One more thing on the relevance of truth to the qualities of a fictional work: take Saul Bellow’s late novel Ravelstein (2000). It quickly became apparent that the enonymous character was based on philosopher-critic Alan Bloom.
Is Ravelstein Bloom? A difficult question to answer and I’m inclined to say he is for a small in-the-know audience and not for others (there are probably indeterminate populations of readers).
But for anyone interested there is certainly a real and relevant relation between them. It matters to the literary effect of a novel set in or based on London whether the city of portrayed as it is or as it might in some fantasy be. It’s the same with people.
Ravelstein’s manners, life style and intellectual tastes are, I understand, close to those of Bloom, though some of this Bloom was quiet about in his public life.
But suppose Ravelstein’s death from AIDS had been pure invention on Bellow’s part, with enough verisimilitude to the rest of Bloom’s life to maintain the importance of the Bloom-Ravelstein relationship. This deviation would surely have a literary significance. The event would stand out in a colour different from the rest and we would reasonably ask about this particular authorial choice of a deviation.
Bellow might have chosen differently, with Ravelstein dying in a terrorist bomb attack, a car accident or by suicide. These different choices would have different meanings for the work’s interpretation. Perhaps the car accident would be a studied denial of meaning, but what if that actually happened to Bloom? Conforming to truth and deviating from it are both choices, but they are different choices, with different consequence for interpretation. Truth matters to the interpretation of fiction.