Forgot to link to Emily's last week post, which (nested in a discussion of how much characters can change) contains the kind of description of the ontology of fictional works that one finds all over the place when one attends to what actual writers say about their craft
Mark Ohm and I develop our theory of fictions as thought experiments in part because of our reading of how writers of fiction describe their craft (downloadable here; I should note that Mary Sirridge and Eva Dadlez made similar suggestions, which we take ourselves to be building on, and that we have another partially completed paper which I'll be happy to send anyone who is interested). Our theory hasn't gone over well either with philosophers or narratologists. As far as I can tell, there are four interconnected widely accepted presuppositions which pretty strongly militate against it, the first related to Emily's post.
- Anti-intentionalism- The original "intentional fallacy" concerned whether an author's interpretation of her own art-work was in any way authoritative. People who thought that it was not were anti-intentionalists. This was a good professional move for academic critics, who ran a nice shop during the heydey of "theory." But it always smacked a little bit of psychic revenge of failed artists, a kind of revenge made most explicit in so-called "reader-response theory," where the real creative force is not the author, but rather the trained reader, e.g. the literature professor. Again, nobody much believes this any more, but we're all a little bit hungover. If we are no longer the sole decoders of a text's meaning, general questions about the ontology of art-forms and genres are still thought of as best left to the professionals.
- Focus on Marginal Cases- Artists are very creative at playing with, and rebelling against, prior conceptions of what their given art-form is supposed to be. They're very good at seeing just how much they can get away with. But then if you think that the purpose of the philosophy of art is to come up with necessary and sufficient conditions that encompass all art (or all of some form or genre), then the weirdest, most out there, experimental stuff ends up seeming central, precisely because that stuff is often designed as counterexamples by the artists in question. The fact that the necessary and sufficient conditions game is ultimately a mug's game doesn't change the centrality of marginal cases, in part because familiarity with such cases is a kind of badge of hipness and not liking them of philistinism. Because of this, any theory of fiction that takes seriously what Steven King says about his writing process is going to be dismissed because it doesn't capture, for example, Gilbert Sorrentino's late novels, or Kathy Acker, or William S. Burroughs, or Finnegan's Wake, etc. etc. etc. However, the escape from the mug's game is to realize that the task of the ontologist of art should be to give necessary and sufficient conditions for the subclass of art-works that are genetically necessary for all of the art-works in the genre in question. Experimental fiction is parasitic on the kind of thing Steven King does. To really explain it, you have to first have a theory of what Steven King is up to. But people who think you have to define the whole genre at one go are constitutively unable to do this.
- Marginalization of Normative Questions- This is probably more widespread among narratologists and metaphysicians who write about fictionalism than in analytic philosophy of art, though it's there too. Noel Carroll likes to contrast classificatory theories of art with commendatory ones, arguing that the former is primary and that the latter is a part (and not the major one) of criticism. But artists do not make this division in their own thinking about art. They are proper Wittgensteinians who recognize that questions about classification in art can only be answered by first attending to the features of prototypical examples, and that this is necessary in part because there is indeterminacy between something being a bad example of something and not being an example of something. Epistemically, can't know what a genre novel is unless you have some kind of idea of what makes a good genre novel a good genre novel. And an answer to the metaphysical status of the being a genre novel will also be dependent upon the properties manifest in the good ones.
- Marginalization of Actual Truth of Artworks- In continental circles the very idea that there is an observer independent truth is often taken to be jejeune. In analytic circles "fictionalism" names positions that take the truth predicate in a given discourse to be radically defective. Both positions are a radical departure from traditional aesthetics, which concerned itself with how actual truths are contained in fictional texts. Both positions thus infect our thinking about fictionality. If, on the other hand, fictionality just is the kind of counterfactuality relevant to thought experiments, the traditional view is restored. I don't know what this ends up doing to the fictionalist project.
As far as I can tell, the above four properties are so widespread anywhere that narrative is academically studied. I wish that I had more time to work on the philosophy of art, but they do strike me as worthy adversaries and I hope that I have time to engage more seriously with them some day.