The Philosophical Investigations
According to the Tractatus language is a picture of reality: language depicts the logical structure of facts. Wittgenstein's repudiation of this view is one characteristic difference between his earlier and later work, between the Tractatus and the Philosophical Investigations. Naturally there was a period of transition, a period when Wittgenstein was moving away from the ideas in the Tractatus before decisively rejecting them, a period when the ideas in the Philosophical Investigations were beginning to take shape. This period is said to have fallen within the years 1930 to 1934. The story of how Wittgenstein came to doubt and then reject the picture theory is this. A Cambridge colleague was the Italian economist Piero Sraffa with whom Wittgenstein often discussed philosophy.1 One day when Wittgenstein was defending his view that a proposition has the same logical form as the fact it depicts, Sraffa made a gesture used by Neapolitans to express contempt and asked Wittgenstein what the logical form
49 of that was. According to Wittgenstein's own recollection, it was this question which made him realize that his belief that a fact could have a logical form was untenable. 1
The repudiation of this beliefhad far-reaching implications for the whole theory expounded in the Tractatus: for the picture theory is a basic one, and, without it, none of the central points of the argument could be sustained. No unbroken line leads from the Tractatus to the Philosophical Investigations; there is no logical sequence between the two books, but rather a logical gap. The thought of the later work is a negation of the thought of the earlier. But of course the Philosophical Investigations is much more than a repudiation of the Tractatus - it is that only incidentally - what gives the second book its importance is that it contains the mature philosophy of the later Wittgenstein. The Philosophical Investigations introduces a new chapter in the history of philosophy. It is not just a continuation or development of the thought of others. It is something wholly original. 2
We have spoken of the well-established idea that a word has meaning in being the name of something. A word, on this view, represents, or refers to something, and to ask what a word means is to ask what it stands for. The word 'apple' is
the name of the fruit that grows on the apple tree, and this fruit is what the word 'apple' means. In the same way the word 'red' is understood to represent the red colour which is seen in various places. No great difficulties are raised by saying that 'a red apple' refers to, and therefore means, a red apple. It becomes more difficult if instead of 'a red apple' I say 'five red apples'. For what does the word 'five' refer to? This is much harder to answer. I can point to an apple and a red coloured patch, but I cannot point to the number five. According to Wittgenstein it would be a mistake to ask what the word 'five' means if it is taken as a question about what the word 'five' names or refers to. Suppose, says Wittgenstein, that I send someone shopping and give him a slip marked 'five red apples'. To the shop-keeper it has the following meaning: he goes to the box marked 'Apples' and opens it; he then looks up the word 'red' in a colour chart and finds a colour sample beside it; then he recites the cardinal numbers up to the number 5 and for each number he takes from the box an apple of the colour which corresponds to the sample. The test of the shop-keeper's understanding what is written on the slip is that he acts as he does. That he understands 'five' is shown by his counting from one to five and stopping (after taking an apple for each number) when he has reached 'five'. If 'four' or 'six' had been written on the slip and he, still in good faith, had acted as he did, it would prove that he had not understood the meaning of these cardinal numbers. What is decisive is how the word 'five' is used. If one asks what the word 'five' names, the question is based on a misunderstanding; the appropriate question is to ask, how the word 'five' is used. 1
Now it is possible to imagine a primitive language-situa-
tion or language-form - what Wittgenstein calls a languagegame - where there would be some reason for maintaining that the meaning of a word is the thing to which it refers. It is conceivable that the conversation between a skilled workman and his mate might consist of names only, that is, the names of the tools needed by the workman and handed to him by his mate every time that he mentions one of them. Such a language - or a language-game - consists of names only, and in order to master the language one must learn what the individual names refer to.