Thursday, February 23, 2017

Plot -- the Math

According to who you read on the matter, there are only a very few basic plotlines in most of literature. For what it's worth, of the groupings at that link I kind of favor the "Seven basic plots" entry. The one-basic or three-basic entries look kind of generalized to me -- by the time you fuzz things up enough to say there is only one basic plot in all of literature you've loosened your definitions enough that you really can make anything sound the same as anything else.

Some mathematicians at the University of Vermont analyzed the words in more than 1,300 works of fiction available at Project Gutenberg and came up with six basic story arcs based on the kinds of words used in different parts of the book. About 85 percent of the works studied fit into one of these six arc patterns, described in the first graphic in this story at Scientific American. By graphing the relative happiness or sadness of the words in the story, the arcs showed up.

Another group of researchers, working in Poland, found that sentence lengths in 113 books repeated in a fractal pattern. That is, they repeated on a larger scale each time when graphed from beginning to end. More stream-of-consciousness novels like Finnegans Wake have extreme repeating patterns, but more traditional narratives show more moderate patterns. The Polish study could represent the first time anything about Finnegans Wake has made a lick of sense to anyone other than author James Joyce himself (and the jury's out on what he knew).

Analysis such as this is possible with computers that can scan and sort massive amounts of data, like the word counts of more than a thousand novels. Does it show anything we didn't already know or suspect? Probably not, as the idea of the basic types of plot is not a new one. Authors who gravitate to certain kinds of experiences or views of the human condition might work with just one or two of the different emotional arcs most of the time. In the end, it's not as if literature can be placed on some kind of graphic scale as the stolid J. Evans Pritchard would have had us believe if he had been real. But as one of the Vermont mathematicians pointed out, the "tons of data" generated by the Human Genome Project has begun to help us understand ourselves. Perhaps the tons of data in this project, he says, may help us to understand stories.

Sounds like a good idea, and I think the data will help us answer a lot of questions about stories we tell, why we tell them and why we seem to come back to the same arcs over and over again. But I bet there's one question it won't answer: What made us able to understand these things before we had the computers and the math to do studies like this? Does the arithmetic support the intuition? Excellent! But from whence came the intuition before the arithmetic was supplied?

It's the kind of thing you might have to write a novel to explain.

No comments: