Wittgenstein drastically underplays the role that simulation (I mean, like, imagining another person’s first-person perspective) plays in knowing others’ mental lives. A lot of actual real-life language games concerned with knowing others’ mental lives revolve around the capacity to have an imaginative experience that counts as a correct imaginative simulation of the other person’s experience. Wittgenstein’s not oblivious to this, but rarely lets it enter into his depiction of what is involved in our conception of knowing others’ mental lives.

And, more problematically, when Wittgenstein discusses the claim that one’s present (e.g.) sensation of pain can constitute the content of what one believes another person feels, he treats it as a metaphysical coup rather than an obvious fact about the role that simulation — in this case not imaginative simulation — plays in the game of knowing one another’s experiences. He does then briefly go on to acknowledge this, but then  goes on to never talking about it again.

(I’m not thinking about simulation as a condition for recognizing expressive behavior, but as an element of a partially independent game that has just as much of a claim to being a game of what it is to know someone’s inner life. Like, I can recognize expressions of experiences I don’t take myself to be able to imagine, and when someone’s having such experiences in front of me then in one ordinary sense I know what this person is experiencing and in another ordinary sense I don’t. When for whatever reason — like more experiences or meditation or a drug trip — my imagination expends, it might expend in a way that counts as now being able to simulate that other person’s experience, and know what the other person is experiencing in a sense in which I didn’t know before.)