Hope

Let's start by individuating the general attitude of hope.Footnote ³ Hope is a propositional attitude, which I will henceforth think of as a relation between a cognitive agent A and a proposition p: A hopes that p iff A stands in the relation of hoping to the proposition that p. Hope is essentially a desire, a pro-attitude, but it also seems to contain a belief component, requiring, at a minimum, the belief that what is desired is in some sense not impossible (see McKaughan (Reference McKaughan2013) ). At least it is hard to imagine hoping for something you believe is utterly impossible. Hope is also therefore often more apt for moral assessment than mere belief. Contrast merely believing that A will kill B with hoping that A will kill B; in the latter, you have a pro-attitude towards the killing, and you seem to believe it is not impossible for it to happen, but in the former you merely believe that it will happen, without any particular pro-attitude towards it. Under normal circumstances it is natural to judge you more harshly when you hope it will happen than when you merely believe it will happen (see Pojman (Reference Pojman1986), 165).

In order to fully grasp the nature of hope, it is useful to contrast hoping with wishing. Wishing, like hoping, is essentially a kind of desire, with a belief component (namely, the belief that what is wished is in some sense not impossible), but wishing is nonetheless logically very different from hoping. Wishing is counterfactual, in the sense that it makes no sense to say that you wish that the world is a certain way; rather, it makes sense to say only that you wish that the world were a certain way. Hoping, on the other hand, is not counterfactual, in the sense that it makes sense to say that you hope that the world is a certain way, but it makes no sense to say that that you hope that the world were a certain way. So, in short, by wishing, your content points to a merely possible world, but by hoping, your content points to the actual world.

The modality involved in hoping and wishing seems to be of different strength. In the case of hoping, the belief component requires that what you hope for is epistemically possible, but in the case of wishing, the belief component requires that what you wish for is logically possible.

The fact that wishing is counterfactual, but hoping is not, fits well with the fact that claim (1) but not (2) seems linguistically or conceptually permissible, when we force a reading in which the knowledge is transparent in the sense that I know that I know that p:Footnote ⁴

1. (1) I know that p, but I wish that not-p.

2. (2) *I know that p, but I hope that not-p.

There seem to be at least two potential sources for (2) not being permissible in the same way that (1) is. First, knowledge is factive, so if you know that p, then p is true. You cannot know falsely. Now, assume that you transparently know that p. Then p is true, but you can still wish that not-p, because the content of your wish points to a merely possible world, not the actual world, so there is no incompatibility between what you know (p, in the actual world) and what you wish (not-p, in a merely possible world). Since the content of your wish points to a merely possible world, not the actual world, there seems also to be a presupposition that what you wish – namely, not-p – is actually false, but there is also no incompatibility between what you know (p, in the actual world) and the presupposition of your wish (not-p is false in the actual world). So, you can wish contrary to your knowledge. But you cannot then hope that not-p, because the content of your hope points to the actual world, not a merely possible world, so there is an incompatibility between what you transparently know (p, in the actual world) and what you hope for (not-p, in the actual world).

Second, if we substitute ‘certain’ for ‘know’, then (1) but not (2) still seems linguistically or conceptually permissible. Now, unlike knowing, being certain is not factive, but like the transparent knowing in (2), being transparently certain leaves insufficient room for doubt. But hoping that p requires room for doubt about the truth of not-p. Recall that it makes little sense to hope for something you take to be impossible. (You can hope you're wrong about what you take to be impossible, but that is a different matter.) So, it seems another source of what's wrong with (2) is that, while the transparent knowledge leaves insufficient room for doubt, hope requires sufficient room for doubt. This in turn makes sense of the fact that, if we read (2) as not being transparent in the above sense, i.e. I know that p, but I don't know (nor am I certain) that I know that p, then (2) seems linguistically and conceptually permissible, i.e. it seems possible to still hope that not-p. This is because there is then sufficient room for doubt. The hope that p requires such room for doubt about the truth of not-p. So, to the extent that the knowledge in (2) removes all such room for doubt, (2) sounds worse, but to the extent that the knowledge in (2) leaves open such room for doubt, (2) sounds better. This all fits well with the belief component of hope, namely, the belief that what you hope for is not epistemically impossible.

The fact that hoping is not counterfactual but wishing is also fits well with the fact that claim (4) but not (3) seems linguistically or conceptually permissible, when we force a reading of them in which the belief is a mere belief in the sense that I believe that p, but I neither know nor am certain that p:

1. (3) *I believe that p, but I wish that not-p.

2. (4) I believe that p, but I hope that not-p.

Belief is not factive, so if you believe that p, p might still be false. You can believe falsely. Now, assume that you merely believe that p in the sense that you believe that p, but you neither know nor are certain that p. Then you believe that p, but there is sufficient room for doubt about your belief. So you can still hope that not-p, because although the content of your hope points to the actual world, not a merely possible world, as far as you can tell there need not be an incompatibility between what you believe (p might be false, in the actual world) and what you hope (not-p might be true, in the actual world). So, on this reading of it, (4) is permissible.

But we can also force a different reading of (4), on which it is impermissible. For example, pace Williamson (Reference Williamson2000), knowledge entails belief, so if I know that p, I believe that p. If we read the belief in (4), not as mere belief, but rather as a component of knowledge, then (4) sounds more like (2) above. For example, if my belief that my wife is cheating on me is just a component of my knowledge (or certainty, for that matter) that she is (if I have seen it with my own eyes, say), then it is impermissible to hope that she is not. So, it seems that it is the room for doubt in the mere belief involved in (4) that makes it permissible to hope against it. To the extent that we force a reading on which we remove this room for doubt, (4) becomes impermissible.

A plausible hypothesis is thus that, the stronger your belief that p is, the less permissible it is to hope that not-p; conversely, the more permissible it is to hope that not-p, the weaker your belief that p must be. It seems to be a matter of degree. So, for example, if your professor tells you that you will fail the exam due to an honour code violation, and you come to firmly believe this, then it is not so permissible to hope that you will still get an A on the exam. But if one of your fellow students tells you, and you don't come to firmly believe it, but merely suspect it could happen, then it is more permissible to hope that you will still get an A on the exam.

Concerning (3), the relevant reading seems more strained. But note that if you merely believe that p in the above sense, i.e. you don't know p nor are you certain of p, it seems to meFootnote ⁵ that it is linguistically or conceptually impermissible to wish that not-p. Recall that the content of your wish points to a merely possible world, not the actual world, so the second conjunct of (3), wishing that not-p, seems to carry with it a presupposition of transparently knowing that p, but the first conjunct of (3), merely believing that p, seems to conversationally implicate not transparently knowing that p. There is thus an incompatibility of a conversational implicature and a presupposition of (3).Footnote ⁶ But, as mentioned above, this is perhaps somewhat strained. Fortunately, it doesn't matter for present purposes, where hope rather than wish is our main concern, so let's simply leave it aside for now.

It's worth noting that the counterfactual difference between hoping and wishing also fits well with the fact that wishing is best directed towards the past, not the future, but hoping is best directed towards the future, not the past. Intuitively, the past is fixed, so you can only wish it were different, not hope it is different, but the future is not thus fixed, so you can hope it will be a certain way, not wish it were a certain way. You can of course hope that the past is different from what you believe it is, but that is a different matter; you cannot hope that the past is different from the way you transparently know (or are certain that) it is. To the extent that you know that the future will turn out a certain way, you can also of course wish that the future were different, but to the extent that you don't transparently know how the future will turn out, which is most of the time, you cannot wish it were different (cf. the discussion above concerning room for doubt).

Note that there are some odd cases of wishing that don't fit the above pattern, e.g. making a wish, or wishing upon a star, but such cases seem to simply use the word ‘wish’ for the concept of hope, or wanting. Note also that there are some cases of hoping that don't fit the above future-directed pattern, e.g. hoping that you didn't make a fool of yourself yesterday, or hoping that your kids had fun at the birthday party earlier today, or hoping that the weather is nice where your wife is right now, but such cases are still cases of hoping that the world is a certain way, a way you just don't know for sure whether it is or not.

Importantly, as always with natural languages, we can force various different readings of one and the same sentence, but hopefully I have identified above some core readings of some interesting sentences involving hope, and thus captured a concept that we will now go on to see is of great religious interest.

Consider the following argument, inspired by what McKaughan (Reference McKaughan2013) calls the modern problem of faith and reason. Assume that a belief that p is justified only if it is adequately supported by argument or evidence. Premise: a belief that there is a God is not adequately supported by argument or evidence. Conclusion: a belief that there is a God is not justified.

If you are, as I believe I am, persuaded by this argument, it does not follow that you cannot centre (central parts of) your life on the proposition that there is a God. You can follow the argument where it leads, but still be rationally justified in centring your life on the proposition that there is a God in virtue of some propositional attitude other than belief. For example, from what was said earlier, you can follow the argument where it leads, and in virtue of that fail to believe that there is a God, but still hope that there is a God (cf. Pojman (Reference Pojman1986) ), and in virtue of that still centre (central parts of) your life on the proposition that there is a God.

But is merely hoping that there is a God enough to centre your life on the proposition that there is a God? As far as I can tell, the answer is yes. For example, consider the so-called ‘liberal’ Quakers. Some of them are theists, but some of them are agnostics and even atheists (cf. Boulton (Reference Boulton2006); Durham (Reference Durham2013); Cresson (Reference Cresson2014) ). Their most central religious attitude is not a matter of a particular belief in a particular God, but, importantly, their lives are nonetheless being guided by what they call ‘the inner light’; their lives are nonetheless centred on the proposition that there is a God independently of any particular belief in that proposition. Now, I am not claiming that the central religious attitude of these Quakers is hope – I have no idea to what extent and how deeply they hope that there is a God, although I suspect they would confirm it if asked – but I am saying that, if agnostic and even atheistic Quakers can centre their lives on ‘the inner light’ and the proposition that there is a God independently of any particular belief in that proposition, then people who don't believe that there is a God, but nonetheless hope that there is a God, can do so, too; some of the Quakers in question do it, so the hopers in question can do it, too.

Poston (Reference Poston2009) asks whether someone who doesn't believe that there is a God can still live a full-blown, fully committed religious life. I believe that some of the ‘liberal’ Quakers answer that question in persona. But, in fact, by simply imagining a ‘liberal’ Quaker who doesn't believe, but merely hopes that there is a God, we can see that the answer is yes. There is no reason to deny such a person her or his full-blown, fully committed religious life (see also McKaughan (Reference McKaughan2016) ).

Above I put forth the plausible hypothesis that the stronger your belief that p is, the less (linguistically or conceptually) permissible it is to hope that not-p; and, conversely, the more permissible it is to hope that not-p, the weaker your belief that p must be. In the case of believing that there is no God, but hoping that there is, this entails that the stronger your belief that there is no God is, the less permissible it is to hope that there is a God; and, conversely, the more permissible it is to hope that there is a God, the weaker your belief that there is no God should be. So, the stronger your atheism is, the less permissible it is to hope that there is a God. But, it is still only a matter of degree; it doesn't seem completely impermissible to hope that there is a God unless you transparently know or are certain that there is no God.

Now, one might think that a person who doesn't believe, but merely hopes, that there is a God must at least occasionally act as if there is a God in order to count as living a full-blown, fully committed religious life. Otherwise, one might suspect, the hope looks too weak to be sufficient. The question then is what it takes to act as if there is a God. For example, is it sufficient to occasionally pray by yourself to something even if you are not clear on what it is? I believe so, but let that be as it may, because in any case there is the deeper question: what are the conditions under which it is rational to act on one's mere hope without an accompanying belief?

Answering this question requires a complete theory of rationality. Unfortunately, I don't have such a theory, but what matters for what's to come is providing a necessary requirement for avoiding irrationality: namely, that it is not irrational to act on one's hope that p only if doing so is better for oneself or others (cf. Jordan (Reference Jordan2016), 114, for a slightly stronger requirement), where acting on one's hope that p means acting as if p is true. So, for example, if you don't believe that p, but you still hope that p, it is not irrational to act as if p only if doing so is better for yourself or others. It is worth noting that this picture is compatible with much psychological research on hope, especially its usefulness for a good life (cf. Snyder (Reference Snyder2000), although this literature often operates with one or more teleological concepts of hope that are narrower than the one I have identified above).

The problem of catching belief

I have argued that you can hope against your belief (but not against your knowledge or certainty), and that it is not irrational to act on such hope only if doing so is better for yourself or others. But Jordan (Reference Jordan2016) objects to such a picture (or at least to something very close to it) by virtue of what he calls the problem of catching belief. Roughly, Jordan argues that, if you keep acting on your religious hope that p, then typically you will come to believe that p, but that way of gaining the belief that p is unjustified, so to keep acting on your religious hope that p is unjustified, too. What's more, acting against your belief also compromises the integrity of your character in a problematic way.

Now, on its face, this argument seems open to counterexample. Assume that I initially have no belief as to whether my neighbour is a nice person (or even that I believe that she is not a nice person), but I hope that she is a nice person, so I keep acting as if she is a nice person. Then, we assume, I will typically catch the belief that she is a nice person. Now, this belief is unjustified, but that gives me no reason to stop acting as if she is a nice person. I should still act as if she is a nice person. What's more, it does not compromise the integrity of my character, at least not if I am open about the matter.

So, I believe this argument fails to refute the position I have sketched, but in order to more clearly see how so, we should first make the argument more precise. Let ‘P()’ be the permissible operator. In possible worlds semantics, P() gives you truth iff its argument is true in at least one accessible possible world. We let the operator P() operate on propositional attitudes, and in particular on the propositional attitudes of belief and hope. Let A be the propositional attitude of not believing that God exists (i.e. agnosticism or atheism), let H be the propositional attitude of hoping that God exists, and let T be the propositional attitude of believing that God exists (i.e. theism). What kind of permissibility is in play here will be further discussed below (it is one of the problems with Jordan's argument), but note for now that P(H) is true iff it is permissible to have the propositional attitude of hoping that God exists, and ~P(T) is true iff it is not permissible to have the propositional attitude of believing that God exists. Finally, assume the conclusion of the argument from justification discussed earlier: T is not justified, because the belief that God exists is not adequately supported by argument or evidence, and the propositional attitude of believing that p is justified only if the belief that p is adequately supported by argument or evidence. Then we can abstract the following valid argument from Jordan (Reference Jordan2016):

1. 1. P(H)  Premise

2. 2. H⇒T  Premise

3. 3. ~P(T)  Conclusion of the argument from justification

4. 4. P(T)  1,2, distribution principle

Here ‘H⇒T’ is not entailment, but rather the empirical claim that H typically leads to T, especially if you keep acting on H, and the distribution principle is the principle that, if P(H) and H⇒T, then P(T).Footnote ⁷ Now, 3 and 4 are a contradiction, so it seems that one of 1–3 must be false. But note that 1 and 3 are mere assumptions for the sake of argument, since the interest of hope at stake is based on not believing that there is a God due to the fact that the belief that there is a God is unsupported by argument or evidence. So that leaves us with 2, and the distribution principle.

Is 2 true? As already mentioned, it is not entailment in the sense that, if you hope that God exists, then you must eventually come to believe it (not even if you keep acting on H). Clearly, it is possible that you hope that God exists (and even act on it) but never come to believe it. But then we might say that you can hope (and keep acting on your hope) that God exists as long as you don't catch the belief that God exists. Jordan's problem with this is that he finds it unlikely that such a consistency will actually typically be the case (although it is not impossible). Jordan thus thinks that 2 is true, and hence that 1 is what's false.

One response is to flat-footedly claim that most actual people shouldn't hope and act on their hopes (against their beliefs), but that those who can resist catching the belief are permitted to do so. What's more, it seems that as long as one is open about this, and about hoping against one's belief, one's character is not problematically compromised. Consider the earlier case of (merely) believing that my neighbour is not a nice person, but hoping and acting on my hope that she is a nice person. As long as I am open about my hope (and hopeful actions) against my mere belief, my character does not seem to be compromised.

In general, I am more optimistic about avoiding catching the belief and compromising one's character than Jordan seems to be, but in any case, I think more turns on the nature of P(). The way it is rendered above to construct a valid argument is as the claim that it is permissible to have the attitude of hoping that God exists (despite not believing that God exists). But in what sense is it supposed to be permissible? In 3 and 4 we are dealing with the attitude of belief, so there the operator P might naturally be thought of as an epistemic operator. But that makes no good sense of premise 1, because epistemic permissibility does not attach well to the attitude of hope, certainly not in the same way as it does in the case of belief. True, the belief component of hope requires such epistemic permissibility, but hope also has an essential desire component, on which the same epistemic permissibility does not operate, witnessed by the structural features of hope discussed in the previous section.

In accordance with our earlier necessary requirement for when it is not irrational to act on one's hope, it is natural instead to read the operator in (1)–(4) as dealing with rational permissibility. But then the distribution principle is clearly false. Even though it is rationally permissible to have the attitude of hoping (and even act on your hope), and that typically leads to theism, it is not thereby rationally permissible to be a theist. Theism is a matter of the attitude of belief, which is aimed at truth, but hope is not such a matter of belief alone. So, when it is rationally permissible to hope (and act on your hope), that is not aimed at truth in the same way that belief is, and especially not if your hope is contrary to your beliefs, so we should not distribute rational permissibility from hope to belief. But then (4) does not follow from (1)–(2), and thus the argument is blocked.

So, in either case, whether we read the operator P() as epistemic or rational permissibility, the argument is blocked due to differences between the propositional attitudes of believing and hoping. Hoping and believing simply cut across each other.