When We Agree, Our Brains Align

New research from Adam Kleinbaum shows how consensus-building conversations bring us closer together.

Conversation is an integral part of everyday life, anchoring all sorts of interactions, from mundane daily planning with one’s partner to collaborative work meetings to jury deliberations. Research has shown that even brief conversations can play an influential role in society, such as by creating support for anti-discrimination laws, or influencing voting behavior, and these effects can spread throughout social networks.

But does conversation actually change the way we think?

That’s the question Tuck professor Adam Kleinbaum set out to answer in a new working paper titled “How Consensus-Building Conversation Changes Our Minds and Aligns Our Brains.”

In it, Kleinbaum and his social-neuroscientist coauthors Beau Sievers, Dartmouth’s Christopher Welker, Uri Hasson, and Dartmouth’s Thalia Wheatley devise a unique experiment using MBA students and functional magnetic resonance (fMRI) imaging. For the study, they showed participants a series of clips from an unfamiliar movie while measuring their neural activity using fMRI scans. By design, the clips were ambiguous in their meaning. After watching the clips, the participants were divided into groups of five to discuss what they saw and to reach a consensus about the narrative. Once they achieved a consensus, they went back into the fMRI machine to watch the clips again, plus other clips from the same movie.


Adam Kleinbaum teaches Social Networks in Organizations and Personal Leadership in the MBA program and Leaders as Network Architects in Tuck Executive Education’s Advanced Management Program.

The researchers were trying to learn more about the process of reaching a consensus, and whether that process had measurable impacts on the subjects’ brains and future ways of thinking. While the researchers could not see the participants’ thoughts, they could measure the participants’ neural alignment, which tracks their shared understanding, through a proxy called inter-subject correlation (ISC) of blood-oxygen level signal from the fMRI. Thus, the study uses ISC to assess whether consensus-building conversation can align thinking within groups.

It seems our brains don’t enjoy getting bulldozed any more than our conscious selves do.
—Adam Kleinabum, Associate Professor of Business Administration

The results were fascinating, Kleinbaum said. First, we found that conversation served not only to align people’s narratives, but also to align the neural activity in their brains. And this increased alignment applied not only to the clips that they discussed together, but also to the new, previously unseen clips. The participants, Kleinbaum explained, applied the group’s narrative to the new movie clips they watched.

A second and perhaps more important finding was that there was significant variation between groups in how they reached alignment. The groups that displayed more equal participation and turn-taking had higher levels of neural alignment than groups where one person dominated the conversation. It seems our brains don’t enjoy getting bulldozed any more than our conscious selves do, Kleinbaum said.

Their findings also shed light on the role certain people play in social networks, and how those roles impact consensus. The researchers found that participants who occupy central positions in the real-world social network—those who are well-connected and have many well-connected friends—play an important role in alignment. This is because people with high “centrality” tend to engage others in the conversation more, and they are more adaptable, which allows them to build consensus by pulling others together and allowing themselves to move closer to the emerging consensus.

Groups are better able to achieve consensus when conversational participation is evenly balanced.

In contrast to the participants with high centrality were those perceived as having high social status. These folks spoke more, gave more orders, and implicitly rejected others’ proposals. While they were rated by their peers as more influential, they were actually less neurally influential. As the authors write, this raises the possibility that their conversation behaviors produced public compliance without private acceptance.

As we know, consensus is a crucial feature of getting things done as a society, or in an organization. The key takeaway from this research is that consensus doesn’t just happen; it must be cultivated in conversation.

Groups are better able to achieve consensus when conversational participation is evenly balanced, Kleinbaum said, and having group members who are highly central in the informal social network of an organization may be instrumental in achieving that balance.