Language Games and Rule-Following

The Philosophical Investigations opens with a quotation from Augustine describing how he learned language as a child: the adults named things, he saw what they pointed at, and he came to understand which sounds stood for which objects. It is a completely ordinary description, which is exactly why Wittgenstein chose it. His entire later philosophy is an attack on the picture buried inside it: that words are names, that naming is the basic function of language, and that meaning is the relation between a word and the thing it stands for.

That picture is also, more or less, the Tractatus. The later work is not a change of mind about a detail; it is the diagnosis of a temptation, offered by someone who had produced the most rigorous version of it available.

The replacement is not a better theory of meaning. It is a redirection: for a large class of cases, the meaning of a word is its use in the language. Stop asking what 'water' names and look at what people do with it: order it, test for it, shout it as a warning, use it in chemistry, use it in poetry. Language is not one uniform mechanism hooking onto reality. It is an enormous collection of different activities, which Wittgenstein calls language games: giving orders, describing, reporting, speculating, joking, thanking, cursing, praying. Words are tools, and asking what a tool really means outside any job is the mistake that generates philosophy.

Then comes the demonstration, and it is the most quoted passage he wrote. Consider what we call games. Board games, card games, ball games, the Olympic Games, a child throwing a ball at a wall. Look for what they all have in common. Winning? Not patience. Competition? Not a child alone with a ball. Skill? Not roulette. Amusement? Not a professional tournament. Do not say there must be something common, he writes: look and see. What you find is a complicated network of overlapping and criss-crossing similarities, like the resemblances among members of a family: build, features, gait, temperament, some shared by some and none by all. The concept works perfectly well without an essence, which means the demand for a definition was never a requirement of meaning.

The deepest part of the book concerns rules. Teach someone to continue the series 'add 2' using the examples 2, 4, 6, 8. They go on: 996, 998, 1000, 1004, 1008. You object. They insist they are doing exactly what you taught. And here is the difficulty: nothing in the finite set of examples excludes their reading. Any finite series is compatible with infinitely many rules. You might try adding an interpretation ('I meant: always add two') but that phrase can itself be interpreted in incompatible ways, and now the regress is running.

Wittgenstein's response is not to find the interpretation that stops it. It is to observe that the regress only starts if you assume that following a rule must be a matter of interpreting it. There is, he says, a way of grasping a rule which is not an interpretation but is exhibited in what we call obeying the rule and going against it. The regress terminates in practice: in a shared form of life, in training, in the brute fact that we go on in the same way. This is either a profound observation about the limits of justification or an evasion, and philosophy has been arguing about which for seventy years.

The private language argument follows from it. Imagine a language whose words refer to sensations only you can have and only you could check; you write 'S' in a diary each time a particular feeling recurs. Wittgenstein argues it could never get going, because there is no distinction available between using 'S' correctly and merely seeming to. Where that distinction collapses, talk of correctness has no content: whatever is going to seem right to me is right means only that here we cannot talk about right. The argument is not a denial of inner experience. It is a demolition of a very old picture of the mind as a private theatre whose contents each of us knows with certainty and nobody else can reach.

What all of this is for is stated repeatedly: philosophy leaves everything as it is. It builds no theories, advances no theses, discovers nothing new. Its work is to assemble reminders about how our words actually function, until a problem that seemed deep is revealed as a knot produced by taking a word out of the practice that gave it sense. His image is a fly in a fly-bottle, buzzing at the glass, and the job is to show it the way out, which is not the same as answering its question.