true.' 'He is always' right means that whatever he asserts is true. As a means of getting rid of 'true' in this sort of case, Ramsey suggests:

1. For all p, if he asserts p, then p is true.

But since "the propositional function p is true is the same as p, as e.g., its value 'Caesar was murdered is true' is the same as 'Caesar was murdered,'" Ramsey thinks he can move from (1) to

2. For all p, if he asserts p, then p.

If the move to (2) is kosher, then 'true' will have been eliminated. Unfortunately, (2) is unintelligible. To see this, try to apply Universal Instantiation to (2). If the variable 'p' ranges over sentences, we get

3. If he asserts 'Snow is white,' then 'Snow is white.'

This is nonsense, because "'Snow is white'" in both occurrences is a name, whence it follows that the consequent of the conditional is not a proposition, as it must be if the conditional is to be well-formed. If, on the other hand, the variable 'p' is taken to range over propositions, then we get the same result:

4. If he asserts the proposition that snow is white, then the proposition that snow is white

which is also nonsense.

Unless I am missing something, it looks as if Ramsey's redundancy theory cannot succeed in eliminating 'true.' It looks as if 'true' is an indisensable predicate, and thus a genuine predicate. This does not, however, show that truth is a genuine property.

Yes, we have the predicates 'true' and 'false' and we apply them to declarative sentences and not to cigars and asteroids. But these predicates are not genuine since they are eliminable: we can say what we want without them. In saying that 'snow is white' is true, we say no more than that snow is white. The truth-predicate adds nothing to the content of what we say. It is dispensable. For this reason, the fact that we have and apply this predicate gives us no reason to suppose that there is a property corresponding to it. And if there is no property of truth, then nothing instantiates or bears this (putative) property, which is to say: there are no truth-bearers in the strict sense of this term.

I now concede that for there to be truth-bearers, strictu dictu, there must be some sense in which truth is a property: 'true' cannot be merely a device for de-quoting a quoted sentence. Consider the equivalence

1. 'Snow is white' is true if and only if snow is white.

Quine comments:

To ascribe truth to the sentence is to ascribe whiteness to snow; such is the correspondence, in this example. Ascription of truth just cancels the quotation marks. Truth is disquotation. (Pursuit of Truth, p. 80)

To put Quine's point precisely and aseptically, appending the phrase 'is true' to a quoted indicative sentence merely serves to remove the quotation marks. Now if this were the only use of 'is true,' then the predicate would be superfluous, and we would have no reason to suppose that it picks out a property of truth. But of course 'is true' also figures in sentences like these:

2. What Mary just said is true.
3. Whatever Mary says is true.
4. Goldbach's Conjecture is true.

Whereas we can delete 'is true' from "'Snow is white' is true" and be left with a sentence, we cannot delete 'is true' from (2)-(4) and be left with sentences. So 'is true' is not superfluous or eliminable and hence does express a property. And since it does express a property, those items that have this property, e.g., what Mary just said, are truth-bearers.

I am assuming that (2) and (3) have the following logical forms respectively:

2F. (x)(x = what Mary just said --> x is true)

and

3F. (x)(Mary says x --> x is true).

But a resolute objector might urge that the logical forms are really these:

2F*. (p)(p = what Mary said --> p)

and

3F*. (p)(Mary says that p --> p)

In the starred propositions the truth-predicate is absent. What we have instead is quantification over propositions. But it is hard to see why someone so parsimonious as to not allow a property of truth into his ontology would allow propositions in. And if propositions are allowed in, how could they not be truth-bearers? There is also a problem with making sense of the starred propositions. Let's try to translate (3F*) into something like English:

3F*T. For any proposition, if Mary says it, then it.

But this is unintelligible. It makes sense only if we add 'is true' or an equivalent after the final 'it.' This yields, 'For any proposition, if Mary says it, then it is true.' But then the truth-predicate has been let back in.

Would a substitutional interpretation of (3F*) help? That would look like:

3F*S. Every substitution instance of 'Mary says that p --> p' is true.

But again, this lets in the truth-predicate.

To sum up. The truth-predicate is eliminable in many cases but not in all. It is not eliminable in the case of 'Whatever Mary says is true.' For if we resort to propositional quantification to get rid of the truth-predicate, we hit a snag whether we interpret the quantifiers objectually or substitutionally. Interpreting the quantifiers objectually leads to unintelligibility, while interpreting them substitutionally lets the truth-predicate back in.

Since the truth-predicate is ineliminable, it is a genuine predicate and therefore expresses a property. The way is thus clear to admit truth-bearers as those items that instantiate the property of truth.

Philosophers attracted by deflationism are likely to be hostile to propositions. Hence their deflationism is likely to assume a sentential cast. A sentential deflationist will say that there is no more to truth than is captured in the following schema:

ES. A sentence 's' is true iff s.

For example,

1. 'Snow is white' is true iff snow is white.

Since deflationism is supposed to be a theory of truth, 'iff' is presumably not to be interpreted as material equivalence, but as something stronger: either logical equivalence (the necessitation of material equivalence) or as expressing a necessary analytic connection between the two sides of the biconditional. Either way, a problem arises.

No instance of (ES) is necessarily true. Thus (1) is not necessarily true. Reading (1) from left-to-right, it is possible that the left-hand side (LHS) be true and the right-hand side (RHS) be false. How? The LHS attaches 'true' to a sentence. But it is a contingent fact that the English sentence 'Snow is white' means what it means; it might have meant that snow is green. In possible worlds in which 'Snow is white' means that snow is green, and snow is green, the LHS is true and the RHS is false, which shows that (1) is not necessarily true.

Reading (1) from right-to-left, if snow is white, but 'snow is white' means that snow is green, then the RHS is true and the LHS is false, which again shows that (1) is not necessarily true.

Since every instance of (ES) is contingent, (ES) itself is contingent which shows that (ES) is not a theory of truth, a theory of how the truth-predicate works. Such a theory must explain how the truth-predicate works in all possible circumstances, not merely in those that happen to obtain.

REFERENCE: Marian David, Correspondence and Disquotation: An Essay on the Nature of Truth, Oxford, 1994, pp. 130-135.

Let me play along with 'Ockham''s demand. I will try to do this without begging any substantive questions. Thus I will not assume that there is a genuine property, truth, into which one might launch a substantive inquiry, and which one might try to analyze or otherwise explicate in correspondence, coherentist, pragmatic, or epistemic terms. Nor will I assume that truth is a genuine property that is indefinable or otherwise inexplicable. In sum, I will not assume that corresponding to the predicate 'true,' there is the property of being true. (It is not obvious that for every predicate there is a property. There are undoubtedly disjunctive predicates; but are there disjunctive properties?)

But is it not a fact that we have and use the predicates 'true' and 'false'? Surely we have these predicates in English and we use them by applying them to some things but not to others. For example, I do not apply 'true' to my cigar in the sense of 'true' that is alone relevant in these investigations. One could call that the sentential sense of 'true.' It is distinct from the sense of 'true' that figures into 'That cigar is a true (i.e.. genuine) Arturo Fuente Curlyhead.'

But we do sometimes apply 'true' and 'false' to sentences:

'Cuba is an island' is true.
'All cigars come from Cuba' is false.

Now it is my view that neither sentence types nor sentence tokens are the primary bearers of the predicates 'true' and 'false.' But that is not to say that sentences are not bearers of these predicates. We also say of some beliefs that they are true, and of other beliefs that they are false. What we usually mean by this is that the content of a belief-state is either true or false. But if a belief-content is true, and I am believing that content, then my belief-state is also true, albeit in a secondary or tertiary sense.

From the foregoing considerations it should be spectacularly obvious that there are truth-bearers. We may disagree about which entities are the primary truth-bearers, but that there must be some truth-bearers or other follows from the plain facts that (i) we have the predicates 'true' and 'false' and (ii) we apply them to some things and not to others.

Note that I am not assuming that there is some property of truth. Of course, I think there is; but I am not assming that there is for present purposes. I am simply pointing out that we apply the predicates 'true' and 'false.' Since we do apply them, there must be something to which we apply them. These items are the truth-bearers. Therefore, truth-bearers exist.

But the foregoing, as pellucid as it is, will not convince our nominalist friend. For it seems he denies that 'true' and 'false' are predicates! He writes,

Your argument is only valid if 'is true' is a genuine predicate. I deny that. It is a verb. If I prefix the word 'that' to the sentence 'snow is white' I get 'that snow is white'. That is a noun phrase which I can use in ways such as 'that snow is white is a fact appreciated by eskimos' &c. So the effect of prefixing 'that' is to remove the main verb of 'snow is white'. The effect of prefixing 'it is true' to the 'that' clause has the effect of putting the verb back. Thus 'it is true that snow is white' is equivalent to the original 'snow is white'. That is all we need to know about 'is true'.

The idiosyncratic use of 'verb' is this passage makes it read like gibberish to me. But some sense can be extracted from this passage. The following equivalence holds:

E. It is true that snow is white iff snow is white.

So instead of talking about the predicate 'true' we can talk about the operator 'It is true that ____.' But this makes no difference to my argument. For the operand of the operation -- that upon which the operator operates -- is a truth-bearer, or, if you insist, a truth-operand. But it may be that 'Ockham' wants to parse (E) as follows:

E*. It is true (that snow is white) iff snow is white.

But how would this show that 'true' or 'is true' is not a genuine predicate? If 'is true' applies or attaches to 'that snow is white,' then 'that snow is white' is a truth-bearer. And what does he mean by 'genuine predicate'? Is he perhaps confusing predicates and properties?

A natural way to read (2) so that it comes out unmistakably true is as

2*. Cognitive conditions are not now ideal.

But then (4) becomes

4*. Cognitive conditions are not now ideal would be accepted in cognitively ideal conditions

where 'now' refers to the same time, say, 2007 A. D.

It is not obvious, however, that (4*) is self-contradictory. For suppose that in the fullness of time, an intrepid band of researchers becomes so well-placed that their actual cognitive conditions are ideal. They would then be in a position to know that cognitive conditions back in 2007 were not ideal.

But I think I can reformulate the argument so as to evade the objection by replacing (2) with

2R. Cognitive conditions are never ideal.

(4) then becomes

4R. Cognitive conditions are never ideal would be accepted in cognitively ideal conditions.

Now (4R) is self-contradictory since it implies that at no time in the actual world are cognitive conditions ideal and at some (future) time in the actual world cognitive conditions are ideal.

But I hear an objection coming: "Cognitively ideal conditions are not future conditions in the actual world; they are future conditions in some merely possible world. Let 'A' be a rigid designator of our world, the actual world, and let 'W' designate a merely possible world that is exactly the same as A up until the present, but then different from A after the present. 'Cognitively ideal conditions' refers to conditions in the future of W. Now the following two propositions are logically consistent: At no time in A are cognitive conditions ideal and At some (future) time in W cognitive conditions are ideal. Therefore, when 'cognitively ideal conditions' is properly construed, (4R) is seen to be non-contradictory."

But if cognitively ideal conditions are never attained, but are merely possible conditions, then it is difficult to see how the actual truth of any proposition could be identified with what investigators in some merely possible circumstances accept. How can what is actually true be explicated in terms of merely possible judgments?

We should recall that the granddaddy of the sort of epistemic approach to truth that we have been discussing was C. S. Peirce. He wrote that "The opinion which is fated to be ultimately agreed to by all who investigate, is what we mean by truth, and the object represented in this opinion is the real." ("How to Make Our Ideas Clear") The talk of fate here suggests that truth is being explicated in terms of future judgments in the actual world rather than future judgments in a merely possible world.

Krausz appreciates that a relativism asserted as nonrelatively true refutes itself. A relativized relativism (RR), however, lacks general force. The issue is whether this lack of general force amounts to an objection to RR. Krausz thinks that it does not. RR may reflect a 'local understanding.'

Well, suppose that among Krausz and his group it is held that truth is relative and that this truth is itself relative to Krausz and his group. It is held by me and my group that truth is nonrelative. What I say about Krausz is that his view is nonrelatively false and rests on a confusion of truth with local understanding. Clearly, he cannot reply in kind since he has tied his own hands by retreating to RR, a retreat that he had to make on pain of falling into contradiction. He cannot, that is, say that my view is true only for me and my group. He cannot say that my nonrelativism is only relatively true. If he said that, he would be assuming that relativism is nonrelatively true and would thereby fall into self-contradiction. His retreat to RR was a forced retreat. He therefore has to admit that my view is possibly nonrelatively true, and not just for me. He has to admit this because my position is a possible position he cannot exclude. He must admit that it is possibly nonrelatively true that truth is nonrelative. But if so, then it is possible that there be one nonrelative truth. Now possibly there is one nonrelative truth is nonrelatively true. It follows that there actually is one nonrelative truth, which suffices to show that nonrelativism is true.

Since nonrelativism is true, it follows that RR is false. This in turn implies that Krausz and Co. are confusing truth with "local understanding."

Let me sum this up in a somewhat different way. The retreat to RR is forced. This retreat insulates relativism from the charge of self-referential inconsistency. But it also makes it impossible for the proponent of RR to exclude as false or incoherent the position of the nonrelativist. Not being able to exclude the position of the nonrelativist, Krausz must admit that the nonrelativist position (NR) is possibly true, indeed possibly nonrelatively true. Surely NR is epistemically possible for Krausz and Co., possible given what they can legitimately claim to know. But this epistemic possibility supports the metaphysical possiblity that NR is true. So he must admit that NR is possibly true. But that NR is possibly true is itself nonrelatively true. After all, Possibly p entails it is true that possibly p. So Krausz must admit not only that NR is possibly true, but also that it is actually nonrelatively true that possibly NR.

And that suffices to show that NR is true, RR is false, and that Krausz is confusing truth with "local understanding."

A. Truth as Idealized Rational Acceptability

One way of suppressing the classical intuition is by offering an epistemic definition of 'true.' One attempts to explicate truth in terms of mental states. Thus someone might suggest that a proposition is true just in case it is believed or accepted by someone. But this won't do, since there are truths that are not accepted by anyone. So one proposes that a proposition is true just when it is acceptable. This proposal, too, is defective inasmuch as what is acceptable to one person will not be acceptable to another. This defect can perhaps be handled by identifying truth with rational acceptability. But what it is rational to accept at one time or in one place may be different from what it is rational to accept at another time or in another place. Much of what we find rationally acceptable would not have been found rationally acceptable by the ancient Greeks. (For example, that the same physics holds both for terrestrial and for celestial bodies.) So one advances to the notion that truth is rational acceptability at the ideal limit of inquiry. One can trace this notion back to C. S. Peirce. In Reason, Truth, and History, Hilary Putnam presents a version of it. Let's consider the theory in the following form:

1. P is true =df p would be accepted in cognitively ideal conditions.

Now we know that

2. Cognitive conditions are not ideal.

From (2) it follows via the trivial equivalence principle *p* is true iff p that

3. Cognitive conditions are not ideal is true.

It follows from (3) via (1) that

4. Cognitive conditions are not ideal would be accepted in cognitively ideal conditions.

But (4) is self-contradictory, whence it follows that

5. The definition of truth in terms of acceptability in cognitively ideal conditions is incorrect.

What I take this argument to show is that the notion of truth as correspondence to the way things are is primary and irreducible. For surely (2) is true. But its being true cannot be explicated in terms of what anyone would accept or assert under ideal epistemic conditions. Therefore, (2) is true in a sense more basic than the sense spelled out in (1).

B. Truth as Coherence

We get a similar result if we try to construe truth as coherence. Suppose

6. P is true =df p would be accepted by a person whose set of beliefs is maximally consistent and coherent.

But we know that

7. No one's set of beliefs is maximally consistent and coherent.

From (7) it follows via the above equivalence principle that

8. No one's set of beliefs is maximally consistent and coherent is true.

It follows from (8) via (6) that

9. No one's set of beliefs is maximally consistent and coherent would be accepted by a person whose set of beliefs is maximally consistent and coherent.

But (9) is self-contradictory, so

10. (6) is incorrect.

Note first that in your example it is not the banana that is true or false, but the banana-on-the-table. A banana by itself is neither true nor false; it simply exists. And of course one could establish a code whereby a banana placed on a certain table at a certain time signifies that Bond is a double agent, or any proposition you choose. Or it might be a banana half-peeled or whatever. But a banana as such could not figure in a code. For to associate the proposition Bond is a double-agent with something physical, the physical item would have to have a structure isomorphic to that of the proposition: the physical item could be not a simple individual, but would have to be a state of affairs.

But the main point is that neither a banana, nor a banana-on-a-table, nor any physical object, nor any collection or arrangement of physical objects intrinsically means anything. You and your conspirators must assign a meaning to the collection of physical objects ab extra. The meaning is not inherent in the collection, but imputed to it from without. It is not intrinsic but extrinsic. In and of itself a collection of physical objects means, signifies, intends nothing; any intentionality it comes to acquire is derived, not original, intentionality. Only mind can mean originally; an arrangement of physical objects means, if it means at all, derivatively.

You will also note that I was careful to say in the passage quoted that not everything could be a primary truth-bearer. That is consistent with saying that uttered sounds, or marks on paper, or bananas on tables could be secondary truth-bearers.

For a more detailed treatment of this topic, see Of Truth and Teddy Bears: Could Anything be a Truth-Bearer?

Example. Having just finished my drink, the thought expressed by an assertive utterance of 'My glass is empty' is true. But the thought is not just true; it is true because of the way things are 'outside' my mind. The glass (in reality) is (in reality) empty. So the realist says something like this: the thought (proposition, judgmental content, etc.) is true in virtue of the obtaining of a truth-making state of affairs or fact. The thought is true because the fact obtains or exists, where 'because' does not have a causal sense but expresses the asymmetrical relation of truth-making. The fact is the ontological ground (not the cause) of the thought's being true.

One might wonder whether this realist theory of truth leads to an infinite regress, and if it does, whether the regress is vicious. Some cryptic remarks in Gottlob Frege's seminal article, "The Thought: A Logical Inquiry," suggest a regress argument against the correspondence theory of truth.

For Frege, a thought (Gedanke) or proposition is the sense (Sinn) of a context-free declarative sentence. 'Snow is white' and its German translation Schnee ist weiss are examples of context-free declarative sentences. When we say that a sentence such as 'Snow is white' is true, what we are really saying is that the sense of this sentence is true. Now could the being-true of a sentential sense consist in its correspondence to something else? Frege rejects this notion: "In any case, being true does not consist in the correspondence of this sense with something else, for otherwise the question of truth would reiterate itself to infinity." (Phil. Logic, ed. Strawson, p. 19) A little earlier, Frege writes,

For what would we then have to do to decide whether something were true? We should have to enquire whether it were true that an idea and a reality, perhaps, corresponded in the laid-down respect. And then we should be confronted by a question of the same kind and the game could begin again. So the attempt to explain truth as correspondence collapses. And every other attempt to define truth collapses too. (Ibid.)

What exactly is Frege's argument here? We begin by noting that

1. Necessarily, for any p, it is true that p iff p.

This equivalence, which I hope nobody will deny, gives rise to an infinite regress, call it the truth regress. For from (1) we can infer that if snow is white, then it is true that snow is white, and iterating the operation, if it is true that snow is white, then it is true that it is true that snow is white, and so on without end. This is an infinite regress all right, but it is obviously benign. For if we establish the base proposition, Snow is white, for example, then we ipso facto establish all the iterations. Our establishing that snow is white does not depend on a prior establishing that it is true that snow is white. In general, our establishing of any proposition in the infinite series does not depend on having first established the next proposition in the series. The truth regress is benign.

Note that if the truth-regress were vicious, then the notion of truth itself would have been shown to be incoherent. For the truth-regress is a logical consequence of the equivalence principle (1) above, a principle that simply unpacks our understanding of 'true.' So if the truth-regress were vicious, then (1) would not be unproblematic, as it surely is.

It follows that if Frege's Regress is to amount to a valid objection to the definition of truth as correspondence, "and [to] every other attempt to define truth," then Frege's Regress must be different from the truth regress. In particular, it must be a vicious regress. But then what is Frege's Regress? Consider

2. Necessarily, for any p, it is true that p iff *p* corresponds to reality.

One can think up counterexamples to (2), but the precise question before us is whether (2) issues in a vicious infinite regress. Now what would this regress (progress?) look like? Let 'T(p)' abbreviate 'it is true that p.' And let 'C*p*' abbreviate '*p* corresponds to reality.' (The asterisks function like Quine's corners.) The regress, then, looks like this:

3. p iff T(p) iff C*p* iff T(C*p*) iff C(T(C*p*)) iff T(C(T(C*p*) iff C(T(C(T(C*p*)) . . .

Is (3) a vicious regress? It would be vicious if one could establish T(p) only by first establishing C*p* and so on. But if these two terms have the same sense, in the way that the first and second terms have the same sense, then (3) will be as benign as the truth regress. Suppose that 'It is true that p' and '*p* corresponds to reality' have the same sense. Suppose in other words that the correspondence theory of truth is the theory that the sense or meaning of these distinct sentences is the same. It would then follow that to etablish that it is true that p and to establish that *p* corresponds to reality would come to the same thing, whence it would follow that the regress is benign.

For the regress to be vicious, the second and third terms must differ in sense. For again, if the second and third terms do not differ in sense, then to establish one is to establish the other, and it would not be case that to establish that it is true that p one would first have to establish that *p* corresponds to reality or to some chunk of reality. But if the second and third terms do not differ in sense, then it appears that the regress doesn't get started at all. For the move from the second term to the third to be valid, the entailment must be grounded in the sense of the second term: the third term must merely unpack the sense of the second term. If, however, the two terms are not sense-connected, then no infinite regress is ignited.

My interim conclusion is that it is not at all clear that Frege's Regress constitutes a valid objection to theories of truth, in particular to the theory that truth resides in correspondence.

REFERENCE: Peter Carruthers, "Frege's Regress," Proc. Arist. Soc., vol. LXXXII, 1981/1982, pp. 17-32.

In context, the passage forwards the following argument:

1. Only sentences can be true or false.
2. "Sentences are elements of human languages. . . ."
3. "Human languages are human creations. . . ."
Therefore
4. Sentences are human creations. (From 2, 3)
Therefore
5. Only human creations can be true or false. (From 1, 4)
6. Human creations "cannot exist independently of the human mind."
Therefore
7. Truth and falsity cannot exist independently of the human mind.
Therefore
8. "Truth cannot be out there."

This argument is dubious because premise (1) is dubious. The primary vehicles of the truth-values cannot be sentences. A written sentence is just a collection of physical marks on paper or in some other medium, and no collection of physical items can, just by itself, have meaning, let alone a truth-value. The same goes for an utterance of a sentence. An utterance is a fleeting physical phenomenon and so cannot possibly be either true or false in the primary sense. It is not the fleetingness, but the physicality that is the problem. Noises emanating from the vocal cords can no more be true or false than can the rattling of a rattlesnake. The same goes for inscriptions, even those that are carved in stone. The problem is simply that physical phenomena lack intrinsic representational power. No physical phenomenon is intrinsically of or about anything; no such phenomenon has non-derivative intentionality. Hence no physical phenomenon is non-derivatively meaningful, whence it follows that no such phenomenon can be non-derivatively true or false. A mind (a possessor of non-derivative intentionality) may interpret a rattling sound as signifying the presence of a dangerous critter, but the rattling itself can do no such thing. Similarly, a mind can interpret an inscription as meaning such-and-such, but the sentence itself, qua physical phenomenon, cannot mean anything.

The meaningfulness of sentence-tokens is derived, not original. In Turkish, 'Seni seviyorum' means what 'I love you' means in English. But 'Seni seviyorum' has this meaning not in itself, not in virtue of its spatial and geometric and physical properties, but only in virtue of the fact that speakers of Turkish use it to express their meaning. And the same goes for 'I love you.' It has the meaning it has only in virtue of the fact that speakers of English use it to express their meaning. As John Searle puts it, "Linguistic meaning is a real form of intentionality, but it is not intrinsic [original] intentionality. It is derived from the intrinsic intentionality of the users of the language." Linguistic meaning and reference is parasitic upon mental meaning and reference.

So it cannot be (declarative) sentences qua physical phenomena that are either true or false, it must be sentences interpreted by minds to mean this or that. This fact by itself, however, does not suffice to show that truth is "out there," i.e., independent of the human mind. For one could maintain the truth and falsity are properties of contents of acts of judging, and that these judgmental contents exist only as the accusatives of acts of judging.

But when I judge that a is F, and my judgment is true, then the item denoted by 'a' in reality possesses the attribute expressed by 'F.' My judgment that the thing coming toward me is a coyote is true only if in fact, "out there," there is something coming toward me and the thing in question exemplifies the property of being a coyote. So even if the truth-bearer is in the mind, the truth-maker, that which makes the truth-bearer true, cannot be in the mind.

To cop a line from the Tractarian Wittgenstein, Die Welt is die Gesamtheit der Tatsachen, nicht der Dinge. "The world is the totality of facts, not of things," where these facts are what make truth-bearers true. Of course, Rorty will strenuously oppose this realism about truth. He will insist that there are no facts in themselves to which any of our sentences correspond.

And yet he tells us that the world is "out there." Well then, what is this world that is "out there" beyond language and mind? Is it undifferentiated mush, a sort of materia prima to be informed as we wish? Then it is a world well lost. Does it instead possess an intrinsic structure? Then, even if propositional truth is in the mind, there must be in the world proposition-like entities, truth-making facts or states of affairs. But then there is a clear sense in which, contra Rorty, truth is "out there." Propositional truth needs grounding in ontic truth, the truth of things. Hence the scholastic tag, ens et verum convertuntur. This convertibility of being and truth is the ground of the world's intrinsic intelligibility. But accepting as he does the death of God, which Nietzsche realized was also the death of truth, Rorty must acquiesce in the death of intrinsci intelligibility. The world gets reduced to a brute meaningless surd on which play out his incommensurable conversations, conversations unto no end, conversations for their own sake. "We are not conversing because we have a goal, but because Socratic conversation is an activity which is its own end." (Ibid., p. 172)

‘Truth is relative’ is elliptical for ‘Truths are relative.’ This is because we encounter truth only in the form of particular truths. A truth is a true truth-bearer. A truth-bearer is any entity that can be characterized correctly as either true or false. Not everything could be a (primary) truth-bearer: a belief, but not a banana, could be reasonably said to be true or false. My belief that a particular banana is ripe is either true or false; but the banana itself is neither true nor false. Of course, one could speak of a ‘true’ (genuine, real) banana as opposed to a ‘false’ (fake, artificial) banana; but these extended uses of ‘true’ and ‘false’ are not relevant to this discussion.

There is an issue in the philosophy of language concerning which entities play the role of primary truth-bearer: sentences in the indicative mood? Sentences in the indicative mood from which all indexical elements (demonstratives, tenses of verbs, etc.) have been removed? Beliefs? Abstract objects such as Fregean propositions? Utterance-events? Some even hold the bizarre view that the primary truth-bearers are sets, sets of possible worlds. We may set aside this technical question for present purposes. All we need is the notion that a truth is a true truth-bearer, whatever truth-bearers turn out to be. Now consider

R. All truths are relative.

(R) is the characteristic thesis of the alethic relativist. The relativist asserts it because he thinks it is true. Think of a multiculturalist who asserts that all truths are culture-relative. One gets the impression that (R) is often asserted as absolutely or non-relatively true, as simply representing the way things are independently of our beliefs and desires, as a sort of insight with a serious claim on our attention, and thus as amounting to the robust

AR. It is true that all truths are relative

and not as amounting to something anemic like

RR. It is relatively true that all truths are relative.

Relativists behave as if relativism is something that needs to be noted, and faced up to, something the recognition of which will make us more ‘enlightened,’ less ‘dogmatic,’ something whose truth will set us free from ‘prejudice,’ and so on; they don’t usually intend something limp-wristed like: it is merely true in my personal opinion that (R).

But (AR) -- absolute relativism -- is obviously self-contradictory. For if (R) is true, then, given that (R) applies to itself, it follows that (R) is relatively true, which contradicts (AR). Of course, one might maintain that (R) does not apply to itself, that (R) does not fall within the scope of the universal quantifier, ‘all.’ (R) would then be a meta-level assertion about a base level of ‘ordinary’ truths. I might discuss this option in a separate post. For now, however, the crucial point that the reader must focus on is that (AR) is self-refuting if taken to apply to itself – which is of course the most natural way to take it.

A logically consistent relativist must therefore retreat to (RR), relativized relativism. This is a forced retreat if the relativist is to avoid self-contradiction. But then his position becomes curiously self-vitiating. One problem is to specify what relative truth could possibly be.

Consider the property of being-believed, whether by an individual or a group of individuals, or the individuals of a particular historical epoch, or social class, etc. This is a relational property: necessarily, to be believed is to be believed by someone. Thus the following ‘theory’ is wholly unproblematic:

B. All believed-propositions are relative.

Note that

B*. It is believed that all believed-propositions are relative

is also unproblematic, unlike (AR).

So if all the alethic relativist means by the relative truth of a truth-bearer is the truth-bearer’s being believed, or accepted, or affirmed, or endorsed, or held, or sincerely asserted, then the alethic relativist is saying something unproblematic. But he is also saying something completely trivial, completely uninformative. In other words, he is not saying anything at all about the nature of truth when he says that truth is relative, but simply replacing truth with something quite distinct, namely, the property of being-believed, a property all will agree is relational. He is simply changing the subject. The topic was truth, and what the alethic relativist seemed to be wanting to say was something about its nature, namely, that it is by its very nature relative. But if relative-truth is just the property of being-believed, accepted, etc., then nothing has been said about truth at all.

To sum up. The aletheic relativist either asserts his thesis (R) as absolutely true or as relatively true. If the former, his thesis is self-refuting. If the latter, then his thesis avoids self-contradiction only to face a dilemma: either relative-truth is the same as the property of being-believed, or it is not. If the former, then the relativity of truth boils down to an uninteresting triviality. If the latter, then it remains wholly unclear what could be meant by the property of relative-truth, and the thesis (R) perishes of semantic indeterminacy.

There is more to it than this, of course. But this is a good start.

The answer to the first question is a resounding No. Since physics has nothing to do with moral questions -- which is not to say that moral questions do not arise in the technological application of physical knowledge or in its dissemination or in the construction of experiments, etc. -- it is quite clear that neither STR nor GTR nor any physical theory has any logical consequences in respect of metaethical doctrines such as moral relativism. And as for cognitive or alethic relativism, far from its being entailed by the Theory of Relativity, I should think that the latter presupposes the absoluteness of truth.

Take the Galilean principle of the additivity of velocities. Suppose I'm on a train moving with velocity v1. I fire my gun in the direction of the train's travel. The projectile's velocity is v1 + v2. But STR implies that the additivity of velocities breaks down at relativistic speeds, speeds approaching the speed of light. Now the proposition that the principle of the additivity of velocities fails at relativistic speeds is not merely true relative to STR, but true absolutely. And the same goes for any number of other propositions of STR and GTR such as the one bearing upon the conversion of mass and energy, E=mc^2, or that the speed of light remains a constant 186, 282 mi/sec. Or consider the proposition that motion and rest are relative to reference-frames. That proposition's truth is not relative to any reference-frame or to any conceptual framework either.

In short, Einstein's Theory of Relativity, far from entailing relativism about truth, presupposes, and thus entails the absoluteness of truth. (Or should I say instead that the Theory of Relativity is logically independent of the question of the relativity of truth?)

As for the sociological question, I suppose one would have to grant that misinterpretations and shallow expositions of the Theory of Rewlativity did contribute to the spread of moral and cognitive relativism. But of course that is not the responsibility of Einstein or modern physics but the responsibility of those shallow-pates we call journalists. (Am I being unfair a second time?)

Now what I have just written is pretty simple-minded. Is this a case where simplex sigillum veri or is there a deeper level of consideration that I am missing?

1. Logical or Propositional Consistency. To say of two propositions that they are consistent is to to say that they can both be true, where 'can' expresses logical or broadly logical possibility. To say of two propositions that they are inconsistent is to say that they cannot both be true, where 'cannot' expresses logical or broadly logical impossibility. I am blogging but not wearing a hat. My blogging is obviously consistent with my not wearing a hat since both propositions are true. But my blogging is also consistent with my wearing a hat since it is possible that I both be blogging and wearing a hat. But I am blogging now and I am not writing now are inconsistent since they cannot both be true. The first proposition entails the second, which implies the impossibility of the first being true and the second false.

Liberals often say that support for fetal rights is 'inconsistent' with support for the death penalty. But surely the following propositions are not logically inconsistent:

P1. Unborn human fetuses have a right to life which cannot be overridden except in rare cases (e.g. threat to the life of the mother).

P2. Capital punishment for certain offences is morally justified.

Where is the inconsistency? I see none. Indeed, they strike me as consistent. Suppose that the right to life is grounded in the potentiality to develop into something to which one accords person-status. This is a possible view, and indeed one actually held by many. Any Aristotelian worth his salt will be able to spell it out coherently and perhaps even persuasively. Suppose also that a similarly coherent case is made for P2. One will then have made a case for the possibility of each. Joint possibility does not straightaway follow from the possibility of each, but if there is no reason to think of them as jointly impossible, then we should consider them jointly possible. On the face of it, P1 and P2 are logically consistent.

2. Pragmatic Consistency. By this I mean consistency in the application of a principle. Consistency in this sense is not a relation between or among propositions, but involves a relation between a principle and cases that it may or may not subsume.

To apply a principle consistently is to apply it to all like cases in a like manner. So if the principle is to cherish, value, uphold human life, then perhaps the thought of those who see inconsistency in support of both fetal rights and capital punishment is that to apply this principle to the unborn but not to certain criminals is to apply it inconsistently. "How can you oppose the killing of the unborn, when you favor the killing of certain criminals? Be consistent!"

The answer to this should be obvious. The cases are dissimilar. The unborn are innocent while the criminals are not. This is a difference that makes a moral difference; it is a difference that justifies a difference in treatment. No doubt principles should be applied to like cases in a like manner. But there is no inconsistency in refusing to apply a principle to cases that are relevantly dissimilar.

Principles obviously allow of exceptions. A person who fancies himself 'consistent' in opposing both abortion and capital punishment on the ground that all human life is worthy of respect will in almost all cases make an exception when it comes to killing in self-defence or killing in a just war. Now wouldn't it be churlish and absurd to accuse such a person of being 'inconsistent'? "If you were consistent in your valuing of human life, you would value the life of the jihadi who has come to decapitate you and your family, and you would not lift a finger against him."

3. Diachronic or 'Emersonian' Consistency. Here is a famous passage from Ralph Waldo Emerson's "Self-Reliance" rarely quoted in full:

A foolish consistency is the hobgoblin of little minds, adored by little statesmen and philosophers and divines. With consistency a great soul has simply nothing to do. He may as well concern himself with his shadow on the wall. Speak what you think now in hard words and tomorrow speak what to-morrow thinks in hard words again, though it contradict every thing you said to-day.(Ziff, 183)

People routinely rip the initial clause of this passage out of its context and take Emerson to be attacking logical consistency. But he is doing no such thing. It should be clear that Emerson’s dictum has nothing to do with logical consistency and everything to do with consistency of beliefs over time. The consistency in question is diachronic rather than synchronic. A “little mind” is “foolishly consistent” if it refuses to changes its beliefs when change is needed due to changing circumstances, further experience, or clearer thinking. It should be clear that if I believe that p at time t, but believe that ~p at later time t*, then there is no time at which I hold logically inconsistent beliefs. Doxastic alteration, like alteration in general, is noncontradictory for the simple reason that properties which are contradictory when taken in abstracto are had at different times. My coffee changes from hot to non-hot, and thus has contradictory attributes when we abstract from the time of their instantiation. But since the coffee instantiates them at different times, there is no contradiction such as would cause us to join Parmenides in denying the reality of the changeful world.

Belief change is just a special case of this. Suppose a politician changes her position for some good reason. There is not only nothing wrong with this, it shows an admirable openness. She goes from believing in a progressive tax scheme to believing in a flat tax, say. Surely there is no logical contradiction involved, and for two reasons. First, the property of believing that a progressive tax is warranted is not the contradictory, but merely the contrary, of the property of believing that a flat tax is warranted. (They cannot both be instantiated at the same time, but it is possible that neither be instantiated.) Second, the properties are had at different times. A logical contradiction ensues only when one simultaneously maintains both that p and that ~p.

Emerson’s sound point, then, is that one should not make a fetish out of doxastic stasis: there is nothing wrong with being ‘inconsistent’ in the sense of changing one’s beliefs when circumstances change and as one gains in experience and insight. But this is not to say that one should adopt the antics of John F. (‘F’ for Flibbertigibbet, or else ‘Flip-Flop') Kerry. Relative stability of views over time is an indicator of character.

Before leaving this topic, let's consider what Walt Whitman has to say in the penultimate section 51 of “Song of Myself” in Leaves of Grass:

Do I contradict myself? Very well then I contradict myself, (I am large, I contain multitudes.)

Here it appears that Whitman is thumbing his nose at logical consistency. If so, the Emersonic and Whitmanic dicta ought not be confused.

4. Ought one be consistent in any of these three senses? I should say it is self-evident that one ought to strive to be logically consistent. For if one's beliefs are inconsistent, then they cannot all be true, which implies that one or more of them are false. The ideal doxastic set would consist of beliefs that are both true and consistent. Logical consistency by itself is not enough since false beliefs can be consistent. But consistency is necessary for doxastic perfection and thus a goal worthy of pursuit.

As for pragmatic consistency, consistency in the application of principles to like cases, this too seems to be a goal that ought to be pursued, although I don't know how I would argue for it.

Diachronic (Emersonian) consistency, however, is not an epistemic desideratum. On the contrary, one ought to change one's beliefs when new facts and arguments so warrant.

One of the marks of a de dicto report is that one cannot existentially generalize on a singular term in 'p.' Thus from (2) one cannot infer that there exists an x such that x died in 1925. A second mark of a de dicto belief sentence is that one cannot replace a referring term in 'p' with a coreferential term salva veritate. (Salva veritate can be translated: in such a manner as to preserve truth.) Thus one cannot replace 'Frege' in (2) with 'The author of the Begriffschrift' salva veritate even though it is true that Frege is the (sole) author of the work in question.

These two points should seem obvious. As for the second, suppose all that Sam knows about Frege is that he is a great logician who had an enormous influence on Russell and subsequent philosophy. It should then be clear that (2) does not entail that Sam believes that the author of the Begriffschrift died in 1925.

A de dicto belief report, then, is one that satifies the two requirements just described. That there are de dicto beliefs seems beyond question. Whether a given belief sentence is to be read as de dicto or as de re may be open to dispute, but that there is a distinction between the readings is clear.

Here are what I take to be two canonical forms of de re belief reports:

3. X is believed by S to be F.

4. S attributes the property of being F to x.

These are equivalent. Examples of (3) and (4) respectively are

5. Frege is believed by Sam to have died in 1925

and

6. Sam attributes the property of having died in 1925 to Frege.

One difference between (2), on the one hand, and (5) and (6) on the other, is that one can existentially generalize on 'Frege' in (5) and (6) but not in (2). A second difference is that in (5) and (6), but not in (2), 'Frege' can be replaced, salva veritate, with the coreferential 'the author of Begriffschrift.' For if Sam attributes the property of having died in 1925 to Frege, and Frege = the author of the Begriffschrift, then Sam attributes the property in question to the author of the Begriffschrift.

A third difference is that 'Frege' in (2) but not in (5) and (6) occurs within the scope of the propositional attitude verb, 'believes.'

Note also that (5) does not entail (2). It could be true that Frege is believed by Sam to have died in 1925 without it being true that Sam believes that Frege died in 1925. How? Well, suppose Sam comes upon Frege's tombstone, but can't read the name due to some obstructing foliage. But he can read the terminal date, 1925. So he forms the belief that the man buried here died in 1925. Now the man in question is Frege. So Sam believes, of Frege, (de re!) that he died in 1925. Suppose further that Sam has never heard the name Frege and so associates no sense with it. Under these suppositions, Sam lacks the belief that Frege died in 1925.

But it is not the possible that is actual that is of concern here, but the possible that is unactual. This is called the merely possible. Is there knowledge of the merely possible? If there is, how could we have such knowledge? Do I know that the cat that is now collared might now have been uncollared? And if I do know it, how do I know it? What is the source of this knowledge, and what is its object?

Although Peter van Inwagen calls himself a modal sceptic, he has no doubt that we do have knowledge of some unrealized possibilities. "I know that it is possible that . . . the table that was in a certain position at noon have then been two feet to the left of where it in fact was." (Ontology, Identity, and Modality, p. 246) Van Inwagen also cites the possibility of John F. Kennedy's having died from natural causes. In both of these cases, we have knowledge about an unrealized possibility that will forever remain unrealized. But there is also modal knowledge of the impossible and the necessary: "it is impossible for there to be liquid wine bottles, and . . . it is necessary that there be a valley between any two mountains that touch at their bases." (246)

So we have (nontrivial) modal knowledge. But how is this possible? I know that JFK might have died from natural causes. But how do I know this? The proposition JFK died of natural causes is known to be false. So how can I know that it is possibly true? It isn't true and never will be true. As I would put it, there seems to be nothing for my knowledge to grab onto.

But my concern at the moment is not how we have (nontrivial) modal knowledge, but whether we do have it even in such mundane cases as that of the possibly uncollared cat and the table that might at time t have been in a different position than the position it was in at t. There may be reasons to be skeptical about even these mundane cases. Consider the following principle according to which possibilities come in world-sized packages:

WSP: State of affairs S is possible only if there is a maximal possible state of affairs W such that W includes S.

I could also express this in terms of (Fregean) propositions instead of in terms of (Chisholmian-Plantingian states of affairs):

WSP*: Proposition p is possible only if there is a maximal possible proposition w such that w entails p.

Think of entailment as the necessitation of material (Philonian) implication. Thus, p entails q iff necessarily (p --> q) where '-->' expresses material implication. A proposition is maximal iff it entails every proposition with which it is consistent. Think of a possibility as a possibly true proposition. The notion that possibilities come in world-sized packages is the notion that a possibly true proposition is possible only if it is entailed by a maximal proposition, a world-proposition. If w is a conjunctive world proposition, and if w entails p, then p is compossible with all of w's conjuncts. So the idea is that every genuine possibility is a com-possibility: there are no isolated possibilities.

So the possibility that the collared cat before me be uncollared now is a genuine possibility only if there is a possible world that includes it or entails it. This seems to make it more difficult for me to know that this possibility is indeed a possibility. For I have to do more than focus on it; I must focus on an entire possible world. If one takes imaginability as the source of the knowledge of mundane possibilities like the one just mentioned, then one has to imagine, not merely that a certain cat is bare of a collar, but a world that includes such a cat. For if the cat were sans collar, then the neck fur would not be matted down, and would not provide as hospitable a residence for fleas, which means that said fleas would have to seek their domicile elsewhere, and that would have the upshot that . . . .

These considerations suggest that modal skepticism may extend even unto mundane possibilities. For if they come only in world-sized packages, and it seems they must, then, given the finitude of our minds, such possibilities will be unsurveyable and unknowable. They will turn out to be mundane in a two-fold sense: ordinary and close-to-home, but also world (mundus) involving. But we will have to think about this some more.

There are epistemic uses of 'possible.' I inquire, "Is Jones in his office?" The secretary replies, "It's possible." I am not being informed that Jones' presence in his office is consistent with the laws of logic, or with the laws of nature; there is no question about that. I am being informed that Jones' presence in his office is consistent with what the secretary knows. Of course, if the secretary knows that Jones is in his office, or knows something that (she knows) entails that he is in his office, then Jones' presence in his office will be logically consistent with what she knows; but in that case she will not say that it is possible that he be there. So 'possible' in its epistemic use conveys both consistency with what one knows and ignorance. When I say that such-and-such is epistemically possible, I am saying that it is possible for all I know, but I don't know all. Letting 'S' range over states of affairs and 'P' over persons, we define

D1. S is epistemically possible for P =_df (i) S is logically consistent with what P knows; (ii) S is neither known by P nor known to be entailed by anything P knows.

What is possible but not epistemically let us call 'ontologically possible' and define as follows:

D2. S is ontologically possible =_df S is objectively or absolutely possible.

The possibility of the ontologically possible does not vary with any finite subject's knowledge or ignorance. Clearly, if S is epistemically possible it does not follow that S is ontologically possible. Suppose Jones dies on the way to his office for his scheduled office hours. If the secretary is unaware of this development, then it will be epistemically possible for her that Jones be in his office despite the objective impossibility of his being there.

We now define 'conceivable':

D3. S is conceivable for P =_df (i) P entertains S attentively; (ii) P finds no logical contradiction in S.

Conceivability as defined does not entail ontological possibility. It follows that conceivability does not entail any type of ontological possibility such as metaphysical (broadly logical) possibility. But conceivability does seem to provide evidence of ontological possibility.

Now what is the difference between conceivability and epistemic possibility? That they are different can be seen if we reflect on propositions which, if true, are necessarily true. The strong or binary Goldbach Conjecture is that all positive even integers greater than or equal to 4 can be expressed as the sum of two primes. This conjecture is such that either it or its negation is necessarily true. But we don't know which. Yet both are conceivable. I bring before my mind the proposition, attend to it carefully, and find no logical contradiction. Then I bring the negation of the proposition before my mind, attend to it carefully, and find no logical contradiction. I conclude that both the proposition and its negation are conceivable. But only one of them is epistemically possible. For only one of them is consistent with what I know. If the Goldbach Conjecture is true then it is consistent with what I know but its negation (since necessarily false) is not, and vice versa.

Conceivability and epistemic possibility are distinct for a second reason. What I know to be actual is conceivable to me, but what I know to be actual is not epistemically possible for me. Since I know that I exist, it is conceivable to me that I exist; but it is not epistemically possible for me that I exist. Clause (i) of (D1) is satisfied, but not clause (ii): it is not possible for all I know that I exist, for I do know that I exist. From 'I exist,' I can infer 'It is possible that I exist' and 'It is conceivable that I exist.' But I cannot infer 'It is epistemically possible that I exist.' For since I know with certainty that I exist, there is no room here for the ignorance which is an essential ingredient in epistemic possibility.

If S is epistemically possible for P, then P must be partially ignorant about S. For epistemic possibility feeds on ignorance. It is because of Smith's not knowing that Zorn's Lemma is equivalent to the Axiom of Choice that their nonequivalence is epistemically possible for him. There are no epistemic possibilities for an omniscient being or for one who knows all about a certain subject matter. But every ontologically possible state of affairs is conceivable to an omniscient being. For God, all things conceivable are possible, and conversely; but nothing is epistemically possible. Here again we see that epistemic possibility and conceivability 'come apart.'

My disembodied existence is both epistemically possible and conceivable. But because epistemic possibility and conceivability are distinct, one who takes conceivability as evidence of ontological possibility need not, pace Shoemaker, be confusing epistemic and metaphysical possibility. Nevertheless, conceivability does not entail ontological possibility.

But this answer seems open to objection. First of all, there is something immediate about my judgment of identity in a case like this: I don't compare the tree-perceived-yesterday, or my memory of the tree-perceived-yesterday, with the tree-perceived-today, property for property, to see how close they resemble in order to hazard the inference that they are identical. Phenomenologically, there is no comparison and no inference. I just see that they are the same. But this 'seeing' is of course not with the eyes. For sameness is not an empirically detectable property or relation. I am just immediately aware -- not mediately via inference -- that they are the same. One problem concerns the nature of this awareness.

Second, Plantinga's answer allows the possibility that the two objects are not strictly and numerically the same, but are merely exact duplicates or indiscernible twins. But I want to discuss this in terms of the problem of how we perceive or know or become aware of change.

Let's consider alterational (as opposed to existential) change. Do we perceive it with the outer senses? A banana on my counter on Monday is yellow with a little green. On Wednesday the green is gone and the banana is wholly yellow. On Friday, a little brown is included in the color mix. We want to say that the banana has objectively changed in respect of color. But what justifies our saying this? Do we literally see that the the banana has changed in color? How do I know that it is numerically the same banana present on Monday, Wednesday, and Friday? How do I know that someone hasn't arranged things so that there are three different bananas, indiscernible except for color, that I perceive on the three different days? On that extraordinary arrangement I could not be said to be perceiving alterational change.

To perceive alterational change one must perceive identity over time. For there is change only if one and the same thing has different properties at different times. But I do not perceive the identity over time of the banana. I perceive a banana on Monday and a banana on Wednesday; but I do not visually perceive that these are numerically the same banana. For it is consistent with what I perceive that there be two similar bananas. I cannot tell from sense-perception alone whether I am confronting numerically the same banana on two different occasions or two different bananas on the two occasions. If you disagree with this, tell me what sameness looks like. Tell me how to empirically detect the property or relation of numerical sameness. Tell me what I have to look for.

Suppose I get wired up on methamphetamines and stare at the banana the whole week long. That still would not amount to the perception of alterational change. For it is consistent with what I sense-perceive that there be a series of momentary bananas coming in and out of existence so fast that I cannot tell that this is happening. (Think of what goes on when you go to the movies.) To perceive change, I must perceive diachronic identity, identity over time. I do not perceive the latter; so I do not perceive change.

I don't know sameness by sense-perception, and pace Plantinga I don't know it by induction. For no matter how close the resemblance between two objects, that is consistent with their being numerically distinct. And note that my judgment that the X I now perceive is the same as the X I perceived in the past has nothing tentative or shaky about it. I judge immediately and with assurance that it is the same tree, the same banana, the same car, the same woman.

What then is the basis of this judgment? How do I know that this tree is the same as the one I saw in this spot yesterday? Or in the case of a moving object, how do I know that this girl who I now see on the street is the same as the one I saw a moment ago in the coffee house? Surely I don't know this by induction.

If S is a subject and p a proposition, we may define infallibility by saying that S's belief that p is infallible just in case S's belief that p entails the truth of p. Given this definition, it is clear that I have some infallible beliefs, among them, the beliefs that I exist, am conscious, have beliefs, seem to see a coyote, etc. I exist, I am conscious, I have beliefs, and I seem to see a coyote are all contingent propositions.

What about necessary propositions? On a standard understanding of entailment, if q is necessary, then for any p, p entails q. Why? Well, to say that p entails q is just to say that there is no broadly logically possible world in which p is true and q false. Now if q is true in every BL-possible world, then no matter what p is or what its truth-value is, there is no world in which p is true and q false. So if q is necessary, then for any p, p entails q.

So if q is necessary, then S's belief that q entails that q, for any q. Given the above definition of infallibility, anyone's belief in any necessary proposition amounts to knowledge of that proposition. But this can't be right. I cannot be said to know a necessary truth just because I believe it. The impossibility of its being false does not insure that I know it. So the impossibility of my being mistaken when I believe a necessary truth does not constitute my knowing the truth.

But this is perhaps not a problem for the stringent conception of knowledge since what is intended by impossibility of mistake is unthinkability of mistake. Although infallibility entails truth, unthinkability of mistake does not entail truth. To give an example of my own, consider set S and a proper subset thereof, T. Jones believes, plausibly, that for every S and every T, T has fewer members than S. Jones moreover finds it unthinkable that he be mistaken about this. He thinks to himself: If one set is a proper subset of another, then it cannot have the same number of members as the other! No way am I wrong about this! But Georg Cantor (and Bolzano and Galileo before him) showed that the principle does not hold for infinite sets.

So unthinkability of mistake does not entail truth. But now another problem arises. Knowledge entails truth. So how, given that unthinkability of mistake does not entail truth, can primary knowledge be understood as unthinkability of mistake?

I began by defending Butchvarov's conception of knowledge, but now I am at least temporarily at an aporetic impasse. But this is a good old Platonic way to end a discussion -- on an aporetic note. So I'll just stop and think about the problem further later. In any case, it is time for lunch and man does not live by philosophy alone. Primum edere deinde philosophari.

Admittedly, knowledge as impossibility of mistake is a very stringent concept of knowledge. Why should we care to set the bar so high? Why is knowledge in the strict sense important? It is important because there are life and death situations in which one needs to know in order to decide on a course of action.

Suppose I am calmly and rationally contemplating suicide. I have lived a full life, my spouse has died, I have no dependents, and I have accomplished all that I can accomplish. But I am old and worn out and I have been diagnosed with a painful form of terminal cancer. Should I do away with myself? The money saved on expensive treatments could be donated to a worthy cause, etc.

If I know that I am just a material being with no prospects of post-mortem survival, then I have nothing to fear from God or an afterlife, and no need to worry about the moral progress I could make by bravely enduring the terminal suffering. For if I am nothing but a complex natural organism, then there is nothing to hope for beyond this life, and that moral progress could do me no good. Now is it not obvious that knowing whether one will or will not survive is very important? Clearly, what one wants here is genuine knowledge, which implies impossibility of mistake. For one is about to make a very important decision, perhaps the most important decision one can make.

Kant asked four great questions: What can I know? What ought I do? What can I hope for? What is man? One makes a joke of the first question if one demands anything less than certain knowledge. And note that what one wants is knowledge with respect to the other questions, not mere conjecture. Butchvarov puts the point as follows:

Where the truths in question are of the greatest importance, as philosophical truths usually are, where what is at issue is the immortality of the soul and the possibility of eternal damnation, the existence of an external world, of other persons, God, or of a real past, mere evidence, however good, is not enough -- it is knowledge, impossibility of error, that we demand. (CK, 271)

People often say things that are literally false in order to convey important information in a lively way. I once heard a doctor on a radio show say seriously that one cannot have too low a cholesterol level. Taken literally, this is spectacularly false, as false as saying that one cannot be too thin or too rich or have too low a heart rate or drink too much water. (See Hyponatremia: How to Kill Yourself by Drinking Water.) But the good doctor was not joking or bullshitting or trying to say something false in order to induce death in his listeners; his aim was to help them by conveying a truth in an emphatic,colorful, attention-getting way. He was exaggerating for a good purpose. He uttered a sentence which, literally interpreted, expresses a false proposition, a proposition so obviously false that no normal listener could take it to be the proposition intended, and which every normal listener would take as exaggerated way of stating the sober truth that most people need to reduce their cholesterol levels.

Or suppose a coach or teacher says to his charges, "There is no limit to what you can achieve." This is plainly false, but the exaggeration is arguably justifiable since it serves to inspire and encourage. So it may be with ordinary knowledge claims. The chess coach who says to his student, "I know you can win the tournament" is exaggerating. He is using 'know' to express his strong conviction that the student has an excellent chance at winning while at the same time encouraging the student to do his best. But if anything is obvious it is that the coach does not know that the student will win.

In the case of the air-traveler who claims to know that he will be in New York tomorrow, he too is saying something literally false: obviously, he does not know he will be in New York tomorrow. What he is doing is exaggerating in a pragmatically useful way. He is building up his own confidence. After all, there is 'no percentage' in a practical person's dwelling on what could go wrong when the chances of disaster are slight. Practical living involves an element of 'bluster'; thus one confidently claims to know all sorts of things that, soberly viewed from a theoretical perpspective, one cannot know.

If this is right, then there is only one legitimate epistemic sense of 'know,' the stringent sense as impossibility of mistake, and no weak epistemic sense. This seems to imply, contra Rhoda, that there are no analogical epistemic senses of the term. But we shall have to think about this some more.

Suppose Socrates Jones has just secured a teaching job at Whatsamatta U. for the 2007-2008 academic year. Suppose you ask Jones, "Do you know what you will be doing next year?" He replies, "Yes I know; I'll be teaching philosophy." But Jones doesn't like teaching; he prefers the life of the independent scholar. So he plays the lottery, hoping to win big. If you ask Jones whether he knows he isn't going to win, he of course answers in the negative. He doesn't know that he will win, but he doesn't know that won't either. Jones also knows that if he wins the lottery, then he won't work next year at a job he does not like. On the one hand, Jones claims to know what he will be doing next year, but on the other he also claims to know that if he wins the lottery, then he won't be doing what he claims to know he will be doing. But there is a contradiction here, which can be set forth as follows.

Let 'K' abbreviate 'knows,' 'a' the name of a person, and 'p' and 'q' propositions. We then have:

1. Kap: Jones knows that he will be teaching philosophy next year.
2. Ka(q -->~p): Jones knows that if he wins the lottery, then he will not be teaching philosophy next year.
3. ~Ka~q: Jones does not know that he does not win the lottery.
Therefore
4. Ka~q: Jones knows that he does not win the lottery. (From 1 and 2)
But
5. (3) and (4) are contradictories.
Therefore
6. Either (1) or (2) or (3) is false.

Now surely (3) is true, so this leaves (1) and (2). One of these must be rejected to relieve the tension. Isn't it obvious that (1) is the stinker, or that it is more of a stinker that (2)? (2) is an instance of the principle that knowledge is closed under known implication: if you know a proposition and you know that it entails some other proposition, then you know that other proposition. This seems right, doesn't it? So why not make the obvious move of rejecting (1)?

Surely Jones does not KNOW that he will be teaching philosophy next year. How could he KNOW such a thing? The poor guy doesn't even KNOW that he will be alive tomorrow let alone have his wits sufficiently about him to conduct philosophy classes. He doesn't KNOW these things since, if we are serious, knowledge implies the impossibility of mistake, and our man can easily be mistaken about what will happen in the future.

Of course, I realize that there is much more to be said on this topic. Tomorrow is another day.

1. "If dreams were not 'qualitatively' different from waking experiences, then every waking experience would be like a dream . . ." (49)

2. If the phrase 'dream-like quality' were "applicable to everything," then "the phrase would be perfectly meaningless." (49)

Therefore

3. If dreams were not qualitatively different from waking experiences, then the phrase 'dream-like quality' would be perfectly meaningless. (From 1 and 2 by Hypothetical Syllogism)

4. The phrase 'dream-like quality' is not perfectly meaningless.

Therefore

5. Dreams are qualitatively different from waking experiences, and "the notion that the whole of our experience might be a dream" is an "absurdity." (From 3 and 4 by Modus Tollens)

This is what is called a Contrast Argument. The idea is that a term cannot be meaningful unless there are items to which it does not apply. The idea has some plausibility: if a term applies to everything, or everything in a specified domain, then there is a 'failure of contrast' that might seem to drain the term of all meaning. So if very experience were dream-like, then it would be meaningless to say of any experience that it is dream-like.

But I see no reason to accept premise (2) above and the contrast principle on which it rests. The principle is

CP. If a term T applies to everything, then T is meaningless.

But 'self-identical' applies to everything without prejudice to its meaningfulness. The fact that nothing is self-diverse does not make it meaningless to say that everything is self-identical. Or consider a nominalist who claims that there are no universals, that everything that exists is a particular. Is our nominalist saying something meaningless? Clearly not.

Or if a Buddhist maintains that all is impermanent, is he maintaining something meaningless? Must there be permanent entities for it to be meaningful to say of us that we are impermanent? It is not even required that it be possible that there be permanent entities. If it is necessarily the case that all is impermanent, then there cannot be any permanent entities. Nevertheless, it is meaningful to make this strong impermanence claim. One understands what is being said.

Could the whole of my experience be dream-like? This supposition may at the end of the day be absurd; but it cannot be shown to be absurd by Austin's contrast argument. Note also that a proponent of the dream-argument can accommodate a modified contrast principle:

CP*. If a term T applies to everything actual and possible, then T is meaningless.

For a proponent of the dream-argument can admit the possibility of veridical experience; he needn't claim that 'dream-like' applies to every possible experience.

None of us really believe that we are BIVs, and only some really believe that we are creatures of God. Pursuing this parallel may shed some light. So consider:

Q1: How do you know you are not a BIV?
A1: I see that I am not.

Q2: How do you know that you are not a creature of God?
A2: I see that I am not.

These question-answer pairs seem on a par. If the atheist says that he sees that he is not a creature of God, or more generally, sees that the manifest world is not a divine artifact, then we point out to him that this is not the sort of state of affairs that can either be seen to be the case or seen not to be the case. For no matter how intently one observes (whether visually or in any other sense-modality) anything in the natural world, one will not be able to discern whether or not it has creature-status. (By 'creature-status' I mean the ontological status of being a product of divine creative activity.) This is true of the objects of outer perception as well as the objects of inner peception, e.g. memories, desires, sensations, etc. And of course nothing is changed by the use of instruments. If I study the moon through a telescope, I will be in no better position to determine that it is a divine artifact than if I view it with the naked eye. Landing on the moon, taking geological lunalogical samples, won't help either. And the same goes for any intrumental extension of the human senses.

In sum, using 'see' to cover inner and outer sense-perception, whether 'naked' or instrumentally extended, one cannot see whether or not our world is a divine artifact. The same is true with respect to our BIV-status: one cannot see whether or not one is a BIV. For what would one look at, or look for, in order to make this out? Would I stare at the bobcat outside my window and try to isolate a property of the critter that would entail that its appearance to me is solely the result of the machinations of an evil scientist? Or would I look for a property that would entail the animal's not being such a result?

Or is there perhaps some particular thing I must search for the finding or nonfinding of which would resolve the question of whether I am a BIV? You may recall Soviet cosmonaut Yuri Gagarin's suborbital flight in April, 1961. He beat astronaut Alan Shepard to the punch by a couple of months. What remains with me from the news reports is Gagarin's claim to the effect that he didn't see God out there in outer space. I thought to myself at the time, "That dumb Commie thinks that God is a physical object floating around in outer space!" Later on it occurred to me that Gagarin might have made his statement just to please the Party ideologues back on the ground in Moscow. Now just as one cannot resolve the God question by poking around in outer space, one cannot resolve the BIV question by poking around in outer space or inner space.

The atheist who thinks that theism can be refuted by sense-perception is not understanding the issue. The same goes for the theist who thinks that theism can be proven by sense-perception. For the question does not concern observables, but the ontological status of observables. In Heideggerian jargon, it does not concern das Seiende, that-which-is, but das Sein, the Being of that-which-is. In particular, it concerns the status, God-dependent existence.

Similarly, one who thinks that the problem of the external world, which is essentially the problem of the ontological status of the objects of outer perception, can be solved by perception, does not understand the issue. One cannot see that one is or is not a BIV, or dreaming, or being deceived by an evil demon, etc. The question is not about observables, but about their ontological status, their dependence on or independence of our minds.

I can see that the animal in front of me is a bobcat. So seeing, I see that the animal is not a dog. But no amount of observation can disclose that the object appearing to me either has or does not have the ontological status of mind-independent existence.

2. Here is where a form of strong local skepticism comes into the picture. This is a skepticism not about knowledge of past events but about the rationality of beliefs about them. The skeptical argument might proceed as follows. The only data available to my consciousness are temporally present data: these present memories. We may grant that the present is not punctual but has a certain temporal spread. Nevertheless, the events recalled fall outside of this (specious) present which we may assume is rather short. How then can present data assure me of the occurrence of these past events? How can I reasonably infer the existence of any past events from data available to consciousness now? We naturally assume that there is a determinate way things were. The bookshop I visited this morning is the Book Gallery on Main Street, in Mesa, Arizona, and there I chatted with its proprietor, Mike Riley, about chess and Texas Hold 'Em, and other things. All of that happened, and it happened as it happened whether or not I remember it accurately. The natural presumption of memory, built into its very intentionality, is that the events remembered are transcendent of present memorial acts. But how can I be reasonably assured of any of this? Is it not possible that the universe began five minutes ago, complete with dusty books, fouled sparkplugs, and memories? Have I never confused memorial-contents with dream-contents? And if I cannot exclude the possibility that the events reported above were dreamt, how can I be so cocksure that present memories reliably inform me of the past?

3. Now what I want to discuss is not some such skeptical argument as the one in whose direction I have been gesturing, but a certain objection to this kind of skeptical reasoning. The objection taxes the skeptic with epistemic self-refutation. A skeptical argument is epistemically self-refuting if the truth of its conclusion implies that the person forwarding the argument could have no justification for accepting its premises. Applied to the present case, the objection might run along the following lines. "Look, any reason you have to doubt what you did this morning is a reason to doubt whether you accurately retain the earlier phases of your skeptical reasoning. The reasoning takes time to deploy, which implies that its earlier phases fall outside of the specious present. If you question the reasonableness of your belief in what you did this morning, then you must also question the reaonableness of your belief in your skeptical argument." The objector's idea is that the skeptic's argument is epistemically self-refuting. For if belief in the past is unreasonable, then belief in the soundness of the skeptical reasoning, part of which is past, is itself unreasonable.

4. But how much of a problem is epistemic self-refutation for the skeptic? Suppose one is a strong global skeptic, one who maintains that no proposition is reasonably believed. Then of course it follows that the proposition No proposition is reasonably believed is itself not reasonably believed. And the same holds for every proposition that figures as a premise in the global skeptic's argument. None of them are reasonably believed. Nevertheless, if the anti-skeptic accepts the premises, and if the argument is valid, then the anti-skeptic must, it seems, accept the skeptical conclusion despite the fact that the conclusion, if true, entails that no premise of the skeptical argument is reasonably accepted!

Now this is surely paradoxical, but is it contradictory? I don't see that it is. A skeptical argument is given the conclusion of which implies that the argument's premises are not reasonably accepted. But this is consistent with the argument's being sound, i.e., being valid in point of logical form, and having true premises.

We'll have to think about this some more.

1. The Entailment Sense. If X and Y are propositions, and X is true, and X entails Y, then there is a clear sense in which X makes Y true: X determines or necessitates Y's being true. This is not the main way 'makes true' is used but it is clear and legitimate. Indeed, one could say of any true proposition that it is a truthmaker of any proposition it entails. Of course, this is not the standard use of 'truthmaker.' See #3 infra.

2. The Causal Sense. If X is an agent-cause or an event-cause, and Y is a proposition, and X brings about a state of affairs that serves as Y's truthmaker in the sense explained in #3 infra, then there is a clear sense in which X makes Y true: X determines or necessitates Y's being true. Suppose Lenny the lion tamer drives the big cat into her cage. Lenny thereby makes true the proposition, The lion is in her cage, by bringing about (causing to exist) the truthmaking state of affairs of the lion's being in her cage.

3. The Ontic Grounding Sense. If X is a state of affairs, or fact, and Y is a proposition, then there is a clear sense in which X, by obtaining, makes Y true. Thus Al's being seated makes true the proposition Al is seated. This is not a case of entailment, since entailment holds between or among propositions, and Al's being seated is not a proposition but a fact. Facts are concrete, propositions abstract. So Al's being seated makes the corresponding proposition true but without entailing it. Admittedly, facts are proposition-like in that they have a structure that mirrors the structure of the corresponding proposition, but they are not propositions. One way to think of propositions is to think of them as the contents of judgments. My judging that Al is seated has as its content the proposition Al is seated. Pretty clearly, Al himself, all 200 surly lbs of him, is not a constituent of the judgmental content; but Al himself, all 200 surly lbs of him, is a constituent of the concrete truthmaking fact.

Nor is ontic grounding a case of causation. Al's being seated does not cause the corresponding proposition to be true. Propositions are abstracta and abstracta are causally inert. We say that Al's being seated is the ontic ground or truthmaker of the proposition, Al is seated. The idea is that a contingent proposition like the one just mentioned cannot just be true; it is true in virtue of the way the world is. 'In virtue of' is a murky phrase, but the idea behind truthmakers, if not crystal clear, is clear enough, and I should think indispensable. When I say of Al that he is seated on the sofa, or is fat, or is a lazy bum, my talking refers beyond itself to a massive chunk of the physical world that exemplifies certain attributes. What I am saying is true depending on whether the thing I am referring to with 'Al' in fact has the attributes picked out by predicates like 'seated,' 'fat,' 'lazy.'

4. A Mixed Case? Suppose I utter, 'I am now talking.' The proposition expressed by this utterance is made true by my utterance. One could say that my utterance is self-verifying in the way that an utterance of 'I am not talking now' is self-falsifying. But 'verify' (from L. veritas, truth) is ambiguous, and clarity will be served if we distinguish its two senses, one epistemological, the other ontological.

In the epistemological sense, to verify a claim is to ascertain whether or not it is true. This sense is illustrated in Ronald Reagan's classic remark, "Trust but verify." In its ontological sense, to verify is to make true. My saying 'I am now talking' makes true the proposition expressed by the sentence; it is not part of any ascertaining of the truth of the proposition expressed.

Should we say that the utterance-event is the truthmaker or ontic ground of the proposition in question? Or should we say that the utterance-event causes a state of affairs that serves as the truthmaker of the proposition in question? Or should we say that this is a mixed case, one combining elements of #2 and #3?

If the skeptic's point is that we 'know' very little, if anything, in an infallible manner, then skepticism is not a very interesting philosophical position since nearly everyone concedes that much. The most interesting forms of skepticism are those that contend that we can't even have justified beliefs, much less knowledge.

Rather than pursue the question whether or not the skeptic about knowledge maintains an 'interesting' position, let's consider skepticism about reasonable belief. In particular, let's consider an argument by Quentin Smith (Language and Time, pp. 14-18) against such skepticism, an argument Rhoda endorses in his post Quentin Smith Contra Skepticism.

There is a difference between seeing a coyote, say, and seeming to see one. I could easily be mistaken about there existing a coyote that I am seeing, but it is hard to understand how I could be mistaken about seeming to see a coyote. Now the question is this: does my seeming to see a coyote render reasonable my belief that there is a coyote that I am seeing? Or are there grounds for skepticism about reasonable believing? Consider a skeptic who says that seeming never supplies adequate grounds for reasonably believing. Such a skeptic, according to Rhoda-glossing-Smith, proposes:

(1) Even if it seems to S that p and S has no reasons (empirical, logical, or other) to disbelieve p, S is not justified in believing p.

Now, asks Smith, what could possibly justify a belief in (1)? Suppose we grant the skeptic that there are no empirical, logical, or other reasons to disbelieve in (1). Obviously, the skeptic cannot justify (1) on the grounds that it seems right to him, since (1) disallows that move. So the skeptic needs to give positive empirical, logical, or other reasons to show that (1) is true. Call these reasons R.

Now, what justification could there be for thinking that R supports (1)? Again, given (1) it can't be that R seems to support (1). So we need further reasons, R', for believing that R supports (1). But now why should we think that R' supports the claim that R supports (1). Again, given (1) it can't be that R' seems to support that claim. So that claim has to be justified by appeal to still further reasons, R'', and so on. An infinite regress ensues. And the regress is vicious since, as a finite mind, the skeptic "cannot comprehend an infinite number of steps in a chain of alleged justifications" (p. 17). Even if he could, says Smith, the skeptic would still not be justified in believing (1) because he would have no justification for believing that the infinite chain justifies (1). A skeptic who appeals to a principle like (1) is, therefore, hoist on his own petard. If (1) is true, then no one (including the skeptic), could be justified in believing it to be true.

The moral of the story: (1) is false. It's seeming to S that p is, in the absence of S's possession of reasons for believing the contrary, sufficient to justify S's believing that p.

It seems to me that there is a mistake in Smith's reasoning and that Rhoda is wrong to endorse it. The question is whether it it reasonable to believe that (1). I grant that (1) is not reasonable to believe on the basis of its seeming to be true. And I grant that there is no R such that R renders reasonable the belief that (1). Why not just conclude that it is not reasonable to believe that (1)? If it is not reasonable to believe that p, for any p, it does not follow that p is false. So it may be that (1) is true even though it is not reasonable to believe it. One cannot infer, as Smith and Rhoda do, that (1) is false from the fact that it is not reasonable to believe it.

Let me put it this way. (1) applies to itself. If every proposition p is such that its seeming to be true is not an adequate reason for believing it, then (1) has the same property. But surely this does not show that (1) is false. Still, why should one accept it? Well, one might give an argument for it. Suppose one argues from some premises P1 and P2 to (1) and suppose the argument is sound. Then one will have been brought to believe by this argument that (1) is true, despite that fact that the truth of (1) entails that it is unreasonable to believe that (1) is true.

This may sound weird and 'paradoxical' but there is no contradiction. Why can't there be a sound argument for a conclusion that it is not reasonable to believe?

Suppose the seeming to see a coyote is due to an evil scientist's electrode operations. If the inability to exclude this possibility destroys the claim to knowledge, then arguably it also destroys the claim to reasonable belief. If I cannot be sure that my perceptions are caused in the right way, or caused at all, or that there is an external world at all, how can my seeming to see a coyote count as evidence that there is a coyote? (Cf. Peter Unger, Ignorance, Oxford 1975, p. 37.) Now whatever one thinks of this argument, it is an argument, and spelled out, would be an argument for (1). But this is consistent with (1)'s being unreasonable to believe if true.

Greco finds Barry Stroud's interpretation of G. E. Moore's proof of an external world implausible:

According to him [Stroud], the question as to whether we know anything about the external world can be taken in an internal or an external sense. In the internal sense, the question can be answered from “within” one’s current knowledge —- hence one can answer it by pointing out some things that one knows, such as that here is a hand. In the external sense, however, the question is put in a “detached” and “philosophical” way.

If we have the feeling that Moore nevertheless fails to answer the philosophical question about our knowledge of external things, as we do, it is because we understand that question as requiring a certain withdrawal or detachment from the whole body of our knowledge of the world. We recognize that when I ask in that detached philosophical way whether I know that there are external things, I am not supposed to be allowed to appeal to other things I think I know about external things in order to help me settle the question.5

According to Stroud, Moore’s proof is a perfectly good one in response to the internal question, but fails miserably in response to the external or “philosophical” question. In fact, Stroud argues, Moore’s failure to respond to the philosophical question is so obvious that it cries out for an explanation -- hence Malcolm’s and Ambroses’s ordinary language interpretations. Stroud offers a different explanation for Moore’s failure to address the philosophical question: “He [i.e. Moore] resists, or more probably does not even feel, the pressure towards the philosophical project as it is understood by the philosophers he discusses.”6 Or again, “we are left with the conclusion that Moore really did not understand the philosopher’s assertions in any way other than the everyday ‘internal’ way he seems to have understood them.”7 The problem with this interpretation, of course, is that it makes Moore out to be an idiot. Is it really possible that Moore, the great Cambridge philosopher, did not understand that other philosophers were raising a philosophical question? (bolding added)

The bolded sentence is what I want to focus on. I fail to see how Stroud's interpretation makes Moore out to be an idiot. To my mind is it obvious that (i) Moore is the opposite of an idiot and that if not obvious, it is at least arguable that (ii) he does not appreciate the problem of the external world. That is essentially what I said earlier in the introduction to a post on Searle's transcendental argument for realism. Regarding the bit about the two hands, I said that it is ". . . if not a joke, then clear proof, not of the external world, but that Moore did not understand the issue." It is nice to discover that I have Stroud on my side.

It is possible to be exceedingly intelligent and yet not have a feel for, or a sense of, or an appreciation of, one or several or all philosophical problems. Otherwise, how would one explain a positivist such as David Stove? The power of his intellect is not in dispute, and he is classified as a philosopher; but, being a positivist, he is arguably bereft of understanding of what philosophers traditionally were trying to do. For Stove, Plato, Plotinus, Augustine, Berkeley, and a host of other philosophers deemed great, are "dangerous lunatics." (The Plato Cult, p. 184) They espouse views that cannot be taken seriously by any sane person. Let us consider what Stove has to say about Berkeley:

Berkeley held that there are no physical objects: that there was no right hand behind his ideas of his right hand, no wig behind his ideas of his wig, and so on. Indeed, he said, there is nothing at all behind any of our ideas of physical objects, except the will of God that we should have those ideas when we do. Yet Berkeley was a physical object himself, after all – born of a certain woman, author of certain printed books, and so forth – and he knew it. p. ix)

Stove’s misunderstanding is so deep that it takes the breath away. Berkeley did not hold that there are no physical objects; what he gave us was a theory of their ontological constitution. That there are physical objects is self-evident, a datum, a starting point; what they are in their ontological constitution, and how they exist, however, are questions open to dispute. To deny the existence of physical objects would of course be lunacy. But to give an ontological analysis of them in terms of ideas, an analysis that identifies them with clusters of coherent ideas, is not a lunatic project. It is no more a lunatic endeavor than that of analyzing thoughts (and mental states generally)in terms of brain states as one kind of physicalist would.

Suppose we explore this comparison a bit. It is prima facie reasonable to hold that our thoughts are identical to complex states of our brains. (I don’t think this is true, but the reasonable and the true are two, not one.) Accordingly, my thinking about Boston is a state of my brain. Suppose a philosopher propounds the following theory: Every (token) mental state is numerically identical to some (token) brain state. Someone who holds such a token-token identity theory is obviously not denying the existence of mental states; what he is doing is presupposing their existence and giving us a theory of what they are in ultimate reality. What he is saying is that these mental states of which we are introspectively aware are really just brain states; they are not states of an immaterial thinking substance, or states of an emergent entity, etc.

Now if the project of reducing the mental to the physical avoids lunacy, then the same goes for the reduction in the opposite direction. If it is not ‘loony’ to say that the perceiving of a coffee cup is a state of my brain, why is it ‘loony’ to say that the accusative of such a state, the coffee cup perceived, is a bundle or cluster of ideas (to be precise: bundle or cluster of accusatives of mental acts)? Either both of these views belong in the lunatic asylum or neither do.

Note also that if one cannot analyze the physical in terms of the mental, or the mental in terms of the physical, on pain of going insane, then one also cannot analyze the universal in terms of the particular, or the particular in terms of the universal. And yet philosophers do this all the time without seeming to lose their grip on reality.

Take the obvious fact that things have properties. These two tomatoes are both (the same shade of) red. That things have properties is a datum; what properties ultimately are, however, and how they exist, is not a datum but a problem. It appears that our two tomatoes have something in common, namely, their being red. This suggests that redness is a universal, an entity repeated in each of the tomatoes. Some philosophers resist this suggestion by maintaining that, although both tomatoes are red, each has its own redness. Trope theorists, for example, hold that the properties of things are as particular (unrepeatable) as the things that have the properties. They then analyze the seeming commonality of redness in some other way, say, as deriving from a mental act of abstraction, or in terms of classes of resembling tropes.

Metaphysical reductions (of the mental to the physical, the physical to the mental, the universal to the particular, the particular to the universal, the modal to the nonmodal, the dispositional to the categorical, etc.) seem to be as meaningful as scientific reductions. The identification of lightning with an atmospheric electrical discharge; of a puddle of water with a collection of H₂0 molecules; of a light beam with a stream of photons –- none of these identifications are intended by their proponents as lunatic denials of the phenomena to be reduced. There are of course interesting questions about when identifications collapse into eliminations; but the point here is that no denial of existence is intended. To analyze what exists is not to eliminate what exists.

Philosophers, like scientists, are not in the business of denying obvious facts; they are out to understand them. The project of understanding aims at the reality behind the appearance. Stove cannot seem to appreciate this. But this does not bespeak a lack of intelligence so much as a lack of philosophical aptitude.

I bring up Stove to show that one can be be highly intelligent and still lack any sense of a philosophical problem. The same may well be the case with Moore. He is is exceedingly intelligent, but arguably has no feel, or sense, or appreciation, of what philosophers who raise skeptical questions are doing. Of course, there is an external world; of course there are hands and feet; of course it is true that this is a hand; of course, I know as opposed to merely believe that this is a hand. The philosopher of knowledge and reality is not denying these commonsense facts asserted by commonsense people in their commonsense meanings. The philosopher is not out to deny the obvious and mundane -- which is why he cannot be countered by insisting on the obvious and mundane. The philosopher is in quest of true being and true knowledge, and this leads him to question whether what counts as knowledge to commonsense really is knowledge.

The nature of philosophy is itself a philosophical problem, so what I have said above will be found controversial. But I think what Greco says about Stroud's interpretation of Moore, that it makes him out to be an idiot, is a clear mistake.

C. If S knows that p, and S knows that p entails q, then S also knows that q.

For example, if Bill knows that he is hiking to Fremont Saddle, and he knows that hiking to Fremont Saddle entails that he is not a brain-in-a-vat, then Bill knows that he is not a brain-in-a-vat. But according to one form of skeptical argument, Bill does not know that he is not a BIV, and since he does know that hiking to Fremont Saddle entails his not being a BIV, it follows by Modus Tollens that Bill does not know that he is hiking to Fremont Saddle.

Now principle (C) may be false. I take no position on its truth-value. If it is false, however, then every version of skepticism that presupposes it is untenable. But it is not clear that every version of skepticism presupposes (C). Descartes' dream argument would seem to work with a rather more specific principle about dreaming, perhaps this:

D. If S knows that p, then S knows that he is not dreaming that p.

Now suppose that I am holding my hand in front of my face, and that in normal circumstances I would say that I know that this is a hand. Still, I might in a philosophical moment in which my epistemic standards are very demanding wonder whether I really know this. I might argue as follows. If I know that this is a hand, then I know that I am not dreaming that this is a hand. For if I am dreaming that this is a hand, then surely I do not know that this is a hand. Similarly, if I am dreaming that I am asleep in bed, then I do not know that I am asleep in bed, even if it is true that I am asleep in bed. If I cannot exclude the possibility that I am dreaming that this is a hand, then I cannot justifiably claim to know that this is a hand. And I cannot exclude this possibility since there is no mark internal to my experience whereby I can be sure that I am not dreaming. Therefore, I cannot justifiably claim to know that this is a hand.

It is therefore not at all clear that every skeptical argument must rely on epistemic closure principles. Panayot Butchvarov (Skepticism and the External World, Oxford 1998, p. 71) approvingly quotes Michael Williams:

The sceptic can put forward knowing that one is not dreaming as an independent and intuitively plausible condition on acquiring perceptual knowledge of the world, so his argument need not involve any general principle of closure [of knowledge under known logical implication] . . . [This] condition seems likely to prove as plausible as any theoretical account of knowledge conjured up to rebut it.

Are there any texts of Moore in which he explicitly discusses epistemic closure principles? If there are, and he rejects such principles, and is right to reject them, then that still does not show that every skeptical argument is defective.