A. Virtue Epistemology > Personal Character and Epistemic Reliability

This post is a response to Kelly’s 9/20 post (which in turn was a response to my 6/20 post). I’ve taken the liberty to create a new thread, since this topic may be of interest to some folks who haven’t kept up with the discussion on the value problem.

Thanks for your comments Kelly. You’re right to think that my paper (“Character, Reliability, and Virtue Epistemology,” which is available at is not aimed at discrediting reliabilism or virtue reliabilism (VR). It is, however, aimed at identifying (a) a problem with present varieties of VR, and (b) certain theoretical challenges that arise from this problem. So I take it that the implications of the argument are not insignificant.

Regarding (a). Early on in the paper, I show that defenders of VR (e.g. Goldman and Greco) *are* in fact resistant to thinking of intellectual character virtues as reliabilist virtues or as reliabilist “knowledge-makers.” While I agree that this is not an objection to VR per se, it is an objection to extant varieties of VR. As to *why* a reliabilist might want to exclude the character virtues, your guess is as good as mine. But here’s one additional conjecture. I think some are attracted to reliabilism (and externalism generally) based in part on a commitment to epistemological naturalism. And I can see why a naturalist might prefer a “mechanism-model” of cognitive subjects to an “agent-model.” But once reliabilists expand their focus to include the character virtues, their conception of cognitive subjects will begin moving in the direction of an agent-model. I’m not saying that naturalism requires a preference for the mechanism-model; but I can see why some naturalists might be attracted to it -– and to that extent be reluctant to bring character virtues into the fold of knowledge-makers.

At any rate, suppose you grant that given their neglect of intellectual character virtues, present versions of VR are incomplete (not incomplete “in theory” or as a matter of principle, but materially so). I think one could still reasonably object: “So what? So virtue reliabilists ought to include the intellectual character virtues in their repertoire of knowledge-makers. How is that supposed to be a problem for VR?” On the one hand, my answer is that it’s not. The fact that reliabilists need to expand their focus doesn’t indicate a problem with reliabilism as such (again, it’s a problem only for the relevant extant versions of reliabilism).

On the other hand, the latter part of the paper -- which addresses (b) above -- is aimed at showing that reliabilists’ inclusion of intellectual character virtues may not be such a smooth endeavor. That is, I attempt to show how doing so gives rise to certain additional theoretical questions and challenges that any fully adequate version of VR (or reliabilism in general) must address. So the “rub” for reliabilism as such is here; it’s not with the broader, “inclusion” point. And even here, the challenge is a reasonably friendly one. It's not meant to motivate the rejection of reliabilism.
September 21, 2006 | Registered CommenterJason Baehr
Excellent. I think it's really useful for epistemologists to be pointing out potential tensions in theories and to explain how important epistemological concerns are not always captured by the attempt to analyze knowledge in its nuts-and-bolts form. My previous post implies that it is sometimes difficult for me to orient my mind this way: Always on the lookout for the 'devastating' objection, one loses sight of other important problems and promising directions of inquiry.

Thanks for the reply.
September 22, 2006 | Registered CommenterKelly Becker
I see a connection here with the John Greco's paper. "Holding defeat to the Fire." There he writes about what he takes to be "an underappreciated problem for reliabilism. The problem concerns how a reliabilist approach in epistemology ought to understand defeating evidence. In more colloquial language: How should reliabilism understand the notion of counter-evidence, or evidence against one’s beliefs? The problem arises because we have certain pre-theoretical intuitions about counter-evidence, and a reliabilist approach in epistemology seems inconsistent with these intuitions."

Let's call this the "Neglected Problem." The paper illuminates how reliabilists typically (and perhaps self-inconsistently) 'go internalist' whenever faced with problems of this sort (presented by Bonjour). Now let me quote in greater length the key conclusion that he draws from his examination of it. In contrast to 'going internalist' in the way most extant forms of reliabilism do,

"The correct strategy is to add a subjective justification condition to reliabilist conditions for epistemic justification and knowledge. According to this strategy, S does not know that p in BonJour’s examples because S is not subjectively justified in believing that p. This strategy can be successful, however, only if it respects three constraints that arise out of the considerations above:

1. the subjective justification condition cannot be ad hoc.

2. the subjective justification condition cannot be a function of necessary relations among propositional contents.

3. the resulting account must discriminate cases of knowing and not knowing in an intuitively acceptable way.

The most promising strategy for satisfying these constraints, I will argue, is to adopt a virtue-theoretic approach to justification and knowledge. According to such an account, intellectual virtues are understood to be reliable cognitive powers or abilities, such as good vision and sound reasoning, and knowledge is understood to be true belief arising out of intellectual virtues. Such an account remains reliabilist: S’s belief is objectively justified only if that belief is reliably formed. But a virtue approach allows for an account of subjective justification as well: S’s belief is subjectively justified only if it is produced by dispositions that are properly motivated, i.e. dispositions that S displays when S is motivated to believe what is true.[5] We may then understand epistemic justification, or the sort of justification required for knowledge, as involving both objective justification and subjective justification so define. In case of knowledge, S’s belief is both objectively reliable and subjectively appropriate. In fact, we may understand these two conditions as organically related: In case of knowledge, S’s belief is objectively reliable because it is subjectively appropriate." [end Greco quote]

So I wonder how far and in what way each of you might draw from or disagree with something in Greco's stance. His language is that of "subjective" and "objective" justification conditions, both cashed out in virtue-theoretic terms, or if this isn't a helpful language? Clearly, at least, its another expression of "epistemic compatibilism" and something I agree with (as Bernecker defines it, the compatibilist claims to satisfactorily retain some of the driving motivations of internalism in one's externalist theory). That subjective and objective conditions are both cashed out 'areteically,' and the one-way dependence of reliability upon responsibility/motivation clarified this way, is a prime reason why Greco, in his Stanford Ency. article on VE, describes VE as "mixed theory." I can't see why something like this shouldn't bring the two of you substantially closer together, but perhaps I'm overestimating the appeal or the plausibility of Greco's stance. It seems paradigmatically like a "moderate VE" stance to me, providing an answer to the theoretical challenge of explaining how the character virtues remain epistemically central even after epistemology has undergone its 'externalist turn.'
September 22, 2006 | Registered CommenterGuy Axtell
Well, Guy, I'm SURE you won't like my reply, given that I don't grant the premise. But so long as I'm part of this conversation, I might as well say what I think.

I don't grant the need for subjective justification, at least not as a necessary condition for knowing. I realize that I'm in the minority here, but my reaction to cases like Bonjour's clairvoyant has always been, "that doesn't show anything." Even IF the clairoyant has very strong reasons for thinking there's no such thing as clairvoyance, I think she can have clairvoyant knowledge. There are a number of possibilities.

First, if she REALLY believes that her clairvoyance-based beliefs are irrational, she won't really (realistically) keep them.

Second, if she only considers that her clairvoyance-based beliefs are irrational but keeps them anyway, then I'm willing to say that she knows.

But, third, what if she really believes that her clairvoyance-based beliefs are irrational AND continues to keep them? Even here I'm willing to bite the reliabilist bullet and say she at least CAN know, but I also get to ask: What are we imagining here? Is it remotely pschologically plausible? (A plea for a sustained discussion of the nature of belief.)

My basic view? Reliabilists get themselves into trouble all the time by trying to satisfy people who think (1) justification is necessary for knowledge because (2) justification rings 'internalist' in almost everyone's ears.
September 25, 2006 | Registered CommenterKelly Becker
Kelly, I might not disagree as far as you think. It is a little surprising that John Greco opts for a general subjective justification condition. I can understand why you resist his lesson that "The correct strategy is to add a subjective justification condition to reliabilist conditions for epistemic justification and knowledge."

It does seem to make a big concession to internalism if its taken as a "general" condition, and that is how Greco states it. I suppose I'm not inclined to take it as a general condition, but rather to say something like that "reflective" knowledge requires reflective justification, but animal or brute knowledge, which would seem to be what the 'unenlightened' clairvoyant or chicken-sexer has, does not. I'm inclined to distinguish 'kinds' of knowledge--though that definately has its problems as well, and to say that low-spectrum knowledge either doesn't have such a requirement of subjective justification, or that its automatically filled to the extent it needs to be filled, in cases of low-range 'brute' or animal knowledge. That may make me sound more like Sosa than like Greco on this, but I don't actually read Greco as demanding that what he calls "subjective justification" makes a similar demand with low-range as with high-range examples. I just think he means its not an issue--not a bar to knowing--insofar as the kind of knowledge in question is of the 'brute' sort.

Jason is critical, I take it, of 'mixed' theories which have the form of the virtues filling dual "subjective and objective" justification requirements (Greco), or again, "responsibility" and "reliability" requirements (Zagzebski). While I know I've used both claims in my papers at times, I think of my DATA account as not really committed to either of these other two ways of framing mixed externalism, but as offering a third way. It has an "areteic" (or 'out of virtue) condition, but one where virtue is thinly defined, so the requirement can be quite different at one end of the spectrum or the other, and an independent "tucheic" or anti-luck condition that can address what the areteic condition misses. But it doesn't impose a "general" subjective justification condition, as if virtue-talk should be all about motives, and anti-luck-talk all about veritic luck.

But that's still all about analysis of knowledge, and I know Jason didn't intend his post to get dragged back into it. So in regards to the connection between personal responsibility and epistemic reliability, which I quoted John's passage becaue it directly addresses, John says "that we may understand these two conditions as organically related: In case of knowledge, S’s belief is objectively reliable because it is subjectively appropriate." That's an interesting point (though perhaps pertinent only to higher-end knowledge). But I wonder if John doesn't concede more than he intends with this point. It seems to me that if John is right about this, he'd be leaving the door open for the neo-Aristotelians to argue that this shows that subjective appropriateness "defines" what a virtue is, while showing that "objective reliability" doesn't define them, but rather only provides the "identifying marks" that allow us to attribute it to individuals. Or am I reading too much into Greco's claim that its the subjective appropriateness of the agent's motivations/desires that leads that (self-reflective) agent to have objectively reliable beliefs?
October 9, 2006 | Registered CommenterGuy Axtell
I'm pleased to have read your response, Guy. I think the 'division of "kinds of knowledge"' strategy is level-headed and allows us to avoid all kinds of talking past each other.

This leaves open questions about what exactly JG is up to. There are a couple of issues, it seems to me. One is whether subjective justification (SJ) is necessary for knowledge. Given what JG says, the answer may still be 'no'. It may simply be what's missing when one doesn't pay heed to defeaters (and so explains why one lacks knowledge in just those cases), and he says only that virtue reliabilism (VR) "allows for" SJ. If this is right, then it seems like a nice approach, except that, again, I'm not one of those reliabilists who tend to go internalist. I think the clairvoyant knows even if she thinks there's something screwy about clairvoyance. (And now we're back to the "brute" vs. "higher" knowledge issue.)

The second issue is the relation between dispositions and motivation. This is related to the first, because it's not clear to me that motivation is necessary. JG says, "S’s belief is subjectively justified only if it is produced by dispositions that are properly motivated, i.e. dispositions that S displays when S is motivated to believe what is true." [5] But this appears to be compatible with one's having those dispositions and forming beliefs from them, but not, in any given case, actually being motivated (from them, by them?). So maybe the idea is that there is a motivational component to forming the dispositions, but that this component is not necessary in every specific case of belief formation. I'm not sure, and so I'm not sure what to think of the account. But it sounds very interesting.
October 16, 2006 | Registered CommenterKelly Becker
Heather mentioned that Jason thinks “that virtues are excellences that admit of different kinds: they can be excellent internally, or excellent because they lead to some valuable end.” This seems like an interesting proposal, if its what Jason intends, and one which ties into the thin concept proposal in the following way: a plurality at the level of the kinds of virtues themselves would undermine our common assumption that either the reliabilist or the neo-Aristotelian perspective on how to define intellectual virtues must be chosen; it would thus potentially undercut the desire to ‘thicken’ our conceptions of the virtues in the way that it seems reasonable to say we do thicken them when we insist that the one way of defining them or the other must be correct. Here then we would have to concede that as Heather asserts, debates between reliabilists and responsiblists about the intellectual virtues have been intractable “because we have thickened the thin concept of virtue in equally legitimate but incompatible ways.”

While I haven’t thought enough about it, I want to add two brief lines of thought that might support such a pluralism at the level of the intellectual virtues themselves (not alternative to, but in addition and complementary to epistemic value pluralism, as discussed in other threads on this blog).

1. A negative argument. The critics of epistemic compatibilism tend to be die-hard proponents of the idea that it’s a winner-take-all game. Either ‘pure’ internalism or ‘pure’ externalism must prevail, their approach to defining the virtues and explaining their value winning out. Let me give examples, quoting from Dancy and Driver and respective advocates of the each view.

Jonathan Dancy: “Non-consequentialists also are unwilling to admit that the consequentialists are right about anything, because they feel that consequentialism is like a cancer—once one has let it in at all it will grow until it has taken over completely. The crucial question is whether the two camps are right at least about this, that no compromise is intellectually acceptable. And I think that they are.”; again he writes, “Consequentialists are wise to seek to give a unified account of all the virtues, because otherwise they will find themselves saying that of the virtues, some are virtues for one sort of reason and others are virtues for another. This position is theoretically unstable, and will always be vulnerable to one which manages to give the same account of why this or that feature is a virtue throughout. Similarly, virtue theorists are right to resist the irruption of a second form of explanation of the status of a character trait as a virtue. Their standard form, which asks how the virtues together contribute to a good epistemic life, is perfectly capable already of capturing the nature and role of the consequence-related virtues. That they are consequence-related does nothing to show that we should accept a consequentialist understanding of them” (82).

Julia Driver expresses similar ‘strategy,’ but of course in support of a stance she calls objective consequentialism allied with “the externalist extreme” of a thoroughgoing reliabilism. This view is similar to that which Firth’s termed “epistemic rule-utilitarianism,” a description Sosa himself picked up and used in his ‘Raft and Pyramid” paper give his initial description of a virtue epistemology. Hence I critcized Dancy and Driver (in ‘Luck in Light of the Virtues”) for having in common the fervent view that epistemic compatibilism (or by extension any sort of via media) is unstable. Like good fundamentalists, they agree in how they hold their beliefs—that if one is right, the other is wrong--but disagree just over who it is that is right.

2. A positive argument. I’ll make this brief, but an example of someone who does hold the pluralism thesis is Vrinda Dalmiya. In “Why Should a Knower Care?” she holds that “caring” should be an intellectual virtue on both responsibilist AND reliabilist scores. Hence instead of contrasting them, she uses the idea that responsibility gives rise to reliability to argue “that caring can be added to the …list of “sight, hearing, memory, introspection, deduction, and induction” (i.e., Greco’s list). She says it’s a ‘responsibilist’ virtue in the sense of deriving such status from motivations, not consequences primarily, but that it does promote reliability means that should be a reliabilist virtue as well. “Caring as an adjective of the knower” is relevant to knowing “because it signals an effort necessary for moth knowledge and things and of selves.”

So I make these comments because they seem to support the direction Jason wants to lead us, from talk about different kinds of VE, to talk about different kinds of virtues.
October 18, 2006 | Registered CommenterGuy Axtell
I’d like to raise a question mainly for Guy (though others should feel free to chime in) concerning his appeal to the notion of thick/thin concepts in connection with a virtue-based analysis of knowledge.

It seems to me that there are at least two ways of understanding this move. On the first, the idea is that something is an intellectual virtue just in case it is a *reliable* personal quality, but that both faculties and intellectual character traits can satisfy this condition – in which case the notion of an intellectual virtue can be 'thickened' in terms of the notion of a cognitive faculty or the notion of an intellectual character trait. Thus where knowledge is understood (roughly) as true belief produced by an intellectual virtue, both reliable faculties and reliable character traits can be viewed as contributing to knowledge. This is the move that I suggest reliabilists should avail themselves of in my “Character, Reliability, and Virtue Epistemology.”

In connection with a second possibility, Guy mentions my claim in “On the Reliability of Moral and Intellectual Virtues” that there are two legitimate conceptions of an intellectual virtue – a ‘competence conception’ and a ‘personal worth conception’ – and that the conditions for either can be satisfied by an intellectual character trait. The idea here is that a virtue is simply an excellence and that there is more than one way in which a character trait can be excellent: it can be such because it’s a reliable way of achieving a certain valuable result (i.e. reliability) or because it contributes to one’s personal (intellectual) worth.

Now how is this relevant to a virtue-based account of knowledge? I take it that Guy’s suggestion is something like this. Instead of applying the thick/thin move to the question of which traits might satisfy the conditions of a reliabilist conception of intellectual virtue, we should apply it instead to the question of what *counts* as an intellectual virtue in the first place. Thus we can maintain that knowledge is (roughly) true belief produced by intellectual virtue and that faculties or character traits can qualify as intellectual virtues – not because an intellectual virtue is a reliable trait and both faculties and character traits can be reliable, but rather because there is more than one way of understanding what it is for a personal quality to be a virtue (i.e. it can be a virtue because it is a competence or because it contributes to personal worth).

So these represent two ways in which the idea of thick/thin concepts might be appealed to in support of a virtue-based account of knowledge aimed at incorporating both faculties and character traits. I take it, Guy, that you prefer something like the second option. I find this option puzzling – though for reasons I have hard time articulating (perhaps others can be of assistance here). The view apparently commits one to saying that in certain cases, a true belief will count as knowledge because it is produced by a reliable personal quality, while in other cases it will be knowledge because it is produced by a personal excellence (i.e. a quality that contributes to personal worth). But the idea that what makes a true belief knowledge might vary in this way strikes me as counterintuitive and problematic (that’s about the most I can say at this point; it admittedly falls short of an argument).

Note that with the first option noted above, the notion of reliability is doing the relevant epistemic work ‘across the board.’ So while (in some sense) one true belief might count as knowledge because it was produced by a certain faculty and another because it was produced by a certain character trait – two very different things indeed – this is because both are *reliable*. Accordingly, this option offers a unified, univocal answer to the question of what makes a true belief knowledge (viz. it’s produced by a reliable personal trait). The second option, as far as I can tell, offers no such answer (for again, it leaves open what the very notion of an intellectual virtue amounts to). And this is what I find puzzling/problematic.

Finally, note that if you were to go with the first option, you’d really just be offering (what I’ve referred to in the past as) as ‘character-enriched’ version of reliabilism, that is, a version of reliabilism that allows intellectual character traits as well as faculties to count as knowledge-makers. The view would be a version of reliabilism because reliability is what’s ultimately doing the epistemic work.

I hope I haven’t confused you too much Guy (or others). Let me know what you think, or how I can clarify my remarks.

October 23, 2006 | Registered CommenterJason Baehr
Thanks for the post. You distinguish usefully between two senses of epistemic compatibilism, both of which could ground responsibilism as a distinctive approach. But I gather there is also a dilemma here for me and for others: Grabbing the second option--two distinct accounts of what *counts* as an intellectual virtue--seems highly problematic (perhaps because it suggests that there is no unitary account to be given of what conditions suffice either for virtue or for knowledge); while grabbing the first option has better initial plausibility because it provides a unitary account, but leaves us with only a "character-enriched version of reliabilism" (instead, presumably, of responsibilism as a distinctive position).

I don't mind saying my position in the previous post probably does equivocate between the two options you clarify. On consideration, I'd probably opt for grabbing the first horn, or a 'between the horns' response. But I wonder first if the dilemma doesn't mark a difference between what I've called "responsibilist externalism" and neo-Aristotelian VE? I'm not sure, but it strikes me that our neo-Aristotelian contingent at JB would be more likely to embrace the second horn, or to say something like a third option, that all virtues are defined by their distinctive motivations, and that reliability is only an indicator but not a defining point about virtues.

Anyway, that response, which seems tied to "pure virtue theory" whether in ethics (Slote) or epistemology (Zagzebski) doesn't appeal greatly to me, personally. So I'm back to grabbing the first horm of the dilemma. I've called my view "responsibilist externalism," and suggested (in 'Two For the Show') that it *might* be a really non-reliabilist form of externalism; but that if reliabilists 'get over' unnecessarily thickening their accounts of virtue and show that they want to and can satisfy the "comfortable home demand" that I characterize responsibilism by (a demand for a comfortable home for research programs into the reflective virtues), then there may be no NEED to so distinsuih itself. This posturing may seem like fence-sitting to you, though?

Maybe it is, but say that we accept Greco's point that "Such an account remains reliabilist: S’s belief is objectively justified only if that belief is reliably formed. But a virtue approach allows for an account of subjective justification as well: S’s belief is subjectively justified only if it is produced by dispositions that are properly motivated, i.e. dispositions that S displays when S is motivated to believe what is true."[5] I'm not sure what is so wrong with this. I think I want to hear more about what is problematic with this general view (the issue of making subjective justification a 'general' necessary condition aside, since I've argued at length that the thin-concept proposal isn't brudened with that). Then I'll know better whether my response needs to be one of grabbing the first horn, or (what it alot more work) going 'between the horns.'

One last query here. Your two options are addressed as pertaining only to the "nature of knowledge" problem. The "personal worth" conception does have a place, I think, but perhaps that place is in relation to the "value of knowledge" question. If we accept the need to balance these two questions (following Kvanvig thus far), might it be that both of your options are valid, but that need to sort them as addressing different questions? Then my stance would be something like this: The personal worth conception is better than instrumentalism at addressing the value question, while externalism (whether we call the position responsibilist externalism or 'character-enriched reliabilism') is better than either internalism or *austere* reliabilism at answering the nature of knowledge question.
October 24, 2006 | Registered CommenterGuy Axtell
Those are some interesting thoughts Guy. Here are a few replies.

I guess I’m not very interested in fussing about how to apply the reliabilist/responsibilist terminology. “Responsibilist externalism” seems to fit your view well enough. I’m inclined to use the “reliabilism” label to describe your view only because, as I understand it, the concept of reliability is what’s ultimately doing the epistemic work (i.e. it explains *why* the traits in question are intellectual virtues). But I can see why, given the association between externalism and reliabilism and that between internalism and the reflective virtues, this might not strike you as entirely satisfactory.

I agree with you that responsibilists like Zagzebski opt for something like a third option relative to a virtue-based analysis of knowledge. (Strictly speaking, I’m not sure it’s a third option given that it, unlike the other two views, doesn’t grant cognitive faculties the status of intellectual virtue.) I also agree that this option is problematic. The main reason is that once one defines knowledge as true belief produced by intellectual virtue and defines intellectual virtues as excellences of intellectual *character,* one makes knowledge too difficult to come by. (Linda is sensitive to this worry, and her account of knowledge doesn’t actually require virtuous character [just virtuous motives and actions]; but at the end of the day I think it’s still objectionably demanding.)

The option you mention about relying on a “personal worth” conception of an intellectual virtue in the context of talking about the value of knowledge and a “competence conception” in the context of talking about the nature of knowledge is interesting. But it’s not clear to me that it will work, given how people tend to understand the relation between the two issues. Parties to the first debate typically think of the “added value of knowledge” as consisting in a *necessary and defining* feature of knowledge. If this is right, then the problem is obvious: if (something like) an exercise of a “personal excellence” is not a necessary or defining feature of knowledge (which, presumably, it would not be if reliability were the governing notion in the analysis of knowledge), then the value of such traits cannot explain the “added value of knowledge.” (As a side note, I can’t help but register my view that this and related ways of thinking about the value problem are deeply problematic; anyone interested can read my “Is There a Value Problem?”)

Thanks again for the interesting discussion Guy.
October 24, 2006 | Registered CommenterJason Baehr
Jason, So I won't fuss further here about the reliabilist/responsibilist terminology. I'm not uncomfotable with what you call a "competence" conception of knowledge, so long as competence, too, is understood thinly-enough that the kinds of competencies pertinent to reflective knowledge (and especially the standards of agent-performance there) can be more demanding than they are at the low end of the knowledge spectrum.

One reason why I do agree we need a fairly unified conception of intellectual virtue is because I hold that knowledge is a spectrum--and my description of it as a family-relations concept implies there's not very much univocality in the items we pick out as instances of knowledge. So if I ALSO say there's little unity to what counts as a virtue, but manifesting virtues is part of what's needed for knowledge, then my account would just look pretty ad hoc. So while I might criticize specifics of the reliabilist's account of competence, and the idea of epistemic value monism it is often wedded with, I wouldn't want to abandon the account where you say 'reliability does the explaining'--not at least until it is clear that there is any really viable alternative to it anyway. I agree with you by and large over the neo-Aristotelian and pure-virtue theories making "knowledge too difficult to come by." But this also makes me think there is not clear alternative to it. An account of this sort seems Aristotelian enough. I don't feel a need to follow Zagzebski's approach of characterizing the intellectual virtues according to the model she imports from Aristotle's ethics.

So, basically, I think responsibilists can live with a competency conception of knowledge, and work within that. Within such a conception, Greco is able to say,

“We should...amend the basic idea of VE, so that the cognitive virtues which are relevant for knowledge have their bases in the epistemic responsibility of the knower” (1993a, pp. 432).

He is able to say that "S’s belief is objectively reliable because it is subjectively appropriate." I want to propose that responsibilists work on developing the IMPLICATIONS of these points for demonstrating the epistemic centrality of the reflective or character virtues, rather than trying to insist upon a restrictive, "personal worth" conception of the virtues. That latter approach may work in ethics, but not in epistemology. Somehow it always leads us to miss out on what is correct and advantageous in the externalist turn in epistemology. Whether its also a confusion of the nature of knowledge question with the value of knowledge question, I can't say, but I'm tempted to suggest that it is.

You understand the connection of these issues far better than I do, but I am dubious of this claim that you say is commonplace today, that whatever the “added value of knowledge” is, it must in turn be taken as as consisting in a *necessary and defining* feature of knowledge. That sounds like a demand to play philosophical Jumanji to me.

But here are two questions to end with. What is the relationship between a competence theory of knowledge, and a "credit theory"? And what is the connection between holding a credit thory of the value of knowledge, and holding a credit theory of the nature of knowledge?

These questions lead me to some things I want to explore in Wayne's paper from the recent Stirling conference, bearing on his similarities and differences from Linda--but I'll perhaps make that a different post. At any rate, I take it that these issues are complex. Kelly, in particular, might agree with a competence conception of knowledge, yet in previous posts he was critical of credit theories of knowledge, thinking that they're inapplicable to instances of low-end or 'brute' knowledge. These issues seem worth trying to sort out.
October 26, 2006 | Registered CommenterGuy Axtell