October 25, 2024, Michigan State University’s fourth annual Charles K. Atkin Distinguished Speaker Series welcomed ComArtSci alumni Chris Carpenter (Ph.D. ’10), Brian Manata (Ph.D. ’15) and Professor Gary Bente — speaking on their research related to superdiffusers, cohesion in groups and what avatars can reveal about perception and nonverbal communication.
Chris Carpenter: 19 Years of Superdiffuser Research
Chris Carpenter, a professor in the School of Communication and Media at Western Illinois University, returned to his alma mater to discuss nearly two decades of work on the concept of “superdiffusers.” This research, which originated from Carpenter’s graduate studies with Professor Emeritus Frank Boster, investigates the power of specific individuals to spread ideas and influence pro-social behavior.
Through their unique blend of specialized expertise or “maven-ness”, persuasive ability, and connectedness, Carpenter noted that these superdiffusers can be critical in health campaigns and social initiatives — reaching peers in ways that traditional media campaigns cannot.
“Persuasion is hard. And there are some things where we’ve done all we really can with mass media,” Carpenter said. “You can only put up so many billboards to tell people to stop smoking or tell people smoking will kill you before they are like, ‘Yeah, we know.’ But, if you can get somebody they know to talk to them about all the ways they can quit smoking, and personally adapt their message to that person, that's probably going to be more effective, especially for those harder-to-change behaviors.”
Carpenter emphasized the model’s adaptability, reflecting on the potential for superdiffusers to inspire a variety of behavioral changes from reducing sugar consumption to encouraging voter turnout. But there’s a catch: “If it really is that effective, it’s the kind of thing you don’t want in the wrong hands,” he said. “Hypothetically, there are bad superdiffusers out there pushing something terrible.”
Indeed, Carpenter plans to explore broader applications of his model moving forward, investigating how superdiffusers can operate in new domains beyond health communication. He also wants to expand the superdiffuser model itself, exploring the conditions in which all three traits together may not be necessary to accelerate diffusion.
He hopes that others will join him in this pursuit, asking ourselves this question: What can superdiffusers do for me?
Brian Manata: On the Problem of Cohesion in Groups
Brian Manata, an assistant professor at Penn State University and another ComArtSci alumnus, presented his research on group cohesion and conflict — a topic relevant to any team or organization. Manata’s work challenges the assumption that cohesion and conflict are inherently opposed. Instead, he explores how the dynamics between these constructs might impact performance for the better.
“People think of conflict and cohesion as different. That typically, if there’s conflict, you lose cohesion in the group,” Manata said. “Which is kind of interesting and a paradox, because you usually need conflict to do well. You need people that have different perspectives to do well.”
In academia, for instance, forming interdisciplinary teams is encouraged both for the unique perspectives and expertise individuals can bring. In theory, these groups can work together to come up with the most innovative ideas.
“It’s hard to be cohesive with people that speak a different language, in a sense, different jargon. The issue here is one of, how do we have conflict without it attenuating cohesion?”
Manata noted that while task conflict and relationship conflict are different — that is, a disagreement about how something might be accomplished versus personality or interpersonal differences — often the two are strongly correlated. This points to a need for leaders, teams and social groups to find ways to use those conflicts productively.
Moving forward, Manata plans to continue investigating how teams from academia to business can foster constructive disagreements without compromising group cohesion, and ultimately building on this research to develop a general theory of group dynamics.
Like Carpenter, Manata also credits his MSU experience as foundational to his research, highlighting studies he conducted on both student and professional teams while he was a grad student. “Everything starts here; this is the Mecca,” he said. “This is where my initial thinking started.”
Gary Bente: The Virtual Mirror — What Avatars Can Reveal About Social Perception and Nonverbal Communication
Gary Bente, a professor in the Department of Communication and director of the CARISMA Lab at ComArtSci, closed the series with a look into the not-so-futuristic world of avatars and artificial intelligence. His team is exploring how embodied conversational agents could replicate human social cues and support relationship-building in virtual environments.
“I was amazed how AI moved forward during the last two years,” he said. “The verbal, the large language models, I think that’s a leap forward. Unbelievable; a qualitative leap. But it’s not happening with the nonverbal.”
Bente has been using avatars to advance the study of nonverbal communication for decades. Now, he issues a challenge: can communication scholars help develop and improve embodied conversational agents?
He notes that while the “conversational” part has been taken care of with large language models, it’s the embodiment part that still needs a lot of work.
“This is our time. I think if we let this go as a communication science, we are stupid. We should dive into that,” Bente said. “If you do deep learning for language models you can probably do it for nonverbals.”
For Bente, it’s not just about creating a digital character that looks human; it’s about designing one that can convey empathy, build trust and engage users on an interpersonal level. Ultimately, Bente envisions these avatars could provide support in areas where resources are scarce, like mental health care or education.
“How can we really improve embodied agents by our communication research? What kinds of means and methods do we have? What kinds of theories can be applied to make this better, to make this more believable?” he asked.
Fans of films like “I, Robot” or “Her” may be divided on whether this is an endeavor worth pursuing, but rest assured — Bente is equally interested in building a research program that answers questions, for better or worse, such as, what are the risks?
The Charles K. Atkin Distinguished Speaker Series is an annual event hosted by the Department of Communication that brings together leading scholars in the field of communication. Established in honor of the late Professor Chuck Atkin, who was a proud Spartan and chair of the department at the time of his passing, the speaker series highlights the work of ComArtSci alumni continuing this legacy of excellence in research.
By: Jessica Mussell