The Priority of KL-Justification Over Justification

Taken from Fantl/McGrath, of course: KL-justification is knowledge-level justification. Background: in Chapter 6, they use a claim about KL-justification to argue for the importance of knowledge.

The claim depends on an equivalence thesis: one is justified in believing iff one has KL-justification.

So, first, those to consign to the sidelines: those who think an adequate understanding of knowledge does not advert to justification. The bottom line here is an attempt to account for the value or significance or importance of knowledge, and the account I want to question can’t get off the ground without the equivalence thesis. Once rid of these scoffers, here’s what I think one should say:

The justification involved in the nature of knowledge is a functional concept. To wit, that which when ungettiered and combined with true belief, yields knowledge. The idea of justification for belief is an analogical notion, related to justification for action. In each case, we specify a goal, and then justification is that which is related in an appropriate way to the goal. No matter what counts as appropriate here: it will survive the analogy.

Based on these data, we can say this. The primitive notion is justification. The derived notion is a functional one, claiming that KL-justification is, precisely, justification.

F/M think the explanation goes the other way, but the only argument I see (in chapter 6, they say they argued this in chapter 5) is this remark in chapter five: it is plausible to think that KL-justification is explanatorily prior.

This matters, because if justification is explanatorily prior, we don’t have an explanation of the importance of knowledge. If, however, KL-justification is prior, we do. In my dissatisfying role as iconoclast regarding explanations of the value of knowledge, I can’t see why we should favor a plausibility claim over the above account.


The Priority of KL-Justification Over Justification — 6 Comments

  1. I suppose that’s an open question, though I don’t think any of the standard approaches to the notion appeal to knowledge in their characterization of what being ungettiered amounts to. That’s compatible with taking it to be a condition of adequacy on an account that if you add it together with JTB, you get knowledge, a condition I favor at least on most days.

    Also, I didn’t characterize justification functionally–I only characterized KL-justification that way.

  2. Jon, the argument is that facts about being justified to be in a doxastic state are grounded by facts about the justification one has for the objects of that state. So, e.g., when you are justified to believe p, this holds in virtue of facts about your justification for p. We argue earlier in the book, on pain of infallibilism about being justified to believe, that no particular degree of justification for p short of probability 1 by itself is sufficient for being justified to believe p. Thus, the facts that ground one’s justification to believe p shouldn’t be taken to be facts of the form “one has degree d of justification for p,” but rather facts of the form “one’s degree of justification for p reaches such and such threshold.” The threshold in question, we argue earlier in the book, is knowledge-level justification.

    So there are two premises used in the argument: (1) the “propositional justification first” claim and (2) the claim that the relevant fact about propositional justification is the fact that one’s degree of justification for p reaches the knowledge-level threshold.

    The “propositional justification first” claim holds that justification for p (i.e., support/evidence for p) is prior not only to doxastic justification (justifiably believing p) but to what is sometimes called personal justification, justification to have a particular doxastic attitude toward p.

  3. Matt, that argument is one about the priority of propositional justification to doxastic, which of course is a position dear to my heart. Since “being justified in believing p” doesn’t require believing p, the equivalence thesis is about propositional justification–no? (Personal justification is but a lambda-conversion from one of the other two, so I don’t think we need to treat it separately.) And in any case, propositional justification comes both as a degree notion and a threshold notion, so I don’t see how the priority of it over doxastic is doing the work here.

    I agree with the threshold point, though I think what you showed in the book is that if you appeal to KL-justification as the threshold, you have an answer to the threshold problem, whereas if you don’t, it isn’t clear what the threshold will be. But a defender of the priority of justification over KL-justification can mimic your theory here by going straight to action itself: a degree of justification is high enough when it can be put to work. And then point out that KL-justification is fundamentally a functional role concept explicable in terms of the threshold idea just explained.

    Your theory wouldn’t be affected at all by this move, right? Except that you’d lose the second of your two explanations of the importance of knowledge, but who cares about that? 🙂

  4. Jon, I was thinking there are three notions: having justification for p (roughly having support/evidence for p), being justified to have a certain doxastic attitude in p, whether you have it or not, and justifiably having a certain doxastic attitude. I was labeling the first “propositional justification,” the second “personal justification’ and the third “doxastic justification.” Maybe the first label would be a non-standard use of the term. Putting aside the label question, I guess I was thinking that when you are justified to believe p, this holds in virtue of facts about your support/evidence for p (and in particular, that it reaches the relevant threshold). I think Fantl and I meant to be endorsing evidentialism with an added twist about what counts as strong enough evidence.

    I agree with your other points.

  5. Right, Matt, I don’t think it’s a matter of terminology here. But I think the argument you aim at is that KL-justification is explanatorily prior to justification, and each of these notions can come in all three of the varieties you mention. And the priority claim that is the most difficult to challenge is the one that says that propositional justification is prior to them all. It is certainly prior to personal and doxastic justification, and so the target of your argument, it seems to me, needs to be the claim that there is some notion of of KL-justification that is prior to propositional justification. Presumably, that will have to be a propositional notion of KL-justification. If it is some other notion, then the Equivalence Scheme will be false, and any claim about the priority of propositional over doxastic will cut the wrong direction. And once we have propositional notions for both justification and KL-justification, an appeal to the priority of propositional to doxastic won’t be relevant.

    In the prior chapter where you used the priority of propositional to doxastic, I thought the argument was good in context. The context is whether pragmatics about belief is prior to pragmatics about justification. Then it is exactly right to point out that pragmatics of belief could only be prior to doxastic justification, and since propositional justification is prior to doxastic and logically independent of belief, there is no hope that pragmatics about belief will be prior to propositional justification.

Leave a Reply

Your email address will not be published. Required fields are marked *