Bouncing off yesterday's post about Orwell, this from Jane Smiley:
Every piece of writing puts forth some logical argument and some theory of cause and effect for the simple reason that words, especially prose words, are sequential. The author and the reader both know that if the author doesn’t provide the logic, the reader will. But the logic of events and people as they exist in the world isn’t self-evident, and narrators of fiction and narrators of nonfiction have different ways of putting together their logical systems.
The promise of nonfiction is that it is accurate, and therefore, like an archeological site, incomplete—here are the stone walls, here is part of a mosaic, here are two goblets. My theory concerns what these objects might mean, how they might be connected to an earthquake for which there is evidence, but I cannot go too far toward completeness or the reader, who might otherwise enjoy my narrative, will cease to be willing to suspend disbelief in its accuracy.
But the history of literature shows that listeners and readers want to know not only what happened, but also how it looked, sounded, smelled, felt, and also what it meant then and what it means now. They want to know but also to experience, and therefore they seek completeness, and so they willingly suspend disbelief in fiction (The Odyssey, the Book of Genesis, Waverley, Flashman).
I am much taken with Smiley's idea that nonfiction, that is, nonfiction rigorous in its adherence to only what can be documented, measured, accurately recorded, is inherently incomplete and we turn to fiction for theories of the missing parts.