Response to Little on ontology in critical realism
So that's one point. In terms of my example, we'd say that it is philosophical reflection that shows that you can't be doing what you and everyone else say improv is and also be a hard determinist.
A second point concerns the epistemic question: "How can you know that one general ontology is correct and another is incorrect?" [E.g., how do you know that the world does not contain wholes that are greater than the sum of their parts? Or how do you know that there is no such thing as a real causal power? Or an essence? Etc.] Bhaskar has two completely different orders of response to this question. One is meta-theoretical, and it is mainly designed to dispel confusion. At that level, he says "Do not be tempted to think that the reflexive epistemic question (i.e., the question of the justification of the ontological commitments that one cannot help but have) -- do not be tempted to think that "How do I know that the world is x-like?" is the SAME question as "Is the world x-like?". To conflate the two is, straight-forwardly, a category mistake, he says. He calls this particular category mistake the "epistemic fallacy." He has a lot of terms that are not especially useful, but this one I think is. I will come back to it. But for the moment I just want to note that that is one order of his response to the reflexive question that you flag. The response is a cautionary, meta-theoretical one: "Don't make this very common post-Kantian category mistake, as you consider the matter." Ok.
The second way he responds to "How do you know that the natural world really does have the general features that it must have, if scientific experimentation is the sort of activity that what we agree it is?" obviously hinges on whether or not we think that science as a practice delivers theories that warrant our belief in them. Bhaskar thinks that belief in scientific theory is, indeed, generally warranted. (I think he also thinks that it is a lot easier to deny this in theory than it is to deny it in practice.) The next question we have to address, then, if we are tracking his thinking, is a general epistemic one about what justifies our belief in the superiority of one substantive explanation over another, this particular scientific theory over that one. His answer here includes the following elements: (a) we can't ever be certain; (b) it is always possible that our best theory will turn out to be wrong [it is hard core Popperian fallibilism, here]; (c) scientific facts are are theory dependent; (c) the better theory will probably explain more of the data; (d) we are likely to eventually have a better theory than the current one.
Now, you might be tempted to say "Well, if that's all the epistemic certainty that you can give me re: our best scientific theory, then identifying the implicit ontology of scientific experimentation is about as significant to me for getting my ontology right as is identifying the implicit ontology of Santa's activities on Dec. 24th." But saying that belief in scientific theory is not rational, such that assuming the intelligibility of experimentation is to begin from a false premise, is a very different kind of objection to RB's argument than is saying that he has over-reached philosophically, or is somehow claiming infallible empirical knowledge of how the world is. It is terribly important to be clear about this. RB even has a term for the mistake of thinking that you can read infallible knowledge off of some purported set of "facts." He calls it the "ontological fallacy."
I'll be quiet in a minute, but I just wanted to go back to the meta-claim that it is a category mistake to conflate the questions: "How do I know if the world is x-like?" and "Is the world x-like?" It is worth noting -- as Bhaskar does -- that although these are logically at different levels of abstraction (so it's a category mistake technically speaking no matter what), nevertheless, if one is a subjective idealist (and probably also if one is a pragmatist) then nothing much hangs on having made this mistake. But that is also to say, of course, that to make it more or less with impunity one must adopt a particular metaphysics. I think that if there is one lesson to be learned from Bhaskar (though he is not alone in the history of philosophy in stressing this), it is that there is no metaphysically neutral ground. The minute you say anything, you have said something about how you take the world to be. Post-Kantians (though one might prefer Descartes as the marker) will emphasize that the minute that you say anything about the world, you have thereby thought something about the world. Bhaskar is not trying to get around this. As I said, he's even got a named fallacy for the effort. No contemporary realist would. Ok, maybe some kind of non-reflective empiricist would. But no dialectical thinker would. As I say, I think it's so important to be clear on what Bhaskar did and did not say. I disagree with some of the things that he said and has gone on to say, and I think everyone else should too . But we have to identify real points of difference. It's so great that you are encouraging this discussion!