A Science of Literature?

My latest Science Progress column just went up--it's a reaction to this intriguing proposal, by scholar Jonathan Gottschall, to remake the ailing field of literary studies with a scientific foundation. An excerpt:

Writing in the Boston Globe ideas section, Gottschall describes in detail what his science of literature would look like, something he can do because he and his colleagues have already performed some early experiments. They've crunched data comparing Western and non-Western literatures to determine if one is more sexist than the other (in the sense of constantly describing whether female characters are attractive). Result: There's no difference. They've used statistical methods to determine whether reader reactions to the personages described in great texts, like the works of Jane Austen, are completely variable or confined within a fairly small set of responses. Result: The latter.

And then there's one of the most impressive literary scientific techniques--"stylometrics," which uses computers to pore over massive texts, compare their phraseology, and thereby determine whether or not they had the same author. We all have ticks in our prose, favorite phrases and flourishes, "stylistic fingerprints" that give us away and make it possible to put literary sleuthing on a firm empirical determination, so as to really determine the authorship of contested texts.

You can read the full column here.

Tags

More like this

I spent a long time in a grad English Department. And, yes there's lots of bs going on there. And maybe more now than 50 years ago. But that's a maybe, there's always been lots of bs in literary studies.

Anyhow, the casual tossoff "ailing field" begs the question, "What do you know of it, exactly?"

On Gottschall: I haven't read him too extensively, but what I have read lacks one fundamental thing I expect from literary criticism: insight. His interpretations of Homer, for instance, are basically commonplaces which he attempts to jazz up with Darwinian interpretations.

So the basic problem is: all Gottschall's Darwinism can do for us, it seems, is put a bit more window dressing on existing interpretations (which have already been pretty capably argued for). In short, Gottschall just doesn't seem to be a very good critic.

The question to be asked is: what precisely is science going to add to literary studies? What is it going to tell us? What new questions is it going to bring to the fore? What are we going to say about Ulysses in light of what we know about the periodic table? Things don't immediately improve a whole lot if we replace the periodic table with Darwin . . . why? Because, translated to human experience, nothing Darwin wrote was particularly novel, deriving as it did from political economy.

Lastly, on the use of statistics: yes, it's been done. One of my literary mentors worked on it quite a bit and found . . . it doesn't really work. Authorship was persistently misidentified in cases where it's already known. That was 20 years ago, though. Perhaps it's gotten better, but it would have to have gotten better in terms of technique, not power.

"Stylometrics" goes back to the 1960s, when Mosteller and Wallace investigated the authorship of The Federalist Papers. It's good solid statistics.

There are aspects of literature that can be studied statistically, but to replace the whole area of literary theory by statistics sounds like a good way to kill off the subject. It really reduces the set of questions that can be answered. And most of the ones it excludes are probably the ones humanists are most interested in.

stylometrics is a new term ... thanks :-)

Hmmm. I'm more keen on the idea of science interacting with philosophy and history. PZ Myer's was discussing this in his recent Bloggingheads:

http://bloggingheads.tv/diavlogs/11305?in=00:08:00&out=00:12:54

I had the same reaction to 90's literary theory that you did. But my way of dealing with it was studying lit in the context of historical periods...

By Jon Winsor (not verified) on 24 May 2008 #permalink