Skip to main content.
March 28th, 2005

Hoping and Wanting

Much to be said about San Francisco, testimony, contextualism, value-laden epistemology, horribly sad news from home and the assorted things that happen over a week. But instead of saying any of them I’ll just make one small grammatical observation. I’m sure this is something that most of you know, and it is well-investigated by linguists, but it was news to me. It’s surprising, from some perspectives, that (1) is grammatical while (2) is not.

(1) John wants Mary to win.
(2) *John hopes Mary to win.

There’s a certain kind of philosophical program (one that no one really adopts but which many people seem to feel the pull of) that wants (hopes?) to have all studies into psychological states be redrawn as studies into the syntax and semantics of words for psychological states. Now I don’t think this is an entirely bad idea. It’s important to learn that not all mental states are reducible to beliefs and desires, for example, and linguistic analysis of e.g. the notion of intention can suggest grounds for anti-reductionism.

But it can also go too far. It would be a very bad mistake to conclude from the distinction between (1) and (2) that there are these two mental states, wanting and hoping, such that one of them is an attitude towards the world and one of them is an essentially first-personal attitude. Sometimes a grammatical rule is just a grammatical rule, and this is one of those times. Just how often this mistake is made is I think a fairly interesting question.

Posted by Brian Weatherson in Uncategorized

31 Comments »

This entry was posted on Monday, March 28th, 2005 at 12:49 pm and is filed under Uncategorized. You can follow any responses to this entry through the comments RSS 2.0 feed. Both comments and pings are currently closed.

31 Responses to “Hoping and Wanting”

  1. Wai-hung Wong says:

    This may be relevant to your point: The Chinese word we use to translate “want” and the one we use to translate “hope” allow the translation of (1) and (2) into two Chinese sentences having the same syntactical structure.

  2. Paul Portner says:

    That’s not the only difference between ‘want’ and ‘hope’. For example:

    (3) *John wants that Mary will win.
    (4) John hopes that Mary will win.

    I think that when a verb can take an infinitive introduced by ‘for’, as both ‘want’ and ‘hope’ can, it is idiosyncratic whether it can drop the ‘for’. So your (1)-(2) may not, by themselves, indicate anything all that important. But I think that whether a verb can take an indicative complement is more semantically telling.

    So perhaps it’s not so crazy to hypothesize a meaning difference between ‘hope’ and ‘want’. Actually I did once propose a difference (in my dissertation from UMass/Amherst, so long ago – 1992 – that I don’t anymore have any opinion of whether it worked well or not). However, it wasn’t the difference that came to your mind, between an attitude towards the world and a first person attitude. Rather it had to do with how temporal semantics is encoded in indicatives and infinitives. Something like the following: While the type of infinitive which occurs with ‘hope’ and ‘want’ necessarily talks about the (relative) future, it doesn’t do so in the same way as ‘will’. ‘Hope’ is compatible with both the kind of tense that exists in indicatives and the temporal meaning of ‘for’, while ‘want’ is only compatible with the latter.

    I don’t know whether a semantic difference explaining (1)-(4), among other things one would hope, would count as “just a grammatical rule” to you or not.

  3. Brian Weatherson says:

    Paul, I agree that there is something different in the lexical entry for ‘hope’ and ‘want’. What I disagree with is that this should be of any interest to philosophers of mind or moral psychologists. The target here (perhaps not clearly expressed!) is someone who says something like “The Humean belief-desire model of the mind is clearly mistaken, because it collapses hopes and wants into the single category of desires, and this data shows they should be kept separate.” I don’t know anyone who makes that argument, but it wouldn’t surprise me to hear it made.

  4. Brendan says:

    Question: I wonder when we take something just to be a grammatical rule, and when we take it to be representative of a difference in, say, mental states?
    Would there be some principle way of deciding? If deciding ends up being mostly arbitrary, I would think this would count against the view you mention above.

  5. Gilbert Harman says:

    An important difference between hoping and wanting is that it is inconsistent to hope for two incompatible things but not inconsistent (and even quite common) to want two incompatible things. It can make sense to want Mary to win and also want her not to win in a way that it does not make sense to hope that she wins and hope that she does not win.

  6. Wai-hung Wong says:

    One way of being ambivalent about a possible state of affairs is to hope that it obtains and to hope that it does not obtain. E.g. I have applied for law school and hope that I will be admitted (so that I will please my parents), but I also hope that I will not be admitted (so that my parents will let me pursue my other interests).

    Of course, the original point in question is not whether wanting and hoping have any difference, but whether such difference (if there is any) can be established by considering some grammatical features of the words “want” and “hope”.

  7. marc moffett says:

    Hi brian, I agree that jumping to the conclusion simply on the basis of (1) and (2) would be a mistake—verbs do occassionally exhibit idiosyncratic selectional properties. But would you agree that it gives us some pretty strong prima facie grounds for being suspicious of the proposed identification? For at least not predicating too much of our philosophy on it until we have investigated the issue further?

  8. Mike says:

    I hope that Illinois wins the NCAA’s and I hope that is not asking too much.
    Frankly, for all I know, those are incompatible hopes. Can’t I sensibly have these hopes?
    Maybe I cannot hope for two things that I know are incompatible.
    But then, suppose I know these hopes are incompatible:I hope you win the NCAA’s and I hope that if you win, you show some grace in winning. It seems like I can sensibly have these hopes knowing that, given your psychology, you will never win graciously. Knowing, in short, that these are (at least psychologically) incompatible.
    Maybe it is logical incompatibility that matters here. But I doubt it. Someone like Martin Luther (and perhaps Descartes) might sensibly hope for two logically incompatible outcomes certain that a perfect being could make it happen.

  9. Jamie says:

    Mike, I don’t understand your first example. What do you mean by saying that for all you know that Illinois wins might be incompatible with that not being asking too much? How could those be incompatible unbeknownst to you — might they be physically incompatible?
    I think your second speculation is probably right: that you can’t hope for two things together knowing them to be incompatible. I think that when you know I won’t win graciously, you can’t really hope that I will. What makes it seem like you can is that the optimism that allows your hope also wedges open a small possibility that you weren’t letting stand in the way of your knowledge, namely, the small possibility that I might, just this once, act out of character and win graciously. You can’t say to yourself, “No, it’s just impossible, I hope it happens.” You have to say, “It’s so out of character, he’ll never win graciously, but well, just maybe he will this one time; I hope so.”

  10. Brian Weatherson says:

    A few quick answers.

    I’m inclined to agree with Gil that we can’t hope for X and hope for ~X. I don’t know how far this extends though, and since I was planning a post on a similar topic I won’t go into it here.

    I think this is a little more like the kind of thing that might suggest hoping and wanting are different mental states, though not through any direct means. The point is that to explain how we can want X and want ~X it seems we have to distinguish prima facie wants from all things considered wants, and certain crude versions of Humeanism might have trouble with that. (But if they do we don’t need fancy grammatical data to show they are bad theories.)

    I don’t really agree with Brendan that we need a principle here. Philosophy is hard – there won’t always be a royal road to the truth. More importantly, languages are arbitrary in ever so many ways, and we needn’t expect the structure of the world to reflect the structure of language in any significant way. Put another way, there are millions of explanations of why the language may have chosen to treat ‘hope’ and ‘want’ differently, and I can’t see why the presence of distinct hope and want boxes in the head would be close to the best explanation of that. If having two boxes is the best explanation of the linguistic distinction, then we should posit different mental states. But there’s no principled way to do inference to the best explanation.

  11. Ralph Wedgwood says:

    I also like to use Wai-hung’s test of translating the target expressions into other languages. When a certain phenomenon shows up in lots of languages, I’m impressed; when it only shows up in English, I tend to think that it has less philosophical relevance.

    In my view, the object of the philosophical inquiry is either the concepts of hoping and wanting, or else those features of the actual states of hoping and wanting themselves that can be discovered by means of investigating these concepts. Of course, these concepts can be expressed in many languages; as philosophers, we’re not interested in the details of the grammar of any particular dialect of English.

    In fact, however, quite a lot of languages distinguish between wanting and hoping. E.g. in French, you say ‘Je veux qu’il vienne’ using the subjunctive, but ‘J’espere qu’il viendra’ using the future tense; there are similar differences in many other European languages as well.

  12. Mike says:

    Jamie

    For all I know the fix might be in, for instance, and Michigan State will win.

    But then you say,

    “_I think that when you know I wonít win graciously, you canít really hope that I will_”

    And you suggest that if I know that you will win and know that if you win you will not win graciously, then it is impossible that you win graciously. But of course this is not true. As everyone knows, I know that p entails p but it does not entail []p. So,then, what I know does not entail that it is impossible that you win graciously.

  13. Gilbert Harman says:

    Brian: “The point is that to explain how we can want X and want ~X it seems we have to distinguish prima facie wants from all things considered wants…”

    You can coherently want A and also want not-A even when you are considering everything.

  14. Wai-hung Wong says:

    You can also coherently hope A and hope not-A when you are considering everything. Besides being ambivalent about A, there may be other ways of hoping both A and not-A, such as the following: Hope, like desire, can be non-final; that is, I can hope A because I hope B where A is a means or necessary condition of B. Now suppose I hope (equally strongly) both B and C, where B and C are not obviously incompatible; but A is necessary for B while not-A is necessary for C. Because of my hoping B and C, I also hope A and not-A.

  15. Wai-hung Wong says:

    I was a little sloppy in my last post. By “hoping B and C” I meant “hoping B and hoping C”; and by “hope A and not-A” I meant “hope A and hope not-A”.

  16. Jamie says:

    Mike,


    For all I know the fix might be in, for instance, and Michigan State will win.

    But that doesn’t make that Illinois will win incompatible with that not counting as asking too much. P doesn’t make not-P incompatible with Q.


    And you suggest that if I know that you will win and know that if you win you will not win graciously, then it is impossible that you win graciously.

    No, I didn’t suggest that. I can’t see what I said that suggested it to you, but in any case I think you’ve misunderstood.
    I may not have been clear. The modality is all epistemic here, as I’m intending the ‘necessary’s and ‘possible’s. Does that clarify what I said?

  17. Matt Weiner says:

    Upthread marc asked:

    But would you agree that it gives us some pretty strong prima facie grounds for being suspicious of the proposed identification? For at least not predicating too much of our philosophy on it until we have investigated the issue further?

    I wouldn’t disagree. “Request” and “ask” I think exhibit a similar difference in less marked form—more discussion on my blog—and I don’t think that gives us any evidence that requesting and asking are different types of acts. Perhaps I should say that even if it turns out that this difference is extremely marked, I wouldn’t think it would give us any evidence that requesting and asking were different types of acts.

    If there were some other reason to support a theory on which requesting and asking were importantly different, then a linguistic difference might support such a theory, but in the absence of such reason/theory I don’t think it gives us much if any reason to look for one.

  18. Matt Weiner says:

    Upthread marc asked:

    But would you agree that it gives us some pretty strong prima facie grounds for being suspicious of the proposed identification? For at least not predicating too much of our philosophy on it until we have investigated the issue further?

    I wouldn’t agree. “Request” and “ask” I think exhibit a similar difference in less marked form—more discussion on my blog—and I don’t think that gives us any evidence that requesting and asking are different types of acts. Perhaps I should say that even if it turns out that this difference is extremely marked, I wouldn’t think it would give us any evidence that requesting and asking were different types of acts.

    If there were some other reason to support a theory on which requesting and asking were importantly different, then a linguistic difference might support such a theory, but in the absence of such reason/theory I don’t think it gives us much if any reason to look for one.

    PS Brian—I think your spam blocker may be set up so as to prevent someone from posting a comment immediately after previewing it.

  19. Matt Weiner says:

    No, either it’s set up so as to post comments when you hit preview or I screwed up. Sorry.

  20. Mike says:

    Well, let’s see.

    1. I hope Ilinois wins
    2. I hope that Illinois winning is not asking
    too much.

    You say,

    But that doesnít make that Illinois will win incompatible with that not counting as asking too much. P doesnít make not-P incompatible with Q.

    Suppose the fix is in. Then it is asking too much (isn’t it?) that Illinois win, since that is something that Illinois cannot do. I’d be asking for something that cannot occur.

    Then you say that you say nothing that suggests that it is impossible to win graciously. You say,

    No, I didnít suggest that. I canít see what I said that suggested it to you . .

    Let’s see. In your earlier post you said this,

    What makes it seem like you can is that the optimism that allows your hope also wedges open a small possibility that you werenít letting stand in the way of your knowledge, namely, the small possibility that I might, just this once, act out of character and win graciously. You canít say to yourself, ďNo, itís just impossible, I hope it happens.Ē

    You are discussing my earlier case where I say (1) I know that you won’t win graciously and (2) I hope that you do. Your explanation above of why I think (2) is sensible despite (1), is that my optimism wedges open a small possibility that you will act out of character. But since you think that I cannot sensibly assert (2) given (1), you presumably think that there is not even this small possiblity. This at least suggests that in your view (1) somehow makes it impossible that you act out of character, as it obviously is not.

  21. marc moffett says:

    Matt, point taken. The ask/request pair is pretty weird because “request” seems to otherwise behave like a typical subject-raising verb, doesn’t it? In contrast, “hope” doesn’t. (For example, “I requested something” is fine, but *“I hoped something” is bad.) The thing about Brian’s example is that the differences he notes tend to cluster with other differences that one might reasonably take to mark semantic differences between properties and relations. Your example is atypical precisely because it doesn’t bring the rest of the cluster with it—which makes we think that “x requested me to come to his office” is ok.

  22. Gilbert Harman says:

    “Ask” and “request”

    I asked a question.
    I requested a question.
    I asked whether Bob had arrived.

    • I requested whether Bob had arrived.
    • I requested a seat in the rear of the restaurant.

    • I asked a seat in the rear of the restaurant.
    • I asked for a seat in the rear of the restaurant.

    • I requested for a seat in the rear of the restaurant.
    • I asked the waiter for the check.

    • I requested the waiter for the check.
    • I requested the check from the waiter.

    • I asked the check from the waiter.
    • Careful, pal, you are asking for it.

    • Careful, pal, you are requesting it.
  23. Matt Weiner says:

    OK, when I said that asking and requesting are the same kind of thing I should have been more careful—I should have said that asking someone to do something and requesting someone to do something are obviously the same kind of thing. I’ll stick to that.

    But Gil Harman’s first few examples (at least) convince me that I need to be more careful about my statements. Maybe that is a case that the linguistic evidence gives us prima facie reason to distinguish the acts involved, until we get down to the formulations that denote the same kind of thing even though they are linguistically different.

  24. Jamie says:


    Suppose the fix is in. Then it is asking too much (isnít it?) that Illinois win, since that is something that Illinois cannot do. Iíd be asking for something that cannot occur.


    Maybe, but that doesn’t make the two propositions incompatible, either. Something that makes Q false doesn’t make P and Q incompatible.


    But since you think that I cannot sensibly assert (2) given (1), you presumably think that there is not even this small possiblity. This at least suggests that in your view (1) somehow makes it impossible that you act out of character, as it obviously is not.


    But, as I noted, the kind of possibility and necessity in question are epistemic. The small possibility is epistemic. But you cannot regard something you know to be false as epistemically possible.
    I can see how interpreting the modalities as alethic, maybe metaphysical, would make it puzzling how your knowing something could imply that you can’t think its negation is possible. But I thought the explicit gloss that the modality is epistemic would erase that puzzle. No?

  25. Mike says:

    Maybe, but that doesnít make the two propositions incompatible, either. Something that makes Q false doesnít make P and Q incompatible.

    Really? Suppose what makes Q false is the fact that it is a contradiction. There is then no world in which P and Q are true. They’re incompatible, wouldn’t you say? In fact, logically so. Now suppose what makes P false is that the fix is in. Every suitably described world in which the fix is in P is false and Q is true—i.e., it is expecting too much that P is true. In each of these worlds (including ours) you can hope that Illinois wins, but your hope is not compatible with also hoping that it is not asking too much.

    But you cannot regard something you know to be false as epistemically possible.

    I can’t see it. I know that water is not XYZ. Does that mean it is not epistemically possible that water is XYZ? That could not have been discovered? You might have a hard time finding many who believe that.

  26. Matt Weiner says:

    I know that water is not XYZ. Does that mean it is not epistemically possible that water is XYZ?

    Yes. It doesn’t mean that it was never epistemically possible, but it means it’s not epistemically possible for you now.

  27. marc moffett says:

    Oops! As Matt pointed out in an email, I intended “object-raising verb”; though I suppose it is actually a control verb. However, I believe the point I was making is unaffected by this difference.

  28. Mike says:

    Well, Matt, somebody better inform the two dimensionalists who seem to be denying exactly what you’re asserting. In claiming that it is epistemically possible that water is xyz they mean that there is a world that, if taken as the actual world, it is true that water is xyz.

  29. Matt Weiner says:

    I always thought the two-dimensionalists meant that it was epistemically possible that water was xyz before we figured out that it was H2O, but I could stand to bone up.

  30. Jamie says:

    Mike,

    Ok, here was your original point:


    I hope that Illinois wins the NCAAís and I hope that is not asking too much.
    Frankly, for all I know, those are incompatible hopes. Canít I sensibly have these hopes?

    I agree that it is possible to understand “those are incompatible hopes” in this context to include the case in which one of the two hopes is impossible, but you didn’t express what you meant in a very natural way.


    I know that water is not XYZ. Does that mean it is not epistemically possible that water is XYZ? That could not have been discovered?


    Yes, it does mean that, as Matt says.


    Well, Matt, somebody better inform the two dimensionalists who seem to be denying exactly what youíre asserting. In claiming that it is epistemically possible that water is xyz they mean that there is a world that, if taken as the actual world, it is true that water is xyx.

    Chalmers, for one, thinks of something as epistemically possible when it is not ruled out a priori. But that is a revisionary usage that he adopted for his own purposes. You might say, Chalmers is speaking of a special case of epistemic possibility: the case in which the knowledge in question is all and only what is known a priori.

  31. Jonathan Lundell says:

    It would be a very bad mistake to conclude from the distinction between (1) and (2) that there are these two mental states, wanting and hoping, such that one of them is an attitude towards the world and one of them is an essentially first-personal attitude.

    I’m not at all sure I follow the distinction. I’m a little uncomfortable with “mental states”, but accepting that language, wanting and hoping are certainly different states.

    In particular, the typical dictionary definition (and ordinary usage) of “hope”, to wish for something with expectation of its fulfillment, to look forward to with confidence or expectation, is distinguished from “want”, which doesn’t imply confident expectation.

    Want‘s grammar is tied to its etymological meaning, “lack”. And notice the radical difference in meaning between:

    (1) I want for nothing.(2) I hope for nothing.

    I want (in both the ordinary and etymological senses) a billion bucks. But I don’t hope for a billion bucks, if I’m sane.

    Hey, they’re different words, with overlapping clouds of meaning. Want has an objective element (lack) and a subjective element (desire). Hope likewise suggests lack and desire, but also expectation. And their grammars are quite different.