Armed with a slideshow and an extensive knowledge of William Shakespeare’s plays, Carnegie Mellon’s Michael Witmore and the University of Strathclyde’s Jonathan Hope gave a presentation this past Tuesday about their collaborative research on Shakespeare’s histories and comedies.

Using a program called Docuscope, developed at Carnegie Mellon, Witmore and Hope were able to statistically distinguish the two Shakespearean genres from each other. Equipped with a specialized dictionary, Docuscope is able to divide texts into strings of words that are then sorted into one of eighteen word categories, such as “Inner Thinking” and “Past Events.” The program turns differentiating amongst genres into a statistical task by testing the frequency of occurence of words in each of the categories for each individual genre and recognizing where significant differences occur.

A point that both Witmore and Hope emphasized was that although Docuscope can calculate numbers with mechanical accuracy, the actual interpretation of the differences in genres takes place after Docuscope spits out the raw data. Without a background in Shakespeare, they said, they couldn’t have accurately assessed Docuscope’s results. “We didn’t really have a hypothesis,” said Hope. Originally, the pair went through a variety of different possible topics before deciding to “put Shakespeare through [Docuscope] to see what would happen.”

What they found was that Shakespeare’s comedies and histories were written with distinctly different diction. For example, the comedies have the most “Interacting” words, which Hope said is plausible because of the witty dialogue that often takes place in comedic plays. A more surprising finding was that comedies had a higher frequency of “First Person” words. This may indicate an interesting twist on the traditional mindset that the “creation of the modern self” occurred in the tragedies, with their many soliloquies and monologues. Tragedies themselves were not found to be statistically distinguishable from the other two genres.

The importance of the human interpretation of Docuscope’s results became clear to Witmore and Hope when the program said that histories have a significantly higher frequency of “Word Picture” words, which primarily includes nouns relating to objects and places such as “chair” and “room.” It wasn’t until Witmore and Hope looked at the results more carefully that they noticed that Docuscope was incorrectly coding the word “king,” often the title of the lead character in Shakespeare’s histories, as a “Word Picture” word. Once they edited out the characters’ names from the directing cues, “Word Picture” words no longer distinguished histories from the other genres.

Hope noted that they may try to create an early Modern English dictionary for Docuscope in addition to fixing problems such as the “king” one, which could potentially produce more accurate results. Some day, Witmore says, he and Hope “want to find what is distinctly different of Shakespeare from other writers.” Although such a project would take many years if done by reading alone, with the help of Docuscope, they hope to be able to more quickly identify a “recipe” for Shakespearean quality. With research such as Witmore’s and Hope’s, Docuscope is poised to play a more prominent role in the literary community in the coming years.