There has been some interesting discussion of 'experimental philosophy' due to a post at the anthropology blog "Savage Minds" (always an interesting read). The post is a bit impressionistic, but it covers a great deal in an intriguing way; it is favorable to the general spirit of 'experimental philosophy', but is critical of certain aspects of it. One of the things I thought was interesting is that the author, Christopher Kelty, hits squarely one of the things that has annoyed me about the movement, at least in a minor way, despite the fact that I am sympathetic to the basic idea. This is that, amidst the interesting work, there has been a great deal of branding and playing of the Game of Authority -- or, as I would have put it, less clearly, rhetorical salesmanship. (Kelty, it should be noted, takes a less cynical view of both than I do.) You'll notice that I keep the phrase in quotes. I do early modern philosophy; there's a great deal of meaning in that phrase, experimental philosophy, and it is absurd, utterly absurd, that anyone expects it now to apply, with its heavy associations, to what is in fact a very, very narrow methodological program. It's not experimental philosophy; it's merely survey-assisted research with respect to one philosophical idea, namely, intuition. And 'experimental philosophers' don't, contrary to their grandiose pretensions, burn the armchair: they get up long enough to do survey and then sit right back down in it. I don't see how anyone looking with an objective eye at the discussion on the Experimental Philosophy blog could say otherwise. I don't think there is anything wrong with this, either the surveys or the armchairs. But I dislike the rhetoric, with its overselling and its rank-pulling.
A second thing I found interesting about the post was that I agree with the suggestion that it would make more sense for experimental philosophy, given its interests, to be oriented "towards anthropology and towards fieldwork—not towards cognitive science, evolutionary biology and statistics". One of the very unsatisfactory things about surveys for what the 'experimental philosophy' crowd claims to be doing is that it is difficult to get an adequate grasp of what is going on in someone's head from a survey. As the old joke goes, surveys just measure behavior in response to surveys. To get a full basket of information relevant to understanding folk intuitions, in the sense(s) to which philosophers have appealed, you need to know not just what people are inclined to say about scenarios under survey conditions; you need to find out what's going on when they reason and act. You need, in other words, not merely surveys but you need the sort of things cognitive scientists on the one side and anthropologists on the other have already been doing for quite some time now. And it seems to me that, for the sort information needed to inform the kinds of philosophical discussion that involve appeal to folk intuitions, the lacuna on the anthropological side is far and away the more serious; and the lack of good ethnography is a more serious lapse in philosophical appeals to intuition than the lack of rigorous surveys. If you are trying to gauge folk intuitions about, say, modality, you need far richer data and analysis than surveys alone can provide. (It goes without saying, of course, that surveys can still provide a useful supplement.)
My primary doubt about Survey-Assisted Intuition Research, however, is something else entirely. I think SAIR is actually a very, very conservative reform of philosophy; even among the more radical branch, those who use it to throw into doubt the appeal to intuitions at all. This is one reason, actually, that the occasionally overblown rhetoric exasperates me slightly: it's not the reform of philosophy, it's surveys. Lovely things, but just one tool in the toolkit. And for it to be formulated at all, SAIR to date has to take for granted that the basic shape of recent non-SAIR philosophical discourse is pretty much right in one respect, namely, the coherence and value of the notion of 'intuition'. Habitual exposure breeds familiarity; it's difficult to stand back and recognize that this notion, 'intuition', and its cluster of associated terms, is not so clear, and never has been clear, and is a remarkably impoverished vocabulary for talking about the functions and operations of thought these terms are used to discuss. Contrast this, for instance, with the eighteenth-century Scottish Common Sense philosophers, and their notion of 'evidence', understood as 'evidentness' or 'obviousness'. Functionally it works much as 'intuition' does, allowing for slightly different conceptual borders. But if we look at how they handled it, they did not take it as a single notion. Beattie, for instance, gives us the following taxonomy of 'evidence':
(I) Relating to abstract ideas and their relations
(a) intuitive evidence (i.e., self-evident principles)
(b) demonstrative evidence (i.e., something's clearly following from principles)
(II) Relating to things that really exist
(a) certain evidence from our own experience, which is of three kinds:
(1) external and internal sense
(2) memory
(3) causal inference
(b) probable evidence from our own experience, which is of two kinds:
(1) according to uniformity (i.e., extrapolation from experienced to unexperienced of the same kind)
(2) according to analogy (i.e., extrapolation from experienced to unexperienced of like kind)
(c) testimonial evidence
Each one of these types of evidentness has its own account; none of them can be conflated with any of the others, and they all have a different role in reasoning, and a different normative force for philosophical inquiry; they are appealed to in different ways and under different circumstances, and despite all being 'evidence' there is a clear recognition that they are different. It's a bit messier than it seems here; Beattie is simplifying and collating because he needs a systematic account for his purposes (in context, attacking Humean skepticism). But you get the idea. And what do we have when we turn to appeals to 'intuition'? Something much less sophisticated. Now, one thing you could imagine SAIRers doing is trying to enrich the account of 'intuition' so as to make it more adequate to what people are trying to do with it, perhaps investigating how far it is one thing and how far it is many different things that have been lumped together through lack of research, etc. There is work in cognitive science that could be adapted to this sort of thing. But we don't really seem to find this; rather, they take the basic notion as a relatively unproblematic primitive, as if it were obvious that there were real things (unified things with a single coherent account) capable of covering (again, in a way that would fall under a single coherent account) the topics with regard to which appeals to intuition are made. Not only is 'experimental philosophy' research on the basis of a limited method, surveys, and into a very limited topic, intuitions, it doesn't even cover the whole field of questions on that topic.
This triad of limited method, limited topic, limited field of questions, makes it rather doubtful that this is more than one of the shifting passions analytic philosophy gets from time to time. One hopes it comes to more, and as I said, I actually like the idea behind it, and would welcome and encourage more SAIR, especially if such researchers toned down their rhetoric a slight bit and focused (as they have largely done, to their credit) on (1) honing their methodology and (2) finding usable results. I'd love to see Survey Methodology for Philosophers become a regular class in graduate philosophy departments (and I think they should start working on putting such classes together without delay). But right now, for all that it seems to promise to rescript our approach to philosophy, all it seems to offer is an iota and a jot. Iotas are useful; but we should not pretend they are alphabets.
