Page 95 - Encyclopedia of Philosophy of Language

to be used—its actual semantics. Private objects play no role.)
Here is a second argument. Let Pol judge, 'A is F.' Now ask: in principle, might Pol be wrong in judging this? Might he have mistaken what his own 'is F' required for truth of A, or whether that requirement was satisfied? If not, then his words are reduced to the level of an inarticulate cry. For they must be governed by the rule: they were spoken correctly exactly wher- ever Pol felt inclined (enough) to speak them. If so, then suppose his words said of A that which is true of A. What might such a fact consist in? Not in Pol's reactions to the words, as exhibited in hisjudgment. But then, evidently, in nothing else either. (This being private language, only his reactions matter.)
Neither of these arguments, one might object, is available to Wittgenstein. Each depends on a principle he explicitly rejects—the first on a rejected epis- temology (see, for example Sect. 84); the second on the rejected idea that facts of some genre, to be such, need to consist in anything else. (See, for example, Sects. 135-37.) The objection vanishes, though, when we ask what entitles Wittgenstein to reject these prin- ciples. In the first case, for instance, there is normally a distinction between real doubts, which might show you do not know, and merely imaginable doubts, which show nothing. Where such a distinction is draw- able, the mere possibility of misremembering what a word means cannot show someone not to know that it has been used correctly. But that distinction rests on a background 'system of natural reactions': we take certain doubts seriously, others not, given which there are facts as to some doubts being real, others not. That background is unavailable in the case of private language. So there can be no appeal there to such a distinction; no such thing as 'the reasonable way of drawing it.' Given such lacunas in the private case, the private linguist lacks the means for resisting either of the otherwise noncompelling arguments.
We normally suppose, rightly, that we can tell a hawk from a handsaw; further that if in some case we have failed to, there will be something to show either that we missed some pertinent fact, or that we judged unreasonably. The above argument shows the import- ance of both suppositions. Without room for some- thing to override our judgments, there would be no judgment. The second supposition is safe just where there are means for drawing a distinction between what is the reasonable view of, or reaction to, a situ- ation, and what is not. A private linguist may seem to have semantic reactions; to take his private language to have one semantics rather than another. Without the resources to distinguish reasonable under- standings of it from others, he can in fact be doing no such thing; nor could his language have a semantics, except perhaps that of an inarticulate cry. Those appear to be resources he would lack.
Frege encouraged the view that if we could just get
words to have the right properties—e.g., to have a proper and univocal sense—then their doing that would settle, effectively, all questions as to how and where the words apply correctly. Such a perfect language, like a perfect machine, would run on forever under its own power; facts as to what bits of it were true would depend in no way on our, or on any, reactions to those bits; the language would have sufficient resources in itself to generate those facts on its own. The illusion that there might be private language rests on the idea of language functioning that way: whether a private word is true of an item depends not at all on anyone else's reactions, and, if the private linguist might be mistaken in such matters, not on his reactions either. Rather, properties intrinsic to the language are conceived as carrying all the weight in determining that the facts about its application are thus. Wittgenstein shows this Fregean ideal for language to be a chimera; no words could have that property. That would be enough to show up private language as an illusion. Conversely, dealing directly with private language is a way of showing why Frege's conception could not be right.
6. Private Language and Rules
Wittgenstein first makes the crucial point about this chimera in his discussion of rules and their require- ments (see Sects. 84-7). For any rule, there are various conflicting things, each of which would count as fol- lowing it correctly, if only this or that understanding of the rule were the right one—the one it in fact bore. We are often capable of seeing what following a rule in a specific case requires. The fact that the rule does require that is not independent of our seeing this. For that fact depends on that understanding of it being the most reasonable one. But in matters of reason- ableness, we, or beings like us, must be the ultimate arbiters. It is no good appealing to anything like a rule to fill the gap we would leave at that point. For a word to have properties which, all on their own, decide that it applies correctly to (in) exactly these cases and no others is for it to be governed by a rule that requires ittobeappliedexactlythere.ButWittgenstein's point is that no rule, in isolation, can do that job. For it does that only given sufficient facts as to its proper understanding; but there are no such facts without a background of natural reactions to the rule, by reason- able beings, for those facts to rest on. One way to see the point is to consider language with all such background cut away. Private language is such language. It fails to be genuine language precisely because it lacks such a background. The conception of semantic properties which Wittgenstein's dis- cussion of rules supports thus becomes compulsory. Language in need of no one (for the standards of its correct use) is language for no one.
See also: Family Resemblance; Rules; Wittgenstein, Ludwig.
Private Language
73