In a society filled with right-brain and left-brain worlds, Dr. Patrick Juola straddles both hemispheres.
A literary specialist, he works and plays in Java as fluently as he does in English, pioneering a trend to bridge the humanities and technology.
Juola, associate professor of computer science at Duquesne University, has put Java together so that it can dissect another language—human language—and examine word usage and speech patterns. The National Science Foundation (NSF) is supporting Juola’s groundbreaking work with a second grant, $1.6 million.
With this three-year grant, Juola and researchers in the new Evaluating Variations in Language (EVL) lab will simplify his program, which determines authorship across a range of fields—from the forensic study of a suicide note to politically critical policies, from Indiana Jones-type questions of biblical authorship to teachers’ work-a-day-questions of possible student plagiarism.
“This new research in a re-emergent field has the potential to really change the way basic scholarship is done,” Juola observed. “Google made it obvious there is money to be made in teaching computers to understand language, and a lot of literary scholars paid attention.”
This work overrides what Juola calls “a purely artificial gulf” that has separated sciences and humanities since a scholarly divide was fostered in the 1950s.
“There is a big movement recently to bridge that gulf now,” Juola said, pointing to creation of the Office of Digital Humanities within the National Endowment for the Humanities (NEH) in 2008, which awarded him a grant in its inaugural year.
This summer, he also received the international Developer’s Challenge Award at the Digital Humanities Conference and instructed the next generation of world-class scholars in authorship, attribution and language variation at the University of Copenhagen.
Duquesne, Juola said, is uniquely positioned to forge ahead in this field because its computer science studies are housed within the college of liberal arts.
In tackling authorship issues, Juola started with a very broad question: “Can we infer personality from writing?”
His answer is obviously affirmative. Millions of minute idiosyncrasies in speech and writing offer telltale clues about who we are, evident in decisions such as choosing the words “large” or “big” over “enormous.” Juola’s Sherlock Holmes-like software successfully connects these clues to decipher the author’s identity.
Spelling and grammar variations, even what prepositions people use are on the program’s radar. “Our research goal is to enable a computer to look at a piece of text and say, ‘Yes, this play was written by Shakespeare’ or ‘Yes, this ransom note was written by a man in his early 40s,’” Juola explained.
Juola, an adjunct scientist in the new Human Language Technology Center of Excellence at Johns Hopkins University, sees programs like his as just the beginning for the field of digital humanities.
“This gives scholars access to more information by letting the computer do some of the rote work,” Juola said. “How many Shakespearean plays can I read? The computer can do that in seconds.
“What happens with the information? That’s where the human comes in.”