By: Tina Cormier

Uncovering the Sociology of Language

Pete Aceves didn’t begin his PhD with the intent of theorizing about the collective cognition of human groups. A combination of curiosity and circumstances led him to a certain line of thinking, and hard work brought to life what began as a hunch.

Aceves, the 2017 winner of the prestigious INFORMS Organization Science Best Dissertation Proposal Competition, is in the process of finishing his PhD with James Evans in the Department of Sociology. His dissertation investigates the role that language plays in affecting patterns of social interaction, and how these patterns can shape problem-solving, creativity, and team performance. Others have explored language structure relative to individual cognition, but Aceves is the first to apply the concept to groups.

“I came into my PhD program intending to focus on international development – thinking about the economy from a sociological point of view,” says Aceves. But with twin one year old children at home and with his wife also starting a PhD program at Michigan, the travel required to study international development was unfeasible.

While working through various options for a new dissertation topic, he came across a fascinating, and ultimately transformative, video by David Deutsch, a Physicist at the University of Oxford. He became transfixed by the concept of information and how the brain can turn a series of complex electrochemical reactions into sound waves that travel through the air, only to again become electrochemical signals in another person’s brain.

“What struck me is that the information in the message remains constant despite the different physical instantiations of this process,” says Aceves. “I ended up in a deep rabbit hole and came out on the other side thinking about how groups of individuals are able to compute information collectively.” Specifically, he questioned how this collective process of computation could be shaped by the structure of the symbolic code used to communicate information between group members, namely language.

The video was the spark that got Aceves thinking about tracing both processes of collective computation through conversations and the effect these processes had on group outcomes. There is a body of work published on how the structure of language, the primary mechanism of information transfer, affects patterns of individual thought, but surprisingly little is known about how language may shape group processes. In his dissertation, Aceves examines whether the structure of the language spoken within a group can affect its approach to problem-solving, its creativity, and ultimately its success or failure.

To answer this question, he tracked the outcomes of over 1800 monolingual trekking expeditions to the Himalayas between 1907 and 2015. His expectation was that the structure of the more than 30 languages represented in the study would affect the success with which each group was able to problem-solve, strategize, plan, and ultimately reach the summit.

Aceves measured a novel language structure attribute that he calls “information density,’ or the average amount of conceptual information contained within the words of a given language. “Consider the following example,” says Aceves. “To communicate, ‘I play soccer and violin’ in Spanish, I would have to say, ‘I play soccer and I touch the violin.’ In this example, the verb ‘play’ has two meanings in English and only one in Spanish,” he explains. Given only this example, English would have a higher information density than Spanish has because on average, each word has more possible meanings.

Aceves estimated the information density of the 33 languages in the Himalayas study (and over 1,000 other languages) and found that higher information density languages are strongly associated with a higher proportion of team members reaching the summit and doing so more quickly than those in groups speaking languages with lower information density. “I was actually surprised by the magnitude of the effect,” he says. “When the information density is standardized on a 1-100 scale, I found a 1-point increase in information density to be associated with a 1.6% increase in the number of team members who reach the summit.” To further investigate the mechanisms behind these results, he is currently running an experimental study in India where he is randomizing bilingual speakers to a single language and tracing how they solve different kinds of tasks.

Encouraged by the strong relationship between language structure and group performance, he hopes to build a new research program to study the implications of his findings. Open questions stemming from his initial work include investigating how language structure affects communication patterns within groups and organizations engaged in different kinds of tasks and within different market environments. In the business world, for example, the global language is English – Could group dynamics be partially based on the composition of home languages within a team?

As he wraps up his dissertation, he reflects on his time at Chicago. “The most influential part of my graduate experience has been my advisor, James Evans. He just always assumed I could do way more than I thought I could, and he was so supportive of my dissertation idea, even though it was unlikely to pay off,” says Aceves. “Constant support for your ideas is sort of priceless.”

With one chapter coming to an end, the next is already beginning; Aceves and his wife are preparing for faculty positions at Bocconi University in Milan, Italy, starting this September, where he hopes to build a research program to answer the many emerging questions within this new field that he calls the “Sociology of Language.”