Sympathy and Insight in Aristotle'sPoetics

Abstract

Usually when we feel we have learned something from a novel or play, it is the story itself that strikes us as illuminating, and in a direct way that needs no help from an outside authority. Anna Karenina does not offer a mere conjecture about how dangerous it can be to fall in love, so that we now need to verify that suggestion by asking an expert. Nor do we feel we are relying on Leo Tolstoy as our authority on love, in the way you might rely on a doctor's authority on a question of health. It is the story itself that seems to teach us something. It draws us in, and by the time we get to the end of it, it is as though what we have been through is not just a reading experience but a series of events encountered at first hand, from which we have emerged wiser than we were before. This is philosophically puzzling, because it suggests that we learn from fiction in something like the way we learn directly from life. In the latter case the learning experience is perceptual: things happen to us and we take them in, thereby discovering what the world is like. But in the fictional case the sequence of events we encounter as we read has been invented by the writer, so even if we feel as though we are experiencing these events at first hand, we are not really perceiving them and therefore not acquiring perceptual knowledge. A natural response to this is to say that we learn from fiction only insofar as it serves as a vehicle by which writers pass on what they have learned through experience or some other means. But then we are confronted by a dilemma. For either we already know, as we read, that what is being conveyed in the writing is true, in which case it teaches us nothing, or we do not know it is true, in which case we have no way of identifying it as the truth, so we still learn nothing.1 It appears that the only way past this problemcall it the problem of verificationis either to say that we accept what the writing conveys on trust or else treat it as conjectural, hence something that remains in need of confirmation. But on neither of these options does the writing stand by itself as a source of knowledge, so accepting either option means abandoning our original intuition that fiction can be a self-sufficient source of knowledge. There are those who will simply accept this consequence. According to them, when we learn from fiction it must be either because we have picked up knowledge which the writer has, intentionally or otherwise, transferred into his writing or because we have formed a conjecture from the writing and gone on to confirm it via a further source.2 In the latter case, however, we have not really learned from the writing itself but from that further source. So if we follow this line of thinking, then whenever we genuinely learn from fiction, it must happen according to the first route: we acquire knowledge from fiction when it serves as a conduit for what the writer already knows. I will call this the conduit view. A defender of the conduit view does not have to say that when we learn from a piece of fiction by accepting a propositional truth implicit in the work or by taking to heart an accurate description of some aspect of the real world that the writer implicitly asserted the proposition in question or presented the description as accurately illustrating an aspect of the real world. The conduit view only requires that in learning from the work we treat the relevant proposition or description as if the writer were asserting something true or illustrating an actual feature of the world. We learn from the writing, on this view, when the proposition or description