What are we talking about when we talk about Sherlock Holmes or Santa Claus? Something that doesn't exist?
Why did human beings develop traditions of storytelling? Of course, any answer to this question is going to be speculative. But it might be reasonable to assume that the capacity for imagination is adaptive (I need to be able to predict what is going to happen as a result of different courses of action), and that engagement with fictions helps to hone the relevant skills. This is, I believe, more or less Gregory Currie's view, and I think it's an entirely plausible one.
Still, this is completely unsatisfying as an answer to why I get so passionate about, say, In Search of Lost Time. The evolutionary account is far too general. It doesn't even really explain why I like novels, let alone (say) first-person novels, let alone (say) In Search of Lost Time. So what do specific fictions do for us?
Most of us are brought up, I think, to answer that question by looking at content. We think that literature is valuable if we learn something from it: we learn what went on in, say, turn of the century Paris (propositional knowledge). We learn what it feels like to be someone like Proust's main character (knowledge by acquaintance). We learn to be more like this character (emulation).
There may be something to these ideas, but they raise all kinds of problems (in part for reasons mentioned by Neil in an earlier post) and above all are monstrously limited. If I wanted knowledge of turn of the century Paris, I could find out more reliably some other way. Do we really go to literary artworks for that? (I'll return to knowledge by acquaintance in a moment.)
Rather than resorting automatically to content, I suggest that we consider two other dimensions: (1) literary form; (2) the process of engagement with fictions. First, literature can provide us with formal models of how best to live, not in what a given work puts on display but in how it presents it. According to the life-as-literature theorists (Alexander Nehamas, and--very differently--Charles Taylor and Alasdair MacIntyre), we can look to well-crafted stories for inspiration for the crafting of our own lives.
Notice that the content of the fictional lives is, for these purposes, no longer important. I can substitute my own content without loss. Notice too that the very pleasure we derive (what Neil called the “this-is-awesome” feeling) is the same satisfaction we stand to gain from crafting our own lives in an aesthetically pleasing way. So here the expansion of formal imagination is at least related to, if not necessarily responsible for, our aesthetic enjoyment.
Now via the process of engagement with fictions, I argue, we stand to hone our skills. This is entirely different from learning facts, and also from gaining a new set of values. Some of the pleasure we gain from an intricately-crafted fiction has to do with the successful deployment of some of our cognitive capacities. Here again pleasure is bound up with longer-term gain (though I'm not making an evolutionary argument).
Some--like Martha Nussbaum--might say that these capacities are, in an ideal case, my moral capacities. I fine-tune my moral skills, the argument goes, by engagement with fictional works. (Richard Rorty, and a caller on the show, would add that by empathizing with characters who are not like myself, I learn to fold all of humankind under the rubric “us,” rather than designating some types of people as “them.”)
Is this true? Well, even if it were, it would be unfortunate to consider this the function (or value) of literature, as Nussbaum often appears to. Further, it's not at all clear that empathy with fictional characters leads to empathy with real people. (Conversely, it's not at all clear that empathy with vicious fictional characters leads to vice. When I watch a mafia movie, I briefly take on the values of the mafiosi in imagination; but I do not go home planning a career in organized crime. And surely I don't end up feeling the pain of pedophiles after reading Lolita. If so, maybe it should be banned!)
Nor is it clear that empathy, in the sense of getting inside someone's head and learning what causes them pain, is guaranteed to yield altruism: this skill is of course extremely useful for sadists and torturers (a point made by Richard Posner in response to Nussbaum).
It seems to me that we should think of fictions as a catalyst for a process which may lead to an increase in altruism, but which may just as well lead to an increase in other-sacrificing perfectionism, or again to morally neutral change, or indeed no change at all. (Alexander Nehamas writes about this unpredictability in his recent work on Beauty.) When we engage with a great work of literature--where the stakes are high, where everything conspires to yield powerful effects--the pieces that compose us are shaken up; we may come to see these pieces more clearly; we may come to imagine new ways of organizing them; we may seek to change some, by bringing them into line with others (second-order and first-order desires, for instance); and we may hone skills necessary to these various ends. What a fiction does for us depends not just on what kind of work it is but also on what we choose or manage to do with it. If you don't lift the weights, the weights are not to blame for your lack of muscle growth.