There is a difference between Rorty and, let's say, me on this, but I don't think it's quite the difference he presents. Philosophical Investigations 500 says that:consider the diﬀerence between the everyday use of epithets like “confused” and “nonsensical” and their technical use by Wittgensteinian therapists. When Descartes mocked the Aristotelian deﬁnition of motion (“the actualization of the potential qua potential”) as unintelligible, he did not try to back up this charge with argument. The term “unintelligible” was just a rhetorical ﬂourish. His point was simply that it would be better to treat “motion” as a primitive term than to try to synthesize mechanism with hylomorphism. When other fans of the New Science called various Scotist and Ockhamite doctrines “nonsense” they did not mean that these authors had failed to attach meaning to the words they used. Rather, they used “nonsense” to mean something like “not worth bothering about, now that Aristotle has been dethroned by Galileo and Newton.” “Useless” would have been as appropriate an epithet as “confused.” (p. 170)
When a sentence is called senseless, it is not as it were its sense that is senseless. But a combination of words is being excluded from the language, withdrawn from circulation.I doubt that any of the New Wittgensteinians would disagree with this, and it's certainly close to Rorty's idea of what 'nonsense' means. As far as meaning is use, meaninglessness (aka nonsense) is uselessness. It's also worth noting that part of the resolute view of nonsense is that 'nonsense' is not a technical term. So I think that Rorty has obscured the difference between himself and the Wittgensteinian therapists.
The real difference might almost be called an ethical one. He cares about usefulness in a historical and more or less consequentialist way (which is one reason why I'm always surprised to find myself agreeing with him). Now that Aristotle has been dethroned his views are nonsense. And that means they are simply not worth our bothering about. The goal, as Rorty says, is "to create a better human future" (p. 169). Therapists, on the other hand, care about individuals or selves. They want to improve themselves and to help, or at least understand, others, and not in a mass way, should anyone think that possible or worthwhile, but as individuals. So it matters to them why Aristotle thought as he did, why Descartes thought as he did, and so on. Understanding comes before cultural change.
This understanding is of a particular kind, though, it seems to me. It is not historical and it is almost egocentric. One puts oneself in the other's shoes, thereby ignoring causal factors. So a Wittgensteinian might want to know why Descartes thought as he did, but she will not do this by taking into account his Jesuit education or the politics of his day. At least, that kind of thing is not likely to be her primary focus. Wittgenstein himself did not go in for that kind of scholarship. His method is suggested by this passage from O. K. Bouwsma's record of conversations with Wittgenstein:
On Thursday evening we met at Black's. It was my turn to introduce the subject. I introduced: Cogito, ergo sum. After I had finished, W. took it up. "Of course, if _______ now told me such a thing, I should say: Rubbish! But the real question is something different. How did Descartes come to do this?" I asked, did he mean what leads up to it in Descartes' thinking, and the answer was: "No. One must do this for oneself." (pp. 12-13)If cogito ergo sum is rubbish then it is because it is useless in something close to Rorty's sense. But the interest of the therapeutic Wittgensteinian (as I understand it--I don't mean to speak as if for other people) is not in what we can use and what should be left behind if we are to improve our culture. The interest is in how an intelligent person can be led to talk nonsense, and how we can avoid the same fate. And that means taking seriously the possibility that Rorty is right when he says of Heidegger's Das Nichts nichtet:
The language game in question is one that Heidegger deliberately and self-consciously created. It is utterly implausible to think that Heidegger might have been led, by a process of elucidation, to ﬁnd himself “confused about his relation to his own words.”Rorty also says that, "Anything will have a sense if you try hard enough to give it one," which is at least close to being true.
But it is possible to struggle to find the right words, and to struggle to think well about something. Heidegger didn't believe that thinking was a breeze, or that whatever he said would be just fine. Neither did Aristotle or Descartes. So what Rorty finds "utterly implausible" is possible, however unlikely it might seem or be. This does not mean that it would be a good use of time to try to imagine how the process might go with Heidegger (although it might be). What it does mean is that there is a way of erring with language that Rorty is (consciously) overlooking. It is not utterly implausible that lesser minds than Heidegger and Descartes might go wrong in this kind of way. And it seems perfectly reasonable--both normal and useful--to call this kind of error "speaking nonsense." So therapeutic Wittgensteinians are concerned with a broader category of nonsense than Rorty is, have a different goal (more personal than political), and therefore a different focus. But I don't think the difference is anything to do with a technical use of 'confusion' or 'nonsense' by the therapists.