Back again, finally, from the many distractions of the past year. With any luck I’ll now be able to catch up on the long list of subjects that has accumulated in the mean time. I was already way behind before this long absence, and can’t catch up all at once. But let’s get started again.
I finally gave in at some point last year and bought Ontology after Carnap (OUP 2016, ed. by Stephan Blatti and Sandra Lapointe). There are some interesting things in it, that I will be commenting on occasionally over the next couple of months if my time doesn’t get away from me again. Right now I want to focus on Appendix A (“Epistemic vs. Pragmatic Interpretations of the Methodology of Intensions”) of a paper by Stephen Biggs and Jessica Wilson, which is just over two pages long (pp. 98-100) and claims to undermine Carnapian explication.
It argues that explication is, or must be, or can only be, an “epistemic” enterprise rather than a “pragmatic” one, by which they mean that it must consist in continual “refinement” of some original predecessor concept rather than the replacement — on ultimately pragmatic grounds — of a vague and informal explicandum by a more precise explicatum in certain clearly defined contexts (in everyday contexts the explicandum will often remain in use). This (though not quite in those terms) is an old controversy with a long history. To oversimplify drastically, there have broadly speaking been two sides to this question: those who think that — to use Poincaré’s expression — knowledge is imposed by nature on us (call them “naturalists”), and those who think it is imposed by us on nature (“constructivists”). Both have a long history reaching back to antiquity, but these days, constructivists are mostly social constructivists who, following Kuhn, tend to think that the series of explications of a given concept is not held together by anything internal to the explications themselves, but is to be explained only by the social processes that give rise to the concept’s successive explications. As far as their actual conceptual content is concerned, the successive explications not only do not share a language, but are “incommensurable” to the point of mutual untranslatability. At the other extreme, naturalists such as Quine (or, more explicitly, Mark Wilson) think that nature ultimately forces us to shape our language and concepts to be adequate to it. We have no choice in the matter. So “refinement” is all there can be; there is no scope to choose our explications according to our convenience and other priorities, as our hand is forced by the facts of the world, which our language comes to reflect more and more accurately.
There is also, to complicate matters, a middle way, represented by Kant, for instance, by Helmholtz, or by Carnap. In all three of these figures, there is a dialectical interchange (taking a somewhat different form in each) between naturalism and constructivism, a give and take. (I go on about this a bit in my paper about Howard Stein, who is the main living protagonist of such a middle way. Also Michael Friedman has a very nice paper drawing parallels between Kant and Carnap in their relations to their principle naturalistic interlocutors, Hume and Quine respectively.)
Within this long-running story, in any case, Biggs and Wilson are evidently with the naturalists, so characterized, though they don’t mention this (or anything else about the larger context just sketched). What they do invoke is Kripke — and yes, “invoke” is the right word here; their argument is by authority, pretty much at the Confucius-say level. But before I can even get there, I need to straighten something out. In their title, Biggs and Wilson allude to Carnap’s “pragmatic interpretation of the methodology of intensions,” and later seek a motivation for “a pragmatic interpretation anywhere in Carnap’s semantics” (p. 99) and refer to Carnap’s “reason for interpreting semantics as a pragmatic enterprise.” But the methodology of intensions is part of semantics, in Carnap’s taxonomy, and explication is part of pragmatics. Semantics and pragmatics differ from each other just as they each differ from syntax, the third component of meta-discourse. Interpreting semantics as a pragmatic enterprise makes about as much sense to Carnap as interpreting syntax as a semantic enterprise. Now of course the choice among semantic (and syntactic) systems for this or that purposes (or, more locally, the choice of explication for this or that explicandum) is a pragmatic one, and the study of such choices belongs to pragmatics, is part of pragmatic meta-discourse, but that doesn’t amount to “interpreting semantics as a pragmatic enterprise.” Within semantics, once a framework is chosen, there is no scope for choice; here everything is “internal” in the ESO sense. The choice of framework (or locally, the choice of explication), on the other hand, is an external matter. At that point we exit semantics and embark on practical meta-discourse, which for Carnap is part of pragmatics (along with all the other practical meta-studies, including the stuff that’s often classified as methodology and epistemology, or as practical or normative reasoning, or, most generally, as the philosophy of philosophy).
The Kripke-invoking critique by Biggs and Wilson also presupposes an essentialism about concepts (completely alien to Carnap), as it employs a sharp distinction between replacement of one concept by another (explicandum by explicatum) and refinement of a concept “without altering its identity.” But wait — what are the identity conditions for a vague concept of ordinary language? Yes, they can sometimes be given, but certainly not within ordinary language. So Biggs and Wilson’s concept of “refinement” is itself not actually specifiable in real time, but only retrospectively, in the past tense. Which means that whatever notion of essence might be at work here, the essence of a concept can’t actually be known until future refinements reveal it to us. By any traditional (e.g. Aristotelian) standard of essence-talk, that seems like serious cheating.
So bracketing these distractions, let’s finally review Biggs and Wilson’s argument. They identify three steps in Carnap’s conception: (1) explication requires replacement rather than refinement; (2) so “there is never a fact of the matter whether an explicatum is correct for an explicandum”; and so (3) “explication is not an epistemic enterprise.” This conclusion would have puzzled Carnap, since although in his view epistemology and explication were both part of pragmatics (see above), he saw explication as primarily a constructive, engineering enterprise (i.e. part of pure pragmatics), while epistemology is largely empirical (he would have been on board with Quine’s “naturalistic epistemology”), and belongs more to descriptive pragmatics. So to say that “explication is not epistemic” would have sounded to him a bit like “logic has no role in ordinary spoken language” or “pure mathematics is not applied in science” — which are not exactly false (if understood a certain way), but seem fatuous and confused.
He would certainly not have understood (nor do I) how (3) could follow from (2), nor do Biggs and Wilson explain this step. Nor do they explain how (2) follows from (1). Their strategy, rather, is rather to point out that Kripke has discredited (1), which is therefore “simply outdated.” No conceptual engineering is possible, so therefore (2) and (3) just collapse. Their argument, in short, goes like this: Carnap’s ideal of explication is functionalist (in the sense e.g. Huw Price uses this term; see also Rick Creath’s 1994 paper on “functionalist theories of meaning”), i.e. it rejects the reducibility of all meaning to “cognitive” meaning, and holds that there can be discourse (and meta-discourse) not amenable to truth or falsity in the ordinary (“cognitive”) sense. But this is just false, as Kripke has shown, they say, so Carnap’s rejection of applying truth and falsehood to such matters is false. Or am I missing something?