There is an ambiguity in what people mean when they talk about 'authorial intention' or 'intent of the author', arising from the fact that 'intention' has been used so long and in so many contexts.
(1) When people started using the term, 'intention' was a much broader term; it meant the act of disposing or orienting something to an end. Thus it included not only what one would like to get out of the action, but also the actual organization and structuring of the action. In the case of a text, saying that the meaning of the text was the intention of the author was saying that the meaning of the text is the organization and disposition, by the author, of words to mean something.
(2) However, there is a narrower sense of 'intention', which here we can just call 'intent', that sprang up at some point as an offshoot. In this case, the intent is merely what the author was aiming at, and not the whole act of aiming and firing at it. 'Intention' was an action term -- it is the orientation and force, the tendency, you actually give the arrow. 'Intent' is more subtle -- it is what you want the arrow actually to do, and, in particular, what you want it to hit.
We see the importance of distinguishing the two especially well in cases like Kierkegaard's pseudonymous works. One cannot assume that a pseudonym completely represents Kierkegaard's view; each of his pseudonyms is put forward as a persona. Thus there is an authorial intention discoverable in the text, since each text is carefully constructed, but Kierkegaard's own intent -- what he's aiming at in the first place -- is deliberately masked by the presentation of the persona and that persona's purported intent. Learning the intention for the text in Kierkegaard is usually not that difficult. You usually need some context so that, for instance, you can figure out that Kierkegaard is adapting terminology already in use, or is alluding to common philosophical topics of the time. But anyone worth their salt can do this, one text at a time, simply by studying the text, both in itself and in its context; this is effectively a study of how, and the way, in which a text means anything at all. Determining Kierkegaard's intent in writing it, on the other hand, is extraordinarily difficult, and requires a critical judgment based on familiarity with how the text fits into the whole of things that Kierkegaard writes about.
Nor is this an artifact of pseudonymity. Kant's book on Swedenborg, Dreams of a Spirit-Seer, is a very readable book: the intention is not difficult to find. It's a well-organized text that can be read without much difficulty, and is certainly much easier to read than most of Kant's other works. On the other hand, it's notoriously difficult to interpret. How can this be? Because no one quite agrees on what Kant's intent was in putting the book forward and putting it forward in the way he does, and what can be gleaned from the text and context leaves a great deal undetermined. Much can be determined about what Kant is doing, and thus we can determine much about the authorial intention; but one part of it, why, exactly, he is doing it, is elusive, and so we argue about intent. We see this as well with Hume. The most irreconcilable interpretations have been put on Hume's work, despite a lot of agreement about the texts themselves, and why? Because people differ about how skeptical his intent actually is.
The distinction is important in another way, in that skepticism about authorial intention is often conflated with skepticism about authorial intent. Intent can be very tricky; it's a matter of motives and expected consequences, and is almost always the hardest thing to determine about the intention of the text. Just as in a conversation we can understand quite well what is said but miss the point of its being said entirely, we can grasp the organization and sense of a text without being aware of why one would write such a thing at all, or why one would write such a thing in this particular way. And the problem is exacerbated by the fact that the intention of the author can go awry, and deviate massively from his intent. Just as an archer can want to hit the bull's-eye, but actual dispose and orient the bow so that it ends overshooting the target entirely, so the intent of a text can be one thing and the whole intention or disposition of the text be heading in an entirely different direction. Clearly, any approach to a text that can't recognize this is radically faulty; but this is precisely what one gets when one fails to make the distinction between intention and intent.
