Two Metaphilosophies on Intuitions, The Folk, and Matters Empirical

(This is a follow-up to the discussion that broke out here.)

Of the many good fruits that the experimental philosophy movement might bear, one that I expect to see ripen quickly is its making analytic philosophers get clearer on their own metaphilosophies.  With each odd empirical finding that gets put on the table, it's as if we're saying to the more traditional intuition-mongering philosopher, "Ok, just what do you take yourself to be doing, such that what you're doing is consistent with that finding?"  And some folks are already rising to the challenge; for example, Brian Weatherson's 2003 paper What Good Are Counterexamples? can be seen as such.  One thing that's becoming clear in the discussion threads here is that there are at least two metaphilosophies of intuition, more or less completely incompatible with each other, that might be operative in analytic philosophy today.  Because of their radical differences with each other, they permit very different responses to the empirical findings.

One of these metaphilosophies takes the philosopher's task to be, at least in part, the mapping out of the structure of 'our' concepts of philosophical matters like AGENCY or KNOWLEDGE.   And philosophers pumping their intuitions is a good way to do this, because the philosophers share in that same concept -- that 'our' is meant to include philosopher and non-philosopher alike.  The analogy here is with linguists studying the syntax of English, and using their own grammaticality judgments as data.  Let's call this, therefore, the "Chomskyan metaphilosophy".

Another view values philosophers' intuitions not because they partake of the folk conception of the matter in question, but precisely because they do not do so.  Philosophers have better judgments about the target concept than the folks, because of some particulars about the philosophers' training and/or knowledge.  Because this metaphilosophy appeals to the superiority of the rational apperceptions of the specially enlightened, we can call this the "Platonic metaphilosophy".

(Three caveats: I don't think that these two are exhaustive of intuition-mongering metaphilosophies, though I suspect they are the main strains, with Frank Jackson as a promoter of the Chomskyan and George Bealer of the Platonic.  Also, these are probably better viewed as two families of metaphilosophies, since clearly there could be a great deal of variation within each. Finally, I'm going to call the respective practitioners "Chomskyans" and "Platonists", but please don't take that to imply that such practitioners or their philosophies are in any other way Chomskyan and/or Platonist!)

Clearly, these two metaphilosophies will be able to respond differently to different empirical challenges.  Consider empirical evidence that philosophers' intuitions diverge significantly from those of the folk.  Such evidence is prima facie; a huge problem for the Chomskyans.  They do have responses.  For example, they can suggest ways of figuring out which philosopher-intuitions are successful at tracking the folk, and give those intuitions more credibility; or they could fall back on a kind relativism, studying now not the folk at large but the philosopher-folk.   But it would be very hard for the Chomskyans to just continue as they were, in the face of such evidence of divergent intuitions between philosophers and nonphilosophers.

But it is not so for the Platonist -- indeed, such evidence might count at least a little bit in favorof the Platonist, since they are committed to their intuitions being better than, and a fortiori different from, those of the unenlightened.

There is a similar difference in responses to empirical data that the folk themselves have divergent intuitions about various cases.  (The term "the folk" is, after all, only something of a convenient fiction to lump together a vast number of very different persons.)  The Chomskyans must now decide which folk(s) they take themselves to be studying -- and what to do about the fact that different philosophers will themselves likely belong to different folk(s).  Again, there are responses, including relativisms; but again, the Chomskyans cannot simply proceed as if these differences were not there.  But prima facie, evidence of divergence across different folk groups is not a problem for the Platonists, since all that shows (they might say) is that lots of people have found lots of different ways to be wrong.

So, is the whole experimental philosophy project just irrelevant to Platonists? Not necessarily. For the Platonists have a theoretical burden to discharge: they owe us an account of what gives the intuitions of the philosophers a higher epistemic status than those of nonphilosophers. And the details of such an account may well be in tension with empirical findings. Here's the abstract form of the concern: suppose the Platonists argue that some philosophers have some characteristic X generally lacked by nonphilosophers, that allegedly renders the former's intuitions more trustworthy.  But then suppose further that there is an empirical result showing that the intuitions in question are highly variable across groups A and B, when there is no reason to think that there is any difference in the degree that X is found in the A and B populations.  Under such supposition, it is reasonable to worry that the mere presence of X may not in fact be sufficient to shield philosophers' intuitions from whatever the influence of A-ness or B-ness might be.  (One possible response the Platonist might have would be to try to study those members of both A and B who are also high in X, to see whether the same variation is found there.  But I take it that there's a certain sort of victory condition here for the experimental philosophers, if we compel the other guys to have to go out & do their own studies.)

One way that this might happen is if the X-factor is in any way sensitive to initial conditions.  Any reflective equilibrium story, to adapt a move from Stich's Fragmentation of Reason (which Adam also uses in his paper available here), has the reflectors (the equilibrizers?) starting with their own intuitions, principles, and/or background knowledge, and reflecting & equilibrating from there.  If members of different groups start with sufficiently different intuitions and such, then we should expect them to RE to different ending positions, too.  So an empirical demonstration of robustly different intuitions across different groups implies a high likelihood of different intuitions at the end of the Grand Equilibration.  Thus, if the proposed X-factor is "being the product of RE", then empirical results showing different intuitions across groups would still, in fact, be a very real threat to Platonic intuition-mongering.

So it will depend a lot on what the Platonists say in defense of their Platonism.  But whatever they say, it will certainly have some empirical content about how one gets to enlightenment from our current unenlightened state.  And that empirical content will itself be scrutinizable by empirical means.  So they are no more immune from empirical challenges than are their Chomskyan brethren -- the challenges just take a different shape for them.

Comments

[tnadelhoffer]

Jonathan,
Thanks for the interesting post. Unfortunately, I am operating under a pending dissertation deadline and have spent too much time lately posting and commenting as it is. Hopefully, others will pick up my slack--thereby keeping the proverbial train moving full steam ahead (it has been nice having active discussion threads for a change). I will try and craft a response on behalf of the Chomskyans soon. Of the two views--I find it much more promising than the Platonist one even if both views seem problematic for different reasons. At least in some areas--e.g., in my primary research area of the philosophy of law--philosophical analyses should be based on the majority of layperson's intutions concerning cases involving foresight, intention, intentional action, and moral responsibility. But I don't presently have the time to put something more coherent together in defense of this claim. Indeed, motivating this claim is the part of the dissertation I am currently working on--so, I should have more to say in the not-so-distant future.

[Brendan]

You might want to check out 'Explanatory Reduction, Conceptual Analysis, and Conceivability Arguments about the Mind' by Brie Gertler (2002). Some of what you say here seems similar to what she talks about if you replace 'intuition' with 'conceptual analysis'. Namely, she discusses how empirical evidence can relate to our concepts: when it should result in revisions etc.
Though I only breezed over her article quickly the other day.

Perhaps there is a third option: call it "Churchlandian Metaphilosophy". Our common sense and philosophical intuitions are probably wrong, therefore, we should readily give them up in the face of empirical evidence. I think that such a view mixed with Chomskyan and Platonic stances here and there seems like a good option.
I for one share the intuition (so to speak), that most intuitions we have a probably wrong.

[Justin]

Let me say to start off that I find the Chomskyan/Platonist framework helpful and clarifying. I want to try to press it a bit. not b/c I think it's fundamentally wrong, but just to help me get clearer on the issues.

A number of Chomskyans are Fregeans or quasi-Fregeans about meaning. A hallmark of quasi-Fregeanism is that concepts are taken to be constituted by the judgments a subject would make *in ideal circumstances*. So, for instance, if I've just taken a mind altering drug and then I tell you that I find it informative that "everyone who goes on vacation for a fortnight goes on vacation for two weeks", you shouldn't yet conclude that "fortnight" and "two weeks" express different concepts in my idiolect. It may be that the drugs are affecting my judgment in a relevant way. If, however, an ideally rational version of myself finds the sentence in question informative, then according to quasi-Fregeans there must be a difference in the concepts expressed by "fortnight" and "two weeks".

With this point in mind, it seems to me that a Chomskyan could try to account for the difference b/w folk and philosophical intuitions by saying that both folk and philosophers have the same concepts, but philosophers are in something closer to ideal circumstances. This would of course need to be spelled out in fuller detail, but it needs to be addressed b/c it seems pretty close to the view held by, say, Jackson.

On the other hand, I know that some Platonists like Bealer basically think that intuitions are like perceptual seemings. Folk intuitions are just as good/reliable/etc. as philosophers' intuitions, it's just that philosophers are better than the folk at accurately describing those intuitions. E.g., the folk (correctly) intuit that there could be a lecturn looking just like this one which was made of ice, but they (incorrectly) say that they have the intuition that this lecturn, which is made of wood, could've been made of ice. It's analogous to the folk reporting that they see that the spoon in the water is bent, while someone more carefully reports that it *looks* bent.

It seems to me that these two points put at least some pressure on the Chomskyan/Platonist distinction. Something seems right about the distinction to me, so I would like to get clearer on the issues involved. (Let me note that the points I've raised here don't in any way address your deeper point, that the psychological evidence on intuitions pose problems for pro-intuition philosophers generally.)

[Neil]

Jonathan,

Maybe we should get finer grained. Perhaps different strategies are appropriate for different areas. Goldman and Pust identify what they call (from memory) the "mentalist strategy", on which Xs= the concept of X. I think that mentalism is much more plausible in some areas of philosophy than others. For instance, I think that "good" should be identified with - suitably regimented - intuitions about goodness (that way of putting it is meant to be neutral between Chomskyan and Platonist metaphilsophies: one possibility is that there isn't a single notion of "good", but many). On the other hand, metaphysical concepts are less plausibly given mentalist analyses: we want to know about the nature of reality, not the nature of people's concepts about reality (of course, I right about "good" only if some kind of non-naturalistic moral realism is false: if it is true, then moral terms are in the same boat as metaphysical).

The second comment I had is more inchoate. I simply want to express my feeling that thinking of knowledge as a distributed enterprise might make a difference to this debate. Once we give up the notion that knowledge claims need to be possessed by any one person, the idea that different people have quite different intuitions become much less threatening. If intuitions are not the final word, but instead invitations to argument then starting from different standpoints need not be much of a worry - so long as we have some reason to think that, under ideal conditions, the differences will narrow over time.

[Jonathan Weinberg]

Some quick replies--

Thomas: I would imagine that some sort of relativization or restriction move would work, in the philosophy of law. E.g., if we're studying the American legal system, then maybe we just don't care what the intuitions of folks in other countries might be. Of course, just restricting to the US of A will probably not get rid of enough of the intuitional diversity, but it's a start. Also, to some extent you're just not a target in the discussion I was offering, in that you freely admit the importance of serious empirical work to your project. But you are right, I think, in worrying that something like these concerns apply to the Chomskyan-empirical-metaphilosophy almost as much as to the Chomskyan-intuitionist-metaphilosophy that I was focusing on.

Brendan: Oh, I'm sure there are lots of other metaphilosophies out there --- I just wanted to consider the two main ones that are meant to license a primary reliance on intuitions. I don't think I could call myself a Churchlandian, but I suppose I'm some sort of neoquinean, at least in part.

Justin: I definitely would not want to object to the appeal to factors like sobriety -- I'll grant that no one has a problem stipulating away the drunk intuitions! -- but the tricky thing, I think you'll agree, is spelling out ideality here such that we really have good reason to think that philosophers are closer to it. I'd suspect, for example, that while one part of ideality would be that the subject is at least minimally reflective, it may well be part of ideality as well that the subject not have been too reflective about the matter at hand. (This is why linguists try to be careful not to trust their own intuitions when things get even a little sticky.) But by all means it would be great for someone to work up such an account, so we could see what empirical stuff it could handle (and, perhaps, what it can't). To whatever extent we move past a notion of mere competency (which everyone can help themselves to, to keep the boozers out), to a notion of ideality , to that extent we do seem to be moving from Chomskyan territory to Platonic. I suppose this might be a matter of degres here; I'm not committed to this being a strict partition or anything like that.

As for Bealer, this move was considered in the earlier thread. Even if you grant the underlying picture, there's a clear problem here of how we are to decide that it is the folk, and not the philosopher, who is doing the misreporting. For such an account to even begin to be a contender, we'd need a much richer account of what it is to get better at intuiting, and what sorts of psychological factors lead the unenlightened to misreport themselves.

Thanks for the Gertler tip!

Neil: I definitely see no reason that different moves might be more appropriate in some places than in others. In one of my own baliwicks, philosophy of psychology, I think sometimes it's basically ok to be some sort of Chomskyan about concepts that are operative inside psychology itself, such as INNATE. (Note that relativizing moves are generally fine here -- if two parts of the literature are just talking past each other, then we should just have two different theories about the two operative concepts.) And I suppose I'm a Platonist about, say, set theory (in several senses, actually, but that's another story). And I'm a could-we-please-stop-using-those-darn-intuitions-so-much-ist in, say, epistemology. Different areas have different ambitions, and indeed can appeal to different empirical situations -- for example, there are grounds for considering mathematicians experts in what they intuit about that that are lacking for philosophers. So I'm in full agreement with you there.

Similarly, I share your sympathy for more distributed epistemic approaches. And basically I don't have any objections to our using intuitions in the "invitations to argument" way you gloss, where they basically bear none of the evidential or argumentative load. I just take it that most of the intuitions that get mongered about in most of epistemology and metaphysics these days are not of that invitational sort, but something more epistemically laden.

[Neil]

Jonathan still disagree, but perhaps less than we initially thought. Part of what we disagree about is how intuitions actually get used in philosophy: he thinks that often they are the court of final appeal, whereas I think that they do - and ought to - carry some evidential weight, but it ought to be quite highly defeasible. But there's a second area of disagreement, which is implicit in a number of Jonathan's comments, for instance the repy to Thomas. Jonathan supposes that philosophers divide into (what he calls) "intuition mongers" and those who "admit the importance of serious empirical work". But I'm a counterexample to the claim, since I think intuitions should carry weight, and that all kinds of empirical work - from that in which he engages through to evolutionary biology, social psychology and neuroscience, and plenty more - should constrain or inform philosophical theory.

[Brendan]

Question: is it really possible for one to be a Chomskyan, after one gets philosophy training? After all, whether one likes it or not, ones training is going to affect ones beliefs. THat is, just in virtue of the training we are no longer "part of the folk", we are philosophers, so to speak. And as philosophers, we can only have the sort of intuitions philosophers have, so we can only follow those intuitions: namely philosopher intuitions. Which seems to make us Platonists about intuitions no matter what.
SO Perhaps (as philosophers) we are all platonists about intuitions, but think that sometimes we should try and follow the intuitions of a Chomskyan, but I do not think we can say we actually are Chomskyans about intuition.
So the idea is, if one is a philosopher, we can only have philosopher intuitions. THey are the only kind we have (our training jades us you might say). thus, we cannot follow the sort of folk intuitions that a Chomskyan would because we are no longer regular folk, and thus do not have the relevant category of intuitions to draw on.

I haven't read all the posts in super detail, so if someone already suggested something like this I apologize.

[tnadelhoffer]

Brendan,
I promised myself I would take a break for a few days--but I wanted to breifly respond to your post. The problem you mention with the intuitions of philosophers being tainted by their theories--a problem that Goldman and Pust call "theory contamination"--is precisely what has prompted some experimental philosophers (including myself) to conduct formal surveys of folk intuitions concerning particular cases. The goal is to catalog the pretheoretical "intellectual seemings" of laypersons. In the event that there seems to be systematic agreemenet, conceptual analyses can be developed that attempt to capture the intuitions in question. Of course, if the intuitions of laypersons are either inconsistent, asymmetrical, or theory-laden, then we have reason to worry about the usefullness of their intuitions and concepts for philosophical purposes. However, we cannot just assume a priori that such conflict exists. With intentional action, for example, there is pretty wide scale agreement (in the 80% range)among laypersons in America (and India) when it comes to particular cases culled from the action theory literature--although much more research needs to be done before any kind of robust analysis can be developed. In the event that the research shows (as I suspect that it will), that there are differences between cultures, genders, ages, socio-economic status (etc.) then philosophers will have to reconsider the approach. But we cannot simply assume these differences exist without doing some of the dirty work.

[Stephen Weinberg]

As an outsider from a different experimental field, experimental economics, I thought I'd relate some of the early history of what was, 30 years ago, a fairly radical and controversial branch of econ. I doubt any of these points will seem surprising to you, but you may be amused to see how the introduction of experimental evidence to a resistant literature played out elsewhere.

Neil argues that many other sources of empirical evidence should operate as constraints on philosophy, which seems to be a hard position to refute once one has made the move of appealing to evidence. However, all the other sources to which Neil alludes seem to be done by specialists in other areas. There are advantages to (a) doing the empirical work yourselves and (b) doing experimental work yourselves.

Some of the elder members of the experimental economics profession discuss an early naivete that, if they simply told psychologists what types of questions were of interest to economists, they would start doing the relevant experiments for us. Of course, it turned out that psychologists are generally interested in very different sets of questions. Relying only on the areas of overlapping may occasionally deliver a big score (e.g. prospect theory), but even then psychologists weren't doing the follow-up work that focused on the particular extensions of most interest to economists. Furthermore, the psych literature had a number of experimental norms that proved unsuitable for economics. For example, economists care a great deal about the incentives facing agents, so most economists were extremely reluctant to accept experiments in which subjects did not have monetary incentives to make the best decisions possible. So experimental economics has devised very different norms for how to compensate subjects. As you see the objections that tend to be raised to philosometrics, you can devise procedures that are most likely to be convincing to your colleagues (and yourselves).

In the early years of experimental economics, a number of standard objections emerged from respondents who didn't want to take the evidence seriously. For example, many economists absolutely insisted that subjects would behave according to economic theory if only they had enough money at stake. Overcoming these objections took a lot of energy for the first generation of experimental economists, and a lot of creativity. For example, classic early experiments were replicated in 3rd world settings in which relatively meager research budgets could provide "stakes" equivalent to several weeks' income.

This brings me to one of the advantages of using experiments, as opposed to other forms of empirical research (which certainly have their own advantages). It is relatively easy to devise and run new experiments to address methodological concerns about existing work. More importantly, as experimental economists disagree among themselves about how to interpret results, experiments offer the advantage of being able to _quickly_ move through several generations of arguments and evidence. Experiments quickly help antagonists identify areas of disagreement, and suggest follow-up work to try to resolve the disagreement. To help with this process, experimental economists have a norm of posting their full procedures on their websites (or printing their full experimental instructions in appendices to their papers), and of providing their data on request. That way, you can quickly replicate each other's work with whatever variations you believe are necessary.

Jonathan alluded to experiments' advantages in crystallizing disagreements in a different context when he said, "With each odd empirical finding that gets put on the table, it's as if we're saying to the more traditional intuition-mongering philosopher, 'Ok, just what do you take yourself to be doing, such that what you're doing is consistent with that finding?'" {By the way, it took decades for experimental economics results to stop being seen as "odd empirical findings"; much of this work was disseminated in a series titled "Anomalies.")

In economics, experiments are now used to test specific theories, to search for behavioral regularities that can be incorporated in building new theories, and to provide "beta-tests" of new market institutions before they're adopted to, say, auction off billions of dollars worth of telecom spectrum. It took decades for experimental economics (and its cousin, psychological economics) to move into the near-mainstream (though the early innovators have started winning Nobel Prizes, so take heart!).

Probably the most important step towards having results accepted as important instead of just "anomalies" was when practitioners used the evidence to build specific new theories and then demonstrated that those theories made additional, useful, testable predictions.

Merely telling the literature "What you're doing is wrong" isn't more than a starting point. Economists, at least, were quick to brush off criticisms when they didn't have any better tools at hand. We needed alternatives that met the new empirical criticisms and _also_ did the work that the old methods did. If you think you've built a case against some common uses of intuitions, you need to show a way to replace those applications with something else that can do the job.

For more information on experimental economics, one of its leading practitioners keeps an extensive web site:
http://kuznets.fas.harvard.edu/~aroth/alroth.html

I don't plan on returning to this site, but if you take umbrage with any of my claims, feel free to pester my brother Jonathan. Lord knows I do.

[Jason Leary]

GEORGE BEELER

I am very much interested in the Platonist epistemology presented by George Beeler . Does anyone know where it is easy to find online texts written (or posted) by Beeler ?

[Jonathan Weinberg]

Jason, it's spelled "Bealer", and you can find a fair amount of his stuff through JSTOR and so on.