Wittgenstein’s Private Language Argument

Author: Ian Tully
Category: Philosophy of Mind and Language
Word Count: 1000

From roughly §243 to §315 in his Philosophical Investigations, the Austrian philosopher Ludwig Wittgenstein lays out what later commentators have dubbed ‘the private language argument.’1 Wittgenstein imagines a case roughly like the following. Suppose a person is stranded on a deserted island and has managed to bring along a diary. One day he decides (maybe in order to keep himself sane) to begin recording a mark – ‘S’ for example – in his diary whenever he experiences a certain sensation. Whenever the sensation occurs, he focuses his attention upon it (in effect, he tries to mentally ‘point’ to it) and marks ‘S’. Wittgenstein’s conclusion is that it is not possible to meaningfully use a term to refer to a private mental state in this way. Thus, there can be no private language. In this essay, I will briefly explain why.

I. Criteria for Correctness

In brief, Wittgenstein’s complaint is that “in the present case [the speaker has] no criterion of correctness...whatever is going to seem right to [him] is right. And that only means that here we can’t talk about ‘right’.” (Wittgenstein 1953: §258, my emphasis). In other words, in our desert island case there would be no criteria for determining when ‘S’ is used correctly, and when it is not. Two questions now present themselves. First, why must there be some ‘criterion for correctness’ in order for a sign to be meaningful? Second, why think that there are no such criteria in our desert island scenario?

The first claim (that for a word to be meaningful one must be able to apply it correctly or incorrectly) seems very plausible. If it were the case that, no matter how I used a word, I could never use it incorrectly, then it would seem that the word had no meaning. For suppose you asked me what my word meant. Any answer that I gave you would be arbitrary; I could just as easily (and correctly) have chosen some other answer. So it would seem that there must be criteria of correctness for a word to be meaningful.

What then could determine the content of our words, and so determine their correctness conditions? There are several possibilities. A natural place to start – and one which Wittgenstein spends a great deal of time investigating — is with the concept of a rule. Suppose that learning a word involves learning a rule which governs the use of that word. The rule might say, in effect, ‘use ‘S’ in these and only these situations’ or ‘use ‘S’ to apply to these and only these objects’. So, perhaps what provides the correctness conditions for our words are the rules that we learn when we learn the various expressions in our language.

The trouble with this approach is that it only pushes the question back a step. As Scott Soames notes, “the problem…is that such rules are themselves made up of words or symbols which must be understood if the rules are to be any use. Obviously this sort of explanation cannot go on forever,” (Soames 2003: 33). In order to know how to follow a rule, we need to know what the rule means. A rule doesn’t carry its meaning on its face; it needs to be interpreted. But in order to do this we need to invoke yet more rules. It would seem, then, that internalized rules cannot play the role of criteria of correctness, on pain of infinite regress.

So much for rules. But why not say that the man with the diary can appeal to his belief that he is now having the same sensation that he had when he originally marked ‘S’ in his diary in order to determine that he is using ‘S’ correctly? Why can’t the beliefs (or intentions) of a speaker determine when she is using an expression correctly and when she is not? The trouble is that beliefs, intentions, and other contentful mental states must — just like rules — get their content from somewhere, and so we run into the same regress problem that bedeviled appealing to internalized rules (cf. Soames 2003: 34).2

II. Wittgenstein’s Externalism

The upshot seems to be that explanations of intentional phenomena need to bottom out in something non-intentional. We can’t use rules, or beliefs, or intentions, in order to explain meaning (content/correctness conditions) unless we are willing to explain the meaning of our rules, beliefs, or intentions in terms of still further rules, etc. Recognizing this fact, Wittgenstein turns to things outside the speaker’s head in order to ground content. More specifically, he thinks that we need to appeal to the practice and “customs” of our linguistic community (§199).3 In other words, the meaning of a given word for a speaker of a language is determined by the pattern of its use in that speaker’s linguistic community, and a speaker uses a word incorrectly when her use is at variance with this pattern.

Given this, Wittgenstein thinks there can be no (meaningful) private language. The reason should by now be clear. If I wanted to use a term to refer to some private mental state, what would be the criteria governing whether I used the term correctly or not? There would be no public criteria (since the state is private) yet all internal criteria have been ruled out. So, there would be nothing to determine when I used the term correctly and when I did not. So, the term would be meaningless.4

III. A Worry

Thus, Wittgenstein’s “community” view, which locates the determinants of content in the patterns of agreement in one’s linguistic community, is incompatible with private language. But here’s a worry for this view.5 Suppose that everyone agrees that a certain shiny yellow metal is gold. Yet they’re wrong: it’s actually fool’s gold. Thus, when they use ‘gold’ to refer to this metal, they’re making a mistake; their term doesn’t apply in this case, even though everyone thinks it does. That would seem to show there’s something wrong with the community view, and perhaps open the door to private language after all.


1The sections in which later commentators have located Wittgenstein’s ‘private language argument’ “[do] not contain a singular critique of just one idea, namely a private language,” but are instead organized as a somewhat loose collection of observations and remarks concerning a variety of issues (Candlish and Wrisley 2012: sec. 4.1). Thus, Wittgenstein’s argument requires a certain amount of reconstruction and interpretation. Whether the result is something Wittgenstein would recognize or endorse is up for debate.

2It is important to see that the problem with appealing to beliefs is not that our man with the diary might simply have a faulty memory and so not remember what his initial sensation, S, was like (and thus never be sure if his later sensations are the same as S or something new). Rather, the problem is that the content of his belief (viz., ‘this is the same sensation as S’) is itself made up of words (so to speak) which themselves would, if this picture is correct, require beliefs or intentions in order to determine their contents, and so on ad infinitum.

3There is some controversy over whether this is really Wittgenstein’s considered view. See Candlish and Wrisley (2012), esp. sec. 4.1.

4It is worth nothing that it is possible to draw a more skeptical conclusion from Wittgenstein’s remarks, as Saul Kripke does in his Wittgenstein on Rules and Private Language. Kripke argues that Wittgenstein’s remarks on rule following provide grounds for a wide-ranging skepticism about meaning. The worry, very roughly, is that nothing about us determines that we mean x by a word rather than some other thing, y, for any rule which we might appeal to is subject to indefinitely many interpretations consistent with our past use of the term (Kripke 1982). For example, suppose I have never calculated the sum of two numbers greater than 57. Now I am asked: ‘what is 68 + 57?’ Applying the rule of addition, I answer ‘125.’ However, Kripke argues that everything about my psychological states and my past use of the term ‘+’ is consistent with my meaning quaddition, rather than addition, where the rule of quaddition states that if x and y are less than 57, they are added together in the normal way, but if x and y are greater than 57, their sum is 5. A full discussion of Kripke’s meaning-skepticism is beyond the scope of this article.

5Cf. Soames 2003: 39


Candlish, Stewart and Wrisley, George, “Private Language”, The Stanford Encyclopedia of Philosophy (Summer 2012 Edition), Edward N. Zalta (ed.), URL = <http://plato.stanford.edu/archives/sum2012/entries/private-language/>.

Kripke, Saul A. 1982. Wittgenstein on Rules and Private Language. Cambridge: Harvard University Press.

Soames 2003. Philosophical Analysis in the Twentieth Century, Vol. 2: The Age of Meaning. Princeton: Princeton Unversity Press.

Wittgenstein 1953. The Philosophical Investigations. Tr. By G.E.M. Anscombe. N.J: Prentice Hall.


Ian is a doctoral student at Washington University in St. Louis. He also holds an M.A. in philosophy from the University of Colorado at Boulder and a B.A. in philosophy from George Washington University. He is interested in ethics, metaethics, moral psychology, and the philosophy of mind. Aside from philosophy, he likes reading fiction, going to rock concerts, and hiking in the Ozarks.