Propositional and Attitudinal Justification

Here’s a somewhat puzzling phenomenon. The justification of beliefs involves two different notions. The believing itself might be justified, which we might call attitudinal justification. Sometimes epistemologists, including me, have called it ‘doxastic justification’, but I now see that this term is a bit myopic, since it blocks us from seeing the similarity between the justification of beliefs, hopes, fears, etc. All of these are capable of attitudinal justification, though only beliefs are capable of doxastic justification.

Back to the puzzling (or strange or unusual) fact about doxastic justification. We can explain how the believing itself is justified by reference to propositional justification: the believing is justified, in part, because the content of the belief is justified. That is, there is propositional justification for the content of belief, so that in the normal case, if the belief itself were absent, such an absence would be irrational (unjustified) because the proposition in question is justified for that person.

Note that this phenomenon does not generalize to other kinds of attitudinal justification. Hopes are not justified because of any justification for their content, and neither are wishes, desires, etc. Perhaps assumptions and presuppositions are, though, leading one to think that the propositional/attitudinal distinction is at home in a way among purely cognitive states that it is not at home among affective states. If that point is right, then one should expect the explanatory story of justification to be different for cognitive states than for affective ones.


Propositional and Attitudinal Justification — 24 Comments

  1. Jon-

    But it looks like whether some affective states are justified (or rather appropriate) does depend on propositional content.

    1. Bob regrets that he hurt Lou.

    Inappropriate. That wasn’t Lou you hurt, it was Sam.

    2. Sue fears the bear behind the sofa.

    Inappropriate. There is no bear there. It’s a dog.

    3. Tom hopes to win the 5k race today.

    Inappropriate. Tom won the 5k yesterday. It’s the 6k that is run today.

  2. I think an example would help me better to see the puzzle, Jon. My first impulse is to say that justification only goes to attitudes, never propositions. If Sue believes or hopes or fears that there is a bear behind the sofa then we can ask (a) is she justified to do so (believe, hope or fear that)? and (b) is there or is there not a bear behind the sofa? But we can’t ask whether or not the proposition is or is not justified as such because “There is a bear behind the sofa” can simply be true or false. (That’s what I thought propositions were for.) It may be the justifiable or unjustifiable content of a fear, hope or belief. But the a proposition is only anything like “justified” outside a propositional attitude in its actual fact-of-the-matter truth or falsity, right?

  3. Thomas, it’s not that propositions are justified in themselves; it’s rather that certain propositions are justified for a person, whether or not they take any attitude toward them. For example, if testimony transmits justification, and I tell you that I’m under 6 feet tall, then, so long as you have no defeating information, the claim that I’m under 6 feet tall is justified for you. You may believe that claim, and you may not. If you believe it, part of the explanation for why your belief is justified is that its content is justified for you. If you believe that I’m not under 6 feet tall, then part of the explanation for why your belief is unjustified is that the claim that I am under 6 feet tall is justified for you.

  4. Thanks, Jon. But isn’t that what being justified in believing something always is? If a proposition p is justified for Jones, then Jones is justified in believing that p. Or, to use your example, if I believe that you are not under 6 feet tall, then THE WHOLE of the explanation for why my belief is unjustified is that the claim that you are under 6 feet tall is justified for me. It’s just another way of saying the same thing.

    I guess now I need an example of some other way of justifying beliefs than giving people reasons to believe in their content.

  5. What do we mean by attitudinal justification for other kinds of attitudes? One answer might be sort of moral—it’s unjustified for me to hope that you slip on the banana peel I left on the ground, because I shouldn’t be hoping for something that will hurt you simply to satisfy my diseased sense of the amusing. Then beliefs won’t have this sort of justification.

    Or I can also imagine a sense in which the justification of hopes and fears does depend on the justificatory status of their content. Suppose I were to hope that the Pirates win the World Series next year. You could say “That’s a ridiculous thing to hope; the Pirates have no chance.” This is perhaps clearer with fears; to argue that fears that Saddam would give nuclear weapons to al-Qaeda were unjustified was to argue that the propositional content of the fear had little support.

    (I just Googled “unjustified fear” and it seemed as though about two-thirds of the top hits were about things you shouldn’t be afraid of because they were ostensibly unsupported by the facts, and one-third things you shouldn’t be afraid of because they were ostensibly not so bad even if they did happen.)

    Wishes don’t seem subject to that criticism; the Pirates’ poor prospects of success doesn’t stop me from wishing they would win the World Series. Maybe we can say that there’s a cognitive dimension and an affective (conative?) dimension to attitudinal justification. Beliefs are purely cognitive, so attitudinal justification for belief just is having your belief based on propositional justifier in the right way. Wishes are purely affective, so justification of a wish is just a question of the worthiness of the affective state. Hopes and fears are both cognitive and affective, so both kinds of justification come into play (though the required basing won’t be exactly the same as for beliefs).

  6. Suppose some shady underworld character thinks I told him I know the World Series is fixed this year (and that the Pirate’s will win) and suppose he has bet a small fortune on this information. It’s all a terrible misunderstanding, of course. Nonetheless, since the bet can be unmade, I’d better hope the Pirates win. And, even given facts about the Pirate’s chances (since the fix is not in), I think my hope here is justified. I would argue that this hope is “affectively” (or, yes, conatively) justified by my fear of the reprisals (certainly not cognitively justified). I haven’t yet thought through what the consequences of that would be for this puzzle (which I still don’t quite get).

  7. That’s a good case, Thomas. It may be unjustified for me to hope that the Pirates win as they stand–I want them to win, I wish they would win, but there’s little point in hoping–but if things were as you describe I would be justified in hoping. So is there a practical environment effect on justification of hope? And similarly for fear–it makes little sense to fear a remote chance of a fairly bad outcome, but it makes sense to fear the same remote chance of a very bad outcome.

  8. Matt, your point about cognitive and conative dimensions is really nice. It could be extended further as well, so that you might have practical justification for holding a belief (such as that you’ll succeed), but here the justification for belief is not explained in terms of the propositional justification for the content.

  9. Jon,
    The examples in (1)-(3) are easily made into case involving the absence of justification.
    1. Bob regrets that he hurt Lou

    But the evidence Bob has that he hurt Lou is scant or non-existent. So there is something irrational about that regret. Similarlt for the fear example in (2). The more interesting case is (3).

    3. Tom hopes to win the 5k race today.

    But suppose Tom has lots of evidence that the race was already run yesterday. It’s irrational to hope for the impossible, I think, and Tom has good reason to believe that it is impossible that he wins the 5k today. It is impossible since, of course, the race has already been run.

    But apart from these, there are some nice examples above.

  10. Thanks Jon! I think the practical consequences of the belief might introduce even a third dimension. (Is that what you meant by extending it further?) Suppose that it is bad for me to fail, it is reasonably likely that I’ll fail, but fearing that I’ll fail will make me even more likely to fail. Then my fear seems to be justified on the conative and cognitive dimensions I’d discuss, but it’s still prudentially bad for me to have that fear.

    That seems to me analogous to the case where I have practical reasons for holding a belief, but lack propositional justification.

  11. Yes, Mike, that’s right–that’s what Matt’s correction involves. And, Matt, yes, that’s what I meant by extending it. So perhaps what we can say is that when the kind of justification is epistemic, then part of the explanation for the justificatory status of an attitude involves propositional justification. (This assumes, of course, that all conative and cognitive attitudes capable of justificatory status take propositions as the objects of the attitude…)

  12. Oh, sorry about that. Missed the correction. I did notice the suggestion that “wishing” seems exempt from the cognitive criticism. But I’m not sure why. Isn’t it irrational to wish for something that you already have? So Smith wishes that he had the the most valuable Matisse. But the appraisal he recently had done clearly indicates that he does have the most valuable work. Given the evidence he has, his wish seems irrational. That’s a cognitive criticism.

  13. I have a somewhat orthogonal question, which I hope you won’t might me asking. What are the grounds for recognizing what you call ‘propositional justification’ as a species of justification at all? Of course, I think I understand the idea that one can be in such a position that one *would* be justified in believing that p and q, despite the fact that one has never reflected on the matter. But it seems wrong to me to think that this fact is best captured (or even appropriately captured) by saying that the content is propositionally justified for me. Here is why:

    I take it that we have a notion of attitudinal justification because thinker’s are subject to certain cognitive limitations, and limiting capacities generally. So while the representational norm which governs belief is truth — because all representation aims at veridicality, and truth is the subspecies of veridicality appropriate to conceptual or propositional content — we need a measure of how well an organism is doing, given its limitations, at achieving its representational norm. And one can be doing perfectly well, in this sense, even though one’s beliefs are largely false.

    The problem is that if this is the reason we have a notion of (attitudinal) justification, then it seems confused to think that a *proposition* can be justified in anything like the way an attitude can. Certainly a proposition can be inferentially supported by other propositions I bear the requisite relations to. But I don’t think it can be appropriate to think of this as a form of *justification*. I think that, precisely because of the point about the origin (as it were) of the concept of (attitudinal) justification, it doesn’t make any sense to think of a proposition as being justified or unjustified. Only a believer or a belief can have these properties. Or so it seems to me.

    What I really need to make this case is an example of confusion created by recognizing a notion of propositional justification. But I’ve not been able to think of one.

  14. Majors, if you reject the idea that there is propositional justification, you’ll have to explain away the use of such in ordinary language. Just as we can say that finding the defendant guilty was justified by the evidence presented (even if the jury refuses to enter a guilty verdict), so we can say that the claim that the defendant was guilty was justified by the evidence presented (even if everyone refuses to believe what the evidence shows). My suspicion, from your first paragraph, is that you actually don’t deny the coherence of such language, but rather than you endorse a version of doxasticism best expressed in print by Alvin Goldman in “What is Justified Belief?” Goldman uses the language of ex post and ex ante justification, but it is the same distinction as what I call propositional and doxastic justification. Doxasticism is the view that claims that the language of propositional justification is to be explained in terms of doxastic justification, the simplest version of which claims that “p is justified for S” is to be understood as “if S were to believe p, the believing would be justified.” The argument Chris Menzel and I give for rejecting doxasticism is developed first in our paper “The Basic Notion of Justification,” though I’ve published versions of the argument in several other places. The argument against doxasticism is that it is bound to commit Shope’s conditional fallacy.

  15. Jon, thanks for those references. As far as I can tell, my original question is much like Majors’, and that seems to make me a doxasticist about justification (at least until I have a closer look). I take it that a doxasticist would not find your puzzle puzzling, i.e., would not accept its premise. Or is there a way of articulating your puzzle in a way that is consistent with doxasticism?

  16. Thomas, no, I think doxasticists will accept the idea that there is both doxastic and propositional justification. The question is which explains the other. The fact that there are these two kinds of justification when we evaluate from a purely cognitive point of view allows us to characterize the difference between epistemic justification and other types on the basis of this duality. Since I think propositional justification is more fundamental than doxastic, I will characterize epistemic justification in terms of the fundamental explanatory role of propositional justification. Doxasticists will need to say something different, though I’m not sure quite what.

  17. Jon, earlier you said, “it’s not that propositions are justified in themselves; it’s rather that certain propositions are justified for a person, whether or not they take any attitude toward them.” But what does it mean to be justified “for a person” if not “as an attitude”? Surely people can only, as it were, “adopt” propositions as attitudes (like beliefs). The proposition, I guess we might say, can’t be related to a person except as a stance that that person might take on the facts in question. That is, for a proposition to be justified for a person can only mean that if a person takes up an attitude towards it, then that person will be justified in doing so. And this will normally involve a “web of attitudes”, i.e., other beliefs, hopes, fears this person might have.

    And here your puzzle seems relevant. For we might ask whether a proposition like “The Pirates will win the Series” is justified for a particular person: but only if we qualify this question with “as a belief”, “as a desire”, “as a hope”, etc. Given one set of facts (about the rest of my attitudes) it may be unjustified for me to believe it but not to hope it (the scenario I sketched above); but given another set of facts (about both me and the Pirates) the opposite might be true.

    But if my sense of this is right, then the difference between cogntive and conative justification cashes out as the difference between cognitive and conative attitudes (which will demand different kinds of justification). When you say that *the believing* might be justified because *the content of the proposition (which just happens to be) believed* might be justified I think you working with an implicitly doxastic notion of content (i.e., propositional content is the sort thing one believes). Explicating this was your point of departure, to be sure, but once you do this, don’t you also bring into the foreground that propositional content is always determined relative to attitudinal content? The mental act that produces “content” is the one we must forget (or sublimate) in order to have anything like content.

    I think the view I’m defending looks like Berkeley’s idealism. (You have to imagine someone believing the proposition though no one does in order for it to have content.)

  18. Thomas, here’s the part of what you say that is, I think, confused and mistaken:

    But what does it mean to be justified “for a person” if not “as an attitude”? Surely people can only, as it were, “adopt” propositions as attitudes (like beliefs). The proposition, I guess we might say, can’t be related to a person except as a stance that that person might take on the facts in question. That is, for a proposition to be justified for a person can only mean that if a person takes up an attitude towards it, then that person will be justified in doing so. And this will normally involve a “web of attitudes”, i.e., other beliefs, hopes, fears this person might have.

    Propositions don’t have to be adopted to be justified by the evidence a person has–that’s how irrationality occurs, by refusing to follow the evidence. The fourth sentence is just the simplistic doxasticist proposal, and it succumbs to fairly easy counterexamples. Here’s two. Suppose you are exceptionally epistemically virtuous so that you won’t belief anything without adequate evidence. Then your proposal implies that every proposition is justified for you. Or again, suppose you know that believing you will succeed makes it more likely that you succeed. Furthermore, suppose your evidence that you will succeed falls just short of justifying that claim for you. Then if you added the belief, your belief would be justified because of the additional evidence that would create. But as it stands the claim that you will succeed is not justified by the evidence you have.

  19. I can’t make sense of your counter examples. Maybe they need a bit of fleshing out.

    Suppose I only believe propositions I have evidence for.

    You say that my proposal now implies that every proposition is justified for me.

    Do you mean every true proposition?

    And do you mean that I would be justified in believing all of them?

    The second one I think is self-referentially problematic. If you know that your belief in success only improves your chances (i.e., does not ensure success) then your belief in success is never evidence for your success. Otherwise the knowledge you refer to should force us to always believe that we will succeed “just in case”. But there will be times when it’s better to know you will fail than believe you will succeed, right? So unless you really can measure the contribution of positive thinking against your physiological powers, say, and say, well, it doesn’t look like I can lift this thing, but it almost looks like I can lift it, and positive thinking will put me over the top, then what you’re really saying is, “It looks like I might be able to lift this thing.”

  20. You said “p is justified for S” must mean “if S were to believe p, that belief would be justified.” I gave you a case where the counterfactual is true for every proposition whatsoever. It follows, on your account, that every proposition is justified for such an intellectually virtuous person.

    On the belief in success example, you’ll have to deny some obvious psychological data to get out of the example. Either that, or deny that evidence can support a claim to a certain degree, or that adding a little more evidence can’t move a claim from being just under the threshold needed for justification to just over it. There’s nothing in here that forces belief or anything like that. It’s just an example where if you were to add the belief, your epistemic condition would improve with respect to the content of the belief. So, by the account of meaning in the first sentence above, you’re committed to the view that the proposition is already justified for you.

    All of this is discussed thoroughly, perhaps ad nauseum, in “The Basic Notion of Justification.”

  21. I think I’m using “mean” here in a less rigorous sense than your counterexample imagines. That is, I was trying to argue that the words “p is justified for S” is an imprecise way of putting something that is better put by saying “if S were to believe p, that belief would be justified.” Maybe I would argue that in the case you propose every p-for-S is justified not every p is justified-for-S. But that would be purely as a matter of definition.

    I normally don’t know what to make of “obvious psychological data”, but you get my objection almost exactly right as a denial that “adding a little more evidence can move a claim from being just under the threshold needed for justification to just over it” (I hope you don’t mind my fixing the double negative: from “denying … that you can’t”, through (presumably) claiming that you can’t, to “denying … that you can”.) In the case you suggest, the words “a little more evidence” means “an infinitessimal amount of evidence”. That is, until you specify the exact measured value of positive thinking in terms that are not relative to the exact measured value of the existing evidence, your counter-example is begging the question. But if you did construct the example in that way (all you would have to do is add absolute values, I guess), I would probably deny that you can quantify the value of each piece of evidence beforehand and then just add them up. Actually, I think that is psychologically obvious.

    But it looks like you are working with a background of conversation in mind that I’m not very familiar with. If you have a sense that you’re just going to have rehash the argument in the paper you’re talking about to establish ground that you’re taking for granted in your puzzle, and you don’t have time for that, it’s cool with me. In any case, good luck with the puzzle; it looks interesting.

  22. Thomas, yes, there are cases where a blog is not the best place for sustained argument, so if you’re interested enough to follow up, there may be more to talk about. I’ll only say this much more here: it would be truly amazing if adding a bit of evidence couldn’t turn an unjustified claim into a justified one. E.g., had the statistical sample been slightly more positive, it would have confirmed that the candidate had won. But as it stands it didn’t.

Leave a Reply

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