It’s a mixture of insight and shortcoming. He had an idea that he wouldn’t have called behavioristic, but it became behavioristic as it later developed. (It was supplanted by better work by Fodor and Lewis.)

Philosophical problems in general, and of the mind in particular, have the character of depth and run as deep as the forms of language and thought that set philosophers on the road to confusion. Wittgenstein makes mention of “illusions”, “bewitchment” and “conjuring tricks” performed on our thinking by our forms of language, and tries to break their spell by attending to differences between superficially similar aspects of language which he feels leads to this type of confusion. For much of the Investigations, Wittgenstein tries to show how philosophers are led away from the ordinary world of language in use by misleading aspects of language itself. He does this by looking at the role language plays in the development of various philosophical problems, from some general problems involving language itself, then at the notions of rules and rule following, and then on to some more specific problems in philosophy of mind. Throughout these investigations, the style of writing is conversational with Wittgenstein taking the role of the puzzled philosopher (on either or both sides of traditional philosophical debates), and that of the guide attempting to show the puzzled philosopher the way back: the “way out of the fly bottle.”

Much of the Investigations, then, consists of examples of how philosophical confusion is generated and how, by a close examination of the actual workings of everyday language, the first false steps towards philosophical puzzlement can be avoided. By avoiding these first false steps, philosophical problems themselves simply no longer arise and are therefore dissolved rather than solved. As Wittgenstein puts it; “the clarity we are aiming at is indeed complete clarity. But this simply means that the philosophical problems should completely disappear.”

Wittgenstein offers no traditional solutions to problems of the mind. His solutions are more like linguistic hat-tricks that some how or another “change the subject” from minds to languages about minds. But minds are still going to perplex us even if our language could explain everything. This is a realist argument against the Wittgensteinian account, and I see no way around it.