Are reasons propositions?

We believe, disbelieve and suspend judgment based on reasons all the time. I think these reasons are best thought of as mental states of the subject. Other people think that they’re best thought of as propositions.

Suppose Peggy believes that God exists (Q) only if there is no unnecessary suffering in the world (~P). And she suspends judgment (i.e. withholds) on whether there is no unnecessary suffering in the world. She consequently withholds on whether God exists. We can represent her transition in thought like so.

B(Q → ~P)  [ed. note: added a missing tilde here]
W(~P)
W(Q)

This seems like a perfectly sensible bit of reasoning.

Peggy’s withholding is reason-based. What are the reasons that form its basis? I say: her belief that (Q → ~P), along with her withholding on ~P. I cite two of her mental states. My view doesn’t endanger the observation that Peggy’s reasoning is perfectly sensible.

Now suppose that you wanted reasons to be propositions instead of mental states. On the most straightforward version of this view, you’d simply identify Peggy’s reasons with the propositional contents of those mental states which I say form the basis of Peggy’s belief. But it seems to me that this view gives a very odd result in Peggy’s case, to wit, it entails that Peggys withholding on Q is based on: [(Q → ~P) & ~P]. But reasoning that way certainly isn’t sensible!

How could someone who thinks that reasons are propositions effectively handle this example?


Comments

Are reasons propositions? — 44 Comments

  1. Here’s a stab.

    She withholds belief concerning Q because she believes there is insufficient evidence either way concerning P. So then, the proposition that is the reason for her withholding concerning Q is, “There is insufficient evidence either way concerning P.”

    You might be tempted to say that it’s not the proposition “There is insufficient evidence…” that is her reason for withholding belief about P. Rather, you might want to say it’s _her belief_ that “there is insufficient evidence…” that is her reason for withholding belief. This would be consistent with something you said in your post, to wit, that her _belief_ that Q -> ~P is one of her reasons for withholding belief about P. But I’d make a distinction between “the reason” and “her reason” for her having a particular mental state. “The reason” for her having the mental state might be some aspect of her upbringing, and similarly, might be her having some particular belief or other. But “her reason” might not be that upbringing, or that belief, but rather, the thing _believed_ (i.e. the proposition) in that belief.

    In other words, it might be that she shouldn’t say “I believe X because I believe Y” (even if that might be true) but rather, that she should say “I believe X because Y.”

    In this case, then, she should say “I withhold belief about Q because there’s insufficient evidence either way about P,” and _not_ “I withhold belief about Q because I withhold belief about P”–even though the latter is true. The latter is true, but if she’s giving her reasons for withholding belief about Q, she shouldn’t say the latter. She should say the former.

    That’s my stab.

  2. It seems to me that “Peggy’s withholding on Q” describes not a proposition but a mental state, so it’s not a statement of the propositional view.

    You need Peggy to believe “possibly not Q” on the basis of “possibly P”. If I’m not mistaken the reasoning the holds. That is, you need modal operators to do the work of mental state ascriptions. Or something like that.

  3. Hi John,

    I take it the “reasons are propositions” view isn’t the whole story about reasons. After all, there’s Carroll’s paradox and there’s the fact that propositions can be hosted in a variety of ways (feared to be true, desired to be true, presented as true, etc…). Your case seems to be handled by filling out the view to include a principle of evidential balance like Feldman and Conee’s EJ. If the fundamental level of evidence consists of assertive propositional contents and Peggy’s evidence for P is counterbalanced, then her withholding is reasonable and consistent with “reasons are propositions” view.

  4. Thomas,

    I meant for “Peggy’s withholding on Q” to name a mental state, and not a proposition.

    Ted,

    The case is supposed to be understood as involving assertive propositional contents throughout. Withholding is withholding on whether the claim is true. So I don’t see how that move helps.

    In any case, I take it you agree that Peggy’s reasons for withholding on Q can’t be [(Q → P) & ~P], right?

  5. I know, but if you want to give the propositionalist a fighting chance he has to be able to translate the situation into the terms of his theory. And that’s all I suggested we do. On the propositional view, Peggy does not withhold a belief; she believes in the possibility of a proposition being false. She then reasons (quite reasonably) that possibly not Q on the basis of

    [(Q → not-possibly P) & possibly P]

    What I’m saying that the additional attitude of “withholding” smuggles a mental state in that the propositionalist (simply?) has to deal with in the ordinary way.

  6. Thanks, Thomas. This is an interesting suggestion. It would handle the case, but I think it falsely presupposes that withholding on P is the same thing as believing that possibly not-P.

    To see this, imagine a situation where we’re convinced that a claim R is either necessarily true or necessarily false (e.g. some complicated mathematical claim, or the claim that some complicated proof is valid), but we lack evidence which. So we withhold on R. But this could not be understood as believing possibly not-R.

  7. I’m with Thomas on this one. One could adopt a classical modal logic without the (N) axiom (which is a theorem of normal modal logics; hence the ‘N’: for ‘normality’) where b is the box modality and T is top:

    (N) b(T)

    In short, I don’t yet see a reason to block viewing this question in terms of some relational structure.

    (BTW, What happened to the LaTeX plugin?)

  8. Kris,

    Sorry, I wasn’t ignoring your comment, but it just showed up now. It must have initially gotten caught in a spam filter.

    I agree with you about the distinction between “the reason” and “her reasons.” The total explanation for why she withholds will include lots of stuff that isn’t among her reasons.

    I also tend to think that the general strategy, of replacing withholding with a special type of belief, is probably the most promising type of response.

    One worry I have about it, though, is that it threatens to over-intellectualize what goes on in a lot of cases. You might simply withhold in response to evidence, without thinking, “this evidence is insufficient.” Or the withholding might be a basic attitude, not based on anything.

  9. Of course there is room for the view that normative reasons are propositions and motivating or basing reasons are mental states. This is a popular view in metaethics, and it’s easy to see why a view in metaethics should generalize to a view in epistemology, and vice versa. I’m not much of a fan of such views myself, but appealing to them here would alleviate this problem.

  10. Hi Errol,

    That doesn’t alleviate the problem, because the opponent I have in mind is claiming that propositions are the motivating reasons.

    (BTW, I think the normative/motivating distinction is basically a mess anyhow, but that’s another story.)

  11. I would counter that “withholding a belief in Q” and “believing that Q might possibly be false” are two ways of simplifying Peggy’s actual situation. They are theory-laden descriptions of a complex empirical reality. They may not be quite commensurable, but we are, at bottom, trying to decide between those two ways of describing her views on God.

    Suppose we ask her whether or not God exists. She says she doesn’t know. We say, “What do you mean you don’t know?” “Well,” she says, “it’s possible that He exists, but if he did there wouldn’t be any unnecessary suffering in the world. There’s sure a lot of suffering. I guess it’s possible that it’s all necessary, but I don’t know. So maybe there’s a God, maybe not.”

    Whether she is withholding a belief in God or believing in the possibility of his non-existence seems to me to be a hermeneutic question. And since the decisive issue (in interpreting her answer) is whether her reasons are (modal) propositions or mental states, we would beg the question here to settle it in favor of withholding.

  12. Right, I wasn’t claiming that you weren’t being clear about who your opponent is. Nevertheless, a theorist who endorses the view I was suggesting can answer Yes to your title question.

  13. Hi John,

    Dropping (N) would drop the requirement that tautologies are necessarily true (w.r.t. some classical modal system0; so, the box modality in such a system could (in principle) be something weak enough to get around examples like the one in #6.

    You cannot do this within a basic Kripke structure, because (N) is a theorem of such normal (i.e., Kripke) systems, and it is one of the main properties that drives omniscience-type objections to normal systems of epistemic logic.

  14. John,

    I’m not quite sure what to make of this sort of case. It seems weird to (me) think that either the fact about the kind of psychological relation to the proposition or a proposition about that psychological relation could be my reason for withholding judgment. If the reasoning is good reasoning, it seems that the subject’s grasp that she should withhold (or something in that neighborhood) would be what rationalizes withholding judgment.

    One worry about the mental state view that I think you favor is that on the reasoning as you’ve construed it, there’s no premise about any of the subject’s mental states so while some will say that anything that is the basis of the conclusion of the reasoning will have to be a premise/represented by a premise, on your account that isn’t so. Unless the reasoning you’ve sketched isn’t the right sketch of the reasoning. If the right sketch of the reasoning is:

    B: Q–>P
    B: I withhold P (I should withhold P?)
    B: I should withhold (Q).

    I don’t see why this sort of reasoning causes trouble either for the deranged, utterly confused, and grossly immoral view that reasons are propositions or the correct view held by the angels that they are always facts. (For a refutation of the reasons as propositions view, see here: http://claytonlittlejohn.blogspot.com/2009/06/reasons-as-facts-or-propositions.html.)

    Alright, I’m off for some kayaking! John, congratulations on the Nous paper!

  15. Thomas, earlier I offered evidence that is entirely independent of the question at hand, so I’m not begging the question (in any significant sense). Let me spell out my reasoning in more detail.

    We cannot generally understand withholding on R as believing possibly not-R, because there are cases like the ones I described earlier: namely, where we withhold on claims that we know are either necessarily true or necessarily false. We might be left with interpretive questions about such cases, but they don’t include whether, in such a case, we’re believing possibly not-R.

    Once we have that on the table, then we’ll be able to construct cases structurally similar to Peggy’s, where it’s obviously wrong to interpret withholding on R as believing possibly not-R. Then your strategy breaks down.

  16. John,

    I forgot a tilde in the bit of reasoning sketched above.

    Why can’t the sense of “possible” in your response to Thomas be an epistemic sense of possibility on which epistemic possibility has to do with what is obviously entailed by what you know?

  17. Thanks, Greg. If I’m catching your drift, then your proposal requires that mathematical truths are tautolgous. Suppose that’s correct. Wouldn’t the next move on my side be to construct an example with a non-tautologous but still necessary claim (perhaps involving essence)?

  18. But we might still ask whether describing them as “withholding belief” is appropriate. If I was trying decide to decide between necessarily true or necessarily false, I don’t think I would describe my mental state as “withholding belief” in it. But we could interpret my state in that way for the sake of argument.

    In Peggy’s case, then, the propositionalist’s strategy is to go back to the pre-theorized situation and suggest an alternative (again, modal) interpretation. I don’t think the propositionalist is barred from making that move.

  19. Clayton,

    I think that something like a “grasp” of propriety might be involved in all good transitions of thought (which doesn’t include all transitions of thought). But her grasp is just another mental state of hers, so that’s okay by me.

    And I do think that the same basic problem will recur at a higher level, only about the more complicated claim “I should withhold on whether P”. Why? Because the subject might either believe that she should withhold, or she might withhold on whether she should withhold. So adverting to the sort of alternative reasoning you sketched (even with all the relevant tildes!) can’t be a general solution to the problem.

  20. Right, Thomas, I agree that handling Peggy’s case that way is consistent with not handling all cases that way. But problem cases of the very same sort remain. And we shouldn’t expect disparate solutions to problem cases of the very same sort.

    Here’s the sort of alternative case I have in mind. Keep Peggy’s case the same structurally, but switch out ‘God exists’ and ‘there is no unnecessary suffering’ in favor of ‘what Goldbach wrote to Euler in his 7 June 1742 letter is true’ and ‘Goldbach’s conjecture is true’.

    If Gregory’s right about tautologies, and mathematical claims are tautologies, then we’ll need a different example, involving a non-tautologous necessary truth.

  21. John,

    The point of my comment was that the “reasons as propositions” view obviously needs filling out. One way it needs filling out is to bring in a principle of evidential balance. Once such a principle is in play, the “withholding” case isn’t a problem for the propositionalist. In any case, it seems you agree that your worry is independent of whether the fundamental level of evidence is propositional. I take it that the propositionalist is committed to a thesis about the fundamental nature of evidence (e.g., evidence consists in assertive propositional contents) and not about inferential principles. In the Peggy case it could be that given the evidence Peggy possesses it’s just a basic movement of her mind to realize that the evidence for P is counterbalanced. That explains her withholding on P in a way that’s consistent with propositionalism. Do you think the propositionalist must include a propositional representation of basic movements of the mind?

  22. Hi John,

    I’m not sure what essences are, but we could stipulate non-tautological necessary* statements (where necessary* is the box modality in our classical system) by adding non-tautological axioms, say.

    All of this is to say that some of the mechanical issues that might lead one to dismiss reasons as propositions might be flimsy grounds if those mecahnical features can be captured within a system of modal logic, such as a weaker system than the normal systems based on Kripke structures. This is because most (or, at least: monotone) modal logics admit of direct translation into first-order logic. So, this approach wouldn’t yield a ‘reasons as propositions’ view per se, but would yield instead a ‘reasons as wffs’ view.

    I don’t have a dog in this hunt; rather, I’m suggesting that a ‘reasons as wffs’ view might be a meatier target than a ‘reasons as propositions’ view.

    I might not get to read the thread for the next few days; apologies in advance for my tardy reply if one is requested.

  23. Greg,

    Thanks again. That’s an interesting variation on the theme, one I’ve not seen defended anywhere.

    Ted,

    To answer your last question: no, I don’t think they’d need to do that, but in line with Clayton’s earlier point, there might need to be some sense in which they “grasp” the propriety of what they’re doing. Perhaps that can be as weak as having a fundamental cognitive disposition to make such transitions, whereupon they get credited with the “grasp” by courtesy.

    Your comment provokes me to get clearer on something I’ve been wanting to get clearer on, but haven’t yet been able to, at least to my satisfaction. When I say “sensible,” I mean something different from “supported by your total evidence.” It’s more procedural. It’s hard to make sense of someone whose reasons for withholding on Q are [(Q → ~P) & ~P). My puzzlement here gets traction prior to any questions about evidence.

  24. I think we may be talking past each other a bit. What I was suggesting is that your way of setting up the propositionalists problem is unfair. You imagined he would

    “simply identify Peggy’s reasons with the propositional contents of those mental states which [you] say form the basis of Peggy’s belief.”

    For the sake of argument, let us grant that much. But you then suggested that this “entails that Peggy’s withholding on Q is based on: [(Q → ~P) & ~P].”

    But to arrive at the basis you just stripped out all the mental state talk (believing and withholding) in the premises of the argument you constructed. What entitles you to leave a mental state ascription “withholding” in the conclusion?

    This move strikes me as akin to just stripping out the modal operators in the premises of an argument, leaving the conclusion untouched, as saying “that’s nonsense!” Yes, it is. But why would anyone construct their argument that way?

    The removal of the mental state descriptions in the premises needs to have provisional consequences for the conclusion before the propositionalist has anything to answer for.

    So, an unsatisfactory, but not nonsensical solution is to interpret “withholding belief” simple as negation (denial).

    Q → ~P
    P
    ————–
    ~Q

    It’s not satisfactory because it fails to model Peggy’s situation (though it does a pretty good job of capturing the spirit of her reasoning).

    Add the modal stuff, says the propositionalist, and you get an even more adequate representation. Maybe we can do more. But I do think it is question begging to tag that word “withholding” onto the conclusion, while stripping it out of the premises.

  25. Thanks for your further thoughts, Thomas.

    When I keep ‘withholding’ in the conclusion state, that’s because I’m modeling her transition in thought. My opponent and I are both trying to answer that question. We both agree about what’s going on in Peggy’s head. Now we ask, what’s the basis for her withholding?

    I keep the attitude operators in my answer; he strips them out of his. That’s just how his view works. And his view doesn’t seem to run into much trouble … until we get to cases like the one I describe.

    I concede that this is the simplest version of propositionalism. Your and Gregory’s remarks are helping nicely to flesh out variations on that basic theme.

  26. But doesn’t the propositionalist have to get out of Peggy’s head? I would have thought that no propositional structure has any hope of supporting an actual (in-the-head) mental state. The whole structure has to be abstracted from the mental states, including, like I say, the formulation of the conclusion (for which we are trying to understand the role of “reasons”).

  27. Thomas,

    I would have thought that no propositional structure has any hope of supporting an actual (in-the-head) mental state.

    If that’s true, then it’s a much more damaging objection than anything I’ve put forward.

  28. Hi John-

    My immediate thought is to take an example from fiction, which you may or may not think is an appropriate move. (But I’ll try it out anyway.)

    The example I am thinking of is Dostoevsky, “The Brothers Karamazov”. In brief, the character named Aloysha withholds judgment on the existence of God for the reason that he also believes God could not allow suffering. To reconcile, he tries to rationalize suffering. Doestoevsky has Aloysha do this by speaking about it at length. The only reasons Aloysha has are propositions.

    While Aloysha does end up saying something equivalent to “God exists only if there is no unnecessary suffering, but there is unnecessary suffering, therefore I will not say whether or not God exists” — although Aloysha says this, the astute reader typically does not throw it away as nonsense (to do so would be to throw away the novel).

    Thus my immediate thought leads me to argue that reasons are best thought of as propositions, instead of mental states, since reasons are better off expressed. It does not seem to matter what Peggy or Aloysha believe if they do not say what they believe; and they ought to say what they believe even if it is not clear. If Peggy or Aloysha do withhold an assertion, they do so because they do not wish to commit themselves to it and to answer for it.

  29. Hello Jared,

    I’m perfectly okay with examples from fiction. In fact, I approve of them!

    I think mental states can, and very often are, expressed. So there’s where I’d take issue with your analysis.

  30. My view isn’t very developed on this point, but I think it is based on a distinction between reasons and causes. Mental states, it seems to me, need to be causally explained, and a merely logical (propositional) structure will never explain why someone believes something (or withholds belief in it). Even “justifying” a mental state can be more complicated (by social factors) than mere propositional support.

    I take it you are saying that having a reason for actually holding a belief (i.e., being in the mental state of holding or withholding belief) means being in certain other mental states (normally, holding or withholding certain other beliefs).

    I think the propositionalist wants a more objectively binding sense of “having reasons”. On this view, you may have reasons to hold a belief without knowing it. Mental states that happen to not impinge on each other causally (say because of other mental states that are not even candidates to be “reasons”, e.g., you might have various reasons to believe something but withhold belief “because” you are drunk), may nonetheless stand in certain logical relations to each other. Such reasons (like the propositions they are reasons for) just ain’t in the head, we might say.

  31. Thanks John-

    Not being an expert on the topic, I suppose I’m confused as to the difference between saying Peggy’s “belief that (Q → ~P), along with her withholding on ~P” is more sensible than saying “Peggy’s withholding on Q is based on: [(Q → ~P) & ~P]”. To my lay ears, the words “along with” sound equivalent to the conjunction “&”. Why aren’t the two equivalent?

    I’m almost convinced by comment #5 (Thomas), to which you responded in comment #6 that Thomas’ suggestion “falsely presupposes that withholding on P is the same thing as believing that possibly not-P.” You draw on the example of someone believing a claim that “is either necessarily true or necessarily false”. I think that example misfires, since nowhere in the original belief is the truth or falsity of the claim considered necessary. It is necessary for the argument’s validity, but not necessary for believing Q (which is why, I presume, one withholds on Q given the difficulty of saying there is no unnecessary suffering, ie all suffering is necessary).

    That’s what I hoped to raise (but failed to raise) with the Dostoevsky example.

    Great post and commentary; The exchange between you (John) and Thomas is extremely thought provoking!

  32. Hi John

    I’m not convinced that the reasoning is „a perfectly sensible bit of reasoning“. Here’s a new example. Old Lady lives in our neighbourhood. Our neighbours‘ dearest gossip topic is whether she is childless. We, however, withhold on whether she is childless. The following transition in thought occurs: I believe that if Old Lady is my mom, she isn’t childless. I withhold on whether she’s childless or not. Therefore, I withhold on whether Old Lady is my mom.

    Obviously, that’s not a sensible piece of reasoning! There are good reasons for believing that someone else is my mom. What went wrong? If it’s appropriate to withhold on whether P and it’s appropriate to believe Q->-P , it can’t be appropriate to believe Q. For, if one believed Q, one had a reason to believe non-P. But, if it’s inappropriate to believe Q, one can either refrain from believing or one can believe that non-Q. Thus, below the line it should read „not-B(Q)“.

    I don’t know whether this has any interesting consequences.

  33. Thomas,

    I agree with the point about causation. It’s one main reason to think of reasons as mental states, rather than propositions. The view that reasons are mental states does not, as best I can tell, make having reasons, or even having good reasons, any less objective than the view that reasons are propositions.

    Jared,

    You draw on the example of someone believing a claim that “is either necessarily true or necessarily false”. I think that example misfires, since nowhere in the original belief is the truth or falsity of the claim considered necessary.

    I agree that the original example doesn’t involve such detail. I need to describe a different case to meet Thomas’s point here head on. I tried to do that in #22 above.

    Tim,

    Nice example about Old Lady. I don’t want to commit to the claim that such reasoning is exceptionlessly sensible. In your example it seems very weird, I think, because we presuppose that you have lots and lots of good evidence about who your mother is. We could add details to make it seem perfectly sensible. But, in any case, I’d be perfectly happy so long as Peggy’s reasoning in my original example seems perfectly sensible. Given its content, do you agree?

  34. Maybe I chose the wrong word. By “objective” I meant independent of Peggy’s (presumably subjective) mental states. I will grant that it may be an objective matter (a straightforward empirical question) whether or not Peggy is in one or another state. But the propositionalist wants the question of whether or not she has reasons to be answered independently of the answer to the empirical question. It’s a “logical” issue to him.

    It’s a bit of a subtle issue actually. All the causal/empirical stuff returns (or remains) in so far as we insist on asking whether Peggy (a real person with real states) has reasons. Reasons, in a sense, come to resemble mental states in so far as they are assigned to particular individuals. But the propositionalist, sticking to his guns, would say the following: it may be that reasons are indexed to mental states (like beliefs and doubts) but there are not identical with them. They are not, to put it another way, reducible to mental states. And that’s because our reasonings are ultimately not reducible to causal accounts of belief formation.

  35. Thomas, I agree that there’s more involved than just mere causation. The relevant causal relations must be facilitated by the agent’s own cognitive dispositions, thus bringing in dimensions of responsibility and evaluation.

    On my view, although Peggy’s reasons just are her mental states, whether her reasons are good or bad depends on lots of stuff other than her mental states (including, perhaps, logical relations among the propositional contents of her beliefs, experiences, etc.). Perhaps that could go some way toward recovering what the propositionalist really wants.

  36. The propositionalist wins if the only determinants of whether reasons are good or bad are logical. Because then reasons would be “in essence” propositions. I’ve got to say I’m starting to talk myself into being a propositionalist (at least about reasons).

    Moreover, I think you can only win the case for reasons = mental states if the good/bad question can be settled without looking beyond such states. If the quality of a reason depends on something other than the mental state then they can’t be identical.

    But I may be lapsing into metaphysics now.

  37. I’ve got to say I’m starting to talk myself into being a propositionalist (at least about reasons).

    Alas, all my good efforts are going to waste! 😉

    About this point: “you can only win the case for reasons = mental states if the good/bad question can be settled without looking beyond such states.” On the most natural understanding of ‘looking beyond such states’, I’d have to say that’s not true. Suppose reasons are mental states. Now suppose that simple reliabilism is true: my perceptual experience E is a good reason for me to believe Q IFF E reliably indicates Q. What makes E a reliable indication of Q will include all kinds of stuff about my environment, which I presume is “beyond” my mental life.

  38. E is a mental state, right?

    Now, either E, being reliable, is the good reason or E’s being reliable is the good reason. If the latter, something other than the mental state is the good reason, namely, whatever makes it reliable. That something may or may not itself be a mental state, if it is, all is well. But if it is assigned to the environment the reasons = mental states thesis falls. Reasons become “reliable mental states” at minimum and the door opens (wide) for the propositionalist, who now says that the difference between a reliable and unreliable mental state is bound to their content, which is, he would argue, propositional.

  39. I wrote my first comment under the vague impression that you thought Peggy’s reasoning was rational because of its form. So your reply was certainly helpful. However, I’m still not convinced that Peggy’s reasoning is rational. Here’s an argument why.

    Suppose I wonder how my colleagues A and B are related agewise. There’re three options: A is older than B, B is older than A, A and B were born the same year. I get to know a reason that speaks against A’s being older than B. Do I have a reason to believe that B is older than A? Obviously not. If there’re three options, a reason against one of them isn’t a reason in favour of one of the other options. Suppose I wonder what attitude is appropriate with regard to proposition p. Again, there’re three options: Believing p, believing not-p, withholding on whether p. I get to know a reason that speaks against believing p. Do I have a reason for withholding on whether p? Since the cases are analogous, I don’t think so. If reasoning is about choosing the appropriate attitude and there’re three attitudes to choose between, eliminating one doesn’t create a reason for one of the options left.

    Of course, this argument isn’t conclusive. But I wonder why the cases are supposed to be different. How *could* a consideration that taken at face value only shows believing Q to be inappropriate be a reason to withhold belief? That’s my challenge.

    An additional worry: What is the propositional content of withholding belief? Do you agree that one can define withholding as (intentionally) not believing p and not believing not-p? If you do, the propositional content of withholding on whether A should be the pair consisting of A and not-A. (Analogy: preferring P to Q has a pair of propositions as its content.) Alternatively, the content of withholding on whether A might be the disjunction . Withholding on whether Q would be appropriate because -P> and don’t help to decide whether Q. At least I’m wondering why the propositionalist should accept that the relevant propositions in Peggy’s case are -P> and . The relevant propositions might as well be -P> and .

  40. Thanks for your further thoughts, Tim.

    About your last point, I think the content of withholding on Q is [Q].

    About whether the argument is rational, I still think the original case differs importantly from your new case. But instead of belaboring the point, I instead want to change the example so that it’s obviously sensible in virtue of its form. So let Peggy’s reasoning be represented like so:

    B(Q ↔ ~P)
    W(~P)
    W(Q)

    The biconditional in the first line does the trick, wouldn’t you say? I don’t know why I didn’t just set it up this way to begin with.

    It certainly wouldn’t be sensible to withhold on Q based on the reasons [(Q ↔ ~P) & ~P], because they obviously entail Q.

  41. Hello.
    Sorry for the philosophy of language, but maybe it could be interesting: the signal ‘B(p)’ refers to a mental state right? It is the mental state of believing. To talk about proposition we must take another way – the proposition that B(p) is expressed by that signal.
    Days ago you were discussing if B(p) refers to a mental state or to a proposition. But this is misguiding the semantic levels of phylosophic talk. B(p) don’t refers to a proposition, but to a mental state (or if you prefer, to a ‘state of things’).
    So, this being recalled: should we indentify Peggy’s reasons with the propositions expressed by the premisses, or with her mental states refered by that premisses.
    This way to put the question may be clearer. I thing it is a very interesting topic.

Leave a Reply

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