http://www.phil-commune.org/phpbb/viewtopic.php?t=560
ZEN
除了象语句与命题这些抽象的对象外，话语与信念也有真值。当我们说一个话语或信念是真的，我们就是在断定该话语与信念的真值，因此，在我看来把“真”当作一个属性并没有坏处——一些信念与话语有而另一些则无。“真”这个词也可被当作真值函项的联结词，在诸如语句“南极洲并不全部在南极圈之内，这是真的”中，陈述的真值取决于其中的否定性子命题。该陈述为真或为假依赖于这个子命题的真假，“这是真的”在交谈中的功能是修辞性的，并没有多余的认识力。
按照我的理解，实在论是这样一种观点，即认为真判断可以解释为与实在的符合关系。这也许是一种不错的主张，如果有人能够以清晰的可理解的途径为真的话语与信念分配对应的实体，并遵循一种可接受的语义学来谈论这些实体。但并不存在这类说明。除非指明对应的方式，我认为没有论据能够支持实在论，同样理由，也无法支持反实在论。在我看来，符合论的真理观和认为话语（或语句）起摹画作用的观点有同样的问题，除非有理解这一观点的人能够像描述真语句的摹画作用那样，指出假语句摹画的是什么。如果真语句不与什么东西符合，那么，真语句也不摹画什么东西。
在这些问题上，罗蒂是对的。在所附的他的论文中与我的主要争点上——真不是探究的目标——他也是对的。罗蒂有时候对我的理解似乎是正确的，探究的目标是获得信念的真值条件，除此以外没有别的办法能使我们的信念坚实。我们不能去做我们知道不能做的。如罗蒂所说，真理并不是有理由的可确定性标准之外的另一个标准。但是说真不是探究的目标却有一个小小的误解：真是这样意义上的目标——我们开始思考时不一定有把握，但我们有更多的证据我们就可能更正确。在股票市场上，有的人根据内部消息投资，有的人则在大家跟风前先去请教掌握内部消息的人。很多时候罗蒂看来赞同把真与有理由的可确定性等同的观点——实用主义者怀疑真与有理由的可确定性之间存在差别。既然他能认可有理由的可确定性是一种标准，为何他不因此接受真也是一种标准呢？
不过，罗蒂认识到说我们的信念是有理由的可确定的与说我们的信念是真的之间存在着差别。如他所说，可确定性联系到更多的事情：证据的有效性、证据的可得性、能否被接受、证据支持的标准等等。而真并不与这些事情相关，我很高兴罗蒂在这个问题上并没有附和他的追随者。当罗蒂谈到对真概念的警惕，我猜想他的意思是提醒人们，可确定的并不必定代表是正确的。实际存在的差别难道不是明显的？如果差别无关紧要又何需提醒？
罗蒂认为我持真理的冗余论，虽然我拒绝符合论，但不能因此把我的观点解释为冗余论，以及否认真是与可确定性不同的标准。不过我仍然反对把真当作具有认识论意义的概念。只要有可能罗蒂就把我描绘成一个冗余论者。不清楚他还想把我带到多远。可能像他发现詹姆斯所建议的“当你知道了关于可确定的一切，你也知道了真的一切”那样。如果真不同于可确定性，如罗蒂所承认的，即使你知道了可确定的“一切”，你也没有掌握真的全部。我认为关于真有很多可探究的。像罗蒂所指出的，我相信只有联系于说话、信念和可评估的意见，才能得到关于真概念我所要寻找的内容。
罗蒂承认，他对于是“把真归结为可确定性”还是接受塔尔斯基的去引号纲领犹豫不定。但他知道真与可确定性并不完全一样，而且他关于塔尔斯基的论述是错误的。《真之理解》内容丰富，并不仅限于去引号。当然你可以说每个T等式也具有去引号的作用。幸好罗蒂虽不怎么情愿但还是允许我保留我完整的真理观。
我希望我的论述是充分的，但我怀疑还未做到。罗蒂想把我的主张——如“我们必须认为任何人的大部分信念与我们相符”和“构成真就是构成对我们的可确定性”——解释成，我们大部分的常识和基础信念是真实的。我同意这些主张，但并不同意他们对我理论的说明。我们所持有的信念是由感知和其他经验引起的知觉信念，这是我们掌握的大部分真理的方式，它们的内容由它们的原因决定。不能在这里展开这个论点。我在关于真的常识性概念中所相信的要点是：正像我们所认为的，存在着人、山、骆驼和星星，这些对象和事件具有我们知觉到的那些特征。我们的信念虽然只是我们的，但这并不意味着它们不是真实的，不能有效描述客观事实。
[原文]
Aside from abstract objects like sentences or propositions, if such there be, the only things in this world that are true are some utterances and some beliefs. When we say an utterance or a belief is true, we predicate truth of that utterance or belief, so I see no harm in holding that truth is property: some beliefs and utterances have it and some do not. The word ‘true’ also operates as a truth-functional connective, as in ‘It is true that Antarctica is not wholly within the Antarctic Circle’: its truth table is like that of negation with the truth values reversed. Because, in this use, it maps true utterances onto true, and false onto false, its function in conversation is rhetorical. Cognitively it is redundant.
Realism as I understand it is the view that predicational use of truth can be explained in terms of a relation of correspondence. This would be an interesting claim if anyone could come up with an intelligible and illuminating way of individuating the entities to which true utterances or beliefs correspond, along with an acceptable semantics for talk about such entities. But there is no such account. Until there is, I see no point in declaring oneself a realist, or, for that matter, an anti-realist. I see no difference between a correspondence view of truth and the idea that utterances (or sentences) ‘represent’, except, perhaps, that if one understood the idea, one could talk of what false as well as true sentences represent. But if there is nothing for true sentences to correspond to, neither is there anything for them to represent.
On these matters, Rorty has me right. He also has me right on the main contention of his present essay, that truth is not a goal of inquiry. This seems to me correct if understood as Rorty sometimes puts it: it is a goal of inquiry to find substantiating evidence for our beliefs, and there is noting more we can do in trying to firm up our convictions. We can’t try to do what we know we can’t do. As Rorty says, Truth isn’t a norm in addition to the norm (norms?) of justification. But it is a little misleading to say truth isn’t a goal: it’s the same goal, since we are bound to think, not always correctly to be sure, that the more evidence we have the more apt we are to be right. People who invest on the basis of inside information on the stock market, or who consult a tout before the races certainly believe so. Much of the time Rorty appears to endorse the view that truth and justification are identical: pragmatists are ‘suspicious of the distinction between justification and truth’. He also thinks justification is a norm. Shouldn’t he conclude that truth is a norm?
However, Rorty knows there is a difference between our beliefs being justified and our beliefs being true. As he says, justification is relative to many things: the availability of evidence, the expense of obtaining it, our audience, our standards of evidential support, and so on. Truth is not relative in these ways, and I applaud Rorty for not going along with many of his followers who would say that it is. When Rorty speaks of the ‘cautionary’ use of the concept of truth, I take him to mean that it is often useful to remind people that being justification isn’t necessarily being right. Is it obvious, then, that there is no sense in which the distinction matters to practice? Why remind someone of a distinction if it doesn’t matter?
Just the same, Rorty may be justified in calling me a quietist with respect to truth, since I reject correspondence theories, don’t think the idea of representation can be cashed in, and agree that truth is not a norm in addition to justification. I also disavow all other attempts to treat truth as an epistemic concept. If that is sufficient, in Rorty’s eyes, to make me a quietist, so be it. But it’s not clear how far he wants me to go. For example, he finds James’s suggestion that ‘once you understand all about justification, you understand all there is to understand about truth’ (Rorty 1995, p. 282) persuasive. If truth differs from justification, as Rorty seems to allow, then there has to be something about truth you don’t understand when you understand ‘all about’ justification. I certainly think there is a lot more to say about truth. As Rorty recognizes, I believe what I call the content of the concept of truth can be brought out only by relating it to speech, belief and the evaluative attitudes.
Rorty confesses that he wobbles between ‘trying to reduce truth to justification’ (ibid.) and embracing some from of minimalism like ‘Tarski’s breezy disquotationalism’ (ibid.). But he knows truth isn’t identical with justification, and he is wrong about Tarski. The Wahrheitsbegriff is about as far from breezy as you can get, and the reason is that it isn’t a form of disquotationalism. You can say if you want that each Tarski-style truth definition has an element of disquotationalism, but none of these definitions is a definition of truth, as Tarski proves. Fortunately, Rorty seems to be willing, perhaps reluctantly, to let me have my more full-blooded view of truth.
I hope this is pretty much all that needs to be said. But I have an uneasy sense that it may not be. Rorty wants to explain away my claim that most of our simplest and most basic beliefs are true as ‘saying that most of anybody’s beliefs must coincide with most of our beliefs’ (ibid., p. 286) or that ‘the pattern truth makes is the pattern that justification to us makes’ (ibid.). I agree with these claims, but do not agree that they give my reason for holding that most of our beliefs are true. The beliefs I have in mind are our perceptual beliefs, the beliefs that are directly caused by what we see and hear and otherwise sense. These I hold to be in the main true because their content is, in effect, determined by what typically causes them. This is not the place to expound the argument. The point is that I believe in the ordinary notion of truth: there really are people, mountains, camels and stars out there, just as we think there are, and those objects and events frequently have the characteristics we think we perceive them to have. Our concepts are ours, but that doesn’t mean they don’t truly, as well as usefully, describe an objective reality.
