Earlier I proposed that the most coherent way to characterize intentional acts was as being directed not at real objects but at possibilities, and furthermore that talk of intentionality as if it were directed at real objects really means that the possibility intended has a unique real world counterpart, not that we are intentionally directed at the real object somehow. To me it seems obvious that language and intentionality go hand in hand. It seems reasonable to view language simply as a tool for sharing intentional states, meaning that by communicating I am attempting to put you in a certain intentional state, such that you consider some object, or state of affairs, or abstract concept, as being a certain way (and possibly to believe that it reflects the actual world as well, although not always). For example, in communicating here I am attempting to put you in a certain intentional state towards language, where you consider it as a way of sharing intentional sates. If we put these two ideas together it seems reasonable to say then that the content of sentence, or any fragment of language, is the possibilities that it conveys. (Of course there is the further issue of whether to characterize the content as the possibilities that it was supposed to convey, the possibilities that the listener took it as conveying, or the possibilities that language speakers on average would understand it as conveying, but that is a separate issue.)
But before I go on perhaps I should clarify what I mean by possibility. Here when I speak about possibilities I am referring to a certain kind of object of thought, one in which some details are filled in but others are left open. Thus we can think of this object as a set of possibilities, in which each member of the set has the properties that are filled in, but vary with regards to all other properties. We can think of this as being equivalent to some description, but it is important to keep in mind that this is only an analogy. The possibilities that are invoked by a language are constrained by the definition of the words as understood by the listener, which is the implied descriptive content of the sentence, but they are also constrained by the context, the possibilities already in mind, and by the listeners pre-dispositions. It is also important to note that the real world counterpart of a possibility is not necessarily a perfect fit to the description. The actual mechanics of how the real world counterpart is determined are a matter of convention, meaning that they might vary from person to person, and is governed by psychological factors. However, generally, it is something like the real object that has the “best fit” to that description, within certain limits (for example, if we have a description X an object that looks X without really being X might be the best fit, given that no object that is really X is present).
If we were to characterize this description of language in more conventional terms the possibilities conveyed would be the sense. And, like any theory of language that attaches to words a sense it easily solves a certain problem about intensionality. The problem I am referring to here is the fact that if two words have the same meaning then they should be able to be substituted for each other without changing the truth value of sentences they are in. For example if the meaning of a word was its reference then Superman would be identical to Clark Kent, because they both refer to the same individual. But it is not true that “Louis Lane believes that Superman can fly” has the same truth-value as “Louis Lane believe that Clark Kent can fly”. But, unlike the customary sense / reference position about language I would hold that the meaning of language is only the sense, the conveyed possibilities, and that the reference, in mostly out of the picture, with a few exceptions which I will detail below.
This is not to say that there is no such thing as truth or reference. But I think that when speaking about truth or reference we invoke an additional concept. Namely we have a conception of how to determine the real world counterparts of the possibilities conveyed by language. We determine the reference of a given piece of language by looking for the real world counterpart of the possibilities that piece of language conveys to us. And likewise we say that a sentence is true when the real counterparts of the content of the sentence are in fact the way the sentence portrays them.
But I would argue that our ideas about the real world counterparts are not necessary for language in general, that they are only invoked when the right kind of situation calls for them. Although I can’t prove this directly there are two considerations that make this seem like the best explanation. The first is that our ideas about real world counterparts are just as conventional as the possibilities that we take a given piece of language as conveying, as I have argued here. And this means that we can allow those correspondence rules to vary arbitrarily and language would work in basically the same way until a situation concerning truth or reference arose. Sentences would still convey the same intentions, and so people could talk to each other in the same way, although whether they thought of themselves as talking about true things or merely possibilities would change. A second reason to believe that our ideas about how real world counterparts are determined aren’t connected to language in general is the fact that we can talk about unreal situations in the same way we talk about real ones. I can tell you about president Bush the third, and I can talk about his positions on international trade, and when you hear me you would have no more trouble understanding me than if I was talking about president Bush the second and his positions on international trade. Of course your reactions will be different, but that is because you already have a mental conception of president Bush the second, and my use of language invokes that conception, while in the case of president Bush the third you are intentionally directed at him only through what I say. But if the real world counterparts of the objects of discourse made a difference this probably wouldn’t be possible at all, because talking about an imaginary person would be radically different than talking about a real person.
Of course we can talk about reference as well, but it is important to remember that when we talk about reference we are dealing with a kind of special case. Not only are we intentionally directed at certain possible states of affairs we are also considering certain possibilities about the world. A good example of this is talk about equality. When we say that one thing is equal, or the same as, another, we mean that they refer to the same object. Which in turn means that we are thinking about a situation in which the two possibilities that capture the meaning of the words have the same real world counterpart. Which means that we are thinking of the world as being a certain way, specifically so that the real world counterpart of both is the same object. And this same line of thinking about equality explains how words with the same referent can be substituted for each other in a sentence and preserve the truth-value of that sentence, given that the truth-value depends on the referents and not the meanings. This is of course in contrast to two words that have the same meaning, which implies that when one is substituted for the other the sentence will retain the same meaning. This is why the sentence “Superman can fly” and “Clark Kent can fly” doesn’t have the same meaning, even though they have the same truth conditions. (Some might try to argue that they do in fact have the same meaning, but that is rather hard to believe, since the citizens of Metropolis think that one is true and the other is false, and thus would assert that they don’t have the same meaning. To assert that they do have the same meaning is to assert that speakers of a language don’t necessarily know when sentences do and don’t have the same meaning, which is ridiculous.)
Have I pulled the rug out from under language? I don’t think so. It is true that language, and intentionality for that matter, may seem separated from the real world at this point. But they are separated from the real world, as the fact that we can make things us, share them with each other, and even treat them as true shows. But this doesn’t cut us off from the truth, or from being able to talk about, and think about, the way things really are. It simply means, in some sense, that we can never be completely sure that we have succeeded. But this little doubt isn’t enough to throw everything into question, the evidence heavily favors the idea that most of the time we successfully talk about, and refer to, the actual world when we intend to.