Sunday Salon: Quantifying Literature

The New York Times published an article which absolutely fascinated me this weekend. Large-scale quantitative analysis and literature are not often mentioned in the same system. Recent advantages in technology, particularly in database capabilities, are shaping a new landscape in the study of literature. Imagine if you could run a search for the word "humane" in every book ever published, or just in every book published by Irish men. Searches such as these and the subsequent analysis can offer great insight into literary and philosophical studies.

The current study revolves around Victorian literature. Two researchers are now studying the 1,681,161 books which were published in Britain around the 19th century. They have access to the full texts of every single one of these books. Impressive right?

Their starting point was an earlier work that focused on the written word as an entry point into the era: Walter E. Houghton’s “Victorian Frame of Mind, 1830-1870,” a landmark book published in 1957 that has shaped generations of scholarship, even as its conclusions have been challenged. Mr. Houghton sought to capture what he called a “general sense” of how middle- and upper-class Victorians thought, partly by closely reading scores of texts written during the era and methodically counting how many times certain words appeared. The increasing use of “hope,” “light” and “sunlight,” for instance, was interpreted as a sign of the Victorians’ increasing optimism. 

This sort of quantifiable information can guide scholars' insights into a culture by analyzing the prevalence and use of key terms. It also offers an opportunity for lesser-known works to come to the forefront as they are included in these larger databases.

The worries are many of course. The size requirement for servers is of course rather astronomical, meaning that larger corporations, primarily Google, have to sponsor these sorts of projects. Then of course we have the question of validity: can you extrapolate hypotheses about a culture from an analysis of their literature? Finally, some wonder if this foray into a more mathy-type analysis of literature will negatively affect or diminish the importance of the typical, more subjective form of literary analysis.

Not to worry, say those who embrace the new methods. There is no need to pit computation against interpretation. If anything, Ms. Jenkins argues, large-scale, quantitative research is likely to highlight “the importance and the value of close reading; the detailed, imaginative, heightened engagement with words, paragraphs and lines of verse.

“Close reading,” she continued, “will become even more crucial in a world in which we can, potentially, read every word of Victorian writing ever published.” 

I can't quite decide where I fall on the issue. I'm definitely of the school of thought which believes that much of what literature does is not easily quantifiable. The words and stories touch my soul and challenge my mind in a way that applying statistical analysis will not affect. But can it only be one way or the other?

What do you guys think?