Good news, English majors! Computers can assess literature now. Your services no longer required.
No, not really. But a fascinating new study from the University of Vermont reveals that computers can analyze literature and, by way of sophisticated data-mining techniques, draw some very interesting conclusions.
RELATED: How Candidates Use Storytelling to Win Elections
To wit: According to the supercomputers at Vermont's Computational Story Lab, there are only six basic kinds of emotional arcs in storytelling. While stories themselves can be infinitely more complicated, these six patterns form the building blocks of all narratives.
The researchers arrived at this conclusion by feeding the full text of more than 1,700 stories, novel and plays into a complex data-mining computer program. The computers then scanned the text, page by page, to perform a kind of "sentiment analysis" using natural language processing. It gets complicated: The system involves industrial-grade cognitive computing terms like optimization, linear decomposition, supervised learning and filtered subsets.
A very helpful followup report over at MIT Technology Review breaks down the essentials. The Vermont program essentially looked for words that have a positive or negative emotional connotation, then monitored the emotional valence ratios as the story progressed. Check out the very fun interactive online charts, and you'll find positive words like "healthy," "wonderful" and "bright" jostling for space with negative words like "foul," "disease" or "fear."
RELATED: Story Time! A.I. Generates Interactive Fiction
As to the results: The six basic emotional arcs in storytelling are as follows, according to the MIT report:
Keep in mind that these are the very basic, elemental building blocks of a narrative arc. Not all stories have standard data-mining "shapes," but researchers say that the majority of Western classics fit into one of these six general arcs.
Check out the Computational Story Lab project page and you can fiddle with the charts for yourself on each individual book. Interesting final note: The research page has a special section dedicated solely to the Harry Potter books. Nothing wrong with that.