Inflate and explode, analyze or explicate

Over at The Splintered Mind, Eric Schwitzgebel identifies what he calls the inflate-and-explode maneuver. Abstractly, the move is this: “Assume that things of Type X must have Property A, and then argue that nothing has Property A.”

Schwitzgebel is especially interested in the case of consciousness. On many accounts, one is supposed to have infallible access to the contents of one’s consciousness. However, one doesn’t have infallible access to anything. Having thus inflated consciousness with the pompous swell of infallibility, one blows it up— there is no such thing as consciousness!

Schwitzgebel conjectures that philosophers who think consciousness involves infallible access are nevertheless more committed to there being such a thing as consciousness than they are to conscious reflection being infallible. If they were to admit the truth of falliblism, they’d give up the inflation rather than accept the explosion.

The upshot of this, Schwitzgebel thinks, is that we ought to accept that consciousness described in modest terms exists. He writes,

“We know some examples of consciousness. We know that these examples have an obvious and important property in common, which we dub… ‘consciousness’… There is not much reasonable doubt about the existence of such examples or the fact that they have this property in common. Definition by example is a relatively safe and theoretically innocent way of characterizing consciousness…”

Of course, I concede that many familiar examples of consciousness describe genuine phenomena. Something’s going on when I hear the honk of a distant goose, when I smell the savory aroma of a cooking ham, when I feel an ache in my knees, when I imagine a rhombus, or when I regret eating so many peanuts.

The strategy Schwitzgebel suggests is to define “consciousness” as the feature that those have in common. Figuring out what that feature is will require serious empirical and philosophical work. If a particular account of that feature ends up not describing anything, then that account should be seen as an unnecessary inflation of consciousness and abandoned.

This shares important similarities with the method of conceptual analysis, which starts with a concept like knowledge and works to provide an account of what comprises knowledge. If a particular analysis fails, that’s too bad for the analysis but no mark against the concept. The flipside is that the concept is taken as given.

The parallel between analysis and Schwitzgebel’s approach to consciousness isn’t perfect.¹ But I had a reaction against Schwitzgebel’s approach which I think parallels my animus toward analysis.

An alternative to analysis is explication.² Instead of taking the concept as given, we begin by noting that the concept both does some useful work and is obscure in many ways. The move of explication is to suggest a successor to the prior concept which will be more clearly specified yet able to do at least some of the work of the prior concept. An explication might be offered just for some specific contexts, and a useful concept might admit of multiple useful explications.

Back to the case at hand: I’m not really comfortable with Schwitzgebel’s claim that examples of consciousness “have an obvious and important property in common.” It’s also assuredly true that these phenomena have something in common. The problem is that any collection of things will have indefinitely many features in common.

Schwitzgebel’s claim can be read as presuming that consciousness is unified, in which case it’s making the analytic assumption that there’s that one thing to be analyzed.

Reading it in an explicative vein, it’s really just a bet about how the explication will turn out. There may be many important properties shared across the homey examples of consciousness experience, but that would lead us to expect multiple explications which capture different important features— each shared by many but perhaps not all of them. Read in this way, Schwitzgebel’s claim claim that the examples “have an obvious and important property in common” is no longer “safe and theoretically innocent.”

To be fair, he writes that it’s “relatively safe and theoretically innocent” (my emphasis). So maybe this is a friendly amendment rather than an objection.

1. There’s the important difference that traditional analysis is supposed to be an a priori enterprise that tells us about our concept but not about the world. Schwitzgebel’s approach is empirical and meant to tell us about the phenomena in the world. Maybe this is what happens to analysis after Kripke. When we can designate a thing in the world in a way that there are necessary truths about it, there’s no reason to settle for just what’s in our conceptual scheme— even if the enquiry becomes a posteriori.
2. The term is Carnap’s, and it gets a lot of use in Carnap-inspired philosophy of science. Quine’s term for it is ‘regimentation.’