The Folk are Intellectualists

John Bengson, Marc Moffett, and I have just written a very short paper that explores folk judgments about know-how and ability. Anti-intellectualist Noë has recently suggested that the folk judge that x knows how to A only if x is able to A. But our research shows that at least in certain cases, the folk overwhelmingly judge that x knows how to A even though x is unable to A. These findings suggest that the folk have strong intellectualist leanings: in the minds of the folk, know-how seems to have very little to do with ability. Link to paper here.

Comments are welcome -- nay, appreciated!

Comments

[Edouard Machery]

Nice, short paper.

A few quick comments:

1/ Footnote 2 is controversial.

I would have thought that the folk' intuitions about knowing how have NO relevance for the truth of the intellectualist accounts of knowing how. The latter is a straigt empirical question. We should accept them if, and maybe only if, our best behavioral theories account for knowing how by means of knowing that (eg, Fodor 1968).

2/ The results of the second experiment are a bit puzzling. 46% of your subjects (how many, BTW?) gave one of the 3 following answers:

1 Jane does not know how and is unable to do

2 Jane does not know how and is able to do

3 Jane knows how and is able to do.

You do not report the distribution of answers for these three options. I guess few people chose 2. If this is so, most of your subjects (that is, among the remaining 46%) chose 1 or 3. But both 1 and 3 are consistent with Noe's position.

If this is correct, 54 % of your subjects gave an answer that is inconsistent with Noe's prediction, while almost 40 something % of your subjects gave an answer that is consistent with Noe's prediction. Not too far from half half. Can you really conclude that people have intellectualist tendencies?

3. how do you explain the difference between your first result (83% give your predicted answer) and your second result (54% give your predicted answer).

Edouard Machery

[Adam Arico]

First, I echo Edouard's curiosities regarding (2) and (3), especially the latter. I wonder if the difference in results between the ski instructor vignette and that of the skater can be attributed to the explicit "back story" in Jane's case. Whereas it's made clear that she has never successfully performed a quadruple (and never attempted a quintuple) Sachow, we're left with the possibility that Pat was once capable of performing the stunts he now teaches to others (though it's clear that he cannot presently perform).

It seems to me that leaving open the possibility (of having previously been able to A) might account for the robustness of the results in Pat's case. If that's so, might your findings support Noe's argument?

[Anibal]

I´m an advocate of the experimental philosophy movement (like an ideology and principle or like current philosophical trend), but a logical caveat in the planning of experiments to lead a certain conclusion (of one sort or another) is that in many cases (if not all) experimental philosophers use and consult their own folk psychology (plus expertise)to asses the patterns of responses from the folk psychology of polled people, so is necessary a very sharp "experimental mind" to untangle this (Knobe, Machery...)

I allude to the unwarranted creations of "back stories" or implicit assumptions attached in many cases when introducing main character´s features that bias responses (if you state that the instructor is the "best" rather than a mere instructor is probably more easy to dissociate the know how from the corresponding ability. Being the best means ideally being capable in achieving that for wich someone is atributed the label "best". And more seriously, if we cannot distinguish between different intensions in the use of verb actions: we can use "know how" as simply the procedural behavioural learning by skeletal musculature non-conciuosly control the most part of the time, or speak of "know how" as know what or -that[semantic knowledge] restrictive to know hows, perhaps intuitively understood the second option in the answers, cited by Machery, is possible. In the former sense of know how there is an anti-intellectual component in the latter not).

A secondary market (philosophy of "experimental pilosophy") is necesary to carry out metanalysis making even more reliable the data, to conduct experiments which analytically pose and define theoretical limits and conditions of possibility with respect of the potential patterns of responses to several linguistic input depending on context.

[John, Marc, and Jen]

Thanks for the provocative comments, Edouard. We'll follow your numbering in our responses.

1. Yes, our observation in note 3 is controversial in certain circles. That's why it's in a footnote. In any event, since our stated aim is to test Noe's prediction, the paper is focused on the question of what the folk would say about know-how and ability. It is a further philosophical point, and one we don't attempt to establish or even apply here, that what they say might have some philosophical relevance.

2. This is a great question. Unfortunately, we lacked the space to report all of the statistics in the paper. But since this is an important issue, we should take the time to elaborate/clarify. As you point out, there are four possible combinations:

(a) Jane knew how but was unable.
(b) Jane knew how and was able.
(c) Jane did not know how and was not able.
(d) Jane did not know how but was able.

The distribution was as follows: out of 57 participants, 31 endorsed (a), 17 endorsed (b), 8 endorsed (c), and only 1 endorsed (d). Given the way the vignette is structured there is only one response consistent with extant intellectualist views: (a). The chance of participants randomly choosing (a) is .25, which means that what we actually found (54%) is significantly higher than chance (p < .001).

The anti-intellectualist, on the other hand, would predict one of two outcomes: (b) or (c). The chance of either getting (b) or (c) is .50. This means that what we actually found (44%) is not significant (p = .427). [Incidentally, nothing changes if we treat (b) and (c) separately: neither come in above chance. Specifically, 30% endorsed (b), which is not significantly above (or below) chance (p = .242). 14% endorsed (c), which is significantly below chance (p = .034).] So, looking at the data this way seems to disconfirm Noë’s prediction, and consequently shows that the folk have intellectualist leanings.

Of course, this story is problematic for the following reason. While (d) is strictly speaking consistent with intellectualism (broadly construed as the thesis that know-how does not require ability), the way the vignette is structured, intellectualists would only predict (a); they would not predict (d). Anti-intellectualism, on the other hand, faces a split vote: the way the vignette is structured, anti-intellectualists would predict either (b) or (c). Because this then raises the bar for the anti-intellectualist (making the chance ratio .5 instead of .25), it may be, in a way, unfairly setting things up in the intellectualists’ favor. For if we had set up the case in such a way as to make it completely obvious that, e.g., Jane was not able to do a quintuple Salchow, then the anti-intellectualist would have been forced to chose (c). This would put intellectualism and anti-intellectualism on even footing, as far as chance goes (either .25 or .5, depending on how the choices were structured). In such a case, the difference between 56%-44% would be non-significant. For these reasons, while the Jane case does not provide support for the anti-intellectualist view, it may not support the intellectualist view either.

That said, it is worth pointing out that the pattern of anti-intellectualist judgments is quite interesting. Given the way the vignette is structured, we should expect the anti-intellectualist to choose (c). After all, though the vignette isn’t perfect, it still makes it pretty clear that Jane is not able to do a quintuple Salchow (e.g., we explicitly state that she could not land a quadruple and has never even tried a quintuple). So, presumably the anti-intellectualist would predict that participants would judge that Jane is not able to do a quintuple Salchow, and since ability is required for know-how, conclude that Jane does not know how either: no ability, no know-how. Yet the majority (68%) of the anti-intellectualist participants chose (b) instead. Anti-intellectualists owe us an answer as to why this is so. One explanation is that the desire to attribute know-how to Jane was strong enough to override the “no-ability” information provided in the vignette. That is, despite the case being such that Jane lacks the ability, participants want to say, in line with intellectualism, that she knows how. On this basis, they then judge that she both knows how and is able -- contrary to what we would expect if the anti-intellectualist idea that ability attributions drive know-how attributions was correct. Consequently, this explanation suggests that even those participants who gave anti-intellectualist responses may have intellectualist leanings.

To tease all of this out, it would be fruitful to run further studies with more carefully worded questions and vignettes. That’s what we hope to do in the coming months.

3. There may be a number of reasons that the Pat case was 83% and the Jane case was 54%. For instance, the following seems like a plausible explanation. The Pat vignette explicitly says that Pat cannot do the jumps. As a result, the judgment that Pat is unable to do the jumps is basically the result of reading comprehension. The open question, then, is whether Pat knows how. This is different from the Jane case -- though it states that she is not able to do the quadruple Salchow and has never tried the quintuple Salchow, it does not explicitly state that Jane is unable to land the quintuple (but, again, we believe that the vignette still makes it pretty clear that Jane is not able to do a quintuple Salchow). Since the Jane vignette is less clear, there is quite a bit more room for participants to judge that she is. That is, in this case whether Jane knows how and whether Jane is able are both relatively open questions, which means that we're bound to get messier results. There may be other reasons for the slightly divergent results, but we suspect that this difference in the vignettes has a lot to do with the difference in percentages.

-----

Adam, thanks for the comments and question. Hopefully we've addressed a few of your worries above. That leaves your comment about the asymmetry in the availability of a back-story between the Pat and Jane cases. You observe that the Jane vignette makes it "clear that she has never successfully performed a quadruple and never attempted a quintuple Salchow" so a back-story in which Jane was once able to do a quintuple Salchow is not available in this case. Since you think that a back-story in which Pat was once able to do the ski jumps is available in the Pat case, you suggest that the fact that the Pat vignette leaves "open the possibility of having previously been able to A might account for the robustness of the results in Pat's case". In short, your suggestion is that the availability of a back-story in which x was once able to A will increase the likelihood that the folk will judge that x is able to A.

Correct us if we're wrong, but it seems that this suggestion would predict that the folk would be more likely to say that Pat is able to do the jumps than that Jane is able to do the quintuple Salchow, since there is a back-story available in the Pat case (in which Pat was once able to do the jumps), but there is no such back-story available in the Jane case. But this is precisely the opposite of what we found. In the Pat case, 7% said that Pat is able to do the jumps. In the Jane case, 32% said that Jane is able to do the quintuple Salchow. Since the apparent availability of a back-story accompanied *less*, not more, attributions of ability, as one would expect if the back-story story was at all explanatory, it seems that the back-story story isn't a viable response to our findings.

So, in response to your question: no, our findings do not support Noe's position.

[Adam Arico]

Thanks, John, Marc, and Jen, for the reply. I wasn't actually suggesting that "the availability of a back-story in which x was once ale to A will increase the likelihood that the folk will judge that x is able to A." Rather, I was alluding to the possibility that the back story for Pat leaves open the possibility that folk will judge that x knows how to A because he _was once_ able to A, even though he's _no longer_ able to A. I take it that this reasoning would be anti-intellectualist, though I could very well be wrong about that.

Not being very familiar with the literature, I assume that "being able" is flexible in temporal application; that is, I expect Noe would classify one as "able to" even if present circumstances prevent it (whether due to restricted space, lack of equipment, broken limb, or old age). I guess what I was emphasizing is that one can deny the present tense "Pat is able" but still hold the belief that Pat, in the past, "was able" (consider our typical notion of coaches as ex-atheletes, past their physical prime, etc. "I know Hank Aaron might not be able to hit a 90 mph fastball anymore, but back in the day...").

If Pat _was_ able, might that influence whether folk judge him to know how? And if so, does that lend support to the anti-intellectualist thesis or yours?

[marc]

Hi Adam, Interesting suggestion.

I believe that the standard neoRylean anti-intellectualism maintains (i) that knowing how to psi is, or requires, a current ability ability to psi, and (ii) that abilities are in some sense robust (e.g., complex dispositions). Condition (ii) is required in order to account for cases where an individual appears to know how to psi, but cannot presently psi. For example, a pianist who knows how to play Rachmaninoff's 3rd, but who has recently lost one of her hands in an accident. We took your suggestion, consequently, to be that a back story in which Pat was once able to perform the ski jumps might be sufficient to ground a present robust ability to perform them. This approach appears to be inconsistent with the data.

Your suggestion, however, was somewhat different, namely, that knowing how to psi requires only the present or past ability to psi. Subjects would then say that Pat knows how in virtue of his (presumed) past ability but still deny that he isn't able. Moreover, this view would be (or, at least, could be) anti-intellectualist insofar as it grounds know-how in ability (and not in, e.g., propositional knowledge). Do I understand you correctly?

If so, then, as far as I can tell (John/Jen can correct me if I am wrong), our data does not rule out your position. Nevertheless, I am doubtful this position is philosophically defensible (unless there is some less naive version floating around). For instance, one implication of the view is that once x knows how to psi, one cannot cease to know how to psi. This seems clearly wrong. Stroke victims, for instance, often cease to know how to do things that they once knew how to do.

Perhaps, however, there is some more sophisticated view: knowing how to psi requires being able to psi or having been able to psi + X, where X specifies the conditions under which one may lose the ability without losing the know-how. I am not aware of any such proposal in the literature. Moreover, the conditions specified in X would have to be such that they remain consitent with a broadly anti-intellectualist position.

In any event, I have my doubts that this could be achieved in reasonably non ad-hoc manner, but I am open to suggestions.

Cheers, Marc

[marc]

Oops! Typing too fast. The second sentence of the second paragraph should read: Subjects would then say that Pat knows how in virtue of his (presumed) past ability but still deny that he is (currently) able.

[Adam Arico]

Hi, Marc. This reading is closer to what I was suggesting, though I should be clear that I'm not so much endorsing this version of intellectualism as throwing out a potential explanation for the variance between the 83% in Pat and the 54% in Jane.

If the intellectualist wants to maintain the "once able, always know how" stance, it doesn't immediately strike me as ad-hoc to include some ceteris paribus (non-stroke, non-amnesia, non-brain-damaged, etc) condition. I'm not sure exactly how those details would play out either--perhaps by specifying that the motor cortex and/or memory regions not suffer damage or massive degeneration--but the idea seems to be possible. More importantly, it seems consistent with (what I know of) Noe's position.

[marc]

Adam,

I see. Your suggestion is that (i) if having had the ability to psi implies (ceteris paribus) knowing how to psi even if (ii) one is not currently able to psi, then if the subjects were telling themselves a back story in the Pat case we would expect a lot of "knows-how, but is unable" judgments. Which is what we found. Since no such backstory is available in the Jane case, we wouldn't necessarily expect this.

Excellent! Again, I don't think that we can rule out the hypothesis on the basis of the evidence we have at hand.

A couple of caveats. First, I don't think that this idea is consistent with Noe's view. Noe, like most neoRyleans, seems to want to identify know how with a current ability. So he couldn't very well deny that Pat has ability but does have the know-how. His view is rather that having the right neural make-up implies having the ability, even if one isn't able to act on the basis of that ability.

Second, suppose that I grant the following conditional: if x once had the ability to psi and x hasn't suffered any major neural degernation or alteration, then x now knows how to psi. Still, while we have located a certain sufficient condition for knowing how to psi, we haven't done much to illuminate the concept of know-how. To veer away from experimental philosophy for a moment, consider a Davidsonesque swampman counterpart to x, call him x*. Presumably, if x know how to psi, then so does x*. But x* doesn't have the right causal-historical connections to past abilities (since x*, ex hypothesi, has no past at all!). If that is right (is it?), then it suggests that there is some nonhistorical property of x--a property shared by x*--which is actually the metaphysical basis of x's know how.

Now, whether or not you think the swampman story is ultimately intelligible, I think it highlights the underlying reason why we would not want to account for know-how in historical terms.

[John]

Hi, Adam. A quick thought: It seems to me to be difficult to determine what Noe's anti-intellectualist view actually is. For instance, when discussing the ski instructor case, he writes, "One good reason to think the ski jump instructor knows how to do the jumps, EVEN THOUGH SHE CAN'T DO THEM, is that she used to do them all the time, with a high level of proficiency" (284). Here we get a sort of "previously able, knows how" view.

But, in the same paragraph, Noe seems to claim that the ski instructor must still have the ability (even if she can't exercise it right now) if she knows how, since ability is necessary for (identical to?) know-how: "The instructor knows how to do the jumps and this knowledge consists in her actual ability to do them...Once again, the linkage between knowing-how and the possession of abilities is left intact" (284). In this passage, Noe is not claiming that the ski instructor had or possessed (past tense) the ability, but that she has or possesses (present tense) the ability. A footnote which appears during this whole discussion seems to summarize his view: there, he writes, "To know how to do something is to have the relevant ability". So, here we get an explicit denial of any "previously able, knows how" view.

Others may have a better grip on Noe's view than I, however. I'd like to make sure I'm clear about precisely what the most plausible non-trivial anti-intellectualist thesis is, if it's not what Marc suggested in his 4:41 post, so I'd be interested in any thoughts you might have about how to cash out its main thesis.

-John

[Adam Arico]

I'll grant you that the evidence doesn't support the type of neoRylanism you describe, but that doesn't change the possibility that the folk see past ability as contributing to current know how. It might not technically make them anti-intellectualists of the Noe variety, but it might eliminate them as pure intellectualists, too. That is, while folk might not respond in a way that fits neatly into Noe's prediction, if they are more likely to attribute "know how" to those who've been previously able, then they wouldn't seem to fit neatly under the intellectualist umbrella. Of course, I'm not at all familiar with the literature, so I could very well be misunderstanding both positions, and thereby wrong about the implications of the experiment.

[Jen]

Sorry, fellas. I don't meant to change the subject, but I just thought I'd follow up on our response to Edouard by giving the same details for the Pat (skier) case as the Jane (salchow) case:

Unlike the Jane case, the Pat case is *very* clear cut. The distribution is as follows (kh/ability)

yes/yes - 4/57 (7%)
yes/no - 46/57 (81%)
no/yes - 0
no/no - 7/57 (12%)

If we split these judgments, as we did with Jane, into intellectualist and anti-intellectualist responses we get 81% vs. 19%. 81% is significantly above chance of .5 (p< .001) and 19% is significantly below chance of .5 (p < .001).

Of course, claiming that the yes/no response is uncontroversially intellectualist may be ignoring some of the subtleties that have emerged thus far in the ongoing debate, but I think I'll just stay the heck out of it!

[Adam Arico]

Before I get too deeply entrenched in the anti-intellectualist position, let me just say that I'm agnostic on the intellectualist/anti-intellectualist debate, per se. Though I've found myself defending the possibility that the evidence could support an anti-intellectualist thesis (which, having had Noe's position briefly elucidated, I'm convinced is not the case), I'm by no means defending anti-intellectualism itself.

I still maintain that part of the reason that 46 of the 57 participants say that Pat knows how AND isn't able could be because the vignette leaves open the possibility that Pat once could. However, I no longer think that this possibility could lend itself to an anti-intellectualist interpretation (thanks to Marc for clarifying why Noe's thesis doesn't allow for this).

I look forward to seeing if the same percentage holds when we fill out Pat's background in a way that explicates that Pat was never able to perform the stunt(s).

[marc]

Adam, I really appreciate your thoughts on the issue. Hopefully, we can run some further tests to control for these issues better. One reason why we would want to do this is that the prediction you make connects up in a certain way with the intellectualist position John and I defend. Basically, our view has two parts. First, the following analysis of know how:

x knows how to psi iff there is a way w of psi-ing such tht (i) x knows w, (ii) x knows that w is a way of psi-ing, and (iii) x has reasonable mastery of the concepts contained in a (possibly implicit) correct conception of w.

Second, we want to claim that a conception of a way w has non-trivial constraints on the perceptual modalities associated with the conception.

Perhaps this second idea is best articulated by illustration. Consider Pat. On our view, Pat must have reasonable mastery of a correct conception of a way of doing the jumps. Now it might be that there is a purely descriptive such conception stated in physiological terms that would underwrite a know how attribution. However, this descriptive conception might be so complex that few, if any, people can actually be said to have reasonable mastery of it (i.e., of the concepts in it). Certainly not a humble ski instructor. If this is right, then Pat's conception of a way of doing the jumps is presumably at least partially demonstrative and kinesthetic. And arguably one can't attain reasonable mastery of such demonstrative/kinesthetic concepts without at least HAVING HAD some physical abilities of the same general sort.

So, thinking aobut your comments in light of my more general views on the matter, I find myself in agreement with you. I think that the subjects must be assuming that Pat is at the very least a highly competent skier who is probably able to do some advanced jumps (even if he can't do the ones he is teaching). And the further we get away from something like this back stroy, the harder it will be to illicit judgments of know how. But this is exactly what (at least our version of) intellectualism predicts!