Monday, March 25, 2013

Problematizing the B Condition on Knowledge

Central to contemporary epistemology is the question of what it is to know something. And the orthodox starting point in discussions of what it is to know something is the "JTB account" of knowledge: Knowledge is justified (J) true (T) belief (B), plus maybe some fourth condition to take care of weird cases where justified beliefs are true merely by accident (so-called Gettier cases). Discussion tends to focus on how to understand the J condition and whether some further fourth condition is necessary. The truth and belief conditions are typically regarded as unproblematic.

In a forthcoming paper, Blake Myers-Schulz and I pick up a mostly-cold torch from Colin Radford (whose seminal work on this topic was in the 1960s) and challenge the belief condition. Can one know that something is the case even if one doesn't believe that it's the case? We offer five plausible cases (one adapted from Radford) along with empirical evidence that our intuitions [note 1] about these cases are not idiosyncratic.

This paper has already drawn several follow-up studies, some critical and some supportive -- but interestingly, even the critical studies can be read as contributing to an emerging consensus that problematizes the belief condition. (I don't predict the consensus will last. They never do in philosophy. But still!)

First, to give you a feel for it, our cases:

1. An unconfident examinee who feels like she is guessing the answer but non-accidentally gets it right;

2. An absent-minded driver who momentarily forgets that a bridge he normally takes to work is closed and continues en route toward that bridge;

3. A prejudiced professor, who intellectually appreciates that her athletic students are just as capable as her non-athletic students but who nonetheless is persistently biased in her reactions to student-athletes;

4. A freaked-out movie-watcher who seems to have the momentary impression that the scenario depicted in a horror film is real;

5. A self-deceived husband who has lots of evidence that his wife is cheating and some emotional responses that seem to reveal that he knows this, but who refuses to admit the truth to himself.

Now maybe not all the cases work, but we think in each case there's at least some plausibility to the thought that the person in question knows (that Queen Elizabeth died in 1603, that the bridge is closed, that athletic students are just as capable, that aliens won't come out of her faucet, that his wife is cheating) but does not believe -- at least not as fully and determinately as she knows. And lots of undergraduates seem to agree with us! So we think the B condition on knowledge should at least be open for discussion. It should not be regarded as uncontroversially nonproblematic.

Follow-up studies (e.g., here, here, here, and here) have added some new plausible cases. Our favorite of these is:

6. A religious fundamentalist geocentrist who aces her astronomy class -- seeming to know that Earth revolves around the sun but not to believe it.
Although some of these follow-up studies are pitched as in agreement with us and others as critique, we think there's actually a pretty clear thread of consensus through it all, from a bird's-eye view:

Knowledge requires some sort of psychological connection to the justified, true proposition -- something broadly like a belief condition; but it doesn't seem to require full-on act-it-and-live-it-and-breathe-it belief. However reasonable it might be to think the Earth goes round the sun, that fact has to register with me cognitively in some way if I am to qualify as knowing it; but the fact needn't play the full functional role of belief as envisioned in behaviorally-rich accounts of belief like my own. But how exactly should we should conceptualize this somewhat weak but broadly beliefish psychological-connectedness condition? At this point, that's wide open.

Blake and I suggest that one must have the capacity to act on the stored information that P; Rose and Schaffer seem to suggest that what's crucial is that the information be "available to the mind"; Buckwalter and colleagues suggest that one must believe but only in some "thin" sense of belief; Murray and colleagues suggest that one need to be disposed to "assent" to the content. None of these approaches are well specified (and I've simplified them somewhat; apologies). Figuring out what's going on with the B condition thus seems like a potentially fruitful task that brings together core issues in epistemology and philosophy of mind.

-------------------------------------------------

[note 1] Yes, I use the word "intuition". Herman Cappelen has me worried about the term. But I stand firm!

32 comments:

Anderson Brown said...

The fundamentalist geocentrist who aces the astronomy exam has mastered the material. One might say "She knows the material," but that would be an idiomatic use of "knows." My intuition is that she can't be said to know that it is true so long as she believes that it is not true. A problem with a couple of the earlier examples is that they equivocate between "believes" and "consciously believes." The insecure student, the forgetful driver and the cuckolded husband might all be described as holding the relevant belief after all. Interesting, but not yet persuasive.

Wesley Buckwalter said...

No doubt 'belief' is polyonymous, as we all know from teaching 101! In our paper we suggest that two fundamental folk psychological conceptions of it are revealed by these results...and also responsible for what seem like intuitions contrary to traditional entailment. (Not that the occurrent/dispositional distinction might not also be important).

But despite what's going on here generally, I've always worried about a few things in Geocentrist (one of the cases with the strongest K/~B data). For one, it invokes religious belief (def one of the ways 'belief' is polyonymous) and moral things like respect for parents. For another, it involves "the earth revolving around the sun", and in some other cases the belief that "2+2=4". Perhaps there is artificially high K because these are things basically everyone just doesn't not-know. Curious to see what would happen to K ascription if the relevant claim was something like the "magnetic field at the center of an arc is (μ0 I φ) / (4 π r)".

Eric Schwitzgebel said...

@ Anderson: The cases are not designed to be compelling on their own, especially to trained epistemologists, whose intuitions will be tuned in certain ways. What you say seems one reasonable way to go. My thought is only: That's not straightforwardly the best way to go. There are other possibilities here that aren't too strongly at variance with intuition.

In my own case, I was drawn to the view by a theoretical position: I have a view of belief as deeply implicated with behavior, so that one doesn't normally believe P unless one lives in a P-ish way; but it seems to me that knowledge can be attributed with less than that, especially when we want to hold someone morally accountable for not acting on information that is available to her as a possible guide for action.

So, with a theoretical view like that in hand, and enough muddiness about the intuitions so that one can't dismiss K-not-B cases as counterintuitive, one is free to deny the B condition if B is strictly construed.

Eric Schwitzgebel said...

Wesley: Yes, you're probably right about some of the factors that might potentially be influencing the geocentrist case. On the other hand, your alternative might swing too far the other way. Another concern about the geocentrism case is that at least in some versions the prompt contains some mental state attribution terms ("accept", "hold") that are probably a little too close to synonymous with "believe".

Wesley Buckwalter said...

Good point, it's definitely potentially worrisome to include those terms as well!

Anonymous said...

I always thought the belief condition was understood broadly, where 'belief' is any attitude with a mind–to–world direction of fit. So it includes explicit beliefs, what someone thinks, implicit beliefs, what is accepted or held by a person, deep-down emotional truths, etc.. Any of these is enough to say that a person believes (providing it is not misleading to use the word without further qualification), and so given other things, does know that P. I figure spelling out this general idea is one of the up shots of the work, which ill read soon probably.

Anyway it would be interesting to see when and why one of these belief-like states rises above the others (whatever they are); when the emotional truth trumps the explicit belief, when the occurrent thought loses to the dispositional, etc.. Even if the goal is to come up with a single answer about what a belief really is, it would clarify how to go about interpreting the data. Perhaps corpus data would be relevant here.

Unknown said...

The Iowa Gambling Task provides another interesting example of apparent knowledge without belief. 'Normal' participants have an unconscious (and correct) inclination to avoid picking from the "high risk" card pile, even though they can't articulate any 'belief' of the kind if you ask them.

It seems to me that JTB represents an account of knowledge appropriate to Plato's time, and making a couple ad hoc modifications won't be enough to save it. Wouldn't it be better to scrap the whole project and try to build a new socio-cognitive account of knowledge that takes advantage of what we've learned in the 2300 years?

Daryl said...

Eric, I believe we can develop your dispositional account of belief into a concept of knowledge which is adequate to the cases stated above, and even join it possibly with your considerations elsewhere of the United States as a conscious entity. Accordingly, I would suggest that knowledge is simply belief writ large upon social bodies. Being dispositional, it would then allow for lapses in belief at the level of individual human beings without any problem.

Conceiving of knowledge in this way has the added advantage of accounting also for the ever provisional state of scientific knowledge in particular: the explicit goal of science is, in effect, to map out systematically what the universe itself believes, and to do so at ever increasing levels of abstraction through ever modifiable dispositional stereotypes.

Unknown said...

What about replacing the B condition with 'acceptance' in Jonathan Cohen's sense? Someone accepts that p if she premises her deliberations with p. So knowledge becomes justified, true (enough) acceptance. Since acceptance is to a large extent voluntary, and since we can hence change what we accept based on context, knowledge will become a very contextual thing.

The cases (in descending order of plausibility?):

4: She knows, because she isn't going to base any of her deliberations on the proposition that aliens will come out of her faucet.

6: She knows, because she premises her deliberations (at least in astronomy class) with the claim that the earth revolves around the sun.

2: She knows in the normal course of things that the bridge is out. (When at home, she remembers and gives directions that involve avoiding the bridge.) But as long as she is forgetting, she doesn't know that the bridge is out.

5: Just the reverse of case (2): He doesn't know that his wife is cheating on him in the normal course of things, but on those irritating occasions when he starts thinking about it, he does.

3: In between (2) and (5). She's constantly flipping back and forth between knowing and not knowing. The football star comes to her office and she treats him like he's simple. Then, she doesn't know. When her colleagues say disparaging things about student athletes, she objects. Then, she knows (that student athletes are capable).

1: No such test-taker exists?





Eric Schwitzgebel said...

Thanks for the interesting continuing comments, folks!

@ Goeff: I like the Iowa Gambling Task example. A nice complement to the prejudiced professor case!

@ Daryl: That's an interesting idea. I always enjoy it when I see my work come together in unexpected ways! I think it's a feature of philosophy that there are harmonies among people's interests and views that are often invisible to the people themselves but which sometimes reveal themselves down the road. One tweak to that picture you suggest: In some cases the knowledge might be embodied, with non-social parts of the environmental context playing an important part of the role, rather than strictly social.

@ Brian: Yes, Cohen and few others have suggested "acceptance" instead of belief. I contemplated mentioning this, but I didn't want to overburden the post. Cohen's "acceptance" (and acceptance more generally) does come apart from belief, but it doesn't map very well onto most of the knowledge-without-belief cases above. In fact, Cohen rejects Radford's examinee case. But the distinction might do well for scientific knowledge which sometimes seems to involve acceptance rather than belief. On your suggestions for the six cases: Your top four seem to me to have substantial merit (though I don't know whether I would ultimately accept them), but 3 and 1, I think, strain credulity. (I briefly develop an objection to 3-like flipping responses in my 2010 paper "Acting Contrary...".)

Daryl said...

Eric, your tweak would be included in the notion of science as the cartography of non-human beliefs. An individual's knowledge would then be understood as a human belief in league with a belief attributed to some part of the universe. However, the individual may not share your attributive tendencies with respect to his own belief or the belief of that part of the universe. That leaves open the question of whose attributions shall dominate, which I think calls forth the sense of confusion attending those special cases stated above.

Anonymous said...

Geocentrist's knowledge is approximately the same as mine about Sherlock Holmes: willing suspension of disbelief. Several of the others hinge on the usual ability to hold mutually contradictory beliefs, that is know one thing at a time

Charles T. Wolverton said...

Since I consider "belief" to be one of those words that causes more trouble than it's worth (at least in formal discussion), rather than just problematize it I'd eradicate it. Like Eric, I favor instead an approach that employs context-dependent behavioral dispositions (CDBDs). So, rather than viewing the survey results in terms of the believing/knowing distinction, I'm going to argue that they suggest an understanding of "knowing" that fails to account properly for context. By "knowing" I mean having a CDBD to assert something and the ability to successfully justify that assertion by achieving consensus acceptance by a relevant community (what I take to be the Sellarsian sense).

In example 2, re commuting to work Ben has two CDBDs, one for the context "bridge open", the other for "bridge closed". While driving that morning, the context included a distraction which caused him to effect the wrong CDBD, a common context error due to inattention. But that has nothing to do with Ben's "knowing that" the bridge is closed, which he clearly does. IMO, this example warranted a very large number of "yes" answers, and that it didn't suggests failure to adequately account for context.

Example 3 is similar. The teacher's behavior in two contexts is described. However, only her behavior in one of those is relevant to the survey question, and all relevant evidence presented is supportive of her assertion. Again, the surveyed students seem to be letting contrary behavior in an irrelevant context dilute the evidence in a relevant context.

Example 1 also involves two contexts and additionally injects "inner dialogue". But the CBDB (writing the correct answer) that was was manifest in Kate's pre-exam performance was in place even in the stressed context of the exam, and if challenged, Kate could have justified her correct answer with ease. The survey results suggest to me that the respondents heavily weighed Kate's pre-test performance and/or appropriately discounted or ignored the inner dialogue.

In example 4, respondents seem to be ignoring the evidence. The described scenario offers only behavior that constitutes evidence against the assertion that Jamie knows that only water will exit the faucet. That so many respondents answered "yes" perhaps suggests an understandable reluctance to accept such kooky behavior by a presumably otherwise normal person as other than aberrational. But an ascription of knowledge based on the evidence seems completely unwarranted.

In example 5, if Tim weren't in denial but instead was asserting P, he certainly could justify that assertion. But he is depicted as asserting ~P, so strictly speaking he can't "know" P. The high level of confidence among the respondents that he does seems to suggest confusion between what they think Tim should know and what he claims to know, a failure to distinguish contexts: a state of denial versus a state of acceptance.

I take justification to occur within a community so that "knowing" is relative to that community. The student in example 6 could presumably justify her geocentrism within her religious community by referring to scripture, religious teaching, et al, and those reasons would be accepted. So, within that community she would be deemed to "know" that our solar system is geocentric. But if asked within her academic community to justify her test answers, she might say that her answers were not assertions but merely the responses required in a context in which the justification for "knowing" course material is, for better or worse, successfully executing prescribed procedures. So, she is being neither hypocritical nor contradictory in meeting the standards for "knowing" required by her two communities, even though the "knowns" in the two communities are incompatible.

Charles T. Wolverton said...

BTW, I agree in principle with Wesley's comment about "2=2=4", only I'd simply describe such an assertion as nothing more than a well-entrenched behavioral disposition. One who thinks such assertions constitute knowledge might want to try getting a legitimate justification for one from a non-mathmatician.

Eric Schwitzgebel said...

Interesting analyses, Charles! I agree, of course, that CDBD's are what's really going on -- so once they are all made clear, it's a pragmatic question of what's the clearest way to deliver the summary to listeners.

Daryl said...

You mean, of course, what is the most effective way to infect listeners with a particular attributive disposition with respect to someone knowing something.

Eric Schwitzgebel said...

Exactly, Daryl!

Michael Schmitz said...

Nice work, very important to challenge the orthodoxy in this area! One way of thinking about the psychological condition necessary for (conceptual, theoretical) knowledge that seems promising to me is in terms of a relation of acquaintance with a fact. I think the subjects in your examples may be said to be all acquainted with a fact even if they lack the conviction and the attended pervasive behavioral dispositions associated with belief - in at least some interpretations of that notion. This acquaintance would still need to be sufficient to ground a capacity to retrieve that information though. If that capacity is lost completely, in all possible contexts, I don't think we would still count the subject as knowing. But it may be weaker than that associated with believing, where believing is more than just a generic notion of a state with mind-to-world dof - the latter notion being widespread among philosophers, but not among laypeople, I would assume. One nice thing about this idea is that it allows us to give an account of factual knowledge that is parallel to non-conceptual theoretical knowledge, for example, to knowing people in the sense of being acquainted with them. We may be said to know somebody in this sense even if we fail to perceive, that is, recognize that person in some contexts - though not, I think, in all. So your cases (or at least some of them) would be analogous to a situation where we know a person, but fail to recognize her, with recognition being the counterpart to belief. Acquaintance of course also entails existence of the person or fact and thus also truth. I suspect that this relational aspect of knowledge is at least as important as that of being justified in the sense of having reasons that the philosophical tradition has been so obsessed with.

Eric Schwitzgebel said...

Michael, that is brilliantly well-said and well-thought! Although I remain a bit nervous about the notion of "acquaintance" (especially because of how it was used by Russell), I am very much tempted, as you say, to compare acquaintance with facts with acquaintance with people, including commitment to a minimal kind of usability or recognitional capacity. Have you written this up anywhere?

Michael Schmitz said...

Thanks, Eric. And right, Russell probably ruined the term 'acquaintance' for several generations of philosophers. I certainly don't want to have much to do with his notion of acquaintance either. Similarly, one might be wary of talking of a 'relational' account because this term has recently been appropriated by disjunctivists. On the other hand, people will tend to misunderstand you in certain ways no matter what terminology you use, and why leave nice words to people with confused theories? So I rather tend to err on the side of trying to re-appropriate them.

I don't have a paper yet where I work this out, but I hope to draft one soon. But there is a paper covering similar ground which might be of interest to you. In this paper I develop an account of actional and perceptual skills plus my version of a dispositional account of belief and intention starting from a critique of Searle's notion of the background:
http://academia.edu/1059670/The_background_as_intentional_conscious_and_nonconceptual

Charles T. Wolverton said...

"If that capacity [the ability to retrieve certain information about an "acquaintance" relation] is lost completely, in all possible contexts, I don't think we would still count the subject as knowing."

The idea of "acquaintance" appears to be another approach to aggregating and the quote highlights why I consider context-dependence and justification critical. Aggregating and ignoring justification results in an unsatisfactory concept of "knowing". "I know but I forgot" is always suspicious.

If in some context I sense that I "know" P but can't justify P in that context, then strictly speaking I don't "know" P. To assume otherwise is to assume implicitly something along the lines of the "belief box" in which there can be a "knowing" access to which can come and go depending on context. But if one "knows" P in all contexts but can justify asserting P only in some, what role is justification playing?

One who specifies that "knowing P" is context-dependent and requires the ability to justify asserting P seems stuck with the conclusion that knowledge can come and go as context changes. The night before the exam, Kate could have - and did - justify her assertion of the date and hence "knew" it. But she admits that during the exam she couldn't have justified her exam answer. In my long comment above, I implied that she did "know" the date because she subsequently could have justified her exam answer. But the survey query asks "did" she know - presumably meaning during the exam. Having reflected a bit more, I'd have to argue that the survey respondents and I were wrong. Consistency with my own definition requires that I judge her not to have "known" the date during the exam.

OTOH, Ben does "know" that the bridge is closed even as he's driving to it - driving is not asserting. If asked by a passenger why he missed the turn for the by-pass, he might slap his forehead and assert "You're right, the bridge is closed" followed by a justifying " A trustworthy friend informed me of that last night". I don't see aggregating in whatever form as necessary or desirable - Ben "knows" the bridge is closed whether he makes the intended turn or not.

Michael Schmitz said...

Charles, it's not about "I know, but I forgot" - that is exactly what the passage you are quoting excludes. It's about "I know, but I can't recall / think of it at the moment", which is perfectly fine, just like "I know, but I don't know why / don't know the reasons" is. Nor is it about a belief box. On my view, it's all about consciousness.

Charles T. Wolverton said...

Michael -

As you suggest, any vocabulary is OK as long as it is agreed up front how each word is to be used. So, I'll interpret your vocabulary using mine and you can correct any misunderstanding.

Your key concepts seem to be "acquaintance with a fact" and "capacity to retrieve that information". Using my preferred vocabulary, I interpret those as suggesting something like "having dispositions to assert a proposition P (AKA, fact) in each of a set of contexts" - what I take Eric to mean by having a "belief". (Grouping context-dependent dispositions and calling the result a "belief" is what I mean by "aggregating".) You also use the phrase "pervasive behavioral dispositions associated with belief" which also suggests aggregating.

I don't know how to interpret "attended", which I infer relates to your focus on "consciousness". Phrases like "retrieving information" and "having access to" suggest to me an architecture in which one entity contains something that another entity ("consciousness" ?) extracts or reads (or attends?) - the same architecture that "belief box" suggests to me. The architecture I envision is more like a web comprising neural configurations that implement behavioral dispositions and are excited (by external sensory stimuli or internal processes) rather than accessed.

You use the phrase "even if they lack the conviction", which suggests to me an inability to justify asserting P in one or more of the underlying contexts.

Finally, I understood the quote as suggesting that one "knows" P as long as there is at least one context in which the subject can justify asserting P, and "I know but forgot" was intended to capture the idea of "knowing" despite being able to justify in only a proper subset of the contexts. This seems an unsatisfactory concept of "knowing" since as I suggested in my last comment, such leniency doesn't seem consistent with the decisive role I understand justification to play in the JTB scheme.

BTW, what's "dof"?

Eric Schwitzgebel said...

Charles: I'm enough of an externalist/reliabilist about knowledge to think that we often know even when we can't justify. Maybe I know that so-and-so is an untrustworthy person, but I can't tell you how I know. I know that "acquaintance" is spelled with a c before the q; I can't tell you how I know. (Well, I can make plausible guesses: I've presumably seen it spelled that way, but those guesses are grounded in my assumption that I know rather than vice versa.)

I agree that to some extent this is just a dispute about how best to build up the terminology.

Eric Schwitzgebel said...

Michael: I've downloaded it and I'm looking forward to reading it. I think our views of belief, but this is my teaching-heavy term and I have a lot on my plate, so it might take a bit.

Michael Schmitz said...

Charles: There are some misunderstandings here, I hope I can clarify all or at least most. My idea is simply that theoretical knowledge may mean acquaintance with (or awareness of) a fact such that this acquaintance grounds a disposition to re-present that fact, that is, remember it, in at least some contexts, but possibly in fewer than are connected with belief, where this is taken to be associated with conviction and attendant more pervasive behavioral dispositions. ("attended" was a typo, sorry about that.)
There is more about this idea of acquaintance grounding a disposition in the paper I linked to yesterday.

"Retrieving information" and many other locutions suggest this picture of unconscious beliefs etc. lying around in the 'belief box' waiting for the light of consciousness to shine on them. But this is exactly the picture I want to reject. On my view that I have a belief even when I am not thinking about what I believe just means that I have thought about it in the past and am disposed to think about it again - in the relevant, convinced sort of way. This is also discussed at length in my paper.

I didn't mean that the subject would necessarily be able to justify what she knows. That would be a further condition on my view. And I am not sure that the ability to justify is required in all cases of knowledge.

I'm still not sure what you are driving at with your talk of 'agglomeration'.

"Dof" means direction of fit.

Michael Schmitz said...

Eric: Take your time! I think it would be interesting to compare notes re dispositional accounts of belief (and 'propositional attitudes' more broadly) at some point. I read your paper on belief a while ago and my sense was that we may have the right amount of common ground and differences to make a discussion productive.

Charles T. Wolverton said...

Eric and Michael -

I'm happy to give up B and any absolutist version of T, but giving up J seems a bridge too far if one wants to retain "knowing" in the formal vocabulary. And not just because that's all that's left of JTB, to which I have no allegiance. As Sellars envisions J, it's a crucial factor in avoiding the Myth of the Given. And a version of "knowing that" with no interpersonal element, past or present, raises the specter of the Given.

However, we may have different ideas about what's required for J. My admission requirements are pretty modest. For example, Rorty points out in CIS that simple assertions like the spelling of common words, one's date and place of birth, 2+2=4, etc, are typically justified implicitly by virtue of nobody's caring to dispute them (unless, of course, the one asserting happens to be the current president). And to some extent, I accept credentials as justification. If Eric cares to provide a definition of his "reliabilism" that is less exhaustive (and potentially exhausting) than the SEP one, I'll be happy to adopt it and will justify my newly acquired "knowing" by quoting him. But anything along the lines of "I can't give any reasons at all, I just know that P" seems obviously a misuse of "knowing that".

Michael -

The description in your second paragraph helps. The view you reject is precisely the one I was attributing to you. And the one you endorse is essentially the behavioral disposition view I hold. Great!

"Aggregating" is what I take Eric to be doing in his paper. Assume that the basic entity is a context-dependent behavioral disposition BD(Ci). A belief is then an aggregation of these {BD(Ci)} for some set of i's that is defined by common features of the resulting behavior. For example, one who is generous in a large number of contexts might be thought to believe "I am supposed to be generous".

Daryl said...

Vocabulary is important, but so is the scaling of context. Scale makes all the difference in the range of aggregated elements counted toward the attribution of a behavioral disposition as a single unit distributed through time. Thus I gather from the comments above that, for Michael, the attribution of a behavioral disposition is attuned to a larger context than Charles would allow for. For Michael, the time before, during, and after the exam add up to a single context, whereas for Charles they stand alone as three separate contexts. But this is why Michael can allow for knowledge even where justification eludes the one who is said to know: that’s only one small part of the overall context.

Charles T. Wolverton said...

Daryl -

I agree with your comment. But in the interim I've rethought the Kate scenario and decided that since the exam was fill-in-the-blank, the odds against getting the answer right by chance should probably be considered adequate implicit justification. But had the test been multiple choice, I'd still be disinclined to credit her with "knowing" the answer.

Daryl said...

That makes sense. A line has to be drawn somewhere, otherwise the aggregation of singular acts counted toward the attribution of behavioral dispositions would lead eventually to the inscription of a subject barely recognizable as a human being. Surely this is the essential fallacy behind, for example, the idea that the poor can "lift themselves up by the bootstraps" if they so choose. The idea isn't wrong, but here the subject is inscribed through an aggregation of singular acts beyond what is typically regarded as belonging to the contingencies of an individual human will, while the human being is reduced to a mere avatar of that subject.

Charles T. Wolverton said...

It's especially interesting to me that you use the will of the poor (or more generally, any disadvantaged group) as an example since that sort of thinking is part of the genesis of my views on behavior. I was young in pre-civil rights Texas when race prejudice was still the norm, but had the good fortune to be largely immune to it due to being reared in a progressive environment. That led me to think in terms of determinism and the influence of environment and personal history. Many years later when I started reading about phil of mind, it was a pretty natural step to "context-dependent behavioral dispositions". All I really had to do was convert life long inclinations into a new vocabulary!

I obviously agree that aggregation has to be limited somehow (and so does Eric). But once I started thinking in terms of CDBDs, trying to resurrect PAs didn't seem to buy me anything. Although that may be because my amateur interests can be much narrower than those of a professional philosopher.