Wittgenstein defines language games in the Blue Book as "ways of using signs simpler than those in which we use the signs of our highly complicated everyday language" (Wittgenstein, 17). The example he gives using the grocer and the apples shows an interaction that uses the same words and produces the same results as we would use and get from our ordinary language, but in the language game, the connection between words and actions is far more transparent.

In the Blue Book, Wittgenstein seems to define language games as more primitive forms of language, or the building blocks of ordinary language. (He will later abandon the idea that language games are incomplete replicas of ordinary speech and will no longer consider instances of language use in language games to be simpler, clearer instances of the same use in ordinary language.) The discussion here of language games is closely linked to two other themes that play a large role in Wittgenstein's philosophy: the anti-scientism expressed in his disparagement of the "craving for generality," and the notion of family resemblance.

It is important to note that Wittgenstein is not anti-scientific in the sense of disparaging the results of science or claiming that they are not valid. His criticism is leveled at the notion that science has given us a complete and satisfying explanation of the way things are, and—most importantly to his argument—at the indiscriminate urge to apply the scientific method even where it does not belong. We saw in the previous section that Wittgenstein disparages psychology because it purports to conduct scientific investigations of the mind, when we have not yet even clarified what the mind is. In this section, Wittgenstein remarks that the scientific method has also been used in philosophy, leading us to look for general rules and strict definitions of words, when such rules and definitions do not apply to language. Certainty of the kind we find with regard to physical phenomena cannot be found in language, and we should not seek it.

Wittgenstein will later refine this idea of a "craving for generality." Here, he sees the craving for generality purely as an offshoot of the scientific method, but he will later suggest that this craving has its source in linguistic confusion.

The notion of family resemblance counters the idea that all uses of a word have an essence or defining properties in common. If we think of all the members of our family, we can remark that they share certain distinctive traits or features, but there is no one trait we could point to that every member of our family shares that distinguishes it from the rest of the human race. There might be a tendency toward a pointed nose or a round jaw, but it is not even necessary that all members of the family have these features. They can still recognizably be members of the same family if they have other characteristic features.

Wittgenstein discusses the words "know" and "unconscious" in talking about "unconscious toothaches." He appeals to the notion of family resemblance. If we want to distinguish between tooth decay that hurts and tooth decay that does not hurt, there is nothing wrong with calling the latter kind of pain "unconscious," or pain that we do not know of yet. The problem arises when we think that this use of "unconscious" must be analogous to the kind of "unconscious" I mean when I refer to being unconscious of someone standing in my peripheral vision, as if an unconscious toothache is a pain that I am not yet aware of feeling. There is a family resemblance between these two uses of "unconscious," but the two uses of the word differ.