Dispatches | November 01, 2006

I recently read an article by Malcolm Gladwell in the New Yorker about a computer program and a data-based fortune teller for successful music and movies. In both cases the developers claim to use not their own opinions or critical views — or even the historical versions of such opinions and views but rather to rely on the objective details of musical pieces and movies to construct models for success.  In music, for example, the components include such elements as tonality, harmony, pitch, tempo, and so on.  The program’s inventor says that there are twenty-some “clusters” of elements that can successfully predict the relative popularity of a piece of music.  Similarly, the movie success forecaster includes such common elements of storytelling as subject, plot, character, and character groupings.

While I don’t know enough about music to have an opinion about “success clusters,” I find it amusing that the guys who’ve devised a tea-leaf reader for movies claim to have discovered that valid and interesting settings are a key component of the relative success of a movie.  Writers have known this for at least 2.5 millennia, probably longer.  In fact, all of the “criteria” of the movie prognosticator strike me as silly, not just because they are taught in creative writing 101, but because-without getting overly postmodern about it — even the language they use is either too self-defining or vague to make much sense.  What is a “strong character,” after all?  One could come up with a number of contradictory answers, ranging from the perspective of a purely aesthetic viewpoint, which might label as “strong” a well-rendered but otherwise “weak” character.  Is Woody Allen in his most neurotic roles a strong or weak character?  And what about when he plays a less than charming but otherwise convincing neurotic character whose movie persona is getting a little worn at the edges.  And so on.

The elements of fiction and film and even the components of those elements are simply too complicated and contingent not just upon each other but upon the relative emphasis of each other to be so easily defined.  If you are Claude Levi-Strauss with twenty years to work out a thousand schemes, you might get a start, but like the great anthropologist you’d probably grow more skeptical about your structures the more you studied them.

Every time I find myself having outbursts with my fellow readers at this magazine — “I’m sick of death stories.  No more dying fathers with loving daughters who do, finally, connect with the old cob!  Let him die!”  The next thing that happens is that we get the best story we’ve gotten in six months about exactly that subject.

If I have learned one thing as an editor it is that pre-set models do not predict artistic success or failure.  Almost every truly great story that we have published in this magazine’s history violates some prejudice or model about practical or even theoretical aesthetics.  I’ve learned the hard way.  Don’t decide in advance, based on what you imagine to be a pattern for failure.  Decide, instead, on the whole story and its effect.