630. Examine these two language games:
- Someone gives someone else the order to make particular movements with his arm, or to assume particular bodily positions (gymnastics instructor and pupil). And here is a variation of this language-game: the pupil gives himself orders and then carries them out.
- Someone observes certain regular processes—for example, the reactions of different metals to acids—and thereupon makes predictions about the reactions that will occur in certain particular cases.
There is an evident kinship between these two language-games, and also a fundamental difference. In both one might call the spoken words “predictions”. But compare the training which leads to the first technique with the training for the second one.
Ludwig Wittgenstein, Philosophical Investigations
Though there’s lots of discussion of intention and willing in Philosophical Investigations, this is the only passage I’ve found that seems to address free will issues, even though Wittgenstein doesn’t appear to acknowledge this, or at least not directly. I think they don’t come up because Wittgenstein is mostly concerned with thinking agents—what they’re doing and what they believe they’re doing—and not so much with the difference between agents and non-agents. (Except to to claim things like it would be silly to say that a stone has feelings.) However, in this passage, there is a clear difference between and agent and a non-agent. One account of the fundamental difference between the two language-games is that the gymnast could always refuse to do what the coach says, but the metal can’t refuse to react to an acid.