« Do you see tinted lenses? | Main | Intuitions and Cross-cultural Variation: Part II, Interpretation »

Comments

Jen Wright

Nice post, Marc. I think your points are right on. Just FYI, I collected some data on the framing effect issue some months back -- I found that only certain vignettes (namely, those that were clearly non-paradigmatic) elicited a framing effect, while the paradigmatic cases did not. In addition, participants appeared to have some level of epistemic access to whether or not their reponses were reliable. If you are interested, I have a post on the data: http://experimentalphilosophy.typepad.com/experimental_philosophy/2006/12/thoughts_about_.html.

Jen Wright

Oops: looks like the url for the post didn't come through. It should be: http://experimentalphilosophy.typepad.com/experimental_philosophy
/2006/12/thoughts_about_.html

Aidan McGlynn

One last try at Jen's link.

jonathan weinberg

Although Marc offers a number of helpful methodological suggestions in here, I'm afraid that they don't come anywhere close to the kind of _rejection_ of the findings that he seems to plump for at the end. They are all ways in which the research could have been done _better_, but none of them are the right sort of reason to doubt the basic findings themselves.

1. First of all, this claim puts much more onto us than we really need: "...the methodology presupposes that every answer (positive or negative) involves an intuition." We don't need anything near that strong. All we really need is that (i) the answers do at least significantly track intuitions, and (ii) there is no reason to expect the degree of this tracking to vary from one of our experimental groups to another. In the WNS paper, we discuss something very close to (i) explicitly in the objections & replies section. And (ii) doesn't seem to me to need much argument. E.g., which population would it be, the European or the Asian, that we should expect to be more intuition-prone?

I have to say, the bit about us making a basic mistake about intuitions-that-not-p and not-intuiting-that-p is not only besides the point, but is also rather gratuitously insulting. (A lot of the tone in the main post, actually, made me want to say, "whoa, dude, chill out!" In the immortal words of James Taylor, "Ain't no need to act like I shot your dog.") Diversity about who does or does not have various intuitions is more than enough to get the challenge off the ground, and indeed if it were the case that a large number of people fail, when confronted with the zombie case, to intuit that it is possible, is a perfectly good reason not to trust that intuition in those who do report it.

Also, I would suggest that we really have very little reason to think that philosophers in general give much more consideration to cases than our subjects have. Here's how it usually goes: you sit in a talk, someone throws a funky case up on PowerPoint, and most everyone forms their take on it right away. It's not like it's standard practice for speakers to pause for a minute or two after presenting their thought-experiment to an audience, to make sure that everyone is intuitionizing. Of course, some people do think about cases more than others do, and about some cases more than other cases, but I don't think that proponents of intuition have any evidence whatsoever that this variation in amount of time spent thinking leads to differences in intuition (which is the finding they would need, in order to lodge this complaint.) I think that there's really interesting empirical questions here, about how philosophers really do or don't approach thought-experiments, and whether ordinary subjects do something already pretty close to this, and if not, what results you'd get if you got them there. But unless one thinks that strong intuitions are just on the whole _totally_ unrelated to relatively-less-considered intuitions, one should take our findings seriously.

2. The basic point about the psychology of filling in seems right to me (and is one that Sosa, for one, has pressed recently). But it doesn't cut at all against our findings unless we have some reason to think our different populations fill in in significantly enough different ways. And I have yet to see such a reason.

Moreover, the filling-in worry is a concern that, if it cuts against us, cuts all the worse against the general method of intuitions -- when we read someone's paper, we have no way at all of determining what sorts of filling in they are or aren't doing! Lean on the possibility of divergent filling-in too far, and you should end up even more worried about the epistemic status of intuitions than I am. I agree about the problem with getting students to understand some of Gettier's actual Gettier cases -- which is of course why we didn't use them in the study! We tried to make the actual vignettes as easy to understand as we could make them. (Which is also why, inter alia, we avoided using philosopher-speak like "justification".)

3. As for the stuff with "really", the point of phrasing the probes that way was precisely to cue subjects away from a subjective-certainty reading and towards a more epistemic reading. I don't understand why you don't think that that is what the effect of it would be. "Knows"/"Doesn't know" is the phrasing that leaves open more interpretations, not "Really knows"/"Only believes". Having said that, of course it would be interesting to do the study again with other probes; and to look at it with Likert scales (which we have been doing more recently) instead of binary choices; and to look at it with some sort of "I just don't know/not enough information" option; and so on.

(Btw: "bizarre"? Or, indeed, "_bizarre_"? Naw, that sort of thing goes on all the time. A lovely & famous example from the development issue has to do with shape or color constancy, and asking children, "How does this look to your eyes?" vs "How does this look really, really?")

I'm not seeing what the _worry_ is here about the ways in which terms like "really" are "frequently used to indicate that the term in question is not being used with its standard meaning." It's not open-ended what kind of meaning is indicated, but seems to be pretty clear. It looks to me like a standards-raiser, and not, say, a metaphor-introducer. I mean, "Hey, does he really know that?", as a response to someone's claim, is completely normal, colloquial English, as is, say, "Do we really know when we need to get to airport?", etc. They all look like decidedly epistemic readings of "knows" that get indicated in such situations. A little googling around on "really know" and "really knows" goes a long way here. (That doesn't help with "only believes", though, because that seems overwhelming to be instantiated with cases of "only believes p (without going on to also believe q)".) If I ask if someone in the classroom is _really_ a student, I'm asking something like: on a fairly strict reading of "student" (so, ruling out auditors, etc.), does this person count as a student?

Perhaps more importantly, there's just no reason at this time to think that the concerns about "really" (or about "only believes (without justification)") would have a differential effect on our subject populations, which is what you really (*ahem*) need to be arguing for here. Now, it certainly would give us a reason not to put too much weight on the absolute scores of any population, as one could reasonably make the prediction that the inclusion of "really" would generally depress people's willingness to attribute knowledge. But we don't put any real weight on the absolute numbers, which indeed do make people look to be strangely skeptical on the whole. (This may have thrown you off somewhat in reading our experiments, since in your own experimental work the absolute numbers are the important thing, and you weren't looking for group differenes, iirc.)

4. I'm now going to indulge in one little accusation of bizarreness of my own here, in that I have to note that I found it just... weird, when you claimed that there's something off-putting or fishy with our needing to use controls of various sorts. No form of any sort of scientific research is going to fail to need such controls. If you want to make the much more radical argument that no possible scientific research could have anything to do with philosophical practice, then go for it; but if your arguments are meant to be more modest than that, then there's just nothing even remotely untoward in our doing so.

Putting that aside, it is of course a fair question to ask whether or not we have used sufficient controls, in our coinflipper case. But I don't see how your concern raised here has any teeth. We didn't generally offer much in the vignettes as to the overall strength of the agents' felt conviction, so any subjects using a subjective-certainty notion of "knows" shouldn't give us much patterning in the data; everyone should either know, or not know, for them. But that's not what we found -- importantly, on some cases the Asian-descended population was _less_ willing to attribute knowledge than the European-descended population. In general, when arguing about uneliminated confounds, etc. one has to argue that the concern in question is something that would produce the results that were found. Otherwise it's not a confound for those results. And the hypothesis that some of our subjects were using the subjective-certainty notion just doesn't do that.

5. Another example of a non-confounding confound is here: "Who knows, maybe East Asians are more likely to tell some such back story than Westerners." Yes, and who knows? Maybe they just are more likely to circle things on left sides of pages when they see the word "car". And, who knows? Maybe they are all secret despisers of analytic methodology, and answered counter-intuitively to help gin up resistance, and who knows...? An experimenter is only responsible for confounds that we have _some_, even tolerably small, reason to think may be the case. And it's not enough that one population be just slightly more likely to "tell some such backstory" than the other -- for such an objection to have force, it needs to be the case that they are sufficiently more likely to do so, that it would produce an asymmetry of the sort our data detected. That's a fairly far-fetched hypothesis, but it would be a fairly easy one to test if someone wanted to do so. (And let me emphasize, again, that I agree that it would make the research _better_ on the whole if we did happen to have data that spoke even to this kind of concern.) But without such a test being done, the priors on it are sufficiently small that it's just not much of an objection.

(A rough analogy: suppose that a refined dispositionalist read your paper with John and Jen, including this part of the Irina case: "whenever she actually attempts to do a Salchow (in accordance with her misconceptions) the abnormality causes Irina to unknowingly perform the correct sequence of moves, and so she ends up successfully performing a Salchow", but they say, "Well, that 'whenever' is consistent with her hardly ever doing it at all -- universally-quantified propositions not entailing existential ones & all that. So, who knows, maybe the bulk of your subjects aren't thinking that she has yet manifested a _reliable_ ability. And so your finding is consistent with my refined dispositionalism." I think that you would be completely justified in responding to such a person, "Yes, well, _maybe_ the subjects are so constituted -- and it would be easy enough to run a version of the study that controlled for that -- but can you give me any reason whatsoever to suspect, with at least a low-but-nontrivial probability, that they _are_ so constituted?")

- - - - - -

At the end of the day, all we've really got here (really!) is some moderate kibbitzing that our data is noisy. Well, _of course_ our data are noisy. And I know that I speak for my co-authors here as well when I say that, absolutely, undoubtedly, there are all sorts of ways in which the research could have been done _better_ than it was done, much of it along many of the lines that Marc has suggested. And we're also at pains to emphasize the preliminary & suggestive nature of the findings reported in that paper. But nothing in Marc's considerations here go very far towards undermining the basic findings, because there's no reason to think that any* of our subject populations would be particularly more or less susceptible to any of these forms of noise, and certainly no reason to think that they are such to the extent found in the differences in our data. It gives us all, on both sides of this debate, a good reason to go do a lot more studies -- but no reason yet to utterly discount the ones that we've already got. If you want to argue against our data, you're going to have to go get some of your own. And I know that you have at least some interest in experimental philosophy, Marc, so I say: go for it! I promise you it'll get a lot of talk in x-phi-land, if you do!


*(Actually, I think one _could_ make a valid critique along some of these lines for our SES findings... which is why I haven't touted them for years.)

Justin

I'd like a bit of clarification about intuitions being only those states with the content that p in which it is presented as necessary that p.

Consider two subjects, one a philosopher, who judges that "S knows that p" for some epistemic case, because it seems necessary to him that "S knows that p." Now consider a non philosopher who makes the same judgment, just because it seems to him that "S knows that p." He has no stance on the necessity of the proposition, because he has no explicit concept of necessity (perhaps he has the same understanding that our undergraduates have prior to our explaining the concept of necessity. They have the background understanding necessary to learn the difference between the necessary and contingent, but have never learned the difference).

Now, how are the philosopher's and the non-philosophers judgments related? My thought is that the only difference is that the philosopher knows that there are these things called necessary truths which are interesting. It seems unlikely that the philosopher's judgment rests on an entirely different source than the ordinary subject's judgment. So the ordinary subject's judgment that "S knows that p" will be good evidence about how the philosopher's intuitions would go.

In short, why not think there are roughly two capabilities here, the capability to judge "S knows that p" and then the capability to classify that as necessary?

marc moffett

Hey, thanks for the comments Jonathan. Sorry if the tone of the post was abrasive, but I can't deny that I was exasperated.

Your response has helped to clarify for me the central point of concern and also why I think these methodological shortcomings are more serious than just a moderate amount of noise. The reason I raised the point about your methodology presupposing a failure to distinguish between intuiting that not p and not intuiting that p is that it underscores your background assumption about the ubiquity of intuitions. One possibility you seem to be routinely overlooking is that there may be very few if any intuitions backing the forced choice responses. Suppose that there aren't. Then, of course, the results don't show us anything of philosophical interest concerning intuitions. This is why I was harping on wording of the vignettes. In a situation where one is not actively in a position to clear up background confusions and inappropriate back stories, it is essential that the vignette itself carry the bulk of that burden. And this is really, really hard, especially with subjects who are unfamiliar with the methodology. So one place we just seem to disagree is that you think intuitions are "cheaper" than I do. Just considering my own case, I think that there are pretty good reasons to doubt that intuitions are so prevalent. (As noted, I don't get the Gettier intuition in the case you give or to the extent I do, it is only because I already know what I am supposed to be looking for. Given this simple first-person fact, I don't see any reason to take your results to bear on intuition.) So I don't agree that I am demanding too much; I want to be sure we are talking about intuitions here and not hunches or anything else on which people routinely base their responses.

Moreover (I am not quite sure how to put this), but I am worried that, even if the subjects are having intuitions, it may be that they "don't know what to do with them." Let me try to explain by example. When I start to introduce students to thought experiment-style counterexamples, I begin by using some cases which are really easy and clear (e.g., argument validity assessment, maybe a married person not being a bachelor). I do this so that I can be pretty confident that the students are, indeed, having an intuition and I can bring it to their attention. It is not, of course, that they are not aware of it; it is just that they don't necessarily get what it is and what sort of import it might have. So there is just a kind of "familiarization" process. (I don't think that this counts as coaching, at least not in any way that is illegitimate. The students are having the intuitions, I am just making explicit their reliance on them in assessing cases.) Because of this, even if intuitions are ubiquitous, the corresponding judgments need not necessarily conform to them because subjects may be relying on other kinds of considerations.

One final point. I think it is common not to worry too much about the details of a vignette because a speaker can generally rely on her audience's competence with this sort of thing--if they can fill in the example in a way that convinces them that is all it takes. So it might be best to think of those slides as counterexample schemas and the audience is invited to fill them in in a way that works if possible. One counterexample, after all, is all it takes. (And I would note that a VERY common question is, "Could you explain a bit more about your examples, because I'm not getting the intuition?" Also, "I think you case is underscribed. If you read it this when, then you get blah, but if you read it that way, you get bleh...")

John Bengson

Hi Marc,

I think you've offered a pretty darning criticism of WNS's attack on IDR. FWIW, it strikes me that the response that the results are merely preliminary is not adequate, given the nature of your objections. (In any case, I find the response a bit puzzling given the way that the results are often presented -- as in, for instance, Jonathan's comment in the exchange you link at the beginning of your post, in which for instance he non-tentatively asserts that it is very rare to find someone who understands the Gettier cases but who doesn't have the Gettier intuitions "Only if you don't ask people who come from other parts of the globe!".)

If I correctly understand the upshot of Justin's comment, he seems to be worried that IDR theorists can't avail themselves of your first objection to WNS since someone (e.g., a non-philosopher) can have the intuition that p without p being presented to them as necessary (because, e.g., they lack the concept of necessity). But it seems that since the IDR theorist doesn't make claims about any and all states that go by the name 'intuition', but only about a certain sort of state (sometimes called 'rational intuition') which contemporary rationalists are at pains to adequately isolate and to show is somehow epistemically relevant, Justin's observation can be accommodated by IDR.

This appears to be a source of deep conflict. Perhaps WNS are simply not satisfied by various rationalists' attempts to distinguish rational intuition from other nearby states. I'd be interested to hear why exactly they are dissatisfied. It can't be that rationalists haven't tried. The efforts of contemporary rationalists to adequately isolate a state of rational intuition that is epistemically relevant cannot be emphasized enough, I think. As you point out, no rationalist worth his salt would identify the epistemically relevant state in question with "response to a question" or "forced-choice snap judgment" or even "that state, whatever it is, which is the source of the non-philosopher's and the philosopher's judgments about hypothetical or actual cases". And this remains so despite what the caricature of intuition driven philosophy which is, at times, proffered by some empiricists misleadingly suggests.

jonathan weinberg

It sounds to me like there's some confusion here about the target of our arguments. Y'all seem to be taking our phrase "intuition-driven romanticism" to be referring to, to a very rough first approximation, George Bealer and people who are engaging directly with & building on his work. But we take IDR to refer to, to a very rough first approximation, a rather vast swath of analytic epistemology, metaphysics, philosophy of mind, and ethics -- basically, any philosophical activity that meets the conditions we articulate in the early part of the paper. And different philosophers and groups of philosophers think about intuitions _very_ differently. (Marc's "[sic!: any] IDR theorists" only works if one restricts the domain of quantification in ways that we sure didn't have in mind.) Among recent defenders of the IDR methodology are not just hard-core rationalists like Bealer, but also (just off the top of my head) Frank Jackson, Brian Weatherson, Henry Jackman, and Tim Williamson. And you will be hard-pressed to find a unified picture across those theorists of the nature & epistemological workings of intuitions. Williamson, for example, assimilated philosophers' armchair judgments to counterfactual judgments more generally, and not a special necessity-appreciating mode. You can even throw people like Devitt in on top of this, who defend intuitions, but in virtue of being generally reliable repositories of _empirical_ information. There are many available ways to go here, and philosophers (as we always do) take all the ways that are available & even some of the ones that aren't.

And we thought (and still think) that our results raise a challenge for all groups of IDR practitioners. Now, some groups may have different resources available to meet that challenge than others. And the neorationalists here _might_, by cranking up the knob on what really counts as an intuition, have a way of dodging some of the bullets. But it's important to see that it isn't the case that WNS have misdescribed or "caricatured" philosophical practice; rather, the neorationalists are advocating a _reform_ of philosophical on the whole practice. They have _a_ picture of how intuitions should work, but it is one picture among many, and despite some significant argumentation on its behalf, it has not achieved any sort of universal status in the field. And maybe neorationalism is a good reform, and philosophers should take it on as a response to challenges like WNS. But I have my doubts -- I am rather dubious of how well we can distinguish _real_ intuitions from merely apparent ones, in ourselves and especially in others. That we can do so _in principle_ is perhaps enough to license a rationalist epistemology; but if we cannot do so _in practice_, then it fails to enable a rationalist _methodology_. And it is the methodology, not the epistemology, that is at stake here.

(Btw, John, I'm not an empiricist in my epistemology, if we take that to mean a complete rejection of the _a priori_. Indeed, my dissertation was a defense of a form of rationalism! Just not a form that would give much comfort to practitioners of IDR.)

jonathan weinberg

I just noticed how Marc's cashed out our phrase "IDR" in the first sentence of the main post -- but in our paper, note that "IDR" stands for "intuition-driven _romanticism_", not "intuition-driven _rationalism_"! Rationalism, again, is just one particular flavor.

Justin

John, you got the upshot of my worry, but I think I can raise the question in a more focused way. The issue here is not how to define intuition, but about the relevance of empirical studies of what we might call "intuition like judgments" to questions about philosophers' intuitions. Even once you admit that what's studied in WNS isn't intuitions in the rationalist's sense, you still have a question of whether WNS's results tell us anything about intuitions in the philosophers' sense.

Here's a simple case for why they might: the experimental subjects and the philosophers are making judgments about the same topic, so they should be using similar cognitive mechanisms. If that's true, then WNS should be relevant to discussion of philosophers' intuitions.

marc moffett

Holy alternate reality, Batman! WSN as a defense of neorationalism? What has the world come to?

A few points, mostly nit-picky. First, I think it is a misleading to say that the neorationalists are advocating a reform of philosphical practice. The basic ideas arguably go back to Plato (if not before) and clearly go back to Descartes, Spinoza and Leibniz and on through Frege. Indeed, it is arguable that even people like Hume are rationalists in the relevant sense. Moreover, I can't even begin to count the number of people that profess not to trust intuitions as part of their philosphical theorizing, but then turn around and rely on them evidentially in the next breath. The practice is very deeply entranced, if not always explicitly acknowledged. So, I think the opposite characterization is the the more accurate one: anti-rationalist empiricism is a reformist position, and very radical one at that. Maybe only the positivists and some of their heirs have come close to actually embracing it in practice.

Second (and relatedly), it is important to distinguish between accepting the neorationalist characterization of intuitions and accepting those intuitions (assuming that the characterization is correct). I think that a large number of people who rely on intuitions methodologically think that there is a distinctive sort of cogntive episode and that episodes of *that sort* are evidentially relevant (i.e., they constitute defeasible reasons for belief). Such thinkers may differ on the best way of analyzing that sort of episode. So it doesn't follow from the fact that someone rejects the neorationalist theory of intuitions that they actually reject neorationalist intuitions (i.e., cognitive episodes of *that sort*). We agree that cognitive episodes of *that sort* have evidentiary status, we disagree about the best way to characterize those sorts of episodes.

This leads me to my final point. While it is true that the neorationalist theory of intuitions is only one among many, I strongly suspect that most of the epistemologists who have considered views on this issue impose pretty significant constraints on what counts as the right sort of cognitive episode. And for this reason, I suspect that defenders of those views will have things to say about WSN that is pretty much parallel to what I have wanted to say. I think it is pretty clear, for instance, that Chalmers and Jackson don't have a permiscuous theory of intuitions.

However, I do grant that there are certain philosophers who want to adopt a more empiricist epistemological stance while nevertheless cleaving to the traditional rationalist reliance on thought experiments and intuitions. (I know, I know, I am being tendentious.) Such philosophers do want to give a deflationary theory of intuitions. And so, perhaps the right moral to draw from WSN is that a deflationary theory of intuitions is implausible. So much the worse, they might say, for intuitions; so much the better, I might respond, for rationalism.

jonathan weinberg

"Holy alternate reality, Batman! WSN as a defense of neorationalism? What has the world come to?" Well, my rationalism is not manifest in the WNS paper; a super-quick taste of it does appear in my paper in the _Epistemology Futures_ volume, though.

"I strongly suspect that most of the epistemologists who have considered views on this issue impose pretty significant constraints on what counts as the right sort of cognitive episode." The authors I listed in the previous post seem to me to indicate that this is probably not so. I guess it depends on what we mean by "significant", but for your purposes here, that needs to mean something like "so significant that it makes sense to think that whatever the WNS subjects are doing, it has no connection whatsoever to real intuitions". And that's a pretty significant sense of "significant". And there are about as many Williamsonians as Bealerians out there, I would reckon.

"I think it is pretty clear, for instance, that Chalmers and Jackson don't have a permiscuous theory of intuitions." I don't think that can be right, at least not about Jackson. Jackson's underlying metaphilosophy is one in which intuitions have their value precisely because they are an expression of the folk theories underlying our conceptual competences. Intuitions, in Jackson, are something very humdrum and everyday and people have them all the time without any special training or anything.

jonathan weinberg

from a post I just put up over at the x-phi blog: http://experimentalphilosophy.typepad.com/experimental_philosophy/2007/09/on-intuition-sn.html

The metaphilosophical view on offer there is that philosophers' reports count in ways that our subjects' reports don't -- not necessarily because philosophers are better than ordinary folk per se, but because the standards for what counts as having an intuition at all is pretty high. Let us (tongue-in-cheekily) call this particular version of elitism, intuition snobbism.

Now, here is the main problem for snobbism: if we crank up the dial on what counts as an intuition, then we accordingly will have to dial down our confidence that anyone is actually having a gin-u-wine intuition at any given time. First, we have to lower our confidence that other philosophers' reports of their intuitions -- their alleged intuitions -- in the journals and talks are, in fact, really reporting intuitions, and not some other form of intellectual seeming. More broadly, we have to surrender our sense that folks at large concur with us on many of our favorite intuitions. So, for example, this infamous bit of Jackson -- which I think reflects a kind of reasoning very common among IDR practitioners -- would have to be sacrificed:

"...I am sometimes asked – in a tone that suggests that the question is a major objection – why, if conceptual analysis is concerned to elucidate what governs our classificatory practice, don’t I advocate doing serious opinion polls on people’s responses to various cases? My answer is that I do – when it is necessary... Everyone who presents the Gettier cases to a class of students is doing their own bit of fieldwork, and we all know the answer they get in the vast majority of cases. But it is also true that often we know that our own case is typical and so can generalize from it to others." (pp. 36-37 of From Metaphysics to Ethics)

We also see this kind of thinking involved in the line that provoked this discussion, Marc's comment on this post over at the Leiter blog:

"After all, while I know plenty of smart linguists who reject [Jason Stanley's] particular theory of the syntactic structure of know-how attributions, it is very rare to find someone who understands the Gettier cases but who doesn't have the Gettier intuitions (i.e., that they are not cases of knowledge)."

But merely comprehending the cases isn't enough, on an intuition snob approach. They need to have, say, the force of necessity; or a strong modal tie to the truth; or to be experienced as internally justifying; etc.

The intuition snobs also have to give up one of their favorite rhetorical moves against the experimental restrictionists: the tu quoque. Here's an example from Marc, just upthread from here: "Moreover, I can't even begin to count the number of people that profess not to trust intuitions as part of their philosphical theorizing, but then turn around and rely on them evidentially in the next breath." Well... how do you know that that is what they were doing? Why do you have any confidence at all that it was an intuition that they were relying on? Maybe it was just a judgment of some other sort, such as a tacit sense of the weight of the empirical evidence. Indeed, speaking for my own case, I'm pretty sure that unless you saw me using DeMorgan's Law or something rudimentarylike that, it wasn't a snob-worthy intuition that I was using.

Even from a Cartesian standpoint, snobbery would require us to become rather less certain that our own reactions to thought-experiments are proper intuitions. There are just too many cases in history of various theorists being utterly certain that they had grasped, via the lumen naturale, some inmost metaphysical truth, which most of us now find unintuitive to begin with. (The Principle of Sufficient Reason is an excellent example, as is most of the arguments for God's existence in the Meditations. Most folks seem unpersuaded by BonJour's appeal to an intuition of a schematic form of an inductive inference at the end of In Defense of Pure Reason, at least in part because what he puts out as intuitive just doesn't strike everyone as such.) And there is too much scientific psychology about the ways in which the human mind is all too happy to play tricks on itself, convincing itself where it ought not be convinced, which is why there is all too many scientific norms that are in place precisely to keep us from being able to pull such autochicanery.

So, given the significant costs that snobbery would incur, we might well ask: why be a snob? There's no descriptive methodological reason to go snobbish -- it's not already built into philosophical practice, and a great many philosophers think that their intuitions can be relied upon precisely because they're already manifested in our quotidian folk practices. For example, a Jacksonian is committed to intuitions being little more than the expression of our shared folk theories underlying our conceptual competences, for example. See also, say, DeRose on the ordinary language basis for his form of contextualism. Marc notes, rightly, that everyone is at least appealing to some-sort-of-intuitive-state-or-other, but snobbery is a non-starter as an account of anything that all or even most analtyic philosophers are appealing to.

This gives us a sense in which intuition snobs and restrictionists are similar: they are both advocating a radical change in philosophical methods. The restrictionists are motivated here by the idea, inter alia, that philosophy would do better to take a few methodology lessons from the sciences. (I think that's what naturalism should really be about, by the way -- not a physicalist ontology, but adding some as-scientific-as-we-can-manage methods to our tool kit.) And of course we're motivated by our particular findings, even as preliminary as they are. But where does snobbery get its motivation? I think it derives mostly from concerns about having a certifiably anti-skeptical epistemology, one in which intuitions can play an important role as a basic (though not infallible or incorrigible) source of evidence. And, actually, I'm perfectly happy to be on board with something like that, as a matter of fundamental epistemology.

But here's the thing: we're not doing fundamental epistemology here. We're doing methodology. And basicness isn't a meaningful category from a methodological standpoint -- we now have oodles & scads & tons of evidence of all sorts that speak to the range & scope of when perception, memory, testimony, & various & sundry forms of intellectual seemings are more, less, or not at all trustworthy. Maybe only a very special sub-class of those intellectual seemings are capable of getting things off the ground in the first place, and need to have conditions like the aforementioned internally-justifying nature in order to qualify as properly basic. But none of that matters after we get off the ground, and indeed we've been flying high for a long time now. Basicness is a starter-motor for the engine of inquiry; once we're in motion, it's finished with its job.

So, where does that leave things? In the context of the debate over philosophical methodology, intuition snobbery comes at a great price: the general loss of being able to determine, in our own practices, in our collaborators, and in our interlocutors, just when we do or don't have a proper intuition on our hands. I suspect that this would be a methodologically disastrous price, but it's clearly something that could be argued about (and argued about empirically). But it's not at all clear that snobbery really buys you very much for that price. All we get is a bit of epistemological machinery that doesn't do us any methodological good. Not, on the whole, a good bargain for philosophy. But isn't it just like snobs to accept only what they take to be the best, regardless of the price? ;-)

Dan Korman

Jonathan --

There's a lot to say about this post, but let me focus on your claim that the intuition snob is advocating a radical change in philosophical methods.

First, you say: “we accordingly will have to dial down our confidence that anyone is actually having a gin-u-wine intuition at any given time” on the grounds that “snobbery is a non-starter as an account of anything that all or even most analytic philosophers are appealing to.” But is that right? Bearing in mind that one doesn’t have to *conceive* of intuitions as intellectual seemings that have the “force of necessity” in order to be having them and reporting them, it seems plausible that this is just the kind of mental episode that we all naturally have, and report having, when asked to consider hypothetical scenarios. Even those who do have a view about intuitions, and think that they’re judgments or inclination to believe, are presumably reporting these same mental episodes when they claim to be reporting intuitions; it’s just that they think (mistakenly, I say) that these seemings are judgments or inclinations. So to worry whether philosophers are reporting how things seem will strike the intuition snob as an idle skepticism (like wondering whether they have inverted color experience).

For the same reason, I doubt that intuition snobbery requires us to “surrender our sense that folks at large concur with us on many of our favorite intuitions.” Now it *may* be that the x-phil results give us reason to worry whether the folk share our intuitions. But not the snobbery.

((You may complain: if we are naturally disposed to report snob-worthy seemings in response to hypothetical scenarios, then that should take care of Marc's worry that subjects aren't reporting their intuitions in the surveys. Let me just say: you're right to complain, but I think there's more to be said on Marc's behalf.))

Second, you say: “The intuition snobs also have to give up one of their favorite rhetorical moves against the experimental restrictionists: the tu quoque.” There’s a serious question of whether (and against whom) the tu quoque is effective, but I don’t see why it’s supposed to be any less effective in the hands of the intuition snob. The way the tu quoque works is: the victim claims that a certain range of facts can’t be know from the armchair (or perhaps can’t be known full stop), and then later helps himself to one of those facts, and I say “how do you know *that*?”, and if all goes well he blushes and admits that he’s been bad. Your suggestion is that the empiricist can sometimes block the tu quoque by saying the fact he helped himself to (unlike the range of facts he impugns) can be known from the armchair by a tacit sense of the weight of empirical evidence. But this at best shows that the tu quoque is sometimes ineffective, not that it’s less effective in the hands of a snob.

Third, you say: “Snobbery would require us to become rather less certain that our own reactions to thought-experiments are proper intuitions. There are just too many cases in history of various theorists being utterly certain that they had grasped, via the lumen naturale, some inmost metaphysical truth, which most of us now find unintuitive to begin with.” But why is this a special problem for the snob? Isn’t it equally a problem for those who blur together seemings, inclinations, and judgments?

John Bengson

Jonathan,

Just a thought about what you pejoratively call 'intuition snobbism', and its implications for Marc's defense of IDR against WSN's criticisms. (I personally would prefer that we have non-pejorative labels for philosophical views, but oh well.) In your comment -- and in your post over at the x-phil blog -- you're focused on a strong version of snobbism according to which genuine intuitions must have the force of necessity, have a strong modal tie to the truth, be experienced as internally justifying, etc. I wonder what you think of a much more modest snobbism (if we can even call it that) that simply denies that intuition is identical to any response to a question, any forced-choice snap judgment, or any output of a cognitive mechanism which produces judgments about philosophically relevant cases (as Justin suggested in his most recent post). This very modest snobbism seems to be all that is required for the IDR theorist to dodge the WSN bullet.

Let me put the thought another way. It seems that one need not deny that intuitions "are something very humdrum and everyday and people have them all the time without any special training or anything" in order to question whether the WSN results impugn intuition. Presumably, all one needs to do to question whether the WSN results impugn intuition is to question, as Marc has, whether WSN have adequately controlled for the possibility that the participants are reporting some mental state other than intuition. I take it that from the perspective of the empirical sciences and its concomitant methodology, such a question is quite reasonable. Correct me if I'm wrong, but in conducting empirical scientific research, the researcher cannot simply assume that his/her study is valid (studies what it purports to study); the researcher has to establish that it is valid. Given this, it strikes me that strong snobbery is not essential to Marc's basic line. All that is essential is that one hold, as is plausible, that it is not the case that any and all responses to questions are reports of intuitions. (Of course, one should at the same time admit that the WNS results might well pose a challenge to a version of IDR, however plausible or implausible it might otherwise be, which *does* identify intuition with such responses.)

(Incidentally, my apologies for suggesting that you were an empiricist in my earlier post. I might be way off, but I seem to remember an exchange over at the x-phil blog awhile back in which you suggested that on your view there is knowledge that we could call 'a priori' not because it's independent of experience, but rather simply because it's prior to rigorous empirical testing -- a la untested scientific hypotheses. Since many philosophers don't appear to consider this to be sufficient for a bona fide version of rationalism, I jumped to what may have been the wrong interpretation of the exchange. Sorry!)

marc moffett

So I am pretty much in agreement with Dan’s and John’s comments, but I will add my two cents anyway.

Before I start, I want to note that I don't take Jonathan’s original response to have adequately addressed the methodological concerns voiced in the post. For instance, I do think the WSN answer choices were bizarre; sufficiently so to, on their own, cast heavy suspicion on the findings. The mere fact that other people do this (albeit in a very different setting) doesn’t make it a good thing to do. Moreover, the “knows/doesn’t know” choice still seems considerably less confounding even if it does allow the subjective certainty reading. In any event, even if it is worse, that doesn’t make the choices you do offer any better. Nor do I think I am just “what if-ing” here. The reading of “only believes” as “believes w/out reason” is pretty darn salient. And I should add here that after consideration I think that my discomfort with the “really knows” answer is that it invites a gradable knowledge reading; that is, it invites a reading on which it contrasts with “sort of knows”. Students reading it this way will be confused by the decision choice because it might seem to them in many cases that the expected answer is “sort of knows” but it doesn’t appear anywhere. Hopefully I will get a chance to revisit some of the other issues when I have the time.

OK. So there are three immediate concerns on the table. The first is that intuitions may be less prevalent than WSNs methodology suggests. The second is that the particular forced-choice responses WSN elicit may not reliably track intuitions. The third is that even if the subjects are having intuitions, these may swamped by other considerations in their thought process.

Now Jonathan seems to think that if I am doubtful that any or many of the subjects in the WSN study were having responses backed by intuitions in any of the cases, I am thereby committed to having very demanding standards for intuitions and so being an intuition snob. But snobbery comes in degrees. Granted, I don’t accept that any old response to a written philosophical vignette, even a poorly written one, counts as an intuition. If that makes me an intuition snob, so be it. (In the same exact sense I am a wine snob: I only buy the finest boxed wines.) In fact, I believe that intuitions are quite common, so common that most routine cases of concept application involve them. Canonical cases look like this: the concepts involved in the example are all paradigmatic in their application as is the concept under consideration. The WSN coin flip case comes pretty close to being canonical (which is presumably why they used it as a control) and I would think that responses here involve and track intuitions.

Moreover, I think it bears emphasizing that these cases do count as intuitions: given the situation, it seems to us that the guy doesn’t know. Nor is this accidental: we couldn’t have exactly that same case and the guy fail to not know. Without this modal constraint, our judgment about the guy’s failing to know would be more like an inference based on incomplete evidence than an intuition. That is, it would go something like this: “In most cases of this sort the guy would not know, so I infer on probabilistic grounds that he doesn’t know.” But this modal claim isn’t highfalutin’. What is required is simply that the case be specified determinately enough that I can see whether or not the focus proposition is true in the situation under consideration. This seems to me to be an utterly familiar, even humdrum, phenomenon.

However, there is a tendency when criticizing IDR to focus on the harder cases first and to simply ignore all the humdrum cases. But if we are interested in the evidential status of intuitions, we must focus on the humdrum cases. Here is a claim: if you just focus on the boring cases, you can make a very strong case that intuitions as a class constitute evidence. (Of course, there is still room for a familiar kind of cultural variation here. If members of culture A are significantly more likely than members of culture B to think that tea leaf readings provide justification, then we would certainly expect to find differences in intuitions between the two cultures. For instance, you can’t build a culture neutral case showing that knowledge is not true belief, by giving a case in which someone believes that p on the basis of tea leaves and p happens to be true. Members of A will take this (or be more likely to take this) as a case of JTB, and they get the intuition that it is a case of knowledge. And there is a straightforward sense in which this is right: *given* that they understand the case as one which involves a justified, true belief with no monkey business, their intuition is dead on. The problem is that they are building in features of the case which are not intended to be there if it is to be understood as a counterexample to K = TB. Plainly, this sort of cross-cultural variation is not of interest and does nothing to undermine IDR.)

But the further we deviate from the canonical cases, the more care is required. And it is precisely here that we need to make explicit what is going on in the humdrum cases. I don’t think it is a particularly rabid brand of snobbery that maintains that as the cases become more cognitively or conceptually demanding or less well specified, intuitions about those cases will be less common. In such cases, unless you are consciously aware of what is going on, it is relatively easy to make a judgment based on hunches, guesses, heuristics, etc. This is especially true in forced-choice decisions. For in forced choice decisions, you have to answer one way or the other even if it is utterly unclear to you which way to go. So again, I hardly think it counts as a pernicious form of snobbery to say that someone with a modicum of “training” will be more likely to understand how to resolve ambiguities and shortcomings of a presented case or to recognize that they do not have an intuition at all about the case because they can’t properly resolve it.

So where do we stand with respect to the three complaints? First, I think that even on this unsnobbish version of intuitional snobbery there is reason to doubt that subjects are having intuitions in the Gettier case. That is, I don’t think you have to dial up the snob-o-meter very much at all to start having significant worries about this case. Moreover, even if they are having intuitions, the available responses make me worry about how to interpret the results. I also think that the second concern infects some of the other cases, like the Zebra and Conspiracy cases. (In those cases, I think that the students are likely reading an occurrence of “possibility” as (and this is commonplace) “likely” or “probable”, and this causes a significant confound. But even if that is not so, if they are reading “really knows” as a gradable concept, it may be that for them “only believes” is a better response even if they think the guy “sort of knows”.)

jonathan weinberg

(Just wanted to say I'm not blowing y'all off -- just a bit swamped right now! I'll hope to start responding to everyone piecemeal over the weekend!)

jonathan weinberg

Responding to Dan: you offer that "it seems plausible that this is just the kind of mental episode that we all naturally have, and report having, when asked to consider hypothetical scenarios." And basically I do take the line that you then suggest, "if we are naturally disposed to report snob-worthy seemings in response to hypothetical scenarios, then that should take care of Marc's worry that subjects aren't reporting their intuitions in the surveys." In short: the view that you are offering here isn't a form of snobbism! Snobbism requires positing that intuitions are fundamentally unrelated to people's ordinary reactions to hypothetical cases. I think what you've offered here is a form of "intuition populism".

On the subject of tu quoques, you write "...this at best shows that the tu quoque is sometimes ineffective, not that it’s less effective in the hands of a snob." It's not that snobbists are ineffective at tq's in general. Rather, I am claiming just that snobbism imperils this particular (and particularly popular) version of the tq, in which the restrictionists are accused of needing to use intuitions in order to make their argument. But once intuitions are made out to be a very special form of intellectual seeming, then that should attenuate the sense that whatever we're doing, we must be using intuitions to be doing it. I think maybe we're not disagreeing here -- maybe you just take me to be rejecting a broader class of argument than I intended.

Finally, in response to my raising the worry that the snobbist should become unsure about whether their own seemings are proper intuitions, you write, "Isn’t it equally a problem for those who blur together seemings, inclinations, and judgments?" I don't think it would be, because such a person isn't putting any _methodological_ weight on those distinctions. If someone were to blur those together (though I certainly don't think that anyone _should_ do so, in their working theory of the mind), then presumably they won't make it part of their methods to preference one of them over the others.

jonathan weinberg

Responding to John:
"I wonder what you think of a much more modest snobbism (if we can even call it that) that simply denies that intuition is identical to any response to a question, any forced-choice snap judgment, or any output of a cognitive mechanism which produces judgments about philosophically relevant cases (as Justin suggested in his most recent post). This very modest snobbism seems to be all that is required for the IDR theorist to dodge the WSN bullet." I'm afraid it isn't. What IDR theorist needs is a way out of thinking that the WNS study so much as even bears on the question who has what intuitions when. That the particular judgments we recorded are not themselves exactly identical to the seemings that philosophers pursue, is not really relevant -- no more so, anyway, than the fact that the numbers recorded from a sphygmomanometer aren't going to tell us exactly the condition of your arteries, nor is it the case that the results of AST, bilirubin, etc. tell us exactly what's going on in your liver. Those latter measurements are taken as good tests of those things, though, because they bear a significant non-random relationship to the things we really do want to know. Now, just how tight a relationship there is between the survey answers and intuitions proper is a matter of debate, and future empirical inquiry, but it certainly doesn't need to be anything close to identity (which is all your proposal would deny). Yet what the snobbist is arguing is that there is no relation between them. And that's what gets him into the sorts of trouble I described above.

Addressing your latter formulation, you write that "all one needs to do to question whether the WSN results impugn intuition is to question, as Marc has, whether WSN have adequately controlled for the possibility that the participants are reporting some mental state other than intuition." This is not so. What is needed is that the mental state being reported bears no indicator relation whatsoever to the subjects' proper intuitions.

Here's an illustration: if I'm studying perception, I might do a study where I give my subjects only an extremely brief exposure to the stimulus. And one might well wonder whether that short exposure is sufficient for them to form as complete a percept as we are trying to investigate, and we would do well to explore just what the relationship is between what subjects get on a quick glimpse, and what they get in more ideal viewing conditions. But it would be strange to claim that, whatever the subjects are having, it not only isn't any sort of perception whatsoever, but moreover isn't even related to perception. And it would be strange to take it as easily given that any between-group differences found in the quick-glimpse study do not make it even somewhat likely that we would find some parallel between-group differences in more ideal forms of perception.

Regarding the rationalism thing, I do sort of remember the thread in question, and I think I was trying to sort out someone else's use of "a priori" in a comment. But no apologies necessary -- it's not a view that I'm at all well-known for holding, and really my protestations of non-empiricism were meant to be more on a piece with such speech acts as: "There is something you do not know -- I am not left handed!" (as spoken by Inigo Montoya, that is). There is a deeper point lurking beneath, though, which is that sometimes proponents of IDR try to take comfort in the thought that surely at least some intuitions are in good standing, such as those of basic math and logic. But I do not think that that should offer them nearly the comfort that it often seems to, because it is an entirely defensible position (or, at least, a position at least one philosopher has tried to defend) that even though those particular sorts of intuitions really are in perfectly good order, nonetheless intuitions about far-flung hypothetical cases are not so.

marc moffett

Jonathan writes in response to John: "[Y]ou write that 'all one needs to do to question whether the WSN results impugn intuition is to question, as Marc has, whether WSN have adequately controlled for the possibility that the participants are reporting some mental state other than intuition.' This is not so. What is needed is that the mental state being reported bears no indicator relation whatsoever to the subjects' proper intuitions."

Hmmmm. And I would have thought that what was required is that WSN show that the mental state does bear some interesting relation to intuitions.

jonathan weinberg

"And I would have thought that what was required is that WSN show that the mental state does bear some interesting relation to intuitions." Well, gosh, what do you think the priors are, here? Which is more antecedently likely: that we just happen to have two utterly unrelated ways of having seemings about hypothetical cases (one of which, remarkably, fails to manifest at all in most ordinary cognition), or that what goes on in intuition proper is significantly cognitively continuous with what goes on in ordinary categorizings? Pretty obviously, the latter is the more likely initial hypothesis. This was the point of my perception science example; it's also Justin's point from earlier, I think.

marc moffett

But the perception analogy is extremely misleading. Granted that under very simple visual stimulation it is antecedently plausible that quick glimpses track judgments under more ideal circumstances. (This would be more closely analogous to, say, the coin flip case.)

But the more appropriate analogy to many of the other cases is to a quick glimpse of a complex scene and the judgment concerning a non-observable property of that scene. [Look at this party scene quickly. Is Suzy really mad at Joe or only drunk?] Here I think it is not particularly plausible to think that the subjects' judgments track the judgments they would have under more ideal circumstances nor that between-group differences would be maintained.

So, no. I don't think that the priors are at all high. Worse still, I think that the posteriors are quite low.

And where, precisely, did you establish that it *seemed* to the subjects that their answers were true? Again, I don't think that most IDR-ists think that intuitions/seemings are as sloppy as you think. Here is a quote from Jackson: "Also, following on from the point just made, we should note that the method of possible cases needs to be applied with some sophistication. A person's first-up response as to whether or not something counts as a K may well need to be discounted. [He then goes on to list a number of issues.] ... In sum, the busines of extracting the cases that counts as K's from a person's responses to possible cases is an exercise in hypothetico-deduction. We are seeking the hypothesis that best makes sense of their responses taking into account all the evidence" (pp. 35-36, From Metaphysics to Ethics).

jonathan weinberg

I like the case you suggest as a better analogy, and it works for my line of argument very well. Psychologists would generally think, as a matter of course, that differences between subjects at the quick processing would predict similar differences between subjects at the slower levels. The continuity between 'thin slice' judgments and longer-observation judgments has been quite extensively observed in phenomena as sophisticated as evaluating quality of teachers, or the outcome of negotiations, and indeed the example you gave of interpreting affect. Psychologists have observed, in many of these areas, that there is not a speed-accuracy trade-off: they are judgments that, if you're going to get them right at all, you're going to get them right away.

In general, even outside of the thin-slicing research, psychologists will expect a _continuity_ between quick versions of a task and slower versions, because very often it is either the same suite of mechanisms involved in the two tasks, or if there are distinct mechanisms involved, then the outputs of the initial ones will serve as part of the input to the later ones and/or influence the attentive processes that will significantly determine the inputs to the later ones. This is not any kind of absolute law, but again, it's where the priors are, which is all I need.

And it's really important to keep in mind that the question at hand is one of whether between-group difference in the survey reports have predictive value towards possible between-group differences in the intuitions. This sets the bar even lower. For even if any individual's survey report turned out to be only very weakly indicative of their particular later intuition, nonetheless the differences between groups could be expected to be predictive across the two tasks, for the reasons just mentioned. (That is, even if performance on the quick task tends to be largely different from performance on the slow task, the differences in the one would still be expected to track differences in the other.) You keep focusing on ways in which any particular person's survey response might happen to fail to reflect what their more reflective intuition would be. But that's just not generally relevant to the question at hand -- unless one had some substantive reason to think either (i) that such worries actually apply to nearly all the subjects nearly all the time (which is what the snobbist would be arguing) or (ii) that some of these factors apply in a lopsided way between our groups (which would be what it would be like to have a real, and not just bare-what-if, confound to suggest). Jackson clearly isn't a snobbist, in the passage you cited -- he's talking about an instance-by-instance elimination of data points, not a blanket discarding. And I've still yet to see anything that should make us seriously suspect any particular asymmetry for these sorts of factors between European-descended NJ kids and Asian-descended NJ kids.

marc moffett

Ahhh. That's what I get for playing the analogy game!

But do you think that hunches, guesses, gut-feelings, etc are part of the same "suite of mechanisms" as intuitions (and so track them) or that these are not methods the subjects were likely to use in responding? For my part, I don't think it is antecedently likely that hunches and what not track intuitions and I think that it is fairly likely that these are the mechanisms at work in some of the cases you report. Let me put that the other way around. I think it is clear that we shouldn't really evidentially on hunches, guesses or gut-feelings. Consequently, if its seemed antecedently plausible that these things tracked intuitions, then IDR wouldn't even be antecedently plausible.

Regarding Jackson. Is your position this: that the difference between being a snob and not is that non-snobs call every response an intuition but discount those responses that snobs don't count as intuitions? Seriously, though. In any event, snobbist or no, Jackson clearly wants (as I asserted earlier) signficant constraints on what counts. Moreover, if we have reason to believe (as I maintain is the case in WSN) that the experimental situation was flawed, then those reasons will apply in a blanket way to those subjects. This establishes my contention that at least Jackson-Chalmers style IDRists will not want to equate intuitions (or at least, evidentially serious intuitions) with forced-choice responses.

jonathan weinberg

"But do you think that hunches, guesses, gut-feelings, etc are part of the same "suite of mechanisms" as intuitions (and so track them) or that these are not methods the subjects were likely to use in responding?" Other than true guesses (as opposed to weak hunches), it seems to me to be a pretty open empirical question. Certainly there is wide variation as to the phenomenology of different seemings here. But I don't think that we have any reason to think that these are anything like _overwhelmingly_ the processes in play when subjects answer questions like the ones we posed. They all have plenty of experience sorting knowings from non-knowings, and the particular cases where we found interesting between-ethnicity variation were not wacky to the point of incomprehension. (I do worry about whether our low SES subjects could have been responding with little more than guesses to the more complicated cases that we found the main effects with for them; as I flagged earlier in the conversation, that's why I don't put much stock in those results now.) I think that Williamson is probably right that it's all a piece with ordinary counterfactual thinking. So even if _some_ of our subjects were out to lunch (and, hey, everyone needs a story about the 20% or so of WE subjects who said that we have knowledge in the Gettier case), there's just no reason to suspect that _on the whole_ they were out to lunch. You just wouldn't predict the data that we found, for example; you'd expect all the groups on all the questions to be closer to a 50/50 split, and we certainly didn't see anything like that. That's part of the rules of the game: if the proposed problem doesn't match the data at least as well as the target hypothesis, then it doesn't count as a serious challenge.

(This is also why the worries about our probes isn't so worrisome. Even putting aside my contention that we have very good reason to think that "really knows" is a perfectly good piece of colloquial English, there's nothing about any conjectured weirdness in it that would at all predict the patterns in our data. As I said, we can reasonably expect that it pushed _everyone_ somewhat towards the "only believes" response, but that just means that it isn't relevant to interpreting our observed between-groups effects.)

"Is your position this: that the difference between being a snob and not is that non-snobs call every response an intuition but discount those responses that snobs don't count as intuitions?" Snobbism is the position that asserts radical psychological discontinuity between intuitions proper and ordinary categorizations. As I said in response to John, merely distinguishing intuitions from the ordinary judgments isn't enough (which is why it just doesn't matter whether anyone would "equate intuitions with forced-choice responses"). Snobbism goes much further than that mere distinguishing, and that's what lets the snobbist reject our findings outright as irrelevant (but at, I tried to argue, an untenable cost). Jackson doesn't seem to hold that view, and indeed he _can't_ hold that view, since what he wants to get at is exactly the theories implicit in our ordinary categorizings. And our findings are still evidence that there will be between-group differences in those ordinary categorizings.

The comments to this entry are closed.

August 2011

Sun Mon Tue Wed Thu Fri Sat
  1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30 31      
Blog powered by Typepad