§49. But what does it mean to say (1) that we cannot define (that is, describe) these elements, but only name them? This might (2) mean, for instance, that when in a limiting case a complex consists of only one square, its description is simply the name of the coloured square.
Here we might (2) say (1) —though this easily leads to all kinds of philosophical superstition—that a sign "R" or "B", etc. may be sometimes a word and sometimes a proposition (4). But whether it 'is a word or a proposition' depends on the situation in which it is uttered (1) or written. For instance, if A has to describe complexes of coloured squares to B and he uses the word "R" alone, (5) we shall be able to say (1) that the word is a description—a proposition. But if he is memorizing the words and their meanings, or if he is teaching someone else the use of the words and uttering them in the course of ostensive teaching, we shall not say (1)that they are propositions. In this situation the word "R", for instance, is not a description; it names an element—but it would be queer to make that a reason for saying that an element can only be named! For naming and describing do not stand on the same level: naming is a preparation for description. Naming is so far not a move in the language-game—any more than putting a piece in its place on the board is a move in chess. We may (2) say (1): nothing has so far been done, when a thing has been named (3). It has not even got a name except in the language-game. This was what Frege meant too, when he said that a word had meaning only as part of a sentence.


  1. The repeated use of 'say' is important throughout the Investigations - and easy to overlook. W. is determined that we should not see his work as outside the ordinary language game, since this would mean that he was working within some alternative, extraordinary, form of life, with an extraordinary language game, which could be called 'philosophy'. There is no such form of life, and no such language game. He therefore wants all the meaningful sentences in his work to be ones that, if ordinarily said, would make sense. It is as though he is trying out how sentences would sound if they were ordinarily spoken.
  2. The repeated use of 'might say' is also interesting, in a way related to our comment on 'say'. We could read it, at least sometimes, as an elision of "might, if we could do it without eliciting misunderstanding...". What kind of misunderstanding? Specifically that which stems from taking his statements as exceptionless generalisations (abstractions; laws) - this being the typical, fanciful, realm of the philosopher. In this way W. is trying to remember to take his own medicine. He does not always succeed; sometimes he is inappropriately (by his own lights) definitive. But he has a pretty good try.
  3. Compare §7, where he writes that "the processes of naming the stones and of repeating words after someone might also be called language-games". (Note the 'might'!) The passages are made compatible by supposing that what W. wants to emphasise is that naming is different from describing, and is commonly a precursor to it. Naming is helpfully seen as one of the stages of getting the game set up (hence the analogy with setting out the pieces ready for the game of chess), rather than playing the game. But (2) this should not be taken by the reader to mean that W. thinks that 'language game' refers to a precisely delineated set of entities, such that naming must be either part of the game, or not. The substantial point is not whether we choose to call naming part of playing the language game, or part of setting up the game. It is an insubstantial in the case of naming as it is in the case of chess (As the players lay out the pieces, might they be said to be "Playing the game"? If not, might the laying out be said to be "part of the game" but not "playing the game"? Who cares? "Does it matter which we say, so long as we avoid misunderstandings in any particular case?" - say, if we have just set out the pieces with a friend when the telephone rings, and we say "Sorry, I can't talk now, we are playing chess".
  4. One "superstition" could well be the (ridiculous) puzzle that would arise if a sign working as a name has one kind of metaphysical link to things, while the same sign working as a proposition has a different metaphysical link. How then could the same sign have two metaphysical links?
  5. Where we already have done the naming, and "know what 'R' means" (as one might say!).