Robert Audi has written very concisely on the relation of belief and intention. In particular he sets forth an interesting thesis, viz., that hoping and intending are distinct, and that the important difference lies in the degree of believed likelihood of the thing intended or hoped for. If a subject believes that the event undertaken is likely or probable then that subject may be said to intend to realize the occurrence of that event by performing some action. This way of putting it may at once appear too precise, by distinguishing events and actions, as well as too vague for not spelling out, among other things, what is meant by "perform." Suppose, then, we consider what Audi, himself, says on the matter:

For if I believed the chances of my helping the motorist by driving to the station were very slim, I could be properly said only to hope to help him by doing so. To distinguish intending to bring about φ by A-ing from merely hoping to bring about φ by A-ing, we need to require that S at least believe her A-ing will be a probable way to achieve φ. ( Action, Intention, and Reason Cornell University Press. 1993. p. 57).

Later Audi, somewhat mysteriously, adds:

This is not to deny that one can both intend and hope to A. (op. cit. p. 58)

If it is the case that my believing φ is probable, upon A-ing, involves intention and not hope, merely, then how is it that I can both hope and intend to A? In particular how is this possible without both believing and not believing that A is probable? One might care to attempt to distribute the intending to the A-ing and the hoping to the φing, but this is not what Audi appears to be saying. What Audi says (ibid) is that such a thing is possible when the subject is "somewhat uncertain" of success. But, then, the agent doesn't believe A is probable, and so, by his earlier statement, cannot intend it. Towards the end of his discussion in this chapter, however, Audi does entertain a case where the sort of distribution I've suggested appears to be regarded with some plausibility: A girl, Sarah, is target shooting and a fox is seen about fifty feet behind the target. She intentionally aims at the rabbit, says that she is trying to hit the rabbit but, knowing that she is a poor shot, has little expectation of hitting her target. However, she does hit her target(op. cit. p. 71). But, according to Audi, "Sarah shoots at the fox with the hope of hitting it; she does not have the intention of hitting it. Here, then, it seems that intention applies to the aiming and firing, but hope applies to hitting the intended target, i.e., the fox. This is not a new argument by Audi, but, rather, a restatement of his position. With this in mind I will not introduce any argument that I have not already introduced, but I would suggest to the reader to consider, first, that what we are supposed to accept, on the basis of his previous arguments, is that Sarah and intentionally fired at the fox; she said that that shooting the fox is what she was trying to do; but, when she hits the fox we are to believe that the rules of language prevent us from saying that she intended to shoot the fox. Second, we need to consider the motivation aside from linguistic intuitions for making this seemingly counterintuitive move. What is to be gained? Just as when philosophers go about introducing a technical vocabulary or when they use old words for new tasks they are expected to justify themselves, so too, one needs to ask what the gain is, not in distinguishing hope and intention, but in accepting what most native speakers of the language would thoughtfully reject. Audi's "out" is to say that this is a discovery of how natural language works and that it is not counterintuitive; but aside from the reasons I've given, and will give, for rejecting Audi's position, there is I think warrant for denying the acceptability of any linguistic judgment that what Sarah did she did not do intentionally. Let us set aside the matter of linguistic intuitions, for the time being. Intuitions differ and it is neither intuition or restatement of Audi's conclusions in the context of further examples that need concern us. Audi may have some simple and persuasive answers. Instead of pursuing this issue, let us consider the general claim that hoping and intending are to be distinguished in the way Audi proposes. To begin with, it seems there are far more simple reasons for distinguishing the two concepts, even though, as I shall attempt to establish, hoping does often imply intention.

'Hope' and 'intend' differ in the semantics of tense. Thus it makes sense to say, looking backward into the past, 'I hope she left'; but, nothing comparable can be said using 'intend', such as 'I intend to have left', unless an adverbial is adjoined creating expectation, such as 'I intend to have left by then', where 'then' is at some time after the utterance. In addition, there is the feature of person, which, also, can be used to distinguish these concepts. I may say, 'I hope she will leave'; but, I cannot say 'I intend that she will leave'. I may say 'I intend for her to leave'. But as long as the verb contained in the complementizer, 'for', takes a complete functional complex containing a matrix nonpassive verb of action a further intention, her's, is introduced. The point is that I may hope someone leaves but not in the sense I intend someone to leave. Moreover, even on those occasions where 'I intend for her to leave' is not degraded, there is an implicit further action, 'By A-ing, I intend for her to leave'. This is not so in the case of 'hope'. Indeed, more plausible than the proposal that I cannot intend what I merely hope for is the idea that a hope is an intention where the expectation of success is attenuated, but to make my case against Audi, this is an unnecessary, albeit, plausible manuever.

Audi's claim trades on degrees of belief in the success of the action in determining its status as something intended or hoped for. Thus, he speaks of what is "probable," "chances...were very slim," "likelihood," as well as "somewhat uncertain" in characterizing the belief of the agent with respect to his or her success. Doing so carries with it certain dialectical dangers, and here I mean in the Platonic and not the Hegelian sense. To see one of them, at least, consider that while it makes sense to say "I began to hope that..." it makes little, if any, sense to say "I began to intend that'. Now I may intend something, but upon encountering difficulty come to merely hope; as in such a case where I at first intend to lift a large stone I soon, having little success, come to hope, merely, that my subsequent efforts will be successful; but here intending and hoping are compatible to begin with, and it may be more correct to say, only, that I believed I could, and had the intention to move the stone, but now, even though it is my intention the best I can hope for is that on my next effort I move the stone. Still, there is a better reason for rejecting Audi's proposal, one having to do with the asymmetry we just pointed out between hoping, which may come about, and intending which is never a gradual thing. Suppose I am a tailgunner during WWII. I get the target in my sights; I believe I am nearly out of ammunition and the plane I am in is low on fuel. I doubt that I can down the enemy plane. But as I fire I come to realize that I am not low on ammunition and, further, that I have a spare fuel tank. Not only this, there is smoke billowing from the enemy plane. Are we now prepared to say that his hope has become an intention; or, are we to say that whereas I had no intention to down the plane that I acquire an intention I did not previously have, even though I am doing nothing different? Audi's claim appears to force this on us, and so, perhaps, ought to be rejected. There are other problems with Audi's account.

Audi has said that in order to be said to intend, and not merely hope, I must believe in the likelihood of success at what I am about to do etc. But this neglects to take into account those occasions where I am absolutely uncertain either of the likelihood or the unlikelihood of success; I simply have no belief in this regard. If Audi is right, then it must be the case that I can only hope; but it seems clear that even when I have no such opinion I can do something with intent. In addition, to this there is a further consideration. Suppose I am at a party and Mike Tyson challenges me to arm wrestle. I have no such belief that I can win. We get down to business. Just as the signal calls for action, Mike sees a girl walk in the door wearing nothing. Quickly, I use all the leverage I can muster, even though I am convinced that most likely I will not win. But as it turns out I beat Mike! Now do I want to say, when asked how I managed such a thing, that I didn't do it intentionally? When I do something successfully on a hope, retrospectively the attribution of intentionality is almost always a certainty; but how can this be so, given Audi's account? One other thing, before drawing this to a conclusion. Anscombe, rightly, made very heavy weather out of the locution 'doing x with the intention y'. Isn't it reasonable to say that whenever I say 'By doing A I hoped for B' that part of what I mean is, "I intend to do B"? If so, the link between hoping and intending is, once again, demonstrably closer than Audi would most likely care to maintain, assuming of course I am reading him correctly. There are other problems. Perhaps Audi has an answer for all these seeming difficulties. He is a very clever philosopher, whose words are very closely tied together; perhaps too much so. There is finer line than many think, particularly among linguists, between prescriptive and descriptive grammar (as well as natural language semantics).