In my previous remarks about Ladyman and Ross I wondered whether the differences between them and Carnap were (largely or entirely) a matter of terminology. What Ladyman and Ross call “metaphysics” Carnap entertained as a programmatic constraint on the language chosen, or developed, as an agreed common language of science, i.e. that it enable us to unify all the disperate knowledge from all the special sciences into a single coherent story. Does it matter whether we call this “metaphysics”? For Carnap, it didn’t really. He obviously was wary of the word “metaphysics” but was quite clear that a good deal of traditional metaphysics (he mentions Aristotle, Leibniz, Kant, Peirce, and Whitehead) could easily be interpreted as engineering work on our conceptual apparatus, rather than taken at face value in the material mode, i.e. as pertaining to some sort of ultimate “reality.” He would no doubt have taken Ladyman and Ross in the same spirit, and interpreted their commitment to objective modality and their “ontic structural realism” as a proposal for constraining the language of science — a proposal Carnap would undoubtedly have seen a lot of merit in.
Still, he would have wondered what was at stake in their insistence that their proposal is a form of “realism,” of what van Fraassen calls “weak metaphysics” that takes the grammatical form of assertions about what is and is not. Here too, though, they wriggle out by explaining, with reference to van Fraassen, that their weak metaphysics is “stance-relative” (i.e. to be understood relative to their own “scientistic” stance, which they see as a fusion of van Fraassen’s “empiricist” and “realist” stances). Their apparent assertions, they say, “should be understood not as doctrines but as proposed provisional commitments for living out the stance.” (All Things, p. 65)
This casts a rather strange light on their overall argumentative strategy. Everything they actually assert is to be taken as cogent only relatively to a stance they do not seriously attempt to give reasons for.
Now this strategy is, in fact, often attributed to Carnap himself; theories, languages, frameworks — all the components of our whole cognitive apparatus — are to be chosen by the standards of a value system that can’t itself be justified or even discussed rationally; it is simply chosen arbitrarily by personal or collective whim. This is what Carnap’s supposed “emotivism” about values is often taken to amount to. Now I’ve always thought this is a complete misunderstanding of Carnap, and as I’ve recently argued (now with a Carnap document from 1958 to back me up), Carnap’s overall conception does have the resources to make it possible to argue rationally for or against comprehensive value systems (such as what van Fraassen and Ladyman-Ross call “stances”). So if you accept that (and of course you may well not, but please tell me why), then it should be obvious that this Carnapian conception is more rational, in the sense of being more in the Enlightenment spirit that Ladyman and Ross recommend and embrace, than Ladyman and Ross’s own strategy, which requires the acceptance of the scientistic stance without rational justification or even discussion.