First, a disclaimer: if anybody ever tells you that an interpretation of a work is invalid because "that's not what the author intended," throw them off an overpass; they're better off coming back as a lobster.
Second, the opposite view is equally dogmatic and flawed. This is the harder point to make and I'll start by extolling some of the virtues of this view before kicking it down a notch.. A peer of mine once brilliantly said about analyzing Lolita "If we were to bring Nabokov back from the dead to ask him what he intended, he'd probably just pop out of his grave and start lying." This is quite true on many levels; on a quite literal level, there's no way you can guarantee an honest statement of intent; just because an author is saying it as opposed to writing it or phrasing it in a direct manner rather than an indirect manner (i.e. telling a story, what a concept!) that doesn't make it any more of a foundation since it isn't verifiable.
I can take this further by saying that intent simply isn't verifiable. It has no objective existence; you can tell your best friend what your intent about writing a story is and then lie to everybody else, but what you said to your best friend doesn't have any objectively verifiable basis; it only exists in your head! This seems somewhat banal, but it's a less metaphyiscal way of stating the fact that subjective ideas can only be mapped, they cannot be verified.* So it would seem that talking about intent is silly; it doesn't objectively "exist", it's something we can only infer in some unquantifiable way by privileging someone's "direct" statements or by looking at the biography of an author and using our idea of what their life was like to suggest how some events may have informed their work.
But this argument seems a bit weak. Is anything in literary interpretation ever objectively verifiable? Answering "yes" to that question seems absurd (if you think otherwise, feel free to speak up; I'm just saying that right now I really don't see any good argument.) So what are we doing then in analyzing literature? We're making sense of its impact and its relationship to the world by constructing a narrative of our own. While others may have strongly disagreed with me on the following point, I'll still make it: narratives are fundamentally about people and primarily reflect our existence as social creatures. When we read a novel or watch a TV show, we socially construct characters from the words on the page or the actors on the screen despite what's most likely a complete poverty of information; we may not know the entirety of their (imaginary) life experience, but just like a famous author, we construct a narrative from a limited biography (and body of canonical works; After the Quake isn't representative of Haruki Murakami, I swear!) and create a being with a life of its own.
One of the most important factors in our social construction of human beings is the idea of intentionality. We infer intentions when it comes to everything people do; perhaps as a way of masking all of the noisy details and deviations of a person's behavior or maybe because we really can learn who to trust and who not to.** Without ascribing intentions, we can't construct a picture of a person or empathize with someone; thus the reason why a heroic or tragic story on the news will captivate us but a statistic can only glance off the (somewhat) rational surface of our minds. The same goes for literature; not just in constructing characters but also in how we construct the story as a whole. There is always a narrative voice telling the story, however passive or indirect; and just as we listen closely to a personal story told to a trusted friend, we "listen" to a piece of literature in order to figure out what to make of the story it contains.
To put it another way, every story always has a storyteller; implicit or otherwise. Our idea of the storyteller is informed and constrained by many things; social norms, the conventions of genre, the idea that they're trying to entertain us (think about the last mystery/thriller you saw; you have a pretty good idea of why the most obvious suspect wasn't the traitor/murderer/villain,) and so forth. A former professor of mine rightfully responded to this point by bringing up the point that many of these ideas are different than a mere statement of intent because we can create a more objectively verifiable case about things like social customs and generic conventions. I agree with them insofar that making blanket statements about what someone intended isn't a good way to make an argument; our idea of someone's intention is as subjective as the author's own intent and an argument requires definite common ground. I still find it necessary, however, to acknowledge that we ultimately create an interpretation of the story that is itself a narrative and in order to do that we create an intent behind the storytelling. But I digress.
All storytellers are actually implicit; even if we know the author or are listening to an orator right in front of us, we don't know every last detail of that person's life, we've constructed a simplified version in order to relate them to the story that they're telling. The point is that this implicit storyteller informs our own effort to make sense of the story and in order to let such a storyteller inform us about the story we endow them with intentionality.*** Storytelling is fundamentally a social enterprise; it depicts complex social relationships (to the point that we can hate Nina Meyers for killing Teri Bauer) with very little information and it makes an impression on us by allowing us to read into how the story is told. From an evolutionary (and completely hypothetical) perspective, stories began as a means of communication about complex social relationships and so we're always scrutinizing the storyteller who must have borne witness to the events and who must have some motive behind telling us the story. But I (once again) digress.
We can't read a story without inferring something about the storyteller's intent and we can't have any understanding of a story without imagining a common ground between ourselves and the author. This is why Beyonce's songs will never do it for me. She talks about guys leaving her in so many of her songs (Single Ladies, Why Don't You Love Me, Say My Name...) but she's been dating Jay-Z for practically all of her adult life and is now married to him. I can't see veyr much authenticity in what she's saying. Jay-Z never left her and she started dating him when she was 21, so it seems unlikely that she has much to be going off of.
People have asked me why I continue to enjoy Lady Gaga despite the seeming artificiality of her songs about partying and seduction (she was a workaholic in school and is more of one now.) That's a good question. I suppose that most of Lady Gaga's songs in The Fame seemed somewhat reflexive and ironic to me. I know that's not a very sophisticated critique, but the difference is that Beyonce, for all of her singing and dancing talent and her picture perfect looks, just comes off to me as too damned earnest for her to have any implicit commentary in her lyrics. Of course, this is just the Beyonce that I've imagined for myself; the truth is that I don't know the first thing about her, and neither does anybody else outside of her personal life.
*My amateur knowledge of phenomenology and semiotics causes me to think that this statement suggests the respective roles of both schools. Semiotics is the study of mappings whereas phenomenology is the study of where mappings come from.
**To be fair, narratives were much more reliable back in the Pleistocene when the world wasn't so damned interconnected. I take no credit for this idea; see The Black Swan by Nassim Taleb.
***A stronger argument that I'm tempted to make is that this idea of the implicit storyteller is necessary to create any sort of narrative context; generic conventions, social customs and even the specific language/dialect that we're reading in is some subjective phenomena independent of the physical text (how could it be a physical property of the text?) that we see as an act of communication between teller and listener. To put this in a more familiar perspective, Peirce concluded that all signs require three elements: a signifier, a singified and an interpretant. Without a social construction that binds the storyteller (imagined or not) and the listener, there's no code with which to link sign and signified.