Abstract: In the Philosophical Investigations, Wittgenstein argues that a language whose words refer to a speaker’s private sensations, and which no one else could learn, is impossible. This article outlines what he means by “private language” and reconstructs his two main arguments: the public check argument, which says a private language has no criterion of correctness, and the training argument, which says rule-following requires public training. I argue that both arguments fail. The sceptical standards they use would, if applied consistently, also undermine public language, and Wittgenstein gives no good reason why only public agreement can ground rule-following. At most he shows that a private language is useless for communication, not that it is impossible.
Introduction
Ludwig Wittgenstein argued that a language only one person could ever understand is impossible. I think his arguments don’t work, and here is why.
Imagine inventing words for your own inner sensations, words that nobody else could ever learn. Could that be a real language? In the Philosophical Investigations, Wittgenstein says no. In this article I explain what he means by a “private language” (PL), set out his two main arguments against it, and argue that both fail.
What is a “private language”?
Wittgenstein’s attack on private language runs through §243–71 of the Investigations (Wittgenstein, Philosophical Investigations, 2009, p. 99). By a private language he means words that refer to one’s ‘immediate private senses’, which only the speaker can know and which cannot be taught (2009, p. 95).
He starts from a simple observation. When I say “I’m in pain”, that is not a description of my crying. It is a more articulate version of the natural expression of pain. As he puts it, ‘the verbal expression of pain [only] relaces crying’ but ‘does not describe it’ (2009, p. 95).
Sensations are also private in a basic way. Only I can know whether I am really in pain, and nobody can cry my cries for me (2009, p, 97). So if other people use sensation words, I can’t know what those words refer to, because I only have access to my own sensations. That is the sense in which sensation talk looks like a private language.
Pointing isn’t enough to give a word meaning
You might think I could give a private word a meaning just by pointing inwardly at a sensation and naming it. Wittgenstein disagrees. Linking a word to an object, called an ostensive definition, is at most a preparatory step. It doesn’t settle how the word is to be used.
Think of the queen in chess. Knowing which piece is called “the queen” tells you little. To understand the queen you need to know how it relates to the other pieces and how it moves on the board. Wittgenstein says a private language lacks this “stage setting”, so we can’t make sense of defining a sensation word in private (2009, p. 98).
The diary argument
This is the heart of the case. Suppose I keep a diary and write “S” every time I have a certain sensation. I fix the meaning of S in private, by focusing my attention on the sensation and naming it. Wittgenstein argues that this doesn’t work, and he gives two lines of attack.
(i)The public check argument
We can’t simply assume that S is just another sensation word. “Sensation”, Wittgenstein notes, ‘is a word of our common language’, so we have no right to treat S as the name of a sensation until we explain how S connects to that word. And there is ‘no criterion of correctness’ for using such a word (2009, p. 98). Nothing separates using S correctly from merely seeming to.
He adds a practical point. A private sensation report is an idle ‘ceremony’ because it achieves nothing: nobody else can correct it or understand it (2009, p 98). Understanding a word should make some practical difference, and in private it never does. Fogelin calls this the “public check argument”: language needs both rules for correct use and a practical role, and a private language has neither (cf. Fogelin, 1987, p. 179).
(ii)The training argument
Language is governed by rules, but what would it mean for one person, alone, to follow a rule? Wittgenstein argues that any action can be made to fit a rule under some interpretation or other. That leads to an endless chain of interpretations, which he rejects. His conclusion is that there must be a way of grasping a rule that isn’t a matter of interpretation, “but which is exhibited in what we call ‘obeying the rule’ and ‘going against it’ in actual cases” (2009, p. 87).
In real cases, someone who follows a rule has been trained to react in a certain way. Through that training they join a shared practice. So language must involve public training, and nobody can train themselves into one. A language that is essentially private, as defined in §243, is therefore impossible for human beings.
standards
My central objection is that the same sceptical standards, applied to public language, would seem to show that public language is impossible too. It is arbitrary to use these standards against private language only. If Wittgenstein can defend public language, the private language user can borrow his defence.
The public check argument
Wittgenstein says we can’t talk about “right” or “wrong” when I report a sensation. This isn’t just memory scepticism. It is a question about normativity: how do we tell reports that are correct from those that only seem correct?
But his own answer for public language is that, at some point, we simply have to accept “the given”. ‘So, what has to be accepted, the given, is – so one could say – forms of life’, and ‘Does it make sense to say that people generally agree in their judgments of colour? What would it be like for them not to’ (Wittgenstein, 1958, p. 226). We can’t imagine everyone in a practice being wrong together. But if we ask what guarantees that the concepts in a practice make sense, the answer is nothing but the practice itself (Fogelin ,1987, p.183).
Put another way, a person’s use of language is meaningful only if it can be checked against public use, yet no further demand can be made of public use itself. The same sceptical question can be asked there: there is a difference between seeming correct and being correct. Agreement across a community doesn’t guarantee that its words are objectively correct. It only shows that everyone agrees.
If sceptical worries don’t sink public language, there’s no reason to deny the same defence to private language. A private speaker can say their words are correct when they are consistent with the seeming recurrence of a sensation, and when they serve a purpose in a private practice, such as helping them predict and tell apart their sensations. So the claim that there is ‘no criterion of correctness’ is unwarranted.
The training argument
Blackburn argues there is no metaphysical difference between rule-following by an individual and by a community. He raises the problem of dispositions: sceptics say that a finite set of dispositions in one person can’t fix what they mean (Blackburn, 1984, p. 289). Kripke shows this with addition. A finite set of past examples doesn’t make it true that I mean “plus” rather than some bent rule like “quadission”, once I face a number too large to have added before (Kripke, 1982, p. 26-27).
But a community’s dispositions are finite too. Wittgenstein grounds correct rule-following in a person’s disposition ‘to think that certain cases fall under the rule and others do not’, and in ‘the fact that the genuine rule follower agrees in his reaction to examples with the members of some community’. Yet, as Peacocke asks, does this really ‘distinguish him from someone who falsely thinks he is following a rule’? (Peacock 1985, p.73).
If rule-following depends on agreement within a community, why trust the community any more than the individual? Suppose Crusoe is alone but uses symbols consistently over time. His past self shapes his present self, so he keeps labelling the same sensation S in the same way. That looks like a kind of self-training. If a community’s practice counts as rule-following despite its finite dispositions, why not Crusoe’s? On this view, a person can train themselves in a language.
If there is no real difference between private and public rule-following, then it is unfair to ask the private speaker to explain the difference between truly following a rule and merely seeming to. They can reply that it amounts to whatever it amounts to in public. Just as a public practice produces shared judgments that can be true or false, so does Crusoe’s.
Part of the problem is Wittgenstein’s claim that language rests on agreement in a “form of life”. He assumes this agreement must be public, with no good reason given. That doesn’t rule out an internal form of life, built on agreement within a rich inner mental life, which would make private rule-following and private language possible.
At most, then, Wittgenstein shows that a private language is useless for communicating with others, since only the speaker can appreciate the words. He doesn’t show that private language or private rule-following is impossible.
Conclusion
My problem with Wittgenstein’s argument is that he applies his sceptical standards to private language without showing why they don’t also threaten public language. If we grant that public language exists, in his sense, then he must also allow that private language is possible. So his argument against private language fails.
References
Blackburn, S. 1984. The Individual Strikes Back. Synthese, 58(3), 281–301
Fogelin, RJ., 1987, Wittgenstein, Taylor & Francis Group, Oxford
Kripke, S.A. 1982. Wittgenstein on rules and private language: An elementary exposition. Cambridge, Mass.: Harvard Univ. Press.
Peacocke, C., 1981. Rule-following: The nature of Wittgenstein’s arguments. In: S.H. Holtzman and C.M. Leich, eds. Wittgenstein: To follow a rule. Boston: Routledge, pp. 72–95.
Wittgenstein, L. 2009, Philosophical Investigations, John Wiley & Sons
Wittgenstein, L. 1953. Philosophical Investigations. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell.
