A professor in the University of Montana’s department of psychology may have figured out a way to identify terrorist organizations and when they’re going to carry out a violent act.
Professor Luke Conway uses a psychoanalytic technique known as complexity to evaluate statements published on terrorist organizations’ websites and blogs.
When planning an attack, these terrorist groups “try to keep their communication the same,” Conway explained. “But their complexity drops without meaning to.”
Complexity is a technique with two basic parts that can be used to analyze a person’s subconscious. The system rates how strongly a statement recognizes different sides to one problem and how those sides all tie together. It is useful because it can be applied to literally anything a person says.
“Complexity is interesting because we don’t think about it, but it is sort of a window into someone’s mind,” Conway said.
The study, funded by the U.S. Department of Homeland Security, analyzed statements released by groups in the Middle East that had the potential to be terrorist organizations. Conway also wanted to see if those statements could be used to predict the general timeframe of an attack.
His research team identified several organizations in the Middle East that had similar beliefs. Some were terrorist organizations and others were known to be non-violent. Conway then removed any words that identified an organization or revealed a date, and passed the pages on to student “coders,” like current Ph.D. student Shannon Houck.
“I read a paragraph and was trained to look for linguistic markers of complexity,” Houck said.
Coders then assigned each paragraph a complexity score.
These complexity scores revealed that terrorist organizations could be identified through significantly less complex rhetoric compared to other groups of similar ideology.
The results for predicting a violent act weren’t as cut and dry as identifying a terrorist group, but Conway said complexity can determine a general time frame.
“We can use complexity as a marker of when a group is about to engage in terrorism,” Conway said.
When coders analyzed statements released by a terrorist organization four months before a known attack, they found that terrorists’ complexity would notably decline in the month before an attack.
A terrorist organization would emphasize their ideology more than normal, but would be less willing to recognize alternate sides to an opinion. The trend suggests when this happens, a violent act is in the near future.
“It was such a cool study,” Houck said.
But the researchers may never know the broader impacts their results may have. After completing the work, they sent the results to Homeland Security, where it disappeared into classified use.
Houck thinks the results could be useful, but she thinks the logistics would be difficult because the process of coding statements immediately after terrorist organizations release them is a “time consuming and ongoing process.”
There are computer programs that can automatically run a statement and give it a complexity score, but Houck said they’re not 100 percent reliable and probably won’t replace human coders in the foreseeable future.
Conway is planning to carry out a similar study in Southeast Asia on Christian groups that may have terrorist intentions. Research is slated to begin around January 2014. If his findings are similar to past results, his method could be used to identify terrorist organizations with any ideological background.