A New Paradigm for Understanding Social Cognition

We consider humans to be at the apex of social cognition. But we’re not the only animals that closely interact with each other. Marmosets, for example, are highly social creatures. In the wild, they choose to work together, sharing food and helping each other raise young.

In a new study, published Aug. 28 in Current Biology, Yale researchers, collaborating across Yale School of Medicine (YSM) and the Faculty of Arts and Sciences (FAS), created a novel approach for closely observing marmoset behavior. Combined with computational modeling, the study revealed sophisticated cooperative strategies utilized by marmoset pairs.

This new system could help improve the way psychologists study social cognition and lead to a better understanding of disorders in which it is impaired, such as autism spectrum disorder.

“This particular study established a paradigm that allows us to obtain high throughput behavioral data,” says Monika Jadi, PhD, associate professor of psychiatry at YSM and co-principal investigator. “So, we can now start probing the brain areas that are involved in social cognition and understanding how they are involved.”

The current established paradigm for studying marmoset social cognition is cumbersome and limits the amount of data researchers can obtain. “It’s not ideal for studying complex behavior dynamics,” says Steve Chang, PhD, associate professor of psychology in FAS and co-principal investigator.

Furthermore, the paradigm involves manually observing and scoring behaviors. “It’s a slow and error-prone process, which does not provide mechanistic insights into what the animals are doing,” says Anirvan Nandy, PhD, associate professor of neuroscience and co-principal investigator.

In the new study, the team used a new apparatus that they developed—Marmoset Apparatus for Automated Pulling—which requires a pair of marmosets to work together, pulling levers within a certain timeframe in order to receive a reward. The researchers also set up cameras around the apparatus, enabling them to track moment-by-moment where the animals were looking.

This rich set of behavioral data allowed the team to build a computational model for understanding the various cooperative strategies that the marmosets utilized.

The researchers discovered that marmosets used different types of strategies to cooperate. In some cases, one marmoset would monitor what the other was doing to help time its pull, which researchers refer to as a “social gaze-dependent strategy.” In other instances, the animals got into a rhythm and synchronized their pulls without looking at each other, using a “social gaze-independent strategy.”

The marmosets showed flexibility when choosing strategies. When the duo was performing well, they tended to use social gaze-independent rhythmic pulling. But at times when they struggled to synchronize, they switched to social gaze-dependent strategies. In other words, the animals would look at each other more as they strived to find a rhythm.

The primates also chose different strategies based on their partner. For instance, if one marmoset struggled to pull the lever within the allotted time frame, the researchers found that its partner would adapt, such as pulling the lever twice to give the first marmoset more time. But when paired with a new partner who did not struggle with timing, the marmoset would revert to pulling only once.

“Pulling strategies of the same monkey can alternate back and forth,” says Chang. “They are using partner identity to adjust how they are interacting with the device.”

Furthermore, the dominant member of the pair tended to pull the lever after the subordinate. “The dominant was monitoring what the subordinate was doing and timing their pulls,” says Nandy.

The findings highlighted how marmosets proactively looked for ways to cooperate with one another. “They’re motivated to figure out the best way to work together for mutual benefit,” says Chang. “And they can adjust their strategies to achieve that benefit.”

The researchers are now combining their new system with neural recording to better understand how neurons encode cooperative strategies. Down the road, this type of research could lead to better understanding of disorders such as autism spectrum disorder, in which individuals often struggle with behavioral rigidity.

“This work opens up a whole realm of possibilities around studying complex social cognition at large,” says Nandy.

Olivia Meisner, PhD, a former graduate student from YSM's Interdepartmental Neuroscience Program, and Weikang Shi, PhD, a Wu Tsai Institute postdoctoral fellow, were co-first authors of the study.

This work was supported by the National Science Foundation (award DGE2139841), the National Institute of Mental Health (award R21MH126072), the National Eye Institute (award P30EY026878), and Yale University. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or National Science Foundation. Additional support was provided by the Simons Foundation Autism Research Initiative and the Wu Tsai Institute.