Tuesday, February 16, 2010

Parts Fitting Together

Every event that occurs in a novel sheds light on all the novel's other events. This annoys Thomas Kurton, the scientist character in Generosity by Richard Powers --

“... fiction's perpetual mistaking of correlation for causation drives Kurton nuts. Even Camus can't help deploying bits of his characters' histories as if they explained all subsequent behaviors and beliefs.”

What Kurton sees as spurious correlation, Marshall Gregory in Shaped by Stories sees as a big part of fiction's appeal --

“Too much of my life – and yours – gives us the sense that the parts don't fit together. In contrast, the analogies of fiction and other narratives provide us with points of comparison that let us see what greater focus, organization, and unity our lives might possess.”

It's in our nature to look for patterns; when it comes to spotting connections, a false positive is likely to be less hazardous than a false negative... at least until you become clinically paranoid...

How much of our enjoyment of fiction stems from its providing more opportunities for connection-spotting than our real lives do?
Posted by James Warner at 2/16/2010 06:30:00 PM

2 comments:

  1. I have not read Powers's GENEROSITY, but the claim of his character, Thomas Kurton, that fiction makes the perpetual mistake of conflating correlation and causation seems, on the face of it, to make the error of attributing to fiction what is in fact a universal error of human cognition generally. Fiction REPRESENTS the tendency of human beings to conflate correlation and causation, but fiction does not invent this error. Cognitive science research, in addition to recent work in evolutionary psychology, suggests very powerfully that one evolved feature of human cognition is the over-attribution of causal agency to events that have no agents behind them at all. The claim is that this tendency of human minds conferred an adaptive advantage on human beings on the Pleistocene plains in the following way: it simply was safer to conclude that any sudden changes in the environment--sudden changes in movement, sound, or light--were caused by predators and to take evasive measures of escape (running away, in other words) rather than to stick around for empirical testing, philosophical theorizing, or holding suspended judgment. Those who ran away had a better chance of sending their genes forward in time than those who were more scientific or philosophical in their response to environmental changes. The rest of us, to make a long story short, are the heirs of this tendency, which is much less advantageous for us now than it was then, but it goes a long way to explain why human beings in general tend to conflate correlation and causation.

    It remains the case, however, that for human beings meaning and pattern are inextricably entangled. For us, there can be no meaning that is not pattern; pattern is by definition the contrary of chaos, and fiction is our most comprehensive and ubiquitous way--not our only way, but our most common way--of generating the patterns that yield meaning. To take fiction to task for doing this, if, indeed, that is what Powers's character is doing, misses the point that we cannot help doing this as part of human cognition, and that fiction only represents this tendency but has not created it.

    I appreciate the reference to my book in your discussion of Powers's point.

    Marshall Gregory
    author of SHAPED BY STORIES
    http://marshallgregory.com
    ReplyDelete
  2. Thanks for these thoughts! I’m enjoying your book.

    I guess Powers’s idea is that Kurton is comparing science to literature. (Or maybe, in effect, he is comparing science to everything apart from science.)

    Scientific methodology can be seen as a set of techniques to overcome our natural tendency to conflate correlation and causation.

    However scientific methodology is of limited use for “making sense” of our individual lives – for this we still seem to need art or religion. Big topics. More later.
    ReplyDelete

Subscribe to: Post Comments (Atom)