Wednesday, June 24, 2009

STORIES ARE DATA COMPRESSION

Last week I watched 'Good Will Hunting' for the first time in about five or six years. This reminded me of the fact that whenever I watch an old favorite that I know well but haven't seen for a long time, I'm struck by the economy of the storytelling. The story always seems so full in my head that I'm shocked that crucial pieces are accomplished with so little screen time. For example, the whole establishment and development of Will and Skylar's relationship in 'Good Will Hunting' is told in four scenes, yet you come away with a very clear sense of who they are together. In Star Wars, the whole galaxy is set up in maybe 10 minutes.

Here's the half-baked thought of the week.
Stories are like mp3's: both rely on data compression. When you watch a good movie, they don't show you everything. Instead, they show you just enough of just the right things for your brain to automatically fill in all the pieces. It's like how an mp3 only plays the loudest sound when multiple ones are in the audio stream at once; you don't notice the others are missing for a few milliseconds because your brain automatically fills them in.

This is a good thing in movies because we don't have time to watch everything happen in a time-line, piece by piece, and it wouldn't feel like a story if we did. Here's a bold claim: If there's no data compression, no selection of salient pieces, there's no story. Or at least not a good one (think of a 4 year-old's listing of every little detail, "and then x, and then x, and then x...").

I think all this data compression is a useful strategy for dealing with the limits of human memory and processing capacity. We can only fit so much explicit information in our heads at one time, so narrative algorithms help us make the most of it.



FREE BONUS FEATURE:
I found this article on how the brain treats tools as temporary body parts. I can't remember if I've linked to this Ed Yong science-blogging fellow before, but he's really clear.

2 comments:

  1. I'm 100% sure you're wrong.

    ReplyDelete
  2. Are you referring to my Bold Claim? There-in lies the difficulty with Bold Claims; complicated issues collapsed into true-false propositions... also issue of conserving processing resources. I'd be happy to discuss it if you'd like.

    ReplyDelete

Note: Only a member of this blog may post a comment.