Friday, June 10, 2011

Computer analysis of fiction

I'm wondering if maybe this kind of misses the point of reading fiction.


green_knight said...

It's not literary analysis as we know it, but most writers I know do similar things. I've marked up texts with the balance between dialogue/description/action/internalisation because I felt that someone else's writing worked much better than mine; and things like asking how much screentime a character gets etc are common questions that help me to understand how an author is creating a certain efect. I have a novel with a main character who doesn't turn up until 2/3rds into the book; so I looked at how other authors kept people in the reader's mind without their physical presence.

All Moretti is doing is codyfing it and working on a much larger scale. The conclusions that he draws will still have to come out of his own understanding.

Edittorrent said...

Something about this quantitative analysis makes me think of mid-century structuralists trying to escape the binary analysis. Not sure it's entirely workable. I mean, look at the way he references Aristotle and Forster before dismissing their models. But why does Horatio's consistent presence mean that he's neither a protagonist nor a secondary character, neither flat nor round? Neither character classification system points to time on the page, but rather to impact on the page. Different issues altogether.


Adrian. said...

The best "network theory plot analysis" ever.

It's pretty geeky, but I found that when I did something similar for the plot of my work in progress (a mystery), I uncovered and fixed several plotting flaws.

Edittorrent said...

I think it's interesting, but I think all sorts of things are interesting. :) I wonder if it's more useful for writers than readers?

In the end, there's magic. I can't get past that-- sometimes writers just have magic on the page, and we can analyze it, but we can't replicate it.